Springer Series on Atomic, Optical and Plasma Physics 95

Shi Nguyen-Kuok

Theory of Low-
Temperature
Plasma Physics
Springer Series on Atomic, Optical,
and Plasma Physics

Volume 95

Editor-in-chief
Gordon W.F. Drake, Windsor, Canada

Series editors
James Babb, Cambridge, USA
Andre D. Bandrauk, Sherbrooke, Canada
Klaus Bartschat, Des Moines, USA
Philip George Burke, Belfast, UK
Robert N. Compton, Knoxville, USA
Tom Gallagher, Charlottesville, USA
Charles J. Joachain, Bruxelles, Belgium
Peter Lambropoulos, Iraklion, Greece
Gerd Leuchs, Erlangen, Germany
Pierre Meystre, Tucson, USA
The Springer Series on Atomic, Optical, and Plasma Physics covers in a
comprehensive manner theory and experiment in the entire field of atoms and
molecules and their interaction with electromagnetic radiation. Books in the series
provide a rich source of new ideas and techniques with wide applications in fields
such as chemistry, materials science, astrophysics, surface science, plasma
technology, advanced optics, aeronomy, and engineering. Laser physics is a
particular connecting theme that has provided much of the continuing impetus for
new developments in the field, such as quantum computation and Bose-Einstein
condensation. The purpose of the series is to cover the gap between standard
undergraduate textbooks and the research literature with emphasis on the
fundamental ideas, methods, techniques, and results in the field.

More information about this series at http://www.springer.com/series/411

Shi Nguyen-Kuok

Theory of Low-Temperature
Plasma Physics

123
Shi Nguyen-Kuok
Laboratory of Plasma Physics
National Research University “MPEI”
Moscow
Russia

ISSN 1615-5653 ISSN 2197-6791 (electronic)

Springer Series on Atomic, Optical, and Plasma Physics
ISBN 978-3-319-43719-4 ISBN 978-3-319-43721-7 (eBook)
DOI 10.1007/978-3-319-43721-7
Library of Congress Control Number: 2016949105

© Springer International Publishing Switzerland 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Presently, low-temperature plasma is widely used in science and engineering. Many

technological problems such as cutting and welding, melting of metals, welding and
spraying of refractory, wear-resistant, corrosion-resistant coatings, obtaining of
spherical and dispersion of powders, thermal surface treatment of refractory
materials, plasma polishing of optical glass and metals, recycling of heat-resistant
plasma chemical waste and others may be successfully solved using
low-temperature plasma. Different types of devices for obtaining of plasma flows
are developed—plasma torches, where the high-frequency induction (RF), the
high-frequency capacitance, ultrahigh-frequency (UHF) and optical discharges are
used as the plasma sources.
Effective use of these types of plasma torches requires extensive studies of
plasma processes in order to: identify their laws; receive the plasma stream with the
desired properties; determine the optimal operating conditions and the development
of new structures of plasma torches; direct control of the plasma; and others. These
problems can be solved on the basis of experimental studies and mathematical
modeling of processes in the plasma.
Field experiments (with the ability to implement them) give detailed information
about the plasma. Many phenomena in plasma have been predicted by theoretical
calculations, and discovered and confirmed experimentally, in particular, the results
of field experiments can be used for comparison with the experimental data in order
to analyze the physical processes and the direct control of the plasma.
Unfortunately, many experiments are expensive and experimental methods are
still far from perfect. The errors of the experimental methods are often large
enough, and often the experimental data of different authors have large discrep-
ancies and contradictory. Some experiments are difficult to carry out in conditions
of low-temperature plasma, and their data provide only its integral characteristics.
An important problem of determining the spatial distributions of quantities such
as temperature and composition of the plasma, gas flow rate, the electromagnetic
field inside the plasma flow and others, their relationships with the external
parameters of the plasma torch is a difficult experimental problem. The solution of
this problem becomes more difficult when you study the properties of multivariate,

v
vi Preface

nonequilibrium, transient, nonsymmetric plasma flows that is determined solely by

a wide variety of physical processes in the plasma and by the complex nature of its
components interaction.
Such difficulties in obtaining experimental data for a wide range of plasma flows
parameters make actual the development of theoretical modeling techniques.
Theoretical studies of the plasma by mathematical simulation methods allow to
eliminate many of the natural experiment limitations, complement and extend the
possibilities in the study of plasma, as well as replace complex, expensive, and
sometimes impracticable principal experiment. Based on the developed mathe-
matical models you can study the physical processes occurring in the plasma, set
optimal modes of plasma sources, to determine changes of the external adjustable
parameters, measuring point and the minimum number of measured physical
quantities needed for solving the problem of plasma control. In particular, progress
in the study of plasma and use cannot be without an in-depth theoretical studies.
One of the main tasks of theoretical studies of plasma processes is to obtain local
physical parameters of the plasma and the establishment of their links with external
power and the geometric parameters of the plasma torch. Typically, in engineering
design it is required to make some form of temperature distribution and plasma
flows, corresponding to the special conditions of practical use. This is typically
achieved by altering the current, plasma gas flow rate, the operating mode of the
generator, and the geometric characteristics of the plasma torch, i.e., only control
the external parameters of the plasma torch.
Significant progress in the use of the plasma can be achieved by providing direct
control of plasma flows. This requires a structural parametric identification of the
plasma. In this case, you first need to examine and define the properties and
parameters of the plasma, to find the place of measurement, to build functional
relationships, etc. This problem can be solved with the help of mathematical
modeling as a real object for study using different mathematical models describing
physical processes in the plasma. However, the theoretical description of the plasma
is often carried out with the help of highly nonlinear differential equations in partial
derivatives. The solution of these equations belongs to the class of rigid and
ill-conditioned problems, and is a serious mathematical and information problem.
The obtained results allow us to establish the flow pattern and gas heating, of
energy in the plasma, to determine the area of cold channel, thermal and ionic
imbalance in the plasma region preference models, create the necessary conditions
for solving the problem of structural parametric identification and on the basis
of the decision to make the transition to the system automatic control of plasma
parameters. You can also use the results as the necessary material for the study and
design of plasma torches, optimization of various plasma processes and the creation
of simulation models of plasma, which have been free from stability problems of
computational procedures.
To learn the basics of theory of low-temperature plasma readers are invited to
first familiarize themselves with the contents of the first five chapters. In Chap. 1 the
general approach to the description of plasma processes, the basic model of the
plasma, their assumptions and applicability are outlined. Models and cross sections
Preface vii

for the interaction of particles in the plasma are considered using the classical model
of elastic collisions in Chap. 2, and with the help of quantum-mechanical theory of
scattering particles in Chap. 3.
Determinations of the composition, thermodynamic functions, transport coeffi-
cients of the plasma on the basis of model and the mean free path are given in
Chap. 4. In Chap. 5 the solution of the Boltzmann kinetic equation by the
Chapman–Enskog approach and calculation of the transport coefficients of the
plasma are provided.
In Chap. 6 the main method of control volume for solving the problems of heat
transfer and plasma dynamics is presented along with the various numerical
schemes (methods) of solutions, their analysis in terms of accuracy, and the ease of
implementation. Stability analysis and determining the cause and the remedy of the
instability are also discussed in this chapter. We introduce the discrete analogues of
various equations of plasma mathematical models. The structural organization
of the computational procedures for the simulation of plasma processes is also
given.
In Chap. 7 the methodology and the main results of the simulation of the RF
plasma torches is set out. As an addition to those obtained in Chap. 6 algorithms,
there are the method of calculating the two-dimensional electromagnetic field in the
RF plasma torches and simulation of the RF plasma torches based equilibrium and
nonequilibrium models.
Similarly, in Chap. 8 the methodology and the main results of the simulation of
Arc plasma torches, different design and technological applications (Arc in the
channel for cutting and spraying, free Arc welding and melting of metals, etc.) are
presented.
In Chap. 9 the modern models and algorithms for the calculation of the
near-electrode processes taking into account the working conditions of the majority
of electric Arcs are described. The calculations for thermionic cathodes based on
pure and thoriated tungsten in a wide range of parameters of the cathode and
plasma, as well as algorithms for calculating the heating of the cathode and anode
processes are provided.
In Chap. 10 the issues of heat transfer and dynamics of solid particles in plasma,
model of heat and motion of particles in the plasma and the main results of the
calculations with respect to the jet of the RF plasma torches are described. The
results of the interaction model and the dispersed flow with the plasma are given.
Description of the experimental stand, the main methods of diagnosis, and the
results of measurements of plasma parameters in the induction and Arc plasma
torches are described in Chap. 11.
I would like to specially thank Profs. E. Son and S. Bobashev for carefully
reading the manuscript and for helpful comments during the preparation of this
book, as well as the staff of the Department of Electrical Engineering and Energy,
Saint-Petersburg State Polytechnic University and the General Physics and Nuclear
Fusion, National Research University «Moscow Power Engineering Institute» for
the help in the preparation of material of the book.
viii Preface

This book was written after my many years of working in the field of
low-temperature plasma physics. However, this would not have been possible if it
were not for many years of creative collaborations with Prof. S. Dresvin. He was
who introduced me to the world of low-temperature plasma physics, when I was
still a student of the Electromechanical Faculty of St. Petersburg State Polytechnic
University.
I would like to emphasize the special contribution to my creative extension
of the work of the Prof. O. Sokolov. During the many sleepless «White nights» in
St. Petersburg, we had a conversation with him about a systematic approach to
solve a multidimensional system of nonlinear differential equations. It is a great idea
that the internal feedback about physical processes found their realization in the
simulation model for a sustainable solution of a system of differential equations for
plasma.
I have many excellent professors, colleagues, friends, my talent graduate stu-
dents in colleges and universities of St. Petersburg and Moscow, and I would like to
emphasize their help and support in the process of working on the material of this
book.
I am infinitely grateful to my parent, my wife, and my sons for their help and
support. Their daily smiles are the driving motor of the writing of this book.
Hence, this book was written on behalf of all these wonderful people.

Moscow, Russia Shi Nguyen-Kuok

Reviewers

Dr. Sc., Professor, Corresponding Member of the Russian Academy of Sciences

E. Son (Joint Institute of High Temperature of Russian Academy of Sciences);
Dr. Sc., Professor, S. Bobashev (Ioffe Physic-Technical Institute)

Shi Nguyen-Kuok
Theory of Low-Temperature Plasma Physics, 2017 - p.
ISBN

The basic approaches to the theoretical description of the plasma, mathematical

models, modeling techniques, numerical methods, and modes of ensure the stability
of solutions of plasma equations systems are represented in this book. The main
methods of calculating the cross sections of plasma particles interaction and solu-
tion of the Boltzmann kinetic equation for determine the transport coefficients of the
plasma are also represented in the book. The results of calculations of the com-
position, thermodynamic properties, transport coefficients, the particle interaction
cross sections of the equilibrium and two-temperature nonequilibrium plasma. The
results of simulation and calculation of plasma parameters in RF and Arc plasma
torches are presented. The complex physical processes in the induction plasma and
Arc plasma, internal and external parameters of the plasma torches, near-electrode
processes, heat transfer and the flow of solid particles in the plasma are considered
in the book etc.
The book is intended for professionals involved in the theoretical study of
Low-temperature plasma and design of plasma torches, and will be useful for
students and postgraduates of corresponding profiles.

ix
Contents

1 The Theoretical Basis of the Low-Temperature Plasma . . . . . . . . . . 1

1.1 The Basic Approaches to the Description of Plasma. . . . . . . . . . 1
1.1.1 The Plasma Relations for the Full Thermodynamic
Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.2 The Model of Local Thermodynamic Equilibrium
in Plasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.1.3 The Model of Partial Local Thermodynamic
Equilibrium in Plasma . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Model of Equilibrium Plasma . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2.1 The Energy Balance Equation . . . . . . . . . . . . . . . . . . . . 12
1.2.2 The Momentum Equations of the Plasma Flow . . . . . . . 13
1.2.3 The Continuity Equation of the Plasma Flow . . . . . . . . 14
1.2.4 The Maxwell’s Equations . . . . . . . . . . . . . . . . . . . . . . . 14
1.2.5 The Equations of Equilibrium Plasma Model
in a Cylindrical Coordinate System . . . . . . . . . . . . . . . . 15
1.3 The Two-Temperature Model of Nonequilibrium Plasma . . . . . . 18
1.3.1 The Violation of Ionization Equilibrium in Plasma.
Equation of Ionization Equilibrium . . . . . . . . . . . . . . . . 23
1.4 The Two-Speed Plasma Model . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.5 Gas Dynamic Models of Plasma Turbulence . . . . . . . . . . . . . . . 26
1.5.1 The Reynolds-Averaged Navier–Stokes Equations . . . . 27
1.5.2 The Main Gas Dynamic Model of Turbulence . . . . . . . 31
1.6 Models of Light Emission and Absorption in Plasma . . . . . . . . . 36
1.6.1 Radiation in the Continuous Spectrum . . . . . . . . . . . . . 37
1.6.2 Absorption in the Continuum . . . . . . . . . . . . . . . . . . . . 42
1.6.3 Radiation in the Spectral Lines . . . . . . . . . . . . . . . . . . . 45
1.6.4 The Absorption in Spectral Lines . . . . . . . . . . . . . . . . . 49
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

xi
xii Contents

2 Classical Theory of the Particle Scattering . . . . . . . . . . . . . . . . .... 53

2.1 Classical Consideration of the Particle Scattering . . . . . . . . .... 53
2.2 Determination of Cross Sections for Plasma Particles
Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 59
2.2.1 Coulomb Scattering: The Cross Section
of the Interaction of Charged Particles . . . . . . . . . .... 59
2.2.2 Ramseur Effect and the Resonance Scattering
of the Electrons by Atoms . . . . . . . . . . . . . . . . . . . .... 64
2.2.3 The Cross Sections for the Interaction
of Molecule–Molecule, Atom–Atom . . . . . . . . . . . .... 65
2.2.4 The Cross Sections for the Interaction
of the Atom–Ion . . . . . . . . . . . . . . . . . . . . . . . . . . .... 66
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 70
3 Quantum Mechanical Theory of the Particle Scattering. . . . . . .... 71
3.1 The Schrödinger Equation . . . . . . . . . . . . . . . . . . . . . . . . . .... 71
3.2 Solution of the Schrödinger Equation for the Elastic
Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 74
3.3 Determination of the Phase Shift . . . . . . . . . . . . . . . . . . . . .... 78
3.4 Born Approximation for Calculating the Amplitudes
of the Scattered Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 81
3.5 Determination of Differential and Total Cross Sections
of Elastic Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 85
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 92
4 Determination of the Composition, Thermodynamic
Properties, and Transport Coefficients on the Basis
of the Mean Free Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.1 The Plasma Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.2 Thermodynamic Properties of the Plasma . . . . . . . . . . . . . . . . . . 98
4.3 Transport Coefficients of the Plasma. . . . . . . . . . . . . . . . . . . . . . 102
4.3.1 Determination of the Electrical Conductivity
and Thermal Conductivity . . . . . . . . . . . . . . . . . . . .... 103
4.3.2 Determination of the Viscosity, the Ambipolar
Diffusion Coefficient, and Integrated Emission . . . .... 108
4.4 The Coefficients of Triple Recombination and Impact
Ionization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 114
4.5 The Frequency of Particle Collisions, the Current Density,
and Heating Capacity of Plasma . . . . . . . . . . . . . . . . . . . . . .... 117
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 120
5 The Boltzmann Kinetic Equation and Calculation
of the Transport Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
5.1 Introduction of the Boltzmann Kinetic Equation . . . . . . . . . . . . . 123
5.2 The Transport Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.2.1 Principle of the Equilibrium Systems . . . . . . . . . . . . . . 130
Contents xiii

5.3 Solution of the Boltzmann Kinetic Equation

by Chapman–Enskog Approach . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.4 Determination of Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.4.1 Determination of the Elements in the Bracket
Integral {B, B} by Sonine Polynomials . . . . . . . . . . . . . 141
5.4.2 The Bracket Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . 143
5.5 Determination the Coefficients of Diffusion, Thermal
Diffusion, and Thermal Conductivity . . . . . . . . . . . . . . . . . . . . . 146
5.5.1 Calculation of Diffusion and Thermal Diffusion
Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
5.5.2 Determination of the Heat Flow . . . . . . . . . . . . . . . . . . 147

5.5.3 Defining Elements of the Bracket Integrals A; ~ A ~
and fA; Dg Using the Sonine Polynomials . . . . . . . . . . 148
5.5.4 The Bracket Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.6 Determination of the Electrical Conductivity . . . . . . . . . . . . . . . 156
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
6 Numerical Methods of the Plasma Physics . . . . . . . . . . . . . . . . . . . . 159
6.1 Basis of the Numerical Methods of the Plasma Physics . . . . . . . 159
6.1.1 Solution of the Generalized Differential Equation . . . . . 159
6.1.2 Solution of the Momentum Equations and the
Continuity Equation of the Gas Flux . . . . . . . . . . . . . . . 171
6.1.3 Conversion of Different Equations in Generalized
Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
6.2 The Stability of the Difference Methods and Computational
Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
6.2.1 Stability of Difference Methods . . . . . . . . . . . . . . . . . . . 184
6.2.2 Analysis of the Computational Procedures Stability . . . 187
6.3 The Structural Organization of Computational Procedures . . . . . 191
6.3.1 The External Feedback in the Structure of
Computational Procedures . . . . . . . . . . . . . . . . . . . . . . . 191
6.3.2 Installing Internal Feedback in the Computational
Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
6.3.3 The Mathematical Algorithms with Internal
Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
6.3.4 The Stability Region of the Computational Procedures
of Plasma Processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
7 The RF Plasma Torches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
7.1 Characteristics of the RF Plasma Torches . . . . . . . . . . . . . . . . . . 205
7.2 Electromagnetic Field in the RF Plasma Torches . . . . . . . . . . . . 214
7.2.1 Equation of the Vector Potential and Its Solution . . . . . 219
7.2.2 Boundary Conditions of the Equation
of the Vector Potential. . . . . . . . . . . . . . . . . . . . . . .... 222
xiv Contents

7.2.3 Calculation of the Electromagnetic Field and Electric

Characteristics of the RF Plasma Torches . . . . . . . . . . . 227
7.3 Simulation of the Equilibrium Plasma in the RF Plasma
Torches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
7.3.1 Algorithms of the Mathematical Model . . . . . . . . . . . . . 241
7.3.2 Analysis of the Simulation Results . . . . . . . . . . . . . . . . 249
7.4 Disturbance of the Thermal and Ionization Equilibrium
in Plasma, Caused by the Gases Movement . . . . . . . . . . . . . . . . 258
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma
Torches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
7.5.1 Discrete Analogues of the Mathematical Model . . . . . . 267
7.5.2 Comparative Analysis of the Simulation Results . . . . . . 272
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
8 The Arc Plasma Torches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
8.1 The Features of the Arc Plasma Torches . . . . . . . . . . . . . . . . . . 285
8.2 The Arc Plasma Torches for Cutting and Spraying. . . . . . . . . . . 292
8.3 Simulation of the Arc Plasma Torches in the Channel . . . . . . . . 299
8.3.1 Models of the Arc Plasma Torches in the Channel . . . . 299
8.3.2 Analysis of the Simulation Results . . . . . . . . . . . . . . . . 307
8.4 Two-Dimensional Electromagnetic Problem for the Arc
Plasma Torches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
8.4.1 Solution Using the Electric Potential . . . . . . . . . . . . . . . 318
8.4.2 Solution Using the Function of the Electric Current . . . 322
8.4.3 Calculation of the Electromagnetic Field of the Arc
Plasma Torches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
8.5 The Free Burning Arc Plasma Torches . . . . . . . . . . . . . . . . . . . . 334
8.5.1 The Arc Plasma Torches for Welding and Melting
of Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
8.5.2 Algorithms of the Mathematical Models . . . . . . . . . . . . 337
8.5.3 Comparative Analysis of the Simulation Results . . . . . . 346
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
9 Electrode Processes in the Arc Plasma Torches . . . . . . . . . . . . . . . . 367
9.1 Model of the Cathode Processes . . . . . . . . . . . . . . . . . . . . . . . . . 368
9.2 Calculation of the Cathode Processes . . . . . . . . . . . . . . . . . . . . . 379
9.3 Thermal Problem of Heating the Cathode . . . . . . . . . . . . . . . . . . 391
9.4 Model of the Anode Processes . . . . . . . . . . . . . . . . . . . . . . . . . . 395
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
10 Heat Exchange and Movement of the Solid Particles
in the Plasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
10.1 The Basic Approaches to the Description of the Heat
Exchange and Movement of the Particles in the Plasma . . . . . . . 401
10.1.1 The Criteria of Heat Transfer and Gas Dynamics
of Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
Contents xv

10.1.2 The Non-gradient Heating and Movement

of the Particles in the Plasma . . . . . . . . . . . . . . . . .... 403
10.1.3 The Gradient Heating Particles in the Plasma . . . . .... 407
10.1.4 Evaporation of Particles in the Plasma . . . . . . . . . .... 410
10.2 Calculation of the Heating and Movement of the Particles
with the Given Parameters of the Plasma Flow . . . . . . . . . .... 415
10.2.1 Installation for Plasma Treatment
of Powder Materials . . . . . . . . . . . . . . . . . . . . . . . .... 415
10.2.2 Solution for Equations of the Heating
and Movement Particles . . . . . . . . . . . . . . . . . . . . .... 418
10.3 The Model of Plasma, Loaded the Flow of Solid Particles . .... 425
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 430
11 The Features of the Experimental Methods and Automated
Diagnostic Systems of RF and Arc Plasma Torches . . . . . . . . . . . . . 431
11.1 The Experimental Setup and Automated Diagnostic System
for the Study of the RF and Arc Plasma Torches . . . . . . . . . . . . 431
11.1.1 Adjusting Elements in Optical System. . . . . . . . . . . . . . 435
11.2 The Basic Methods of the Plasma Diagnostics . . . . . . . . . . . . . . 438
11.3 Measurements of Plasma Parameters in the RF
and Arc Plasma Torches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Appendix B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Appendix C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
Appendix D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
Appendix E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
Chapter 1
The Theoretical Basis
of the Low-Temperature Plasma

1.1 The Basic Approaches to the Description of Plasma

There are three approaches to the theoretical description of plasma and plasma
processes: the microscopic description using the drift approximation; the statistical
description based on the solution of the Boltzmann equation, together with
Maxwell’s equations for the description of the field, and the macroscopic
description of the equations of magnet- and gas dynamics (MGD).
The microscopic description reduces to determining the tracks of particles in the
plasma fields, produced by the particles and external sources. At the same time
completely neglect the mutual collisions of particles. Obviously, this simple
approach does not allow getting a complete quantitative description of the prop-
erties of plasma, but the results of this method are clear, so it largely enables a
qualitative understanding of many phenomena in plasma physics. For example,
microscopic description applied to the ensemble of charged particles, allows
establishment the presence of macroscopic charges and currents, which create the
additional magnetic and electric fields, etc.
The basis of the statistical description are the Boltzmann kinetic equation with
respect to the distribution function and the Maxwell equations for the description of
the field, and the particle distribution function f ð~ r;~
v; tÞ in the general case is
seven-dimensional, and the kinetic equation should be drawn up for each ith type of
plasma particles. In the case of elastic interaction of particles in plasma the kinetic
Boltzmann equation has the form:

Z1 Z2p Z1 h
@fi @fi ~ @fi X      i
þ~
vi  þ Fi  ¼ fi v0i fj v0j  fi ðvi Þfj vj gij bdbdedvj
@t @~
r @~
vi j
0 0 0
ð1:1Þ

© Springer International Publishing Switzerland 2017 1

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_1
2 1 The Theoretical Basis of the Low-Temperature Plasma

     
Here fi ðvi Þ; fj vj ; fi v0i ; fj v0j —the particle distribution function of ith and jth type
before the collision and after it, respectively; vi ; vj ; v0i ; v0j —the velocities of particles
~
of ith and jth type before the collision and after it, respectively; ~ Fi ¼ eE—the
mi
external force acting to the unit of mass of ith type particles; ~ E—the electric field
intensity; mi—the mass of ith type of particles; gij—the initial relative velocity of ith
and jth types of particles; b—the impact parameter; e—the azimuth angle of ori-
entation of the particle scattering plane.
In solving the Boltzmann kinetic equation some mathematical difficulties arise
because of the need to determine the collision integral in the right side of (1.1), that,
in its turn, depends on the specific conditions of plasma and the distribution
function, i.e., from the solution of the kinetic equation.
In many practical problems, the distribution function is close to Maxwell
equilibrium distribution function (~Fi ¼ @f @fi
@t ¼ @~
i
r ¼ 0).

 m 3=2  mv2 
fM ðvÞ ¼ 4pv2 n exp  ð1:2Þ
2pkT 2kT

Here n, m, v—the concentration, mass, and velocities of particles, respectively;

T—the temperature of particles; k—the Boltzmann constant.
At the same time, the detailed information about the distribution function of
particles is not always need. In solving the practical problems not the distribution
functions, but such integral quantities as temperature, velocity, pressure, obtained
by averaging of these parameters for the whole types of particles, are typically of
great interest. In general, for the calculation of these quantities it is necessary to
solve the Boltzmann kinetic equation and to find the distribution function of the
plasma particles, then to average and to determine the unknown quantity. However,
this direct method is rather difficult because of the mathematical complexity of the
Boltzmann equation solution, and in practice it is preferable to solve this problem in
a different way.
It is well known that the equations of magnet- and gas dynamic can be obtained
by the average of the kinetic equation, such as the continuity equation, the
momentum transfer equation, the equation of energy conservation, and others. The
most of mathematical models describing plasma processes depending on the specific
physical conditions is constructed by these equations. These models are called
magnet- and gas dynamic (MGD or MHD—magnetic- and hydro- dynamic) models.
In some practical applications for study the most important processes and building a
simplified model it is necessary to neglect by various phenomena because of the
complexity of the processes occurring in the plasma. Such MGD models are often
used in the analysis of individual physical processes, or in determining the degree of
influence of external parameters on the plasma properties. Some conservation laws
with some preset known variables are usually used in the open model. So the number
1.1 The Basic Approaches to the Description of Plasma 3

of unknown variables can be kept to a minimum, and many additional variables can
be neglected or can be determined by indirect or special way.
For example, to determine the temperature field of the arc in a long cylindrical
channel and the heat flux, carried away by convection, one can apply the
one-dimensional model of mass, current, and energy balance. The axial component
of velocity in the channel is determined from the total gas flow rate, and a
one-dimensional electromagnetic field—from a given current and the estimated
temperature field. Based on this model, a qualitative picture of gas heating in a long
channel of arc can be obtained, the temperature, heat flows, and distribution of the
electromagnetic field inside the channel can be found. Of course, such a greatly
simplified model does not allow determination the heating degree of the individual
components of plasma, as well as consideration the near-electrode processes and
analyze their impact for the formation of the physical processes in the Arc column.
Here are the other examples. To estimate the heating degree of the injected gas
according to its quantity and the predetermined value of electrons temperature, the
one-dimensional model of gas heating in half-space is widely used. To determine
the composition and properties of plasma the model of thermodynamic equilibrium
is applied. In this model the influence of the electromagnetic field to the distribution
of the various components of the velocity and excitation degree are neglected, and
the emission is optically thin for the plasma region where the mean free path of a
photon is more than the characteristic size of the plasma. Such similar models are
often used in practice and are very useful, especially when you want to obtain
quickly a qualitative picture of the different physical processes, occurring in plasma,
to analyze the influence degree of external parameters of plasma torches on plasma
properties, and to align them with the power supply, to determine the optimal area
for regulation of plasma torches parameters.
In some cases, the accuracy of the simplified models may not meet to the
requirements due to many unaccounted factors, and the plasma parameters can be
very different from real ones. Therefore, it is necessary to introduce some limita-
tions of applicability of various models of plasma processes. Historically, the
original models for description of plasma are based on the notion of thermodynamic
equilibrium of plasma.
Obviously, the plasma in laboratory and in nature can almost never be in a
complete thermodynamic equilibrium. However, it is extremely important to know
the ratio in case of thermodynamic equilibrium. First, many of the plasma are in a
state which is not much different from the complete thermodynamic equilibrium.
This condition is called, for example, the local thermodynamic equilibrium (LTE),
when the internal and external energy states of individual particles in plasma are the
same as for the complete thermodynamic equilibrium, but the blackbody radiation
is completely absent or weakened. Second, very often the ratio for full equilibrium,
are applied by using the principle of detailed balance, which means that in a
thermodynamic system in a state of complete equilibrium, each process has the
same velocity as the reverse one. The number of important relations for the plasma
in this state should be called.
4 1 The Theoretical Basis of the Low-Temperature Plasma

1.1.1 The Plasma Relations for the Full Thermodynamic

Equilibrium

1. Equality of temperatures for all groups of particles: electrons Te, ions Ti, atoms
Ta, molecules Tm: Te = Ti = Ta = Tm = T.
2. The quasi-neutrality of plasma in a volume of space with radius more than
Debye radius rD:
X
ne ¼ ðz  1Þns;z1 ð1:3Þ
s;z

The summation in the right side of (1.3) is over all stages of ionization z and on
all elements of s that are present in the plasma.
3. The Boltzmann distribution of particles in the excited levels. The ratio of the
particles of this type at the upper kth level of energy to their concentration on the
lower ith level is determined by the Boltzmann’s formula:
nk gk
¼ exp½DEki =ðkTÞ; ð1:4Þ
ni gi

Here gk, gi—the statistic weights (degeneracy) for kth and ith levels; DEki —the
difference of energies between the kth and ith levels.
For atoms and atomic ions the statistic weights gi that defines a degenerate of
levels depend on the quantum numbers of the total orbital momentum L, the
total spin S and total angular momentum of the electrons Ji:

gi ¼ ð2L þ 1Þð2S þ 1Þ; for the LS term

ð1:5Þ
gi ¼ 2Li þ 1; for ith level with Ji

In the literature, for example in tables of Moore [1], almost all elements and
values of Ji and ionization degrees Ei can be found.
Accordingly, the ratio of the concentration of particles ni on the ith quantum level
with the energy Ei and the total concentration of particles of the same kind will be:
ni gi
¼ exp½Ei =ðkTÞ ð1:6Þ
n Uin ðT Þ

The statistic sum Uin(T), internal to the bound states U(T) is defined as:
X
Uin ðT Þ ¼ gi exp½Ei =ðkTÞ ð1:7Þ
i

In (1.7) the summation is over all levels. However, the levels that are very close
to the ionization are not threshold, may not be realized, because the quantum
orbits of highly excited energy states perturb by the surrounding particles. Energy
1.1 The Basic Approaches to the Description of Plasma 5

of the last particle level that must be taken into account is Ei ¼ Eiþ  DEiþ ,
DEiþ —the reduction of the ionization potential. Practically, the ground state and
a large number of levels near the ionization limit make significant contribution to
the statistical sum. It is convenient to carry the calculation of statistical sums for
three components: the statistical weight of the ground state g1; the sum that
includes the weights of all the lower levels (from 2 to m); the sum of the weights
of highly excited hydrogen-reduced levels up to the ionization limit.

X
m
Uz1 ðT; ne Þ ¼ g1 þ gi exp½Ei =kT 
i¼2
 3=2
2 z2 E H
þ
X

þ þ exp Ez1 =kT gz;k exp Ez;k =ðkTÞ
3 DEz1 k
ð1:8Þ

Here EH is the ionization energy of hydrogen atoms.

The third term in the right side of (1.8) is generalized to the case where the
system of atomic levels is based on the different kth quantum levels of the
original ion. It should be noted that in many cases the static weight of the ground
state g1 is a good approximation for the statistical sum of the weights. The
statistical sum of electrons, atoms, and ions of single, double and triple ion-
ization are given in “Appendix A.1”.
The internal statistical sum of the weights of the molecules is determined by the
kind of internal degrees of freedom: Uin ¼ Ue Uv Ur , here Ue, Uv, and Ur—the
electronic, vibrational, and rotational statistical sum of the weights of the
molecules respectively.
Determination of electronic sum of the weights Ue is carried out in the same way
as for atoms. Calculation of the vibrational statistical sum of the weights is
increasingly done in an approximate of description of molecular vibrations by
the harmonic oscillators. For diatomic molecules Uv ¼ 1expð1:44m=T1
Þ, here ν,
−1
cm —the frequency of oscillator.
In practice the model of rigid rotor is most commonly used during the calcu-
lating the rotational static sum of weights. So, for diatomic molecules
Ur ¼ rhcBkT
r
, here Br —the rotational constant; r ¼ 1 for heteronuclear and
homonuclear molecules; h—the Planck’s constant; c—the speed of light in
vacuum. The calculation of the partition function of polyatomic molecules can
be found in the special books.
Note that there is a statistical sum of the weights of the translational motion of
particles with mass m that is defined as Utr ¼ 2pmkT
h2 . The total sum of the statistical
weights of the internal and translational motions of the particles UðTÞ ¼ Utr Uin .
4. Implementation of the law of acting mass for the ionization equilibrium in
plasma. If you compare the populations of two successive stages of ionization
particles of sth type, you will see that the application of Boltzmann formula (1.4)
6 1 The Theoretical Basis of the Low-Temperature Plasma

to the ionization particles leads to the equation of Saha for the ratio of the
concentration of particles on the (z − 1) stage and the concentration of particles
on the z stage ns;z .
 !
þ þ
ne ns;z Us;z ðT Þ 2pme kT 3=2 Es;z1  DEs;z1
¼2 exp  ¼ Ss;z1 ðT; ne Þ
ns;z1 Us;z1 ðT Þ h2 kT
ð1:9Þ

Here Us;z ðT Þ, Us;z1 ðT Þ—the statistical sum of the weights for the particles of
sth chemical type for two stages of ionization zth and (z − 1)-st; the value z = 1
corresponds to the neutral particle, z = 2—to the once ionized particle, etc.;
þ
Es;z1 —the ionization potential of the particles of sth chemical grade of the zth
stage.
5. The reduction of the ionization potential for the particles of sth chemical grade
þ 2
on the zth stage is DEs;z1 ¼ z 4pee0 rD , here e—the charge of electron; ε0—the
electrical constant.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
6. The radius of the Debye screening rD ¼ h P
e0 kT i. The summa-
2
2 e ne þ ðz1Þ ns;z1
s;z

tion in this formula is carried out for all zth stages of ionization and all the sth
chemical elements.
7. The Maxwell velocity distribution for each ith grade of plasma particles:
 m 3=2 m v2 
i i i
fM ðvi Þ ¼ 4pv2i ni exp ð1:10Þ
2pkT 2kT

8. The equation of state or the Dalton’s law for partial pressures. For low-density
plasma the equation of state of an ideal gas is right. In plasma physics, this
equation for a unit volume can be written as:
!
X
p¼ ne þ ns;z1 kT ð1:11Þ
s;z

The summation in the right side of (1.11) is carried out for all zth stages of
ionization and all the sth chemical elements.
9. The law of Kirchhoff. Photons have integer spin and obey to the statistics of
Bose–Einstein. If photons and plasma have the same temperature T, then
according to this statistics, the number of photons, distributed in plasma:
nm ¼ exp½hm=1ðkT Þ1. In this case, the plasma radiation has a continuous spectrum of
a blackbody and the equilibrium value of the spectral density of radiation and is
given by equation of Planck.
1.1 The Basic Approaches to the Description of Plasma 7

9
Bm ðTÞ ¼ exp½2hm =c 3 2
=
hm=ðkTÞ1
ð1:12Þ
2hc =k ;
2 5
Bk ðTÞ ¼ kc2 Bm ðTÞ ¼ exp½hc= ðkkT Þ1

Here λ—the wavelength.

The volumetric spectral density over all solid angles X is described as
9
R4p 8phm3 =c3 >
>
um ¼ 1
c Bm dX ¼ exp½hm=ðkkT Þ1
>
=
0
ð1:13Þ
R4p 8phc=k5 >
>
uk ¼ 1
c Bk dX ¼ >
exp½hm=ðkkT Þ1 ;
0

The integral bulk density of blackbody radiation is given in accordance with the
R1
Stefan–Boltzmann law: u ¼ um dm ¼ aT 4 , here a ¼ 7:56  1016 J/(m3 K4).
0
Consequently, the integrated flux of radiation from the surface of blackbody:
cu ca 4
P¼ ¼ T ¼ rT 4
4 4

Here the coefficient of the Stefan–Boltzmann r ¼ 5:67  108 W/(m2 K4). The
maximum intensity of blackbody radiation exists at kT ¼ 0:2898 cm K (Wien’s
displacement law).
In scientific literature in addition to the considered model of the complete
thermodynamic equilibrium the most known models are: the model of LTE, the
partial local thermodynamic equilibrium (PLTE), the coronal approximation model
(CAM), the collision radiation model (CRM). We restrict ourselves to a qualitative
description of the plasma models, details can be found out in [2–4].

1.1.2 The Model of Local Thermodynamic Equilibrium

in Plasma

The complete thermodynamic equilibrium plasma is rarely observed, but it is often

in a state close to LTE. The term LTE was originally intended to describe the state
of the plasma in the inner parts of the stars, where the radiation field was close to
the equilibrium. In stellar atmospheres the temperature gradient and, consequently,
the power flow always exist. In spite of this fact the local state of the plasma can be
quite clear described, assuming that the thermodynamic equilibrium takes place.
The LTE condition can be often found in laboratory plasmas with the sufficiently
high density of electrons, and where the collision processes involving the electrons
play a dominant role.
8 1 The Theoretical Basis of the Low-Temperature Plasma

The LTE model generalizes the case when deviations from the thermodynamic
equilibrium are relatively small and the model is based on the following
assumptions:
1. All of the distribution functions, except the distribution function of radiant
energy, are the Boltzmann functions with the temperature T;
2. The volumetric spectral density of radiation is not equal to this parameter for a
blackbody um ;
3. The concentration of particles in plasma is sufficient to the approximately sat-
isfaction of the detailed balance of particles in each volume and the Maxwell
distribution on velocities for all particles—the functional row of the particles
distributions in the phase space coordinates is preserved, but the particle con-
centration, that are average over the distributions, change with the change of the
spatial coordinates;
4. In the real cases plasma is optically transparent for many frequency bands, and a
significant fraction of the radiation leave the plasma volume free, thereby vio-
lating the conditions of complete thermodynamic equilibrium. In other words, in
LTE the detailed balance of processes of emission and absorption in the plasma
is broken. The speed of collision processes, leading to the settlement and release
of different levels of energy of substances have to exceed the applicable rates of
radiation processes—it is necessary for the standing of plasma in a state of LTE.
The probability of collision transition should be in an order of magnitude greater
than the probability of spontaneous emission. In line with these arguments Griem
[5], McWhirter [6], and Wilson [7] independently of one to another brought a
simple criterion for the existence of LTE in optically thin plasma with a predom-
inant role of collisions:

ne  1014 Te1=2 ðEk  Ei Þ3 ð1:14Þ

Here, ne is measured in cm−3, Te ; Ek ; Ei —in eV. The most energy “gap” Ek  Ei

usually equals E2 − E1. For the atoms like hydrogen with Te ¼ 1 eV the condition
of LTE can be obtained for ne  1017 cm−3. Note, that radiation-dominated plasma
in laboratory conditions is very rare. It can only be in stellar atmospheres and in
objects of astrophysics, so we do not give the conditions when we can expect the
existence of LTE radiation-dominated plasma.

1.1.3 The Model of Partial Local Thermodynamic

Equilibrium in Plasma

The model of PLTE corresponds to the further compared with the model of LTE
increase of the role of transport processes and external influences. This trend is
increasing with decreasing plasma density, especially of electrons. Because of the
1.1 The Basic Approaches to the Description of Plasma 9

smallness of the energy transfer d in one elastic collision, that is proportional to the
ratio of masses of the colliding particles (for electron and atom collisions
d ¼ 2me =ma ), the electron temperature Te can be very different from the tempera-
ture of the heavy components T—the gas is in a state of PLTE, both types of
particles can have Maxwell distribution on velocities in different temperatures. In
this case, the populations of the excited levels are still described by the Boltzmann
distribution, but at the temperature Te.
The presence of a set of different temperatures is a characteristic feature of PLTE
approximation. It is based on the difference in relaxation times–times of the
establishment of the stationary distributions within the different degrees of freedom
for particles of each types and, as a consequence, on the difference between the
corresponding average energies and temperatures. Accordingly, temperature of
electrons, ions, and neutral particles is used to describe their velocity distributions,
rotational, oscillation temperatures, and electronic levels to describe the distribution
of the particle populations on the corresponding levels and temperatures of ion-
ization to describe the distribution concentrations of ions of one chemical type and
different multiplicities. Convection and diffusion of electrons from one area of
plasma to another one can also lead to a violation of the ionization equilibrium in
plasma, etc. As a result, the usage of LTE and PLTE models for research most of
plasmas in channels of plasma torches is very important and we will consider them
in details in subsequent sections of this book.
The model of coronal approach has its name because of its use to describe the
conditions in the solar crown. It is the rarefied plasma with concentration
ne ¼ 106  1010 cm−3 and the average kinetic energy of electrons ee ¼ 0:1  1 keV.
For such case for very small densities of atoms and electrons the deviation from
thermodynamic equilibrium is so much that a detailed balance does not hold for
almost all processes. A simple description of the state of such plasma is possible only
in very special conditions, namely, when:
1. For all optical transitions plasma is optically thin;
2. The electron density and the density of radiant energy are so small that with a
probability close to unity, the excited atom is deactivated due to spontaneous
emission;
3. The only mechanism of recombination of ionized atoms is photorecombination;
4. Particles (atoms and ions) are ionized only by direct electron impact from the
ground state. Virtually, no collision excitation or de-excitation of the particles
from the excited levels and processes are carried out in three-body
recombination.
If we denote by Sz1;1 and az;1 —ionization and recombination coefficients from
the ground state (the index “1”), the balance between the processes of ionization
and recombination will be nz1;1 ne Sz1;1 ¼ nz;1 ne az;1 . Hence we can write
S ðT Þ
well-known equation of “coronal” equilibrium nnz1 z
¼ az1;1 e
z;1 ðTe Þ
. The function in the
right side is not universal as in Saha equation and depends on the type of atoms or
ions.
10 1 The Theoretical Basis of the Low-Temperature Plasma

Similarly balance
P for the collision excitation and spontaneous emission is
nz;1 ne Ki\k ¼ nz;k i\k Aki , here Ki\k —the coefficient of excitation in the transition
from the ground state to the ith level; Aki —the probability of an optical transition
from kth to the ith level in spontaneous emission.
For estimations and determine the conditions of application the model of the
coronal approach, the following expression can be used [8]:

ðz  1ÞTe
1=4
Eþ
Sz1;1 ðTe Þ ¼ 2:34  107 þ 7=4
exp 1:16  104 z1
Ez1 kTe
ðz  1Þ2 ð1:15Þ
az;1 ðTe Þ ¼ 2:8  1011 1=2
Te

4 f Ek  Ei
Ki\k ðTe Þ ¼ 1:7  10 1=2
exp 1:16  10 4
ðEk  Ei ÞTe kTe

Here f—the oscillator strength for the transition from the ith to the kth level;
þ
energies of the corresponding levels of excitation Ek ; Ei and ionization Ez1 can be
expressed in eV; Sz1;1 , az;1 , Ki\k —in cm /s.
3

In addition to use in the spectroscopy of the solar crown, this model was ana-
lyzed with respect to its application to more dense plasmas like plasma in
TOKOMAK reactor and beam discharges.
The collision–radiation model is applied to describe the plasma with a strong
disequilibrium when the above models can’t be used. This model includes a scheme
of full description of particles levels together with the equations for the kinetics of
the continuous spectrum of energy and equations of electrodynamics to find the
field. The balance of the populations of the quantum levels of particles of each
chemical types is described by the general kinetic equations like

dnk X X
¼ ðni mik  nk mki Þ þ Sak þ Qk ð1:16Þ
dt i a

Here mik , mki —the frequencies of collision and radiation transitions between the kth
and ith levels; Sak —the source function that describes the creation and annihilation
of particles with the concentration nk and with creation the particles of the other
chemical type; Qk —the traffic carried by different particles.
Thus, in the considered plasma models the Boltzmann and Saha equations are
often applied to plasmas with complex processes of collision and radiation, leading
to different degrees of deviation from the thermodynamic equilibrium.
The above-mentioned plasma models, of course, can be regarded as special cases
of CRM. To solve the problems of theoretical analysis plasma in laboratory and
astrophysics is more often expressed with regard to its deviation from the tem-
perature, ionization, and radiation equilibrium with the other specific conditions of
collision–radiation processes.
1.1 The Basic Approaches to the Description of Plasma 11

Most equations in these models are integral–differential with distributed

parameters. They can be obtained by converting the kinetic Boltzmann equation to
the system of equations for the moments corresponding to different conservation
laws; for example, for a system of 13 MHD equations [2, 9] allows description the
evolution of plasma on a macroscopic scale. On the basis of kinetic equations
enough rigorous theory of transport phenomena for multicomponent plasma is now
established that justify the hydrodynamic approximation and represent the appro-
priate magnet-hydrodynamic equations for plasma.
Among the conditions of application of the hydrodynamic approximation the
usual restrictions for gas dynamics includes. The main among them is the
requirement that the mean free path and the time between collisions for all plasma
components should be much less than the characteristic spatial and temporal scales
of the macroscopic changes. The fulfillment of these conditions allows distraction
from the discrete structure of the plasma to treat it as a continuous medium, sig-
nificantly simplify the equations of moments of the kinetic theory and use the linear
transport relations. The coefficients of proportionality in these equations, as a rule,
are corresponding coefficients to the transfer coefficients, and the equations them-
selves—to the laws of conservation of energy, momentum, and mass.
Thus, the MHD model for plasma description is the most widely used in the
study of plasma processes, and a mathematical description adequately meets the
challenges of modeling the physical processes occurring in the channel of plasma
torch. Because of the wide practical use of MHD models in the study of plasma
their basic positions are given in following.

1.2 Model of Equilibrium Plasma

The model of equilibrium plasma is based on the assumption of thermodynamic

equilibrium in plasma. The great number of theoretical and experimental data
indicates that the conditions of complete thermodynamic equilibrium in plasma
(1.3)–(1.13) are almost not carried out for real plasma. However, the model of
thermal equilibrium in plasma is easy to calculate the heating of the gas and plasma
in the plasma torches. This fact is explained by the relative simplicity of the
equilibrium plasma model and lots of real physical factors specific for the plasma.
Here are some of them:
The dense thermal equilibrium plasma creates in channels of most plasma tor-
ches that operate at atmospheric pressure or close to atmospheric pressure. The
current density in this plasma is j > 102 A/cm2 and the concentration of charged
particles ne [ 1016 cm−3. In these cases, the elastic collisions between electrons
and heavy particles (ions and atoms) in plasma occurs with a frequency of
m [ 1011 s−1 and so effective that the difference in temperature groups of electrons
and heavy particles is negligible and the assumption of the equality of the tem-
perature for all groups of particles will be justified [10];
12 1 The Theoretical Basis of the Low-Temperature Plasma

The mean free path of the plasma particles is often much smaller than the spatial
dimension of its flow that allows us to consider the plasma as a continuous medium
having electrical conductivity, thermal conductivity, viscosity and other charac-
teristics, as well as the thermodynamic functions of plasma depend only on its
temperature and pressure. In this case, the elementary processes of interaction
between groups of particles and between particles of each group are taken into
account in determining the transport coefficients and thermodynamic functions in
terms of averaged over the distribution function the interaction cross sections;
The plasma radiation is optically thin for a wide range of frequencies and
photons go out of its volume free, without participating in the activation of pro-
cesses such as “recolonization” of the various excited levels of particles.
The model of equilibrium plasma typically consists of: energy equation that
express the law of energy conservation; the momentum equations—the law of
momentum conservation; Maxwell’s electromagnetic equations—the law of the
field conservation; the continuity equation for hole gas—the law of mass conser-
vation; the system of equations defining the dependence of the transport coefficients
and thermodynamic functions of the plasma on temperature.

1.2.1 The Energy Balance Equation

The energy balance equation is the fundamental equation in the analytical

description and mathematical modeling of plasma and plasma torches. It allows
obtaining the distribution of the plasma temperature in the channel of plasma torch
and connects this temperature with external electric circuit parameters: current I,
voltage U, channel diameter d, flow of the heated gas G. This circumstance is most
important to the plasma torch engineering and technology, because it gives the
opportunity to see that external parameters of the plasma torch affect to the tem-
perature and velocity of plasma, and to see the connection between them, it is
necessary for solving the problems of regulation and control of plasma processes.
The complete change of the plasma temperature in space and time
DT ¼ @T þ ð~vrÞT—the substantial derivative is determined by the combined effect
Dt @t
of energy sources and its sinks. The power rE 2 in the form of Joule heating is
allocated in the unit of plasma volume. In addition, at the high speed of plasma flow
and its large gradients heat is added released to the power source because of the
forces of the viscous friction Udiss (dissipation function).
The retraction of energy from the volume unit occurs due to the influence
temperature gradients divðkgradTÞ to the thermal conduction and due to convection
associated with the gas movement divðq~ vHÞ, here H—the enthalpy of plasma. The
radiation is optically thin and photons leave the plasma volume free without
arousing the different degrees of freedom of the plasma particles. Thus, the power
loss Urad ðTÞ due to radiation can be represented as a function of temperature and
enter it at the energy equation with the sign “minus”:
1.2 Model of Equilibrium Plasma 13

DT
qcp ¼ divðkgradTÞ þ rE 2  Urad þ lUdiss ð1:17Þ
Dt

Here q; cp ; k; r; l—the density, heat capacity, thermal conductivity, electrical

conductivity and viscosity of plasma, respectively; E—the electric field intensity.
For ideal gases, liquids, and plasma at the constant pressure: cp gradT ¼ gradH.
Substituting this expression into Eq. (1.17), we get another entry of the energy
equation with the enthalpy H:

@ðqHÞ k
þ divðq~
vHÞ ¼ div gradH þ rE 2  Urad þ lUdiss ð1:18Þ
@t cp

The synopsis of the energy equation with enthalpy (1.18) is rather than with
temperature (1.17), because the enthalpy characterizes all kinds of plasma energy
that allows account accurately for all channels and influx of heat removal from the
unit of plasma volume. However, the solution of the energy equation with enthalpy
is inconvenient because of the fact that the properties of the plasma including the
thermodynamic transfer function are often represented as a function not of enthalpy
H, but the plasma
temperature T.
d H
If cp ¼ dT ¼ const is taken at the constant pressure p for a small plasma
p¼const
volume, the energy equation can be written with temperature:

@ðqTÞ k 1
2 
þ divðq~
vTÞ ¼ div gradT þ rE  Urad þ lUdiss ð1:19Þ
@t cp cp

The energy balance equation for the equilibrium plasma, written in the form
(1.19), is convenient for solving, and all possible types of stored energy are taken
into account through the total heat capacity cp of plasma.

1.2.2 The Momentum Equations of the Plasma Flow

An important property of plasma generated by plasma torch, in addition to its high

temperature T, is a great ~ v. The plasma velocity in modern plasma torches of
various designs in different modes of operation may be from a few meters to several
kilometers per second. The high temperature and velocity of plasma provide the
basic technological advantages of plasma torches for the metal cutting, spray
coating, treatment (melting) of fine refractory particles, plasma chemical synthesis
and decomposition of substances, and others.
The movement of the plasma flow is described by the momentum conservation
law. Complete change of momentum of the system in time and space is the result of
the action of various forces applied to the system: ~
gq—gravity, here ~g—the accel-
eration of gravity; gradp—the force of static pressure gradient p; ~
FB ¼ ~j  ~
B—the
14 1 The Theoretical Basis of the Low-Temperature Plasma

electromagnetic force, here ~j—vector conduction current density, ~

B—the magnetic
induction vector of the electromagnetic field; ~
Fl —the viscous friction forces.

@ ðq~
vÞ
þ ðq~
vrÞ~ gq  rp þ ~
v ¼~ FB þ ~
Fl ð1:20Þ
@t

Gas medium that include low-temperature plasma are compressible. The viscous
friction force is determined by the tangent and some related with volume (in
compression and expansion of an elementary volume of plasma) and shear defor-
mations with additional requirements:

Fl ¼ div~
s ð1:21Þ

Viscous stress tensor is defined as [3]:

 
@vi @vj 2
sij ¼ l þ  dij div~
v ð1:22Þ
@xj @xi 3

Here vi ; vj —the velocity components on coordinates xi, xj; l—the plasma viscosity;
dij —the delta function of Cronecker, dij ¼ 1 i  j and dij ¼ 0 for i 6¼ j.
The first term in (1.22) is the shear viscous friction, and the second—the second
volume friction with the viscosity that characterizes the dissipation of the kinetic
energy of heavy components in thermodynamic disequilibrium and volumetric
strain [3].

1.2.3 The Continuity Equation of the Plasma Flow

To describe the motion of plasma flow another fundamental law—the law mass of
conservation is needed. In general, this law is expressed by the equation of con-
tinuity of the entire gas:

@q
þ divðq~
vÞ ¼ 0 ð1:23Þ
@t

1.2.4 The Maxwell’s Equations

The Maxwell’s equations are the basic equations for a wide variety of different
types of electromagnetic problems for plasma torches. In the energy equation
rE 2 —power of Joule heating in the unit of plasma volume is a function of the
source expression and in the momentum equations ~ FB ¼ ~j  ~
B—electromagnetic
1.2 Model of Equilibrium Plasma 15

force in plasma is the result of the interaction of the plasma current with the
magnetic field. Thus, at first it is necessary to define the electromagnetic field in the
channel of the plasma torch solutions for the solving of momentum equations and
energy equation. Maxwell’s equations include four basic equations:

X @~
B q
~¼
rotH ~j; rot~
E¼ ; div~
B ¼ 0; div~
E¼ c ð1:24Þ
@t ee0
P
Here ~j—the sum of the current densities of the plasma: current conduction r~ E,
current transfer qc~ ~ ~
v and bias current @ D=@t; H—the magnetic field strength;
~ ~
B ¼ ll0 H—magnetic induction; ~ ~
E—the electric field strength; D—the electric
shift; qc ¼ eðZni  ne Þ—the density of electrostatic charges; (Z-ion charge; for
spatial scale greater then the Debye radius (r [ rD ) and quasi-neutral plasma
ðZni ¼ ne Þ and (qc ¼ 0); l0 ; e0 —magnetic and dielectric permittivity in a vacuum;
l; e—relative magnetic and dielectric permeability of plasma (in many cases, they
take the same values as in a vacuum, i.e., l ¼ e ¼ 1).

1.2.5 The Equations of Equilibrium Plasma Model

in a Cylindrical Coordinate System

In most cases, the design of channels is a cylindrical symmetric system. Therefore,

for convenience of the solutions, it is necessary to write equations of the model of
equilibrium plasma in a cylindrical coordinate system. The equation will be sym-
metric in a cylindrical coordinate system around the axes r and z:
The energy balance equation:

@ @ 1@ @ k @T
ðqT Þ þ ðqvz T Þ þ ðrqvr T Þ ¼
@t @z r @r @z cp @z
 ð1:25Þ
1@ k @T 1
2 
þ r þ rE  Urad þ lUdiss
r @r cp @r cp

In (1.25) the dissipative function is:

"  2   # 
2  
@vz @vr vr 2 @  vu  2 @vu 2
Udiss ¼2 þ þ þ r þ
@z @r r @r r @z
 2  2
@vz @vr 2 @vz 1 @ ðrvr Þ
þ þ  þ
@r @z 3 @z r @r

In the cases when the velocity of the plasma is much less than the speed of
sound, the power of thermal sources is much less electrical power Udiss  rE 2 due
to viscous friction and it can be omitted from the energy equation.
16 1 The Theoretical Basis of the Low-Temperature Plasma

The momentum equations of the gas relative to the three components of velocity
vz ; vr ; vu :

@ @ 1@ @p
~ ~ @ @vz
ðqvz Þ þ ðqvz vz Þ þ ðrqvr vz Þ ¼  þ j  B z þ2 l
@t @z r @r @z @z @z
      ð1:26Þ
1@ @vz @vr 2@ @vz 1 @ ðrvr Þ
þ rl þ  l þ þ qgz
r @r @r @z 3 @z @z r @r

@ @ 1@ @p
~ ~ 2@ @vr
ðqvr Þ þ ðqvz vr Þ þ ðrqvr vr Þ ¼  þ jB rþ rl
@t @z r @r @r r @r @r
2      
vu @ @vz @vr 2vr 2 @ @vz 1 @ðrvr Þ
þq þ l þ l 2  l þ
r @z @r @z r 3 @r @z r @r
ð1:27Þ

@  @   1@  
qvu þ qvz vu þ rqvr vu
@t @z r @r
  ð1:28Þ
@ @vu 1@ @vu vu @ vr vu
¼ l þ rl  2 ðrlÞ  q
@z @z r @r @r r @r r

The continuity equation:

@q @ 1@
þ ðqvz Þ þ ðrqvr Þ ¼ 0 ð1:29Þ
@t @z r @r

If in the task the analyzed time interval is longer than the time of different
processes of ionization, recombination and dissociation in the plasma, and the speed
of fluctuations for different physical processes is much less than the reciprocal speed
of their setting, the processes in channels of plasma torches can be stationary.
In the specific case Maxwell’s electromagnetic equations (1.24) can be different.
For example, for a two-dimensional electromagnetic field of RF plasma torches
because of the cylindrical symmetry of the limit inductor current and conduction
currents in the plasma (transfer and offset currents equal to zero), we have the
following relations for the components of intensities of the electric and magnetic
fields: Ez ¼ Er ¼ 0; Eu 6¼ 0; Hu ¼ 0; Hz 6¼ 0; Hr 6¼ 0. Then the Maxwell’s equa-
tions will be:

@Eu @Hr @Hr @Hz

¼ l0 ;  ¼ rEu ;
@z @t @z @r
ð1:30Þ
1@   @Hz @Hz 1 @
rEu ¼ l0 ; þ ðrHr Þ ¼ 0
r @r @t @z r @r
1.2 Model of Equilibrium Plasma 17

For two-dimensional electromagnetic field of Arc plasma torches Eu ¼ 0;

Ez 6¼ 0; Er 6¼ 0; Hu 6¼ 0; Hz ¼ Hr ¼ 0, Maxwell’s equations have the following
form:

@Hu @Er @Ez @Hu

 ¼ rEr ;  ¼ l0
@z @z @r @t
ð1:31Þ
1@   @Ez 1 @
rHu ¼ rEz ; þ ðrEr Þ ¼ 0
r @r @z r @r

In addition to energy equation, momentum equations, continuity equation, and

Maxwell’s equations the model of thermal equilibrium plasma include Saha
equation and Dalton’s law of partial pressures for determining the composition of
the plasma, as well as a system of equations for the calculation of transport coef-
ficients and thermodynamic functions depending on the temperature of the plasma:
rðTÞ, kðTÞ, lðTÞ, Urad ðTÞ, qðTÞ, cp ðTÞ, HðTÞ, and others.
Algorithms for the solution of Eqs. (1.25)–(1.31) are given in a number of
textbooks and monographs on the computational mathematics and mathematical
physics. There are some features in the solution of such a system. The equations
describing the thermal equilibrium plasma (1.25)–(1.31) together with a system of
equations for determining the composition, the transport coefficients and thermo-
dynamic functions of plasma form a closed system of nonlinear differential equa-
tions, i.e., we have a model with distributed parameters that depend on the solution
of equations. Thermodynamic functions and transport coefficients depend on the
temperature of plasma that should be found from the energy equation. However, the
solution can be found if the solutions of Maxwell’s equations (for calculation of
specific power rE 2 in plasma) and the momentum equations (for determination of
the velocity of plasma particles) are known. But they cannot be solved without
knowledge about the temperature field, etc.
Thus, to solve the closed system of equations of the model plasma iterative and
recursive procedures that allow for the nonlinearity of the equations and the
interconnectedness between these equations should be applied. Experience shows
that such a sequential iterative procedure for solving the system of equations for the
plasma is very unstable, especially in the global iterative loop in the transition from
solutions of an equation to solve the other. A small change in the iterative procedure
of solving a single equation often causes large changes in the solution of the
following equations, which contributes to instability of the solutions of equations in
general. Thus, one of the key issues in modeling highly nonlinear systems has been
and remains the question of the sustainability of computing and iterative proce-
dures. In computational mathematics, besides the widely used method of
under-relaxation that can partially compensate for the lack of information after each
decision is still in general, there is no stable algorithm for solving systems of
nonlinear differential equations, including those in the simulation of the plasma.
18 1 The Theoretical Basis of the Low-Temperature Plasma

1.3 The Two-Temperature Model of Nonequilibrium

Plasma

The state of thermodynamic equilibrium (1.3)–(1.13), that is required as a basis for

constructing a model of equilibrium plasma, is very rare. Application of the equi-
librium plasma model for description of the physical processes in plasma torches is
not always justified. As it is shown by numerous experiments, plasma in its different
manifestations in the channels of plasma torches in nature has some spatial regions
where the equality of temperatures of electrons and heavy particles (atoms, ions,
molecules) is violated. Also it is found that there is no a single Maxwell’s velocity
distribution and the Boltzmann’s distribution over the population of the excited
energy levels for all types of particles, there is no the ionization equilibrium. In such
cases, plasma is nonequilibrium, and the model of equilibrium plasma is insufficient
for its description.
If the thermal plasma will be considered like a continuous medium with a
common temperature, we have deliberately refused to consider two main processes:
ionization and heating of the plasma particles. Ionization from the collision process
turned into a thermal process described by the law of active mass, and the exchange
of energy between electrons, atoms, and ions is generally not included in the energy
equation (the law of energy conservation). However, plasma in plasma torch
becomes an effective means of thermal impact only if it is heated by heavy com-
ponents: atoms, ions, molecules. In order to obtain such plasma, a mechanism, that
provides efficient energy transfer from electrons to atoms, molecules and ions, is
needed. Such mechanisms are the elastic collisions of electrons with atoms, ions
and molecules, accompanied by the transfer of energy and momentum. The chal-
lenge is that it is required to determine not only the mechanisms of energy in the
plasma, but also their most favorable conditions for the flow and the degree of
plasma heating subject to further its practical use.
In the two-temperature plasma model it is assumes that the temperature of
ionization, excitation, and settlement are close to the kinetic electron temperature Te
and the temperature distribution is identified with the temperature T of the heavy
components of the atom–ion gas. The validity of these assumptions was confirmed
in many researches [11–13]. However, it should be noted that due to the spatial
heterogeneity and temporal changes in the relevant portable and kinetic processes
of balance, the overall temperature for individual or all degrees of freedom can not
be installed. To describe multistage dynamic plasma systems, in general, it is
necessary to introduce several temperatures, concentrations, and distributions [14].
Moreover, the nonequilibrium plasma conditions can lead to a more complicated
form of the distribution functions that are impossible to characterize any temper-
ature [12]. Therefore, consistent description of nonequilibrium plasma requires the
use of sophisticated methods of physical kinetics that if applied to research prob-
lems of gas heating, greatly complicates their decision.
In such a situation it is essential to explore the most typical types of nonequi-
librium state and identify the conditions when plasma is in the state of local thermal
1.3 The Two-Temperature Model of Nonequilibrium Plasma 19

Fig. 1.1 Scheme of energy exchange in the two-temperature plasma model

equilibrium. In this respect, the most interesting is analysis of the thermal and
ionization nonequilibrium plasma. Conditions for the creation of such a plasma and
its save for a quite long time are determined by the energy balance of the electron
gas and heavy components and the processes of ionization and recombination,
bipolar diffusion, thermal diffusion, radiation and other processes, occurring
simultaneously in plasma.
The scheme of energy exchange in the two-temperature model of nonequilibrium
plasma is shown in Fig. 1.1 (for simplicity, the case of the atomic plasma is
considered). Specific power of the supplied electric field re E 2 is directly perceived
by the easiest and the most moving electron gas, and atoms and ions are heated
primarily as a result of elastic collisions with electrons [4, 13]. Efficiency of energy
transfer in one elastic collision is proportional to the ratio of the masses of the
colliding particles mi ; mj :

2mi mj
d¼ 2 ð1:32Þ
mi þ mj

For the collisions electron–atom: d ¼ 2me =ma , atom-ion d ¼ 1=2, etc. The
noticeable energy metabolism in a single collision occurs between the particles with
similar mass. That is why it is possible to assume that the kinetic temperatures of
atoms, ions, and molecules in plasma are the same.
A relatively small fraction of energy transfers in collisions of electron–atom and
electron–ion so the local thermal equilibrium establishes separately for the electron
gas and for the heavy components (atoms, ions) with temperatures Te and T,
respectively. The presence of a possible low pressure, as well as the convection,
diffusion and other transport processes with the continuous energy input from the
field to the electron gas can also enhance the thermal nonequilibrium in plasma.
This fact makes it possible to introduce into consideration the two-temperature
model of thermally nonequilibrium plasma, where the temperature of heavy
20 1 The Theoretical Basis of the Low-Temperature Plasma

components (atoms and ions) is considered to be approximately equal

(Ta ¼ Ti ¼ T), and the electron temperature Te may be different from the temper-
ature of the heavy components. Thus, the specific power re E2 produced by the
electrons from the field is consumed for heating of the heavy components of plasma
(the ions and atoms) for ionization and activation of different atomic and molecular
levels in plasma, on convective, conductive heat transfer flows and to radiation.
Similarly, for the heavy components the energy that derived from the elastic
electron impact is used for convective and conductive heat transfer processes
(Fig. 1.1). As in the case of equilibrium plasma flow to subsonic plasma in the
energy equation energy Φdiss dissipation can be neglected.
Here: He ; Hai ; ke ; kai —the enthalpy and heat conduction of electron and atom-ion
gas, respectively, kai ¼ kaic þ kint , here kaic ; kint —the contact and internal thermal
conductivity of the atom-ion gas; re —the electrical conductivity of electrons;
Urad —the power loss due to radiation.
For many frequency bands the radiation can be considered optically thin, and
photons can easy go out of the spatial region of the discharge. The kinetic energy of
the electrons, transmitted to atoms and ions by elastic collisions per unit of time per
unit of volume:

3
kdnemðTe  T Þ ¼ bðTe  T Þ ð1:33Þ
2

Here: m ¼ mea þ mei ¼ ve ðna Qea þ ni Qei Þ—the sum of the average frequency of
electron collisions with atoms and ions with the momentum transfer (the average
number of collisions of electrons with atoms and ions per unit of volume and per
qffiffiffiffiffiffiffi
unit of time); ve ¼ 8kTpme —the average thermal velocity of the electron; na ; ni ; ne —
e

the concentration of atoms, ions, and electrons; Qea ; Qei —the average effective
collision cross sections of electron–atom and electron–ion;
qffiffiffiffiffiffiffi
b ¼ 2 kdnem ¼ 2 k ma ne pme ðna Qea þ ni Qei Þ—the coefficient of energy exchange
3 3 2me 8kTe

between electrons, atoms and ions.

Equation (1.33) can be used to determine the effectiveness of the heating of
heavy components of the plasma and it is a major one in the physics of plasma
heating.
The presence of two temperatures in plasma requires the use of two energy
equations instead of a single one: for electrons and for heavy components. The
relationship between these equations is established through the energy of elastic
interaction in collisions between electrons and heavy particles.
Taking into account the convective and conductive transport streams the energy
equation of electron gas can be written as follows:

@ ðqHe Þ
þ divðq~
ve He Þ ¼ divðke gradTe Þ þ re E 2  bðTe  T Þ  Urad ð1:34Þ
@t
1.3 The Two-Temperature Model of Nonequilibrium Plasma 21

Note that the speed of electron gas

ve ¼ ~
~ v þ~
vD þ~
vA þ~
vT

is the sum of the velocity ~

v of total gas flow, the drift of electrons in an electric field,

~j
~
vD ¼  ;
ene

the bipolar diffusion

Te þ T gradne
~
vA ¼ Damb ;
2T ne

the thermal diffusion [15]

gradðTe þ T Þ
~
vT ¼ DT :
2T

Heat transfer due to bipolar diffusion and thermal diffusion of electrons is often
expressed through the reactive thermal conductivity and it is taken into account when
determining the overall thermal conductivity of the electron gas, ke ¼ kec þ kR , here
kec ; kR —contact and reactive heat conduction of electrons (Sect. 4.3.1). The exit of
the drift electron flow energy in the direction of the electric field re E 2 is determined
by the longitudinal component of the electric conductivity (re ¼ rek ).
Light and mobile electrons that receive energy from the electric field, play a
major role in the processes of ionization, recombination and emission of plasma
therefore the ionization energy should be included in the general formula for the
electron enthalpy:

1 5
He ¼ kTe þ Ei ne
q 2

Similarly, in the energy balance of two-temperature plasma model the radiation

power takes into account. The power losses on the total radiation of plasma can be
logically included in the energy equation for electrons, but not to the energy
equation for the heavy components, although the radiation is largely the result of
interaction of electrons in the free–free, free–bound and bound–bound transitions.
With these assumptions, the energy equation in the steady state for electron gas will
be as follows:
 
5
div kTe þ Ei ne~
ve ¼ divðke gradTe Þ þ re E 2  bðTe  T Þ  Urad ð1:35Þ
2
22 1 The Theoretical Basis of the Low-Temperature Plasma

In a cylindrical coordinate system and taking into account the cylindrical symmetry
of the energy equation of the electron gas (1.35) can be written as:
    
@ 5 1@ 5
kTe þ Ei ne vez þ r kTe þ Ei ne ver
@z 2 r @r 2
  ð1:36Þ
@ @Te 1@ @Te
¼ ke þ rke þ re E 2  bðTe  T Þ  Urad
@z @z r @r @r

With the help of (1.33) a source of energy in the energy equation of heavy (atom–
ion) components is considered. This equation has the following general form:

@ ðqHai Þ
þ divðq~
vHai Þ ¼ divðkai gradTÞ þ bðTe  T Þ ð1:37Þ
@t

Here Hai ¼ q1 52 kT ðna þ ni Þ—the enthalpy of the atom–ion component; kai —heat
conductivity of atom–ion component. Atom–ion gas moves with the total gas flow
velocity. Given the stationary energy equation for atom–ionic component has the form
 
5
div kT ðna þ ni Þ~
v ¼ divðkai gradTÞ þ bðTe  T Þ ð1:38Þ
2

In a cylindrical coordinate system the energy equation for atom–ionic gas is

written as
   
@ 5 1@ 5
kT ðna þ ni Þvz þ r kT ðna þ ni Þvr
@z 2 r @r 2
  ð1:39Þ
@ @T 1@ @T
¼ kai þ rkai þ b ð Te  T Þ
@z @z r @r @r

For plasma in a state of ionization equilibrium, the concentration of particles is

determined by the law of action mass (1.9). According to this law, the amount of
energy spent on ionization of the unit of plasma volume at any point of space and
per unit of time equals to the amount of energy released in the same volume in the
reverse recombination reaction. For the plasma flow this model does not correspond
to the real picture of the physical processes in plasma. Because of its heterogeneity
the region, where the energy release and active field ionization mainly exists in
plasma, i.e., the energy of system is mainly spent on ionization, and vice versa,
there is an area, where the reverse recombination reaction of particles and release of
energy recombination are predominant. Thus, there is a direct connection between
the particle and energy balance or between ionization and thermal equilibrium states
of plasma. In other words, there is a connection between the laws of conservation of
energy and particles. We draw attention to the fact that for low-temperature plasma
the ionization energy of particles is usually more than their thermal energy, such as
argon single ionization energy Eiþ ¼ 15:755 eV compared with the thermal energy
1.3 The Two-Temperature Model of Nonequilibrium Plasma 23

kT ¼ 1  3 eV for most plasma in RF, Arc and other types of plasma torches.
Consequently, in determination of the energy and particles balance in plasma, it is
important to consider its model, taking into account not only a violation of the
temperature equilibrium, but also the ionization one.

1.3.1 The Violation of Ionization Equilibrium in Plasma.

Equation of Ionization Equilibrium

Ionization equilibrium is violated if the processes of ionization and recombination of

particles occur with the different speeds, i.e., the number of ionization acts is not
equal to the number of recombination acts per unit of time and per unit of plasma
volume. In the steady-state conditions the radiation from plasma volume, the dif-
fusion of charged particles and forced convection of heat contribute to a violation of
the ionization equilibrium. To neglect the first of these mechanisms a high con-
centration of electrons ne  1016 cm−3 requires for example at the atmospheric
pressure. The role of diffusion (bipolar diffusion and thermal diffusion) clearly
reveals for plasma flows in the channel of plasma torch with large gradients of
electron density and temperature. In this case, the diffusion of charged particles from
the axial zone reduces the electron density on the axis and on the periphery of their
increase in the comparison with the equilibrium model. The forced convection can
play a major role in particles balance and it increases with increasing of the gas flow,
blown through the discharge. The special analysis is carried out in [16], it shows that
the convection can lead to a significant deviation of plasma from the ionization
equilibrium under a stream of cold gas through the discharge with specific con-
sumption qv [ 0:1 g/(cm2 s) at temperature Te = 10000 K and pressure p = 105 Pa.
However, the main mechanisms of the appearance and disappearance of charged
particles in low-temperature plasma, especially in dense plasma of high pressure are
the ionization of atoms by direct electron impact and the triple recombination [3, 4].
The velocity of charged particles formation as a result of the ionization of atoms by
direct electron impact is proportional to the concentrations of electrons ne, atoms na,
and some ratio of ionization of this process ki (it is determined by the ionization
cross section) and is equal to ki ne na .
The step ionization is very significant along with the direct ionization of atoms
from the ground state in collisions with electrons: the average thermal energy of the
electrons in the normal plasma torch is about 1 − 3 eV, and the ionization energy
of atoms from the ground level for most gases Ei > 5 eV. Therefore, the number of
electrons with such energy in the low-temperature plasma can be determined by the
Maxwell distribution function, and this number is very small. In dense plasma of
high pressure the ionization from the excited level (the step ionization) is the most
probable process of ionization in plasma torches. In this case, the ionization
coefficient ki takes into account the total probability (i.e., the effective cross section)
of processes of the direct and stepwise ionizations.
24 1 The Theoretical Basis of the Low-Temperature Plasma

In a dense plasma the recombination of electron–ion pairs often occurs with the
participation of the third particle (the Thomson theory)—atom, ion, and electron
[2–4]. Analysis of specific conditions of recombination in plasma torches at a high
concentration of charged particles is given in [3, 4], it shows that the cross section of
three-body recombination is extremely high in conditions of the presence of electron
as the third particle. This process is particularly effective with the participation of
slow electrons in the field of ions for a long time. First of all the capture of an
electron to a distant ion orbit occurs, and the role of the third electron is in the
selection of energy released during the deceleration of electron in the field of ion.
Next, the recombination occurs in stages, covering all the excited states of an atom
with radiation transition, and the quanta of radiation energy is emitted in the spectral
lines, i.e., shock radiation or triple recombination occurs. In this case, the recom-
bination velocity of particles is proportional to the concentration of ions ni, and to the
square of the electron density n2e (two electrons take part in the recombination
reaction simultaneously) and is proportional to the coefficients kr, characterized the
probability of triple recombination process, i.e., it is equal to kr ni n2e .
Thus, the velocity of charged particles formation as a result of two opposing
processes: the electron impact ionization and triple recombination, involving the
third electron n_ e ¼ ki na ne  kr ni n2e .
Taking into account the transport of charged particles due to the convection and
bipolar diffusion, the equation of ionization equilibrium of plasma (the single
ionization) is:

@ne
þ divðne~
ve Þ ¼ divðDamb gradne Þ þ ki na ne  kr ni n2e ð1:40Þ
@t

For quasi-neutral plasma (ne = ni), using the principle of the detailed balance [2, 3],
the equation of ionization equilibrium can be written as follows:
 2
@ne n n2
þ divðne~
ve Þ ¼ divðDamb gradne Þ þ kr ne na e  e ð1:41Þ
@t na na

Here n e ; n a —the concentration of electrons and atoms in ionization equilibrium (it

is determined by the Saha equation).
In the case of stationary and cylindrical axe-symmetry the equation of ionization
equilibrium can be written as:
 
@ðne vez Þ 1 @ðrne ver Þ @ @ne 1@ @ne
þ ¼ Damb þ rDamb
@z r @r @z @z r @r @r
 2 ð1:42Þ
ne n2e
þ k r n e na 
na na

The sources function  in equation of ionization equilibrium is often referred

n 2 n2e
n_ e ¼ kr ne na n  na .
e
a
1.3 The Two-Temperature Model of Nonequilibrium Plasma 25

As in the case of equilibriums plasma, in the model of nonequilibrium plasma it

is considered, that the particle velocity distribution is the Maxwell one and the
“population” distribution—the Boltzmann one. The two-temperature model must
also contain the following equations: the momentum equations [(1.26)–(1.28)]; the
equation of continuity for all gas (1.29); the Maxwell equations describing the field
[(1.30), (1.31)]; the system of equations for determination of two temperature
transfer coefficients and the thermodynamic functions of plasma; the Dalton’s law
of partial pressures; equations for calculating the coefficients of impact ionization
and triple recombination of charged particles and others.

1.4 The Two-Speed Plasma Model

The difference in mass of electrons and other particles not only leads to disruption
of the equilibrium temperature between electrons and heavy components in plasma,
but also to the difference of velocities of flow. Because of the small mass electrons
drift in the field with a directional speed different from the speed of the total gas
flow, despite the fact that one effective electron blow can transfer almost all of its
directional momentum to heavy particles. Unlike the electrons, ions have greater
inertia, their mobility is much smaller than the mobility of electrons, and they move
with the same speed as the other neutral particles that are weight comparable in
plasma. This fact makes it possible to build a two-speed plasma model with the
different mechanisms of acceleration and deceleration of the particle fluxes.
So consider the unit volume of plasma where the electrons move in the field
under the influence of the sum of forces: the electric and magnetic forces
 
ne e ~ E þ~ve  ~
B , and the electric field only makes a work on the electrons; static
pressure gradient rpe , here pe ¼ ne kTe —the partial pressure; the frictional force
caused by the scattering of electrons on heavy components qe~ ve mm , here mm —the
average effective collision frequency of electrons with heavy particles and the
convective forces ðqe~ ve rÞ~
ve .
The equation of motion of the electron gas in a two-speed plasma model has the
form:

@ ðqe~
ve Þ  
þ ðqe~ ve ¼ ne e ~
ve rÞ~ ve  ~
E þ~ B  rpe  qe~
ve m m ð1:43Þ
@t

Comparable in mass heavy particles move in the flow at the same speed ~ v. The flow
of heavy components is determined by the action of them the sum of forces: the
 
electric and the magnetic forces ni e ~ v ~
E þ~ B ; acceleration forces qe~
ve mm caused
by the scattering of electrons on the heavy components; static pressure gradient
rpai , here pai ¼ ðna þ ni ÞkT—the sum of the partial pressures; viscous friction
force ~Fl ¼ r ~ ~s, consisting of the tangent and the associated bulk and shear
26 1 The Theoretical Basis of the Low-Temperature Plasma

deformations (1.22) additional friction and force convection ðqai~

vrÞ~
v. In this
case, the equation of motion of heavy components can be written as

@ ðqai~
vÞ  
þ ðqai~ v ¼ ni e ~
vrÞ~ v~
E þ~ B þ qe~
ve mm  rpai þ r ~
s ð1:44Þ
@t

Here qai ¼ ma ðna þ ni Þ—the density of atom–ion gas.

In the model it is necessary to include the continuity equation and the energy
balance equation for the electron gas and for heavy components. Thus, the
two-speed plasma model allows not only to investigate the mechanisms of accel-
eration and deceleration of its streams, to find the effective conditions for accel-
eration or deceleration of plasma flows, their gas dynamics, but also to determine
the energy transfer in two-temperature (two-fluid) model of a nonequilibrium
plasma (Sect. 1.3). In practice, the two-speed movement of the plasma flow is very
common in the case where the electrons drift in the field coincides with the total gas
stream direction. So in an electric arc with a longitudinal electric field the velocity
of electron gas may be significantly different with the total gas flow rate. In the
approximate model the velocity of electrons is taken equal to the sum of the total
gas flow velocity, the electron drift velocity in the field, the bipolar diffusion and
thermal diffusion. In the two-speed model drift in the field is determined by the
acceleration of electrons in the field and by their scattering on the heavy particles,
and the velocity of bipolar and thermal diffusion is determined by gradient of static
pressure. The difference is only the fact that in the two-speed models these
mechanisms can be considered simultaneously.
Earlier we have considered the model of laminar plasma flow. The presence of
electric and magnetic fields in plasma contributes to regulate movement of the
plasma flow. However, the mechanisms of diffusion and thermal diffusion, the
presence of the velocity and temperature gradients in the gas mixture, in many
cases, can lead to disruption of the plasma stream moving order and occurrence of
turbulence, particularly in places where the field is weak effect, i.e., in the plasma
jet and in the wall zone of plasma torch. Consider the basic models of turbulence of
the computational fluid dynamics based on the terms of a continuous medium taken
to the plasma flow.

1.5 Gas Dynamic Models of Plasma Turbulence

One of the important applications of computational fluid mechanics (Computational

Fluid Dynamics—CFD) modeling of plasma flows is the modeling of turbulent
flow in streams and in the near-wall zone of the plasma torch. Turbulent flows in the
channels and in the jets of plasma torches are characterized by different fluctuations
of the velocity field. Orderly and calm steam, called as laminar, occurs only at a low
flow rate v, small size of the region L, and the high viscosity of the liquid l.
Quantitative measure of the fulfillment of these conditions is the numerical value of
1.5 Gas Dynamic Models of Plasma Turbulence 27

the Reynolds rate Re ¼ vL l . With increasing Reynolds rate the number the transition
from laminar to turbulent flow occurs. It was at first demonstrated in experiments of
O. Reynolds, performed in 1883 at Manchester University. The fundamental sig-
nificance of the results is that the order of Reynolds rates, where a loss of stability
and transition to turbulence take place, is the same in the areas of different sizes in
fluid of different physical nature.
Currently, the description of turbulent flows methods, based on the solution of
the Navier–Stokes equations, are mainly used.
Direct numerical simulation (DNS) [17] assumes the full solution of unsteady
Navier–Stokes equations and the continuity equation. The difficulties of DNS
model are in the limited computing resources even now. In solving the Navier–
Stokes equations, approach based on the use of time-averaged quantities is often
used. As a result of the decision of the modified equations requires less computer
resources, but there are additional variables. The different turbulence models are
used to close the obtained equations. Today, there are a large number of models.
But there is not universal model for all classes of engineering problems. Selection
of the optimal model of turbulence depends on the type of course, a specific class of
the problem, the required accuracy of the solution, available computing resources,
etc. From a practical standpoint, the statistics obtained from the DNS, could be used
to test and calibrate the models, based on the Reynolds-averaged equations.

1.5.1 The Reynolds-Averaged Navier–Stokes Equations

The turbulence models used in engineering calculations are usually based on the
solution of the Reynolds-averaged Navier–Stokes equations (RANS). All the
characteristics of the turbulent flow (velocity, pressure, temperature, concentration
of mixture components) have the random fluctuations around the average values. In
this regard, the value of the velocity components (uj ; j ¼ x; y; z) and scalar quan-
tities (U) represent the sum of the mean and fluctuating components (the bar atop
denotes the averaging over the time, and the bar corresponds to the fluctuating
component):

uj ¼ uj þ u0j ð1:45Þ

Uj ¼ Uj þ U0j ð1:46Þ

The intensity of the velocity fluctuations is characterized by the turbulent kinetic

energy

1
k ¼ u0j u0j ð1:47Þ
2
28 1 The Theoretical Basis of the Low-Temperature Plasma

The moving of a continuous medium (liquid or gas) is described by the Navier–

Stokes equations, which consists of the continuity equation

@q @ðquj Þ
þ ¼0 ð1:48Þ
@t @xj

and the momentum equation

@ðqui Þ @ðqui uj Þ @p @sij

þ ¼ þ þ Fi ð1:49Þ
@t @xj @xj @xj

The momentum equations in the form (1.49) represent the three transport
equations for the velocity components, where p ¼ p þ p0 —the pressure, the mean
pressure and the pulsation pressure, respectively; sij —the components of tensor of
the shear stress; Fi—volumetric forces acting on the fluid.
For Newtonian liquids according to the Stokes hypothesis there is a linear
relationship between the components and the rate of deformation tensor

@ui @uj
Sij ¼ þ =2 ¼ 
Sij þ S0ij ;
@xj @xi

Here
 
Sij ¼ @ui @uj @u0i @u0j
þ =2; S0ij ¼ þ =2;
@xj @xi @xj @xi

i.e.,

1
sij ¼ 2l Sij  Skk dij ð1:50Þ
3

After the substituting (1.50) into (1.49) we obtain the momentum equations in
the form of:
 
@qui @quj ui @ 2 @uk @ @ui @uj
þ ¼ pþ l þ l þ þ Fi ð1:51Þ
@t @xj @xj 3 @xk @xj @xj @xi

It is easy to show that the average value of the result of multiplication uj ui is

defined as
 
uj ui ¼ uj þ u0j ðui þ u0i Þ ¼ uj ui þ u0j u0i ð1:52Þ
1.5 Gas Dynamic Models of Plasma Turbulence 29

Substituting (1.45) into (1.51) and averaging, and taking into account (1.52),
gives the transport equation for the average velocity components:
  
@q uj 
ui @q ui @ 2 @uk @ @ui @uj
þ ¼ pþ l þ l þ  qu0j u0i þ Fi
@t @xj @xj 3 @xk @xj @xj @xi
ð1:53Þ

In (1.53) qu0j u0i correspond to the value of shear stress sij in Eq. (1.49), the
correlation u0j u0i is called the Reynolds (or turbulent) stresses. In the physical
interpretation of the Reynolds stress the analogy with viscous stresses, that char-
acterize the transport of momentum due to molecular collisions, is used. The
Reynolds stresses also characterize the transport of momentum due to mixing in a
pulsating turbulent flow. They form a second-order tensor (3 × 3 matrix), the
components of it are symmetric with respect to the diagonal: u0j u0i ¼ u0i u0j .
Similarly the averaged continuity equation can be written as:
 
@q @ quj
þ ¼0 ð1:54Þ
@t @xj

So, with the help of decomposition of Reynolds and averaging procedure we obtain
a system of Eqs. (1.53) and (1.54) for the mean velocity field. These equations are
called the Navier–Stokes equations, Reynolds-averaged (Reynolds averaged
Navier–Stokes equations or equation RANS). They can not be solved, as yet
unknown turbulent stresses. In other words, the system-averaged Reynolds equa-
tions are not closed, and to address the circuit u0j u0i must be determined.
In solving problems of mathematical modeling of plasma processes not only the
definition of the vector velocity field, but scalar fields, such as temperature or
concentration requires. General transport equation for the generalized scalar value
is:
  
@qU @ quj U @ @U
þ ¼ qD þ Fi ð1:55Þ
@t @xj @xj @xj

Here D—the diffusion coefficient.

Averaging this equation is similar to the averaged equations for each component
of the velocity. As in the equations of motion, the unknown correlation between the
velocity fluctuations and a scalar quantity U arise during the averaging of nonlinear
expression for the convective flow:
  
uj U ¼ uj þ u0j U þ U0 ¼ uj U þ u0j U0 ð1:56Þ
30 1 The Theoretical Basis of the Low-Temperature Plasma

The averaging transport equation takes the form:

  
@qU @ quj U @ @U 0 0
þ ¼ qD  quj U þ Fi ð1:57Þ
@t @xj @xj @xj

Here u0j U0 is the turbulent flow of sense scalar quantities U that is a result of the
turbulent mixing.
As in the case of the equations of motion for u0j U0 the transfer equation that
contain unknown higher order correlations can be obtained. Thus, in using the
Reynolds averaging there is a problem of closing the system of averaged
Eqs. (1.53) and (1.57). To solve this problem the physically based mathematical
model requires that will allow expression of the desired correlation by the mean
values and the correlation of a lower order. The goal is achieved by means of a
particular model of turbulence.
Experiments have shown that the turbulence is most intense in areas of high
gradients of average speed. In 1877, J. Boussinesque suggested thatthe relationship

between the turbulent stresses u0 u0 and the strain rate tensor Sij ¼ @ui þ j =2 of
@u
j i @xj @xi
the averaged flow is similar to the corresponding communication that takes place
between the viscous stress sij and components Sij (1.50) for a Newtonian fluid:
 
0 0 2  1
q uj ui  kdij ¼ 2lt Sij  Skk dij ð1:58Þ
3 3

The coefficient of proportionality lt is called the turbulent (or eddy) viscosity.

Equation (1.58) can be written as:
 
@ui @uj 2 @uk
qu0j u0i ¼ lt þ  qk þ lt dij ð1:59Þ
@xj @xi 3 @xk

In view of (1.59) the Eq. (1.53) for the mean velocity takes the form:
   
ui Þ @ quj ui
@ ðq @ 2 @uk
þ ¼ p þ qk þ ðl þ lt Þ
@t @xj @xj 3 @xk
  ð1:60Þ
@ @ui @uj
þ ðl þ l t Þ þ þ Fi
@xj @xj @xi

By analogy with the concept of turbulent viscosity, the gradient diffusion allows us
to express the turbulent flow u0j U0 of scalar quantity U with gradients of mean values:

@U
u0j U0 ¼ Dt ð1:61Þ
@xj
1.5 Gas Dynamic Models of Plasma Turbulence 31

Here Dt —the coefficient of turbulent diffusion. In view of (1.61) the averaged

Eq. (1.57) takes the form:
    
@ qU @ quj U @ @U
þ ¼ qðD þ Dt Þ þ Fi ð1:62Þ
@t @xj @xj @xj

Thus, by using the concepts of turbulent viscosity and gradient diffusion for
solving the problem of the closure of averaged equations it is necessary to determine
the turbulent viscosity and turbulent diffusion coefficient Dt. Note that the concept of
turbulent viscosity lt and diffusion gradient Dt are extremely useful in the con-
struction of approximate models of turbulence and is still widely used in numerical
simulations. However, we must remember that the relevant concepts have no strictly
physical basis and in some cases contradict the experimental data. As a consequence,
the application of the concepts of turbulent viscosity and gradient diffusion can lead
to the large errors in cases of parietal areas, flow separation, and swirling currents.

1.5.2 The Main Gas Dynamic Model of Turbulence

Using the concept of averaged Navier–Stokes equations for turbulent kinetic energy
k ¼ 12 u0j u0j and its dissipation rate e ¼ 2mS0ij S0ji , here m ¼ lq—the kinematic viscosity
and thermal dissipation rate of turbulent kinetic energy x ¼ ke , we can get different
turbulence models based on transport equations of turbulent flow. The object of the
application of the model is to determine the turbulence Reynolds stresses. If the
concept of turbulent viscosity is adopted, this problem is reduced to the finding of
the turbulent viscosity through other flow characteristics (field of average speed for
the calculation of it, for example, the turbulent kinetic energy and its dissipation
speed are involved).
In general terms, the transport equations of turbulent flow, are used in the models
with one or two differential equations, they can be written as follows:
   
@ ðqUÞ @ quj U @ @U
þ ¼ ðl þ C U Þ þP  DþA ð1:63Þ
@t @xj @xj @xj

Table 1.1 The parameters of the generalized transport Eq. (1.63)

Determines the amount U P D CU
lt
Kinetic energy k sij @
ui
@xj
qe rk
or b qkx
The dissipation rate of kinetic energy ε ce1 ke sij @
ui 2
ce2 q ek
lt
@xj re

The specific rate of dissipation ω a xk sij @

ui
@xj
bqx2 rx lt
32 1 The Theoretical Basis of the Low-Temperature Plasma

Table 1.2 The coefficients for the models of turbulent flow with two differential equations
Models Coefficients (K = 0.41)
ke ce1 ¼ 1:44 ce2 ¼ 1:92 rk ¼ 1:0 re ¼ 1:3 cl ¼ 0:09
[19]
kx b ¼ 0:09 b ¼ 0:075 a
0:556 rk ¼ 0:5 rx ¼ 0:5
[20]
pffiffiffiffiffi
kx b 1 ¼ 0:09 b1 ¼ 0:075 a1 ¼ b1 =b 1  rx1 K 2 = b 1 rk1 ¼ 0:85 rx1 ¼ 0:5
(SST) pffiffiffiffiffi
b 2 ¼ 0:09 b2 ¼ 0:0828 a2 ¼ b2 =b 2  rx2 K 2 = b 2 rk2 ¼ 1 rx2 ¼ 0:856
[21]

Designations of the parameters in (1.63) for each type of equations are shown in
Table 1.1 [18].
In the case of the model of Shear Stress Transport (SST) for x, the last term of
(1.63)

1 @k @x
A ¼ 2ð1  F1 Þqrx2 ð1:64Þ
x @xj @xj

In other cases A = 0. The mass function F1 defined as in [18].

Constants for the models with two differential equations are summarized in
Table 1.2.
We give a brief overview of the main gas dynamic models.
Laminar model is used to calculate the laminar flows. In this case, a direct
solution of the Navier–Stokes equations without any additional any turbulent model
occurs. We can assume that the laminar model is required to describe the laminar
flow where the Reynolds stresses disappear. The condition of this model is the low
value of the Reynolds rate, it would otherwise need to apply turbulence model below.
Zero-order model (the algebraic model) where the turbulent viscosity is deter-
mined from a simple algebraic equation. The model of Baldwin–Lomax (BL) is a
typical representative of this one. The algebraic models of turbulence differential
equations do not contain transport and do not employ the nonlocal characteristics
such as boundary layer thickness, wall friction, etc. They are very nonuniversal,
difficult to implement in modern codes, and are not used at the moment.
The model of the first order, where the eddy viscosity is determined by solving
a differential equation (models with a differential equation). Differential transport
equation can be written with respect to different variables: the turbulent viscosity;
turbulent kinetic energy or the like. As the linear scale the distance from the wall is
commonly used. The most successful of these models is the Spalart–Allmaras
(SA) model. This one contains a single differential equation for “high Reynolds”
turbulent viscosity associated with the turbulent viscosity with the algebraic rela-
tion. This is a one-parameter model was developed for aerospace applications. It
contains a number of amendments expanding its scope: the amendment on the
curvature and rotation; correction for roughness and others. This model gives good
results for the boundary layers, characterized by a positive pressure gradient.
1.5 Gas Dynamic Models of Plasma Turbulence 33

The model of the second order, where the eddy viscosity is determined from
two differential equations. These models are models that have become the industry
standard and widely used in the practice of computing, for example, the model
Shear Stress Transport (SST). Consider the basic model with two differential
equations (k  e and k  x).
Models k  e can be divided into high- and low Reynolds and by the method of
describing the effect of the near-wall effects. In high Reynolds models (“standard”
model, Reynolds Normalization Group—RNG) use the parietal function, and in low
Reynolds and (Launder–Sharma)—special damping function. The model k  e uses
two differential transport equations for calculating the turbulent kinetic energy k and
the dissipation rate ε. This model is stable, does not require significant computing
resources and has long been the industry standard, and has worked well in the
calculation of internal flows, but when it is used, there are problems in the calcu-
lation of flows with large pressure gradients and in the case of flow separation.
Model k  e is often used to calculate the free flow. Start of separation is defined
too late and the sizes of separated flows are too small. The model k  e gives too
optimistic results for the flow that is detached. Due to incorrect predictions of the
separation point model k  e is gradually replaced by more sophisticated models.
Models k  x. The experience of calculations of separated turbulent flows over
bluff bodies showed that the models based on the equation for the dissipation rate ε
predict the onset of flow separation greatly downstream than that is observed in the
experiments, with settlement size of the recirculation zone and traffic intensity in it
are too low. At the same time, it appeared that another class of models, where instead
of the transport equation for ε the equation for the specific dissipation rate of turbulent
kinetic energy ω is used, resulting in better agreement with the experimental results
with this flows, if the separation zone is small. The model k  x was developed in the
works of Wilcox [22] and was the basis of a combined model of shear stress SST,
created later by Menter [23] and combining the models k  x and k  e.
In the standard model k  x Low Reynolds effects are accounted, the influence
of the compressibility and the shear perturbation, but it is significantly inferior to
the breadth of application of the models k  e. Models k  x can be used to
calculate the wall turbulence without the introduction of special functions and have
extremely high sensitivity to the boundary conditions in the external flow.
Menter’s model of Shear Stress Transport (SST) includes the models k  x of
Wilcox for the wall region and the converted model k  e for the region far from the
wall [23]. Thus, in calculating the total flux the SST model gives good results,
especially in cases of separation of the flow and large pressure gradients therein. In
addition, this model proved to be reliable and not demanding the computing power.
Therefore, it combines the advantages of models k  e and k  x, but in terms of
computation this one yields to the models with only single equation. SST model is a
new industrial standard. It gives very good results even in the calculations of
phenomena such as flow separation by blowing air into the boundary layer, or in the
process of turbulent heat transfer.
In the Reynolds Stress Models—(RSM) the turbulent viscosity is not used
(1.59). Instead, this class of models assumes the decision of a single transport
34 1 The Theoretical Basis of the Low-Temperature Plasma

equation for each component of the Reynolds stress tensor u0j u0i and the transport
equations for the dissipation rate e. As a result the possibility of taking into account
the anisotropy of turbulent fluctuations opens, it significantly expands the potential
range of application of these models, including for flows with large curvature of the
streamlines, swirled flow and large deformation of the velocity field.
Reynolds stress model is potentially superior the model of turbulent viscosity in
calculation abilities of different currents. They are much more complicated than
models of turbulent viscosity and have a number of problems as computational and
physical (the principle of feasibility) characters. Taking into account the symmetry
of the Reynolds stress tensor in RSM with differential equations six transport
equations for different components u0j u0i and one equation for isotropic dissipation
rate in three-dimensional space must be considered. Algebraic Stress Models
(ASM) include two transport equations for the scales and six transcendental
equations for the stresses and explicit algebraic Reynolds stress model (EARSM)—
2 transport equations for the scales and explicit expressions for the stresses.
The main problem of the usage of RSM models is the reliability of approximate
simulation of many unclosed correlations arising in the derivation of transport
equations for u0j u0i . Additional difficulties arise in the formulation of the boundary
conditions near the solid surfaces. Despite the potential benefits, RSM model
Reynolds stresses are used less often than the models of turbulent viscosity.
Large Eddy Simulation (LES) is becoming an increasingly popular method of
calculation of the turbulent flows not only in research, but also in engineering
practice. LES computing strategy differs greatly from the use of the Reynolds
averaging (RANS). Instead of the solutions of averaged equations, that are deter-
ministic, the nonstationary solutions, representing a random process are paid here.
The aim of LES method is to reproduce numerically the largest vortex structures of
the long wavelength part of the inertial range of the energy spectrum. Therefore, the
numerical solution arises as a result of the fluctuations in a turbulent flow. In the
contrast to the direct numerical simulation the shortwave (High Frequency) part of
the spectrum does not play, but its impact is taken into account by using the
approximation models of the SubGrid-Scale (SGS) viscosity. In accordance with
the hypotheses A. Kolmogorov the small-scale turbulent fluctuations are spatially
isotropic and statistically universal. In other words, the dynamics of small
(sub-grid) vortices does not depend on the geometric characteristics of the flow and
characteristics of the formation of large vortex structures. This fact simplifies the
problem of creating a universal model sub-grid. In addition, the increasing of the
resolution of the calculation algorithm (primarily at the expense of reducing the size
of the grid cells, as well as in the case of increasing the accuracy of approximation
of differential equations by algebraic ones) leads to the decrease of the sub-grid
turbulence and expand the range of oscillations, explicitly permitted.
1.5 Gas Dynamic Models of Plasma Turbulence 35

Detached Eddy Simulation (DES) [24] The characteristic dimension of tur-

bulent eddies in the boundary layer near the solid surface is much smaller than in
the developed turbulent flow far away from it. In this case the explicit permission of
even the largest eddies in the boundary layer requires grinding the calculated mesh
that may lead to an unacceptable increase in the duration of computation. At the
same time, for near the wall surface flows and flows with a small separation quite
reliable and cost-effective model based on Reynolds averaging exist. For these
currents the RANS models can adequately assess the situation of flow separation,
resistance and the lift force without excessive grid refinement near the surface.
However for a wide range of practically important problems, accompanied by a
large area of flow separation, the existing models RANS lead to large errors in the
calculation of these quantities. Moreover, the principle possibility of a foundation
of universal model based on the Reynolds averaging is questioned, because this
model does not account for the large-scale turbulent flow disturbances, caused by
the boundary conditions, the shape and the dimensions of the computational
domain. At the same time the LES method has the same aim. In this case of creating
the idea of a hybrid approach, where the RANS model is activated for the wall
region, appeared and the LES model is activated for the region far from the wall.
The most successful and widespread hybrid approach was the method of Detached
Eddy Simulation (DES), proposed in 1997 by P. Spalart and later developed in the
works of M. Strelets, M. Shur, A. Travin and others. The ability of the selection of the
specified turbulence method generates a whole family of the DES model and some of
them are available in the software package Ansys Fluent and Ansys CFX.
DES method gives good results for the flows with sufficiently thin boundary
layers, where near-wall grid cells are strongly elongated along the solid surface
(where the model is activated RANS), and extensive zones of separation, for that
the grid is close to isotropic (in this region model LES is activated). Thus, the
method of DES is a hybrid approach where the switching models RANS and LES is
used for the different regions of space. Unified structure of averaged equations and
switching models and others open other than DES, approaches to the construction
of hybrid models. In particular, it is possible to use average values of Reynolds and
residual stresses (or turbulent and sub-grid).
These turbulence models give different results in the simulation of separated
flows, boundary layers and so on. The accuracy of different models also depends on
the width of the computational grid in the wall layer of cells (that are usually
characterized by y+, it can be regarded as a local Reynolds number in the cell). At the
same time a large number of engineering hydrodynamic problems can be solved
using commercial software packages (Ansys Fluent, Ansys CFD, FlowVision et al.).
In these packages the computational grid is created directly in the package, or
imported from another editor. In many cases, the kind of computational grid deter-
mines the further calculation algorithm. Currently, two types of grid: tetrahedral and
Cartesian are used in software packages. The tetrahedral grid allows you to create
cells, similar in form to the boundaries of the computational domain and regions with
36 1 The Theoretical Basis of the Low-Temperature Plasma

large gradients of the unknown quantities and this fact enables a good approximation
of the boundary layers. At the same time, the construction of a tetrahedral grid is very
difficult. Cartesian grid allows only create rectangular cells, that may lead to worse
approximation of the boundary layer, but its construction is simple.
There are also a number of ways to approximate the area of the boundary layers
with high gradients. Examples of such methods include ALRG technology (adaptive
local refinement grid) and the technology of sub-grid approximation of geometric
shape, implemented, for example, in a software package FlowVision. ALRG
technology allows splitting the selected cells in all directions (adapted to a prede-
termined level). In this cell, close to the fragmented crushed that the size of two
adjacent cells does not differ by more than 2 times. Grinding cells can be set both in
the bulk and on the surface. At the same time, the technology can automatically
sub-grid approximation more accurately reproduce the shape of the surface.
All of these models and techniques, of course, can be taken in order to simulate the
plasma stream flows, especially in the edge zone of the plasma torches and plasma jets.

1.6 Models of Light Emission and Absorption in Plasma

Dense low-temperature plasma is a powerful source of radiation. Emission and

absorption in plasma are the result of various physical processes occurring inside it
in the interaction of particles: electrons, ions, atoms and molecules. Power emission
and absorption directly affect the energy balance and temperature of plasma, so the
calculation of emission and absorption of light by plasma is the major problem in
plasma physics and gas discharge. Light quanta appear and disappear during the
radiation transition of electrons from one energy state to another. The interaction of
the electrons with the ions play an important role in the formation of the light flux
of radiation, and electron–electron collisions are not contribute to the flow of the
emission and absorption of light.
Consider a system that consists of electron and a positive ion. Figure 1.2 shows
a diagram of the energy states of electron. The energy is measured from the
boundary of the continuous spectrum, separating free and bound states. The lower
level corresponds to the ground state of atom. According to the circuit shown in
Fig. 1.2, the following transitions of electrons can be implemented: (1) free–free
(ff), (2) free–bound (fb) and bound–free (bf), (3) the bound–bound (bb). In the
process of transition from top to bottom the quanta hx the quanta emitted that is
equal to the difference between the energies of the upper and lower levels; during
the transition from the bottom to up level the quanta hx the quanta absorbed.
Free–free transitions are bremsstrahlung during the Coulomb collisions of
electrons with ions and inverse bremsstrahlung absorption processes. Free–bound
transitions are the processes of photo– and bound–free transitions (the photo–
ionization process). Free–free, free–bound and bound–free transitions provide the
continuous spectra (continuum) because either the initial or final, or both electron
states lie in the continuous energy spectrum. Bound–bound electron transitions
1.6 Models of Light Emission and Absorption in Plasma 37

photorecombination(fb), bremsstrahlung(ff)
E photoionization (bf ) emission/absorbtion

•e meve2
continuum Ek =
2
E∞
EI
En+1
En

E3
E2

bound-bound transitions (bb)

E1
Fig. 1.2 Diagram of energy levels and transitions of electron in the field of ion

between discrete energy levels in atoms lead to the emission and absorption of
photons in the spectral lines. The band spectra of molecules are the variation of the
line spectra that are also formed during the bound–bound transitions. For each state
of electrons in the molecule corresponds a set of oscillation and rotational levels,
and there are many possible transitions between them. The same transitions occur in
the interaction of the electron–atom in plasma.

1.6.1 Radiation in the Continuous Spectrum

1.6.1.1 Bremsstrahlung (ff) Radiation

In the motion of a free electron in the continuum area a quantum hx that is equal to
the difference between the kinetic energies of the initial ve and final v0e speeds:

me v2e me v02
hx ¼  e
ð1:65Þ
2 2

To describe the ff-spectrum of the differential cross section of emissions is intro-

duced, that depends on the speed of the electron ve :
38 1 The Theoretical Basis of the Low-Temperature Plasma

drffx
drffx ðve Þ ¼ xdx ð1:66Þ
dX

Here x ¼ 2pm—the cyclic frequency, radian/s.

The power emitted by one electron in the frequency domain x, x þ dx in the
total solid angle 4p in the medium with a concentration N of atoms or ions
(electron–electron interaction does not lead to bremsstrahlung) is determined
depending on this section:

dWxff ¼ hxNve drffx ðve Þ ð1:67Þ

By Kramers, the differential cross section for interaction of electron and ion is
equal to:
16p Z 2 e6
drffx ðxÞ ¼ pffiffiffi 2 2 3 dx ð1:68Þ
3 3 m e ve c 
hx

Here Z—the charge of the ion.

Bremsstrahlung intensity in plasma in the range dx of 1 cm3 is obtained by
integration (1.67) on the velocities of electrons with the distribution function f ðve Þ,
normalized to the unity. The maximum speed when the electron can emit a photon
qffiffiffiffiffiffi
hx is v e ¼ 2mhxe . Then, the spectral intensity is defined as:

Z1
Ixff ¼ hxNne f ðve Þdrðve Þve dve ð1:69Þ
v e

The results of Maxwell function f ðve Þ averaging and N ¼ Ni is:

 
Z 2 N i ne hx
Ixff ¼ C pffiffiffiffiffi gff exp  dx ð1:70Þ
Te kTe

16 2p 1=2 e6
C¼ 3=2
¼ 1:08  1045 J cm3 K1=2 ð1:71Þ
3 3 me c3 k1=2

The (1.70) the factor gff called Gaunt factor is introduced. It refines the result of
the semiclassical consideration and is calculated by quantum mechanics. If you
design x ¼ mv2e 2h ¼ Ek =h the boundary frequency, which can radiate electron
with the kinetic energy Ek , gff will depend on the dimensionless parameter
x ¼ x=x . Gaunt factor values depending on the parameters x are listed below [25]:

x 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

gff 1 2.01 1.61 1.34 1.13 0.97 0.81 0.68 0.53 0.36 0
1.6 Models of Light Emission and Absorption in Plasma 39

For small in comparison with kT quanta, in particular in relation to the infrared

region [26], we can write
pffiffiffi 
3 4kTe 1
g ¼
ff
ln ¼ 1:27 3:38 þ lg Te  lg ne ð1:72Þ
p e2 n1=3
e
3

Here Te, K; ne, cm−3. In the dense low-temperature plasma: gff ¼ 1:5  2:5.
The emission spectrum is determined by the exponential factor. According to
(1.70) the spectral intensity of the light isotropic emitted (by averaging over a large
number of electron–ion pairs) in the angle 4p that corresponds to the total radiation
power per unit volume of plasma.
These results can be used to calculate the intensity of transitions due to collisions
with neutral atoms, if in the calculation of the cross sections to replace the Coulomb
scattering cross section rC on the transport rtr one. The Coulomb cross section and
its dependence on the rate determined by the Coulomb radius rC , when the kinetic
energy of the electron is equal to the potential energy of interaction:

me v2e Ze2 4pZ 2 e4

¼ ; rC ¼ prC2 ¼  2 ð1:73Þ
2 rC me v2e

The ratio of the cross sections for bremsstrahlung scattering on ions and neutral
atoms will be

drffx;i pa20 Z 2 2IH 2
¼ ð1:74Þ
drffx;n rtr Ek

m v2 2 4
Here Ek ¼ e2 e —energy of electron, IH ¼ 2p he2 me ¼ 13:6 eV—the ionization

potential of a hydrogen atom, a0 ¼ h2 ð4p2 me e2 Þ—the first Bohr radius,
rtr ¼ rC ð1  hcos hiÞ, hcos hi—the
  averaged cosine of the scattering angle.
In a typical relation rtr pa20
1  10: electrons with energies of a few
electron volts per heavy particle emitting in 10–100 times more energy on ions than
on neutral particles, i.e., the intensity of bremsstrahlung (and absorption) for
electron scattering on ions and neutral particles is comparable with the degree of
ionization d
ð1  10Þ %. In a weakly ionized plasma d  0:1 %, in addition it is
typical for it the lack of multiply charged ions. As a result electron bremsstrahlung
is predominant in collisions with neutral particles.

1.6.1.2 Recombination (fb) Radiation

In the radiative electron capture the rate ve with the transition to a bound state of
negative energy En the quanta is emitted:
40 1 The Theoretical Basis of the Low-Temperature Plasma

me v2e
hx ¼ jEn j þ ð1:75Þ
2

The cross section for this process rffn can be found using the following method.
The validity of (1.68), if we consider it for the differential
 bremsstrahlung cross
section is limited to a maximum photon hxmax ¼ me v2e 2 but electron stay still free.
The formula (1.68) can be applied for large emission of photons. Electron in such
cases will still have negative energy, i.e., remains bound. If the excess hx over

hxmax is small, an electron is transferred to the upper  levels of the excited atom.
But at a fixed initial energy of electron e ¼ me ve 2 quanta with discrete energy
2

values are emitted (1.75), but drfb x —this section of the photons emission in the
continuous spectral range dx. To fix, the fact that the upper levels of the atom are very
close to each other is used, and the total cross section for the emission of photons in a
small, but finite interval Dx around x will be important. Electron falls on one level at
a distance of Dn belonging to the energy interval DE ¼ hDx around En ¼ e  hx.
If DE  jEn j, then the locks on all these levels are almost equally. Consequently,

drx drx DE
Dx ¼ rfb
n Dn; n ¼
rfb ð1:76Þ
dx dx hDn

We will consider the hydrogen-like system. Electron in such

 a system is moving
in the field of a point charge Z, and for it En ¼ IH Z 2 n2 , here n—the main
 
quantum number. But DE=Dn
jdEn =dnj ¼ 2IH Z 2 n3 , and IH ¼ me4 2h2 and
using (1.68), we obtain the equation for the capture cross section in cm2:
2
16p e10 Z 4 1 2:1  1022 ðIH Z 2 Þ
rfb
n ¼ p ffiffi
ffi ¼ ð1:77Þ
3 3 c3 h4 me v2e x n3 ehxn3

Equation (1.68) for calculation of rfb n have to be applied in the conditions asso-
ciated, in particular, with the quasi-classical motion of the electron in orbits with large
quantum numbers n, and is consistent with the principle of spectroscopic stability.
The spectral intensity of radiation by electrons, moving in a range of speeds ve ,
ve þ dve , in their capture to a level with n per unit of plasma volume will be
determined:
fb
Ixn ¼ hxNi ne rfb
n f ðve Þve dve ð1:78Þ

Here Ni—the ion concentration.

Substituting rfbn (1.77) in Eq. (1.78) and replacing ve in accordance with (1.75), to
x, i.e. me ve dve ¼ 
hdx, taking into account the Maxwell function f ðve Þ we can get:
1.6 Models of Light Emission and Absorption in Plasma 41

 
Z 2 Ni ne 2xn hx
fb
Ixn ¼ C pffiffiffiffiffi exp  expðxn Þdx ð1:79Þ
Te n kTe
  
Here xn ¼ ðme e4 Z 2 Þ 2h2 n2 kTe ¼ ðIH Z 2 Þ ðkTe n2 Þ ¼ En =ðkTe Þ, C—the con-
stant (1.71).
Quanta of the certain energy hx can be emitted in the capture of electron on the
different levels, and the law of conservation of energy is respecting (1.75).
Thus, for the same reasons (the decrease of electron density is because of reduced
Maxwell function), and for the bremsstrahlung recombination radiation intensity
decreases exponentially with increasing frequency. However, this applies to the
photoelectron capture with its transition to a bound state with a fixed n. Radiation
with the same frequency x can also occur in the capture of electron to other levels
with different En in the conditions of (1.75). Therefore, the intensity of the spectrum
is determined with the addition of (1.79) on the possible quantum numbers n:

X  1 
1
Z 2 Ni ne hx X 2xn
Ixfb ¼ fb
Ixn ¼ C pffiffiffiffiffi exp  expðxn Þdx ð1:80Þ
n Te kTe n n

The lower limit of the addition n is determined by the condition jEn j  hx. At
the different frequencies of recombination spectrum a different number of levels of
grip gives contribute to the intensity of spectrum. With the increasing frequency the
number of such levels increases, and apparently it is manifested in the “sawtooth”
shape of the spectrum (Fig. 1.3). Typically, at not so high temperatures of electrons
the neighboring quantum numbers n are summed up.
The total intensity of radiation, J/(s cm3) is the sum of the brake (1.70) and
recombination (1.80) components and it is determined by the following approxi-
mate equation:

Z 2 Ni ne hx
Ix ¼ C 1=2
W dx; ð1:81Þ
Te kTe

Fig. 1.3 Continuum of

helium plasma with
ne ¼ 5  1016 cm3 : points
the experimental data, lines
the calculated results
42 1 The Theoretical Basis of the Low-Temperature Plasma

8 
>
< 1; x ¼ hx=kTe \xg ¼ Eg ðkTe Þ;
WðxÞ ¼ eðxxg Þ ; xg \x\x1 ¼ I=ðkTe Þ; ð1:82Þ
>
: ðxxg Þ
e þ 2x1 eðxx1 Þ ; x [ x1 :

Here Eg —the energy of the lowest excited state, I—the ionization potential. In the

small quanta region (hx\ Eg ) the contribution of free–bound and free–free
transitions in emissivity are characterized by the relation

Ixfb : Ixff
fexp½hx=ðkTe Þ  1g : 1 ð1:83Þ

Consequently, the photons hx\0:7kTe are emitted mainly in the brake mech-
anism, and  hx [ 0:7kTe —at recombination one. At Te ¼ 104 K photons with
hx ¼ 0:7kTe ¼ 0:6 eV (k ¼ 2000 nm) corresponds to the
 infrared region of the
spectrum. These values are significantly less than Eg that for hydrogen, for


example, is I 22 ¼ 3:37 eV (n = 2) (k ¼ 370 nm—in the near ultraviolet region).
According to (1.83), nearly all the spectrum emission, except for infrared region,
occurs by the recombination mechanism. This situation is typical of the
low-temperature plasma. Brake processes play a small role there and come to the
fore only in the emission in a rather far infrared and microwave bands.

1.6.2 Absorption in the Continuum

1.6.2.1 The Bremsstrahlung (ff) Absorption

The above information about bremsstrahlung is valid for the description of

absorption with regard to the principle of detailed balance. Full balance of ff process
consists of spontaneous emission, absorption and stimulated emission. We can
introduce the coefficients of the bremsstrahlung absorption affx and stimulated
emission bffx , corresponding to absorption and by stimulated emission coming under
the influence of external radiation per pair of interacting electron–ion or electron–
atom.
In the electron moving in the field of continuum the absorption of a quantum hx
occurs that is equal to the difference of kinetic energies in the final v0e and initial ve
speeds. The law of energy conservation in bremsstrahlung (ff) absorption is:

me v02 me v2e
e
¼ þ hx ð1:84Þ
2 2

Take into account the Einstein relation

1.6 Models of Light Emission and Absorption in Plasma 43

 
v0e bffx v0e ¼ ve affx
 
 0  p2 c2 v0e drffx v0e ð1:85Þ
bffx ve ¼
x2 dx

The absorption coefficient calculated per electron with speed and one ion can be
found with the Kramers Eq. (1.68):

16p3 Z 2 e6
affx ðve Þ ¼ pffiffiffi 2 ð1:86Þ
3 3 me chx3 ve

Multiplying affx ðve Þ and Nne by averaging this relation on velocity ve with the
Maxwell distribution, we obtain the equation for the coefficient of the brake light
absorption in plasma, cm−1:

Z 2 Nne gff
vffx ¼ C11=2
;
Te x3
 ð1:87Þ
16p2 2p 1=2 e6  3 1=2 
C1 ¼ 3=2
¼ 1:45  10 10
cm 5
s K
3 3 me ck 1=2 h

where N—the concentration of ions or atoms; here is similar to (1.70), we introduce

a correction Gaunt factor gff.
The photoionization cross section at the bf absorption rbf n (an inverse process
for the process of photorecombination) can be obtained from (1.77) taking into
account the conditions of detailed balance (Milne equation) [27]:

gZ hx 2 fb
n ¼
rbf rn ð1:88Þ
gZ1 me ve c

Here gZ , gZ1 —the statistical weights of ions corresponding to the degree of ion-
ization. For a neutral atom: Z ¼ 1, for a hydrogen atom statistical weights of levels
with the main quantum number n will be: gZ1;n ¼ 2n2 , gZ ¼ 1:

8p e10 me Z 4
n ¼ pffiffiffi
rbf ð1:89Þ
3 3 ch6 x3 n5

If the threshold frequency of photoionization of atoms from the level with

 
xn ¼ jEn j=h, jEn j ¼ me e4 Z 2 2h2 n2 , the photoionization cross section, cm2:

18 n xn 3
n ¼ 7:9  10
rbf ð1:90Þ
Z2 x

With increasing frequency, from the threshold value, photoionization cross section
is proportional to x3 . Equations (1.88) and (1.77) are obtained on the basis of the
44 1 The Theoretical Basis of the Low-Temperature Plasma

semiclassical approximation by the generalizing of formula (1.68) for the brems-

strahlung the results should be as closer to the exact quantum solution as n is
higher.
As in the case of bremsstrahlung, in the quasi-classical equation for the pho-
torecombination and photoionization introduce appropriate corrective factors of
Gaunt gfb and gbf . The total absorption coefficient vx is the sum of the coefficients
of the brake vffx and photoionization vbfx absorption: vx ¼ vx þ vx .
ff bf
bf
The coefficient vx is found by the sum of the photoionization cross sections for
P1
x ¼
all levels of atom: vbf bf
n Nn rn , where the weighting factors are the popula-
tions of levels Nn. Substituting equation (1.89), the Saha equation (1.9) and the
Boltzmann equation (1.4), using (1.87) in the last equation, we obtain the
absorption coefficient without stimulated emission in the photoionization:

X
1    
z2 Nne gbfI hx
vbf
x ¼ Nn rbf
n ¼ C1 1=2 exp  exp  1  nbf
n Te x3 kT kT
    ð1:91Þ
2a4 kTk3 gi I hx
¼ 3=2 2 2 na exp  exp  1  nbf
3 p e ga kT kT

Here for many-electron atoms in the calculation of photoionization absorption

coefficients a factor nbf , that takes values from 0.1 to 3 s with increasing wave-
length k from 200 to 1000 nm, is introduced.
Accordingly, the total absorption coefficient vx , cm−1 is written as [26]

Z 2 N i ne Z 2 Ni ne expðxÞWðxÞ
vx ¼ C 1 expðxÞWðxÞ ¼ 4:05  1023 ð1:92Þ
1=2
Te x3 Te
7=2 x3

Here x ¼ hx=kTe ; C1 is determined by (1.87), and Wð xÞ—by (1.82).

The contribution
of photoionization and braking mechanisms to the absorbance
at 
hx\ Eg is described by a ratio vbfx : vx
½expðhx=kTe Þ  1 : 1.
ff

At the not high temperatures,

  when the degree of ionization is small and the
quanta are not large hx\ Eg , we can obtain for χω, cm−1 the equation of
Unsold–Kramers [26]:
 
 exp  Ihx
2 6
16p Z e kTe N gi I  hx 2
Z N gi kTe
vx ¼ pffiffiffi 4 3 exp  ¼ 1:95  107 2
3 3  h cx g a kT e T e g hx=kTe Þ3
a ð

ð1:93Þ

Here gi, ga—the statistical weights of ion and atom and N—the atom concentration.
1.6 Models of Light Emission and Absorption in Plasma 45

1.6.3 Radiation in the Spectral Lines

Radiation in the spectral lines has a significant energy contribution to the radiant
loss of plasma. In bound–bound transitions of electron in atom or the ion between
different energy levels photon in the spectral lines is emitted (or absorbed). Thus, in
spontaneous transition of atom or ion from the upper kth energy level to lower ith
the photon hxki ¼ Ek  Ei is emitted. The intensity (W/cm3 sr) in the spectral line
at such a transition of atom or ion from the upper level k of to the lower level i is
expressed by the relation:

Z1
1 1
Iki ¼ hmki Aki nk ¼ hxki Aki nk ¼ iki ðkÞdk; ð1:94Þ
4p 4p
0

Here mki —the frequency of the line center (mki ¼ m0 ¼ c=k0 ); Aki —the probability of
optical (spontaneous) transition from the kth to ith level (the first Einstein coeffi-
cient); nk—concentration of the emitting particles of the upper kth level; iki ðkÞ—
distribution of the spectral intensity at a wavelength that is often fixed by spec-
trometer. Figure 1.4 shows the distribution of the spectral intensity of argon plasma
(free burning arc) in the wavelength range λ = 400–450 nm.
The relation (1.94) is exact at low optical thickness of the plasma layer with the
depth L:

Fig. 1.4 The emission spectrum of the argon plasma (open free burning arc) in the wavelength range
λ = 400–450 nm at a distance z = 2 mm from the cathode tip to the observation area, (I = 80 A,
a diffraction grating with 600 lines per 1 mm): ArI—the atom of argon, ArII—the ion of argon
46 1 The Theoretical Basis of the Low-Temperature Plasma

ZL
so ¼ k0 dl0 \1
0

Here k0—the absorption coefficient at the center of the line.

The probability of optical transitions Aki ¼ 107  108 s1 and corresponds to the
lifetime of atom (or ion) in the kth excited state with ski ¼ 1=Aki ¼ 107  108 s.
The spectral measurements for the extended objects of finite optical thickness,
however, require the consideration of reabsorption and reemission of light.
Spontaneously re-radiated photon can have a different frequency, polarization, and
the direction of propagation, dissipate in the quenching process. The totality of this
kind of nonlocal processes in the phase space is considered in the framework of the
general problem of radiation transfer.

1.6.3.1 The Spectral Lines Outline

The part of the spectrum corresponding to the transition between the bound states of
particles is understood as the spectral line. Mechanisms of line broadening and
shaping their contours can be associated with the interactions between the particles,
external fields and radiation.
The excited state of atom is not strictly stationary and has a finite lifetime s.
Energy level En according to quantum mechanical uncertainty principle is
“smeared” on Dx  DE=h  s1
A1 ki . In this case, the width of the spectral line
outline, existing independently of external influences, is called natural, it is dkn

105 nm. The interval dk between these two points of the outline k0 and k00 , where
iðk0 Þ ¼ iðk00 Þ ¼ iðk0 Þ=2 is called the half-width of the line outline.
Motion of the emitting particle along the beam z at a rate vz (vz  c; where c—
the speed of light) leads to the Doppler frequency shift by the value m0 vz =c where
m0 —the frequency of the radiation of the particle at rest, when vz ¼ 0. The average
displacement of the observed frequency even at room temperature is much greater
than the natural width dkn . In plasma the particles are in the thermal motion, and if
plasma is optically thin, and disturbances of emitting atoms can be neglected (for
example, rarefied plasma in weak fields), the form of the spectral lines with the
Doppler broadening depends on the function f(v) of the particle distribution on
velocities. If this function is the Maxwell one, the line outline takes a Gaussian
form:
!1=2 !
Mc2 Mcðk  k0 Þ2
imn ðkÞ ¼ exp  ð1:95Þ
2pkTk20 2pkTk20
1.6 Models of Light Emission and Absorption in Plasma 47

Table 1.3 Values of the Atom M λ0 (nm) δλD (nm)

Doppler width at T = 104 K
H 1 656.3 0.047
He 4 587.6 0.021
Ar 40 426.6 0.0049
Hg 200 546.1 0.0028

Here M—the relative atomic (molecular) mass of radiating particles. We have

R1
imn ðkÞdk ¼ 1.
0
The width of such outline is called the Doppler width and it is defined as:
rffiffiffiffiffiffiffiffi
k0 2kT
dkD ¼ 2  ln 2 ð1:96Þ
c M

As an example, the values of the Doppler width of the lines outline at temperature
T = 104 K are represented at Table 1.3.
However, the effect of Doppler is not always the main mechanism of the spectral
lines broadening in plasma with a high density, where the indignations of radiating
atoms due to collisions with other particles arise very often. These collisions lead to
the Lorentz broadening. The shape of lines is the dispersion one (the term ðk  k0 Þ2
is present in denominator):

imn ðk0 Þdk2L

imn ðkÞ ¼ ð1:97Þ
4ðk  k0 Þ2 þ dk2L

When the Lorentz width δλL becomes more than natural, it can be expressed in the
frequency of the “broadening” collisions:
rffiffiffiffiffiffiffiffi
k2 rN 2kT
dkL ¼ 0 ¼ ¼ ð1:98Þ
2pc pc pM

Here σ—the cross section of the “broadening” collisions; N—the concentration of

the perturbing particles.
The cross sections of “broadening” collisions are usually higher than the
gas-kinetic ones. According to (1.98) at atmospheric pressure and T  104 K the
Lorentz width, roughly estimated, is in times less than the Doppler one. From (1.98)
it also follows that in a weakly ionized plasma (ne =N  1) the Lorentz width
pffiffiffiffi
dkL  p= T .
The dispersion (Lorentz) and Doppler (Gaussian) shapes of lines with the same
width are compared on Fig. 1.5.
It is essential that the decrease in intensity in the wing of the dispersion (Lorentz)
line is much slower than in the wing of Doppler (Gauss).
48 1 The Theoretical Basis of the Low-Temperature Plasma

Fig. 1.5 Comparison of

Lorentz (lower curve 1) and
Gaussian (upper curve 2)
outlines

The collisions with electrons, as a rule, play the main role in the mechanism of the
spectral lines broadening. It is defined as large cross sections of these collisions and
high-speed electrons. Cross section depends essentially on the nature and type of
interaction between the emitting atom and electrons. The discovery of the phe-
nomenon of splitting and shift of atomic levels under the influence of external
electric field with the intensity E by R. Stark in 1913 has been called the Stark effect.
The splitting of the levels in high electric fields is proportional to E and it corre-
sponds to the linear Stark effect, in weak fields it is proportional to E2 and leads to the
quadratic Stark effect. Both circuits are described in the first approximation in the
form (1.97). More details of these interactions are presented in [28]. The most typical
values of half-width of the lines outlines are broadened as a result of the quadratic
Stark effect, for the low-temperature plasma are equal to 0.0005–0.5 nm.
In practice, profiles of the spectral lines that are registered with the spectral
equipment present a convolution of the Gaussian and dispersion outlines. The result
of the simultaneous broadening of the Lorentz line shape and the Gaussian one is
the convolution of the last lines:

Zþ 1
IV ðDkÞ ¼ I0V IG ðDk  Dk0 ÞIL ðDkÞdðDk0 Þ ð1:99Þ
1

It is known as the Voigt profile. After the transformation (1.99), we obtain the
convenient for numerical calculations equation

Zþ 1
a expðy2 Þdy
IV ðDkÞ ¼ I0V ð1:100Þ
p ðu  yÞ2 þ a2
1

Dk G 2 G 2 1=2
Here u ¼ dk G; a ¼
dkL
2dkG R ¼ ½ðdkD Þ þ ðdkA Þ 
; dkG A —the width of the hard-
; dkG
R R
ware outline.
1.6 Models of Light Emission and Absorption in Plasma 49

Fig. 1.6 The Holtsmark

function

Lorentz, Gaussian and, as a consequence, Voigt outlines have the symmetrical

appearance. The presence of ions in plasma leads to a further displacement of the
result outline and to the asymmetry of the last one. This mechanism of broadening
by ions is described by statistical theory [28]. The influence of it to the lines
outlines is represented in experimental proceedings. However in plasma with a high
concentration of ions, simultaneous impact to atoms of a large number of ions,
effects to the line broadening, and it should be taken into account.
The basis of the mechanism of broadening by ions is the Holtsmark function [28]
that is the probability of the presence of the field with the relative value β (caused
by the presence of ions with a concentration ni):

Z1 "  #
2 x 3=2
HðbÞ ¼ x sin x exp  dx ð1:101Þ
pb b
0

Here b ¼ E=E0 , E0 ¼ Ze=r02 —the average electric field intensity created by ions, Z
—the ion charge, r0 ¼ ð4pni =3Þ1=3 —the average distance between the ions.
The function HðbÞ, calculated by the numerical methods is shown at Fig. 1.6.
In practical terms, the statistical broadening by ions (manifested in the statistical
wing of the spectral line) should be expected for lines with large values of the
constants of the quadratic Stark effect. This asymmetry of the spectral lines has
been frequently observed in the Arc discharge.

1.6.4 The Absorption in Spectral Lines

The decrease in the light intensity with a spectrum in a narrow range of wavelengths
ðk; k þ dkÞ at the passing of a homogeneous medium with the layer thickness l and
the atomic transition line from the kth to ith level (k—the upper level) occurs
exponentially in Bouguer–Lambert–Berr law:
50 1 The Theoretical Basis of the Low-Temperature Plasma

I ðk; z ¼ lÞ ¼ I ðk; z ¼ 0Þ exp½vik ðkÞl ð1:102Þ

Here vik ðkÞ—the spectral absorption coefficient; vik ðkÞl—the spectral absorption
index or the measure of the optical density.
Similarly to the case of radiation the absorption coefficient in the lines is defined
as follows:

 Z1
gi nk
vik ¼ ðhmki =cÞBik ni 1  ¼ xik ðkÞdk ð1:103Þ
gk ni
0

Here mki —the frequency at the center of the radiation line outline; Bik —the Einstein
coefficient for absorption; ni and nk —the concentrations of particles in the lower
and upper levels; xik ðkÞ—the spectral intensity of absorption depending on the
wavelength k.
The Einstein coefficients for spontaneous Aki and induced Bki radiation can also
be used:

gi gi 8phm3ik
Bki ¼ Bik ; Aki ¼ Bik ð1:104Þ
gk gk c 3

Note that according to (1.103), using the measured relative changes in the
intensity of light passing through the object, the absolute value of the absorption
coefficient and respectively, the density of absorbing states (neglecting the induced
transitions) can be obtained.

References

1. C. Moore. Atomic energy levels. As derived from the analyses of optical spectra. —
Washington, U.S.G., 1958. – 309 p.
2. V. Golant, A. Gilinskiy, S. Sakharov. Basics of plasma physics. – M.: Atomizdat, 1977.
3. M. Mitchner, C. Kruger. The partially ionized gases — M.: Mir, 1976 — 496p.
4. Y. Raiser. Physics of gas discharge. — M.: Science, 1987 – 592 p.
5. H. Griem. Phys. Rev., V. 131, p. 1170, 1963.
6. R. McWhirter, A. Hearn // Proc. Roy. Soc. 1963, V.82, p. 641.
7. Wilson R., J. Quant. Spectr. Radiative transfer – V.2, 1962, p. 477.
8. R. Huddlestone and S. Leonard. The diagnostics of plasma: Mir, 1967.
9. M. Alievsky, V. Zhdanov. The transport equations for non-isothermal multicomponent
plasma. – 1963, №5 – p. 11 – 17.
10. Shi Nguyen-Kuok. The modeling of equilibrium plasma in RF- and Arc plasma torches //
Proceedings of the Int. scientific. conf. “Electro-physical and electrochemical techniques”
— St. Petersburg, 1997, pp. 63—66.
11. V. Kolesnikov The arc discharge in inert gases // Physical Optics: Proceedings of Phys. Inst
named by P. Lebedev, the USSR Academy of Sciences. – M.: 1964. - T.30. — pp. 66 — 157.
12. L. Biberman, V. Vorobev, I. Yakubov. Low-temperature plasma with nonequilibrium
ionization. // Successes of physical sciences. — 1979. — T. 128, №2. — pp. 233 — 271.
References 51

13. E. Velikhov, A. Kovalev, A. Rakhimov. Physical phenomena in gas-discharge plasma. – M.:

Science, 1987 – p. 160.
14. V. Goldfarb, E. Ilyina. The nonequilibrium low-temperature plasma sources. The populations
of the levels and spectral diagnostics. // Proc. of the Siberian department of the USSR
Academy of Sciences — №3, V.1, 1975, pp. 28 — 38.
15. J. Beulens, D. Milojevic, D. Schram, P. Vallinga. A two dimensional nonequilibrium model
of cascaded arc plasma flows. // J. Phys. Fluids, 3, 2548, 1991.
16. S. Dresvin, J. Amouroux, Nguen Quoc Shi. Analisis of deviation from thermal and ionization
equilibrium in an Argon plasma flow. // J. High temperature Material Processes. — V. 1, №3,
1997 — pp. 369—381.
17. I. Belov, S. Isaev. Modeling of turbulent flow — St. Petersburg: Publishing House of the
Baltic State Technical University, 2001 — p. 108.
18. A. Yun. Development and Analysis of Advanced Explicit Algebraic, Turbulence and Scalar
Flux models for Complex Engineering configuration — Doctor thesis. Darmstadt. 2005.
19. W. Jones, B. Launder. The prediction of laminarization with a two-equation model of
turbulence. // J. Heat Mass Transfer, 15: pp. 301—314, 1972.
20. D. Wilcox. Turbulence modeling for CFD — California, 1994.
21. F. Menter Zonal two Equation k-ω Turbulence Models for Aerodynamic Flows // AIAA paper
1993. P. 2906.
22. D. Wilcox. Turbulence Modeling for CFD. La Canada, California: DCW Industries Inc.,
1998.
23. F. Menter. Two-equation eddy-viscosity turbulence models for engineering applications //
AIAA Journal. 1994. Vol. 32, No 8. pp. 1598–1605.
24. A. Snegirev. High performance computing in technical physics. Numerical modeling of
turbulent flows: Proc. allowance. — St. Petersburg: Polytechnic University Press, 2009.
25. V. Ochkin. Spectroscopy of low-temperature plasma. — M.: FIZMATLIT, 2006.
26. Y. Raiser // Success of Phys. Sciences — 1969 — T. 99, № 4 — p. 545.
27. Optical pyrometer of plasma. // Ed. N. Sobolev. — M.: Publishing House of Foreign. lit.,
1960.
28. B. Smirnov. Ions and excited atoms in plasma. — M.: Atomizdat 1974.
Chapter 2
Classical Theory of the Particle Scattering

2.1 Classical Consideration of the Particle Scattering

For the determination of the plasma transport coefficients (electric conductivity,

thermal conductivity, viscosity, the ambipolar and thermal diffusion coefficient) it is
necessary to know the cross sections of particles interaction in plasma. The cal-
culation of the interaction cross sections should be carried out using the quantum
mechanical theory, but in the case of low-temperature plasma, when you need to
take into account the large number of types of interactions between the particles, the
bulky size and the complex method of calculation significantly limit its practical
application. In practice, for the calculation of the interaction cross sections in
plasma classical theory of binary collisions is often used according to the different
functions of the potential energy (potential functions) /ðrÞ that can be chosen using
the experimental data about the transfer coefficients. The paired elastic collisions of
neutral–neutral, ion–neutral, electron–neutral, and the Coulomb interaction of
charged particles are determined for transport phenomena in low-temperature
plasma.
The Classical Problem of Particles Paired Collisions The classical problem of
particles with the mass mi scattering on the particle with the mass mj in a spherically
symmetric potential field /ðrÞ is often reduced to an equivalent problem of the
mm
scattering of a particle with the reduced mass m ¼ mi þi mj j and the relative initial
velocity ~g ¼~vj ~vi in the same potential field. Schematically, the formulation of
this problem is shown in Fig. 2.1.
To analyze the dynamics of movement in equivalent case we write the equations
of conservation of angular momentum and energy in the form:

mgb ¼ mr 2 a_ ð2:1Þ

© Springer International Publishing Switzerland 2017 53

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_2
54 2 Classical Theory of the Particle Scattering

Fig. 2.1 Driving motion of the particle with mass mi in a spherically symmetric potential field
/ðrÞ in a collision with a particle of mass mj . a the angle that determines the direction of a particle
motion (polar angle); am the angle, corresponding to the maximum convergence of particles; b the
impact parameter; h the scattering angle; r the current distance between the particles; rm the
distance of the maximum particle convergence

1 2 1 2 1 2 2
mg ¼ m_r þ mr a_ þ /ðr Þ; ð2:2Þ
2 2 2

where a_ ¼ da
dt ; r_ ¼ dt .
dr

If we put a value a_ from (2.1) to (2.2), we get:

1 2 1 2 1 2 b2
mg ¼ m_r þ mg 2 þ /ðr Þ ð2:3Þ
2 2 2 r

The Eq. (2.3) does not contain a variable a and it can be used to describe the
one-dimensional motion of a particle with mass m with and a total energy 12 mg2
2 2
moving in an effective potential field /eff ðr Þ ¼ /ðr Þ þ 12 mg2 br2 , here 12 mg2 br2 —the
centrifugal potential.
The scattering angle hðg; bÞ is defined as:

hðg; bÞ ¼ p  2am ð2:4Þ

here am —the angle, corresponding to the maximum convergence of particles.

Taking into account (2.1) and (2.3), we have the following relation:
2.1 Classical Consideration of the Particle Scattering 55

da=dt b=r 2
da ¼ dr ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dr ð2:5Þ
dr=dt 1  /ðrÞ  b
2
1=ð2mg2 Þ r2

Integrating (2.5), we obtain an expression for am

Z1
b=r 2
am ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dr ð2:6Þ
/ðrÞ
rm 1  1=ð2mg 2 Þ  r2
b2

Substituting (2.6) into (2.4), we have

Z1
dr=r 2
hðg; bÞ ¼ p  2b qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi; ð2:7Þ
/ðr Þ
rm 1  1=ð2mg 2 Þ  r2
b2

where the distance of the maximum convergence of particles rm is determined from

d r /ðrm Þ b2
the condition dt  ¼ 0, i.e., by solving the equation 1  1=ð2mg 2 Þ  r 2 ¼ 0.
r¼rm m

If we take into consideration the binary collisions only, some of plasma transport
coefficients can be expressed using the deflection angle. The Eq. (2.7) is true for
any spherically symmetric interaction potentials /ðrÞ, so the knowledge about the
potential function /ðrÞ is important for determination of the binary collision
dynamics.
Differential, Efficient, and Averaged Effective Interaction Cross Sections
The differential cross section is the ratio of the differential probability of particles
scattering to the flow of incident particles in the elementary solid angle
dX ¼ 2p sin hdh. The differential cross section is dr ¼ 2pbdb and
rðE; hÞ ¼ ddXr ¼ 2p2pb db . The effective cross section QðlÞ ðEÞ for interaction of order
sin hdh
l, depending on the given initial relative power E ¼ mg2 =2 is defined as:

Z1 Zp
ðlÞ
   
Q ðEÞ ¼ 2p 1  cos h bdb ¼ 2p
l
rðE; hÞ 1  cosl h sin hdh ð2:8Þ
0 0

In the particular case l = 1 we have the cross section of the momentum transfer
or the diffusion cross section, for l = 2—the viscosity section, etc.
The average cross section depending on the temperature, using the Maxwellian
velocity distribution, can be determined as:
56 2 Classical Theory of the Particle Scattering

Z1  
ðl;sÞ 2ð l þ 1Þ E
h i ec c2ðs þ 1Þ QðlÞ ðE Þd
2
Q ðT Þ ¼
l
ðs þ 1Þ! 2l þ 1  ð1Þ 0 kT
Z1
2ð l þ 1Þ
¼ h i ex xs þ 1 QðlÞ ðkTxÞdx ð2:9Þ
l
ðs þ 1Þ! 2l þ 1  ð1Þ 0
rffiffiffiffiffiffiffiffiffi
4ð l þ 1Þ 2pm ðl;sÞ
¼ h i X ðT Þ;
ðs þ 1Þ! 2l þ 1  ð1Þ l kT

2
Here x ¼ c2 ¼ mg
2kT ¼ E=ðkTÞ—the initial relative velocity of the colliding particles.
Accordingly, the collision integral Xðl;sÞ ðT Þ of the order ðl; sÞ, depending on the
temperature, is determined as follows:
rffiffiffiffiffiffiffiffi Z1 Zp
pkT  
Xðl;sÞ ðT Þ ¼ ex xs þ 1 rðE; hÞ 1  cosl h sin hdhdx
2m
0 0
h i
rffiffiffiffiffiffiffiffiffi Z1 rffiffiffiffiffiffiffiffiffi
kT ðs þ 1Þ! 2l þ 1  ð1Þl kT ðl;sÞ
¼ x s þ 1 l
e x Q ðkTxÞdx ¼ Q ðT Þ ð2:10Þ
8pm 4ðl þ 1Þ 2pm
0

In the future, data about the averaged effective cross sections Qðl;sÞ ðT Þ are used
to calculate coefficients of plasma transport. The procedure of calculating the
transport coefficients usually begins with a definition of the function of the inter-
action potential of the particles /ðrÞ and the differential cross section rðE; hÞ. These
data dependent on the chemical composition and density of plasma, difficult col-
lisional and radiative processes in it as well as on the specific conditions of plasma
existence. There are a huge number of literatures on these issues.
For a gas that consists of the neutral particles, the interaction potential is
short-range at the long distance (r [ R) that decreases so rapidly that the integral
R1
/ðr Þdr tends to zero. As an example, we note that at large distances the two
R
charged particles interact by the Coulomb law /ðr Þ  1=r, charged and neutral
particles are attracted because of the interaction of charged particles and the induced
dipoles /ðr Þ  1=r 4 and two neutral particles are attracted due to the interaction of
instantaneous dipoles. In all cases, the particles are attracted to large distances to
each other, i.e., the potential energy is negative, and repel at short distances due to
the overlap of the wave functions of the particles. Here are some potential functions
depending on the intermolecular distance r that are often used in calculation of the
interaction cross sections of particles and plasma transport coefficients [1]:
2.1 Classical Consideration of the Particle Scattering 57

Z Z e2
(1) /ðr Þ ¼ i rj expðr=rD Þ—the potential of the Coulomb interaction. Here
Zi ; Zj —the charges of the particles, rD —the Debye radius;
(2) /ðr Þ ¼ /0 expðr=q0 Þ—the interaction potential between atoms and mole-
cules. Hereh /0 , q0 —the some constants;
r12 r6 i
(3) /ðr Þ ¼ 4e r  r —the Lennard–Jones potential with polynomials of
6th and 12th degrees, that is often used to describe the induced dipole–dipole
interaction of molecules. The value r—distance r, where /ðr Þ ¼ 0, e—the
maximum energy of attraction or the depth of the potential well that is reached
pffiffiffi
at r ¼ 6 2r;
(4) /ðr Þ ¼ 1, r  r; /ðr Þ ¼ 0, r  r—model of hard spheres with a diameter r;
(5) /ðr Þ ¼ dr d —a model of the point center of repulsion. The value d ¼ 9  15
is called as the
h repulsive
h  index, i d—is a constant;
r 6 i
(6) /ðr Þ ¼ 16=a a exp a 1  rm  r
e 6 r m
; where r  rmax ; /ðr Þ ¼ 1,
r  rmax —the modified potential of Buckingham that is used to describe the
induced dipole–dipole interaction of molecules. Here rm—the equilibrium
distance between the nucleus of the molecule that is formed with the repulsive
particles; rmax —the distance r, that corresponds to the relative maximum /ðr Þ.
The ratio rmax =rm is determined by the smallest root from the transcendental
equation
 
 
rmax 7 rmax
exp a 1   1 ¼ 0;
rm rm

(7) /ðrÞ ¼ e½expð2bðr  rm Þ=rm Þ  2 expðbðr  rm Þ=rm Þ—the Morse potential

where b = 1.6739 Å−1.
(8) /ðrÞ ¼ A=r n —the inverse-repellent potential. Where A—a constant.

The Trajectories of the Particle in a Pair Collision The scattering angle is p for
the small values of the impact parameter, because the direct reflection of the par-
ticles occurs (Fig. 2.2, 3).
For small impact parameters the scattering angle is small and positive (Fig. 2.2, 1).
At the large values of the impact parameter, the scattering angle of impact is small and
negative due to the attraction of particles (Fig. 2.2, 2).
Between these two cases the function hðg; bÞ if it is continuous, it has at least one
zero value (Fig. 2.2, 4).
As a result, the dynamics of interaction between colliding particles can be treated
several times around the scattered particles (Fig. 2.2, 5).
In the experimental determination of the angular dependence of the scattered
particles the Ramseur device is used, that consists of multiple detectors positioned
58 2 Classical Theory of the Particle Scattering

Fig. 2.2 Possible tracks of particle motion in a pair collision: S the screen

at different angles and recording the scattered particles. These detectors cannot
differ the particles that made several turns.
The function hðg; bÞ for some impact parameter can tend to infinity. This one
corresponds to the fact that the attraction is compensated with the repulsion, i.e., the
particles are captured or “orbiting” (Fig. 2.2, 6).
In practice, there are two cases of dispersion that are known as “radiance” when
hðg; bÞ ¼ 0 (Fig. 2.2, 4) and “rainbow scattering” when @h@b ðg;bÞ
¼ 0 (Fig. 2.2, 7).
For the case of “radiance” particle does not change its direction of movement,
despite the fact that part of the track in the field of the potential action, for example,
2
when the condition /ðr Þ ¼  12 mg2 br2 is performed.
The problem of light propagation in the atmosphere drops leads to the optical
effect “radiance”.
The Conditions for Application of the Theory of Classical Mechanics for the
Scattering of Particles In classical mechanics particles are considered infinitely
small. In quantum mechanics a wave packet with the de Broglie wavelength cor-
responds to a particle. Provisions of classical mechanics, especially violated in the
scattering at small angles because of occurrence of diffraction. The small angle at
the classical scattering can be defined as the ratio of the transverse and longitudinal
pulses:
2.1 Classical Consideration of the Particle Scattering 59

Dp? b @/
hclass   :
p pv @r
In the scattering of particles in the power center the cross change in momentum
is defined as Dp?  h=a, where a  b—the size of the gap. Quantum scattering
angle or the diffraction angle is estimated as

Dp? h k
 k

hdif     :
p pa a b

The condition of the classical mechanics application is as follows:

 
b @/ @ / k

¼b  : ð2:11Þ
pv @r @r 2E b

ðbÞ
As a rough approximation (2.11) has the form: /2E  bk. In this condition, we can
restrict with the classical consideration. k   1=m and the classical description is
mainly affects for electrons. In this case, quantum effects play a significant role, and
the scattering of electrons in plasma should be considered by the theory of quantum
mechanics.

2.2 Determination of Cross Sections for Plasma Particles

Interaction

2.2.1 Coulomb Scattering: The Cross

Section of the Interaction of Charged Particles

Among all the forces of interaction between atomic particles the Coulomb forces
has the slowest decrease with distance (*r 2 ) and potential energy of interaction
/ðr Þ  r 1 . Coulomb forces have the greatest long-range, i.e., collision occurring at
large impact parameters, make a major contribution to the loss of a particle
momentum. The low-temperature plasma with the degree of ionization of more than
0.1–1 % a collision of charged particles with each other play the decisive role for
the drift and diffusion of charges and energy exchanges. The small-angle scattering
in the short-range potential can be considered, that is valid for almost all impact
parameters. The scattering angle θ in this case can be determined by the relation of
the longitudinal cross-pulse Dp? pulse to the pjj of dissipating particles (Fig. 2.3).
Given the fact that r 2 ¼ b2 þ v2 t2 we have:
60 2 Classical Theory of the Particle Scattering

Fig. 2.3 To the derivation of Rutherford’s equation

Z1 Z1 Z1  
Dp? 1 2 b 2b @/ 1
h ¼ F? dt ¼ F dt ¼  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dr
pjj mv mv r mv @r v r  b2
2
0 0 b
ð2:12Þ
2 Z1
Zi Zj e b dr
¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
mv2 =2
b r 3 1  ðb=r Þ2

where Zi ; Zj —the charges of the particles. If we change the variable b=r ¼ sin h,
we have r ¼ b= sin h; dr ¼  b cos hdh. Substituting the last relation in (2.12) and
sin2 h
integrating, we obtain

Zi Zj e2 Zi Zj e2
hðE; bÞ ¼ ¼ ð2:13Þ
bmv2 =2 Eb

As a result,

Zi Zj e2
bðE; hÞ ¼ ð2:14Þ
Eh

The differential cross section for the Coulomb potential is

 2
dr 2pb db Zi Zj e2 1
rðE; hÞ ¼ ¼ ¼ ð2:15Þ
dX 2p sin h dh E h4

Equation (2.15) is called the Rutherford’s equation. The rapid growth of the
differential cross section with decreasing scattering angle (r  1=h4 ) and inversely
proportionality of θ to the square of kinetic energy of the relative motion is a
characteristic of Coulomb scattering.
Rutherford’s equation in up to a factor coincides with the equation obtained for
the differential cross section in accordance with the quantum mechanical theory.
2.2 Determination of Cross Sections for Plasma Particles Interaction 61

This is due to the condition of classical mechanics application hv=e2 1, that

expresses the smallness of the de Broglie wavelength k  compared to the charac-
teristic size of the force field r0 , i.e., the distance between the particles, when the
potential energy of the Coulomb interaction /ðr Þ ¼ e2 =r is comparable to the
Z Z e2
kinetic energy of the relative motion E ¼ mv2 =2. The value r0 ¼ mv i j
2 is called as

the Coulomb radius. The inequality hv=e 1 is equivalent to k

2
 r0 . It imposes a
restriction on the kinetic energy: E me4 =ð2h2 Þ. For electrons in the ion field the
condition E IH ¼ 13:6 eV must be satisfied. In most cases that are interested in
the physics of low-temperature plasma: E IH . Thus, the calculation of the
electrons scattering in the field of the other charges in up to a factor can be
considered in accordance with the theory of classical mechanics. For ions it can be
fair to energies of tens of keV.
For small scattering angles (1  cos h  h2 =2) scattering cross section of the
vehicle in the Coulomb potential

Zhmax  2  2
ð1Þ Zi Zj e2 1 h2 Zi Zj e2 hmax
Q ðEÞ ¼ 2p 4 2
hdh ¼ p ln
E h E hmin
hmin ð2:16Þ
 2
Zi Zj e2
¼p ln K;
E

where hmax ; hmin —the maximum and minimum scattering angles; ln K ¼ ln hhmax min
¼
ln rmax
r0 —the Coulomb logarithm; r max —the maximum radius of the scattering.
Using the Coulomb potential of a charged particle /ðrÞ ¼ Zi Zj e2 =r without
shielding by surrounding charged particles would lead to a divergence of the
integral (2.7). In order to get the final value of the cross section, many authors
consider it possible to limit the upper limit of integration in (2.7) with the screening
parameter, taking it equal to the radius of the Debye rD , or even the average
interatomic distance (r ¼ n1=3 ).The different values of the effective interaction
cross sections follow from here.
As for the potential function, interaction of charged particles with small scat-
tering angles are widely described by a screened Coulomb potential:
Z Z e2
/ðrÞ ¼ i rj expðr=rD Þ. Such a potential function shows that the averaged
effective cross sections of the electron–ion interaction in the fourth order of
approximation have the form [2]
62 2 Classical Theory of the Particle Scattering


ð1;sÞ 4p 1
Q ¼ b ln K   2c þ wðsÞ ;
2
s ð s þ 1Þ 0 2
12p 2
Qð2;sÞ ¼ b ½ln K  1  2c þ wðsÞ;
s ð s þ 1Þ 0

 ð2:17Þ
12p 2 7
Qð3;sÞ ¼ b0 ln K   2c þ wðsÞ ;
s ð s þ 1Þ 6


16p 2 4
Qð4;sÞ ¼ b0 ln K   2c þ wðsÞ ;
s ð s þ 1Þ 3

Zi Zj e2
Here K ¼ 2rb0D ; b0 ¼ 2kT —the
distance of maximum convergence of colliding
sP
1
particles; c ¼ 0:5772—Euler’s constant; wðsÞ ¼ 1=n, (wð1Þ ¼ 0).
1
The averaged cross section of the electron–ion interaction, obtained from (2.17)
for l = 1, s = 1 [2] is shown in Fig. 2.4.
In the literature, there are other equations for calculation of the effective cross
section of interaction of charged particles. One of them [3, 4] has the form
!
a
1010 b
104 Te
Qei ¼ ln ; ð2:18Þ
Te2 1=3
ne

Here a, b—are numerical coefficients. These factors vary among different

authors. Thus, in [3] a = 5.85, b = 5.35; in [4] a = 2.79, b = 5.98. These

Fig. 2.4 The effective cross

section of the electron–ion
Qei ðTe Þ at the thermal
equilibrium. 1–6, 8
calculation of equations,
respectively (2.20), (2.21),
(2.18) with a = 5.85,
b = 5.35, (2.18) when
a = 2.79, b = 5.98, (2.19),
(2.22), (2.23); 7 data [2]
2.2 Determination of Cross Sections for Plasma Particles Interaction 63

differences are generally related to the choice of impact parameters in the Coulomb
logarithm. Most often, the quantity on the logarithm, is the ratio of Debye radius to
sustainable medium parameter of the collision, resulting to the dispersal of particles
to the angle of 90°.
Michner and Kruger [5] found a relatively simple equation to calculate the
effective cross section of the electron–ion:

Qei ¼ 6pb20 ln K; ð2:19Þ

Here K ¼ rbD0 . In this case, the average sighting distance b0 and the Debye radius is
calculated as:
rffiffiffiffiffiffiffiffiffi
b0 ¼ e2 e0 kT
; rD ¼ :
12pe0 kT ne e 2

In [6] the cross section Qei for two-temperature plasma is calculated by the
equation
" #
8:78
1010 2:79
104 Te
Qei ¼ ln 1=3 : ð2:20Þ
Te2 ne ð1 þ Te =Ta Þ1=3

In [7] for the cross section Qei the equation is:

!
6:58
1010 2:57
104 Te
Qei ¼ ln 1=3 : ð2:21Þ
Te2 ne ð1 þ Te =Ta Þ

In [8] using the methods of quantum field theory to derive the expressions are
given using the renormalized scattering probability that automatically limit the
interaction at the large impact parameters and eliminate the need to artificially
introduce a lower limit of integration. Thus obtained expression for the cross
section of electron scattering on the ions is given by [9]:

  
7
1010 6 T
Qei ðTÞ ¼ lg 2:8
10 p ffiffiffiffi
ffi þ 3 : ð2:22Þ
T2 ne

Results of the effective cross section calculation of the electron–ion of different

expressions (Fig. 2.4) shows that the latter expression gives the most consistent
with the data Devoto [2]. For more matches with the data [2] we suggest the
following calculation equation obtained by approximation (2.22):

  
6:4
1010 6 Te
Qei ¼ lg 2:8
10 p ffiffiffiffi
ffi þ 3 : ð2:23Þ
Te2 ne
64 2 Classical Theory of the Particle Scattering

These calculations [10] shows that in a dense low-temperature plasma the mass
difference between electrons and ions do not significantly affect the angle of dis-
persion for interactions of electron–ion, electron–electron, and ion–ion and average
ðl;sÞ ðl;sÞ ðl;sÞ
effective cross sections of the highest order Qei ðTÞ; Qee ðTÞ; Qii ðTÞ are dis-
tinguished from each other by not more than 3–7 %, in the future, you can take
ðl;sÞ ðl;sÞ ðl;sÞ
them equal: Qei ðTÞ ¼ Qee ðTÞ ¼ Qii ðTÞ. By (2.17) we also obtain that the
average effective cross sections are proportional to the square of the charge of the
ions.
Electrons play the dominant role in the dynamics of the process of screening, the
temperature in the equations for the cross section of the interaction of charged
particles is often taken equal to Te (ions are considered to be fixed in comparison
with electrons).
The results obtained by author in the calculation of two-temperature section
Qei ðTe ; Ta Þ also show that it depends mainly on the electron temperature Te.

2.2.2 Ramseur Effect and the Resonance Scattering

of the Electrons by Atoms

Elastic scattering of electrons on neutral particles can be described only by means of

quantum mechanical theory. Currently, it is possible to calculate analytically the
effective scattering cross sections for simple atoms only. For more complex parti-
cles, there is a considerable discrepancy between the experimental data and the
results of theoretical calculations, especially for low-energy electrons (0.1–2 eV).
In this range, the Ramseur minimum effect significantly affects the determination of
the effective interaction cross sections of electron–atom, such as the scattering of
electrons in argon. The wave properties of the electron appear in the existence of
minimums. In some of the wave energy that describe the motion of the electron (de
Broglie wavelength) and dispersed potential field, interfere and cancel each other
out. The probability and the scattering cross section depend on the intensity of the
scattered wave and it is very small. Such a situation occurs when a low-energy,
different for different gases, when the de Broglie wavelength of an electron is
comparable with the size of the atom. At the same time as a result of the diffraction
of electrons per atom there is an amplitude maximum of the electron wave and the
electron as “skirts” atom without scattering on it.
Results of calculation of the effective cross section of the electron–atom inter-
action, depending on the power using the so-called phase shift [11] gl are in good
agreement with the experimental data. Consequently, experimental data remain a
reliable source for determining the effective interaction cross section of electron–
atom in low-temperature plasma. Results of calculation of the average cross section
of the interaction of the electron–atom [2] for l = 1, s = 1 are shown in Fig. 2.5.
2.2 Determination of Cross Sections for Plasma Particles Interaction 65

Fig. 2.5 The average

effective interaction cross
sections for argon depending
ð2;2Þ
on the temperature. Qaa ðTÞ,
ð2;2Þ ð1;1Þ
Qai ðTÞ, Qea ðTÞ the
interaction cross sections
atom–atom, atom–ion,
electron–atom, respectively;
ð1;1Þ
Qai:per ðTÞ the exchange cross
section of atom–ion

2.2.3 The Cross Sections for the Interaction

of Molecule–Molecule, Atom–Atom

The potentials /ðrÞ of the heavy particles are divided into long-range and
short-range. Long-range potentials between the heavy particles determine the values
of the effective cross sections only at low energies of particles in the conditions of a
temperature not more than 3000 K. Below this limit as a function approximating
the interaction potential,
h the Lennard–Jones potential with polynomials 6th and
r 12
 r 6 i
12th degree /ðrÞ ¼ 4e r  r [1] or a modified capacity of Buckingham [10]
well established itself. At T > 3000 K the contribution of the short-range potential
in the effective cross section becomes significant, and it is necessary to use the
exponential type of function for approximation:

/ðrÞ ¼ /0 expðr=q0 Þ; ð2:24Þ

Here /0 ; q0 —some constants. For example, the potential of atom–atom interaction

in argon at /0 ¼ 7100 eV, q0 ¼ 0:258 Å [2] (or /0 ¼ 32300 eV, q0 ¼ 0:224 Å
[10]) /ðrÞ ¼ 0:19  14:5 eV, (r ¼ 2:716  1:598 Å).
Another approximation is the Morse potential:

/ðrÞ ¼ efexp½2bðr  rm Þ=rm   2 exp½bðr  rm Þ=rm g ð2:25Þ

here rm—the equilibrium distance between the kernels of molecules formed of

repulsive particles. Detailed tables of collision integrals for the Morse potentials can
be found, for example, in [12, 13].
In selection of appropriate potential functions the polarity of the particles,
especially with unfilled outer electron shells and others, should be also considered.
66 2 Classical Theory of the Particle Scattering

Fig. 2.6 The results of calculations of collision integrals with the order l = 1 (a) and l = 6 (b) for
the atom–atom interaction cross sections

The Average Effective Interaction Cross Section of Atom–Atom Theoretical

differential scattering cross sections were calculated by A. Phelps and others [14–18]
using the scattering phase shifts estimated by approximation of G. Ventzela,
H. Kramers, and L. Brillouin. The theoretical results of asymmetric scattering of
argon atoms in an Argon plasma are approximated by the following function:

dr 1 1
ðhÞ ¼ a½1 þ 2g  cos h2 þ b ð2:26Þ
dX 2 2

Here
8 h  0:8  e 1:3  e 2:2 i1
>
> aðeÞ ¼ 0:27e0:2 1 þ 9e þ 350 þ 800 ;
>
>
< h       i1
0:5 1:1 2
bðeÞ ¼ 3:0e0:2 1 þ 15e þ 800e
þ 10 e000 ; ð2:27Þ
>
> h i1
>
>      
: gðeÞ ¼ 0:002e0:75 1 þ e 1:15
þ 80e
1:95
þ 500
e 2:5
:
6

dX ¼ sinðhÞdhdu—differential solid angle of scattering.

The collision integrals to l = 6, s = 11 were obtained by direct integration of the
Eqs. (2.8) and (2.9) for the quantum elastic scattering differential cross sections
(Fig. 2.6).

2.2.4 The Cross Sections for the Interaction

of the Atom–Ion

When the ion closes to molecule or atom, it polarizes them by their Coulomb field,
and this ion experiences the induced dipole moment in the molecule. The inter-
action potential of the atom–ion is a more long-range than the potential of the
2.2 Determination of Cross Sections for Plasma Particles Interaction 67

atom–atom, as, in addition to these components, there are electrostatic attraction of

charge-induced dipole and pull-type charge-induced dipole.
Located at a distance r from the molecule, the singly charged ion induces a
dipole moment d ¼ ae=r 2 in the molecule by its electric field (E ¼ e=r 2 ), here a—
the polarizability of the molecule. Positive ions attract electrons that are molecular
closer to him, and the negative pushes them away. Therefore, the polarizing nature
of the interaction is attractive. The electric field intensity of the induced dipole in
the point, where the ion is located, E ¼ 2d=r 3 ¼ 2ae=r 5 , its potential
uðrÞ ¼ ae=ð2r 4 Þ. The force that acts to ion is F ¼ 2ae2 =r 5 , the potential energy
of interaction between the ion and the molecule /ðrÞ ¼ ae2 =ð2r 4 Þ, and the energy
is negative regardless of the sign of the ion charge.
Approximation of the ion–atom interaction potential using (2.24) or (2.25) can
be written as

/ðrÞ ¼ /0 fexp½2C ðr  re Þ=r  2 exp½Cðr  rm Þ=rg ð2:28Þ

Here C ¼ r ln 2=ðre  rÞ; σ—the distance of the maximum convergence of two

particles. For example, to approximate the interaction potential of the atom–ion in
argon /0 ¼ 1:25 eV, re ¼ 2:434 Å, C = 3.382, r ¼ 2  2:9 Å [2]. Results of
calculation of the average effective interaction cross section of atom–ion for argon
ð2;2Þ
plasma Qai ðTÞ using the function (2.28) are shown in Fig. 2.5.
Determination of the average effective atom–ion cross section is extremely
important for the transport coefficients. The low-temperature plasma with ion
concentration ni [ 1014 cm−3 (p = 105 Pa = 1 bar) [19] a significant process of
atom–ion recharge, when the ions move in their own gas, is observed. The positive
ion is accelerated by electric field between collisions or just wandering in the gas
passing by a neutral atom, the latter selects the outer electron is neutralized and
continues the inertia path in the form of a neutral atom. Recharging takes place even
in their flying at large impact parameters that occurs very often. For electron
transfer the contact time it is sufficient, as it goes from an atom to ion with a speed
about its rate in atom (106 m/s), that is much larger than the velocity of the ions. If
the process is in an electrostatic field, the new formed ion begins to accelerate in the
direction of the field with an average zero directed velocity. So, in an act of
recharging occurs total loss of directional momentum of the ion. Recovery is
possible in the case of heterogeneous particles, but it most often occurs one time
and as a result of not always feasible conditions on the energy state of the particles
with different ionization potentials.
Recharge in a gas is interesting because of their result of energy transfer is
equivalent to elastic collision, although it accompanied by a transformation of
particles as an inelastic collision. The recharging processes are quantum mechanical
in nature, so their cross sections are more than elastic scattering cross section of
atom–ion interaction. It is extremely difficult to investigate the overcharging slow
ions experimentally and the spread of the experimental data is more than 100 %.
Typically, the atom–ion charge sections are measured at ion energy of more than
68 2 Classical Theory of the Particle Scattering

1 eV or more. In this case, the theoretical equation is considerable interesting but its
calculation error is roughly 50 % [19]:
 rffiffiffiffiffiffi
EH 2 100v0 Ei
Qai:per ¼ pa20 ln K ð2:29Þ
Ei vi EH

Here a0 = 0.529 × 10−10 M—the Bohr radius of the hydrogen atom; Ei—the
ionization potential of atoms (molecules) when the ion is formed; EH = 13.6 eV—
the ionization potential of the hydrogen atom; v0 ¼ 2:19
106 m/c—the speed of
qffiffiffiffiffiffi
the electron in the first Bohr orbit in the hydrogen atom; vi ¼ 8kT
pmi —the average
i

thermal velocity of the ion; K—a correction coefficient.

In practical use of the Eq. (2.29) to calculate the cross section the coefficient
K should be selected and the follow calculated effective viscosity and thermal
conductivity, where the charge-section are included, must coincide with the
experimental ones. For the argon plasma K = 0.5. These calculations of recharge
section with K = 0.5 are shown in Fig. 2.5 and they have a good agreement with
the experimental data [20].
The Average Effective Interaction Cross Section of Atom–Ion in Argon For the
interaction of the atom–ion in argon (Ar–Ar+) the processes of elastic collisions
should be taken into account, as well as the collision processes with the charge
ðl;sÞ
transfer. They correspond to the collision integrals Xij of odd l order. Another
approximation was proposed by E. A. Mason and others [2, 14, 21–23], it does not
take into account the contribution of elastic collisions there:

QðlÞ ¼ 2rex ; ð2:30Þ

where rex —the cross section of the charge transfer.

The collision integrals Qðl;sÞ of odd l—order have been obtained using the
Devoto equation [2, 11, 24]:
 2 !
ðl;sÞ Bx Bf B2 p2 X sþ1
1
Q ¼ A  ABx þ
2
þ ½Bx  2A þ  þf 2
2 2 4 6 n2
 
 2
n¼1
ð2:31Þ
B T B T
þ ½Bðx þ fÞ  2A ln þ ln ;
2 M 2 M

sP
þ1
Here x ¼ lnð4RÞ, f ¼ 1
n  c, and c  0:57721—the Euler’s constant. The
n¼1
parameters A and B are the approximating coefficients of the charge transfer section
2
(its dimension Å ) as a function of the collision velocity v, cm/s:
2.2 Determination of Cross Sections for Plasma Particles Interaction 69

1
rex ¼ ½A  B ln v2 ð2:32Þ
2

Estimated values for these coefficients were determined by Cressault [25] for the
þ þ
even and odd divisions 2 Rg=u and 2 Pg=u states of the molecular ion Ar2þ , that
correlates with the ground state of the atom Arð1 sÞ and ion Ar þ ð2 pÞ:
(
þ
2
Rg=u ) 26:39; B ¼ 1:12;
2 þ ð2:33Þ
Pg=u ) 18:96; B ¼ 0:83:

Results of the collision integrals calculation for the interaction cross sections of
atom–ion argon at l = 1; 5 and l = 2; 6 are shown in Fig. 2.7 and 2.8.

Fig. 2.7 Results of the collision integrals calculation for the interaction cross sections of the
recharge atom–ion Ar–Ar+, l = 1 (a) and l = 5 (b)

Fig. 2.8 Results of the collision integrals calculation for the interaction cross sections of the
atom–ion Ar–Ar+, l = 2 (a) and l = 6 (b)
70 2 Classical Theory of the Particle Scattering

Full collision integrals were obtained by averaging of the pair over all states
using statistical weights equal to the multiplicity of states:
P ðl;sÞ
xn Qn ðl;sÞ ðl;sÞ
QR þ 2QP
Qðl;sÞ ¼ P
n
¼ : ð2:34Þ
xn 3
n

References

1. J. O. Hirschfelder, Ch. F. Curtiss, R. B. Bird. Molecular theory of gases and liquids / John
Wiley and Sons, New York. 2nd Edition: 1964.
2. Devoto R.S. Transport coefficients of ionized argon // Phys. Fluids. 1973. Vol. 16. No. 5.
P. 616–623.
3. Hoffert I., Lien H. Quasi one-dimensional, nonequilibrium gas dynamics of partially ionized
two-temperature argon // Phys. Fluids. 1977. Vol. 10. No. 8. P. 1769–1777.
4. Finkelberg V., Mekker G. Electric Arcs and thermal plasma. M.: Publishing House of Foreign
Literature, 1961.
5. M. Mitchner, C. Kruger. Partially ionized gases. M.: Mir, 1976.
6. Kolesnikov V.N. The Arc discharge in inert gases // Physical Optics (Proc. of Phys. Inst of
Acad. Sc. of USSR.). M.: Publishing PhIAN, 1964. V. 30. pp. 66–157.
7. Polyansky V.A. Diffusion and conductivity of a partially ionized gas mixture multitemper-
ature// J. of Technical Physics Letters. 1964. № 5. pp. 11–17.
8. Konstantinov O., Perl B.// J. of Experimental Technical Physics. 1964. V. 30. P. 1664.
9. Ismailova D. Study of the electric Arc in Argon/ Ed. by A.F. Jeenbaev. Frunze: Ilim 1963.
10. Kulik P.P. Essays on physics and chemistry of Low-temperature plasma/ Ed. by L.S. Polak.
M.: Nauka, 1971.
11. Devoto R.S. Transport coefficients of partially ionized argon// Phys. Fluids. 1967. Vol. 10.
No. 2. P. 354–364.
12. Samuylov E.V. Tsitelauri N.N.// J. Thermal physics of High Temperatures. 1964. V. 2.
P. 565.
13. Samuylov E.V. Tsitelauri N.N. Research on Physical hydrodynamics. M.: Nauka, 1965.
14. Phelps A.V., Greene C.H., Burke J.P.// J. Phys. B 3. 2000. Vol. 33. P. 2965.
15. Aziz R.A. Interatomic potentials for rare-gases: pure and mixed interaction// Inert Gases,
Springer Series in Chemical Physics. Berlin: Springer-Verlag. 1984. Vol. 34. P. 5–86.
16. Aziz R.A., Chen H.H.// J. Chem. Phys. 1977. Vol. 67. P. 5719.
17. Aziz R.A., Slaman M.J.// J. Chem. Phys. 1990. Vol. 92. P. 1030.
18. Aziz R.A.// J. Chem. Phys. 1993. Vol. 99. P. 4518.
19. Райзер Ю.П. Физика газового разряда. М.: Наука, 1987.
20. Nichols B.J., Witteborn F.C. // NASA TN D. 1966. P. 3265.
21. Mason E.A., Vanderslice J.T., Yos J.M. // Phys. Fluids. 1959. Vol. 2. P. 688.
22. Mason E.A. Higher approximations for the transport properties of binary gas mixtures.
I. General formulas // J. Chem. Phys. 1957. Vol. 27(1). P. 75–84.
23. Mason E.A. Transport properties of gases obeying a modified Buckingham (exp-six) potential
// J. Chem. Phys. 1957. Vol. 22(2). P. 169–186.
24. Devoto R.S. Simplified expressions for the transport properties of ionized monatomic gases //
Phys. Fluids. 1967. Vol. 10. No. 10. P. 2105–2112.
25. Cressault Y., Gleizer A. // J. Phys. D. 2004. Vol. 37. P. 560.
Chapter 3
Quantum Mechanical Theory
of the Particle Scattering

3.1 The Schrödinger Equation

The fundamental equation of quantum mechanics is the Schrödinger equation. It is

intended to describe the dynamics of particles motion without spin at speeds much
less than the speed of light. In the case of fast particles and particles with spin its
generalizations are applied (the Klein–Gordon, Pauli, Dirac equations, and others).
In quantum physics, the wave function w (complex-valued) is used to describe
the state of a moving particle. The most common Copenhagen interpretation of the
wave function is related to the probability of finding a particle in a pure state (the
square of the wave function is a probability density). Let the wave function to be set
in an n-dimensional space and ~ r ðx1 ; x2 ; . . .; xn Þ at each point at the time t, it will
look like wð~r; tÞ. For the three-dimensional space in spherical we have~ r ðr; h; uÞ. In
general, the Schrödinger equation can be written as:

h2 @
 Dwð~
r; tÞ þ /ð~ r; tÞ ¼ ih wð~
r Þwð~ r; tÞ; ð3:1Þ
2l @t

Here h ¼ 2p
h
, h—the Planck’s constant; μ = m—the particle mass. In the center of
mass system for the particles scattering μ—the reduced mass of the particles;
/ð~r Þ—the external to the particle potential energy in the point ~r ðx1 ; x2 ; . . .; xn Þ.
The solution of Schrödinger equation with respect to the time should be simple,
because the time is included in this equation using the first derivative in its
right-hand side. Indeed, a particular solution for the case /ð~
r Þ is not a function of
time and it can be written as

wð~
r; tÞ ¼ wð~
r Þ expðiEt=hÞ; ð3:2Þ

Here the function wð~

r Þ should satisfy the equation

© Springer International Publishing Switzerland 2017 71

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_3
72 3 Quantum Mechanical Theory of the Particle Scattering

h2
 Dwð~
r Þ þ /ð~
r Þwð~
r Þ ¼ Ewð~
r Þ; ð3:3Þ
2l

This expression is obtained from Eq. (3.1) by substituting of (3.2) in it. The
resulting Eq. (3.3) does not contain the time and called stationary Schrödinger
equation.
Expression (3.2) is a particular solution of the time-dependent Schrödinger
Eq. (3.1), the general solution is a linear combination of particular solutions of the
form (3.2). The dependence on time for wð~ r; tÞ is the elementary function, but its
dependence on the coordinates has not always an elementary form, since the
Eq. (3.3) with one form of the potential function /ð~ r Þ is quite different from the
same equation for its other. In fact, the Eq. (3.3) can be solved analytically only for
a small number of special types of functions /ð~ r Þ. The interpretation of the value E
in the Eq. (3.2) is very important. The time dependence of the function wð~ r; tÞ in
(3.2) is exponential, and the coefficient before t in the upper index of exponent
function is selected so that the right side of (3.3) contains only a constant factor E.
In the left part of (3.3) the function wð~ r Þ is multiplied by a function of potential
energy /ð~ r Þ. Consequently, the value E should have the dimension of energy. The
one value with a dimension of energy that is constant in mechanics is the total
energy of the system. According to the physical interpretation of the Schrödinger
equation E is the total energy of the particle motion, described by the function
wð~r; tÞ.
The function of a plane wave can be represented as A exp½iðkz  xtÞ, where
A—the amplitude; k—the wave number; x ¼ 2pm ¼ Eh —the angular frequency; m—
the frequency.
Let us call the sum of the incoming and reflected waves as the incident wave.
Then, the total wave function w is a sum of functions of incident winc and wscatt
waves (Fig. 3.1):

w ¼ winc þ wscatt : ð3:4Þ

During an elastic collision, the wave energy does not change in time, so you can
only consider only that part of the incident wave function that depends on the
coordinates in the space.

winc ðr; hÞ ¼ A expðikzÞ: ð3:5Þ

The scattered wave is a spherical wave that has the asymptotic form far from the
center of scattering:

A
wscatt ðr; hÞ  f ðhÞ expðikr Þ; ð3:6Þ
r

Here f ðhÞ—the scattering amplitude that is related to the differential cross section
by the equation
3.1 The Schrödinger Equation 73

Fig. 3.1 To the description of the wave–particle scattering

rðhÞ ¼ jf ðhÞj2 : ð3:7Þ

Thus, the total wave function can be written as


expðikrÞ
w ¼ A expðikzÞ þ f ðhÞ : ð3:8Þ
r

The amplitude A is often assumed to be equal to unity.

The quantum mechanical theory of particles scattering based on the solution of
the Schrödinger wave Eq. (3.1) or (3.3) in view of the potential energy /ð~ r Þ of
interaction between the particles. In practice, quantum mechanical calculation of the
particles scattering is determined as follows:
(1) Functions of incident and scattered waves;
(2) Total wave function;
(3) Phase shift;
(4) Amplitude of the scattered wave;
(5) Differential cross sections for the particles interaction;
(6) The total cross section of collisions between particles.
According to the Heisenberg uncertainty principle DxDp  h2, here Dx and Dp—
the standard deviations of the measured values of the position and momentum, in
quantum theory both the position and velocity of the particles, i.e., the angle of the
particles scattering after the collision, cannot be accurately determined. Quantum
mechanical description allows getting only the probability of scattering (i.e., the
differential cross sections) of the particles at a certain angle h. This probability is
usually expressed using the phase shift gl of the radial wave function wð~ r Þ. The
phase shift gl depends on the wave number k and function U ðr Þ, that are defined as
74 3 Quantum Mechanical Theory of the Particle Scattering

pffiffiffiffiffiffiffiffiffi
2p p 2lE 2l/ðr Þ
k¼ ¼ ¼ ; U ðr Þ ¼ ; ð3:9Þ
k h h h2

Here p, μ, E—the momentum, the reduced mass, and the total energy of the system,
respectively.
Equation (3.3) can be written as
 
r Þ þ k2  U ð~
Dwð~ r Þ wð~
r Þ ¼ 0: ð3:10Þ

In a spherical coordinate system with the symmetry (3.10) has the form
   
1 @ 2 @w 1 @ @w

r þ 2 sin h þ k2  U w ¼ 0: ð3:11Þ
r @r
2 @r r sin h @h @h

3.2 Solution of the Schrödinger Equation for the Elastic

Interactions

The solution of the wave Eq. (3.11) is carried out using the method of variables
separation. For this the wave function wðr; hÞ is represented as a result of multi-
plication of two independent one from the other functions, one of them depends
only on the coordinates r, and the other—from h

wðr; hÞ ¼ Lðr ÞY ðhÞ; ð3:12Þ

Here Lðr Þ—the radial function that depends on the coordinates Y ðhÞ—the spherical
harmonic function.
Substituting (3.12) into (3.11), we obtain
 
 
1 d dL   1 1 d dy
r2 þ r 2 k2  UðrÞ ¼  sin h : ð3:13Þ
L dr dr Y sin h dh dh

The left side of (3.13) depends only on r, and the right—on h. Both parts of the
equation must be equal to each other r и h, each of them must be equal to a
constant, for example lðl þ 1Þ, where l—the orbital quantum number.
 

d 2 dL lðl þ 1Þ
r þ r k  UðrÞ 
2 2
L ¼ 0; ð3:14Þ
dr dr r2
 
1 d dy
sin h þ lðl þ 1ÞY ¼ 0: ð3:15Þ
sin h dh dh
3.2 Solution of the Schrödinger Equation for the Elastic Interactions 75

Equation (3.15) is the Legendre equation of the second order and has two lin-
early independent solutions, each of them can be written as the sum of the power
series for cos h. Both solutions are equal to infinity (i.e., unacceptable) for h ¼ 0, if
l is not zero or a positive integer number. Taking into account (3.12), the general
solution of Eq. (3.11) in accordance with the cylindrical symmetry of the problem
and the physical condition that the wave functions are finite at r = 0 can be written
in the form [1]:

X
1
wðr; hÞ ¼ Al Ll ðrÞPl ðcos hÞ: ð3:16Þ
l¼0

Here Al —is a constant; Ll ðrÞ—the solution of (3.14); functions Pl ðcos hÞ—the

Legendre polynomial, is a solution of Legendre Eq. (3.15).
The solution of Legendre equation is as usual

Y ¼ a0 Fðcos hÞ þ a1 Gðcos hÞ; ð3:17Þ

Here the function F contains only even degrees of cos h, and G—only odd degrees,
a0 and a1—some constants. To simplify (3.17) instead of F and G we write the
function, given that for l = 0, 2, 4, … constant a0 corresponds to it, and for l = 1, 3,
5, …—constant a1.
The expression for Ll ðrÞ is represented as

Gl ðrÞ
Ll ðrÞ ¼ : ð3:18Þ
kr

Substituting (3.18) into (3.14), we obtain the equation


d 2 Gl ðrÞ lðl þ 1Þ
þ k 2
 UðrÞ  Gl ðrÞ ¼ 0: ð3:19Þ
dr 2 r2

At large values of r the function UðrÞ tends to zero as the term containing 1/r2.
Consequently, the asymptotic solution has the form:

Gl ðrÞ  sinðkr þ Cl Þ; ð3:20Þ

Here Cl —is a constant, that is determined by the boundary condition that the wave
function of the relative motion must vanish when the distance between the particles
r, is greater then rmax, that subsequently assumed to be infinity.
At large values of r the expression (3.14) has an asymptotic solution
 
1 pl
Ll ðrÞ  sin kr  þ gl ; ð3:21Þ
kr 2
76 3 Quantum Mechanical Theory of the Particle Scattering

Here gl —the phase shift of the lth partial wave due to the influence of the potential
energy /ðr Þ.
In the ideal case of noninteracting particles /ðr Þ ¼ 0, gl ¼ 0, i.e., there is no
phase shift if there is no scattering. The phase shift depends on the orbital quantum
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
numbers l (or angular momentum h lðl þ 1Þ) and k (or momentum of the relative
motion hk). You can draw an analogy with the classical case when the scattering
angle h is calculated as a function of the impact parameter b, and the relative
collision velocity g.
Substituting Eq. (3.21) to (3.16), we obtain an asymptotic expression for the
total wave

X
1  
1 pl
wðr; hÞ ’ Al sin kr  þ gl Pl ðcos hÞ: ð3:22Þ
l¼0
kr 2

To determine the constant Al the full wave function should be represented as a

sum of functions of incident winc and scattered wscatt waves. For the incident wave
propagating along the axis z, we have

winc ðr; hÞ ¼ expðikzÞ ¼ expðikr cos hÞ: ð3:23Þ

Expression (3.23) can be expanded in a series of Legendre polynomials;

expansion coefficients contain Bessel functions
rffiffiffiffiffiffiffi
X
1
p
winc ðr; hÞ ¼ ð2l þ 1Þi Pl ðcos hÞ
l
Jl þ 1=2 ðkrÞ
l¼0
2kr
ð3:24Þ
X
1
¼ ð2l þ 1Þi Pl ðcos hÞjl ðkrÞ;
l

l¼0

Here jl ðkrÞ—the spherical Bessel function that has the asymptotic solution for large
values of r the following:
 
1 pl
jl ðkrÞ  sin kr  : ð3:25Þ
kr 2

Thus, the asymptotic form of the incident wave at large r is expressed as

X
1  
1 pl
winc ðr; hÞ  ð2l þ 1Þil Pl ðcos hÞ sin kr  : ð3:26Þ
l¼0
kr 2

Substituting (3.26) into (3.22), we find the expression for the asymptotic form of
the scattered wave:
3.2 Solution of the Schrödinger Equation for the Elastic Interactions 77

wscatt ðr; hÞ ¼ wðr; hÞ  winc ðr; hÞ

X1
   
1 pl pl
 Pl ðcos hÞ Al sin kr  þ gl  ð2l þ 1Þi sin kr 
l
:
l¼0
kr 2 2
ð3:27Þ

To determine the coefficient Al, given that sinðxÞ ¼ expðixÞexpðixÞ

2i the latter
expression can be transformed as follows:

X1
1 1
wscatt ðr; hÞ  Pl ðcos hÞ
l¼0
kr 2i

 
pl  
ð3:28Þ
 exp i kr  Al expðigl Þ  ð2l þ 1Þil
2

  
pl  
 exp i kr  Al expðigl Þ  ð2l þ 1Þi : l
2

Here, according to the physical meaning, the term with the exponential function
corresponds to a negative power of the incident wave and the exponential term with a
positive power—reflected. Since there is no term in the expression for the function of
the incident wave, the second term (in square brackets) must be equal to zero, i.e.,

Al expðigl Þ  ð2l þ 1Þil ¼ 0;

ð3:29Þ
Al ¼ ð2l þ 1Þil expðigl Þ:

Thus, we have obtained the expression for the coefficient Al. Substituting it into
(3.22), we obtain an expression for the total wave function

X
1  
1 pl
wðr; hÞ ¼ ð2l þ 1Þi expðigl Þ sin kr  þ gl Pl ðcos hÞ:
l
ð3:30Þ
l¼0
kr 2

Equation (3.27) after the substitution (3.29) takes the form

X1
1
wscatt ðr; hÞ  Pl ðcos hÞð2l þ 1Þil
l¼0
kr

   
pl pl
 expðigl Þ sin kr  þ gl  sin kr  ð3:31Þ
2 2
expðikr Þ X 1
1
¼ ð2l þ 1Þ½expð2igl Þ  1Pl ðcos hÞ
kr l¼0
2i

The expression for the scattering cross section can also be obtained by using
(3.6) and (3.31). It will look like
78 3 Quantum Mechanical Theory of the Particle Scattering

expðikr Þ expðikr Þ X1
1
f ðhÞ ¼ ð2l þ 1Þ½expð2igl Þ  1Pl ðcos hÞ: ð3:32Þ
r kr l¼0
2i

Consequently,

1X 1
1
f ðhÞ ¼ ð2l þ 1Þ½expð2igl Þ  1Pl ðcos hÞ
k l¼0 2i
ð3:33Þ
1X 1
¼ ð2l þ 1Þ expðigl Þ sin gl Pl ðcos hÞ:
k l¼0

3.3 Determination of the Phase Shift

The phase shift can be used to determine the differential cross section in the fol-
lowing form:
2
dr 1 X1

¼ 2 ð2l þ 1Þ expðigl Þ sin gl Pl ðcos hÞ : ð3:34Þ
dX k l¼0

According to the classical theory of angular momentum, the particle in the

collision depends on the pa, where p—momentum of the center of mass of the
system; a—the range of the radius, or “black disc,” where the wave is completely
absorbed. In this way
pa
l ¼ ka ¼ L: ð3:35Þ
h

The phase shift can be found by using the equation for the calculation of the
scattered waves and full waves [2, 3]:

kj0l ðkaÞjl ðk1 aÞ  k1 jl ðkaÞj0l ðk1 aÞ

tggl ðkÞ ¼ ; ð3:36Þ
kn0l ðkaÞjl ðk1 aÞ  knl ðkaÞj0l ðk1 aÞ

where jl —Bessel function; nl —Newman function j0l , n0l —derivatives of the argu-
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2mðE þ /0 Þ
ment of the Bessel and Newman functions; k1 ¼ h —the wave number
based on potential field; /0 —the depth of the potential barrier (/0 [ 0 for attractive
potential).
For the scattering cross section of particles using the modified effective range
theory (MERT) phase shifts for the first time were used in the study of neutron–
neutron and neutron–proton scattering [1, 4]. Processing scattering of particles is
usually made by comparing the phase shifts that were determined from the
3.3 Determination of the Phase Shift 79

experimental interaction cross sections and theoretical calculations using nuclear

potentials. It was necessary to do a lot of calculations for it and the semiempirical
equations for a given nuclear capability were introduced.
The experimental data are used to find some characteristics of the nuclear
potential, semiempirical methods allow relatively easy to determine which char-
acteristics of nuclear capability can be obtained from experimental data. The values
of the phase shifts change with the energy of the incident particle for each nuclear
capability. The most famous potentials are square, Gauss, exponential, and
Yukawa. Energies of nuclear potentials are expressed as a function of the depth of
the potential well and the inner distance, defining the boundaries of existence
potential. The internal distance cannot be found by direct measurement, it must be
obtained from the experimental data by entering some values of the active length r0
and scattering length A. The active length also depends on the depth of the potential
pit and decreases with its increase. The relative values of the current domestic and
distances depend on the accepted form of the potential well. At low energies in the
scattering process the s-waves are dominated.
The basis of the approach using the existing distance lies in the formation of the
assumptions of the phase shifts functions in addiction of the energy in the form of
simple functions, using the most common assumptions about nuclear potentials.
These functions include some uncertain parameters. This allows the comparison of
theoretical and experimental results that are not obtained by phase shifts in the
parameters and functions that reduces the amount of computation required. Thus,
for a given potential energy it is necessary to find three parameters instead of a large
number of phase shifts. The nature of the scattering is determined by the parameters
of the functions found. To obtain the functional form of the phase shifts and errors
the method of least squares is used. Designating the phase shift g and the wave
number k, we can write the following expression for kg:

1 1 2
kctg g ¼  þ r0 k ; ð3:37Þ
A 2

Here r0 —the active length; A—the scattering length obtained for zero energy; factor
1/2 is introduced because the function value begins to decrease near the edge of the
potential pit.
Equation (3.37) contains only two parameters—the length of the scattering and
the given length. It is known as independent of the form of the potential approx-
imation relations, since in any case you can choose these parameters so that the
form of the function fit for any shape, depth, and size of the potential pit.
The scattering cross section at zero energy is defined as Q0 ¼ 4pA2 . For
potentials acting on a short distance, marking the angular momentum L, the
Eq. (3.37) can be written as

1 1
k2L þ 1 ctg gðLÞ ¼  þ r0 ðLÞk2 : ð3:38Þ
AðLÞ 2
80 3 Quantum Mechanical Theory of the Particle Scattering

Note that (3.37) is independent of the potential type and gives similar results for
the scattering of any type. The utility of the MERT theory lies in the fact that the
scattering amplitude for a range of energy with a certain accuracy, expressed in
terms of a small number of parameters. To obtain such experimental parameters
points that lie in the area can be used. For a scattering neutron–proton of the order
of 10‒15 m at a certain value r0, and we can assume that the approximation that does
not depend on the type of potential, it can be used at energies up to 10 meV.
There are strict limits on the use of MERT theory to the nuclear field to the field
of atomic scattering. The modified effective range theory that is briefly presented
above is ill-suited to describe of the electron–atom scattering, since the forces are
proportional to 1=r 4 at the long distances. Nevertheless, it managed to change the
way that it gave acceptable results for the scattering of electrons by neutral particles
that can be polarized. Coefficients used in it are chosen that the resultant values
agree with the experimental sections in the lower energy range (several electron
volts), including the results of Ramseur–Collat experiments. Next there is the
extrapolation to zero energy.
For the s-wave can apply Eq. (3.38) with l = 0, at sufficiently low energy the
follow approximation is used [4–6]:
pa 2
tg g0  g0  Ak  k ; ð3:39Þ
3a0
pa 2
tg g1 ¼ k  A1 k 3 ðl ¼ 1Þ; ð3:40Þ
15a0

pak2
tg gl ¼ ðl [ 1Þ: ð3:41Þ
ð2l þ 3Þð2l þ 1Þð2l  1Þa0

If the value A is positive, such as helium and neon, the phase shift will be in the
form of a monotonically decreasing function. If the value of A is negative, the phase
shift will initially increase by a positive factor p at zero energy to a maximum value,
and then decrease, passing a value of zero when A  ðp=3a0 Þak. The scattering
length in Ramseur minimum can be found from the last equation. To obtain A it is
necessary to have data points that lie in the region of the minimum sections. For the
momentum transfer cross sections of some inert gas the Ramseur–Townsends effect
is also observed, and the minimum of the cross section of the energy can be
determined in this way. In this case it is necessary to take into account an additional
factor g1 even for the s-waves as opposed to total scattering cross section, where
this ratio is only present in the case of p-waves. Buckman and Lohmamn [5] have
analyzed the scattering cross sections by using MERT theory data, but in their
approach, they used two additional members with k of a higher degree
 
4a 2 pa 2
tg g0 ¼ Ak 1 þ k lnðka0 Þ  k þ Dk 3 þ Fk 4 ðl ¼ 0Þ: ð3:42Þ
3a0 3a0
3.3 Determination of the Phase Shift 81

Table 3.1 The coefficients for the approximation MERT4

Parameter References
[10] [5] [6] [7] Author
A −1.609 a0 −1.442 a0 −1.436 a0 −1.492 a0 −1.55 a0
A1 8.094 a30 9.229 a30 8.57 a30 9.104 a30 7.8 a30
D 60.52 a30 70.784 a30 75.2 a30 64.3 a30 61.5 a30
F −59.47 a40 −97.188 a40 −103.9 a40 −77.3 a40 −71.4 a40

For the p-waves (l = 1) and waves of higher degree, the equations have the same
form as (3.40), (3.41). In the analysis of Backman and Lohman [5] it was noted that
the limit of the application of MERT theory at low energies has been set in the
determination of the scattering length for comparison of the relative values of QT
and QM . The phase shifts for MERT theory (l [ 0) in the electron–atom scattering
[4–7] are
pa 2
tg g1 ¼ k  A1 k 3 ðl ¼ 1Þ; ð3:43Þ
15a0

pak2
tg gl ¼ ðl [ 1Þ; ð3:44Þ
ð2l þ 3Þð2l þ 1Þð2l  1Þa0

Here a0—Bohr radius, a—the dipole polarizability of atom, a ¼ 11; 08a30 for
Argon; k—number of the wave associated with the energy E, eV by the equation
E ¼ 13; 6057ðka0 Þ2 ; A, A1, D, and F—the parameters, that should be determined
for a correspondence between theoretical and experimental data. On the basis of the
physical meaning, the parameter A is of greatest interest, which is the expression for
the cross section at zero energy. Data for A, A1, D, and F are shown in Table 3.1.
Cross sections depending on the phase shifts are determined by summing of an
infinite series of higher order partial waves [8, 9].

3.4 Born Approximation for Calculating the Amplitudes

of the Scattered Waves

A finite number of partial waves make the contribution to the scattering amplitude
in the communication of the scattering and incident particles over a short distance.
Their number increases with increasing collision energy. If the particle interaction is
essential at a great distance, theoretically to obtain accurate values of the scattering
amplitude it is necessary to take into account the large number of partial waves. In
practice, the phase shifts of the first few partial waves l can be calculated accurately,
using the asymptotic radial part of the wave function and phase shifts of partial
82 3 Quantum Mechanical Theory of the Particle Scattering

waves of higher orders (from l + 1 to infinity), they are estimated using the
approximation of Born using only interaction at large distances.

Z1  
ð1Þ l 1 h
f ðhÞ ¼  2 r/ðrÞ sin 2rk sin dr: ð3:45Þ
h k sin h2 2
0

As a test of the Born approximation, the calculation of differential scattering

cross section of charged particles in the plasma will be considered using the
function of potential energy based screening

Zi Zj e2 expðr=rD Þ
/ðrÞ ¼ ; ð3:46Þ
4pe0 r

Here rD —the Debye radius.

Note that /ðrÞ reduces to zero very quickly when r  rD . We have

Zi Zj e 2
ð1Þ
2l 4pe0 1
f ðhÞ ¼   2 : ð3:47Þ
h 2
2k2 ð1  cos hÞ þ 1
rD

Thus, for the first (l = 1) Born approximation the differential cross section with
Yukawa potential function is
0 12
  Zi Zj e 2
dr 2l 1
’@
4pe0 A

 2 2
dX h 2
 cos hÞ þ2k2 ð1 1
rD
!2
Zi Zj e2 =ð4pe0 Þ
¼
; ð3:48Þ
4E sin2 ðh=2Þ þ h2 =2mrD2

where the total energy of the system E ¼ k2lh (3.9).

2 2

In the absence of screening (i.e., rD ! 1), this expression reduces to the

well-known formula of Rutherford.
Consider the elastic scattering of a particle of mass μ with energy E ¼ h2 k 2 =ð2lÞ
using a function /ðrÞ. The Thompson expression (for the partial waves) for the
elastic scattering amplitude has the form of

1X 1
1X L
f ðhÞ ¼ ð2l þ 1ÞTl Pl ðcos hÞ ¼ ð2l þ 1ÞTl Pl ðcos hÞ þ DfL ðhÞ; ð3:49Þ
k l¼0 k l¼0

Here the elements of the transition matrix Tl are associated with a phase shift gl by
ratio
3.4 Born Approximation for Calculating the Amplitudes … 83

Tl ¼ expðigl Þ sinðgl Þ: ð3:50Þ

If the values Tl for l [ L are approximated by TlB, that can be found by using of
the Born approximation, we can write

1 X 1
1X L
DfL ðhÞ ¼ ð2l þ 1ÞTlB Pl ðcos hÞ ¼ fB ðhÞ  ð2l þ 1ÞTlB Pl ðcos hÞ:
k l¼L þ 1 k l¼0
ð3:51Þ

Here fB ðhÞ—the total scattering amplitude in the Born approximation,

Z1 X
2l 1
1X 1
fB ðhÞ ¼  2 /ðrÞr 2 ð2l þ 1Þj2l ðkrÞPl ðcos hÞdr ¼ ð2l þ 1ÞTlB Pl ðcos hÞ;

h l¼0
k l¼0
0
ð3:52Þ

Consequently,

Z1
2lk
TlB ¼ 2 /ðrÞr 2 j2l ðkrÞdr: ð3:53Þ
h
0

The long distance interaction only contributes to the phase shifts of higher partial
waves, so it is reasonable to take for /ðrÞ the function of the form /ðrÞ ¼ C=r 4 ,
here C—is a constant. Then we get
!
pak 1 1 h X L
Pl ðcos hÞ
DfL ðhÞ ¼  sin  : ð3:54Þ
a0 3 2 2 l¼1 ð2l þ 3Þð2l  1Þ

Note that the upper limit of summation in (3.51) is L but not infinity, then we
have

1X L
f ðhÞ ¼ ð2l þ 1Þ expðigl Þ sin gl Pl ðcos hÞ
k l¼0
" # ð3:55Þ
pak 1 1 h X L
Pl ðcos hÞ
þ  sin  :
a0 3 2 2 l¼1 ð2l þ 3Þð2l  1Þ

The resulting expression for f ðhÞ allows to find the averaged interaction cross
ðL;SÞ
section Qei (Fig. 3.2).
84 3 Quantum Mechanical Theory of the Particle Scattering

Fig. 3.2 The average effective cross sections of charged particles interaction in an argon plasma,
depending on the temperature: a, b the interaction of electrons with single ions at l = 1 and 6; c,
d the same with a double ions; e, f the same with the three time ions
3.5 Determination of Differential and Total Cross Sections … 85

3.5 Determination of Differential and Total Cross Sections

of Elastic Interactions

In general, the differential cross section can be defined as

dr
r¼ ¼ jf ðh; uÞj2 : ð3:56Þ
dX

At the same time, if the potential depends on the spin, then there may be some
azimuthal function u. The physical meaning of the differential cross section can be
shown as follows:

dr A
dX¼ ; ð3:57Þ
dX B

Here A—the amount of particles dispersed in a solid angle dΩ per unit time;
B—amount of particles passing through a unit area per unit time.
The total cross section is defined as:
Z
dr
rtot ðkÞ ¼ dX: ð3:58Þ
dX

Substituting the expression for ddXr into (3.58), we have

Zp ( )
1 X
rtot ðkÞ ¼ 2p ð2l þ 1Þ exp½igl ðkÞ sin gl ðkÞPl ðcos hÞ
k2 l
(
0
) ð3:59Þ
X  
 ð2l0 þ 1Þ exp igl0 ðkÞ sin gl0 ðkÞPl0 ðcos hÞ sin hdh;
l0

Rp
Using the equation Pl ðcos hÞPl0 ðcos hÞ sinðhÞdh ¼ 2l 2þ 1 dll , where dll —the
0
Kronecker function, we obtain

4p X1
rtot ¼ ð2l þ 1Þ sin2 gl ðkÞ: ð3:60Þ
k2 l¼0

The amplitude at zero scattering angle f(0) is associated with a full cross section
86 3 Quantum Mechanical Theory of the Particle Scattering

Fig. 3.3 The differential elastic scattering cross section for the scattering of electrons by atoms in
argon for different scattering angles: points—the experimental data, curves—the designed data:
filled circle—[12]; filled triangle—[16]; filled square—[17]; space with dash lines—[5]; doted
lines—[6]; dash with doted line—[7]; solid line—results of the author
3.5 Determination of Differential and Total Cross Sections … 87

Fig. 3.3 (continued)

1X 1
Imf ð0Þ ¼ ð2l þ 1ÞIm½expðigl ðkÞÞ sin gl ðkÞPl ð1Þ
k l¼0
ð3:61Þ
1X 1
k
¼ ð2l þ 1Þ sin gl ðkÞ ¼ rtot :
k l¼0 4p

This ratio (the optical theorem) holds even for inelastic processes such as
scattering processes in nuclear physics and particle physics. The form of this
relationship is consistent with the following fact: the total scattering cross section
reflects the decrease in the flow of the incident beam. This phenomenon can be only
as a result of destructive interference that, in turn, can only occur between the
incoming wave and the reflected elastic wave propagating further. This explains the
linear form of f(0) [2, 9].
In our case, the spin effects are not taken into account, these effects may be
relevant for the heavy inert gases. The results of the differential cross sections
calculations for scattering of electrons by atoms in argon using the fitting procedure
MERT4 are generally in good agreement with the experimental data [7, 11–14]
(Fig. 3.3). The author has calculated the scattering angles (θ = 20–100°). In the
calculations for higher energies it is necessary to use a large number of members
ð1Þ ð6Þ
(partial waves). We give expression for QB ðEÞ—QB ðEÞ [8, 15]:
88 3 Quantum Mechanical Theory of the Particle Scattering

9
ð1Þ 4p X
1
>
>
QB ðEÞ ¼ 2 ðl þ 1Þ sin2 ðgl  gl þ 1 Þ; >
>
k l¼0 >
>
>
>
>
>
ð2Þ 4p X 1
ðl þ 1Þðl þ 2Þ 2 >
>
>
QB ðEÞ ¼ sin ðgl  gl þ 2 Þ; >
>
k l¼0 ð2l þ 3Þ
2 >
>
>
>
1
>
>
ð3Þ 4p X 3ðl þ 1Þðl2 þ 2l  1Þ 2 >
>
QB ðEÞ ¼ 2 sin ðgl  gl þ 1 Þ >
>
k l¼0 ð2l  1Þð2l þ 5Þ >
>
>
>
 >
>
ðl þ 1Þðl þ 2Þðl þ 3Þ 2 >
>
þ sin ðgl  gl þ 3 Þ ; >
>
ð2l þ 3Þð2l þ 5Þ >
>
>
>
X1
>
>
>
ð4Þ 4p 2ðl þ 1Þðl þ 2Þð2l2 þ 6l  3Þ 2 >
>
QB ðEÞ ¼ 2 sin ðgl  gl þ 2 Þ >
>
k l¼0 ð2l  1Þð2l þ 3Þð2l þ 7Þ >
>
 >
>
>
>
ðl þ 1Þðl þ 2Þðl þ 3Þðl þ 4Þ 2 >
>
þ sin ðgl  gl þ 4 Þ ; =
ð2l þ 3Þð2l þ 5Þð2l þ 7Þ
X
>
4p 1 5ðl þ 1Þð2l4 þ 8l3  4l2  24l þ 9Þ 2 >
>
ð5Þ
QB ðEÞ ¼ 2 sin ðgl  gl þ 1 Þ >
>
k l¼0 ð2l  3Þð2l  1Þð2l þ 5Þð2l þ 7Þ >
>
>
>
>
>
5ðl þ 1Þðl þ 2Þðl þ 3Þðl þ 4l  2Þ 2
2 >
>
þ sin ðgl  gl þ 3 Þ >
>
>
>
ð2l  1Þð2l þ 3Þð2l þ 5Þð2l þ 9Þ >
>
 >
>
ðl þ 1Þðl þ 2Þðl þ 3Þðl þ 4Þðl þ 5Þ 2 >
>
þ sin ðgl  gl þ 5 Þ ; >
>
ð2l þ 3Þð2l þ 5Þð2l þ 7Þð2l þ 9Þ >
>
>
>

>
>
4p X1
15ðl þ 1Þðl þ 2Þðl 4
þ 6l3
þ l 2
 24l þ 9Þ >
>
sin ðgl  gl þ 2 Þ >
ð6Þ
QB ðEÞ ¼ 2 2
>
>
k l¼0 ð2l  3Þð2l  1Þð2l þ 3Þð2l þ 7Þð2l þ 9Þ >
>
>
>
>
>
3ðl þ 1Þðl þ 2Þðl þ 3Þðl þ 4Þð2l þ 10l  5Þ 2
2 >
>
þ sin ðgl  gl þ 4 Þ > >
>
ð2l  1Þð2l þ 3Þð2l þ 5Þð2l þ 7Þð2l þ 11Þ >
>
 >
>
ðl þ 1Þðl þ 2Þðl þ 3Þðl þ 4Þðl þ 5Þðl þ 6Þ >
>
þ sin ðgl  gl þ 6 Þ : >
2 ;
ð2l þ 3Þð2l þ 5Þð2l þ 7Þð2l þ 9Þð2l þ 11Þ
ð3:62Þ

Total scattering cross section is determined by the equation

4p X 1
QT ðEÞ ¼ ð2l þ 1Þ sin2 gl : ð3:63Þ
k 2 l¼0

ð1Þ
The cross section of the pulse transmission QM ðEÞ ¼ QB ðEÞ.
At very low energies (e\0:1 eV) the partial waves of lower orders make the
main contribution to the scattering amplitude. Next (at 0:1\e\1 eV) it should take
into account the contribution of higher order partial waves. They give a dip in the
scattering cross section of particles that is a manifestation of the Ramseur–
Townsend observed experimentally (Table 3.2). At the same time, measurements of
the differential cross section of elastic scattering in crossed beams can in principle
 2
Table 3.2 The differential cross sections for elastic scattering of electrons by atoms in Argon, Å sr−1 [16]
Scattering The electron energy E (eV)
angle θ (o) 0.1 0.5 1 3 5 7.5 10 15 20
0 5.4888 × 10−1 3.304 × 10−1 2.3103 × 10−1 4.761 × 10−1 1.826 6.0209 1.201 × 10 1.593 × 10 1.327 × 10
5 5.0127 × 10−1 2.529 × 10−1 1.4338 × 10−1 2.968 × 10−1 1.372 5.0407 1.064 × 10 1.481 × 10 1.227 × 10
10 4.5367 × 10−1 1.859 × 10−1 7.9252 × 10−2 1.815 × 10−1 1.017 4.1446 9.241 1.353 × 10 1.120 × 10
15 4.0886 × 10−1 1.313 × 10−1 3.6685 × 10−2 1.269 × 10−1 7.617 × 10−1 3.3325 7.841 1.210 × 10 1.011 × 10
20 3.6405 × 10−1 8.821 × 10−2 1.2294 × 10−2 1.224 × 10−1 5.993 × 10−1 2.6604 6.497 1.053 × 10 8.961
25 3.1925 × 10−1 5.461 × 10−2 3.0244 × 10−3 1.582 × 10−1 5.209 × 10−1 2.1087 5.237 8.877 7.729
30 2.7640 × 10−1 2.996 × 10−2 6.0769 × 10−3 2.246 × 10−1 5.181 × 10−1 1.6915 4.145 7.253 6.441
35 2.3467 × 10−1 1.330 × 10−2 1.8819 × 10−2 3.136 × 10−1 5.713 × 10−1 1.4086 3.192 5.713 5.181
40 1.9687 × 10−1 3.837 × 10−3 3.8646 × 10−2 4.117 × 10−1 6.721 × 10−1 1.2546 2.476 4.285 3.977
45 1.6466 × 10−1 2.487 × 10−4 6.2449 × 10−2 5.125 × 10−1 8.009 × 10−1 1.207 1.946 3.080 2.884
50 1.3890 × 10−1 9.997 × 10−4 8.7373 × 10−2 6.049 × 10−1 9.381 × 10−1 1.2406 1.602 2.086 1.915
55 1.1930 × 10−1 4.537 × 10−3 1.1118 × 10−1 6.861 × 10−1 1.070 1.3218 1.411 1.333 1.143
60 1.0446 × 10−1 9.577 × 10−3 1.3162 × 10−1 7.421 × 10−1 1.173 1.4226 1.333 8.177 × 10−1 5.769 × 10−1
65 9.2974 × 10−2 1.515 × 10−2 1.4842 × 10−1 7.785 × 10−1 1.243 1.5094 1.325 5.125 × 10−1 2.249 × 10−1
70 8.3732 × 10−2 2.084 × 10−2 1.6074 × 10−1 7.869 × 10−1 1.269 1.557 1.341 3.809 × 10−1 6.441 × 10−2
3.5 Determination of Differential and Total Cross Sections …

75 7.5051 × 10−2 2.638 × 10−2 1.6915 × 10−1 7.701 × 10−1 1.243 1.5458 1.344 3.725 × 10−1 6.105 × 10−2
80 6.6370 × 10−2 3.192 × 10−2 1.7363 × 10−1 7.337 × 10−1 1.176 1.473 1.305 4.341 × 10−1 1.669 × 10−1
85 5.7969 × 10−2 3.697 × 10−2 1.7503 × 10−1 6.749 × 10−1 1.067 1.3358 1.207 5.153 × 10−1 3.276 × 10−1
90 4.9567 × 10−2 4.173 × 10−2 1.7279 × 10−1 6.049 × 10−1 9.269 × 10−1 1.1398 1.045 5.769 × 10−1 4.957 × 10−1
95 4.2286 × 10−2 4.593 × 10−2 1.6774 × 10−1 5.209 × 10−1 7.645 × 10−1 9.0733 × 10−1 8.317 × 10−1 5.937 × 10−1 6.273 × 10−1
100 3.6405 × 10−2 4.873 × 10−2 1.5934 × 10−1 4.313 × 10−1 5.909 × 10−1 6.553 × 10−1 5.853 × 10−1 5.545 × 10−1 7.001 × 10−1
105 3.1645 × 10−2 5.041 × 10−2 1.4842 × 10−1 3.417 × 10−1 4.257 × 10−1 4.1726 × 10−1 3.389 × 10−1 4.705 × 10−1 7.029 × 10−1
110 2.8284 × 10−2 5.097 × 10−2 1.3526 × 10−1 2.588 × 10−1 2.800 × 10−1 2.1619 × 10−1 1.375 × 10−1 3.613 × 10−1 6.441 × 10−1
(continued)
89
Table 3.2 (continued)
90

Scattering The electron energy E (eV)

angle θ (o) 0.1 0.5 1 3 5 7.5 10 15 20
115 2.5932 × 10−2 5.041 × 10−2 1.2126 × 10−1 1.851 × 10−1 1.686 × 10−1 8.4853 × 10−2 1.728 × 10−2 2.677 × 10−1 5.433 × 10−1
120 2.3804 × 10−2 4.929 × 10−2 1.0698 × 10−1 1.257 × 10−1 9.969 × 10−2 4.5367 × 10−2 1.683 × 10−2 2.252 × 10−1 4.285 × 10−1
125 2.1563 × 10−2 4.817 × 10−2 9.3534 × 10−2 8.289 × 10−2 8.037 × 10−2 1.1314 × 10−1 1.641 × 10−1 2.722 × 10−1 3.332 × 10−1
130 1.9043 × 10−2 4.733 × 10−2 8.1492 × 10−2 5.797 × 10−2 1.129 × 10−1 2.9404 × 10−1 4.733 × 10−1 4.397 × 10−1 2.940 × 10−1
135 1.6354 × 10−2 4.649 × 10−2 7.057 × 10−2 5.013 × 10−2 1.946 × 10−1 5.8529 × 10−1 9.465 × 10−1 7.449 × 10−1 3.304 × 10−1
140 1.3806 × 10−2 4.593 × 10−2 6.1329 × 10−2 5.741 × 10−2 3.192 × 10−1 9.7454 × 10−1 1.568 1.187 4.565 × 10−1
145 1.1762 × 10−2 4.509 × 10−2 5.2928 × 10−2 7.757 × 10−2 4.789 × 10−1 1.4366 2.305 1.750 6.777 × 10−1
150 1.0474 × 10−2 4.369 × 10−2 4.5647 × 10−2 1.07 × 10−1 6.581 × 10−1 1.9407 3.108 2.406 9.717 × 10−1
155 1.0025 × 10−2 4.201 × 10−2 3.9206 × 10−2 1.423 × 10−1 8.457 × 10−1 2.4588 3.949 3.108 1.319
160 1.0362 × 10−2 3.977 × 10−2 3.3885 × 10−2 1.792 × 10−1 1.025 2.9404 4.733 3.781 1.677
165 1.1314 × 10−2 3.725 × 10−2 2.9404 × 10−2 2.134 × 10−1 1.187 3.3885 5.405 4.397 2.013
170 1.2490 × 10−2 3.529 × 10−2 2.6184 × 10−2 2.414 × 10−1 1.311 3.7246 5.965 4.873 2.285
175 1.3498 × 10−2 3.389 × 10−2 2.4196 × 10−2 2.599 × 10−1 1.392 3.9206 6.301 5.181 2.462
180 1.3862 × 10−2 3.332 × 10−2 2.3523 × 10−2 2.660 × 10−1 1.420 4.0046 6.413 5.293 2.520
3 Quantum Mechanical Theory of the Particle Scattering
3.5 Determination of Differential and Total Cross Sections … 91

be used for the integral scattering cross sections that can also be compared with the
results of measurements of attenuation methods. Although this method is quite
attractive, there may be problems in extrapolating the cross sections for large and
small angles, where the experimental data cannot be obtained. This eventually leads
to the limited accuracy of finding the integral scattering cross sections.
In some cases, such as for light inert gases, methods for determining the phase
shift can be used for differential scattering cross sections by fitting the measured
angular distributions. These techniques can be used to extrapolate the differential
scattering cross sections and for immediate calculation of integral cross sections for
elastic scattering. This technique is mainly used for low energy, i.e., energies below
the first excitation level. Thus it is possible to limit the number of phases required
for approximation and avoid the complicated phase shifts.
The transport cross section QðLÞ is calculated based on the relative kinetic energy
in a collision e by the numerical integration of the differential cross sections for
elastic scattering of electrons by atoms in argon, using the method adopted by Rath
[18–20]:
At 0:02  e\1:0 eV the theoretical values of the collision cross sections are taken
from [21].
At 1:0  e  10:0 eV the experimental values given in [11, 22] are used.
At 10:0  e  100:0 eV the experimental values of [23] are applied for small and
theoretical values of [24–26]—for the large scattering angles.
The reliable source in the determination of the effective cross section of the
electron–atom interaction in low-temperature plasmas are experimental data. The
results of averaged cross section calculations of electron–atom in argon [22] for
l = 1 and l = 6, s = 1–11 are shown in Fig. 3.4.

Fig. 3.4 Results of the effective cross sections calculation of interaction of electron–atom in
Argon for orders l = 1 (a) and l = 6 (b)
92 3 Quantum Mechanical Theory of the Particle Scattering

References

1. Thurston M.O. Gaseous Electronics Theory and Practice. Canada: CRC Press, 2006.
2. Sakurai J.J., Napolitano J.J. Modern Quantum Mechanics. USA, 2010.
3. Gasiorowicz S. Quantum Physics. USA, 2003.
4. Buckman S.J., Mitroy J. // J. Phys. B. 1989. Vol. 22. P. 1365.
5. Buckman S.J., Lohmamn B. // J. Phys. B. 1987. Vol. 20. P. 5807.
6. Ferch J. et al. // J. Phys. B. 1985. Vol. 18. P. 967.
7. Haddad G.N., O’Malley T.F. // J. Phys. 1982. Vol. 35. P. 35.
8. Devoto R.S. Transport coefficients of partially ionized argon // Phys. Fluids. 1967. Vol. 10.
No. 2. PP. 354–364.
9. Devoto R.S. // Phys. Fluids. 1966. Vol. 9. P. 1230.
10. Weyhreter M. et al. // J. Phys. D Atoms, Molecules and Clusters. 1988. Vol. 7. P. 333.
11. Gibson C., Gulley R.J., Sullivan J.P. et al. // J. Phys. B. 1996. Vol. 29. P. 3177.
12. Weyhreter M. et al. // J. Phys. D Atoms, Molecules and Clusters. 1988. Vol. 7. P. 333.
13. Plenkiewicz B. et al. // J. Phys. Rev. A. 1988. Vol. 38. P. 4460.
14. Buckman S.J. // J. Phys. 1997. Vol. 50. P. 483.
15. Meeks F.R. et al. // J. Chem. phys. 1994. Vol. 100 (5).
16. Plenkiewicz B., Plenkiewicz P., Jay Gerin J.P. Pseudopotential calculations for elastic
scattering of slow electrons (0–20 eV) from noble gases. I. Argon.
17. Gibson J.C. et al. // J. Phys. B. 1996. Vol. 29. P. 3177.
18. Transport properties in a two temperature plasma: theory and application / Rat V. et al. //
Phys. Rev. E. 2001. Vol. 64.
19. Rat V., Andre P., Aubreton J. et al. // Phys. D: Applied Phys. 2001. Vol. 34. P. 2191–2204.
20. Transport coefficients including diffusion in a two-temperature argon plasma / V. Rat, et al. //
J. Phys. D: Appl. Phys. 2002. Vol. 35. P. 981–991.
21. Bell K.L., Scott N.S., Lennon M.A. // J. Phys. B. 1984. Vol. 17. P. 4757.
22. Bose T.K., Kannappan D. and Seeniraj R.V. // Warme Sioffubertragung. 1985. Vol. 19. P. 3.
23. Panajotovic R., Filipovic D., Marinkovic B. et al. // J. Phys. B. 1997. Vol. 30. P. 5877.
24. Nahar S.N. and Wadehra J.M. // Phys. Rev. A. 1987. Vol. 35. P. 2051.
25. Wadehra J.M., Nahara S.N. // Phys. Rev. A. 1987. Vol. 36. P. 1458.
26. Itikawa Y. At. Data Nucl. Data Tables. 1974. Vol. 14. No. 1.
Chapter 4
Determination of the Composition,
Thermodynamic Properties,
and Transport Coefficients on the Basis
of the Mean Free Path

Determination of composition and properties of plasma, including the calculation of

the thermodynamic properties and transport coefficients is always a key issue in
plasma physics, in calculation of plasma processes and plasma torches. There are a
large number of sources dedicated to this issue. Relatively simple equations can be
obtained based on the model of the mean free path of the particles.
The phenomena of diffusion, viscosity, and thermal conductivity are physically
similar and they are the transferring of some flows generated by the gradients of
concentration, temperature and velocity in gas. The diffusion is a mass transfer from
one region to another due to the presence of a concentration gradient. It can also be
the result of a temperature gradient (thermal diffusion), and energy transfer can be
the result of a concentration gradient (diffusion heat conductivity). Viscosity—the
momentum transfer through the gas due to the presence of a velocity gradient;
thermal conductivity—the transfer of thermal energy due to the presence of a
temperature gradient in gas. These processes are called as transport phenomena.
This chapter provides an analysis of theoretical methods and the basic method of
calculating the structure, thermodynamic properties, and transport coefficients of
equilibrium and nonequilibrium two-temperature plasma on the basis of the sub-
mission of the mean free path of the particles.

4.1 The Plasma Composition

The calculation of the plasma composition in the equilibrium ionization on the basis of
the action mass law is performed using the Saha equations. In the equations the speed
of ionization and recombination are the same, the source of birth and death of charged
particles are equal to zero, all the flows of transport and dissipation of particles are also
equal to zero. The right side of the Saha equations the electron temperature is usually
taken instead of the temperature of ionization. For a 5-component plasma with atoms,
one-, two-, and three-ions and electrons the Saha equation are
© Springer International Publishing Switzerland 2017 93
S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_4
94 4 Determination of the Composition, Thermodynamic Properties …



ne niþ Uþ 2pme kTe 3=2 Eiþ  DEiþ
¼2 i exp  ;
na Ua h2 kTe



ne niþ þ Uiþ þ 2pme kTe 3=2 Eiþ þ  DEiþ þ
¼2 þ exp  ; ð4:1Þ
niþ Ui h2 kTe



ne niþ þ þ Uiþ þ þ 2pme kTe 3=2 Eiþ þ þ  DEiþ þ þ
¼2 þþ exp  ;
niþ þ Ui h2 kTe

Here ne , niþ , niþ þ , niþ þ þ , na —the concentration of electrons, respectively, single,

double, and triple ions and atoms in the plasma; Uiþ , Uiþ þ , Uiþ þ þ , Ua —the
statistical sums of single, double, and triple ions and atoms in plasma Eiþ , Eiþ þ ,
Eiþ þ þ —the potential of the single, double and triple ionization of atoms; DEiþ ,
DEiþ þ , DEiþ þ þ —the reduction of potentials, respectively, for single, double, and
triple ionization of atoms due to the screening of the charged particles in plasma

2e2 3e2 4e2

DEiþ ¼ ; DEiþ þ ¼ ; DEiþ þ þ ¼ : ð4:2Þ
4pe0 rD 4pe0 rD 4pe0 rD

Taking into account the thermal non-equilibrium of plasma the Debye radius is
defined as
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2e0 kTe T
rD ¼   : ð4:3Þ
e ne þ ni þ 4niþ þ þ 9niþ þ þ ðTe þ T Þ
2 þ

Hereafter, we will use the following notation: Te 6¼ Tai ¼ Ta ¼ Ti ¼ T.

The Saha equations can be used for the calculation of the nonequilibrium plasma
composition, when Te 6¼ Tai and the experimental evidence is required. For this
calculation the other equation can be used. The Potapov equations [1] obtained in
the analysis of the chemical composition of the equilibrium system are the most
famous

þ T=Te


ni U þ 2pme kT 3=2 Eiþ  DEiþ
ne ¼2 i exp  ;
na Ua h2 kT

þ þ T=Te


ni Uiþ þ 2pme kT 3=2 Eiþ þ  DEiþ þ
ne ¼2 þ exp  ; ð4:4Þ
niþ Ui h2 kT

þ þ þ T=Te


ni Uiþ þ þ 2pme kT 3=2 Eiþ þ þ  DEiþ þ þ
ne ¼2 þþ exp  :
niþ þ Ui h2 kT

These equations are valid in cases where the plasma can be regarded as an ideal
gas, consisting of quasi-independent subsystems with different temperatures. When
thermal equilibrium (Te ¼ T) Potapov equation turns into Saha equation (4.1).
4.1 The Plasma Composition 95

To determine the composition of plasma it is necessary to use two additional

equations
The plasma quasi-neutrality condition

ne ¼ niþ þ 2niþ þ þ 3niþ þ þ ; ð4:5Þ

The Dalton’s law of partial pressures

 
p ¼ ne kTe þ niþ þ niþ þ þ niþ þ þ þ na kT: ð4:6Þ

Thus, to calculate the composition of the 5-component plasma we have a system

of nine Eqs. (4.1)–(4.6) with unknown variables ne ; na ; niþ ; niþ þ ;
niþ þ þ ; DEiþ ; DEiþ þ ; DEiþ þ þ ; rD . The solution of such a system can be per-
formed iteratively with the bottom of relaxation to achieve the convergence.
Calculation of the composition of two-temperature plasma is carried out for a given
temperature Te ; T and pressure p. The values Ua ; Uiþ ; Uiþ þ ; Uiþ þ þ ;
Eiþ ; Eiþ þ ; Eiþ þ þ are given as input variables. The results of the composition
calculation for the main gases of equilibrium plasma at atmospheric pressure are
shown in Figs. 4.1 and 4.2.

Fig. 4.1 The compositions of equilibrium plasmas: a Helium, b Neon, c Argon, d Krypton,
nR ¼ na þ ne þ niþ þ niþ þ þ niþ þ þ —the total concentration of particles
96 4 Determination of the Composition, Thermodynamic Properties …

Fig. 4.2 The composition of the equilibrium plasma of molecular gases at atmospheric pressure
[2]: a the Air, b Nitrogen, c Hydrogen

It should be noted that the compositions of nonequilibrium plasma derived on

the basics of Saha equation (4.1) and Potapov equation (4.4) are significantly dif-
ferent from each other in the strong thermal disequilibrium area (Fig. 4.3). There
are different opinions on suitability of the Potapov and Saha equations for calcu-
lation of nonequilibrium plasma composition. Some experimental values ne [3],
obtained for the two-temperature RF-plasma, are close to those calculated by the
Saha equation. However, the lack of reliable experimental data does not allow
judging about the validity of the application of these equations.
In the case where the ionization equilibrium is disturbed, composition of the
plasma should be determined by the equation of ionization equilibrium (1.40).
4.1 The Plasma Composition 97

Fig. 4.3 The concentration of electrons in the two-temperature argon plasma: straight line the
data obtained by the Saha equation; dotted line by the same Potapov equation

Fig. 4.4 Debye radius and reduced argon plasma ionization potentials depending on the
temperature
98 4 Determination of the Composition, Thermodynamic Properties …

On the Fig. 4.4, there are data for the Debye length and decreasing ionization
potentials of the argon plasma in the dependence on the temperature.

4.2 Thermodynamic Properties of the Plasma

Calculation of the thermodynamic properties includes calculation of the density,

heat capacity, and enthalpy of plasma depending on the temperature. They are
easily identified if the plasma consistence is known. So the density of a multi-
component plasma
X
q¼ mj nj ð4:7Þ
j

Here mj ; nj —mass and concentration of the jth component of plasma. The sum-
mation is over all the plasma components. For example,  for atomic plasma with
single, double, and triple ionization density q ¼ ma na þ niþ þ niþ þ þ niþ þ þ .
Plasma energy is defined as the sum of the enthalpies of all its components and
of stored energy. In addition, for each component the possible reserves of energy of
different degrees of freedom are taken into account [progressive
(Ha ; Hm ; He ; Hiþ ; Hiþ þ ; Hiþ þ þ ), vibrational (HV ), rotational (HR ) motion], stocks
of energy dissociation (HD ) and ionization (HI ), the excitation of different levels
(H  ), to perform work in compression advanced (Hp ), etc.
X
HP ¼ Hj ¼ Ha þ Hm þ He þ Hiþ þ Hiþ þ þ Hiþ þ þ þ HV
j ð4:8Þ

þ HR þ HI þ HD þ H þ Hp

In determination of the plasma enthalpy it is generally considered that the law of

distribution of energy over the degrees of freedom is valid: the same kinetic energy
kT=2 corresponds to each degree of the freedom in plasma. So, the enthalpy
 of atomic
plasma with single, double, and triple ionization will be H ¼ q1 52 kTne þ Eiþ
 
niþ þ Eiþ þ nþþ
i þEi
þþþ þþþ
ni þ 52 kT na þ niþ þ niþ þ þ niþ þ þ .
The results of density, enthalpy, and specific heat calculations in equilibrium
plasma of hydrogen H2, helium He, argon Ar, oxygen O2, nitrogen N2, and air are
shown in Figs. 4.5, 4.6, and 4.7.
The density of the two-temperature argon plasma is shown in Fig. 4.8, it shows
that until Te = 11,000 K the plasma density depends mainly on the temperature of
the heavy components.
For two-temperature plasma model of energy expended on ionization is taken
into account when calculating the total enthalpy of the electron gas. Thus, the
enthalpy of the electron gas of monatomic plasma considering single, double, and
triple ionization is determined by the formula
4.2 Thermodynamic Properties of the Plasma 99

Fig. 4.5 Density of equilibrium plasma at atmospheric pressure depending on the temperature: 1
Hydrogen (H2), 2 Helium (He), 3 Argon (Ar), 4 Oxygen (O2), 5 Nitrogen (N2), 6 Air

Fig. 4.6 The enthalpy of the equilibrium plasma at atmospheric pressure for gases depending on
the temperature: the notations as in Fig. 4.5


1 5
He ¼ kTe ne þ Eiþ niþ þ Eiþ þ niþ þ þ Eiþ þ þ niþ þ þ ð4:9Þ
q 2

In two-temperature plasma it should also distinguish the specific heat of elec-

trons and heavy components. The heat capacity of the jth component of the plasma
dH
at a constant pressure can be defined as: cpj ¼ dT j . The results of calculations of
j

enthalpy and heat capacity of a two-temperature argon plasma are shown in

Figs. 4.9 and 4.10.
100 4 Determination of the Composition, Thermodynamic Properties …

Fig. 4.7 Heat capacity of the equilibrium plasma at atmospheric pressure for gases depending on
the temperature: the notations as in Fig. 4.5

Fig. 4.8 The density of the two-temperature argon plasma at atmospheric pressure: straight line
calculation of Saha equation (4.1), dotted line calculation of Potapov equations (4.4)
4.2 Thermodynamic Properties of the Plasma 101

Fig. 4.9 The total enthalpy of a two-temperature argon plasma, depending on the temperature: All
the notations are the same as in Fig. 4.8

Fig. 4.10 The total heat capacity of a two-temperature argon plasma, depending on the
temperature: the notations are the same as in Fig. 4.8 (dashed–dotted and solid lines are different
because of the error of numerical differentiation enthalpy temperature)
102 4 Determination of the Composition, Thermodynamic Properties …

4.3 Transport Coefficients of the Plasma

Calculation of transfer coefficients is needed to determine the flow of heat transfer,

momentum and particle in the simulation of plasma, but it is often a very cum-
bersome and time-consuming operation, especially in the case of multicomponent
plasma with regard to its thermal and ionic nonequilibrium. In the course of the
calculation a number of assumptions of the selection of parameters for calculating
is used, for example, selected different kinds of effective cross sections, taking into
account of certain thermal and electrical phenomena occurring in plasma. The
results of calculations made by different authors are often significantly different, and
as a rule they go far beyond the authors point out errors. Unfortunately, the same
situation is observed for the experimental for data about the transport coefficients.
In these circumstances, it will be useful to compare the results of a simpler
method based on the concept of the mean free path with the data of accurate method
for calculating the Chapman–Enskog-based solutions of the Boltzmann transport
equation and choose a suitable equation for the calculation of the transport
coefficients.
The condition of local thermodynamic equilibrium (LTE) plasma greatly facil-
itates the task of defining its properties. On the basis of the theory, with respect to
thermodynamic equilibrium plasma, two different approaches in determining the
interaction of the particles in the plasma and the plasma transfer coefficients are
used. The first approach is based on the concept of the mean free path of the plasma
particles (atoms, ions, electrons, molecules, excited atoms and molecules, and
others). And the calculations of plasma transport coefficients can be obtained from
averaging over the Maxwellian distribution. The second approach is based on direct
solution of the Boltzmann equation—Chapman–Enskog method with the definition
of the relevant transport streams.
For multicomponent plasma the approximation of the Chapman–Enskog
although may give the error of about 1 %, but the error in the calculation of the
cross sections of particle collisions caused by an error in the knowledge of the
interaction potential of the order is of about 10 % or more [4]. Therefore, large
computational costs in the calculation of the transport coefficients of the above
procedure in many practical problems can be justified.
With regard to the two-temperature plasma, the conditions for calculating the
transport coefficients are: the electron gas and the gas of heavy components are both
in thermodynamic equilibrium; plasma reaches the ionization equilibrium to
achieve thermal equilibrium that is possible when the ionization rate exceedingP the
rate of thermal relaxation of the plasma; plasma is quasi-neutral, i.e., ne ¼ Zi ni ;
basic mechanisms of charged particles are straight and speed two-particle ioniza-
tion, while the three-body recombination—the channel of their disappearance;
plasma lacks all possible convective and diffusive fluxes; radiation and free access
of charged particles from the plasma volume does not violate the ionization equi-
librium of the system (n_ e ¼ 0); radiation is considered to be optically thin.
4.3 Transport Coefficients of the Plasma 103

In accordance with such restrictions, the plasma composition, its thermodynamic

properties, and the coefficients are uniquely determined by the electron temperature
Te, the temperature of the heavy components T, and pressure p.
To calculate the transport coefficients of a nonequilibrium plasma approach
based on the use of simple, yet experimentally proven methods for determining the
properties of the plasma are more appropriate, as well as the results of accurate
calculations obtained by solving the kinetic equation. For example, in [5] the
two-temperature plasma transport coefficients are calculated based on the concept of
the mean free path of particles and taking into account the kinetic corrections for
accurate calculations in thermal equilibrium (T = Te).
Definition of the argon plasma properties is relevant in connection with the
widespread use of argon in the plasma torch technology. The most accurate transfer
coefficients for equilibrium in argon plasma are obtained in [6–11]. The coefficients
of two-temperature plasma are both theoretically and experimentally in many
studies [12–19]. The data of different authors about argon on thermodynamic
properties such as density, heat capacity, and enthalpy, have the satisfactory
agreement. However, the data of different authors on the dissipative transport
coefficients—electrical conductivity, thermal conductivity, viscosity, radiation—
vary greatly.

4.3.1 Determination of the Electrical Conductivity

and Thermal Conductivity

The mobility of electrons is much larger than mobility of ions, so the conductivity
of overall gas is often equated with its electronic conductivity σ ≈ σe.
Methods for the calculation of electrical conductivity for partially ionized gas of
arbitrary composition that are based on the exact theory, are quite complex.
Relatively simple and accurate equation for calculation of the electrical conductivity
of the ionized gas can only be obtained in two limiting cases. For a weakly ionized
gas we have

Z1
4p ne e2 v2 f ðvÞ
re ¼ P eM dv;
3 kTe m mem ðvÞ
0

Here feM ðvÞ—the Maxwell distribution function of the electron velocity;

P
m mem ðvÞ—the sum of the frequency of collisions for electrons with all neutral
particles.
In the case for the fully ionized plasma, consisting of singly charged ions there is
no field, expressions for re ; ke take the following form:
104 4 Determination of the Composition, Thermodynamic Properties …

ne e 2 ne k2 Te
re ¼ 1;975 ; ke ¼ 3;203 ;
me mei me mei
pffiffiffiffi
3=2
2
e2
Here mei ¼ ni 4 3
2p me
kTe 4pe0 me ln K—the frequency of collisions.
These results were obtained by L. Spitzer, taking into account the Debye
screening in determination of the Coulomb potential and, as shown in [5], they are
valid for only large values of ln K.
For the calculation of the electrical conductivity and the electron conductivity
the expression of elementary kinetic theory can be used in the form

k r e 2 ne k r e 2 ne
re ¼ P ¼ qffiffiffiffiffiffiffi ; ð4:10Þ
me ve j nj Qej m 8kTe ðn Q þ n Q Þ
e pme a ea i ei

kke ne k2 Te kke ne k2 Te
ke ¼ P ¼ qffiffiffiffiffiffiffi ; ð4:11Þ
me ve nj Qej m 8kTe ðn Q þ n Q Þ
j e pme a ea i ei

qffiffiffiffiffiffiffi
Here 
ve ¼ pme —the
8kTe
average thermal velocity of the electrons; kr ; kke —the
amendments to the kinetic model of equilibrium plasma (Ta = Te) that are
the functions of the collisions frequency. The coefficients kr ; kke indicate how
the values of re ; ke are calculated by the equations of the kinetic theory, and are
different from the corresponding values obtained by the elementary equations. The
change limits are kr ¼ 1  2:8; kke ¼ 0:3  1:6. The calculation results of re ; ke for
the equilibrium argon plasma (4.10) and (4.11) with kr ¼ 1:05; kke ¼ 1:6 have a
good agreement with data obtained by the Chapman–Enskog [6, 7]. Data for the
calculation of σ (4.10) for the equilibrium argon plasma at atmospheric pressure are
shown in Fig. 4.11. It also shows their comparison with experimental data and the
calculated data of other authors. The calculation results for the electrical conduc-
tivity of various gases in equilibrium plasma at atmospheric pressure are shown in
Fig. 4.12.
The electronic conductivity in [5] with the influence of an alternating electro-
magnetic field is determined as follows:

k r e 2 ne m m
re ¼  :
me x2 þ m2m
P
Here x ¼ eB=me —the gyromagnetic or cyclotron frequency; mm ¼ j mej —the
sum of the collisions frequency of electrons with all the other particles. Electronic
thermal conductivity [5] is calculated by the expression
4.3 Transport Coefficients of the Plasma 105

Fig. 4.11 The electrical conductivity of the equilibrium argon plasma depending on the
temperature: straight line the author’s estimates; dotted line the results of the calculation of Devoto
[6]; filled square, filled circle, plus sign experimental data, respectively [21–23]

Fig. 4.12 The electrical conductivity of the equilibrium plasma at atmospheric pressure for gas:
the notations are the same as in Fig. 4.5

15=ð2pÞ kke ne k2 Te
ke ¼ P :
1 þ pffiffi2 P
mei
m
me j mej
j ej

The calculation using this formula with the amendment kke ¼ 1:2 gives the same
result as the calculation by (4.11) with the amendment kke ¼ 1:6.
In some works, the electronic electrical conductivity was calculated from the
ratio for the mixture [5], when the resistivity r1e ¼ r11 þ r12 is the sum of the resis-
tivity for weakly ionized and fully ionized gases. Similarly k1e ¼ k11 þ k12 . However,
in this approach it should be borne in mind that the additive r1e ¼ r11 þ r12 is broken
at comparable values of the collisions frequency mei and mea .
106 4 Determination of the Composition, Thermodynamic Properties …

The thermal conductivity of the heavy components can be determined as [20]

rffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffi
75k pkTa na 75k pkTi ni
ka ¼ ; ki ¼ :
64 ma na Qaa þ ni Qai 64 mi na Qai þ ni Qii

The results of calculation by equations [5]

pffiffiffi pffiffiffi
7:5 2 k 2 na Ta 7:5 2 k2 ni Ti
ka ¼ qffiffiffiffiffiffiffi ; ki ¼ qffiffiffiffiffiffi
p m 8kTa ðn Q þ n Q Þ p m 8kTi ðn Q þ n Q Þ
a pma a aa i ai i pmi a ai i ii

have good agreement with these data.

However, the calculation for the equilibrium plasma by the last equation gives
some higher results at T > 9000 K compared with the data of [6]. This is mainly
due to the neglect of the atom-ion recharge.
For the temperature Ta < 9000 K, K ¼ 0:4 (2.29)
rffiffiffiffiffiffiffiffiffiffi
75k pkTa n
ka ¼ a ; ð4:12Þ
64 ma na Qaa þ ni Qai þ Qaiper
rffiffiffiffiffiffiffiffiffi
75k pkTi ni
ki ¼   ; ð4:13Þ
64 mi na Qai þ Qaiper þ ni Qii

For the temperature Ta ≥ 9000 K, K ¼ 0:5 (2.29)

pffiffiffi
7:5 2 k 2 na T a
ka ¼ qffiffiffiffiffiffiffi  ; ð4:14Þ
p m 8kTa n Q þ n Q þ Q
a pma a aa i ai aiper

pffiffiffi
7:5 2 k2 ni Ti
ki ¼ qffiffiffiffiffiffi   : ð4:15Þ
p m 8kTi n Q þ Q þ n Q
i pmi a ai aiper i ii

The existence of spatial gradients of temperature and concentration of the par-

ticles in plasma causes the different diffusion fluxes. Thus, the electrons and ions
diffuse from the hot spots in the cold, where they recombine and emit energy spent
before their ionization. There is a special type of heat transfer—reactive thermal
conductivity. Reactive heat conductivity is determined by the corresponding flow of
 
heat transfer like kR ¼ Ei Damb @n
@T p . In [28], the thermal conductivity is determined
e

as

Ei2 na ne
kR ¼ nP Damb : ð4:16Þ
kTa ðna þ ne Þ2
4.3 Transport Coefficients of the Plasma 107

Here nP ¼ na þ ne þ ni —the total concentration of particles.

The total thermal conductivity for the equilibrium argon plasma is shown in
Fig. 4.13. The results show that the total thermal conductivity consists essentially of
reactive and electronic conductivity. In the temperature range of T = 10,000–
15,000 K the reactive thermal conductivity can be greater than the contact (con-
duction atoms, ions and electrons) and provides a maximum total conductivity. The
values of the total thermal conductivity of the equilibrium plasma at atmospheric
pressure in various gases are shown in Fig. 4.14. The molecular gas first maximum
thermal conductivity is achieved mainly at T < 10,000 K due to the dissociation of
molecules (Fig. 4.14).
Electrical and thermal conductivity of two-temperature argon plasma are
designed using the above equations, the results of these calculations are shown in
Figs. 4.15 and 4.16.

Fig. 4.13 The total thermal conductivity of the argon plasma depends on the temperature:
experimental data: filled square [24]; filled triangle [25]; cross line [23]; filled circle [26]; open
square [27]; the other symbols are the same as in Fig. 4.11

Fig. 4.14 The thermal conductivity of the equilibrium plasma at atmospheric pressure: the
notations as in Fig. 4.5
108 4 Determination of the Composition, Thermodynamic Properties …

Fig. 4.15 The electrical conductivity of two-temperature argon plasma: the notations are the same
as in Fig. 4.8

4.3.2 Determination of the Viscosity, the Ambipolar

Diffusion Coefficient, and Integrated Emission

With thePconcept of the mean free path of the particles the viscosity may be defined
as l ¼ i mi ni  vi li , where vi —the average thermal velocity for the particles of ith
grade; li —the mean free path for the particles of ith grade.
Based on the concept of the mean free path it can also be set that the ratio of the
viscosity of the electrons le to the viscosity of the heavy particles lai have the order
of [5]


le ne me Te 1=2
 :
lai na þ ni ma Ta
4.3 Transport Coefficients of the Plasma 109

Fig. 4.16 The total thermal conductivity of two-temperature plasma depending on temperature:
the notations as in Fig. 4.8

It follows that the plasma viscosity is mainly determined by the momentum

transfer at the contact movement of the heavy particles (atoms, ions, molecules),
therefore, the influence on the viscosity of the electrons can be neglected.
In particular, the viscosity of plasma is caused mainly by ion pulse transfer, so it
is determined by the relation [5]
pffiffiffiffiffi
0:406ð4pe0 Þ2 mi ðkTi Þ5=2
li ¼ ð4:17Þ
Z 4 e4 ln K

Here Z—ion charge.

There are other equations for the viscosity calculation in the references, e.g.,
110 4 Determination of the Composition, Thermodynamic Properties …

" #
na ni
lai ¼ 6  1025 Ta0:5  þ   :
na Qaa þ ni Qai þ Qaiper na Qai þ Qaiper þ ni Qii
ð4:18Þ

The calculation results using the Eq. (4.18) have some discrepancy with the data
of Devoto [6] at T < 12,000 K.
Viscosity is related to the thermal conductivity as

4
lai ¼ ma kai ; ð4:19Þ
15k

and in accordance with the expressions (4.12)—(4.15) it can be written that

For T < 9000 K
" #
5 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi na ni
lai ¼ pkTa ma  þ   ; ð4:20Þ
16 na Qaa þ ni Qai þ Qaiper na Qai þ Qaiper þ ni Qii

For T > 9000 K

rffiffiffiffiffiffiffiffiffiffiffiffi" #
kTa ma na ni
lai ¼  þ   : ð4:21Þ
p na Qaa þ ni Qai þ Qaiper na Qai þ Qaiper þ ni Qii

As in the case of the thermal conductivity calculation, in the viscosity deter-

mination it is necessary to consider the recharging process of ions. The results of
calculation for the equilibrium viscosity of argon plasma using (4.20), (4.21) and
taking into account exchange cross section [(2.29) with K = 0.5] at a temperature

Fig. 4.17 Viscosity of the equilibrium argon plasma at atmospheric pressure, depending on the
temperatures: Experimental data: filled square [29]; filled circle [30]; open square [31]; the other
symbols are the same as in Fig. 4.11
4.3 Transport Coefficients of the Plasma 111

Fig. 4.18 Viscosity of the equilibrium plasma at atmospheric pressure: the notations as in Fig. 4.5

Fig. 4.19 Viscosity of the two-temperature argon plasma, depending on the temperatures: the
notations as in Fig. 4.8

T < 20,000 K are in good agreement with the data of [6] (4.17). There are some
experimental data for comparison in Fig. 4.17. The results of viscosity calculation
for various gases of equilibrium plasma at atmospheric pressure are shown in Fig.
4.18. By (4.20) and (4.21), we can calculate the viscosity of the two-temperature
argon plasma (Fig. 4.19).
112 4 Determination of the Composition, Thermodynamic Properties …

Ambipolar diffusion coefficient for nonequilibrium plasma (Fig. 4.20) is deter-

mined using the equilibrium data [7] according to the following expression

Damb ¼ 3:97  1010 Ta0:63 ðTe þ Ta Þ: ð4:22Þ

The calculation of the integral characteristics of plasma by the emission spec-

trum is an important task in the preparation of the energy balance, the accuracy of
this calculation is often associated with the correct model described the plasma. The
model of partial local thermodynamic equilibrium (PLTE) and the equations
Maxwell–Boltzmann–Saha are most commonly used (1.2), (1.4), (1.9). For deter-
mination of the integral radiation ratio, the absolute intensities of the spectral lines
and continuum are the main source of data. For example, the estimated linear and
continuous emission of argon plasma at atmospheric pressure in the test temperature
range and wavelength k [ 200 nm in many cases, are optically thin.
The intensity of the spectral lines, the transition of an atom, or ion to the level of
the upper kth level to the lower ith energy level is expressed by

Z1
1
Iki ¼ hmki Aki nk ¼ iki ðkÞdk; ð4:23Þ
4p
0

Here mki —frequency of the photons radiation in the center of the line
(mki ¼ m0 ¼ c=k0 ), Aki —the probability of spontaneous transition k ! i (the first
Einstein coefficient), nk—the population of emitting (upper) level of the transition.
The calculation of the ratio (4.23) is exact, this ratio is used if the optical
thickness of the plasma layer with depth L is small

ZL
s0 ¼ k0 dl  1;
0

Here k0—the absorption coefficient at the center of the line.

Full volume emission, such as argon, can be determined by [37]

 

þ 
40
pffiffiffiffiffi DEi þ hm þþ hm
Urad ¼ 1:14  10 Te exp  ne ni exp þ 4ni exp ;
kTe kTe kTe
ð4:24Þ

Here hm ¼ 2:85 eV; hm þ ¼ 8:2 eV; DEi ¼ 4pee0 rD —the decrease in the ionization
2

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Te þ Ta Þ—the Debye radius.
potential; rD ¼ e2 nee0ðkT e Ta

The Eq. (4.24) describes the radiative losses due to brake and recombination
radiation, as well as discrete transitions between excited states of atoms and ions.
The results of calculation by this equation for the equilibrium argon plasma and
4.3 Transport Coefficients of the Plasma 113

Fig. 4.20 Ambipolar diffusion coefficient of a two-temperature argon plasma, depending on the
temperature

Fig. 4.21 The total radiation of equilibrium argon plasma: Experimental data: filled square [32];
filled circle [10]; plus sign [33]; cross sign [34]; curves estimates: single dot [35], double dot [36],
straight line author
114 4 Determination of the Composition, Thermodynamic Properties …

Fig. 4.22 Full two-temperature argon plasma emission in depending on temperature: the
notations as in Fig. 4.8

compared with data obtained by other authors and some of the experiments are
given in Fig. 4.21, and the results of the calculation of the radiation
two-temperature plasma of argon are shown in Fig. 4.22.

4.4 The Coefficients of Triple Recombination and Impact

Ionization

In case of deviation from the plasma ionization equilibrium its composition must be
determined by solving the equation of ionization equilibrium (EIE) (1.40). The main
mechanisms of birth and disappearance of charged particles in a dense low-
temperature plasma are stepwise ionization and triple recombination. For solving the
EIE, it is necessary to find the coefficients of recombination kr and ionization ki that
are connected with each other by the principle of detailed balance [5, 38]. To calculate
the rates of these processes it is sufficient to know one of these factors.
4.4 The Coefficients of Triple Recombination and Impact Ionization 115

It should be emphasized that both the experimental and theoretical data on these
factors are very scarce, and the main method of their determination is a theoretical
calculation.
The best-known analytical expression for the calculation of kr obtained by
Hinnov and Hirchberg [39], is

kr ¼ 3:4  1034 ðTe =1000Þ9=2 : ð4:25Þ

The result of the kr calculation by (4.25) is in good agreement with the exper-
imental data (Fig. 4.23), but this expression is applicable only in a very narrow
range of temperatures (Te < 3000 K).
Some estimates of kr for hydrogen (Fig. 4.23) were represented in the works of
Bates [44, 45]. In the work of A. Gurevich and L. Pitaevskii the kr factor is defined
as
pffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffi
4 2p3=2 e10 Z 3 ln Z 2 þ 1
kr ¼ ¼ 0:614  1020 Te9=2 ð4:26Þ
9ð4pe0 Þ5 1=2
me ðkTe Þ9=2

In the work of Biberman [46] the following formula for the calculation of kr is
given:

kr ¼ 4:3  1044 ðEH =Te Þ9=2

" #2
ðgi =ga ÞðDE1 =Te Þ1 þ b ðEH =Te Þ3=2 expðE2 =Te Þ ð4:27Þ
 1þ ;
13:3a

Here gi ; ga —the static weight of the ions and atoms; EH —the ionization potential of
the hydrogen atom; DE1 ¼ E1  E2 —the potential difference of two energy levels
of corresponding atom (E1 ¼ Ei ). When Te 0:07DE1 the coefficients a = 0.032,
b = 4/9; Te
0:07DE1 —a = 0.25, b = 5/6.
Note that the calculation Eq. (4.27) gives low results (three decimal orders)
compared with the experimental data in the temperature range Te = 7000–
20,000 K. Further, in the calculation according to (4.27) there is sharp rise in the
values of kr at Te  0:07DE1 (Fig. 4.23). Ismailova [47] gives a different expression
for determining the rate of recombination



44 E Ei  E 
kr ¼ 1:29  10 þ 2 exp ; ð4:28Þ
kTe kTe

Here E  ¼ 11:61 eV.

However, the value kr, obtained using (4.28) does not coincide with the results of
calculations by (4.27), and all such equations give lower results compared with the
experimental data (Fig. 4.23). Good agreement between the calculated and exper-
imental data, however, provides a relatively simple equation of Raiser [48]
116 4 Determination of the Composition, Thermodynamic Properties …

kr ¼ 5:2  1035 ðTe =1000Þ9=2 ¼ 1:644  1021 Te9=2 : ð4:29Þ

From the comparison of the calculated and experimental data (Fig. 4.23) it
follows that at Te 3200 K the formula of E. Hinnov and J. Hirchberg (4.25) is
recommended, at 3200 K Te 3800 K—the L. Bibermans formula (4.27) and at
Te
3800 K—the formula of Y. Raiser (4.29).
Impact ionization coefficient ki is determined by kr based on the principle of
detailed balance

n2
ki ¼ kr e
; ð4:30Þ
na

Fig. 4.23 Triple recombination coefficient depending on electron temperature: Experimental data:
cross sign for He [40]; filled circle for Cs [40]; filled square for Ar [41]; filled triangle for Ar [42];
open square for H [43]; Estimated results: 1 by (4.25), 2 by (4.26), 3 by (4.29), 4 for H [44], 5 by
(4.27)
4.4 The Coefficients of Triple Recombination and Impact Ionization 117

Fig. 4.24 Impact ionization coefficient depending on temperature: straight line calculation on the
Saha equation (4.1); dotted line calculation on the Potapov equations (4.4)

Here ne ; na —the concentrations of electrons and atoms in the ionization balance,
calculated by the Saha equation. The calculation results for ki are presented in
Fig. 4.24.

4.5 The Frequency of Particle Collisions, the Current

Density, and Heating Capacity of Plasma

For improving the plasma heating and to design the plasma torches or analyze of
the plasma flow properties in the channel of plasma torches, it is important to know
the frequency of the particles collisions, the current density and the plasma power
under heating. The results of calculation of the average total frequency of collisions
are shown in Fig. 4.25 and prepared according to the following equation:
118 4 Determination of the Composition, Thermodynamic Properties …

Fig. 4.25 The average total frequency of elastic collisions of electron–atom–ion in argon plasma:
the notations as in Fig. 4.24

rffiffiffiffiffiffiffiffiffi
8kTe
meai ¼ ðna Qea þ ni Qei Þ: ð4:31Þ
pme

In Fig. 4.26 there are the results of determining the minimum current density to
for the stationary combustion of the argon plasma at atmospheric pressure, esti-
mated from the following expression:
rffiffiffiffiffiffiffiffiffi
3 2me 8kTe
j ¼ k
2
ne ðna Qea þ ni Qei ÞðTe  Ta Þr: ð4:32Þ
2 ma pme

Figure 4.27 shows the calculation results of the power transmitted through the
elastic collisions between electrons and heavy particles of argon plasma and is
defined as
rffiffiffiffiffiffiffiffiffi
3 2me 8kTe
Seai ¼ k ne ðna Qea þ ni Qei ÞðTe  Ta Þ: ð4:33Þ
2 ma pme
4.5 The Frequency of Particle Collisions, the Current Density … 119

Fig. 4.26 The minimum current density for steady burning of argon plasma: the notations as in
Fig. 4.24

Fig. 4.27 The specific power transmitted in argon plasma at atmospheric pressure by the heavy
particles (atoms and ions) through elastic collisions of electrons: the notations as in Fig. 4.24
120 4 Determination of the Composition, Thermodynamic Properties …

References

1. A.V. Potapov. Chemical equilibrium multitemperature systems // J. Thermal physics of High

Temperatures. 1966. T. 4. No 1. pp 55–58.
2. Ochkin V.N. Spectroscopy of low-temperature plasma. M .: FIZMATLIT 2006.
3. Goldfarb V.M., Dresvin S.V. // J. Thermal physics of High Temperatures. 1965. V. 3. No 3.
PP. 420–424.
4. Eletskii A.V., Palkina L.A., Smirnov B.M. Transport phenomena in weakly ionized plasma.
M .: Atomizdat 1975.
5. M. Mitchner, C. Kruger. Partially ionized gases. M.: Mir, 1976.
6. Devoto R.S. Transport coefficients of ionized argon // Phys. Fluids. 1973. Vol. 16. No 5.
P. 616–623.
7. Devoto R.S. Transport coefficients of partially ionized argon // Phys. Fluids. 1967. Vol. 10.
No 2. PP. 354–364.
8. Devoto R.S. Simplified expressions for the transport properties of ionized monatomic gases //
Phys. Fluids. 1967. Vol. 10. No 10. P. 2105–2112.
9. Li C.P., Devoto R.S. Fifth and sixth approximations to the electron transport coefficients //
Phys. Fluids. 1968. Vol. 11. P. 448–450.
10. Evans D.L., Tankin R.S. Measurement of emission and absorption of radiation by an argon
plasma // Fhys. Fluids. 1967. Vol. 10. No 6. P. 1137–1144.
11. Drellischak K.S. Partition functions and thermodynamic properties of high temperature gases
// AEDS TRD-64. 1964. Vol. 10. P. 1–148.
12. Polyansky V.A. Diffusion and conductivity of a partially ionized gas mixture multitemper-
ature // J. of Technical Physics Letters. 1964. No 5. pp. 11–17.
13. Zhdanov V.M. Transport phenomena in multicomponent plasma. M.: Energoizdat 1982.
14. Kannappan D., Bose T.K. Transport properties of a two-temperature argon plasma // Phys.
Fluids. 1977. Vol. 20. No 1. P. 1668–1673.
15. Miller E.J., Sandler S.I. Transport properties of two-temperature partially ionized argon //
Phys. Fluids. 1973. Vol. 16. No 4. P. 491.
16. Hsu K.C., Pfender E. Calculation of thermodynamic and transport properties of a
twotemperature argon plasma // Proc. of V Int. Symp. Plasma Chem., Edinburgh. 1981.
Vol. 1. P. 144–152.
17. Bohadore M.N., Soo S.L. Nonequilibrium transport phenomena in partially ionized argon //
J. Heat and Mass Transfer. 1966. Vol. 9. P. 17–34.
18. Dresvin S., Nguyen Quoc Shi. Thermodynamic and transport properties two temperature Ar
plasma // Abst. of the IV European Conference on Thermal plasma Processes—Athens.
Greece, 1996. P. 120.
19. Shi Nguyen-Kuok. The study of RF and Arc plasma torches: Dissertation of D.Sc. St.
Petersburg State Technical University, 2002.
20. Lui W.S., Whitten B.T., Glass I.I. Ionizing argon boundary layers. Part 1. Quasi-steady
flat-plate laminary boundary layer flows // J. Fluid Mech. 1978. Vol. 87. No 4. P. 609–640.
21. Lau J., Canad. // J. Phys. 1964. Vol. 42. P. 1548.
22. Lin S.C., Resler E.L., Kantrowitz. A. // J. Appl. Phys. 1955. Vol. 26. P. 95.
23. Emmons H.W. // Phys. Fluids. 1967. Vol. 10. P. 1125.
24. Asinovsky E.I., Kirilin A.V. // J. Thermal physics of High Temperatures. 1965. V. 5. P. 677.
25. Bues I., Patt H.J., Richter J.Z. // Angew. Phys. 1967. Vol. 22. P. 345.
26. Knopp C.F., Cambel A.B. // Phys. Fluids. 1966. Vol. 9. P. 989.
27. Yos J.M. Transport Properties of Nitrogen, Hydrogen, Oxygen, and Air. RDS-TM-63-7.
1963. AVCO/RAD. Wilmington, Mass.
28. Donskoy A.V., Klubnikin V.S. Electrical plasma processes and installation engineering. L.:
Engineering, 1979.
29. Aeschliman D.P., Cambel A.B. // Phys. Fluids. 1970. Vol. 13. P. 2466.
30. Schreiber P.W., Hunter A.M., Bendetto K.R. // Phys. Fluids. 1971. Vol. 14. P. 2696.
References 121

31. Soo S.L., Bahadori M. // AGAR Dogaph. 1964. No 84. P. 1013.

32. Krey R.U., Moris J.C. // Phys. Fluids. 1970. Vol. 13. P. 1483–1489.
33. Abu-Romia M., Blair D. // AIAA Paper. 1969. P. 601–602.
34. Barzelau M. // AIAA Paper. 1966. Vol. 4. P. 815–819.
35. Biberman L.M., Norman G.E. // Successes of physical sciences. 1967. No 91. PP. 193–218.
36. I.T. Yakubov // Optics and Spectroscopy. 1965. No 19. PP. 193–199.
37. Kozlov G.I., Kuznetsov V.A., Masukov V.A. Radiant loss of Argon plasma and radiative
model of a continuous optical discharge // J. Experimental technical physics. 1974. V. 66.
PP. 954–964.
38. Golant V.E., Zhilinski A.P., Sakharov S.A. Basics of plasma physics. M.: Atomizdat 1977.
39. Hinnov E., Hirchberg J.G. Electron-Ion recombination in dense plasmas // Phys. Rev. 1962.
Vol. 126. P. 795.
40. Aleskovsky Y.M. Study volume recombination in a Cesium plasma // J. Experimental
technical physics. 1963. V. 17. P. 570.
41. Cool T.A., Zukoski E.E. Recombination, ionization and nonequilibrium electrical conduc-
tivity in seeded plasmas // Phys. Fluids. 1966. Vol. 9. P. 780.
42. Desai S.V. Corcoran W.H. Recombination of electrons and ions in an atmospheric argon
plasma // J. Quant. Spectrosc. Radiat. Transfer. 1969. Vol. 9. PP. 1371–1386.
43. Craggs J.D., Hopwood W. // Proc. Roy. Soc. 1946. Vol. 186a. P. 257.
44. Atomic and Molecular Processes / Ed. by D.R. Bates. M.: Mir, 1964.
45. Bates D.R., Kingston A.E. Collisional-radiative recombination at low temperatures and
densities // Proc. Phys. Soc. 1963. Vol. 83. P. 43.
46. Biberman L.M., Vorobiev V.M., Yakubov N.T. Nonequilibrium Low-temperature plasma IV
ionization and recombination functions // J. Thermal physics of High Temperatures. 1969.
V. 7. No 4. PP. 593–603.
47. Ismailova D. Study of the electric Arc in Argon / Ed. by A.F. Jeenbaev. Frunze: Ilim 1963.
48. Y.P. Raiser. Physics of the gas discharge. M.: Nauka, 1987.
Chapter 5
The Boltzmann Kinetic Equation
and Calculation of the Transport
Coefficients

5.1 Introduction of the Boltzmann Kinetic Equation

The kinetic theory of gases is based on the knowledge of the distribution function,
defined in the phase space (μ-space) of a molecule. This function is the probable
number of gas molecules fi ðr; vi ; tÞdrdvi of ith grade in the elementary volume dr
(*—For convenience, in this chapter, vectors in bold) around the point r, with the
phase velocities in the element dvi near vi .
Consider a mixture of monatomic gases in nonequilibrium state. Assume that the
mixture is sufficiently sparse, so for the description of the properties of such mixture
the pair, rather than the triple collision of particles should be considered. For
generality, assume that the external force Fi acts as the mass of particles per unit of
ith type. We believe the strength of Fi —the function of coordinates and time,
independent of the velocity. It is believed that the external force is much smaller
than the forces acting on the molecules during the collision. Intermolecular forces
are far superior than gravitation forces and are valid only for the moment of
collision.
The basis of the statistical description of plasma is the Boltzmann Kinetic
Equation (BKE) with respect to the distribution function fi ðri ; vi ; tÞ for each ith
grade of the plasma particles that is generally seven-dimensional. In the case of
elastic interaction of particles, the BKE is

@fi @fi @fi

þ vi  þ Fi 
@t @r @vi
Z1 Z2p Z1

X      
¼ fi v0i fj v0j  fi ðvi Þfj vj gij bdbdedvj ð5:1Þ
j
0 0 0
X
¼ jðfi fj Þ;
j

© Springer International Publishing Switzerland 2017 123

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_5
124 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

Fig. 5.1 The particle scattering in a pair collision: 1, 2—particles of ith and jth grades

     
where fi ðvi Þ; fj vj ; fi v0i ; fj v0j —the particle distribution function of ith and jth
type before and after collision, respectively; vi ; vj ; v0i ; v0j —the molecular velocities
of particles of ith and jth grades before and after collision, respectively; Fi ¼ emi Ei —
the strength of the electric field acting on a unit mass of the particle; E—the
strength of electric field; ei —the charge of ith component;
gij ¼  gji ¼ vi  vj ¼ g—the initial relative velocity of the particles of ith and jth
grades; b—the impact parameter; e—the angle of azimuthal orientation of the plane
where the scattering of particles takes place (Fig. 5.1).
We accept the following notations:

XZ Z Z     
1 2p 1
 
fi v0i fj v0j  fi ðvi Þfj vj gij bdbdedvj
j
0 0 0 ð5:2Þ
X ZZZ   X
¼ fi0 fj0  fi fj gbdbdedvj ¼ jðfi fj Þ
j j

The Boltzmann kinetic equation is also often written in the form of a differential
operator:
  X ZZZ  
@
þ vi  rr þ Fi  rvi fi ðr; vi ; tÞ ¼ fi0 fj0  fi fj gbdbdedvj
@t j
X ð5:3Þ
¼ jðfi fj Þ
j

The left side of BKE—the total change of the particle distribution function in
time, space, and the phase space velocity, respectively, the right side—the sum of the
collision integrals jðfi fj Þ, it is called the operator of collision. The right side of (5.3)
5.1 Introduction of the Boltzmann Kinetic Equation 125

depends on the change of the distribution function due to the interaction of the
particles with each other and with the walls. From elementary volume eliminated the
particles, have experienced a collision, their number is equal to the number of
particles in the phase space, multiplied by the probability of the particle output
process from a given volume. Into the same volume the particles are brought after
inverse collisions. The right-hand side of Eq. (5.3) for a particle in an unbounded
space has a form determined by the process of direct and inverse collisions.
For finding the sum of the collision integrals for all components, interaction
between identical components (when i = j) is considered as well as between the
different, which speeds up to and after the collision, they are considered as inde-
pendent variables in the process of integration. In the references during the decision
of BKE, often the index «i» is omitted (i.e., vi → v and fi → f, j = 1). The
Boltzmann kinetic equation for the one-component gas can be written as
  ZZZ
@  0 0 
þ v  rr þ F  rv f ðr; v; tÞ ¼ f f1  ff1 gbdbdedv1 ¼ jðff1 Þ: ð5:4Þ
@t

Equation (5.1) is valid only for the low-density gas. If the mean free path of the
particles in plasma is small compared with the dimensions of the plasma torch
channel, the gas behaves as a continuous medium and the plasma is represented as a
complex multicomponent chemical system, where interaction of the components
described various functions of potential /ðrÞ.
Gas in any initial state that has not external perturbations for a long time reaches
eventually a steady state. If the system is adiabatically isolated and is not affected
by external forces, such stationary state is homogeneous state where all of the
distribution functions fi ðr; vi ; tÞ are reduced to the Maxwell distribution:
 m 3=2  m v2 
i i i
fi ¼ 4pv2i ni exp  : ð5:5Þ
2pkT 2kT

Using the H-theorem it can also be proved that the integrand of the right side of
each integral in the BKE (5.1) equals to zero. This is the principle of detailed
balance, when in equilibrium state, the number of molecules of the ith grade, with
the rate of nonspecified interval velocities, due to collisions with molecules of jth
grade, exactly compensated by inverse collisions by the number of molecules with
velocities corresponding to the specified interval.
When the system is not in equilibrium, the distribution function can be deter-
mined from Eq. (5.1).
Usually we are interested in the properties of gases in the environment, only a
little different from the equilibrium. Indeed, only in those circumstances where the
flows are linear with respect to some derivatives, usual calculation of the transfer
coefficients can be used. This limitation of the distribution function is close to
Maxwell, and the Boltzmann kinetic equation can be solved by perturbation theory
developed by Chapman and Enskog [1]. The solutions are then used to get of the
expressions for the fluxes and transport coefficients. In these expressions, it is
126 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

considered that mass transfer occurs in the presence not only of the concentration
gradient (diffusion), but the temperature gradient (thermal diffusion), and vice versa.
These and other phenomena of the second order that cannot be described in
terms of a simplified kinetic theory, of course found out only when a more rigorous
approach is used. For example, the diffusion coefficient is numerically equal to the
particle flux density in this class of particles; the viscosity is numerically equal to
the momentum flux density produced by a unit of the velocity gradient, and thermal
conductivity—to the energy flux density caused by a unit temperature gradient.

5.2 The Transport Equations

The kinetic theory of rarefied and dense gases provides a transport equation. These
equations are differential equations describing the concentration, flow rate, and
temperature as a function of coordinate and time.
Hydrodynamic transfer equation—the equation of mass conservation, momen-
tum conservation, and energy conservation—can be obtained directly from the
BKE. In deriving these equations, certain values expressed in terms of the distri-
bution function appear. We can show that they are the flows of mass, momentum,
energy, and directly related to the speed of the diffusion, tensor of pressure, and
heat flux.
The linear velocity of the particle of ith grade relative to the fixed coordinate
system is denoted by vi , and the components of the velocity vix —through viy , viz .
The absolute value of the linear velocity is denoted as vi . The average speed vi of
the particle of ith kind, with concentration of particles ni is determined as follows:
Z
1

vi ðr; tÞ ¼ vi fi ðr; vi ; tÞdvi : ð5:6Þ
ni

The average speed is a function of coordinate and time, and it is a macroscopic

velocity for the particles flow of ith grade.
The average mass velocity is determined by the equation
Z
1X
v0 ðr; tÞ ¼ mi vi fi ðr; vi ; tÞdvi ; ð5:7Þ
q i
P
Here qðr; tÞ ¼ ni mi —the total density of the gas.
i
The thermal velocity for the particle of ith grade is defined as the velocity of the
particle in the coordinate system moving with the mean mass velocity v0 :

Vi ðvi ; r; tÞ ¼ vi  v0 : ð5:8Þ
5.2 The Transport Equations 127

The diffusion rate of the ith component of the chemical gas is the velocity of the
flow of particles with ith grade in the coordinate system moving with the mean
mass velocity of the gas

i ðr; tÞ ¼
vi  v0 :
V ð5:9Þ

To simplify some expressions a quantity called the reduced speed Wi is used, it

is defined as
rffiffiffiffiffiffiffiffi
mi
Wi ¼ Vi : ð5:10Þ
2kT

The reduced initial relative velocity c is also used:

rffiffiffiffiffiffiffiffi
l
c¼ g; ð5:11Þ
2kT

where l—the reduced mass of the two colliding particles.

The temperature in the kinetic theory is defined in terms of the average kinetic
energy of all the particles of the thermal motion:
Z
3 1X 1
kT ¼ mi v2i f ðvi Þdvi ; ð5:12Þ
2 n i 2
P
where n ¼ ni —the total concentration of particles.
i
The mass transfer is expressed by the vector of mass flux density:
X X Z X
Wm ¼ wmi ¼ mi Vi fi dVi ¼ ni m i V i : ð5:13Þ
i i i

The momentum transfer can be described as

X X Z X
WP x ¼ WPxi ¼ mi Vix Vi fi dVi ¼ ni mi Vix Vi : ð5:14Þ
i i i

The vector WPxi has components proportional to Vix Vix , Vix Viy , Vix Viz . Similar
vectors of flows can be obtained for pulses of y- and z-components. These three flux
vectors completely characterize the transfer of momentum. Nine components of
these vectors form a symmetric tensor pi of second range:
R
ðpi Þxx ¼ mi Vix Vix fi dVi ¼ ni mi Vix Vix ;
ðpi Þxy ¼ ðpi Þyx ¼ ni mi Vix Viy ; ð5:15Þ
. . .:
128 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

Symbolically, the tensor pi, characterizing the partial pressure of the jth com-
ponent of the chemical gas, can be written as

pi ¼ ni mi Vi Vi : ð5:16Þ

Sum of partial pressures tensors for all sorts of particles can be described by gas
pressure tensor of mixture:
X X
p¼ pi ¼ ni mi Vi Vi : ð5:17Þ
i i

The physical meaning of the pressure tensor is that it represents the momentum
flux of the gas. The individual components of the pressure tensor are expressed in
the following: the diagonal elements pxx, pyy, pzz are equal to the normal stresses;
e.g., pxx is equal to the force acting on a unit area normal to x-axis, etc.
The voltage or pressure is received as the pressure tensor components that can be
measured by devices moving in the flow with velocity v0. The pressure measured in
the stream using the fixed gauge depends on v0 and on the location of manometer.
In a system in equilibrium, the shear stresses are equal to zero, and the normal
stresses are equal to each other. In this case, the force acting on any selected surface
element in the gas is constant and independent of the orientation of the surface
perpendicular to it, i.e.,

pxx ¼ pyy ¼ pzz ¼ p;

ð5:18Þ
pxy ¼ pyz ¼    ¼ 0;

where p—the hydrostatic pressure for the equilibrium of the system.

The transfer of kinetic energy for the molecules of ith grade is characterized by
the vector
Z
1 1
Wq i ¼ mi Vi2 Vi fi dVi ¼ ni mi Vi2 Vi ¼ qi : ð5:19Þ
2 2

The sum of such vectors for all the components of the gas mixture is equal to the
heat flow density vector:
X X1
q¼ qi ¼ ni mi Vi2 Vi : ð5:20Þ
i i
2

The physical meaning of the components qx, qy, qz of the vector of heat flux
density q is that they represent the flux density of the kinetic energy in the direction
of the axis x, y, and z, respectively.
The transfer of energy and momentum by collisions of particles (internal sources
of transfer) is not taken into account in the above equations’ transfer streams.
5.2 The Transport Equations 129

The general transfer equation is

Z

@fi @fi @fi

wi þ vi  þ Fi  dvi
@t @r @vi
X ZZZZ   ð5:21Þ
¼ wi fi0 fj0  fi fj gij bdbdedvi dvj ;
j

therefore,
Z

@fi wi @ @
þ  ðvi fi wi Þ þ  ðFi fi wi Þ dvi
@t @r @vi
Z

@wi @wi @wi

 fi þ vi  þ Fi  dvi
@t @r @vi
    ð5:22Þ
@ ni wi @   @wi @w @w
¼ þ  ni w i vi  ni þ vi  i þ Fi  i
@t @r @t @r @vi
X ZZZZ  
¼ wi fi0 fj0  fi fj gij bdbdedvi dvj :
j

Equation (5.22) is known as a general Enskog transport equation for wi ,

corresponding to the particles of the ith grade.
The equations like (5.22) for any wi are little used because of the presence of
very difficult integrals in their right parts, but if wi —the molecular mass (for par-
ticles) of ith type, given that the mass of individual particles is not changed by the
collision, these equations can be written as
X ZZZZ  
mi fi0 fj0  fi fj gij bdbdedvi dvj ¼ 0: ð5:23Þ
j

The mass, momentum, and energy of the particles are conserved in collisions, we
can show that when wi ¼ mi , mi Vi , and mi Vi2 =2
X ZZZZ  
wi fi0 fj0  fi fj gij bdbdedvi dvj ¼ 0: ð5:24Þ
j

Relations (5.23) and (5.24) make it possible to simplify the general transport
equation. The validity of them can be set as follows. Integral
X ZZZZ  
wi fi0 fj0  fi fj gij bdbdedvi dvj ð5:25Þ
j
130 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

is equal to the integral

X ZZZZ  
w0i fi fj  fi0 fj0 g0ij b0 db0 dedv0i dv0j ; ð5:26Þ
j

that is record to reverse collision.

Since

gij ¼ g0ij ; b ¼ b0 ; dvi dvj ¼ dv0i dv0j ; ð5:27Þ

the integral (5.25) can be represented as

ZZZZ  
 w0i fi0 fj0  fi fj gij bdbdedvi dvj : ð5:28Þ

Consequently, we can write

X ZZZZ  
wi fi0 fj0  fi fj gij bdbdedvi dvj
j
ZZZZ ð5:29Þ
1   0 0 
¼ wi  w0i fi fj  fi fj gij bdbdedvi dvj
2

In the case of wi ¼ mi (5.29) Eq. (5.23) follows. Equality wi  w0i ¼ 0 expresses

the fact that the masses of the individual particles are not changed during the
collisions.

5.2.1 Principle of the Equilibrium Systems

If we consider only elastic collisions—the system without changing the internal

energy and conversion of the external energy into internal one and back, we can
distinguish a hypothetical model in a conservative enclosure system, where the
particle of the ith grade is scattered by the particle of the jth grade, the momentum
ΔP and the kinetic energy ΔE are losing and this particle of the ith grade is deflected
by an angle h in the area of solid angle dX. If the particle of ith kind is moving back
from the final state and is scattered by the particle of jth grade and obtain the
momentum ΔP and the kinetic energy ΔE, it returns to the original state:
ZZZZ
1X   
wi  w0i fi0 fj0  fi fj gij bdbdedvi dvj
2 i;j
ZZZZ    ð5:30Þ
1X
¼ wj  w0j fi0 fj0  fj fi gji bdbdedvj dvi :
2 i;j
5.2 The Transport Equations 131

Replacing the particle of the ith grade to the particle of the jth grade, it can be
written as
ZZZZ  
1X
wi fi0 fj0  fi fj gij bdbdedvi dvj
2 i;j
ZZZZ    ð5:31Þ
1X
¼ wi þ wj  w0i  w0j fi0 fj0  fi fj gij bdbdedvi dvj :
4 i;j
 
The sum of the collision invariants wi þ wj  w0i  w0j ¼ 0. Therefore,
Eq. (5.24) is observed at wi ¼ mi , mi Vi , and mi Vi2 =2.
The different transport equations can be obtained from the total Enskog equa-
tion (5.22), assuming wi ¼ mi , mi Vi , and mi Vi2 =2. In particular, wi ¼ 1, we have

@ni @
þ  ni vi ¼ 0: ð5:32Þ
@i @r

Multiplying (5.32) on mi and summed over all types of particles, we obtain the
equation of continuity for the entire gas.

@q @
þ  qv0 ¼ 0: ð5:33Þ
@t @r

If we substitute wi ¼ mi Vi in (5.22) and if we sum over all types of particles, the

collision integral in the right-hand side of the equation disappears. After some
transformations, we have the following equation of the gas motion:

@v0 @ 1 @ 1X
þ v0  v0 ¼   pþ ni Fi : ð5:34Þ
@t @r q @r q i

Substituting wi ¼ mi Vi2 =2 in the general transport equation with a conversion

that are similar to that performed in the derivation of the previous equation, we get
" #
X1 @ni Vi2 @ @Vi2 @Vi2 @Vi2
mi þ  ni Vi vi  ni
2  ni vi   ni Fi  ¼ 0: ð5:35Þ
i
2 @t @r @t @r @vi

Introducing the pressure tensor (5.16) and the vector of heat flux density (5.20),
the energy balance equation can be written as

@ ^ @ X
qUtrans þ ^ trans v0 þ @  q þ p : @ v0 
 qU ni mi Fi  Vi ¼ 0; ð5:36Þ
@t @r @r @r i
132 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

where U^ trans —the internal energy of the gas per unit of mass that is equal to the
energy of translational motion of the particles,
X1
^ trans ¼ 1
U ni mi Vi2 : ð5:37Þ
q i 2

In (5.36) and then mark « : » means the dual tensor multiplication.

The internal energy is equal to the total energy per unit of the gas mass in the
coordinate system moving with the mean mass velocity v0 (kinetic energy of the
flow is not included in this energy), without taking into account the potential energy
of the gas in an external field. Using ratios for U ^ trans and T, Eq. (5.36) can be
transformed to
 
3 @T @T @ @
nk þ v0  ¼   q  p : v0
2 @t @r @r @r
! ð5:38Þ
X 3 @ X
þ ni mi Fi  Vi þ kT  ni Vi :
i
2 @r i

The resulting equation is valid for a gas mixture consisting of particles that have no
internal degrees of freedom.

5.3 Solution of the Boltzmann Kinetic Equation

by Chapman–Enskog Approach

The Boltzmann kinetic equation (5.1) is a nonlinear integral–differential equation,

and its exact analytical solution is very difficult. Chapman–Enskog approach
received the widespread among the existing methods for solving the Boltzmann
kinetic equation to determine the transfer coefficients and the essence of its method
is a logical expansion of the distribution function on the small parameter.
According to the method of Chapman–Enskog the solution of (5.1) for the
distribution function fi of ith grade of particles is sought in the form
fi ðvi Þ ¼ fi0 ðvi Þ½1  vi Ui ðvi Þ, where Ui ðvi Þ—a vector of the perturbation that is
proportional to the external electric field E. Then the unknown vector Ui ðvi Þ is
decomposed into a series in Sonine polynomials and this expansion of the distri-
bution function is inserted into the Boltzmann kinetic equation (5.1). Executing a
sequence of approximate solutions with different initial and boundary conditions,
such as conditions for normalization of each component of the gas, we get various
approximations of the distribution function of particles. Knowing the distribution
function, we can identify the different streams of particles and the corresponding
5.3 Solution of the Boltzmann Kinetic Equation by Chapman–Enskog Approach 133

transport coefficients that are often expressed through the collision integrals, taking
into account the interaction of groups of particles:
sffiffiffiffiffiffiffiffi Z1
kT
Xðl;sÞ ¼ expðc2 Þc2s þ 3 QðlÞ ðgÞdc;
2pl
0

where g—the relative velocity of the particles in the system of the mass center;
c2 ¼ 2kT
l 2
g —square of the reduced relative velocity for the particles of ith and jth
grades with the reduced mass l; QðlÞ ðgÞ—the effective cross section for collisions
of particles of the ith and jth grades, calculated by Eq. (2.8).
The final result of the method is that we can express all the transport coefficients
through the system of integrals Xðl;sÞ . With the help of these integrals the dynamics
of molecular collisions and, therefore, the law of intermolecular forces are taken
into account.
Solutions in accordance with Chapman–Enskog approach, for example, show
that in a first approximation, to calculate the diffusion coefficient it should be taken
that l = 1, s = 1, and for calculation the viscosity and thermal conductivity—l = 2,
s = 2. The difference between the values of integrals is not very large. To calculate
the transport coefficients of higher approximations it is necessary to know the
collision integrals for other values of l and s.
The above-mentioned Chapman–Enskog approach is based on various
assumptions that to some extent limit the application of the results. Consider each
constraint separately and discuss the conditions when these limitations become
significant.
Chapman–Enskog approach consider only the binary collisions, it does not apply
at high densities, when the triple collisions begin to play a significant role. This
method makes it possible to introduce the distribution function in a series. The first
approach that is considered here is rightly in the case where all the gradients of
physical quantities are small. Higher approximations give correction for large
gradients. In a first approximation, the flows are proportional to the first derivative
of the density, velocity, and temperature; the resulting equations that describe the
change in density, velocity, and temperature in time (transfer equation) are called
the Navier–Stokes equations. Higher approximations give the terms proportional to
higher powers of derivatives and derivatives of less high physical quantities. The
transport equations, corresponding to the second approximation are known as the
equations of Barnett. Quantitatively, the first approximation is valid when the rel-
ative change in density, velocity, and temperature in the region of the mean free
path is small compared to unity.
Using Chapman–Enskog approach it is assumed that the volume of the vessel
containing the gas is large compared to the mean free path of the particles, so the
surface layers take the negligible proportion of the total volume. At very low
densities the particles more often collide with the walls of volume containing the
gas than with each other; therefore, in this case, the mechanism of establishment of
134 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

the local equilibrium in the gas is ineffective. In such a gas the concept of local
density, velocity, and temperature lose their original meaning, and the gas even
approximately is not a continuous medium, but rather exhibits the properties of the
discrete environment. Gas in highly diluted conditions is known as Knudsen, and
the behavior of such a gas is well known.
Strictly speaking, Chapman–Enskog approach is only applicable to monatomic
gases (molecules without internal degrees of freedom, for them the interaction
potential is spherically symmetric). Inelastic collisions are possible between
molecules with internal degrees of freedom. In these collisions the kinetic energy
changes, and the mass and momentum are clearly preserved. Therefore, the vis-
cosity and diffusion insignificantly depend on the internal degrees of freedom, and
the theory of atomic gases can be successfully applied to polyatomic molecules, if
they are not too different from the spherical ones.
Consider the case of a gas mixture of two components 1 and 2 with the distri-
bution functions f1 and f2, respectively. In Chapman–Enskog approach, the distri-
bution functions are presented in the form of series

ð0Þ ð1Þ ð2Þ

f1 ¼ f1 þ f1 þ f1 þ    ; ð5:39Þ

ð0Þ ð1Þ ð2Þ

f2 ¼ f2 þ f2 þ f2 þ    ; ð5:40Þ

where the zero-order approximation corresponds to the local equilibrium of

Maxwellian distribution:
 m 3=2  m 
ð0Þ 1 1
f1 ¼ 4pV12 n1 exp  ð v1  v0 Þ 2 ; ð5:41Þ
2pkT 2kT
 m 3=2  m 
ð0Þ 2 2
f2 ¼ 4pV22 n2 exp  ð v2  v0 Þ 2 ; ð5:42Þ
2pkT 2kT

where m1, m2—particles masses of the gas components: k—the Boltzmann con-
stant; n1, n2—the concentrations of the gas particles: T—temperature of the mix-
ture; v0 —the speed of the gas mixture; v0 ¼ M1 x1 v1 þ M2 x2 v2 ; v1 , v2 —the velocity
of gas components; M1 ¼ m1 =m0 , M2 ¼ m2 =m0 ; m0 ¼ m1 þ m2 ; x1 ¼ n1 =n,
x2 ¼ n2 =n—ratio of the concentrations of the components (molar fractions);
n ¼ n1 þ n2 —the concentration of the mixture.
Chapman–Enskog approach for determination of the transport coefficients of the
Boltzmann kinetic equation is described in detail in the paper of Chapman and
Cowling [1].
The boundary condition of BKE has the form
Z
fi dvi ¼ ni ; ð5:43Þ
5.3 Solution of the Boltzmann Kinetic Equation by Chapman–Enskog Approach 135

X Z
mi vi fi dvi ¼qv0 ; ð5:44Þ
i
Z
1X 3
mi ðvi  v0 Þ2 fi dvi ¼ nkT: ð5:45Þ
2 i 2

With the use of the following additional conditions:

Z
ðrÞ
fi dvi ¼ 0; r ¼ 1; 2; 3; . . .; ð5:46Þ

X Z
ðrÞ
mi vi fi dvi ¼ 0; r ¼ 1; 2; 3; . . .; ð5:47Þ
i
Z
1X ðrÞ
mi ðvi  v0 Þ2 fi dvi ¼ 0; r ¼ 1; 2; 3; . . . ð5:48Þ
2 i

ðrÞ
all the functions fi can be clearly defined. If the obtained distribution functions
satisfy the conditions (5.43)–(5.45), they may be considered as solutions of the
BKE.
If limit the further consideration by the first approximation (up to f(1)), the
distribution function can be written as the sum of the  equilibrium
 distribution
ð0Þ ð1Þ ð1Þ
functions and small deviations from it, that is fi ¼ fi 1 þ Ui . Functions f1
ð1Þ
and f2 take the form

ð1Þ ð0Þ
f1 ¼ f1 U1 ; ð5:49Þ

ð1Þ ð0Þ
f2 ¼ f2 U2 ; ð5:50Þ

where U1 and U2 —the perturbation functions.

For any vectors F and K, we have
ZZZ
 
n2i Ii ðFÞ ¼ fi0 f 0  fi f gbdbdedv
ZZZ ð5:51Þ
ð0Þ  
¼ fi f ð0Þ Fi þ F  Fi0  F 0 gbdbdedv;

ZZZ  
ni nj Iij ðKÞ ¼ fi0 fj0  fi fj gbdbdedvj
ZZZ ð5:52Þ
ð0Þ ð0Þ
¼ fi fj ðK  K 0 Þgbdbdedvj :
136 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

Expressions (5.51), (5.52) are the linear functions, so Ið/ þ wÞ ¼ Ið/Þ þ IðwÞ
and Iða/Þ ¼ aIð/Þ regardless of the values of the constants (a—the arbitrary
constant). The right side of the BKE can be linearized as follows:

ð1Þ
D1 ¼ n21 I1 ðU1 Þ  n1 n2 I12 ðU1 þ U2 Þ;
ð1Þ
ð5:53Þ
D2 ¼ n22 I2 ðU2 Þ  n1 n2 I21 ðU1 þ U2 Þ;

wherein

ð0Þ ð1Þ ðr1Þ  

ðrÞ @ r1 fi @ r2 fi @ 0 fi @ @ ðr1Þ
Di ¼ þ þ  þ þ vi  þ Fi  f : ð5:54Þ
@t @t @t @r @vi i

The left part of the BKE can be written as

 

ð1Þ ð0Þ 5 1
o
D1 ¼ f1  V1  r lnðTÞ þ x1 d12  V1 þ 2 W1 W1 : rv0 ;
W12 ð5:55Þ
2
 

ð1Þ ð0Þ 5 1
o
D2 ¼ f2 W2  V2  r lnðTÞ þ x2 d21  V2 þ 2 W2 W2 : rv0 ;
2
ð5:56Þ
2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi o o
where Wi  mi =2kT Vi —the reduced speed, Vi  vi  v0 , W1 W1 and W2 W2 —
the dual tensors.
The value d12 can be determined by any of the following expressions [1–3]:
 

q1 q2 1 1
d12 ¼ F2  rp2  F1  rp1 ; ð5:57Þ
qp q2 q1

n1 n2 ðm2  m1 Þ q q
d12 ¼ rx1 þ rlnðpÞ  1 2 ðF1  F2 Þ; ð5:58Þ
nq qp

where q1 and q2 —the mass density of the components of mixture; q ¼ q1 þ q2 —

the mass density of mixture; p1 and p2 —the partial pressures of the components;
p ¼ p1 þ p2 —the pressure of mixture.
Since rx2 ¼ rx1 , xi ¼ nni , it is possible to show that d21 ¼ d12 . Functions U1
and U2 then can be rewritten as [1, 3, 4]

@ lnðTÞ @
U1 ¼ A1   D1  d12  2B1 : v0 ;
@r @r ; ð5:59Þ
@ lnðTÞ @
U2 ¼ A2   D2  d12  2B2 : v0 ;
@r @r
5.3 Solution of the Boltzmann Kinetic Equation by Chapman–Enskog Approach 137

where Ai and Di —vectors of velocity, Bi —tensors without sources such that

o
Ai ðVi Þ ¼ Vi Ai ðVi Þ; Di ðVi Þ ¼ Vi Di ðVi Þ; Bi ðVi Þ ¼ Vi Vi Bi ðVi Þ; ð5:60Þ
o
where the double tensor is defined as Vi Vi ¼ Vi Vi  13 Vi2 U; U—the unit tensor.
Making the appropriate transformations, we obtain the following pairs of
equations for the coefficients Ai ,Di , and Bi of the disturbance function U:
 
ð0Þ 5
f1 W1  V1 ¼ n21 I1 ðA1 Þ þ n1 n2 I12 ðA1 þ A2 Þ;
2
ð5:61Þ
2
 
ð0Þ 5
f2 W2  V2 ¼ n22 I2 ðA2 Þ þ n1 n2 I21 ðA1 þ A2 Þ;
2
ð5:62Þ
2
ð0Þ
x1
1 f1 V1 ¼ n1 I1 ðD1 Þ þ n1 n2 I12 ðD1 þ D2 Þ;
2
ð5:63Þ

ð0Þ
x1
2 f2 V2 ¼ n2 I2 ðD2 Þ þ n1 n2 I21 ðD1 þ D2 Þ;
2
ð5:64Þ

ð0Þ o
f1 W1 W1 ¼ n21 I1 ðB1 Þ þ n1 n2 I12 ðB1 þ B2 Þ; ð5:65Þ

ð0Þ o
f2 W2 W2 ¼ n22 I2 ðB2 Þ þ n1 n2 I21 ðB1 þ B2 Þ: ð5:66Þ

Coefficients Ai and Di have to satisfy the following boundary conditions:

Z Z
ð0Þ ð0Þ
f1 m1 V1  A1 dv1 þ f2 m2 V2  A2 dv2 ¼ 0; ð5:67Þ

Z Z
ð0Þ ð0Þ
f1 m1 V1  D1 dv1 þ f2 m2 V2  D2 dv2 ¼ 0: ð5:68Þ

Thus, the distribution function can be written as

ð0Þ o
f1 ¼ f1 1  A1 ðV1 ÞV1  r lnðTÞ  D1 ðV1 ÞV1  d12  2B1 ðV1 Þ V1 V1 : rv0 ;

ð5:69Þ

ð0Þ o
f2 ¼ f2 1  A2 ðV2 ÞV2  r lnðTÞ  D2 ðV2 ÞV2  d12  2B2 ðV2 Þ V2 V2 : rv0 :

ð5:70Þ
138 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

Taking into account (5.61)–(5.66) denote

Z   Z  
ð0Þ 5 ð0Þ 5
n2 fA; ag ¼ f1 W12  V1  a1 dv1 þ f2 W22  V2  a2 dv2 ; ð5:71Þ
2 2
Z Z
ð0Þ ð0Þ
n2 fD; ag ¼ x1
1 f 1 V 1  a 1 dv 1  x 1
2 f2 V2  a2 dv2 ; ð5:72Þ

Z o
Z o
ð0Þ ð0Þ
n fB; bg ¼
2
f1 W1 W 1 : b1 dv1 þ f2 W2 W2 : b2 dv2 ; ð5:73Þ

where a—the vector having two vector components—vectors a1 and a2 , that are
defined in the phase velocity space; b—the double speed tensor; fF; Gg—the
parenthesis integrals that are defined as follows:

n2 fF; Gg  n21 ½F; G1 þ n1 n2 ½F1 þ F2 ; G1 þ G2 12 þ n22 ½F; G2 ; ð5:74Þ

where
Z Z
½F; G1  G1 I1 ðF Þdv1 ;½F; G2  G2 I2 ðF Þdv2 ; ð5:75Þ

Z Z
½F1 þ G2 ; H1 þ K2 12  F1 I12 ðH1 þ K2 Þdv1 þ G2 I21 ðH1 þ K2 Þdv2 : ð5:76Þ

Because the symmetry and linearity have collision integrals,½F; G1 ¼ ½G; F 1 ,
½F; G2 ¼ ½G; F 2 , and ½F1 þ G2 ; H1 þ K2 12 ¼ ½H1 þ K2 ; F1 þ G2 12 , and so fF; Gg ¼
fG; F g, fF; G þ H g ¼ fF; Gg þ fF; H g, and fF; aGg ¼ afF; Gg (a—the arbitrary
constant).
These relations need to find the transport coefficients. The coefficients A, B, and
D in the future can be calculated using the expansion by Sonine polynomials:

X
n ðm þ n Þ X
n
ðm þ nÞ!
ðnÞ np
Sm ðxÞ ¼ ðxÞp ¼ ðxÞp
p¼0
ð pÞ!ðn  pÞ! p¼0
ð p Þ!ð n  p Þ! ð m þ p Þ!
ð5:77Þ
X
n
C ð m þ n þ 1Þ
¼ ðxÞp ;
p¼0
Cð p þ 1 ÞCð n  p þ 1 ÞCð m þ p þ 1 Þ

ð0Þ ð1Þ
wherein Sm ðxÞ ¼ 1 and Sm ðxÞ ¼ m þ 1  x; Cð xÞ—the gamma function. Thus,
factors A, B, and D are unnormalized functions of the Sonine polynomials, that
were originally used in the gases kinetic theory of D. Barnett in 1935. Note also the
property of orthogonal Sonine polynomials:
5.3 Solution of the Boltzmann Kinetic Equation by Chapman–Enskog Approach 139

Z1
ðpÞ ðqÞ Cðm þ p þ 1Þ
expðxÞSm ð xÞSm ð xÞxm dx ¼ dp;q ; ð5:78Þ
p!
0

where dp;q —the delta function of Kronecker.

5.4 Determination of Viscosity

During the determination of viscosity it can be noted that the pressure of the
two-component mixture at least order of approximation is reduced to a simple
hydrostatic pressure of the mixture. For a first-order approximation to this system, it
is necessary to add the pressure tensor pð1Þ , defined as [1, 2]
 ð1Þ  ð1Þ
pð1Þ ¼ n1 m1 V1 V1 þ n2 m2 V2 V2 ð5:79Þ

This tensor characterizes the deviation of the system pressure from the hydro-
static one. In the notation of approximation functions of the first order (5.69) and
(5.70) it has the form
Z Z
ð0Þ ð0Þ
pð1Þ ¼ 2m1 f1 V1 V1 ðB1 : rv0 Þdv1  2m2 f2 V2 V2 ðB2 : rv0 Þdv2
 Z Z 
2 ð0Þ o ð0Þ o
¼ m1 f1 V1 V1 : B1 dv1 þ m2 f2 V2 V2 : B2 dv2 S
5
4
¼  kn2 T fB; BgS;
5
ð5:80Þ

where S—the rate-of-shear tensor, element of this tensor is determined by equation

 
1 @v0b @v0a 1
Sab ¼ þ  dab ðr  v0 Þ:
2 @a @b 3

If we write an expression for the viscosity

2
l ¼ kn2 T fB; Bg; ð5:81Þ
5

then by substituting it into (5.80), the it takes the form

pð1Þ ¼ 2lS: ð5:82Þ

140 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

Then (5.82) will be the expression for the tensor of viscous pressure in any
environment.
The pressure tensor p is defined as
Z o
p ¼ m f VV dv: ð5:83Þ

Using the conditions

Z
f ð0Þ AðW ÞW 2 dv ¼ 0 ð5:84Þ

the pressure tensor can be rewritten as

p ¼ pU  2lS; ð5:85Þ

where p—hydrostatic pressure U—the unit tensor.

For a monatomic gas the viscosity l is defined as
Z
2kT o o
l¼ f ð0Þ WW : WW BðW Þdv; ð5:86Þ
5n

or in another form

2kT
l¼ ½BðWÞ; BðWÞ: ð5:87Þ
5

In Eq. (5.87), the term in brackets is called the square bracket integrals,
defined as
Z
½BðWÞ; BðWÞ  BðWÞ : I ðBðWÞÞdv: ð5:88Þ

Consequently, the viscosity can be calculated as

Z Z1
4kT ð0Þ 16kT  
l¼ f W BðW Þdv ¼
4
exp W 2 W 6 BðW ÞdW: ð5:89Þ
15n 15p1=2
0

Using the Sonine polynomial expansion, the bracket integral and viscosity of the
gas can be determined.
5.4 Determination of Viscosity 141

5.4.1 Determination of the Elements in the Bracket Integral

{B, B} by Sonine Polynomials

Chapman–Enskog coefficients B1 and B2 depending on the viscosity of the

two-component mixture can be expressed as a sum of

X
þ1
ðpÞ
X
þ1
ðpÞ
B1 ¼ bp b1 ; B2 ¼ bp b2 ; ð5:90Þ
p¼1 p¼1
p6¼0 p6¼0

where bp is the coefficient that is used for both expansions, the member p = 0 is not
included in the sum and functions bðpÞ are provided in two phase spaces of the
velocity:
9
ðpÞ ð0Þ ðpÞ ð0Þ >
b1 ¼ b1 ¼ 0; b2 ¼ b2 ¼ 0; ðp ¼ 0Þ; >
=
ðpÞ ðpÞ
b1 ¼ 0; b2 ¼ 0; ðp\0Þ; ð5:91Þ
o o >
>
ðp [ 0Þ: ;
ðpÞ ðp1Þ ðpÞ ðp1Þ
b1 ¼ S5=2 ðW Þ W1 W1 ; b2 ¼ S5=2 ðW Þ W2 W2 ;

From Eq. (5.73) the following expression can be obtained:

n o
B; bðqÞ ¼ bq ; ð5:92Þ

wherein
Z o
Z o
ð0Þ ð0Þ
n2 bq ¼ f1 W1 W1 : b1 dv1 þ f2 W2 W2 : b2 dv2 : ð5:93Þ

Integrating (5.93), we have

5n1 5n2
b1 ¼ ; b1 ¼ ; bq ¼ 0ðq 6¼ 1Þ ð5:94Þ
2n2 2n2

From (5.90)–(5.93) we obtain the expression

X
þ1
bp bpq ¼ bq ; ð5:95Þ
p ¼ 1
p 6¼ 0
142 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

where
n o
bpq ¼ bðpÞ ; bðqÞ : ð5:96Þ

If the values bpq are known, all the values bp can defined by solving the system of
algebraic equations (5.95).
Because bq ¼ 0ðq 6¼ 1Þ (5.94), the expressions for the bracket integrals are
X n o
fB; Bg ¼ bp B; bðpÞ ¼ b1 b1 þ b1 b1 : ð5:97Þ
p

In accordance with the work of Chapman and Cowling [1] and with the des-
ignations adopted by them, we note that for two-component gas mixture the vis-
cosity approximation of the order m can be expressed by the Sonine polynomials:
 
ðmÞ ðmÞ
½lm ¼ p x1 b1 þ x2 b1 ; ð5:98Þ

ðmÞ ðmÞ
where b1 and b1 —the expansion coefficients are determined by solving a system
of algebraic equations (5.94) and (5.95).
Using the matrix notation this system of equations can be written as

Bb ¼ b; ð5:99Þ

where
2 3 2 3
bmm  bm1 bm1    bmm bm
6 .. .. .. .. .. .. 7 6 .. 7
6 . . . . . . 7 6 . 7
6 7 6 7
6b  b11 b11    b1m 7 6b 7
B ¼ 6 1m 7; b ¼ 6 1 7 and
6 b1m  b11 b11    b1m 7 6 b1 7
6 . .. .. .. .. .. 7 6 . 7
4 . 5 4 . 5
. . . . . . .
bmm  bm1 bm1    bmm bm
2 3 ð5:100Þ
0
.
6 .. 7
6 7
6b 7
6 1 7
b¼6 b 7
6 1 7
6 . 7
4 .. 5
0

By increasing the order of approximation (expansion) m in the matrixes there are

additional rows and columns.
ðmÞ ðmÞ
Thus, for obtaining the expansion coefficients b1 and b1 or viscosities with
this order of decomposition it is necessary only to form the matrix 2m  2m (5.100),
5.4 Determination of Viscosity 143

and invert it. The elements of the matrix B in (5.100) are obtained from the com-
binations of bracket integrals containing relevant Sonine polynomials. For all
(p, q) it is easy to show that bpq ¼ bqp and it is possible to imagine that

ðp1Þ   o ðq1Þ   o
bpq ¼ bqp ¼ x21 S5=2 W12 W1 W1 ; S5=2 W12 W1 W1

1 ð5:101Þ
ðp1Þ  2  o ðq1Þ  2  o
þ x1 x2 S5=2 W1 W1 W1 ; S5=2 W1 W1 W1 ;
12

ðp1Þ   o ðq1Þ   o
bpq ¼ bqp ¼ x1 x2 S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 ; ð5:102Þ
12

ðp1Þ   o ðq1Þ   o
bpq ¼ bqp ¼ x1 x2 S5=2 W22 W2 W2 ; S5=2 W12 W1 W1 ; ð5:103Þ
21

ðp1Þ   o ðq1Þ   o
bpq ¼ bqp ¼ x22 S5=2 W22 W2 W2 ; S5=2 W22 W2 W2

2
ð5:104Þ
ðp1Þ  2  o ðq1Þ  2  o
þ x1 x2 S5=2 W2 W2 W2 ; S5=2 W2 W2 W2 :
21

Analyzing the expression for the bracket integrals, we can see that Eqs. (5.103)
and (5.104) are identical to Eqs. (5.102) and (5.101), and have the difference only
in the indexes that indicate the components of the mixture. Thus, in general, a
complete solution of the BKE using Chapman–Enskog approach for the viscosity of
two-component gas mixture requires the determination of only three types of
bracket integrals:

ðpÞ   o ðqÞ   o
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ; ð5:105Þ
1

ðpÞ   o ðqÞ   o
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 ; ð5:106Þ
12

ðpÞ   o ðqÞ   o
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 : ð5:107Þ
12

5.4.2 The Bracket Integrals

In order to complete the calculation of viscosity

n it is
o necessary to calculate bracket
integrals’ parentheses used in (5.74) for bðpÞ ; bðqÞ : First of all, it is necessary to
144 5 The Boltzmann Kinetic Equation and Calculation of the Transport …
h i h i
ðpÞ ðqÞ ðpÞ ðqÞ
find the square bracket integrals in (5.75), (5.76): b1 ; b1 , b1 ; b1 and
h i 1 12
ðpÞ ðqÞ
b1 ; b2 . Integration is performed for all variable parameters of the collision,
12
information about the intermolecular interaction potential is required for it.
However, the potential affects only the results of the integration of using variables
g and b, they determine the scattering angle h. The other six integrals can be used
without data on the potential. For obtaining these two integrals, we use Eqs. (5.65)
and (5.66). Then we substitute the polynomial approximation of Sonine solutions
B1 and B2 , doing multiplication with additional Sonine polynomials and then
integrating the equation. Use the property of orthogonal Sonine polynomials (5.78),
we obtain a set of equations including matrix coefficients needed for decisions. In
the right-hand sides of equations a constant bq is included, and the left side is a
combination of bracket integrals corresponding to the desired ratio bpq of the matrix
used for the determination of values bp . These values are in turn used to find the
viscosity according to Eq. (5.98). Thus, expressions for the integrals include the
bracket average effective interaction cross sections, the ratio for them were obtained
earlier (Chap. 2).
To find the viscosity the following quantities are used:
o
Wi ¼ Wi Wi : ð5:108Þ

The value Qpq

ij is defined as [5–7]

X
v n h i
Qpq
ij ¼ ni nl dij SnðpÞ ðWi2 ÞWi ; SnðqÞ ðWi2 ÞWi
il
l¼1
h i ð5:109Þ
þ djl SnðpÞ ðWi2 ÞWi ; SnðqÞ ðWl2 ÞWl ;
il

where l and v—conditioned sort of particles of the gas mixture.

Formulas for the bracket integrals in determining viscosity of plasma are given
in Appendices B and C [4, 8].
The elements of matrix that are necessary for the determination of the viscosity
can be found using Qpq
ij :

rffiffiffiffiffiffiffiffiffiffi
2pmi pq
ppq
ij ¼ Q ; ð5:110Þ
kT ij

Equation for the calculation of the viscosity of a sixth-order approximation is

5.4 Determination of Viscosity 145

Fig. 5.2 The calculation

results of the argon viscosity
using Chapman–Enskog
approach (6-order), as well as
their comparison with the data
of other authors and the
results of calculation by the
classical model: (1)—[9, 10];
(2)—[11]; (3)—[12]

 00 pffiffiffiffiffi 
 pij p01 p02 p03 p04 p05 ni mi 
 ij ij ij ij ij
 p10 p11 p12 p13 p14 p15 0 
 ij ij ij ij ij ij
 p20 0 
5ð2pkTi Þ  ij30
 1=2 p21 p22 p23 p24 p25
ij ij ij ij ij

lij 6 ¼   pij p31 p32 p33 p34 p35 0 ; ð5:111Þ
2j pj  40 ij ij ij ij ij 
p p41 p42 p43 p44 p45 0 
 ij ij ij ij ij ij
 p50 p51 p52 p53 p54 p55 0 
 ij ij ij ij ij ij
 n 0 0 0 0 0 0 
j

where
 00 
 pij p01 p02 p03 p04 p05 
 10 ij ij ij ij ij 
p p11 p12 p13 p14 15 
pij 
 ij ij ij ij ij
 p20 p21 p22 p23 p24 p25 
 ij 
j pj ¼  ij30 ij ij ij ij
:
 pij p31 p32 p33 p34 ij 
p35
 40 ij ij ij ij
 pij p41 p42 p43 p44 p45 
 ij ij ij ij ij 
 p50 p51 p52 p53 p54 55 
pij
ij ij ij ij ij

The results of the viscosity calculation by Eq. (5.111) are represented in Fig. 5.2.
For comparison, data of the other authors and the results of calculation by the
classical model are also shown in [10–14].
The results of calculations using the Boltzmann kinetic equation by Chapman–
Enskog approach the best way match to the experimental data than the results of the
calculation by the classical model.
146 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

5.5 Determination the Coefficients of Diffusion,

Thermal Diffusion, and Thermal Conductivity

5.5.1 Calculation of Diffusion and Thermal Diffusion

Coefficients

The diffusion occurs in a gas mixture if components of the mixture have the

1  V
different average speed,
v1 
v2 ¼ V
2 6¼ 0. This implies that
Z Z

1  V
V
2 ¼ n1 f1 V1 dv1  n1 f2 V2 dv2 6¼ 0: ð5:112Þ
1 2

Using the function of the first-order approximation (5.69), (5.70), and

R ð0Þ R ð0Þ
n1
1 f1 V1 dv1 þ n1
2 f2 V2 dv2 ¼ 0, the velocity of diffusion in one direction
can be determined by equations
 Z Z 
1

1 ð0Þ ð0Þ


vd ¼ ð V 1  V2 Þ ¼  n1
1 f 1 V1  D 1 dv 1  n 1
2 f 2 V 2  D 2 dv 2 d12
3 3
 Z Z 

ð0Þ ð0Þ
þ n1 1 f 1 V 1  A 1 dv 1  n 1
2 f 2 V 2  A 2 dv 2 r ln T

1
¼  n½fD; Dgd12 þ fD; Agr ln T :
3
ð5:113Þ

For the case where the gas mixture does not have gradients of temperature and
pressure, and there are no external forces acting on a molecule, the expression for
the diffusion coefficient can be obtained from Eq. (5.113). At the same time, d12 ¼
rx1 ¼ n1 rn1 and r lnðTÞ ¼ 0, so

1

1

vd ¼ ðV 1  V2 Þ ¼  fD; Dgrn1 : ð5:114Þ
3 3

Equation (5.114) can be written as

  n
vd ¼ D12 n1

1
1 rn1  n2 rn2 ¼ D12 rn1 ; ð5:115Þ
n1 n2

where the proportional constant D12 is the diffusion coefficient. And the diffusion
coefficient should be
n1 n2
D12 ¼ fD; Dg: ð5:116Þ
3n
5.5 Determination the Coefficients of Diffusion, Thermal Diffusion … 147

The presence of a temperature gradient means that the second term of

Eq. (5.113) contains a thermal diffusion coefficient, that can be expressed as
[1, 4, 15]
n1 n2
DT ¼ fD; Ag: ð5:117Þ
3n

The above equation allows us to rewrite (5.113) as

 
n2 n2 rT

d ¼ 
v fD12 d12 þ DT r ln T g ¼  D12 d12 þ kT ; ð5:118Þ
n1 n2 n1 n2 T

where kT ¼ DT =D12 ¼ fD; Ag=fD; Dg.

5.5.2 Determination of the Heat Flow

Assume that the particles in the mixture can exchange with the kinetic energy (i.e.,
cannot exchange with the energy stored by the internal degrees of freedom),
whereas the expression for heat flow may be written as [1, 4, 15]
Z Z
1 1
q¼ f1 m1 V12 V1 dv1 þ f2 m2 V22 V2 dv2 ; ð5:119Þ
2 2

wherein
Z  
q 5

1 ð0Þ 5
 ð n1 V 1 þ n2 V 2 Þ ¼  f1 W1  fðV1  D1 Þd12 þ ðV1  A1 Þr ln Tgdv1
2
kT 2 3 2
Z  
1 ð0Þ 5
 f2 W2  fðV2  D2 Þd12 þ ðV2  A2 Þr ln Tgdv2
2
3 2
1 2
¼  n ½fA; Dgd12 þ fA; Agr ln T :
3
ð5:120Þ

From (5.120) q can be obtained in the form of

5
2 Þ  1 kn2 T ½fA; Dgd12 þ fA; Dgr ln T :

1 þ n2 V
q ¼ kT ðn1 V ð5:121Þ
2 3
148 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

After the transformation of (5.121), we have

5
1 þ n2 V
q ¼ kT ðn1 V
2 Þ þ nkT ðV
1  V
2 Þ fA; Dg  krT
2 fD; Dg
ð5:122Þ
5
1 þ n2 V

2 Þ þ nkTkT ðV
1  V

2 Þ;
¼ krT þ kT ðn1 V
2

where

1 fA; DgfD; Ag 1  
k ¼ kn2 fA; Ag  ~ A
¼ kn2 A; ~ ; ð5:123Þ
3 fD; Dg 3

 
~ A
A; ~ ¼ fA; Ag  fA; DgfD; Ag ¼ fA; Ag  2kT fA; Dg þ k2 fD; Dg; ð5:124Þ
fD; Dg T

wherein

~ 1 ¼ A1  fA; Dg D1 ¼ A1  kT D1 ;
A ð5:125Þ
fD; Dg

~ 2 ¼ A2  fA; Dg D2 ¼ A2  kT D2 :
A ð5:126Þ
fD; Dg

If we assume that diffusion does not occur (i.e., V
1 ¼ V
2 ¼ 0), we obtain

q ¼ krT.
The first term on the right-hand side of Eq. (5.122) corresponds to the transfer of
heat due to inequality of the mixture temperature, i.e., in the presence of a tem-
perature gradient. The second term is the case because the heat flow is measured
with respect to v0, this term corresponds to the heat flux carried by the molecules in
the diffusion. The third term determines the effect of thermal diffusion.
The thermal conductivity can also be represented as
Z  
2k 2 T ð0Þ 5
k¼ f W AðW Þ W  dv:
2 2
ð5:127Þ
3mn 2

 
5.5.3 ~ A
Defining Elements of the Bracket Integrals A; ~
and fA; Dg Using the Sonine Polynomials

To find the diffusion coefficient, thermal diffusion, and thermal conductivity, it is

 
necessary to calculate the bracket integrals A;~ A
~ , fA; Dg, and fD; Dg. Using
5.5 Determination the Coefficients of Diffusion, Thermal Diffusion … 149

~ 1, A
Chapman–Enskog approach, the coefficients A ~ 2 , D1 , and D2 can be expanded
as a sum of the Sonine polynomials:

X
þ1 X
þ1 X
þ1 X
þ1
~1 ¼
A
ðpÞ~2 ¼
a p a1 ; A
ðpÞ
ap a2 ; D1 ¼
ðpÞ
dp a1 ; D2 ¼
ðpÞ
dp a2 ;
p¼1 p¼1 p¼1 p¼1

ð5:128Þ

where in accordance with the notations of Chapman and Cowling [1]

ðpÞ ð0Þ 1=2

a1 ¼ a1 ¼ M1 q2 W1 =q ðp ¼ 0Þ; ð5:129Þ

ðpÞ ð0Þ 1=2

a2 ¼ a2 ¼ M2 q1 W2 =q ðp ¼ 0Þ; ð5:130Þ

ðpÞ ðpÞ  
a1 ¼ S3=2 W12 W1 ðp [ 0Þ; ð5:131Þ

ðpÞ
a2 ¼ 0 ðp [ 0Þ; ð5:132Þ

ðpÞ
a1 ¼0 ðp [ 0Þ; ð5:133Þ

ðpÞ ðpÞ  
a2 ¼ S3=2 W22 W2 ðp [ 0Þ: ð5:134Þ

For the determination of coefficients of the expansion ap in (5.128), using (5.71),

we obtain
n o
A; aðqÞ ¼ aq ; ð5:135Þ

wherein
Z   Z  
ð0Þ 5 ðqÞ ð0Þ 5 ðqÞ
n ap ¼
2
f1 W1  V1  a1 dv1 þ
2
f2 W2  V2  a2 dv2 :
2
ð5:136Þ
2 2

Integrating (5.136), we find that ap ¼ 0 when q 6¼ 1 and

rffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffi
15 n1 2kT 15 n2 2kT
a1 ¼  ; a1 ¼ : ð5:137Þ
4 n2 m1 4 n2 m2

Similarly, to determine the coefficients of the expansion dp in (5.128) using

(5.72), we obtain
n o
D; aðqÞ ¼ dq ; ð5:138Þ
150 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

wherein
Z Z
ð0Þ ðqÞ ð0Þ ðqÞ
n2 dq ¼ x1
1 f1 V1  a1 dv1  x1
2 f2 V2  a2 dv2 : ð5:139Þ

Integrating (5.139), we have

rffiffiffiffiffiffiffiffi
3 2kT
d0 ¼ ; dq ¼ 0ðq 6¼ 0Þ: ð5:140Þ
2n m0

By combining (5.128)–(5.134), (5.138)–(5.140), write the system of equations as

X
þ1
dp apq ¼ dq ðq ¼ 0; 1; 2; . . .; 1Þ; ð5:141Þ
p¼1

where
n o
apq ¼ aðpÞ ; aðqÞ ¼ aqp : ð5:142Þ

If the values apq are known, all the values dp can be found by solving a system of
algebraic equations (5.141). Next, to determine the coefficients of the expansion ap
it is necessary to substitute (5.129)–(5.134) in (5.128). Since A ~ ¼ A  kT D then
n o n o n o
~ aðqÞ ¼ A; aðqÞ  kT D; aðqÞ ¼ aq  kT dq :
A; ð5:143Þ

Since dq ¼ 0 ðq 6¼ 0Þ, it can be written as

X
þ1
ap apq ¼ aq ðq ¼ 1; 2; . . .; 1Þ: ð5:144Þ
p ¼ 1
p 6¼ 0

From the system of Eqs. (5.144) and (5.142) the expansion coefficients ap can be
obtained. Based on the above equations, it can be shown that
 
~ A
fD; Dg ¼ d0 d0 ; fD; Ag ¼ d1 a1 þ d1 a1 ; A; ~ ¼ a1 a1 þ a1 a1 : ð5:145Þ

In accordance with the work of Chapman and Cowling [1] as well as designa-
tions adopted by them, we note that for a two-component gas mixture the diffusion,
thermal diffusion coefficients, and thermal conductivity can be expressed in the
5.5 Determination the Coefficients of Diffusion, Thermal Diffusion … 151

form of approximation of any given order m through Sonine polynomials as

follows:
rffiffiffiffiffiffiffiffi
1 2kT ðmÞ
½D12 m ¼ x1 x2 d ; ð5:146Þ
2 m0 0
rffiffiffiffiffiffiffiffi 
5 2kT 1 ðmÞ 1 ðmÞ
½DT m ¼  x1 x2 x1 M1 2 d1 þ x2 M2 2 d1 ; ð5:147Þ
4 m0
rffiffiffiffiffiffiffiffi 
5 2kT 1=2 ðmÞ 1=2 ðmÞ
½km ¼  kn x1 M1 a1 þ x2 M2 a1 ; ð5:148Þ
4 m0
ðmÞ ðmÞ ðmÞ ðmÞ ðmÞ
where d1 ; d0 ; d1 can be found from the system of Eqs. (5.141), a1 ; a1 —
from (5.144).

5.5.4 The Bracket Integrals

To find the diffusion, thermal diffusion coefficients, and thermal conductivity it is

ðmÞ ðmÞ ðmÞ ðmÞ ðmÞ
necessary to determine the coefficients of expansion d1 ; d0 ; d1 and a1 ; a1 .
For this purpose, it is necessary to calculate the bracket integrals, that are repre-
 
sented in (5.74) for aðpÞ ; aðqÞ . In turn, it is necessary to calculate integrals in
h i h i h i
ðpÞ ðqÞ ðpÞ ðqÞ ðpÞ ðqÞ
square brackets (5.75), (5.76), i.e., a1 ; a1 , a1 ; a1 , and a1 ; a2 .
1 12 12
Integrating by all the variables defining the parameters of the collision, and
information about the intermolecular interaction potential are also required.
Using the property of orthogonal Sonine polynomials, a number of equations
containing the matrix coefficients can be obtained, they are necessary for the
determination of solutions. Right side of the equations contain constants dq ; aq , and
the left—a combination of bracket integrals corresponding to the required elements
of the matrix apq that is used to determine the coefficients of the expansion dp ; ap .
Using matrix notation this system of equations can be written as

Dd ¼ d and Aa ¼ a; ð5:149Þ
152 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

where
2 3 2 3
amm  am1 am0 am1    amm dm
6 .. .. .. .. .. .. .. 7 .
6 . 7
6 . . . . . . . 7 6 . 7
6 7 6 7
6 a1m  a11 a10 a11    a1m 7 6 d1 7
6 7 6 7
D¼6
6 a0m  a01 a00 a01    a0m 7 7; d ¼ 6 d0 7;
6 a1m 6 7
6  a11 a10 a11    a1m 7 7 6 d1 7
6 .. .. .. .. .. .. .. 7 6 . 7
4 . 4 . 5
. . . . . . 5 .
a  am1 am0 am1    amm dm
2 mm3
0
6 .. 7
6 . 7
6 7
607
6 7
d ¼ 6 d0 7;
6 7
607
6 . 7
4 .. 5
0
ð5:150Þ
2 3 2 3
amm    am1 am1    amm am
6 .. .. .. .. .. .. 7 .
6 . 7
6 . . . . . . 7 6 . 7
6 7 6 7
6 a1m    a11 a11    a1m 7 6a 7
A¼6 7; a ¼ 6 1 7;
6 a1m    a11 a11    a1m 7 6 a1 7
6 . .. .. .. .. .. 7 6 . 7
4 . 5 4 . 5
. . . . . . .
amm    am1 am1    amm am ð5:151Þ
2 3
0
.
6 .. 7
6 7
6a 7
6 1 7
a¼6 7:
6 a1 7
6 . 7
4 .. 5
0

ðmÞ ðmÞ
In (5.150), (5.151) superscript (m) for the expansion coefficients dp and ap
was omitted, as it was done by Chapman and Cowling.
ðmÞ ðmÞ ðmÞ ðmÞ ðmÞ
Thus, for obtaining the expansion coefficients d1 ; d0 ; d1 and a1 ; a1 and
the coefficients of diffusion, thermal diffusion, and thermal conductivity, respec-
tively, in this order of decomposition, it is necessary only to form matrices
ð2m þ 1Þ  ð2m þ 1Þ and 2m  2m and invert them. Elements of the matrix apq in
(5.151) are obtained by combining the bracket integrals containing relevant Sonine
polynomials. As for any p, q it is easy to show that apq ¼ aqp , then for apq it can be
written [1, 4, 15] as
5.5 Determination the Coefficients of Diffusion, Thermal Diffusion … 153

h  i
ðpÞ   ðqÞ 
apq ¼ aqp ¼ x21 S3=2 W12 W1 ; S3=2 W12 W1
h  i
1
ð5:152Þ
ðpÞ   ðqÞ 
þ x1 x2 S3=2 W12 W1 ; S3=2 W12 W1 ;
12
h  i
ðpÞ   ðqÞ 
apq ¼ aqp ¼ x1 x2 S3=2 W12 W1 ; S3=2 W22 W2 ; ð5:153Þ
12
h  i
ðpÞ   ðqÞ 
apq ¼ aqp ¼ x1 x2 S3=2 W22 W2 ; S3=2 W12 W1 ; ð5:154Þ
21
h  i
ðpÞ   ðqÞ 
apq ¼ aqp ¼ x22 S3=2 W22 W2 ; S3=2 W22 W2
h i 2
ð5:155Þ
ðpÞ  2  ðqÞ  2 
þ x1 x2 S3=2 W2 W2 ; S3=2 W2 W2 :
21

Analyzing the expression (5.152)–(5.155), it can be noted that (5.154) and

(5.155) are identical to (5.153) and 5.152) and have the difference only in the indexes
that indicate the components of mixture. Thus, in general, for a complete solution of
the BKE using Chapman–Enskog approach for determination of the diffusion,
thermal diffusion coefficients, and thermal conductivity of a two-component gas
mixture, the determination of only three types of bracket integrals is required:
h  i
ðpÞ   ðqÞ 
S3=2 W12 W1 ; S3=2 W12 W1 ; ð5:156Þ
1
h  i
ðpÞ   ðqÞ 
S3=2 W12 W1 ; S3=2 W22 W2 ; ð5:157Þ
12
h  i
ðpÞ   ðqÞ 
S3=2 W12 W1 ; S3=2 W12 W1 : ð5:158Þ
12

Let us write the final equations for determination of the diffusion, thermal dif-
fusion coefficients, and thermal conductivity. We obtain
rffiffiffiffiffi
~ pq pq nj mj pq
Q ij ¼ Qij  Q dp0 dq0 ; Wi ¼ Wi ; ð5:159Þ
ni mi ii

where Qpq
ij is defined as [10, 16]

X
v n h i
Qpq
ij ¼ ni nl dij SnðpÞ ðWi2 ÞWi ; SnðqÞ ðWi2 ÞWi
il
l¼1
h i ð5:160Þ
þ djl SnðpÞ ðWi2 ÞWi ; SnðqÞ ðWl2 ÞWl :
il
154 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

Formulas of the bracket integrals in determining the diffusion coefficient, ther-

mal diffusion, electrical conductivity, and thermal conductivity of the plasma are
given in Appendices D and E [4, 15].
The elements of the matrix will look like
rffiffiffiffiffiffiffiffiffiffi
2pmi ~ pq
qpq
ij ¼ Q : ð5:161Þ
kT ij

We write the expression for the definition of thermal conductivity for a

sixth-order approximation.
 00 
 qij q01 q02 q03 q04 q05 0 
 ij ij ij ij ij
 q10 q11 q12 q13 q14 q15 ni 
 ij ij ij ij ij ij
 20 q21 q22 q23 q24 q25 0 
1=2  qij
75k ð2pkTÞ  30 ij ij ij ij ij

½k6 ¼   qij q31 q32 q33 q34 q35 0 ; ð5:162Þ
8 j qj  40 ij ij ij ij ij 
 qij q41 q42 q43 q44 q45 0 
 ij ij ij ij ij
 q50 q51 q52 q53 q54 q55 0 
 ij p
ij ffiffiffiffiffi ij ij ij ij
 0 nj = mj 0 0 0 0 0

where
 00 
 qij q01 q02 q03 q04 q05 
 10 ij ij ij ij ij 
q q11 q12 q13 q14 15 
qij 
 ij ij ij ij ij
 q20 q21 q22 q23 q24 q25 
 ij 
jqj ¼  ij30 ij ij ij ij
:
 qij q31 q32 q33 q34 ij 
q35
 40 ij ij ij ij
 qij q41 q42 q43 q44 q45 
 ij ij ij ij ij 
 q50 q51 q52 q53 q54 q 55 
ij ij ij ij ij ij

Figure 5.3 shows the results of the thermal conductivity calculation by solving
the Boltzmann kinetic equation using Chapman–Enskog approach, the data of other
authors as well (for comparison) were calculated using the classical model using the
expression obtained in [13]. Figure 5.4 shows the results for the components of the
thermal conductivity calculations using the classical model and by solving BKE by
Chapman–Enskog approach. Calculation by the classical model was obtained using
the equations from [13].
The total thermal conductivity is the sum of the components:

ktotal ¼ ka þ ki þ ke þ kR ¼ kc þ kR ; ð5:163Þ

where ka , ki , ke , kR —are, respectively, atomic, ionic, and reactive components of

the thermal conductivity; kc ¼ ka þ ki þ ke —the contact thermal conductivity.
5.5 Determination the Coefficients of Diffusion, Thermal Diffusion … 155

Fig. 5.3 The calculation

results of the thermal
conductivity for electronic in
the Argon by Chapman–
Enskog approach (2- and
6-order) and comparison with
calculation results using the
classical model and
experimental and theoretical
data of other authors: (1)—
[11]; (2)—[9, 10]

Fig. 5.4 Comparison of the

thermal conductivity
calculation results for argon
components obtained by the
classical model and by
Chapman–Enskog approach
(6-order)

Here is the equation for determination of diffusion and thermal diffusion coef-
ficients of the sixth-order approximation [1, 2, 4, 9, 10, 15].
 00 pffiffiffiffiffi 
 qij q01 q02 q03 q04 q05 ni = mi 
 ij ij ij ij ij
 q10 q11 q12 q13 q14 q15 0 
 ij ij ij ij ij ij
 20 0 
 3qð2pkTi Þ  qij
1=2 q21
ij q22
ij q23
ij q24
ij q25
ij

Dij 6 ¼  q30 q31 q32 q33 q34 q35 0 ; ð5:164Þ
2njqj  ij40 ij ij ij ij ij 
 q q41 q42 q43 q44 q45 0 
 ij ij ij ij ij ij
 q50 q51 q52 q53 q54 q55 0 
 ij ij ij ij ij ij
 1=m 0 0 0 0 0 0 
j
156 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

Fig. 5.5 The calculation results of thermal diffusion (a) and diffusion (b) coefficients for Argon,
obtained by Chapman–Enskog approach (6-order): (1)—[9, 10]
 00 
 qij q01 q02 q03 q04 q05 0 
 ij ij ij ij ij
pffiffiffiffiffi
 q10 q11 q12 q13 q14 q15 ni mi 
 ij ij ij ij ij ij
 20 0 
1=2  qij q21 q22 q23 q24 q25
15ð2pkTi Þ  30 ij ij ij ij ij

½DTi 6 ¼  qij q31 q32 q33 q34 q35 0 ; ð5:165Þ
4j qj  40 ij ij ij ij ij 
q q41 q42 q43 q44 q45 0 
 ij ij ij ij ij ij
 q50 q51 q52 q53 q54 q55 0 
 ij ij ij ij ij ij
 n 0 0 0 0 0 0 
j

where jqj is determined as above.

Figure 5.5 presents the calculation results of these coefficients by Eqs. (5.164)
and (5.165). Also for comparison, here are presented the data of [9, 10].

5.6 Determination of the Electrical Conductivity

Electrical conductivity is one of the main parameters that affect the properties of the
plasma. The electrical conductivity is determined by the scattering of electrons on
atoms, ions, and electrons, and can be found using the diffusion coefficients [1, 2].
We write the expression for calculating the electric conductivity [9, 10] as
 00 pffiffiffiffiffi 
 qij q01 q02 q03 q04 q05 ni = mi 
 ij ij ij ij ij
 10 q11 q12 q13 q14 q15 0 
qffiffiffiffiffi  qij ij ij ij ij ij

 3e kTi  qij
2 2p 20
q21 q22 q23 q24 q25 0 
ij ij ij ij ij

rij 6 ¼  q30 q31 q32 q33 q34 q35 0 ; ð5:166Þ
2jqj  ij40 ij ij ij ij ij 
q q41 q42 q43 q44 q45 0 
 ij ij ij ij ij ij
 q50 q51 q52 q53 q54 q55 0 
 ij ij ij ij ij ij
 n 0 0 0 0 0 0 
j

where jqj is determined as above.

5.6 Determination of the Electrical Conductivity 157

Fig. 5.6 The calculation

results of the electrical
conductivity of argon,
obtained by Chapman–
Enskog approach (6-order)
and the classical model, and
compare with those of other
authors: (1)—[9, 10];
(2)—[12]; (3)—[11]

Matrix elements are calculated in the same way as for the coefficient of diffusion,
thermal diffusion, and thermal conductivity (5.159–5.162). For comparison, cal-
culation was performed using equations of classical model [13, 14]. Figure 5.6
shows the results of calculation of the electrical conductivity of the Argon plasma
by Chapman–Enskog approach and their comparison with data of other authors.

References

1. Chapman S., Cowling T.G. The Mathematical Theory of Non-Uniform Gases. Cambridge:
Cambridge University Press, 1970.
2. J. O. Hirschfelder, Ch. F. Curtiss, R. B. Bird. Molecular theory of gases and liquids / John
Wiley and Sons, New York. 2nd Edition: 1964.
3. Ferzinger J.H., Kaper H.G. Mathematical Theory of Transport Processes in Gases,
North-Holland. Amsterdam, 1972.
4. Loyalka S.K., Tipton E.L., Tompson R.V. Chapman-Enskog solutions to arbitrary order in
Sonine polynomials I: Simple, rigid-sphere gas // Phys. Rev. A. 2007. Vol. 379. PP. 417–435.
5. Transport properties in a two temperature plasma: theory and application / Rat V. et al. //
Phys. Rev. E. 2001. Vol. 64.
6. Rat V., Andre P., Aubreton J. et al. // Phys. D: Applied Phys. 2001. Vol. 34. P. 2191–2204.
7. Transport coefficients including diffusion in a two-temperature argon plasma / V. Rat, et al. //
J. Phys. D: Appl. Phys. 2002. Vol. 35. P. 981–991.
8. R.V. Tompson, E.L. Tipton, S.K. Loyalka. Chapman–Enskog solutions to arbitrary order in
Sonine polynomials V: Summational expressions for the viscosity-related bracket integrals //
European Journal of Mechanics B/Fluids 29 (2010) 153–179.
9. Devoto R.S. Transport coefficients of ionized argon // Phys. Fluids. 1973. Vol. 16. № 5.
P. 616–623.
10. Devoto R.S. Transport coefficients of partially ionized argon // Phys. Fluids. 1967. Vol. 10. №
2. PP. 354–364.
11. Kulik P.P. Essays on physics and chemistry of Low-temperature plasma / Ed. by L.S. Polak.
M.: Nauka, 1971.
158 5 The Boltzmann Kinetic Equation and Calculation of the Transport …

12. Boulos M.I., Fauchais P., Pfender E. Thermal plasmas, Fundamentals and Applications. New
York: Plenum Press, 1994. Vol. 1.
13. Shi Nguyen-Kuok. Modeling of equilibrium plasma in RF and Arc plasma torches // Proc. of
the International Scientific and Technical. Conf. “Electrophysical and electrochemical
technology” SPb. 1997. pp. 63-66.
14. Zhdanov V.M. Transport phenomena in multicomponent plasma. M.: Energoizdat 1982.
15. R.V. Tompson, E.L. Tipton, S.K. Loyalka. Chapman–Enskog solutions to arbitrary order in
Sonine polynomials IV: Summational expressions for the diffusion- and thermal
conductivity-related bracket integrals // European Journal of Mechanics B/Fluids 28 (2009),
695–721.
16. Devoto R.S. Simplified expressions for the transport properties of ionized monatomic gases //
Phys. Fluids. 1967. Vol. 10. № 10. P. 2105–2112.
Chapter 6
Numerical Methods of the Plasma Physics

6.1 Basis of the Numerical Methods of the Plasma Physics

6.1.1 Solution of the Generalized Differential Equation

Problems of mathematical simulation of plasma processes are the most difficult

tasks among of the plasma research problems. Numerous physical processes such as
the exchange of energy and momentum, transformation of particles, ionization and
recombination, dissociation of molecules, radiative processes, chemical reactions of
particles activation and deactivation, and others occur simultaneously in plasma. All
of them usually happen with the active interaction of plasma with an external
electromagnetic field.
Numerical methods are the main tools for solving the problems of mathematical
simulation. Effective using of them is possible only in the case of the best-fit of
mathematical models and physical processes occurring in the plasma.
The model description of plasma and plasma processes (Chap. 1) is often
expressed using nonlinear differential equations in partial derivatives. These equa-
tions have a similar structure, and they can be written in the same form of a
Generalized Differential Equations (GDE). For low-temperature plasma the sta-
tionary case is of the greatest interest, when the flow of plasma maintains its state for
a period of time greater than its relaxation time. In this case, the GDE has the form

divðq~
vUÞ ¼ divðCgrad UÞ þ S; ð6:1Þ

where U—a generalized variable; C—coefficients that is included in the flow

C grad U—that can be called as the diffusion coefficient; S—some source function.
The GDE consists of three groups of members: divðq~ vUÞ—convection;
divðC grad UÞ—diffusion; S—the source function. If there are other members in the

© Springer International Publishing Switzerland 2017 159

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_6
160 6 Numerical Methods of the Plasma Physics

differential equation that cannot be expressed in this form, they can be combined
into a common of the source function.
Most designs of the plasma torches channels have a symmetrical cylindrical
shape and it is convenient to solve GDE (6.1) using the two-dimensional cylindrical
coordinate system and this equation can be written in a form



@ ðqvz UÞ 1 @ @ @U 1@ @U
þ ðrqvr UÞ ¼ C þ rc þ S: ð6:2Þ
@z r @r @z @z r @r @r

Presentation the equations of the plasma model in the form of GDE (6.1) or (6.2)
allows the development of general mathematical and software algorithms for
solving this equation, as well as the design of the computational procedures for
modeling of plasma.
The Analytical Solution of One-Dimensional GDE Let us analyze the simplest
solution to the GDE for example on one-dimensional problem of steady processes
of the cold gas purging and heating in a half-space of heated plasma (Fig. 6.1).
Using the model of equilibrium plasma (Sect. 1.2) one-dimensional energy
balance equation describing the heating of the gas has the form


dðqvz T Þ d k dT 1 2 
¼ þ rE  Urad : ð6:3Þ
dz dz cp dz cp

Fig. 6.1 The diagram of the cold gas heating in a half-space of heated plasma
6.1 Basis of the Numerical Methods of the Plasma Physics 161

The one-dimensional GDE with one-dimensional gas flow takes the form


dðqvz UÞ d dU
¼ C þ S: ð6:4Þ
dz dz dz

The one-dimensional continuity equation of the gas is written as

d
ðqvz Þ ¼ 0 or qvz ¼ const: ð6:5Þ
dz

The exact analytical solutions of one-dimensional energy balance equation and

GDE may be obtained only in the case where k=cp = const and Г = const, and with
a relatively simple form of the source function. For a simple analysis of the solution
(6.3) and (6.4) we assume that the source function is equal to zero. Divide the
region of gas heating into plurality sub-domains with the length L (Fig. 6.1).
Because of the smallness of each subregion it can be assumed that the factors
k=cp = const and Г = const. We assume that the unknown variables at the begin-
ning and in the end of each subregion with length L will be T0 ; TL and U0 ; UL
respectively (Fig. 6.1), and then changes TðzÞ and UðzÞ in each subregion with such
boundary condition can be obtained from the analytical solutions
9
expðPeDz=LÞ  1
TðzÞ ¼ T0 þ ðTL  T0 Þ ; ð0  Dz  LÞ; >
>
=
exp Pe  1


 ð6:6Þ
dn T TL  T0 Pe n
PeDz >
>
¼ exp ; ;
dz n exp Pe  1 L L
9
expðPeDz=LÞ  1 >
UðzÞ ¼ U0 þ ðUL  U0 Þ ;>
=
exp Pe  1

n
 ð6:7Þ
d U
n
UL  U0 Pe >
; >
PeDz
¼ exp ;
dz n exp Pe  1 L L

where Pe—the grid Peclet number, defined as the ratio of the intensities of the
convective F and diffusive D fluxes: Pe ¼ DF ¼ C=L
qvz
: Dependencies UðzÞ for different
Pe at constant ratios are shown in Fig. 6.2.
The resulting one-dimensional analytical solution of the GDE with constant
coefficients Г in the form (6.7) can then be used to describe the change in the target
variable between the nodes of the computational grid.
The method of control volume is known as a modern method that combines the
advantages of two well-known techniques that are commonly used in solving dif-
ferential equations in partial derivatives: method of the finite elements and method
of the finite difference. As in method of the finite elements, representation of the
target variable between the nodes of an elementary volume in the form of a
polynomial is used at method of the control volume. However, unlike the method of
finite element, this functional relationship is not arbitrary, and it is obtained by
162 6 Numerical Methods of the Plasma Physics

Fig. 6.2 Dependencies UðzÞ

for different values of the grid
Peclet number Pe

solving a physical problem, such as (6.7). At the same time, the functional rela-
tionship is used to determine the appropriate flow through the faces of the ele-
mentary volume. As a result, we obtain a finite-difference equation in the form of
discrete analogue for the unknown variables in the nodes, as it was done in all
methods of the finite difference.
To use the result of a one-dimensional analytical solutions of GDE in the form
(6.7), we apply the certain grid to computational domain (as in the method of finite
differences), dividing the study area into a plurality of disjoint sub-domains so that
each sub-domain contains only one estimated point. These sub-domains can be
called as the control volume. Thereafter, for each control volume it is necessary to
make a balance of physical flows by integrating the corresponding equation for the
control volume. The result of this operation will be the discrete analogue that
expresses the relevant law of conservation for this control volume at the same way
as a differential equation expresses the law of conservation for the infinitesimal
volume.
Thus, an important feature of the method of control volume is that it contains the
foundations of exact integral conservation of such quantities as mass, momentum,
and energy in any group of the control volume and therefore everywhere in the
studied area, that is very important in numerical simulations of plasma processes.
This property can be manifested using any quantity of the grid points, but not only
at a very large number of them. The solution of equation using an approximate grid
may correspond to the exact integral balances.
The Numerical Solution of the One-Dimensional GDE Consider the
one-dimensional GDE and continuity equation in the form (6.4) and (6.5).
One-dimensional control volume and the location of the grid points are shown in
Fig. 6.3.
6.1 Basis of the Numerical Methods of the Plasma Physics 163

Fig. 6.3 The control volume and grid points for the numerical solution of one-dimensional GDE

The differential equation in the form (6.4) expresses the corresponding law of
conservation in the infinitesimal volume, and to obtain an appropriate balance in the
finite control volume (Fig. 6.3) it is necessary to integrate it across all the control
volume:

Zze Zze
 Zze

dðqvz UÞ d dU
dz ¼ C dz þ Sdz;
dz dz dz
zw zw zw



dU dU
ðqvz Þe Ue  ðqvz Þw Uw ¼ Ce Cw þ SM ðze  zw Þ;
dz e dz w

Here the indices e, w correspond to the values on the respective faces of the control
volume; it is assumed that the source function S in the control volume takes an
average value at the point M.
Substituting the analytical solution (6.7) in the last equation we get
2   3
Pee ðze zM Þ
exp zE zM 1
ðqvz Þe 4UM þ ðUE  UM Þ 5
exp Pee  1


Ce ðUE  UM Þ Pee ðze  zM Þ Pee
 exp
zE  zM zE  zM exp ðPee Þ  1
2   3
exp PewzMðzz
w zW Þ
1
 ðqvz Þw 4UW þ ðUM  UW Þ 5
W

exp Pew  1


Cw ðUM  UW Þ Pew ðzw  zW Þ Pew
þ exp ¼ SM ðze  zw Þ:
zM  zW zM  zW exp Pew  1

Ce Cw
Designating Fe ¼ ðqvz Þe ; Fw ¼ ðqvz Þw ; De ¼ ; Dw ¼ ; Pee ¼
zE  zM zM  zW
Fe Fw
; Pew ¼ ; Last equation can be written as
De Dw
164 6 Numerical Methods of the Plasma Physics

 
UM  UE UM  UW
Fe UM þ  Fw UM  ðUM  UW Þ  ¼ SM ðze  zw Þ:
exp Pee  1 exp Pew  1

And we have the discrete analogue for a one-dimensional GDE

aM UM ¼ aE UE þ aW UW þ bM ; ð6:8Þ

where

aE ¼ De AðPee Þ; aW ¼ Dw AðPew Þ þ Fw ;
aM ¼ aE þ aW ; bM ¼ SM ðze  zw Þ:

The function AðPeÞ ¼ exp Pe Pe1 features the approximating method for the nec-
essary variable in the control volume using this exponential computational scheme.
In the derivation of discrete analogue the continuity equation is used as Fe ¼
Fw ¼ qvz = const: As a result of the approximate solution using an iterative pro-
cedure, this condition can be fulfilled only in the final decision. However, using this
condition makes it possible to obtain the discrete analogue (6.8) with positive
coefficients satisfying the criterion of stability (6.49). In this case, the coefficients
De [ 0; Dw [ 0; Fe ¼ Fw [ 0 and the requirement of the positivity of factors in
discrete analogue [the stability criterion (6.49) will be considered the bottom]
makes it necessary to maintain the function A(Pe) as positive.
The Numerical Solution of the Two-Dimensional GDE Consider the algorithm
for solving the two-dimensional GDE (6.2) using the method of control volume.
A typical two-dimensional control volume for the solution of the GDE (6.2) and the
location of the grid points of the grid are shown in Fig. 6.4.
As it was noted, the GDE (6.2) expresses the corresponding law of conservation
at the infinitesimal volume, and to obtain an appropriate balance in the finite control
volume it should be integrated through all the control volume (Fig. 6.4)

Fig. 6.4 Control volume and

the grid points for solving the
two-dimensional GDE
6.1 Basis of the Numerical Methods of the Plasma Physics 165

Zrn Zze Zze Zrn Zrn Zze


@ ðqvz UÞ 1@ @ @U
dzrdr þ ðrqvr UÞrdrdz  C dzrdr
@z r @r @z @z
rs zw zw rs rs zw
Zze Zrn
 Zze Zrn
1@ @U
 rC rdrdz ¼ Srdrdz
r @r @r
zw rs zw rs

The elementary control volume dV ¼ drdzrdu; du—the single rotation angle. In

the cylindrical axial symmetry the variable Ф is independent of the angle u, Rso the
integration in this elementary volume dV both sides of equation reduces to du .
The resulting integral equation can be divided into three groups
– for the convective fluxes

Zrn Zze Zze Zrn

@ 1@
Ic1 ¼ ðqvz UÞdzrdr; Ic2 ¼ ðrqvr UÞrdrdz;
@z r @r
rs zw zw rs

– for the diffusion fluxes

Zrn Zze
 Zze Zrn


@ @U 1@ @U
Id1 ¼  C dzrdr; Id2 ¼  rC rdrdz;
@z @z r @r @r
rs zw zw rs

– for the source function

Zze Zrn
IS ¼ Srdrdz:
zw rs

Further integration is carried out separately for convective and diffusive fluxes
through the corresponding verges of the control volume e, w, n, s. Note that the
differential of the same order is situated under the first integral of the equations for
convective and diffusive fluxes, so these integrals can only be exactly once inte-
grated. We obtain four subgroups of convective and diffusive fluxes through the
verges of the control volume e, w, n, s in the form:
166 6 Numerical Methods of the Plasma Physics

Zrn Zrn
Ice ¼ ðqvz Þe Ue rdr; Icw ¼  ðqvz Þw Uw rdr;
rs rs
Zze Zze
Icn ¼ ðrqvr Þn Un dz; Ics ¼  ðrqvr Þs Us dz;
zw zw

Zrn
 Zrn

@U @U
Ide ¼  C rdr; Idw ¼ C rdr;
@z e @z w
rs rs
Zze
 Zze

@U @U
Idn ¼  rC dz; Ids ¼ rC dz:
@r n @r s
zw zw

Integrands with subscripts e, w, n, s should be computed along each corre-

sponding verge of the control volume. In most of plasma processes associated with
convective and diffusive fluxes the integrands are smooth. And for each of them at
its range of integration the average value can be found when the integral is equal to
the product of the average value and the range of integration. The average value of
the integrand along each side e, w, n, s is accepted as an approximation. For
convenience, we use the same notations e, w, n, s for the sides of the control volume
and intersection points of the grid lines (Fig. 6.4). The recent subgroup integrals
thus can be written as

Ice ¼ ðqvz Þe Ue ðrn rs Þ2ðrn þ rs Þ ; Icw ¼ ðqvz Þw Uw ðrn rs Þ2ðrn þ rs Þ ;
Icn ¼ ðqv
r Þn r
n Un ðze  zw Þ; Ics ¼ ðqvrÞs rs Us ðze  zw Þ;
@U ðrn rs Þðrn þ rs Þ ðrn rs Þðrn þ rs Þ
Ide ¼ Ce @z e 2 ; Idw ¼ Cw @U@z w 2 ;
@U  @U
Idn ¼ Cn @r n rn ðze  zw Þ; Ids ¼ Cs @r s rs ðze  zw Þ:

Note:

Fe ¼ ðqvz Þe ðrn rs Þ2ðrn þ rs Þ ; Fw ¼ ðqvz Þw ðrn rs Þ2ðrn þ rs Þ ; ð6:9Þ
Fn ¼ ðqvr Þn rn ðze  zw Þ; Fs ¼ ðqvr Þs rs ðze  zw Þ:

De ¼ zECze M ðrn rs Þ2ðrn þ rs Þ ; Dw ¼ zMCzw W ðrn rs Þ2ðrn þ rs Þ ;

ð6:10Þ
Dn ¼ rNCrn M rn ðze  zw Þ; Ds ¼ rMCr
s
S
rs ðze  zw Þ:

Fe Fw Fn Fs
Pee ¼ ; Pew ¼ ; Pen ¼ ; Pes ¼ ; ð6:11Þ
De Dw Dn Ds
6.1 Basis of the Numerical Methods of the Plasma Physics 167

Here Fe , Fw , Fn , Fs —the mass flow through the corresponding verges e, w, n, s of

the control volume (intensities of the convective flux); De , Dw , Dn , Ds —intensities
of the diffusion flux through the same verge (always positive); Pee , Pew , Pen , Pes —
the grid Peclet numbers at the verges of control volume e, w, n, s, respectively.
Then it can be written

Ice ¼ Fe Ue ; Icw ¼ Fw Uw ; Icn ¼ Fn Un ; Ics¼ F  s Us ;

Ide ¼ De ðzE  zM Þ @U
@z e ; I dw ¼ D ð
w Mz  z W Þ @U
;
@U  @U@z w
Idn ¼ Dn ðrN  rM Þ @r n ; Ids ¼ Ds ðrM  rS Þ @r s :

The further compilation of balance fluxes through the control volume is pro-
duced by adding the appropriate convective and diffusive fluxes through every
verge of the control volume. The assumption of the constancy of some variables on
the verges of the control volume provides an opportunity to apply the result of a
one-dimensional analytical solution of GDE (6.7) for the definitions of these fluxes.
We introduce a function of the grid Peclet number Pe that characterizes the cal-
culation circuit [1]
Pe
AðPeÞ ¼ ¼ AðjPejÞ þ maxðPe; 0Þ: ð6:12Þ
exp Pe  1

Then the amount of fluxes through the verges of control volume will be
8
>
> Ice þ Ide ¼ Fe UM þ ðUM  UE ÞDe AðPee Þ;
<
Icw þ Idw ¼ Fw UM þ ðUM  UW Þ½Dw AðPew Þ þ Fw ;
ð6:13Þ
>
> I þ I ¼ Fn UM þ ðUM  UN ÞDn AðPen Þ;
: cn dn
Ics þ Ids ¼ Fs UM þ ðUM  US Þ½Ds AðPes Þ þ Fs :

The source function is constant inside the control volume. It takes the value
S ¼ const ¼ SM in the central grid M, i.e.,

Zze Zrn
IS ¼ Srdrdz ¼ SM DV; ð6:14Þ
zw rs

where DV ¼ ðrn rs Þ2ðrn þ rs Þ ðze  zw Þ—the volume of the two-dimensional control
volume.
The corresponding balance for the control volume is the sum of all incoming and
outgoing fluxes through the verges of control volume (6.13) and the source function
(6.14)

Ice þ Ide þ Icw þ Idw þ Icn þ Idn þ Ics þ Ids ¼ IS

168 6 Numerical Methods of the Plasma Physics

Substituting (6.13) and (6.14) into the last balance equation, we have

ðUM  UE ÞDe AðPe Þ þ ðUM  UW ÞðDw AðPw Þ þ Fw Þ

þ ðUM  UN ÞDn AðPn Þ þ ðUM  US ÞðDs AðPs Þ þ Fs Þ ð6:15Þ
þ ðFe  Fw þ Fn  Fs ÞUM ¼ SM DV:

On the basis of (6.15), some discrete analogue of GDE (6.2) can already be
obtained. However it does not guarantee the performance of the stability criterion
(6.49) and this fact considerably restricts its application. To display the discrete
analogue of GDE the continuity equation and the condition of the fulfillment of the
law of mass conservation can be used inside the control volume.
The two-dimensional continuity equation in cylindrical coordinates has the form
@ 1@
ðqvz Þ þ ðrqvr Þ ¼ 0:
@z r @r

The mass balance in the continuity equation can be fulfilled using the integration
inside the control volume (Fig. 6.4).
Analogically to the procedure of integration of GDE (6.2) we have
Zrn Zze Zze Zrn
@ 1@
ðqvz Þdzrdr þ ðrqvr Þrdrdz
@z r @r
rs zw zw rs
Zrn Zze
 
¼ ðqvz Þe ðqvz Þw rdr þ ðrqvr Þn ðrqvr Þs dz ¼ 0:
rs zw

Using the assumption of the average value of the integrands and of the constancy
of the fluxes through the verges of control volume e, w, n, s we obtain

ðrn  rs Þðrn þ rs Þ ðrn  rs Þðrn þ rs Þ

ðqvz Þe  ðqvz Þw
2 2
þ ðqvr Þn ðze  zw Þrn  ðqvr Þs ðze  zw Þrs ¼ 0;

or taking into account (6.9) we have

Fe  Fw þ Fn  Fs ¼ 0: ð6:16Þ

Equation (6.16) is the continuity equation in an integral form for this control
volume. Taking into account (6.16) the discrete analogue of GDE (6.2) can be
obtained from the balance Eq. (6.15), it has the following form:

aM UM ¼ aE UE þ aW UW þ aN UN þ aS US þ bM ; ð6:17Þ
6.1 Basis of the Numerical Methods of the Plasma Physics 169

where
8
< aE ¼ De AðjPee jÞ þ maxðFe ; 0Þ  0; aW ¼ Dw AðjPew jÞ þ maxðFw ; 0Þ  0;
a ¼ Dn AðjPen jÞ þ maxðFn ; 0Þ  0; aS ¼ Ds AðjPes jÞ þ maxðFs ; 0Þ  0;
: N
aM ¼ aE þ aW þ aN þ aS [ 0; bM ¼ SM DV:

Here the function A(Pe) (6.12) was used, and the corresponding computational
scheme is called exponential and it is based on the analytical solution (6.7).
Calculation of the exponential function in (6.12) is time-consuming, so the other
scheme based on the approximation of the exponential function in the form [1] can
be used

Pe
AðPeÞ ¼  max½0; ð1  0:1 PeÞ5 : ð6:18Þ
exp Pe  1

It should be noted that the implementation of the law of mass conservation in an

integrated form (6.16) makes it possible to guarantee the positivity of the coeffi-
cients of the discrete analogue in (6.17), i.e., aE ; aW ; aN ; aS ; aM  0 and
aE þ aW þ aN þ aS ¼ 1 that is necessary for compliance of the stability criterion (6.49).
aM

The obtained discrete analogue of GDE (6.17) expresses the appropriate balance
for the finite control volume exactly the same as the original differential equation
for the infinitesimal volume.
The Numerical Solution of the Three-Dimensional GDE Generalized differen-
tial Eq. (6.1) in a three-dimensional cylindrical coordinate system is
 
@ ðqvz UÞ 1 @ 1 @ qvu U
þ ðrqvr UÞ þ
@z r @r r @u




@ @U 1@ @U 1 @ C @U
¼ C þ rC þ þ S: ð6:19Þ
@z @z r @r @r r @u r @u

The corresponding balance for the control volume is obtained by integrating

(6.19) in the three-dimensional control volumes (Fig. 6.4 with additional grid points
T, B and verges t and b in the direction u).

Zut Zrn Zze Zut Zrn Zze


@ ðqvz UÞ @ @U
dzrdrdu  C dzrdrdu
@z @z @z
ub rs zw ub rs zw
Zze Zut Zrn Zze Zut Zrn

1@ 1@ @U
þ ðrqvr UÞrdrdudz  rC rdrdudz
r @r r @r @r
zw ub rs z w ub rs
170 6 Numerical Methods of the Plasma Physics

Zrn Zze Zut   Zrn Zze Zut


1 @ qvu U 1 @ C @U
þ rdudzdr rdudzdr
r @u r @u r @u
rs z w ub rs z w ub
Zut Zze Zrn
¼ Srdrdzdu:
ub z w rs

The integration procedure (6.19) is similar to the process of integration of

two-dimensional GDE, and the discrete analogue is

aM UM ¼ aE UE þ aW UW þ aN UN þ aS US þ aT UT þ aB UB þ bM ; ð6:20Þ

where
8
>
> aE ¼ De AðjPee jÞ þ maxðFe ; 0Þ  0; aW ¼ Dw AðjPew jÞ þ maxðFw ; 0Þ  0;
>
> aN ¼ Dn AðjPen jÞ þ maxðFn ; 0Þ  0; aS ¼ Ds AðjPes jÞ þ maxðFs ; 0Þ  0;
<
aT ¼ Dt AðjPet jÞ þ maxðFt ; 0Þ  0; aB ¼ Db AðjPeb jÞ þ maxðFb ; 0Þ  0;
>
>
>
> a M ¼ aE þ aW þ aN þ aS þ aT þ aB [ 0; bM ¼ SM DV;
:
DV ¼ ðrn rs Þ2ðrn þ rs Þ ðze  zw Þðut  ub Þ:

Fluxes through the verges of the control volume are determined using the
equations
9
Fe ¼ ðqvz Þe ðrn rs Þ2ðrn þ rs Þ ðut  ub Þ; >
>
>
>
Fw ¼ ðqvz Þw ðrn rs Þ2ðrn þ rs Þ ðut  ub Þ; >
>
>
=
Fn ¼ ðqvr Þn rn ðut  ub Þðze  zw Þ; ð6:21Þ
Fs ¼ ðqvr Þs rs ðut  ub Þðze  zw Þ; > >
>
>
Ft ¼ qvu t ðze  zw Þðrn  rs Þ; >
>
  >
;
Fb ¼ qvu b ðze  zw Þðrn  rs Þ:
9
De ¼ zECze M ðrn rs Þ2ðrn þ rs Þ ðut  ub Þ; >>
>
>
Cw ðrn rs Þðrn þ rs Þ
Dw ¼ zM zW ðut  ub Þ; >
>
>
>
2
>
D ¼
n
Cn
rN rM nr ðu  u Þðz  z Þ; =
t b e w
Cs ð6:22Þ
Ds ¼ rM rS rs ðut  ub Þðze  zw Þ >
>
>
>
Dt ¼ rM ðuCu ðze  zw Þðrn  rs Þ; >
>
>
t
MÞ >
>
;
T
Cb
Db ¼ rM ðu u Þ ðze  zw Þðrn  rs Þ:
M B

Fe Fw Fn Fs Ft Fb
Pee ¼ ; Pew ¼ ; Pen ¼ ; Pes ¼ ; Pet ¼ ; Peb ¼ : ð6:23Þ
De Dw Dn Ds Dt Db

The function AðjPejÞ can be determined as (6.18).

6.1 Basis of the Numerical Methods of the Plasma Physics 171

6.1.2 Solution of the Momentum Equations

and the Continuity Equation of the Gas Flux

The Main Methods for Solving the Momentum Equations of the Gas Flux The
momentum equations of the plasma flux for the velocity components vz ; vr ; vu in a
cylindrical coordinate system can be written in a convenient form for decision
8
vvz Þ ¼ divð2l grad vz Þ  @p
< divðq~ @z þ Sz ;
vvr Þ ¼ divð2l grad vr Þ  @p ð6:24Þ
: divðq~ @r þ Sr ;
vvu Þ ¼ divðl grad vu Þ þ Su ;
divðq~

where  @p @z ,  @p
@r —the components of the static pressure gradient;
Sz ¼ FBz þ qgz þ Flz ; Sr ¼ FBr þ Flr ; Su ¼ Flu —the source functions in the con-
tinuity equation; FBz ; FBr —the components of electromagnetic force ~ FB ¼ ~j  ~
B;
qgz —the gravity force; Flz ; Flr ; Flu —the additional viscosity members.

 
 9
1@ @vr @vz 2@ @vz 1 @ ðrvr Þ >
Flz ¼ rl   l þ ; >
>
r @r @z @r 3 @z @z r @r >
>

 
 >
>
@ @vz @vr 2@ @vz 1 @ ðrvr Þ vr =
Flr ¼ l   l þ  2l 2 ;
@z @r @z 3 @r @z r @r r > >

 >
>
vu l @l >
>
Flu ¼ þ þ qvr : >
;
r r @r

The momentum Eq. (6.24) can be written in the similar form of GDE (6.1), i.e.,
the discrete analogue of the GDE (6.17) or (6.20) can be used to solve them. Thus,
the moment equation for the gas flux in the cylindrical coordinate system can be
represented as



@ 1@ @ @vz 1@ @vz @P
ðqvz vz Þ þ ðrqvr vz Þ ¼ 2l þ r2l  þ Sz ; ð6:25Þ
@z r @r @z @z r @r @r @z



@ 1@ @ @vr 1@ @vr @P
ðqvz vr Þ þ ðrqvr vr Þ ¼ 2l þ r2l  þ Sr ; ð6:26Þ
@z r @r @z @z r @r @r @r



@   1@   @ @vu 1@ @vu
qvz vu þ rqvr vu ¼ l þ rl þ Su : ð6:27Þ
@z r @r @z @z r @r @r

If the gradient of static pressure is taken into account through a common source
function, the momentum Eqs. (6.25–6.27) are written in the form of GDE (6.19)
and to solve them we can use the discrete analogue (6.20).
172 6 Numerical Methods of the Plasma Physics

However, the difficulty in the solutions of the momentum Eqs. (6.25) and (6.26)
is the unknown pressure field p (for its definition there is no clear differential
equation). In computational mathematics, there are two ways to solve this problem
and, accordingly, two main methods of the momentum equations solutions. The
first method is based on the elimination of pressure from the momentum equation.
Thus, in the case of two-dimensional problems pressure is eliminated from the two
momentum equations through the cross differentiation of each equation using the
vortex intensity x=r and the current function ψ. For two-dimensional fluxes of gas
flow, this method is based on the known method of solving of the momentum
equation with variable current function and the vortex intensity, described, for
example, in [2–4].
This technique has several features. The pressure is not one of the dependent
variables, instead of two momentum equations and continuity equation for the gas it
is necessary to solve two equations for the current function ψ and the vortex
intensity x=r. Conditions for the vortex intensity on the wall of the plasma torch are
very hard, and because of its approximate representation it is often difficult to obtain
convergent solutions. The pressure has been so successfully excluded from the
formulation of the problem; it is often final desired or intermediate and it is nec-
essary to calculate the thermal properties of the plasma. In these cases, the diffi-
culties encountered in determining the pressure field shifted the benefits of using the
variables of the current function and the vortex intensity. In addition, the main
disadvantage is the inability to apply it in the case of three-dimensional problems
where there is no the current function ψ.
The second method is based on direct solution of the momentum equations and
the continuity equation with physical variables—components of the velocity and
pressure. The pressure field is determined from the continuity equation (integrated
form) after substituting to it the discrete analogues of the momentum equations.
Next, the resulting pressure field is used for solution of the momentum equations
and calculation of the plasma composition, its transfer coefficients and thermody-
namic properties. Thus, the momentum equations and continuity equation are
solved together, the pressure plays a linking role for the parameters of the
momentum and continuity equations, and the resulting velocity field automatically
satisfies the two laws of conservation: of momentum and mass. This method is
described in some of papers [1, 3, 5].
Using the continuity equation to determine the pressure has some difficulties. In
many problems the pressure is known only at one boundary—at the exit from the
plasma torch, or the reactor chamber. On the other boundaries (on the wall, at the
entrance and at the axis of symmetry of the plasma torch) the pressure is generally
unknown.
However, components of the velocity, temperature, gas or gradients normal to
these boundaries are usually known at these boundaries. Thus, in the entrance of
plasma torch or gas inlet chamber the gas flow velocity can be measured or
determined from the total gas flow Q. On the wall of the plasma torch gas as it
6.1 Basis of the Numerical Methods of the Plasma Physics 173

“sticks”, and its speed can be equated to zero. In the plasma torch axis of the radial
component of the plasma velocity is equal to zero by the condition of axial sym-
metry. The temperature and density of the gas at these boundaries are easily
measured or can be determined by simple models. Later, we will show that the
pressure on the borders of these can be found with the defined components of the
velocity and density of the gas, if used in solving shifted chess grid.
Consequently, both of the above methods require joint solutions of momentum
equations and continuity equation (to exclude or determine the pressure). The
second method allows directly obtaining the pressure field required for finding the
composition and properties of the plasma. It describes the physical processes more
accurately and has greater clarity than the first method. In the future, we focus on
the second method, used for the development of algorithms for solving of
momentum equations and continuity equation for the plasma flux. Note that this
method is more time-consuming. It requires performing the additional calculations,
compared with the first method. However, its advantage in the direct determination
of the pressure is it compensates the additional complexity.
The Structure of the Shifted Chess Grids To implement the numerical methods it
is not necessarily to solve all the equations, using the same grid, but each equation
can be solved using its own grid. The boundary conditions for the equations are set
differently on the borders and to use them effectively we need to apply the different
grids.
Therefore it has been developed, the so-called shifted chess grid in which the points
are displaced relative to each other and are staggered. For example, the grid points
for determination of components of the velocity and pressure may be shifted from
each other for a half pitch and the velocity components can be obtained at the
verges of the control volume during the determination of the pressure (Fig. 6.5).
The main advantage of the shifted chess grid structure compared with a grid for
solving of all the equations is to eliminate the need to determine the pressure
directly on the borders, where it is not known (at the entrance, the wall, and the
symmetry axis of the plasma torch). It offers the possibility of finding its value in
terms of known components of the gas velocity.
Shifted chess grid allows obtaining the pressure difference that is necessary for
the calculation of the pressure gradient, using the points on the verges for each
component of the velocity (Fig. 6.5). It does not require a special algorithm for
determining or approximating, as needed in the case of solutions of equations using
the same grid. Similarly, for obtaining the solution of the continuity equation is not
necessary to perform additional approximation on the faces of the velocity com-
ponents of the control volume; they were obtained for these verges (Fig. 6.5) using
its grids. In particular, the use of shifted chess grids lies at the basis of the developed
solution procedures of the momentum equations SIMPLE and SIMPLER [1].
To solve the momentum equation and the continuity equation for the compo-
nents of the velocity vz, vr and the pressure p the developed shifted grid that is
shown in Fig. 6.5 is used, and the typical of the control volume are shown at
Figs. 6.6, 6.7 and 6.8.
174 6 Numerical Methods of the Plasma Physics
 6.1 Basis of the Numerical Methods of the Plasma Physics 175

b Fig. 6.5 The structure of the shifted chess grids for simulation RF-plasma torches and Arc in the
channel (a), and as well as the free Arc (b): •, grid points [z1, r2] to solve the equation of energy
balance; →, grid points [z2, r1] for solution of momentum equation for vr; ↓, grid point [z1, r2] for
solution of momentum equation for vz; ×, grid point [z2, r2] for solving the continuity equation
and determine the pressure p; grid point [z1, r1] is used to solve the momentum equations for vφ
and Maxwell’s equations; three-dimensional momentum equations for vz, vr, vφ are solving in the
grid point [z1, r2, φ2], [z2, r1, φ2], [z2, r2, φ1] respectively

Fig. 6.6 The control volume and the grid [z1, r2] for solving the momentum equation (for vz) and
the equation of the energy balance: →, points for determination of the component vz; ↑, the same
for the component vr; ×, the same for the pressure p

Fig. 6.7 The control volume

and the grid [z2, r1] for
solving the momentum
equation (for vr): the notations
are the same as in Fig. 6.6

In solving the momentum equations and continuity equation for RF- and Arc
plasma torches are generally known the following values: the geometric dimensions
of the channel of the plasma torch; field of study of plasma processes (the inductor
176 6 Numerical Methods of the Plasma Physics

Fig. 6.8 The control volume

and grid [z2, r2] for solving
the continuity equation (for
p): the notations are the same
as in Fig. 6.6

zone, the Arc column zone, the plasma jet zone) and the location of the plasma
torches (vertical, horizontal); outlet pressure on the plasma channel, chamber or
reactor; kind and gas consumption; flow speed of the gas on the inlet of the plasma
torch; its temperature at the entrance, the wall or the free surface of the plasma
torch. Usually in the calculation it is assumed that the system has cylindrical axial
symmetry with two-dimensional or three-dimensional flow of the gas stream.
In view of the above, a system of shifted grids can be built for the effectively use
of the information about the boundary conditions of the problem and the regime of
the plasma torch operation. In constructing the shifted grids we will be guided by
the following: boundary points to determine the unknown variables will be located
directly on the border, where their values are known, or set. For unknown quan-
tities, or values for a known gradient at the boundaries of these points, because of
the shifted grid would move inward by a half pitch calculation area.
So, we can offer the following structure shifted grids for simulation of RF-
plasma torches and electrical Arcs in the channels (Fig. 6.5a).
1. At the entrance to the channel the gas velocity can be determined from its
predetermined flow Q, the temperature of the input stream is often known, so
the initial points for determination of the speed vz and temperature T are located
at this boundary.
2. Due to the cylindrical axis symmetry at the center of the plasma torch vr = 0,
and therefore grid points for velocity vr are on the symmetry axis.
3. The grid for determination of the velocity vz and temperature T is displaced by a
half pitch from the axis and it is possible to find the corresponding streams
using a boundary condition on the axis vr ¼ vu ¼ @v @T
@r ¼ @r ¼ 0.
z

4. The grid for determination of the velocity vr is shifted by a half pitch down the
z axis.
5. The grid for determination of the pressure p is shifted by a half step down the
z axis compared to the grid for vz and a half step on the r axis compared with grid
6.1 Basis of the Numerical Methods of the Plasma Physics 177

for vr, thereby there is an opportunity to find the components of the velocity on
the verges of the discrete analogue for solving the continuity equation.
6. The pressure at the exit from channel of the plasma torch is commonly known,
and the output grid points for determination of the pressure p in the channel are
located directly on this boundary.
7. Consider that gas «sticks» to the wall, i.e., its velocity at the wall of the plasma
torch is equal to zero. The grid for determination of vr is constructed so that the
points are located on the wall.
8. The gas temperature T on the wall of the plasma torch is usually defined or
specified on the inside or the outside. The latter condition is more accurate, and
the grid for the decision of energy balance equation is structured to cover the
wall of the plasma torch and had the grid points on its outside. The decision of
energy balance equation may give the additional information about the tem-
perature distribution inside the wall of the plasma torch and its degree of
heating.
9. Since virtually every equation is solved with the use of a separate grid for
calculating the coefficients of the discrete analogue it should take the appro-
priate approximation of the temperature T, the velocity components vz and vr
using the grids, other than the main one.
10. Some of the convenience for the approximation and determination of the dis-
crete analogue coefficients can be achieved in the solution of energy balance
equation and the momentum equation for vz using the same grid.
The structure of a shifted grid for the simulation of a free electric Arc (Chap. 8)
using its boundary conditions is represented on the Fig. 6.5b. Below, we consider
the use of the structure shifted chess grid for discrete analogue of the momentum
equation and the continuity equation of the plasma flow.
The Discrete Analogue of the Momentum Equations The procedure for
obtaining the discrete analogue of the momentum Eqs. (6.25–6.27) is similar to the
procedure for the obtaining a discrete analogue of the GDE (6.2) with a difference
of integration and static pressure gradients  @p @p
@z and  @r . The discrete analogue for
them can be written as (6.17)
9
aM vzM ¼ aE vzE þ aW vzW þ aN vzN þ aS vzS þ bzM þ ðpw  pe ÞDSr ; >
>
>
>
DSr ¼ ðrn  rs Þðrn þ rs Þ=2; =
aM vrM ¼ aE vrE þ aW vrW þ aN vrN þ aS vrS þ brM þ ðps  pn ÞDSz ; ð6:28Þ
>
>
DSz ¼ ðze  zw Þðrn þ rs Þ=2; >
>
;
aM vuM ¼ aE vuE þ aW vuW þ aN vuN þ aS vuS þ buM :
The coefficients aE ; aW ; aN ; aS ; aM ; bzM ; brM ; buM are determined using their
grids by analogy with coefficients of the discrete analogue (6.17); ðpw  pe Þ and
ðps  pn Þ—the difference of the pressure in grid points to obtain vz and vr ,
respectively; ðpw  pe ÞDSr and ðps  pn ÞDSz —the pressure forces on the surfaces
with the areas DSr and DSz respectively.
178 6 Numerical Methods of the Plasma Physics

For further change it is convenient to introduce the following notations:

8 P
> aE vzE þ aW vzW þ aN vzN þ aS vzS ¼ P am vzm ;
>
<a v þa v þa v þa v ¼ a v ;
E rE W rW N rN S rS Pm rm ð6:29Þ
> aE vuE þ aW vuW þ aN vuN þ aS vuS ¼ am vum ;
>
: DSr DSz
aM ¼ d zM ; aM ¼ d rM ;

Here the summation is over all the nearest mashes of the central point M of this
control volume, m = E, W, N, S. The discrete analogues of the momentum
Eq. (6.28) can be written as
8 P
>
> ¼
am vzm þ bzM
þ ðpw  pe ÞdzM ;
> v
< zM P aM
am vrm þ brM
vrM ¼ þ ðps  pn ÞdrM ; ð6:30Þ
>
> P aM
>
: am vum þ buM
vuM ¼ aM :

Note that the notations are strictly defined for each respective grid. Note
P P
am vzm þ bzM am vrm þ brM
^vzM ¼ ; ^vrM ¼ : ð6:31Þ
aM aM

Here ^vzM and ^vrM , the partial rate velocities for vz and vr , accordingly, and there
is no the pressure difference in the equations for determination of them. Discrete
analogues of the momentum equations for vz and vr can also be written in the form
convenient to insert into the continuity equation
(
vzM ¼ ^vzM þ ðpw  pe Þdzm ;
ð6:32Þ
vrM ¼ ^vrM þ ðps  pn Þdrm :

The Discrete Analogue of the Continuity Equation for the Pressure According
to a second method of solving the momentum equations, the pressure is determined
from the decision of the continuity equation for all the gas flow. At the same time,
the following information can be used: the correct pressure field after the substi-
tution to the momentum equation (equation of the momentum conservation) gives
the velocity field that would satisfy the continuity equation (the law of mass con-
servation). For this reason, the momentum equation and the continuity equation
should be solved jointly. The two-dimensional continuity equation in cylindrical
coordinates is

@ ðqvz Þ 1 @
þ ðrqvr Þ ¼ 0: ð6:33Þ
@z r @r
6.1 Basis of the Numerical Methods of the Plasma Physics 179

The continuity equation for determination of the pressure is solved using grid
[z2, r2] (Fig. 6.5). The mass balance for the control volume that is shown in
Fig. 6.8, can be carried out by integrating (6.33) for this control volume

Zrn Zze Zze Zrn

@ ðqvz Þ 1@
dzrdr þ ðrqvr Þrdrdz ¼ 0:
@z r @r
rs zw zw rs

 ðrn  rs Þðrn þ rs Þ 
ðqvz Þe ðqvz Þw þ rn ðqvr Þn rs ðqvr Þs ðze  zw Þ ¼ 0: ð6:34Þ
2

Using (6.32), components of the velocity vz and vr can be represented on the

verges of the control volume (Fig. 6.8) in the form
(
vze ¼ ^vze þ ðpM  pE Þdze ; vzw ¼ ^vzw þ ðpW  pM Þdzw ;
ð6:35Þ
vrn ¼ ^vrn þ ðpM  pN Þdrn ; vrs ¼ ^vrs þ ðpS  pM Þdrs ;

Here the partial rate velocities and ratios are defined as

8 P P
> am vzm þ bze am vzm þ bzw
>
< ^vze ¼ ; ^vzw ¼ ;
ae aw
P P ð6:36Þ
>
> am vrm þ brn am vrm þ brs
: ^vrn ¼ ; ^vrs ¼ :
an as
8
> ðrn  rs Þðrn þ rs Þ ðrn  rs Þðrn þ rs Þ
>
< dze ¼ ; dzw ¼ ;
2ae 2aw
ð6:37Þ
>
> ðz  zw ÞðrN þ rM Þ ðze  zw ÞðrS þ rM Þ
: drn ¼ e ; drs ¼ :
2an 2as

All the neighboring points in the corresponding grid for determination of vz (grid
[z1, r2]) and vr (grid [z2, r1]) are marked using the index «m». It depends on the
location of these grids points e, w, n, s of the control volume for determination of
the pressure p (grid [z2, r2]). Substituting (6.35) into (6.34), we obtain a discrete
analogue of the continuity equation for the pressure p

aM p M ¼ a E p E þ a W p W þ a N p N þ a S p S þ b M ;

where
8
>
> a ¼ qe dze ðrn rs Þ2ðrn þ rs Þ ; aW ¼ qw dzw ðrn rs Þ2ðrn þ rs Þ ;
< E
aN ¼ qn drn ðze  zw Þrn ; aS ¼ qs drs ðze  zw Þrs ;
ð6:38Þ
>
> a ¼ aE þ aW þ aN þ aS ;
: M
bM ¼ ðqw^vzw  qe^vze Þ ðrn rs Þ2ðrn þ rs Þ þ ðrs qs^vrs  rn qn^vrn Þðze  zw Þ:
180 6 Numerical Methods of the Plasma Physics

Thus, with the help of the shifted chess grid by substituting the solutions of the
continuity equation for the velocity components vz and vr into the integrated con-
tinuity equation we obtain a new kind of discrete analogue of the continuity
equation for the pressure (6.38). Together with (6.28) we have four systems of the
discrete analogues of the momentum equations and the continuity equation for all
the gas flow for the four unknown variables vz, vr, vφ and p.
The algorithm of the joint solution of the momentum equations and the conti-
nuity equation is
1. The shifted chess grids for the solution of the momentum equation and the
continuity equation are built using the boundary conditions and the features of
this problem.
2. Some suggested initial velocity is introduced in dependence on the coordinates
of the computational grid. In many cases, for a rapid convergence the zero
velocity is selected as the initial velocity (except the boundary points).
3. The coefficients of the momentum equation are calculated using their grids and
then the partial velocities are calculated (6.31) by substituting the values of the
velocities in neighboring points.
4. The coefficients of discrete analogue of continuity Eq. (6.38) are calculated and
(6.38) is solved for the pressure.
5. The system of discrete analogs of the momentum equation is solved using the
pressure (6.28) and a new velocity is found.
6. The calculations are repeated from position 3 as long as there is no convergence
of the solution.

6.1.3 Conversion of Different Equations

in Generalized Form

Equation of the Energy Balance for the Equilibrium Plasma If we assume that
the total specific heat capacity of the plasma cp in the control volume is constant,
the equation of the energy balance for the equilibrium plasma can be written in the
form of GDE (6.1) as


k 1 2
vTÞ ¼ div
divðq~ gradT þ rE  Urad : ð6:39Þ
cp cp
There is another form of the equation of the energy balance using enthalpy H

divðq~
vH Þ ¼ divðkgradT Þ þ rE 2  Urad : ð6:40Þ

For quasi-neutral atomic plasma with single, double, and triple ionization the
electron density, the plasma density, and enthalpy are determined by the equations
6.1 Basis of the Numerical Methods of the Plasma Physics 181

 
ne ¼ niþ þ 2niþ þ þ 3niþ þ þ ; q ¼ ma na þ niþ þ niþ þ þ niþ þ þ ;

1 5
H¼ kTne þ Eiþ niþ þ Eiþ þ niþ þ þ Eiþ þ þ niþ þ þ
q 2

5  þ þþ þþþ
 ð6:41Þ
þ kT na þ ni þ ni þ ni
2


1 5 þ þ þþ þþ þþþ þþþ 5k
¼ kTne þ Ei ni þ Ei ni þ Ei ni þ T:
q 2 2ma

Therefore, the equation of the energy balance (6.40) for the atomic plasma has
form:


2ma 2ma
divðq~
vT Þ ¼ div kgradT þ
5k 5k


 
5
 rE  Urad  div
2
kTne þ Eiþ niþ þ Eiþ þ niþ þ þ Eiþ þ þ niþ þ þ ~
v :
2
ð6:42Þ

Note that the solution of (6.42), there is no need to determine the specific heat
capacity cp .
Equation of the Energy Balance for the Electron Gas in Two-Temperature
Plasma Consider the equation of the energy balance for the electron gas of atomic
neutral plasma with single,
 double,
 and triple ionization.
 Using the continuity
equation in the form div ma na þ niþ þ niþ þ þ niþ þ þ ~v ¼ 0, we get the fol-
lowing relations:
8  þþ
  
< divna þ ni
> þ niþ þ þ ~v ¼ div niþ ~
v ;
  
div na þ niþ þ niþ þ þ ~v ¼ div niþ þ ~
v ;
>
:    
div na þ niþ þ niþ þ ~v ¼ div niþ þ þ ~
v :
Consequently,

 
5  
divðq~
vHe Þ þ div kTe na  niþ þ  2niþ þ þ  Eiþ niþ  Eiþ þ niþ þ  Eiþ þ niþ þ þ ~v
2


5 5  
¼ div kTe ne þ Eiþ niþ þ Eiþ þ niþ þ þ Eiþ þ þ niþ þ þ þ kTe na  niþ þ  2niþ þ þ
2 2
      
þ Eiþ na þ niþ þ þ niþ þ þ þ Eiþ þ na þ niþ þ niþ þ þ þ Eiþ þ þ na þ niþ þ niþ þ ~ v


5   5k
¼ div kTne þ Eiþ þ Eiþ þ þ Eiþ þ þ na þ niþ þ niþ þ þ niþ þ þ ~ v ¼ divðq~
vTe Þ:
2 2ma

The equation of energy balance for the electron gas (1.34) for the stationary case
and ~
ve ¼ ~
v can be written in the form of GDE (6.1)
182 6 Numerical Methods of the Plasma Physics


2ma 2ma 
divðq~
vTe Þ ¼ div ke gradTe þ re E 2  Urad  bðTe  Ta Þ
5k 5k

 
5  þþ

div kTe na  ni  2niþ þ þ  Eiþ niþ  Eiþ þ niþ þ  Eiþ þ þ niþ þ þ ~
v
2
ð6:43Þ
qffiffiffiffiffiffiffi 
Here b ¼ 32 k 2m
ma ne
e 8kTe
pme na Qea þ niþ Qeiþ þ niþ þ Qeiþ þ þ niþ þ þ Qeiþ þ þ —the
coefficient of the energy exchange of electrons with atoms and ions;
Qea ; Qeiþ ; Qeiþ þ ; Qeiþ þ þ —averaged effective cross sections of the electron with
atoms, single, double, and triple ions, respectively (Sects. 3.4 and 3.5).
Equation of the Energy Balance for Atom–Ion Gas in Two-Temperature
Plasma For the atom–ion gas we have enthalpy

1 5  þ þþ þþþ
 5k
Hai ¼ kTa na þ ni þ ni þ ni ¼ Ta :
q 2 2ma
The equation of energy balance for atom–ion gas (1.37) for the stationary case
can be written in the form of GDE (6.1) as


2ma 2ma
divðq~
vTa Þ ¼ div kai gradTa þ bðTe  Ta Þ: ð6:44Þ
5k 5k

The Continuity Equation of the Chemical Elements and the Equation of the
Ionization Equilibrium In general, the steady continuity equation, or the equation
of mass conservation of the ith chemical elements has the form GDE (6.1) as

divðq~
vxi Þ ¼ divðCi gradxi Þ þ Si ; ð6:45Þ
Here xi ¼ qi =q—the mass concentration of ith chemical element; Ci ; Si —the dif-
fusion coefficient and the source function of the ith chemical element.
Stationary equation of ionization equilibrium for atomic plasma is written as:

2 
ne n2e
divðne~
vÞ ¼ divðDamb gradne Þ þ kr ne na  ; ð6:46Þ
na na

Here Damb —the bipolar diffusion coefficient; kr —the coefficient of the triple
(impact-radiative) recombination; ne ; na —the concentration of electrons and atoms
of equilibrium plasma, determined by the Saha equation.
Equation (6.46) can be written in the form of GDE (6.1) as


n2 n2e
divðq~
vxe Þ ¼ divðqDamb gradxe Þ þ kr me ne na e
 : ð6:47Þ
na na
6.1 Basis of the Numerical Methods of the Plasma Physics 183

Taking into account, boundary conditions for the solution of the equation of
ionization equilibrium, and a link between the particle concentration and temper-
ature in the determination of the transfer coefficients and the thermodynamic
properties of the plasma, it is desirable to solve the equation of ionization equi-
librium (6.47) using the same grid [z1, r2] (Fig. 6.5), like the equation of energy
balance.
For the convenience, the parameters of different equations in the form of GDE
(6.1) are shown in Table 6.1.
Note that the static pressure gradient is included to the source function of the
momentum equation for vz and vr , this gradient is then substituted into the conti-
nuity equation of the gas flux. The momentum equations have the discrete analogue
(6.28), and the continuity equation has the discrete analogue (6.38). For all other
equations in the form of GDE (6.1) the discrete analogue can be applied the form

Table 6.1 The equations in the form of GDE (6.1)

Equations Parameters
Φ C S
Equation of the energy balance for
the equilibrium plasma model:
Taking into the specific heat T k=cp S ¼ ðrE2  Urad Þ=cp
2ma



Without the specific heat T k 2ma 5
5k rE2  Urad  div kTne þ Eiþ niþ
5k 2
 
þ Eiþ þ niþ þ þ Eiþ þ þ niþ þ þ ~v
Equation of the energy balance for Te 2ma
ke 2ma 
5k re E 2  Urad  bðTe  Ta Þ
the electron gas 5k 

5  
 div kTe na  niþ þ  2niþ þ þ
2
 
 Eiþ niþ  Eiþ þ niþ þ  Eiþ þ þ niþ þ þ ~
v
Equation of the energy balance for Ta 2ma
5k kai 2ma
5k bðTe  Ta Þ
the atom–ion gas


Momentum equation for vz vz 2l @p ~ ~ 1@ @vr @vz
 þ jB zþ rl 
@z r @r @z @r


2@ @vz 1 @ ðrvr Þ
 l þ þ qgz
3 @z @z r @r


Momentum equation for vr vr 2l @p ~ ~ @ @vz @vr
 þ jB rþ l  
@r @z @r @z


2@ @vz 1 @ ðrvr Þ vr
 l þ  2l 2
3 @r @z r @r r
 
Momentum equation for vu vu l  ru lr þ @l
v
@r þ qvr

Continuity equation for all the gas 1 0 0

flux
Continuity equation for the ith xi Ci Si
chemical elements
 
Equation of ionization equilibrium xe qDamb n2 n2
kr me ne na e
na  nea
184 6 Numerical Methods of the Plasma Physics

(6.17) or (6.20). The above transfer equations of the turbulent flow of the models
k  e and k  x (Sect. 1.5) have also the form of GDE (6.1), their basic coefficients
are presented in Tables 1.1, 1.2.

6.2 The Stability of the Difference Methods

and Computational Procedures

6.2.1 Stability of Difference Methods

Property of the difference methods that is able to cause a computational system to a

single final decision under certain perturbations of the input data is called stability.
If Uh —the solution of kth differential equations and nh —the input data for some
difference method. These parameters depend on the integration step or the grid step
Δh. And if we change Δh we get a sequences of decisions fUh g and input data
fnh g. In mathematical physics it is accepted that the problem and its difference
method of solution are correct if the following conditions are carried out:
1. The problem is uniquely solvable and the solution exists and is unique for all the
input data nh from some admissible family (the region of stability).
2. The solution of the difference problem continuously depends on the initial data
nh , and this dependence can be uniform for the grid steps Δh.
The condition (2) implies that a constant value M > 0 exists and it is independent
of Δh, and for sufficiently small step jDhj  h0 , the inequality [6] occurs
 0   
U  Uh   M n0  nh  ; ð6:48Þ
h ð1hÞ h ð2hÞ

where U0h —the solution of the difference problem with the input data n0h , and (1h)
and (2h)—the norms in a variety of gridlines functions defined on this grid.
The property of continuous dependence of the solution of the difference problem
on the input data expressed by inequality (6.48) is called the stability of difference
methods for input data, or simply stability. Thus, if the difference method is stable,
it converges; and the establishing of the convergence and accuracy of the solution
of the problem are connected with the choice of the approximation method and
stability of the system.
Stability Criterion After the sampling of the differential equation using a
finite-difference
P method a system of discrete analogues can be obtained in the form:
aM U M ¼ am Um þ bM . Then a sufficient condition for the convergence of solu-
tions of such a system in the iterative methods of Gauss–Seidel is a diagonal
dominance, i.e.,:
6.2 The Stability of the Difference Methods and Computational Procedures 185

X
j aM j  jam j for all equations;
X
j aM j [ jam j at least; for one of the equations: ð6:49Þ
Conventionally, we call the difference method for obtaining the discrete ana-
logues with the property (6.49) as stable. Note that the stability criterion (6.49)
exists only in the solution of linear algebraic equations. In solving the nonlinear
systems, the stability criterion does not exist, and the condition (6.49) is not suf-
ficient for convergence of the solutions.
Stability criterion (6.49) has the following features:
1. The criterion (6.49) is a sufficient but not a necessary condition. Therefore the
convergence may be even at the partial violation of (6.49).
P
am Um þ bM
2. At the recording of a discrete analogue in a form UM ¼ aM , it follows
that if the value aM higher, the value UM is the more slowly changing from
iteration to iteration, and the less it depends on the values of Um in neighboring
nodes, i.e., on the disturbances of the iterative procedure, it contributes to the
sustainability
P of solutions.
3. If aM ¼ am , as is typical for most of the difference methods, for the imple-
mentation of the stability conditions (6.49) the requirement that all the coeffi-
cients of the discrete analogue have the same sign should be complied. In this
case, the second condition for the stability criterion is automatically met if at
P one value of the required variable is known at the border, i.e., aM ¼ 1,
least
am ¼ 0.

The Convergence Condition Suppose that UM is the value of the unknown
variable in the node M, obtained in the previous iteration, and UM —the same as
found in the current iteration, then the condition for solutions convergence in an
iterative procedure of solving the inequality:
 
UM  UM 
max   e for all M: ð6:50Þ
UM 
Here e ¼ 105  103 —some small value, depending on the conditions of the
problem. If the value of Φ have a large gap, the following condition of convergence
can be used:
 
UM  U 
max M
 e for all M: ð6:51Þ
maxjUM j

A very slow convergence of (6.50) and (6.51) may prove to be incorrect and
insufficient to stop calculating and making the intermediate value as a final solution.
It so happens that the solution is far from steady-state value, but the conditions
(6.50) or (6.51) already occurs. In this case, it is necessary to define some residual
R for the solution of differential equation and sufficient condition for the conver-
gence of solutions in this case is an extremely small value R.
186 6 Numerical Methods of the Plasma Physics

To analyze the stability of difference methods, applicable to the solution of

problems of gas dynamics and heat transfer in plasma using numerical solution of
one-dimensional GDE in the form (6.8).
Using the function A(Pe) in (6.8) allows the compact construction of all the
numerical method in one record for their analysis without changing the type of
discrete analog. For example, the exponential scheme AðPeÞ ¼ expPe Pe1 for central
difference scheme AðPeÞ ¼ 1  0:5Pe for the scheme upstream AðPeÞ ¼ 1, etc.
In sampling the condition Fe ¼ Fw ¼ qvz ¼ const is used. Because of the
approximate solution this condition cannot be satisfied immediately in an iterative
process. However, its use makes it possible to obtain the discrete analogue (6.8)
with coefficients satisfying stability criterion (6.49). In this case, the coefficients
De [ 0; Dw [ 0; Fe ¼ Fw [ 0, so the requirement of positiveness of the coeffi-
cients in the discrete analogue[the 3-d consequence of the stability criterion (6.49)]
leads to the requirement positiveness for the function A(Pe).
It is easy to note that for |Pe| > 2 the using of a central difference scheme A
(Pe) < 0 (Fig. 6.9), we get the negative coefficients of the discrete analogue (6.8)
and violation of the stability criterion (6.49). Consequently, the central difference
scheme in this case is unstable and its use is not desirable.
To prevent A(Pe) < 0 by approximation using the central difference scheme,
a combined scheme ½AðPeÞ ¼ maxð0; 1  0:5PeÞ can be used.
The scheme «upstream» is stable if A(Pe) > 0 for all Pe, but it is not accurate for
small values of |Pe|. In this case, according to the «up flow» the average value of the
unknown variable on the verges of the control volume still has the value of this
variable in the near node down the stream, that is true only for a large predominance
of convective flow under the diffusion one, i.e., at the large value of Pe (Fig. 6.2).
For the exponential scheme A(Pe) > 0 for all Pe (Fig. 6.9), and all the coeffi-
cients of the obtained discrete analogue are positive that satisfies the stability cri-
terion (6.49). In terms of accuracy and stability, an exponential scheme can be used
to construct the discrete analogue. However, the exponential calculation usually
requires more computation time, so this scheme is often approximated by different
polynomials, one of them is available in [1]—the scheme with a power-law
Pe
(Fig. 6.9): AðPeÞ ¼  max½0; ð1  0:1PeÞ5 . The exponential scheme
exp Pe  1
is accurately approximated using this function by and it guarantees the positive
values of the coefficients of the discrete analogue.

Fig. 6.9 The function A(Pe)

for different computational
schemes: 1 the exponential
scheme; 2 the central
difference scheme; 3 the
scheme «upstream» ; 4 the
combined scheme; 5 the
scheme with a power-law
6.2 The Stability of the Difference Methods and Computational Procedures 187

6.2.2 Analysis of the Computational Procedures Stability

From the above calculation schemes (Sect. 6.2.1), the best sustainable ones can be
selected. However, these schemes are stable only in the solution of linear problems
(i.e., with constant coefficients). For nonlinear stability problem solving procedure
is not guaranteed in the application of even the most stable of the scheme.
A divergence or oscillation process of solutions with increasing amplitude is usu-
ally the sign of the instability of the computational procedure.
Based on the fact that the object modeling—the low-temperature plasma is in a
stable state, it can be assumed that the stability of calculation procedures may be
lost due to the inconsistency of the structural model of computational procedures
with natural processes, i.e., due to loss of information. Stability of the computa-
tional procedure is particularly disturbed at the decision purely nonlinear differential
equations, where the loss of information is essential and cannot be recovered or
compensated.
In practice the iterative methods are often used as the main tool of exclusion of
the instability of computational procedures for solving nonlinear problems. Using
iterations greatly simplifies the calculations, instead of the simultaneous solution of
all the equations their decision is consistently carried out. Starting with some initial
decisions their new approximate solution can be found. The old decision should be
replaced by the new one, and the process should be repeated until the convergence
is reached. Links of the physical processes that cannot be established at this stage of
solution are recovered in solution of equations with repeated clarifications and
translation of the relevant factors.
This structure (iterative model) of solutions of differential equations as a result of
iterative procedures is shown in Fig. 6.10, here there are two levels of iterative
procedures—local for solving of each equation and global for consistent solutions
of equations (to clarify and enumeration of relevant factors, i.e., to restore the
connection between the equations). The coefficient α is the coefficient of external
feedback and it is implemented using the known algorithms of top or bottom
relaxation.

Fig. 6.10 The structure of the sequential solution of differential equations as a result of iterative
procedures: I1, I2, …, In, local iterative procedures; III, global iterative procedure
188 6 Numerical Methods of the Plasma Physics

Despite the popularity of the use of iterative procedures, they have a significant
drawback: the information about the connection of some physical processes (be-
tween the laws of conservation) may be only partially restored at the end of new
iteration and is not reflected in the solution of the equation. The coefficients of each
equation are determined and corrected only in the transition from decision to
decision, and in the process of solving each equation remain unchanged. Therefore,
in the computational procedure of the simulation model, there is always some delay
due to the processes of simulation compared to actual physical processes.
Moreover, if the physical connection is simulated for local iterative procedures, the
time delay is equal to the working time of one local procedure, and if the con-
nections between the physical processes exist at the global procedure, the time lag
even further—it is equal to the time of one global calculation procedure.
The harder the physical processes are associated, the greater the delay in the
management of computational procedures becomes and the greater the distortion of
the simulation model becomes, compared with the mathematical model. In the case
where the external feedback circuit cannot compensate the delay in the control, the
calculation process becomes unstable or diverging.
The successful use of traditional iterative structures for modeling of some of
nonlinear physical processes only notes that for these processes, the natural com-
munications are not so strong, and the information that restores by the iterative
procedures is sufficient to compensate the lost information inside the solution. At
this structural representation, the nonlinear and rigid connections of the physical
processes cannot yet lead to the instability of the system. Therefore, to ensure the
stability of the calculation procedures, it is necessary to simulate and determine not
only the external but also the internal communication of the physical processes in
the computational structure.
For example, analyze the stability of the modeling procedures for established long
plot Arc in the channel. For a long Arc, where the radius of the channel that is much
less than its length, the established section occupies a large space of the
 channel
 and
has the parameters that almost do not change along the channel @T @z ¼ 0 . The
 
motion of the gas is one-dimensional vr ¼ vu ¼ 0 . Isolation of energy rE2 is
consumed in the form of heat transfer by conduction  in the radial direction,
 and
radiation. The electric field has only one component Eu ¼ Er ¼ 0; Ez ¼ E . In this
case, in the model of the steady plot of the long Arc include only two equations: the
equation of the energy balance for determination of the temperature and the Maxwell
equation for calculating the strength of electric field:


1d dT I
rEz2  Urad ¼ rk ; Ez ¼ ; ð6:52Þ
r dr dr RR
2p rrdr
0

Here I—current of the Arc; R—the radius of the Arc channel.

6.2 The Stability of the Difference Methods and Computational Procedures 189

Fig. 6.11 The control volumes and grid for solution of the one-dimensional equation of the
energy balance of the Arc in the channel

The solution is carried out by method of the control volume. The balance of
power and the discrete analogue for the equation of energy balance are obtained by
the integrating of the equation of energy balance in the control volume (Fig. 6.11).
Discrete analoguefor the equation of energy balance (6.52) takes the following
form:

aM TM ¼ aN TN þ aS TS þ bM ; ð6:53Þ

where
(
aN ¼ rNrnr
kn
[ 0; aS ¼ rMrsr
ks
[ 0; aM ¼ aN þ aS [ 0;
M S  
bM ¼ 0:5SM ðrn  rs Þðrn þ rs Þ; SM ¼ rEz2  Urad M :

The standard solving procedure consists of the following:

1. Some grid is constructed and the initial temperature field is given on this grid for
the first iteration, for example, T = 10,000 K.
2. The strength of electric field E, the released (specific) power rE2 and coeffi-
cients of the discrete analogue are determined by (6.53).
3. The system of discrete analogs (6.53) (with constant coefficients) is solved in
order to find a new field of temperature.
4. The convergence condition is checking (6.50).
5. If the convergence condition is satisfied, then the solution is over, if not, a new
field of temperature is adopted as the current and the calculation is continued,
starting from position 2.
Solution using this procedure and grid with the steps Dr ¼ 0:001  0:1 mm is
unstable. The dynamics of the iterative procedure (a change of temperature and
released power from iteration to iteration) is shown in Fig. 6.12. The rare changes
in temperature and released power from iteration to iteration indicate the instability
of computational procedure.
190 6 Numerical Methods of the Plasma Physics

Fig. 6.12 The change in temperature T (a) and released power P = σE2 (b) in the Arc from
iteration to iteration at the unstable solution for different points (r = 0 and r = 0.8 mm) of the Arc
channel with radius R = 1.5 mm and current I = 90 A

This iterative procedure has the following features: the coefficients of the dis-
crete analogue remain constant inside the iterations (position 3) regardless of its
decision; conversion of the coefficients is performed after obtaining of the new field
of temperature. Links between the various physical processes and the field tem-
perature (i.e., by the decision of the equation of the energy balance) are corrected or
restored only after solving of the discrete analog, so there is a delay in the
administration of the simulation model for the computational procedure for some
duration, at least, one iteration in comparison with natural processes .
This equation of energy balance
 consists of three members: the transfer of power
by heat conduction 1r drd rk dT dr ; the losses of power due to radiation Urad and the
emitted power rE2 . Beside the established link of power transmitted by heat
conduction with the temperature in the discrete analogue (6.53) the knowledge of
the relations between the losses of power due to radiation Urad ¼ const and emitted
power rE 2 with temperature is required. These links are not established inside the
iteration. They are restored after each of iterations through the conversion of the
strength of electric field, emitted power, and the losses of power due to radiation
using the resulting field of temperature.
It is difficult to set the links between Urad and rE2 and temperature inside the
iteration, because they depend on many factors. However, the rare change in the
decisions of equation of the energy balance and released power from iteration to
iteration (Fig. 6.12) indicates a rigid connection between the released power and the
temperature of plasma. The exact dependence rE 2 (or rE 2  Urad ) on the tem-
perature cannot be determined, but based on physical reasoning it can be assumed
that it has a decreasing character, as well as for any sustainable process.
Indeed, if it is not so, then any perturbation (raising or lowering of temperature)
may cause an additional increase or respectively decrease of the emitted power, and
it contributes to further increase or decrease of the temperature. This process does
not occur in nature as the increase or decrease in the temperature cannot continue
indefinitely, always in nature, particularly in plasma, there is a negative feedback
mechanism for the stable existence.
6.2 The Stability of the Difference Methods and Computational Procedures 191

Thus, to eliminate the delays in the administration of the simulation model, as

compared to the physical processes beside the iterative procedure, it is necessary to
establish a special internal communication directly inside the solutions each of
equation. This connection can be realized with the help of some dependence
between the source function S and the unknown variable Ф.

6.3 The Structural Organization of Computational

Procedures

6.3.1 The External Feedback in the Structure

of Computational Procedures

Most of the tasks at the simulation of plasma processes are solved using two
iteration loops: the internal (local) for solving the individual equations and the
external (global) for the consistent solutions of equations (Fig. 6.10).
The traditional structure of the local iterative procedure with constant coeffi-
cients is shown in Fig. 6.13. It has the following features:
The decision of any equation begins with some initial (approximate) field
1. 
 U 0 .
M
2. In solving each of equations the solution of other equations remains unchanged.
Besides the physical processes described in this equation, the communications
of the other processes and the laws of conservation are not established in its
decision.
3. The stability condition of the iterative procedure (6.49) exists only for the con-
stant coefficients aM ; aE ; aW ; aN ; aS ; bM and the block of definitions of these
factors is beyond the scope of an iterative procedure. The coefficients of a discrete
analogue are determined again after each of the global iteration cycle, taking into
account the decisions of other equations. Thus, the dependence of some physical
processes can be simulated only in the external (global) feedback loop (Fig. 6.10).

Fig. 6.13 The local iterative procedure with the external determination of the coefficients
192 6 Numerical Methods of the Plasma Physics

4. The new solution after each of iteration cycles replaces the old one. The solving
process is repeated until the convergence condition is satisfied (6.50).
5. The recursive solution is carried out for all grid points using the discrete ana-
logue of the differential equation. For multidimensional problems the rapid
convergence of solutions for the system of algebraic equations can be achieved
through a combination of two methods: direct in the same direction of coordi-
nates and recursive in other directions (the method of alternating directions).
6. Repeat of the solutions is realized by the circuit with the external feedback using
the factor α, the regulation of this factor can be slowed down or speeded up the
convergence of solutions, if it is necessary.
Control of the iterative procedure using external feedback is implemented using
the following algorithm:

ðnÞ ðn1Þ
UM ¼ aUM þ ð1  aÞUM ; ð6:54Þ

ðnÞ ðn1Þ
Here UM and UM —the solutions of the point M in two nearest iterations n and
n − 1.
Specifically, some algorithms, well-known in computational mathematic, of the
lower ða\1Þ or top ða [ 1Þ relaxations are implemented using α. The algorithm of
lower relaxation (change of U from iteration to iteration is made smaller than in the
iteration process) is often used at the solving of hard and ill-conditioned problems
for artificial dilation of the solution, that is capable for convergence.
The analysis of the stability of the computational procedures based on the
decisions of equation of energy balance for the long Arc in the channel (Sect. 6.2.2)
showed that for objects with highly nonlinear properties such as plasma, the
divergence or instability of the solution can be obtained after the first iteration, even
if we use the tiny grids. Therefore, using only the external feedback (as a lower
relaxation) to slow the convergence after each of iterations cannot always solve the
problem of the stability of computational procedures.

6.3.2 Installing Internal Feedback in the Computational

Procedures

The procedure of solving the system of differential equations consists of many

substructures (subsystems) (Fig. 6.10), each of them provides a solution of single
equation. Therefore, the stability of all the computational system depends on the
stability of the subsystems, and for the stability of all the system it is necessary to
ensure the stability of each of subsystem.
It can be done using the principle of self-sustaining of physical processes, that is
reflected in many of the structural representations. For example, suppose that a
result of various thermal processes, the temperature of plasma in some volume of
the plasma torch channel increases. Then for the steady state of the plasma the
6.3 The Structural Organization of Computational Procedures 193

release of energy in the form of Joule heat rE 2 in this volume should be reduced,
and at the same time the flow of energy transfer from this volume the other one
have to increase due to different dissipative processes. As a result, the temperature
in the volume cannot be increased indefinitely. The stable plasma, due to internal
reverse (negative) links, is regulated itself by initiating the various feedback pro-
cesses that may lead to the counter to any disturbance.
A similar picture can be obtained by considering the changes in gas-dynamic,
thermodynamic, electromagnetic, and other physical processes. We see that
everywhere and for every object the internal feedback exists for self-regulation and
maintaining of the steady state of the object. Thus, in simulations these connections
must be established in the model as a negative feedback.
Based on the principle of sustainability and self-regulation of natural object the
internal feedback of different physical processes can be restored and set in a
decreasing dependence of source function S of the target variable Φ. In various
studies [1, 7] it is proposed to establish this relationship inside the procedure of
solving of each equation as a linear source function S:

S ¼ Sc þ Sp U; ð6:55Þ

where the coefficients Sc ; Sp depend on the coordinates and on Φ. After each

iteration cycle, Sc and Sp and be converted with other factors of a discrete analogue
using the new values of Φ.
The dependence of the source function S on Φ (6.55) is set inside the iterative
solving procedure and it is an internal feedback for output value. Linearization of
the source function is performed on the grounds that, first, the system of the discrete
analogues allows only the formal linear relationship, and second, for the iterative
procedure the depending of the source function S on Φ can be formed approxi-
mately. Mistakes in the approximation of this function can be eliminated with the
help of external feedback of iterative procedure. Third, the structure of the GDE
(6.1) shows that the inner connection of various physical processes can be set in the
discrete analogue only using the source function, and other members have already
been successfully used in the sampling of the differential equation.
Although the adoption of a linear dependence S(Φ) allows us to describe some
connection inside the iterative procedure, a method of determination of its coeffi-
cients is unknown. In addition to the trivial cases when the source function can be
explicitly expressed as a function of the independent variable, for the linearization
of the source function for most of the equations for description of plasma when only
the values S and Φ from the previous iteration are usually known. For the con-
struction of the linear dependence it is necessary to have additional information.
This creates some difficulties in the formation of such a function.
The proposal about the maximum possible range of the target variable Φ [1] very
often leads to a small value of the coefficient Sp , i.e., it fails to effectively control of
the course of solutions that in some cases can negate the effect of feedback in
solving of stiff systems of equations. In this connection, we will offer the following
algorithm to establish the connection and control the speed of solving the equations:
194 6 Numerical Methods of the Plasma Physics

Fig. 6.14 Illustration of

linearization for the source
function S ¼ Sc þ Sp U
(S ¼ const when
b ¼ 1; Sp ¼ 0)

because of the possible approximate solution in an iterative procedure the rela-

tionship can be constructed where the solution obtained in the following procedure
does not exceed the some certain value. If the source function in the point M takes
the value SM for UM in this iteration, it can be requested that in the next iteration the
resulting solution does not exceed the value bUM , where β—some coefficient
(Fig. 6.14). This can be achieved using the following algorithm [7, 8]:
 
S
Sp ¼  ; Sc ¼ S  Sp U: ð6:56Þ
bU

Coefficient β determines the speed of change of the function S, depending on Φ.

It can be called the coefficient of internal feedback in solving of this equation.
Basically the coefficient β varies over a wide range ðb ¼ 0  1Þ and it is deter-
mined depending on specific conditions of the problem. Note that always Sp  0. It
is in accordance with the above principle of the stability of the physical process.
The meaning of the input coefficient β can be identified by (6.55) and (6.56). The
smaller β is, the greater the angle of the dependence S(Φ) is and greater the source
function S reacts to a change in the solution of the equation and internal feedback
effect is more; wherein, the relationship S(Φ) is not the same for all control volumes
and it is determined by the values of the source function S and the variable Φ at
each point M, obtained in the previous iteration. Using the proposed the lin-
earization algorithm for the function S(Φ) and the installation of internal feedback
(6.56) in contrast to the algorithms [1] there is no need to predetermine the max-
imum range of variation of the target variable Φ. This algorithm is universal,
effective, and useful in solving the equations with any possible range of variation
for the variables Φ, so the changes of the target variable from iteration to iteration
can be effectively controlled.
6.3 The Structural Organization of Computational Procedures 195

In this case, the coefficient β is introduced to establish the internal negative

feedback between the source function and the variable Φ, but given its regulatory
role, we will be able to artificially mimic the relationship (on the principle of
sustainability of the physical processes) not only for the source function, but also
for other physical processes with variable Φ. Thus, with the introduction of the
internal feedback coefficient β of we were able to fully restore the information lost
as a result of the assumption of the constancy of the coefficients in iterative pro-
cedures, simulating the natural physical processes in computational procedures,
making them identical, and thus achieve the stability of these procedures.
Dynamics of the Iterative Procedures with Internal Feedback Results of the
iterative procedures are approximate and vary from iteration to iteration, if the
iterative procedures are stable, they, as the results of the computational processes,
can satisfy the mathematical model and the gradient of the steady solution in time
(actually in iteration) is equal to zero.
Thus, the models of iterative procedures are always unsteady, even for stationary
physical processes. If the real processes are sustainable, the iterative procedures still
may be unstable. If we take one of iteration per unit time, the change of the solution
in the computational procedure by iterations and the real time of the processes are
identical. In this case, the iterative procedure simulates the time-dependent model of
physical processes and the equation for them has the form of GDE

@ ðqUÞ
þ divðq~
vUÞ ¼ divðCgradUÞ þ S: ð6:57Þ
@t

Here the generalized variable Φ is regarded as the vector of state for all unknown
variables and the generalized value S—as vector of state for all the source functions.
Therefore, the linearization of the generalized value S is equivalent to the lin-
earization of the source function, and the installation of internal feedback loop of
the global iterative procedure is equivalent to its installation at all local iterative
procedures.
Equation (6.57) is the general equation for describing the dynamics of plasma
processes and the global iterative procedure. In particular, at the convergence of the
solutions @U
@t ¼ 0, and the set values of the variables are the parameters in stationary
processes of the Eq. (6.1).
Equation (6.57) describing the dynamics of iterative procedures using internal
feedback (6.55) and (6.56) can be represented as

@ ðqUÞ
 Sp U ¼ divðCgradUÞ  divðq~
vUÞ þ Sc : ð6:58Þ
@t

Character of the transients for iterative procedure from the initial decision to
installed one in the system is determined by the view of the left side of Eq. (6.58).
At the negative factor Sp all the coefficients in the left side of (6.58) are positive,
i.e., the equation satisfies the Hurwitz stability criterion [9]. It follows that the
196 6 Numerical Methods of the Plasma Physics

linearization of the source function (6.55) using the algorithm (6.56) can lead the
system to stability. Note also that at the large negative coefficient Sp, the transition
(iterative) process is faster, i.e., the convergence condition
  (6.50) is performed
faster due to the exponential function UðtÞ ¼ A exp Sp =q þ C, where A and C—
are some constants.
To analyze the dynamics of iterative procedures with internal feedback a simple
model of new long Arc in the channel is used (6.52). Recall that the traditional
structure of iterative procedures with constant coefficients without installing of the
internal feedback (6.53) led to a divergence of solutions already almost after the
first iteration (Fig. 6.12). For this case, individual feedbacks of the released power
rE 2 and power losses due to radiation Urad with temperature can be set. However,
for simplicity, combine them in a common source function: S ¼ rE 2  Urad , the
internal feedback S(Ф) is implemented using the algorithm (6.55 and 6.56).
The discrete analoguefor the equation of energy balance (6.52) with internal
feedback is the following:

aM TM ¼ aN TN þ aS TS þ bM ;

where
8
< aN ¼ rN nrnM [ 0; aS ¼ rMsrs S [ 0;
r k rk

a ¼ aN þ aS  0:5SpM ðrn  rs Þðrn þrs Þ [ 0;  ð6:59Þ

: M
bM ¼ 0:5ScM ðrn  rs Þðrn þ rs Þ; SM ¼ rEz2  Urad M :

The dynamics of an iterative procedure for the solution of (6.59) with different
coefficients of internal feedback β and the initial approximations for the same grid
are shown in Fig. 6.15. The various initial conditions were selected for analysis
(T0 = 8000, 10,000, 13,000 K) with appropriate coefficients of internal feedback
β = 0.1; 0.5. The computational grid with steps Dr = 0.01—0.1 mm was used.

Fig. 6.15 The change of temperature T (a) and released power rE 2 (b) from iteration to iteration
for the stable solutions with internal feedback for the point on the axis (r = 0) of the Arc channel
(I = 90 A, R = 1.5 mm): 1 β = 0.5, T0 = 10,000 K; 2 β = 0.1, T0 = 10,000 K; 3 β = 0.1,
T0 = 8000 K; 4 β = 0.1, T0 = 13,000 K
6.3 The Structural Organization of Computational Procedures 197

The results (Fig. 6.15a) show that for the procedure with internal feedback the
convergent solutions can be obtained on the computational grid with the steps
Dr = 0.01–0.1 mm with a coefficient of internal feedback b  0:5. If b ¼ 0:5,
convergence is provided for all the points but with some hesitation (the two-way
convergence), hence b ¼ 0:5—a boundary point of the stability for the procedure.
If b ¼ 0:1 the solution converges without hesitation (unilateral convergence) for all
central points with different initial conditions. Small fluctuations, as can be seen,
can be obtained for the points in the area of sharp temperature gradient with the
initial condition T0 = 13,000 K. Thus, the dynamics of the iterative procedure
depends on its structure—the coefficient of the internal feedback and the input data
of the problem: current and power of the Arc, its geometry, type of gas in plasma,
etc., as well as on the initial conditions and the step of the computational grid. All of
the solutions are stable, however, the rapid convergence is obtained by the uni-
lateral convergence b ¼ 0:1 and the initial condition T0 = 10,000 K.
Figure 6.15b shows the variation of the released power by iteration.  It is seen
that due to the inputted coefficient β and obtained negative feedback Sp \0 , the
value of the released power vary from iteration to iteration at reasonable limits, it
promotes the convergence of solution.
This example clearly shows that the new structure of iterative procedures with
internal feedback has the property of stability compared to the conventional
structure (without internal connections). The introduction of negative feedback on
the output values provides the additional information about the nonlinearity and
interdependence of some physical processes inside the procedure of the iterative
solving of equation that is not in the traditional structure. From the point of view of
the theory of automatic control the feedback for the time derivative with the output
value is a powerful tool to ensure the stability [10].

6.3.3 The Mathematical Algorithms with Internal Feedback

As shown above (Sect. 6.3.2), the installation of additional internal feedback loop
allows recovering of the lost information inside the iterative procedure and ensuring
the manageability and stability of the system as a whole. Consider the discrete
analogue of GDE (6.1) considering internal feedback loop and linearization algo-
rithms of the source function for concrete equations of plasma.
The Discrete Analogue of the GDE Based on Internal Feedback Generalized
differential Eq. (6.1) using the linearization of the source function (6.55) has the
following form:

divðq~vUÞ ¼ divðCgradUÞ þ Sc þ Sp U: ð6:60Þ

Using the method of the control volume and the corresponding computational
scheme (Sects. 6.1.1 and 6.2.1), the discrete analogue of the two-dimensional GDE
198 6 Numerical Methods of the Plasma Physics

(6.2) for the control volume (Fig. 6.2) with the internal feedback (6.55) and
(6.56) is

aM UM ¼ aE UE þ aW UW þ aN UN þ aS US þ bM ;

where
8
> aE ¼ De AðjPee jÞ þ maxðFe ; 0Þ; aW ¼ Dw AðjPew jÞ þ maxðFw ; 0Þ;
>
< a ¼ D AðjPe jÞ þ maxðF ; 0Þ; a ¼ D AðjPe jÞ þ maxðF ; 0Þ;
N n n n S s s s
¼ þ þ þ  DV; ¼ DV; ð6:61Þ
>
> a M a E a W a N a S S pM b M S cM
:
DV ¼ ðze  zw Þ ðrn rs Þ2ðrn þ rs Þ :

Similarly the discrete analogue for three-dimensional GDE (6.19) has the form

aM UM ¼ aE UE þ aW UW þ aN UN þ aS US þ aT UT þ aB UB þ bM ;

where
8
> aE ¼ De AðjPee jÞ þ maxðFe ; 0Þ  0; aW ¼ Dw AðjPew jÞ þ maxðFw ; 0Þ  0;
>
>
>
< N ¼ Dn AðjPen jÞ þ maxðFn ; 0Þ  0; aS ¼ Ds AðjPes jÞ þ maxðFs ; 0Þ  0;
a
aT ¼ Dt AðjPet jÞ þ maxðFt ; 0Þ  0; aB ¼ Db AðjPeb jÞ þ maxðFb ; 0Þ  0;
>
> aM ¼ aE þ aW þ aN þ aS þ aT þ aB  SpM DV [ 0; bM ¼ ScM DV;
>
>
:
DV ¼ ðze  zw Þ ðrn rs Þ2ðrn þ rs Þ ðut  ub Þ:
ð6:62Þ

In the calculation of the coefficients for the discrete analogues of (6.61), (6.62)
calculating the Eqs. (6.9)–(6.11), (6.18), (6.21)–(6.23) should be used.
Algorithms of the Installation of the Internal Feedback Loop The first lin-
earization algorithm (6.56) for the absolute values of the variables is illustrated in
Fig. 6.14. The increment of the target variable is formed in absolute value of ΦM,
multiplied by β, that is set based on the analysis of the possible increments Φ in
iteration by simultaneously solving the equations. This algorithm provides a greater
increment for large absolute values of variables, and vice versa. This pattern is
logically justified and is particularly effective in solving the equations and the value
of variables may vary over a wide range, for example, the value of the temperature
or the mass concentration of particles in the plasma.
However, this algorithm has some flaw: in the case when the absolute values of
the variables inside the iterative procedure are small enough, and the source
function takes a very high value, the convergence of the solution can be achieved
very slowly because of the large coefficient Sp. Moreover, when the current values
of the variables are equal to zero, the algorithm cannot work because of the division
by zero. For example, in solving the momentum equations when the components of
the plasma velocity at some points may be equal to zero, or very close to zero. Thus
it is possible to adopt Sp = 0, Sc = S, as well as in the traditional iterative procedure
6.3 The Structural Organization of Computational Procedures 199

without internal feedback loop. However, the experience shows that the simulation
of highly nonlinear objects such as plasma and plasma processes, even the use of a
structure without internal feedback loop in one equation at a single iteration can
lead to loss of stability of the computational procedures.
At the same time in many cases, it is difficult to find the initial values in the
nodes that are sufficiently close to the final values. It is convenient to put some of
the main variables (e.g., temperature) as constant, and the other equal to zero,
especially when they can take both positive and negative values (for example, the
components of the gas velocity). In this case, a second algorithm can be recorded by
for the linearization of the source function not for the absolute value of the variable,
but for a given their increment ΔΦ (Fig. 6.16) [8].
 
 S 
Sp ¼  ; Sc ¼ SM  Sp U: ð6:63Þ
DU

This linearization algorithm of the source function allows a fully getting rid of
the first algorithm lack (6.56) and enabling the zero crossing for the variable inside
the iteration. The source function is linearized for all possible values of the vari-
ables, without exception, in the formation of its coefficients not the absolute values
of the variables, but their increments ðDU 6¼ 0Þ are used. Note that because of
constancy DU the algorithm (6.63) may not be effective for the equations where the
variables can take values over a wide range.
Thus, the third linearization algorithm of the source function can be obtained by
combining the first two algorithms (6.56) and (6.63):
   
S  S 
Sp ¼   for jUj  DU; Sp ¼   for jUj\DU; Sc ¼ S  Sp U:
  ð6:64Þ
bU DU

If the source function is explicitly analytic function SðUÞ, then there are many
variants of linearization for the source function S ¼ Sc þ Sp U. The main thing is

Fig. 6.16 Linearization of

the source function using the
given increment of variables
200 6 Numerical Methods of the Plasma Physics

that it is necessary to choose the appropriate terms Sc ; Sp depending on the physical

formulation of the problem. For example, if the source function can be written as
S ¼ S1  S2 U, then the linearization can be carried out according to the algorithm

Sp ¼ bS2 ; Sc ¼ S  Sp U: ð6:65Þ

In the fourth linearization algorithm of the source function (6.65) the coefficient
β plays the role of internal feedback and is regulated. It shows how many times the
real coefficient Sp effectively increases. For example, the source function
For the equation of the energy balance of atom–ion gas in the two-temperature
plasma model has the form
rffiffiffiffiffiffiffiffiffi
6me 8kTe
S¼ ne ðna Qea þ ni Qei ÞðTe  Ta Þ:
5 pme

The linearization of the source function can be performed as

rffiffiffiffiffiffiffiffiffi
6me 8kTe
Sp ¼ b ne ðna Qea þ ni Qei Þ; Sc ¼ S  Sp Ta :
5 pme

The linearization of the source function is possible even when it is equal to zero.
When using the above four algorithms the linearization of the source function fails
only on condition that if the source function S 6¼ 0. If S ¼ 0 or it is close to zero, it
is possible to artificially create an internal feedback, expanding the source function
into two equal parts S ¼ S  S , where S should be selected on the basis of

physical considerations for  case. The first function S is further a
  the particular
subject for linearization S ¼ Sc þ Sp U using one of the above algorithms.
All of the above algorithms have been tested and successfully implemented in
the simulation of different plasma processes.

6.3.4 The Stability Region of the Computational

Procedures of Plasma Processes

An important task for the practical application of internal feedback is to find a

region of stability in the simulation of plasma processes. In a closed structure with
internal feedback stability of the system is directly related to the installation of this
connection, and in the future we can use the coefficient β to determine the region of
stability of the computational procedures. This area has the closed borders and
depends on the initial conditions and the steps of integration (steps of the grid). For
the iterative procedure the lower limit of the coefficient β is close to zero. In this
case, the source function is linearized using a very large coefficient Sp with the
feedback effect that excessively slows the decision and achieving of the
6.3 The Structural Organization of Computational Procedures 201

convergence. The upper limit of β depends on the conditions of a specific task. The
region of stability for computational procedures can be determined as the multi-
dimensional enclosed space, whose dimension is determined by the number of
equations, included in the model, and each coordinate of the space is the range of
variation for the corresponding coefficient of internal feedback β.
As a result of the simulation of plasma processes the information about the
region of calculation procedures stability at the solving of the basic equations for
plasma in a wide range of input date for RF- and Arc plasma torches is obtained.
Thus, in the solution of the equation of the energy balance it was obtained:
P = 0.5–1 kW/cm3—for RF-plasma torches, P = 1–10 kW/cm3—for free Arc
column and most Arcs in the channel and the anode Arc spots, P = 50–
100 kW/cm3—for the compressed Arc in the channel and for the cathode Arc spots.
Similarly, at the decision of the equation of momentum F = 0.005–1 N/cm3, at the
decision of the continuity equation Δp = 20–5000 Pa and at the decision of the
equation of the energy balance for the electron gas P = 0.5–100 kW/cm3. Steps of
the grids are regulated in the range Δh = 0.001–10 mm for the power of modern
computers. Figures 6.17, 6.18, 6.19 and 6.20 show the stability regions for the
simultaneously solving the different equations in various models of plasma.

Fig. 6.17 The stability

region of iterative procedure
at solving the equation of the
balance energy together with
other equations for the model
of equilibrium plasma
(analogue D-decomposition
of Y. Neumark [10])

Fig. 6.18 The stability

region of the iterative
procedure at solving the
momentum equation together
with other equations of
equilibrium plasma
202 6 Numerical Methods of the Plasma Physics

Fig. 6.19 The stability

region of the iterative
procedure at solving the
continuity equation together
with other equations of
equilibrium plasma

Fig. 6.20 The stability

region of the iterative
procedure at solving the
equation of the balance
energy for the electron gas
together with other equations
of nonequilibrium plasma

Thus, using the internal feedback the sustainability of computational procedures

for solving the nonlinear and related equations can be fully achieved, and the ability
to control the internal feedback coefficient β allows the monitoring the solving of
the equations and managing of the computational processes.

References

1. Patankar S. Numerical methods for solving problems of heat transfer and fluid dynamics: M.:
Energoatomizdat 1984.
2. Numerical Methods research viscous liquid / A.D. Gosmen, V.M. Pan, A.K. Ranchen, etc.
M.: Mir, 1972.
3. Fletcher K. Computational Techniques for Fluid Dynamics. M.: Mir, 1991. V. 1, 2.
4. Paskonov V.M., Polezhaev V.I., Chudov L.A. Numerical modeling of heat and mass transfer.
M.: Nauka, 1984.
5. D. Anderson, D. Tannehill, R. Pletcher. Computational Fluid Mechanics and Heat. M.: Mir,
1990. V. 1.
6. A.A. Samarski. The theory of difference schemes. M.: Nauka, 1989.
References 203

7. Shi Nguyen-Kuok. Ensuring the sustainability of the computational process in the simulation
of heat and gas-dynamic systems with distributed parameters // Proc. of the International
Scientific and Technical. Conf. “Information technologies in modeling and management.”
SPb. 1996. PP. 68–70.
8. Shi Nguyen-Kuok. A stable solution of system of differential equations for modeling the
Low-temperature plasma processes // Proc. of the I International scientific-practical
conference. Conf. “Differential Equations and Applications”. SPb. 1996. PP. 234–236.
9. Besekersky V.A., Popov E.P. The theory of automatic control systems. M.: Nauka. 1972.
10. The theory of automatic control / Ed. by A.V. Netushin. M.: Higher School. 1976.
Chapter 7
The RF Plasma Torches

7.1 Characteristics of the RF Plasma Torches

RF plasma torches—plasma devices that are widely used in science and technology.
RF discharge does not require the electrodes, there is a ring current is one of the few
sources that generate clean plasma electrode materials are not contaminated. These
plasma torches have a relatively large space plasma flow and have a virtually
unlimited service life in any gaseous medium. RF plasma torches (Fig. 7.1) allow
obtaining the pure plasma with a temperature of 7000–11,000 K at atmospheric
pressure. The heating of the plasma without electrodes in the RF plasma torches of
the induction type is described using the law of induction and conditions of the
heating in a conducting medium and in a high-frequency alternating electromag-
netic field. The conducting medium, in this case, is ionized gas—plasma.
High-frequency magnetic field of the inductor generates an alternating magnetic
flux dФ/dt variable and induces an electromotive force (EMF), whereby currents
having annular iu differing in amplitude and phase and having a spatial distribution
of the plasma in the induction.
There are the following types of discharge chambers of the RF plasma torches
(Fig. 7.2): the quartz chamber that is not cooled by water; the quartz water-cooled
chamber; the metal chamber with cuts; the porous chamber.
The stabilization of the plasma in the discharge chamber of the RF plasma
torches is generally carried out by specially formed gas flows. There are three ways
of such stabilization: the axial gas stabilization; the gas vortex stabilization; and
stabilization by unformed gas flow.
The temperature of the RF plasma torches The investigations of the temperature
distribution in RF plasma were carried out by many authors [1–11]. The spatial
distribution of temperature and electrical conductivity for the RF Argon plasma are
shown in Fig. 7.3a. The radial temperature distributions in argon plasma and in
plasma of molecular gases in the area of maximum temperature, according to
different authors, are represented at Fig. 7.3b, c. The results shown in Fig. 7.3 are

© Springer International Publishing Switzerland 2017 205

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_7
206 7 The RF Plasma Torches

Fig. 7.1 The installation of the RF plasma torch (f = 27.12 MHz, the discharge power
P2 = 3 kW) (a) and an Argon plasma in the inductor with gas consumption QAr = 30 l/min (b)

obtained for the plasma torches with the quartz chamber when the gas vortex
stabilization of the RF plasma. The average temperatures in the central part of
plasma depending on the power of plasma discharge and the gas consumption are
shown in Fig. 7.4.
The results of the general analysis of temperature represented in Figs. 7.3 and
7.4 show that the maximum of temperature in the RF plasma torches is located not
on the axis of the plasma torch, but in the zone of maximum current density and the
released power. The values of temperature obtained by the different authors in
various conditions at certain frequencies of generator, diameters of plasma torches,
the powers of discharge are not very different to each other. Thus, for RF Argon
plasma the maximum of temperature is 9000–11,000 K, for Air and Nitrogen
6500–8500 K, for oxygen 8500–9500 K. Such a coincidence of the results is due to
the fact that in all experiments, the power per unit of plasma volume is changed
slightly.
Thermal nonequilibrium plasma in the RF plasma torches is found by
researchers [12–14]. The first attempt to quantitatively measure the effect of
nonequilibrium was made by author in [12]. The electron temperature in Argon was
measured by the absolute intensity of the recombination spectrum of ArI in the
wavelength λ = 430–480 µm. The measurements were carried out (on the cut) of
the RF plasma torch with a tube diameter of 28 mm with a power of discharge in
plasma 7–8 kW and the frequency of generator 17 MHz.
The temperature of the atom–ion gas was determined by hot-wire anemometer,
i.e., the heat flow going to the water-cooled copper cross-flow tube with diameter of
1.35 mm. These measurements of the total heat flow via a tube from the center to
the edge of the plasma jet by means of conversion were leaded to the values of the
heat flux in the local point. Heat flow in accordance with the theory of heat transfer
7.1 Characteristics of the RF Plasma Torches 207

Fig. 7.2 The main types of discharge chambers in the RF plasma torches [1]: a the quartz
chamber that is not cooled by water (axial gas stabilization); b the quartz water-cooled chamber
(unformed axial flow); c the quartz chamber without water cooling (vortex stabilization); d the
metal cooled split chamber with a quartz sheath; e the metal cooled chamber with refractory
inserts; f the quartz water-cooled chamber (axial stabilization); g the same (vortex stabilization);
h the metal split cooling chamber (axial stabilization); i the same (vortex stabilization)

is related to the enthalpy (i.e., the temperature of the heavy component) and it
allows obtaining of the value Te and Ta in the same local point. It was found that on
the axis of the plasma torch the temperature of electrons is not equal to the tem-
perature of the ions and atoms: (Te = 10,200 K; Ta = 7000–8000 K) [12].
In [13] the relaxation method for measuring of Te and Ta was used. It is based on
the temperature difference between the relaxation times of the atom–ion and elec-
tron gases. It was estimated [13], the relaxation time for Te is τe = 5 × 10−6 s and
for Ta − τa = 10−2 s. Measurements of Te and Ta in such method [13] were made at
the following conditions: frequency f = 26 MHz, the plasma discharge power of 3–
6 kW, the diameter of quartz chamber d = 2R = 20 mm. The grid modulation
scheme (quenching) of the high voltage and automatic recording of the intensity of
the spectral lines and the continuum on a fast spectrograph were applied. The results
of temperature measurements by this method are presented in Table 7.1.
208 7 The RF Plasma Torches

Fig. 7.3 Temperature T and electrical conductivity σ in the axial plane of the RF plasma torch in
argon at a discharge power P2 = 2.5 kW, the inside diameter of the plasma torch d = 30 mm,
f = 26 MHz, Q = 15 l/min [2] (a) and radial temperature distribution in an argon (b), nitrogen,
oxygen, and air (c) RF—plasmas according to various authors: b 1—[3]; 2—[4]; 3—[5]; 4—[6]; 5
—[7]; 6—[8]; 7—[2]; 8—[9]; c 1, 2, 3—[11]; 4—[10]

Fig. 7.4 The average temperatures in the RF plasma torches from the gas consumption
Q (a) (P = 30 kW) and from the power of plasma P (b): triangle Q = 250 l/min; square
Q = 75 l/min; the inner diameter of the quartz plasma torch 60 mm; f = 5.28 MHz

Table 7.1 The results of The coordinate Te Ta Te − Ta

measurements of the electron r/R (103 K) (103 K) (103 K)
and atom–ion temperatures
using the gas relaxation 0.9 9.3 6.65 2.63
method [13] 0.5 10 8.3 1.7
0 10 8.7 1.3

One of the carefully the temperature of electrons and ion–atom gas in the RF
Argon plasma was measured by authors in [14]. In this work, in addition to hot-wire
anemometer, the method of enthalpy probe was used. RF discharge was excited in
7.1 Characteristics of the RF Plasma Torches 209

Table 7.2 The values of Te U1 Q (l/min) Te Ta P2

and Ta in induction argon (kV) (103 K) (103 K) (kW)
plasma at the exit from the
plasma torch according to 3 5 9.5 9.0 13.5
[14] 3 10 9.5 6.5 15.8
3 15 9.5 5.0 19.0
3 20 10.0 4.0 19.5
3 30 10.5 4.0 20.5
2 5 8.5 5.5 6.0
2 10 9.0 4.0 8.2
2 15 9.5 3.0 9.4
2 20 10.0 3.0 9.5
2 30 10.5 3.0 9.5

the metal water-cooled chamber with an internal diameter of 32 mm. Generator

with the frequency of 6 MHz was used, power dissipated in plasma was about 4–
6 kW. The plasma-forming gas was axially supplied into the discharge and its
consumption was varied from 1 to 30 l/min.
The measurement results for Te and Ta for axial point r = 0 in various regimes
are shown in Table 7.2.
It also shows the total power of the induction plasma torch P2. Values of Te at
the axis of plasma torch changing with the changing P2 and Q vary slightly in the
wide diapason. This is confirmed by many measurements of different authors, and
the temperature of the ion–atom gas is strongly dependent on consumption of argon
and voltage in the inductor (Fig. 7.5a). With increasing consumption Q, the tem-
perature of atoms and ions falls sharply. For example, for Q = 20–30 l/min, Ta is
about 3 times less than Te. At low consumptions (1–5 l/min), when U1 = 2.5–3 kV
temperature Te at the axis coincides with Ta. However, at the voltage across the
inductor U1 = 1.6–2 kV, Ta, even at low flow consumptions Te is well below.
At Fig. 7.5b radial temperature distribution in the measured cross section for two
cases is shown: (1) on an axis Te = Ta (Q = 5 l/min); (2) Te ≠ Ta, (Q = 20 l/min).
The figure shows that even in the case where the plasma at the axis of the discharge is
equilibrium, in the periphery region there is a significant discrepancy of Te, and Ta.
From the data [15–17] it follows that the primary mechanism of disorders of the
thermal equilibrium is the penetration of a “cold” plasma gas in plasma clot. The
“cold” gas incoming into plasma is almost no change the electron temperature, but
significantly reduces the atom–ion temperature, but the consumption of gas,
“captured” by plasmoid is exclusively determined by the value of the magnetic
overpressure (magnetic pinch effect), and very weakly depends on the consumption
of gas supplied to the plasma torch. Gas that is not captured by plasmoid wraps
around its front part. In the measurement results [12–14] it is shown that a large
temperature difference between the electrons and the atom–ion gas is observed only
in argon plasma. In the plasma of molecular gases Te and Ta have only the slightly
difference.
210 7 The RF Plasma Torches

Fig. 7.5 The electron and atom–ion temperatures (Te and T) at the cut of the RF plasma torch on
the various voltages of inductor U1 (a) and gas consumption Q (b): a open diamond, filled
diamond—U1 = 3 kV; open square, filled square—U1 = 2,5 kV; open circle, filled circle—
U1 = 2 kV; open triangle, filled triangle—U1 = 1,6 kV; b 1—Q = 5 l/min; 2—Q = 20 l/min

Gas dynamics in the RF plasma torches In numerous studies of gas dynamics of

the RF plasma torches [12, 18–22], the speed of the plasma jet, flowing into the
open space, was measured. These experiments are interesting from the point of
view of the use of the plasma jet for the technological processes, but do not contain
enough information for the analysis of the characteristics of the RF discharge.
Research of [23, 24] showed that the plasma torch with the vortex stabilization has
a very complex distribution of velocities. Tangential entrance of the gas into the
discharge chamber leads to the formation of the vortex nucleus zone where there is
a reverse current from the central zone of the plasma discharge. In the area of
inductor the gas flow near the wall is directed to the exit of the plasma torch, in the
axial region—in the opposite direction. Due to this fact there is a high recirculation
of plasma in plasma torch with vortex stabilization, and it leads to decrease in the
proportion of power supplied to the jet.
However, analysis of the experimental data of different authors shows that both
in tangential and axial entering of the plasma-forming gas the backflow zone is
observed in the inductor. The mechanism of gas-dynamic processes occurring in the
induction plasma torch, based combustion model, was suggested by [25]. The area
of active energy release in it—is a skin layer, where the heating and acceleration of
the gas flow occurs in direction from the wall towards the axis. The lines of the gas
current are refracted in the skin layer, and cold gas penetrates to the central region
of the plasma torch through the side surface of the discharge. Productivity of
pumping through the skin layer depends primarily on the power released at him. If
the supplied gas does not have time to pass through the plasma, the plasmoid is a
7.1 Characteristics of the RF Plasma Torches 211

tube, and as a result of the formed “affluent” gas may be ejected from the area of
inductor and the plasma torch is extinguished. If the gas is fed too little, then there
is an internal recirculation of the gas. Then the appearance of local flows in the
direction opposite to the jet stream perhaps. However, the experimental verification
of the model showed that gas flow direction inside the plasmoid is absent. This
means that neither the thermal mechanism nor the combustion mechanism [25], are
not basic in this case.
Depending on the position of researchers the magnetic pressure in the RF dis-
charge takes a decisive role in the dynamics of plasma flows [20], or it is ignored
[26]. Chase [20, 21] pointed to the decisive role of the magnetic pressure in the
induction plasma when the gas is accelerating. Conclusions of J. Chase are con-
firmed in [27], here the calculation results of the plasma jet velocity are obtained by
solving the energy balance equations, the Maxwell equation, and the momentum
equation. The calculation results are in good agreement with the experimentally
measured values of the velocity of plasma jet. The measurements of air plasma
velocity at the exit of the RF plasma torch are shown at Fig. 7.6. The experimental
[28] data and the data of calculations [27] are in good agreement that plasma
acceleration (in axial direction) occurs at the boundary of plasmoid and open space
at the side of open space due to overpressure in the bunch (magnetic pinch effect).
Cold gas, that is penetrating in a radial direction into a central discharge zone, leads
to decrease in temperature on the axis of the plasma torch in the radial direction.
Magnetic pressure gives a significant acceleration in the discharge of the heated gas
toward the exit and the proportion of the power transmitted in the stream in the RF
plasma torch will be higher than for vortex plasma torches. The mechanism of the
gas acceleration by the electromagnetic forces, of course, exists also in the vortex
plasma torches.

Fig. 7.6 Radial velocity

distribution of the RF Air
plasma torch (f = 5.28 MHz)
at z = 20 cm from the cut at
various discharge powers:
filled diamond—P = 53 kW,
filled square—P = 39 kW,
filled circle—P = 25 kW,
Q = 50 l/min
212 7 The RF Plasma Torches

The gas acceleration in the RF plasma torch can be described as follows: RF

discharge glows inside the inductor and the induced currents iφ in the discharge
interact with the magnetic field (for example, with a longitudinal component Bz of
the magnetic field induction), a electromagnetic force Fr occurs as a result, com-
pressing the plasma, and plasma is driven in the direction of the axis (Fig. 7.7a). In
the future the mechanism of the thermal acceleration leads to the motion of plasma
stream on the basis of the law of mass conservation. The amount of gas is drawn
into the center of plasma torch in radial direction, and the same is supplied to the
exit of plasma torch in the axial direction. The external “hot” layer of the gas
penetrates into plasma in the radial direction, is reacted with plasma, is rapidly
heated and accelerated (axially) as the penetration into discharge.
Thus, the electromagnetic force plays an important role in shaping of the plasma
flow movement in the RF plasma torch. This force accelerates the plasma in the
direction of the center axis of the plasma torch, leads to the emergence of front and
aft vortex flows at the top and bottom (with sufficient force of inductor current)
zones of plasma torch (Fig. 7.7b). The formation of such areas is known in the
theory of induction heating as the effect of electromagnetic stirring.
However, the component of the electromagnetic force Fr is able not only to carry
away the gas in the center of discharge, followed by heating and acceleration, but at
the same time form an excessive magnetic (gasostatic) pressure in the discharge,

Fig. 7.7 The action of electromagnetic force (a) and qualitative picture of gas flow in the RF
plasma torch (b): I1—current of inductor; Fr—electromagnetic force that occurs as a result of
interaction of the terminal plasma current with a longitudinal magnetic field; iφ—the induced ring
current
7.1 Characteristics of the RF Plasma Torches 213

that has a maximum on the discharge axis. Thus the gradient of gas-static pressure

@P
@r and component of electromagnetic force Fr are in opposite directions. There
are just two mechanisms: the acceleration of plasma flow by the action of the
electromagnetic force Fr and braking (preventing to the penetration of the gas flow
into discharge) due to the effect of the gradient force gas-static pressure. Thus, the
resulting force Fr
@P@r causes to acceleration of plasma stream in the radial
direction that deliver a small amount of gas into discharge. Our most careful cal-
culations based on a full simulation of the RF plasma torches show that the
resulting force of the plasma acceleration Fr
@P @r —is only (5–10) % of the elec-
tromagnetic force Fr . It allows enter into discharge in the radial direction to the
external hot layer of gas (T [ 3000 K) that is a small fraction of the total number of
blown gas. Such a general mechanism is called as magnet-gas-static acceleration in
the channel of plasma torch.
Therefore, assumption of some researchers about the equality of the electro-
magnetic force and the gradient of static pressure or disregard of them in solving the
equations of plasma motion are erroneous. This undoubtedly leads to an erroneous
determination of the resulting force and acceleration of plasma to an inaccurate
calculation of the plasma flow velocity.
The answer to the question how much gas can penetrate into the discharge is
found due to the interaction of two processes occurring simultaneously in the
channel of plasma torch: acceleration and gas heating at a determination of dis-
charge power. We see that for a given power the steady burning of plasma allows
amount of gas to penetrate into plasma and this amount warms up to the high
plasma temperature, that provides the conductivity and efficient allocation of energy
in the discharge.
We assume that for some reason, more gas with some power than discharge can
heat penetrates into the discharge. Then the external skin layer of discharge cools
automatically and the temperature and electromagnetic force Fr acting on the
plasma flow decrease. As a result the amount of gas entering to plasma reduces.
Thus, the discharge automatically adjusts and maintains the steady burning in
electromagnetic field of inductor and keeps the high temperature inside the channel
of the RF plasma torch. This is possible due to the internal mechanism of negative
feedback of thermodynamic and electromagnetic processes in the discharge. Due to
the installation of such links it is possible to simulate these processes in the
mathematical model and to ensure the stability of computational procedures for
simulation of plasma processes (Sect. 6.3). Excessive amount of cold gas moves
past near the plasma flow and forms a natural thermal protection layer along the
wall of the plasma torch.
We see that the movement and heating of the plasma flow in the channel of
plasma torch are the result of interaction of electromagnetic, thermodynamic and
plasma processes in the RF discharge. The picture of its movement and heating can
be based on the simulation of all these processes.
214 7 The RF Plasma Torches

7.2 Electromagnetic Field in the RF Plasma Torches

The allocated power in the RF discharge and the degree of gas heating, as well as
electromagnetic force acting on the plasma flow and plasma acceleration, have
direct connection with the distribution of electromagnetic field; the calculation of
the electromagnetic field inside the channel of plasma torch is also an important
issue in the simulation of the RF plasma torches.
The RF plasma torch is a complex electromagnetic system. Plasma flow with
different temperatures can be regarded as a conductive medium in the RF—plasma
torch. It has a certain distribution of electrical conductivity along the length and
radius of the channel of plasma torch and the magnetic field of inductor induces the
terminal currents in plasma that are different to each other in amplitude and phase.
The interaction of two fields generated by the current of inductor and plasma
currents, it defines a single electromagnetic field inside the RF plasma torch.
The first works where the RF discharge was regarded as electrodynamic system
with distributed parameters, should be considered as works by J. Thomson [29, 30].
Applying the solution of equations of electrodynamics for a long solenoid, he
justified the magnetic induction nature of discharge without electrodes and defined
the ignition threshold of discharge and certain conditions of its existence.
The number of works that are devoted to experimental measurements of the
distribution of electric E and magnetic H fields and the current density j in the RF
plasma torches is small. This is because of significant difficulties due to conse-
quence of high temperatures to the sensors of measuring systems and due to plasma
disturbances by the sensors at the change of its parameters. Therefore the tem-
perature distribution and one of the parameters (E or H) are the most often deter-
mined. Other characteristics of discharge are calculated on the basis of these
measurements using one of mathematical models. This is the approach that is used
in [31–33] for study of the magnetic field topology in the RF plasma torches using a
water-cooled magnetic probe. The scheme for measurement of the parameters of the
RF plasma torch (f = 2.6 MHz, power 25 kW) are shown in Fig. 7.8, and the
measurements of magnetic field in the central section of the inductor [32] are
presented in Fig. 7.9. Measurements were carried out for two cases: in empty and
loaded with plasma inductors at the same values of current in inductor I1 = 68 A,
the voltage at the inductor U1 = 5.7 kV. Results of conductivity measurements and
determination of electric field and current density in a cross section are shown in
Fig. 7.10.
Unfortunately, the lack of information about the distribution of longitudinal
magnetic field makes it impossible to obtain the integral parameters of the RF
discharge. However, experimental data [31–33] allow making the estimate calcu-
lations and obtaining a qualitative picture of electromagnetic field inside the RF
plasma torches.
Calculation of the electromagnetic field of the RF plasma torches is usually
based on the solution of Maxwell equations
7.2 Electromagnetic Field in the RF Plasma Torches 215

Fig. 7.8 Scheme of the field

measurement using a
magnetic probe [32]:
D1 = 155 mm—the internal
diameter of discharge
chamber, D2 = 199 mm—the
internal diameter of inductor,
D3 = 206 mm—diameter of
the inductor on the coil axis,
Li = 134 mm—the length of
inductor, d = 3 mm—
diameter of the magnetic
probe

Fig. 7.9 Measurements

result [32] of the magnetic
intensity in unloaded inductor
(1) and in the RF Argon
plasma torch (2)
216 7 The RF Plasma Torches

Fig. 7.10 Radial distribution

of the electric field in the RF
Argon plasma torch [32]: 1—
the electrical conductivity of
plasma σ; 2 the current
density j; 3 the electric
intensity E

~ ¼ r~ ~ @~
B
rotH E þ @ D=@t; rot~
E¼
; div~
B ¼ 0; div~
E ¼ qs ; ð7:1Þ
@t

where r~ ~
E, @ D=@t—current density of conduction and of displacement in plasma; in
the area larger than the Debye radius (r > rD) and quasi-neutral plasma ðZni ¼ ne Þ
and qs ¼ 0.
Depending on the design of inductors there are two types of models of the
electromagnetic field (the electromagnetic problem) of the RF plasma torches:
• one-dimensional problem that is used to calculate the length of inductor with
plasma: the existence of only one component of the magnetic field Hz is allowed
(Hr ¼ Hu ¼ 0).
• two-dimensional problem, that can be used to calculate the inductor with
plasma: two components of the magnetic field Hz , Hr , (Hu ¼ 0) are considered.
In both cases, the cylindrical shape of the inductor and the ring currents in the
plasma allow only one component of the electric field Eu , (Ez ¼ Er ¼ 0).
One-dimensional electromagnetic problem Method of solving the one-dimensional
problem of calculation the length of inductor with plasma is quite simple. We briefly
outline it.
For the one-dimensional problem using Hz , (Hr ¼ Hu ¼ 0) and
Eu (Ez ¼ Er ¼ 0) the system of Eq. (7.1), written in the cylindrical coordinate
system, can be reduced to the following form:
7.2 Electromagnetic Field in the RF Plasma Torches 217

Fig. 7.11 The grid and control volumes for solving the one-dimensional electromagnetic problem
of long inductor with plasma


1d r dH_ z

jxl0 H_ z ¼ 0; ð7:2Þ
r dr r þ jxe0 dr

Here H_ z  H—the
_ complex intensity of the magnetic field.
The solution of Eq. (7.2) is carried out by the method of control volume (Sect. 6.
1.1). Balance of field and the discrete analogue are obtained by integrating (7.2) of
the control volume (Fig. 7.11).

Zrn
 Zrn
1d r d H_ _
rdr
jxl0 Hrdr ¼ 0:
r dr r þ jxee0 dr
rs rs

After integration we will have



r dH_ r dH_


0; 5jxl0 H_ M ðrn
rs Þðrn þ rs Þ ¼ 0;
r þ jxe0 dr n r þ jxe0 dr s

Here the first derivative of the magnetic field at the edges of the discrete analogue is
determined as follows:
 
dH_  H_ N
H_ M dH_  H_ M
H_ S
¼ ; ¼ :
dr n rN
rM dr s rM
rS

Consequently, the discrete analogue for Eq. (7.2) at internal points is:

aM H_ M ¼ aN H_ N þ aS H_ S þ bM ; ð7:3Þ
218 7 The RF Plasma Torches

where

aN ¼ ðrn þ jxer0nÞðrN
rM Þ ; aS ¼ ðrs þ jxer0sÞðrM
rS Þ
aM ¼ aN þ aS þ 0:5jxl0 ðrn
rs Þðrn þ rs Þ

_
For efficient use of the boundary conditions on the symmetry axis (ddHr  ¼ 0)
r¼0
and on the boundary of a plasma clot (r ¼ R2 ; H_ ¼ HR2 ) the computational grid is
constructed so that the first point is at a distance of a half-step from the symmetry
axis (rM ¼ Dr=2), and the last—in the region R2 (Fig. 7.11). Then, for the first
control volume in accordance with the boundary condition on the axis of symmetry
we have the discrete analogue (7.3) with as ¼ 0. For points at the distance R2,
H_ ¼ HR2 .
Wherein
sffiffiffiffiffiffiffiffiffiffiffi
I1 W R1 0:5h
z 0:5h þ z
HR2 ¼ arctg þ arctg
ph R2 ðzÞ R1
R2 ðzÞ R1
R2 ðzÞ
sffiffiffiffiffiffiffiffiffiffiffi!
 ð7:4Þ
R2 ðzÞ 0:5h
z 0:5h þ z

1
arctg þ arctg ;
R1 R1 R1

where R1—the radius of inductor; W—the number of turns in inductor; h—the

length of inductor; z—coordinate of the plasma cross section, measured from the
geometric center of inductor.
Two-dimensional electromagnetic problem The one-dimensional problem and
the assumption of one-dimensional electromagnetic field (Eu , Hz ) are justified, if
the edge effects in inductor are neglected. Physically, it is possible, if the ratio of
length to diameter of inductor is much larger than unity and a current in the inductor
is of relatively low frequency. However, for most real designs of the RF plasma
torches the length of inductor is commensurate with its diameter (h  D), and the
frequency of the current in the inductor can vary from a few to hundreds of
megahertz. In this case, two-dimensional and edge effects play a significant role and
consideration of two-dimensional electromagnetic field is necessary.
For two-dimensional electromagnetic field in the channel of the RF plasma
torches, due to the symmetry of the ring current and inductor current in plasma
(transfer current is equal to zero) we have the following relations for the compo-
nents of electric and magnetic fields Ez ¼ Er ¼ 0; Eu 6¼ 0; Hu ¼ 0; Hz 6¼ 0;
Hr 6¼ 0, and Maxwell equations can be written in the two-dimensional cylindrical
system coordinate as
)
@E
@Hr
@z
@H
@r ¼ rEu þ e0 @t ;
z u @Hz 1 @
@z þ r @r ðrHr Þ ¼ 0;
@Eu @Hr @ @H
ð7:5Þ
@z ¼ l0 @t ; 1r @r rEu ¼
l0 @t z :
7.2 Electromagnetic Field in the RF Plasma Torches 219

There is an effective method for solving this system of equations using the
magnetic vector potential ~
A, its advantage is the reduction in the number of
equations to a single equation for ~
A that simplifies greatly the solution of the
problem.

7.2.1 Equation of the Vector Potential and Its Solution

General method for solving the two-dimensional electromagnetic problem is to find

the distribution of the (magnetic) vector potential ~ A (~B ¼ rot~A) [34, 35] that for
symmetric ring current has only one component A  Au , Az ¼ Ar ¼ 0. Using the
vector potential allows reduction of the number of equations that are necessary to
solve to a minimum. Instead of solving the Maxwell equations (7.5) for the
intensities of electric and magnetic fields Eu , Hz , Hr , it is necessary to solve only
one equation for the vector potential. For solving of general equations of the
electromagnetic field for the vector potential ~ A, imagine that in each elementary
region of the RF plasma torch, including possible areas of inductor, plasma, and
whole space, there are the circular currents of conduction (current in plasma r~ E and
~ ~
in inductor jc ), and the displacement current @ D=@t, i.e., the total current density
~j ¼ r~E þ~jc þ @ D=@t;
~ and it can be assumed that the current density in inductor ~jc is
determined as the ratio of the current in inductor ~ Ic to the area S of the conductive
section of inductor. For RF plasma torch the high-frequency current is substantially
only on the internal side of inductor, since there is minimum inductance resistance
in inductor, so a conductive sectional area of the internal side of the inductor should
be considered. Presentation of total current density by the sum of current densities
of conduction in inductor, in plasma, and displacement current allows obtaining the
convenient equation for determination of the current density in any area of research,
and not only inside the inductor or plasma flux. Thus, the total current density
inside the plasma flux~j ¼ r ~ ~
E þ @ D=@t, in inductor~j ¼ ~jc þ @ D=@t,
~ and in the other
space, such as in the gap between the coils or between inductor and plasma.
Equation of the vector potential has the form [34]

@~
A @ 2~
A
D~
A
l0 r
e0 l0 2 ¼
l0~jc : ð7:6Þ
@t @t

This equation, that describes the change of vector potential in space and time,
can be used in the future to determine the dependence of the parameters of elec-
~
tromagnetic field from the current in inductor and in plasma. Here the term l0 r @@tA is
2~
due to the current of conduction in plasma, e0 l0 @@tA2 —is due to the displacement
current l0~jc —is due to the current in inductor. Thus, the equation of the vector
220 7 The RF Plasma Torches

potential (7.6) describes the change of the field in all the space, and not only inside
the RF discharge.
Equation (7.6) in a cylindrical coordinate system based on only one component
(Au ¼ A) has the form


@2A 1 @ @A A @A @2A
þ r

l r
e 0 l ¼
l0 jc : ð7:7Þ
@z2 r @r @r r2 0
@t 0
@t2

The solution of this equation is carried out by means of submission of the

required variable A in the complex plane. By definition, the image of a sinusoidal
vector potential A ¼ Am sinðxt þ wa Þ in the complex plane has the
_ jx t , A_ ¼ Aejwa , where j2 ¼
1; A—the
form:A ! Aejðx t þ wa Þ ¼ Ae _ complex current
value of the vector potential; A—the current value of the vector potential; wa —the
initial phase of the vector potential with for the current in inductor, the initial phase
of the current in inductor is set equal to zero wi ¼ 0.
Equation (7.7) after the substitution of complex value of the vector potential
becomes

_

@ 2 A_ 1@ @A 1 2 _
þ r
þ jl xr
e 0 l x A ¼
l0 jc : ð7:8Þ
@z2 r @r @r r2 0 0

For simplicity we denote A  A_  A_ u

Vector potential Eq. (7.8) is obtained from Maxwell equations, so it like
Maxwell equation expresses the law of conservation of the electromagnetic field in
an infinitesimal volume. Therefore, for the implementation of this law of conser-
vation of the field in the final control volume and in all the study area, we should
integrate the Eq. (7.8) for each control volume (Fig. 7.12) and make up the balance
of electromagnetic field for each control volume that will ensure the implementation
of the law of conservation of electromagnetic field in the entire region. The result of

Fig. 7.12 The control

volume and grid points for the
solution of the equation of the
vector potential
7.2 Electromagnetic Field in the RF Plasma Torches 221

integration is the discrete analogue for equation of vector potential for the control
volume. Integrate the vector potential Eq. (7.8) in two-dimensional control volume
(Fig. 7.12).

Zrn Zze Zze Zrn


@2A 1@ @A
dzrdr þ r rdrdz
@z2 r @r @r
rs zw zw rs
Zze Zrn
 Zze Zrn
1

þ jl0 xr
e0 l0 x2 Ardrdz ¼
l0 jc rdrdz:
r2
zw rs zw rs

At the same time, we note that under the integral of the first and second terms
there is a return differential of the same order, so we get

Zrn


  Zze



@A @A @A @A

rdr þ r
r dz
@z e @z w @r n @r s
rs zw
ð7:9Þ
Zze Zrn
 Zze Zrn
1

þ jl0 xrr
e0 l0 x2 r Adrdz ¼
l0 jc rdrdz:
r
zw rs zw rs

The third and fourth terms of this equation does not depend on the coordinates,
so take them integrands constant in the central point M.

Zze Zrn

1
þ jl0 xrr
e0 l0 x r Adrdz
2
r
zw rs


1
¼ þ ðjrM
e0 Þl0 xrM ðze
zw Þðrn
rs ÞAM ; ð7:10Þ
rM
Zze Zrn
l0 jc rdrdz ¼ l0 Ic rM :
zw rs

Further integration of the first and second members (7.9) should be made on
each side of the control volume. The average values of the integrand in points e, w,
n, s are accepted as an admission. The derivatives under the integral can be
approximated as:
   
@A AE
AM @A AM
AW @A AN
AM @A AM
AS
¼ ; ¼ ; ¼ ; ¼ : ð7:11Þ
@z e zE
zM @z w zM
zW @r n rN
rM @r s rM
rS
222 7 The RF Plasma Torches

Substituting (7.10) (7.11) in (7.9) we get



AE
AM AW
AM ðrn
rs Þðrn þ rs Þ AN
AM AS
AM
þ þ rn þ rs ðze
zw Þ
zE
zM zM
zW 2 rN
rM rM
rS


1

þ ðjrM
e0 Þl0 xrM ðze
zw Þðrn
rs ÞAM ¼
l0 Ic rM :
rM

And the discrete analogue of Eq. (7.8) is

aM AM ¼ aE AE þ aW AW þ aN AN þ aS AS þ bM ; ð7:12Þ

where
8
ðrn
rs Þðrn þ rs Þ ðrn
rs Þðrn þ rs Þ
>
> aE ¼ 2ðzE
zM Þ [ 0; aW ¼ 2ðzM
zW Þ [ 0;
>
>
>
>
>
> a ¼ rnrðNze
r

zw Þ
[ 0; as ¼ rsrðMze
r

zw Þ
[ 0;
< N M S
aM ¼aE þ aW þ aN þ aS þ Sp [ 0;
>
> Sp ¼ r1M þ ðjrM
e0 xÞl0 xrM ðze
zw Þðrn
rs Þ;
>
>
>
>
>
> bM ¼ l0 Ic rM for the points on inductor;
:
bM ¼ 0 for the points out of inductor:

Solution of the system of discrete analogues (7.12) for all control volume gives
the distribution of the magnetic vector potential ~ A for the coordinates r, z. The
~
resulting distribution A, in turn, can be used to further define the parameters of the
field in the RF plasma torch. Note, that for solving the system (7.12) it is necessary
to define the boundary conditions for the vector potential.

7.2.2 Boundary Conditions of the Equation of the Vector

Potential

The boundary conditions for solving the equation of the vector potential is generally
defined in the initial section, or at the inlet of the plasma torch, on its wall, the axis
of symmetry and at the exit. Consider ways to determine the vector potential at
these borders
The zero boundary condition at infinity border The (magnetic) vector potential
~
A has the direct communication with the electric field intensity ~
E, and at an infinite
distance from the sources of the current value of the vector potential and electric
field are equal to zero: ~A¼~ E ¼ 0 if l ¼ 1. For RF plasma torches due to the
attenuation of the electromagnetic field far from inductor the approximate zero
boundaries can be taken at a distance of a few diameters of inductor, and the step of
7.2 Electromagnetic Field in the RF Plasma Torches 223

the grid out of inductor can be made larger due to lack of currents of conduction in
this area.
This way of defining the boundaries is simple but it is necessary to solve the
equation of the vector potential for a much larger area, because the distribution of it
should be determined, not only inside the inductor, but also outside it, although the
parameters of the field of inductor are not of great interest in simulation of plasma
processes inside the channel of plasma torch.
The boundary condition for the sum of the currents The second method of
finding the vector potential at the boundary is based on the principle of superpo-
sition of the field as a sum of the external and internal currents in inductor and
plasma. In this case, the domain for determination of the field is usually limited by a
cylindrical channel of the plasma torch and the solution of the equation for the
vector potential of (7.8) is carried out only for this area. The internal side of the wall
of the plasma torch can be chosen as the external border. Determination of the value
of vector potential for points on the boundary using the superposition principle,
however, requires a large volume of computation, because the value of the vector
potential in each such point depends on the cumulative effect of all the currents in
the inductor and in plasma. The currents in plasma are obtained from the solution of
Eq. (7.8) and change in the iterations. In this case, we have the solution of (7.8)
with simultaneous approximation of its boundary condition.
We assume that the current in inductor is known and is constant for all the turns
of inductor. The currents in plasma have different amplitudes and phases. The phase
shift of the currents in plasma are different from the current in inductor, i.e., current
in inductor can be presented as a real value, and the currents in plasma—the
complex values. To further find the vector potential at the boundary of the plasma
torch according to the principle of superposition of fields it is necessary to deter-
mine the field of loop with current.
The electromagnetic field of one circuit with the current The electromagnetic
field outside the circuit with the current (where there is only the displacement
current) is described by the magnetic vector potential

@ 2~
A
D~A
e0 l0 2 ¼ 0: ð7:13Þ
@t
The solution of this equation is:
Z ~
~ l jðtÞdV
A¼ 0 ; ð7:14Þ
4p r
V

Here the integral is taken over the volume where the current ~i flows.
At Fig. 7.13 the origin of coordinates 0 is placed in the center of circuit with
current. The axis 0z is directed perpendicular to the plane of the loop. Here Rc —the
224 7 The RF Plasma Torches

Fig. 7.13 To the definition of electromagnetic field with a loop with current

radius of the loop current; R—radius of the line for vector ~

A, with center at the point
P; r—distance from the point M where the vector potential is determined, to the
element of circuit d~l(r ¼ MQ); φ—angle of rotation 0Q for 0N or angle between the
vector potential at the point M and the vector of element with a current circuit d~l;
~j—the current density vector; elementary volume dV ¼ dSdl. We write
Z ~ Z Z Z Z
~ l jðtÞdV l0 d~l l0 ~iðtÞd~l l0~i d~l
Au  ~
A¼ 0 ¼ ~ ~
jdS ¼ ¼ :
4p r 4p r 4p r 4p r
V l S l l

The integration is performed around the loop with current. The magnetic vector
potential has only one component A, i.e., the ring current ~i has only one component
iu . Because of the symmetry to the axis 0z the lines of the vector potential with
equal values must be the circumference, parallel to the loop with current. The
projection of the vector d~l to the vector ~
A is equal to cos u dl, and the value of
magnetic vector potential at the point M is calculated as [36]
7.2 Electromagnetic Field in the RF Plasma Torches 225

I rffiffiffiffiffi
li cosu dl l0 i Rc
A¼ 0 ¼ FðkÞ; ð7:15Þ
4p r 4p R
l

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where k ¼ 4RRc
z2 þ ðR þ Rc Þ2
is independent of l and r, 0  k  1 The function F ðkÞ is
defined as

ð2
k2 ÞKðkÞ
2EðkÞ
FðkÞ ¼ ; ð7:16Þ
k

Here K ðkÞ и Eðk Þ—the full normal elliptic Legendre integrals of the I and the II
types, the values of them can be found in the tables [37]

Zp=2 Zp=2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

da
KðkÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi; EðkÞ ¼ 1
k 2 sin2 a da: ð7:17Þ
1
k2 sin2 a
0 0

Thus, the calculation of the vector potential A for the loop with current is
reduced to finding the elliptic Legendre integrals K ðkÞ and Eðk Þ The variable
k depends on the coordinates of the point where the vector potential (point M) is
determined and on the geometric characteristics of the loop with current i. Note that
if k ¼ 1 (i.e., z ¼ 0, R ¼ Rc ), A ¼ 1, that cannot be. Therefore, the Eq. (7.15) can
be used to determine the parameters of the field only at a sufficient distance from the
loop with current.
The boundary condition of the vector potential on the wall and on the entrance
of the plasma torch This Eq. (7.15) for the calculation of the vector potential of
the circuit with a current can be used to determine the boundary condition for the
vector potential using the principle of superposition of the field. The value of the
vector potential at the ith point on the boundary can be found as the sum of the
vector potentials for all currents of inductor and currents of plasma using the
following equation:
rffiffiffiffiffi wc
l Ic Rc X l x M X

1 X
N
1
pffiffiffiffi _ 
A_ i ¼ 0 F ðkim Þ
j 0pffiffiffi rk rjk Ajk Sjk F kijk ; ð7:18Þ
2p ri m¼1 2p ri j¼1 k¼1

where M, N—the number of grid points on the axis z and r, respectively; j, k—indices
for axes r and z, respectively, for each of the ring currents in plasma; rk —radius of jkth
loop of the ring current in plasma; ri —the distance to a point where a vector potential
is determined; the complex density _jjk of the jkth ring current with plasma is deter-
mined by the vector potential A_ jk using the equation _jjk ¼ rjk E_ jk ¼
jxrjk A_ jk , and the
226 7 The RF Plasma Torches

complex currents of plasma are found as I_jk ¼ _jjk Sjk , here Sjk —sectional area of jkth
circuit of the ring current.
Note that Eq. (7.18) can be used to determine the value of the magnetic vector
potential on the wall and at the entrance to plasma torch, i.e., where there is no the
current of conductivity.
The boundary condition of the vector potential on the axis of symmetry and at
the exit of the plasma torch The values of the vector potential on the other
boundaries are determined relatively simply. On the axis using the cylindrical
conditions A ¼ 0. At the exit of the plasma torch, where there is no current of
conduction, the value of the vector potential can be found using two discussed ways
or using a “soft” condition of determination the magnetic vector potential at the
border @A
@z ¼ 0. In this case, the discrete analogue for points on exit is

aM AM ¼ aW AW þ aN AN þ aS AS ; ð7:19Þ

where
8 ðrn
rs Þðrn þ rs Þ rn ðze
zw Þ rs ðze
zw Þ
>
< aW ¼ 2ðzM
zW Þ [ 0; aN ¼ rN
rM [ 0; aS ¼ rM
rS [ 0;
aM ¼aW þ aN þ aS þ Sp [ 0;
>
: S ¼ 1 þ ðjr
e xÞl xr ðz
z Þðr
r Þ:
p rM M 0 0 M e w n s

Thus, a complete system of discrete analogues for internal and boundary points
of the Eq. (7.8) is formulated. The solution of such a system can be made using the
standard methods, such as the Gauss–Seidel method or using the method of
alternating directions.
Note that in equation for calculating the value of the magnetic vector potential at
the boundary (7.18) a vector potential at all internal points A_ jk is included.
However, A_ jk is determined from the solution (7.8), i.e., the boundary condition is
closed on the very solution. Thus, the solution of equation for the vector potential
(7.8) should be performed using an iterative method with the approximation and
clarification of the boundary condition (7.18). The iterative procedure for solving a
discrete analogue of (7.12), (7.18), and (7.19) is
1. Set a starting field of the magnetic vector potential inside the channel of plasma
torch. For the fast convergence the zero field can be taken as the initial
condition.
2. The approximate values of the vector potential are determined at the boundary
by the Eq. (7.18).
3. Solve the system of discrete analogues (7.12), (7.19) in order to define a new
field of the vector potential.
4. Check the condition of the solutions convergence and the procedure is repeated,
starting from position 2, if it is not running.
7.2 Electromagnetic Field in the RF Plasma Torches 227

5. Calculate all electromagnetic quantities on the basis of the resulting field of the
vector potential, and to solve other equations of plasma.

7.2.3 Calculation of the Electromagnetic Field and Electric

Characteristics of the RF Plasma Torches

Determination of the inductance and current in the inductor The current in

inductor I1 is usually at the known voltage at the inductor U1, and it is necessary to
determine the inductance of inductor L and its inductive reactance XL = ωL.
The inductance of a single-layer solenoid (empty inductor without plasma) is cal-
culated by the well-known formula of Nagaoka [38]. According to this formula

l0 pD2 W 2
L ¼ KL : ð7:20Þ
4h

The Nagaoka correction factor KL is a function of the ratio of the inductor length
to its diameter, KL ¼ f ðh=DÞ, the value of this function is given in [38].
Correction for the inductance of the inductor for the winding pitch is calculated
by equation

l0 Wh
DL ¼ F : ð7:21Þ
2

The function FðDh=Dc Þ takes into account the ratio of the winding pitch Dh to
the diameter of inductor coil Dc and the shape of the inductor coil (circular or
rectangular). The value of this function is also shown in [38]. Author’s calculations
show that the correction for the step of winding for real inductors of the RF plasma
torch is usually 2–7 % of the inductance, calculated by Eq. (7.20).
The total inductance

l0 pD2 W 2 l Wh
L ¼ KL
F 0 : ð7:22Þ
4h 2

Consequently, the current in inductor can be calculated as

U1 U1
I1 ¼ ¼ : ð7:23Þ
XL xL

To accurate determination of the total inductance of inductor with plasma it is

necessary to know the total magnetic flux. The resulting value of the vector
potential allows easily definition of it. Full magnetic flux is associated with the
228 7 The RF Plasma Torches

inductance of inductor with plasma according to equation U ¼ LI1 . Here I1—the

current in inductor with plasma. We write the following relation between the
magnetic flux Ф and the magnetic vector potential ~
A in inductor with radius Rc
Z Z I
U¼ ~ Bd~s ¼ rot~ s¼ ~
Ad~ Ad~l ¼ 2pRc A:

And the inductance of inductor with plasma

2pRc A
L¼ : ð7:24Þ
I1

Definition of the electromagnetic field Knowing the distribution of the magnetic

vector potential, we can determine the parameters of the electromagnetic field in the
RF plasma torches, including the allocated power rE 2 is used to further solving of
the equation of energy balance, and the electromagnetic forces Fz and Fr , required
for solution of the momentum equations in plasma, as well as all the electrical
parameters that are necessary for solving of the practical problems of coordination
and adjustment of the generator and the RF plasma torches. Here are the necessary
equations for the calculation of the field
1. The electric intensity

_
E_ u ¼
jxA: ð7:25Þ

On the axis due to conditions of the cylindrical axis symmetry E_ u ¼ 0.

2. The components of the magnetic field
 "  #
1 @ _
A 1 1 @ r A_ 1 _
A @ A_
H_ r ¼
; H_ z ¼ ¼ þ : ð7:26Þ
l0 @z l0 r @r l0 r @r

From the condition of the cylindrical symmetry H_ r ¼ 0. To determine the

component of the magnetic field H_ z , on the axis magnetic vector potential can
be presented near the axis of symmetry in the form of a cubic polynome
A_ ðr; zÞ ¼ a_ 0 ðzÞ þ a_ 1 ðzÞr þ a_ 2 ðzÞr 2 þ a_ 3 ðzÞr 3 . A_ jr¼0 ¼ 0 and a_ 0 ðzÞ ¼ 0. We
_ _
have the following relation @@rA ¼ a_ 1 ðzÞ ¼ Ar . Substituting this relation
r¼0 r¼0
into the Eq. (7.26), we obtain a new expression for determination of H_ z on the
axis

_

 2 @ A_ 
Hz r¼0 ¼ : ð7:27Þ
l0 @r r¼0
7.2 Electromagnetic Field in the RF Plasma Torches 229

3. The density of the current of conduction in plasma

_
_ju ¼ rE_ u ¼
jxrA: ð7:28Þ

4. Active power density of discharge in plasma

P ¼ rEu2 : ð7:29Þ

5. The components of the electromagnetic force in plasma

8 h  i h  i
_
< Fr ¼ Re l0 rE_ u Hz ¼ xrIm Aconj A_ @ A_
r þ @r ;
h  i h  i ð7:30Þ
: F ¼ Re
l rE_ H ¼ xrIm Aconj _ @ A_
:
z 0 u r @z

Here Re and Im—real and imaginary parts of the complex value; “*” and conj
—the notations of conjugate complex value.
6. Complete the active current in the plasma

Z l ZR Z l ZR
 
I_2 ¼ _ju drdz ¼
jx _
rAdrdz; I2 ¼ I_2 ; ð7:31Þ
0 0 0 0

where l; R—length and radius of channel of the RF plasma torch.

7. The phase shift of currents in plasma compared with current in inductor

w ¼ arg j_u : ð7:32Þ

8. The total allocated power in plasma

Z Z l ZR
P2 ¼ ReðrE_ u2 ÞdV ¼ 2p ReðrE_ u2 Þrdrdz: ð7:33Þ
V 0 0

9. Active, reactive, and full equivalent resistance of plasma is

Rl RR
2p ImðrE_ u2 Þrdrdz qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P2 0 0
R2 ¼ 2 ; X2 ¼ xL2 ¼ ; Z2 ¼ R22 þ ðxL2 Þ2 :
I2 I22
ð7:34Þ

10. The magnetic flux of the inductor with plasma

230 7 The RF Plasma Torches

Z Z I
U¼ ~ s¼
Bd~ rot~ s¼
Ad~ Ad~l ¼ 2prA:
~ ð7:35Þ

These and other electrical parameters can be used to determine the optimal
operation of generator and RF plasma torch.
Calculation results The following details of the calculation of electromagnetic
field in the RF plasma torches are the result of the joint solution of the equation for
the vector potential (7.8) and other equations from the model of equilibrium plasma:
equation of the energy balance, the momentum equations, the continuity equation
(Sect. 1.2), as well as equations for calculation of the composition and thermody-
namic properties and transport coefficients of the plasma (Sect. 4.1–4.3). The fol-
lowing parameters and characteristics of electromagnetic field are identified
1. The inductance of inductor L, the inductive resistance XL ¼ xL and the current
of inductor I1;
2. The distribution of vector potential A(r, z);
3. The magnetic flux distribution Ф(r,z);
4. The distribution of intensity of the electric and magnetic fields E(r,z), Hz(r,z),
Hr(r,z);
5. The distribution of current density of conduction in plasma j(r,z) and the phase
shift of the plasma current wðr; zÞ;
6. The distribution of power density in plasma P ¼ rE2 ;
7. The distribution of components of electromagnetic force in plasma Fr ðr; zÞ,
Fz ðr; zÞ;
8. The total current in plasma I2 , the total power dissipated in plasma P2 , active R2 ,
reactive X2 and full equivalent Z2 plasma resistance that are necessary for
determination of the agreed regime of operation for the RF plasma torch and
generator.
The general scheme of the RF plasma torch with possible three streams of gases
and the water cooling of the wall are shown in Fig. 7.14.
Initial data for the simulation of the RF plasma torches, the most commonly used
in practice, are presented in Table 7.3
Because of the limited volume of the book we give only the results and analysis
of electromagnetic field of the RF plasma torch (5 MHz); character of the field is
similar at other plasma torches. Some characteristics obtained by their calculation,
are shown in Table 7.4. Initial data for calculation the RF plasma torch with fre-
quency of current in inductor 5 MHz (Table 7.3.) are the following: the voltage at
the 4th turn of inductor U1 ¼ 4:4 kW, that corresponds to the current I1 ¼ 120 A,
two flows of plasma-forming and protective gas (Argon) with consumptions,
respectively Q2 = 20 l/min, Q3 = 40 l/min; Q1 = 0; d1 = 3 mm; D1 = 5 mm;
d2 = 25 mm; D2 = 30 mm; R6 = 23 mm; length of calculated area 120 mm.
The distributions of intensity of electric and magnetic fields in the channel of the
RF plasma torch (5 MHz) are shown in Figs. 7.15, 7.16, 7.17, 7.18, 7.19, 7.20,
7.21, 7.22 and 7.23.
7.2 Electromagnetic Field in the RF Plasma Torches 231

Fig. 7.14 Scheme of the RF plasma torch: 1 central tube for the possible entrance of chemicals, 2
tube for a plasma-forming flow, 3 a wall of the RF plasma torch, 4 water flow for cooling of the
wall, 5 the external casing of RF plasma torch, 6 turns of inductor, 7 water flow for cooling of
inductor, d = 2R5 diameter of the plasma torch channel, d1, D1, d2, D2 internal and external
diameters of the central channel and the plasma-forming flow, respectively, R6 internal radius of a
water-cooled channel of plasma torch, D the internal diameter of inductor, Dc diameter of the coil
of inductor, h the length of inductor, L1 the distance from the entrance of plasma torch to the first
round of inductor, L2 the length of plasma torch, Q1, Q2, Q3 the consumption of plasma-forming
and protective gas, respectively

The intensity of the electromagnetic field quickly decreases as the penetration of

waves into the plasma, and the release of energy in the skin layer (Figs. 7.15, 7.16,
7.17, 7.18, 7.19, 7.20, 7.21, 7.22). In the gap between plasma and inductor, where
there is no current of conduction, there is a slight enhancement of the magnetic field
Hz (Fig. 7.23b) (compared with empty inductor) due to the interaction of magnetic
fields of currents in the inductor and plasma. In this zone the value Hz, as in the case
of empty inductor (Fig. 7.18), should be almost constant for r (Fig. 7.20), because
the induced current of conduction is equal to zero, and the change due to the
displacement current is insignificant. However, starting with the skin layer, where
converting of electromagnetic energy into thermal one occurs actively, and the
magnetic field decreases rapidly (Figs. 7.19, 7.20).
In the absence of plasma, an electromagnetic field is generated only by the
current in inductor, and therefore in Figs. 7.15, 7.18 and 7.21 the field has a
symmetrical shape for the central cross section of inductor.
 232

Table 7.3 Initial data for the RF plasma torches

Parameter Frequency in the inductor (MHz)
0.066 0.12 0.44 1.76 5.00 5.28 13.76 27.12
U1 (kW) 1–3 1.2–2 8–9 6–7 4–4.5 6–6.5 4–6 3–3.5
W 11.0 3.0 4.0 5.0 4.0 4.0 2.5 3.0
D (mm) 130.0 200.0 220.0 96.0 78.6 92.0 58.0 34.0
d (mm) 100 140 165 60 41 54 40 27
Dc (mm) 10 20 22 10 6 10 10 5
h (mm) 260 100 160 100 45 70 34 (54) 25
L1 (mm) 140 150 200 70 3 70 66 30
L2 (mm) 600 500 600 300 78 240 220 100
Q = Q1 + Q2 + Q3 (l/min) 10–20 10–20 6000–16,000 60–300 (8) + 20 + 40 60–80 10–20 7–10
Gas Air Air Air + H2S Air (N2) Ar Ar (N2) Ar Ar
Position Bottom–up Bottom–up Horizontal Bottom–up Up–bottom Horizontal Bottom–up Bottom–up
U1 voltage on inductor; W number of turns; Q total gas consumption
7 The RF Plasma Torches
7.2 Electromagnetic Field in the RF Plasma Torches 233

Table 7.4 The results of calculation of electromagnetic field in the RF plasma torches
Parameters Frequency in the inductor (MHz)
0.44 1.76 5.00 5.28
KL 0.615 0.697 0.563 0.626
L (10−7 H) 28.50 15.40 11.67 11.66
XL = ωL (Ω) 7.878 17.03 36.38 38.68
I1 (A) 1200 350 120 155
Emax (V/cm) 20.3 22.7 12.5 13.2
Hzmax (A/cm) 204.0 128.1 60.9 59.9
Hrmax (A/cm) 60.0 27.5 13.4 6.5
jmax (A/cm2) 160 185 147 137
Pmax (W/cm3) 748 1019 808 769
Pmax/jmax (V/cm) 4.67 5.51 5.50 5.61
Fzmax (10−3 N/cm3) 0.15 1.75 1.90 0.30
Frmax (10−3 N/cm3) −21.7 −13.1 −6.2 −5.4
I2 (A) 5180 1690 400 580
P2 (kW) 400–600 48.4 13.3–15.1 30
R2 (Ω) 0.0150 0.0169 0.0834 0.0875
KL the correction factor of Nagaoka in Eq. (7.20) for determination of inductance, Emax the
maximum intensity of electric field on the wall of the RF plasma torch, Hzmax, Hrmax maximum
components of the intensity of magnetic field, jmax the maximum current density of plasma, Pmax
maximum power density of plasma, Fzmax, Frmax maximum components of the electromagnetic
force in plasma, I2 the total conductivity current of plasma, P2 the total active power allocated in
plasma, R2 the total resistance of plasma

Fig. 7.15 Electric intensity field in the inductor without plasma

As expected, the intensity E and Hz have the peak and the intensity Hr—
minimum in the central zone of inductor. Change of Hzc along the axis r inside
inductor without plasma is slightly (Fig. 7.18) that is also confirmed using the
results of calculations for empty inductor in accordance with the theory of induction
heating. As bigger is the distance from inductor the value E and Hz reduce quickly,
and at the end parts of the first and last turns of inductor Hr reaches maximum
234 7 The RF Plasma Torches

Fig. 7.16 Electric intensity field in the inductor with the plasma

Fig. 7.17 Electric intensity field in the inductor with the plasma E1–E8, in cross-sections z = 4,
10, 16, 26, 36, 48, 80 and 120 mm, respectively

Fig. 7.18 Magnetic intensity component Hz of the inductor without plasma

7.2 Electromagnetic Field in the RF Plasma Torches 235

Fig. 7.19 Magnetic intensity component Hz of the inductor with the plasma

Fig. 7.20 Magnetic intensity component Hz of the inductor with the plasma in the cross-sections
z = 4, 10, 16, 26, 36, 48, 80 and 120 mm, respectively

Fig. 7.21 Magnetic intensity component Hr of the inductor without plasma

236 7 The RF Plasma Torches

Fig. 7.22 Magnetic intensity component Hr of the inductor with the plasma

Fig. 7.23 Changing the electric (a) and magnetic (b) intensity in the central section z4 = 26 mm:
Ec4 ; Hzc4 intensities of the fields without plasma; E4 ; Hz4 also with the plasma; DE4 ¼ Ec4
E4 ,
DHz4 ¼ Hzc4
Hz4 ; ratio of the intensities Ec4 =E4 and Hzc4 =Hz4

(Fig. 7.22). The values Ec4 ; E4 , DE4 ¼ Ec4
E4 , Hzc4 ; Hz4 , DHz4 ¼ Hzc4
Hz4
field and ratio of intensities without plasma and with plasma Ec4 =E4 , Hzc4 =Hz4 in
the central section (z4 = 26 mm) are presented at Fig. 7.23.
Note that the equation for calculation of the magnetic field using the law of the
total current that is often used in the theory of induction heating, gives overestimate
value of Hz. For example, for a point on the wall in the central section of plasma
torch z4 = 26 mm we have a accurate value of Hz = 61 A/cm, while the calculation
using the law of the total current gives Hz ¼ I1hW1 1 ¼ 1204
4:5 ¼ 107 A/cm, that gives
the excess in 1.75 times. This difference is explained by the fact that the expression
for calculating the magnetic field using the law of total current is valid only for an
7.2 Electromagnetic Field in the RF Plasma Torches 237

Fig. 7.24 The distribution of current density of plasma

Fig. 7.25 Changes of current density of plasma j1–j8 in the sections z = 4, 10, 16, 26, 36, 48, 80
and 120 mm, respectively

infinitely empty long solenoid and cannot be used in the case of short inductor with
plasma.
However in calculation of the boundary using Eq. (7.4) with R2 = 16.5 mm (R2
is determined using the temperature at boundary of Argon plasma that is equal to
7000 K) Hz = 62.4 A/cm, this value agrees with the exact calculations of author.
Consequently, this expression can be used to solve one-dimensional electromag-
netic problem.
The distribution of the current density j(r, z) and allocated power density P(r,
z) in plasma are presented in Figs. 7.24, 7.25, 7.26 and 7.27. These data are input
parameters for the solution of the equation of the energy balance. Note that the
resulting distributions j(r, z) and P(r, z) are similar in shape, because there is a direct
link between the allocated power and current density P ¼ rE 2 ¼ j2 =r. The thick-
ness of the skin layer can also be determined using the obtained data, for the case
Δ = 6–8 mm, and 70–90 % of plasma currents and allocates power are concen-
trated in this layer.
238 7 The RF Plasma Torches

Fig. 7.26 The distribution of power density of discharge

Fig. 7.27 Changes of power density in discharge sections P1–P8, z = 4, 10, 16, 26, 36, 48, 80,
and 120 mm, respectively

The phase shifts of plasma currents, compared to the current in inductor, are
shown in Fig. 7.28. Note that between the temperature and the phase shift of a
certain connection: experimental results (Sect. 7.1) and simulation of the RF plasma
torch show that the temperature in the channels of different the RF plasma torches
are not significantly different, and the distribution of phase shifts of currents in
plasma in these channels have approximately the same shape as in Fig. 7.28.
The knowledge of distribution is the electromagnetic force Fr(r, z) and
Fz(r, z) (Figs. 7.29, 7.30) are another important factors in the calculation of elec-
tromagnetic field of the RF plasma torches, they are used to solve the equations of
plasma motion and largely determine the dynamics of the plasma flow movement in
the channel of plasma torch. The component Fr(r, z) provides the movement of gas
in plasma, its maximum is concentrated in the area of the skin layer in the center of
inductor. The results of calculations show that the electromagnetic force in the RF
plasma torch is much smaller than in an electric arc, it leads to the flow of plasma
7.2 Electromagnetic Field in the RF Plasma Torches 239

Fig. 7.28 The phase shifts of currents in the discharge in comparison with the current in inductor

Fig. 7.29 The radial component of electromagnetic force Fr

Fig. 7.30 The axial component of the electromagnetic force Fz

240 7 The RF Plasma Torches

flux in the channel of the RF plasma torch with a relatively low speed (up to
100 m/s).
As it is shown in Table 7.4 the ratio Pmax/jmax for the RF plasma torches does
not change significantly and is 3–6 V/cm that corresponds to the intensity of
electric field in the area of maximum current density and power density.
If to present the electromagnetic system of the RF plasma torch as a model of air
transformer, where plasma is considered as a single turn of second circuit of
electromagnetic circuit, the law of current transformation is well executed
(I2 ¼ kI1 W, where the transformation ratio k  1).
Note that at the same current in inductor the allocated power in the case of
plasma torch with atomic gas is greater than in the case of its operation with a
molecular gas. For example, at operation of plasma torch with 0.44 MHz in Argon
the allocated power P2 = 600 kW, and when it is operating with Air P2 = 400 kW.
This is due to large energy content (enthalpy) of molecular gases because of their
vibrational and rotational excitation.
Thus, the present method of calculation of the two-dimensional electromagnetic
field on the basis of the equation of the vector potential allows the definition of the
necessary distribution of electromagnetic field in the RF plasma torches for the
study of plasma processes. These important parameters such as the distribution of
allocated power and electromagnetic force are used for further solutions of equation
of energy balance and the momentum equation for the plasma flux. The data also
allow determination of the effect of demagnetization of field, the thickness of the
skin layer, the phase shift of plasma currents, the inductance of the inductor—
plasma system, its impedance, allowing a coherent and optimized operation of RF
plasma torch with a generator.

7.3 Simulation of the Equilibrium Plasma in the RF

Plasma Torches

The first work on the calculation of the RF plasma torches are based on
one-dimensional model of equilibrium plasma—a model of gas purge through the
channel of plasma torch with the determination of temperature field of gas and
one-dimensional electromagnetic field. This model was used for a preliminary
estimates of the main parameters of the RF discharge, such as the mean temperature
and the specific power of discharge, the average intensities of the electric and
magnetic fields, the total current and total power in the discharge, and others. Of
course, the one-dimensional formulation of the problem makes it relatively quickly
to get some average values of the characteristics of the discharge, but in-depth
analysis of the physical processes in the discharge is not enough. In the short
inductor the edge effects play a significant role, two-dimensional electromagnetic
field and two-dimensional flow of plasma with reverse vortex of plasma flow are
not uncommon, but the main feature of the RF plasma torch. The main
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 241

characteristics of a discharge are defined on the basics of them—the field of the gas
flow, heat flux, a complete picture of heat transfer in plasma, a single field of the
discharge, the discharge power with the currents of inductor and plasma, and others.
Therefore, the study of the basic physical processes of the RF discharge and cal-
culation of parameters of the RF plasma torch should be carried out using a
two-dimensional model of plasma based on the regime of operation of the plasma
torch and its design features.
The model of thermal equilibrium plasma is formulated above (Sect. 1.2) and is
used for study of dense plasma with a pressure close to atmospheric pressure. It is
based on the assumption of the equality of all groups of temperature for the particles
in plasma (Sect. 1.1), and the justice of Maxwell equations, the Boltzmann equation
and the Saha equations (1.2), (1.4), (1.9). Quasi-neutral plasma flow extends over a
distance much larger than the mean free path of particles, and is presented as a
continuous medium with uniform transfer coefficients and thermodynamic prop-
erties depending on temperature and pressure. In the first approximation, the
radiation around all the frequency range is considered to be optically thin, freely
coming out of the discharge and is represented as a member of the radiation losses
in the balance of power discharge.
Thus, the model of equilibrium plasma in the RF plasma torches include the
following equation: equation of the balance energy (1.25), momentum equations
(1.26), (1.27), continuity equation (1.29). Calculation of two-dimensional electro-
magnetic field in the RF discharge (1.30) is carried out using the equation for magnetic
vector potential (7.8), and the distribution of electromagnetic field and electric
parameters of system inductor-plasma can be determined by Eqs. (7.20–7.35).
Furthermore, the equilibrium plasma model includes various equations for obtaining
of composition, transfer coefficients, and thermodynamic properties of equilibrium
plasma (Sect. 4.1–4.3). In many cases, the transfer coefficients and the thermodynamic
properties of equilibrium plasma (particularly at atmospheric pressure) are known.
These data are available in many reference books and manuals and are often specified
as a discrete function of temperature and gas pressure. In the simulation procedure the
determination of them, using the values of temperature and pressure, is reduced to the
approximation. Methods for solving the problems of approximation are given in many
books on the computational mathematics, and they will not be considered in this book.

7.3.1 Algorithms of the Mathematical Model

As it is shown in Sect. 6.1.2, the solutions of equation of the balance energy, the
momentum equation, the continuity equation and other in the simulation of plasma
processes it is convenient to use not one fixed grid, but several grids displaced from
each other by a half pitch (Fig. 6.5), so that each equation can be solved with a
separate grid. The Shifted Chess Grid should be constructed with maximum use of
the boundary conditions for the solution of equations in the following way: the
242 7 The RF Plasma Torches

boundary points of each displaced grid should be placed directly on the border,
where the absolute value of the necessary variable is known, and they are deter-
mined back down on half a step inside the computational area from those bound-
aries where the gradient of this variable is given.
Such a structure of shifted grids for simulation of the RF plasma torch using the
specific of equations of equilibrium plasma model, boundary conditions, and their
design features plasma torch are shown above (Sect. 6.1.3) in the development of
the joint algorithm of solution the equation of the energy balance, the momentum
equation, the continuity equation, etc., (Figs. 6.5a, 6.6–6.8). As it is shown above
(Sect. 6.1.2), the structure of the shifted chess grids for simulation of the RF plasma
torch and the Arc zone in the channel can be the same because of the similarity of
many boundary conditions on the axis of symmetry, wall, and the exit of the RF
plasma torch.
Discrete analogues of the equation of the balance energy, the momentum
equation, the continuity equation for plasma flux are obtained for internal points
(Sect. 6.1.1, 6.1.2), and their form based on internal feedback of physical processes
are given earlier (Sect. 6.3.2, 6.3.3). Discrete analogue for equation of the magnetic
vector potential and determination of the boundary conditions are given in Sect. 7.2.
To form a complete system of discrete analogues and their matrix for determination
of the coefficients of mathematical model of equilibrium plasma in the RF plasma
torches we have to define the boundary conditions and the corresponding discrete
analogues for the boundary points of the equation of the balance energy, the
momentum equations, and the continuity equation.

Fig. 7.31 Estimated grid and the control volumes for the boundary points at the decision of
equation of the energy balance for the RF plasma torches
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 243

At the solving of various equations using a grid structure and the same boundary
conditions for the discrete analogues for internal and boundary points are the same.
Due to the displacement of the grid points the boundary points of grids in various
cases can be not only on the border of calculated area, and for these points a special
discrete analogues is required. Consider the boundary conditions for the solution of
these equations and obtaining the corresponding discrete analogues in simulation of
the RF plasma torches. In the future, these expressions could be used to simulate the
arc zone in the channel (Sect. 8.2, 8.3).
1. The boundary condition and the corresponding discrete analogues for
equation of the balance energy. As it is shown earlier (Sect. 6.1.1, 6.1.3, the
discrete analogue of two-dimensional equation of balance energy for common
internal point (Fig. 7.31) in a cylindrical coordinate system is

aM TM ¼ aE TE þ aW TW þ aN TN þ aS TS þ bM : ð7:36Þ

Hereafter, all the coefficients of the discrete analogue aM , aE , aW , aN , aS , bM are

determined using the formula (6.61).
The computational grid is shown in Fig. 7.31, for the decision of equation of the
energy balance there is an integral part of offset grid used in the simulation of the
RF plasma torch (Fig. 6.5a). It encompasses not only the inside part of the plasma
torch channel, but the area of the internal wall of the plasma torch. Thus, in the
solution of the equation of energy balance the wall temperature of the plasma torch
the degree of heating that is also an important parameter in the simulation and
design of the RF plasma torches. Moreover, the boundary condition for the internal
side of the water wall of the channel in plasma torch can be approximately deter-
mined using from the measured value of water temperature by thermocouple.
At the entrance of plasma torch (Fig. 7.31), the temperature of the cold gas is
considered to be known T = T0 = 300 K.
On the inner side of the water channel in the wall of plasma torch (Fig. 7.31) can
be set as TW = 350 K.
On the axis of plasma torch we have @T @r ¼ 0 because of the cylindrical sym-
metry. A discrete analogue for the points adjacent to the axis of symmetry is
obtained by integrating the equation of the energy balance (6.39) for the control
volume 1 (Fig. 7.31). The integration procedure is similar to the procedure of
integration of generalized differential equation (Sect. 6.1.1). As @T@r ¼ 0 at r = 0, the
heat flux due to heat conduction through the verge s of the control volume is equal
to zero (ks @T@r ¼ 0). At the same time, at the axis of cylindrical symmetry vr ¼ 0
s

and convective heat flux through the verge s of the control volume is equal to zero
(qs vrs Ts ¼ 0). It follows that a discrete analogue of these points is (7.36) with a
coefficient aS ¼ 0.
At the exit of plasma torch value of T is not specified, but is determined by
decision of the equation of energy balance. However, the exit is often quite far from
244 7 The RF Plasma Torches

the inductor, and we can assume that all processes are established (@T @z  0), and the
convective flow is much greater than the flow due to the thermal conductivity
(AðPee Þ ¼ 0). In this case, the approximate temperature value equal to its value at
the nearest point upstream can be taken. We see that it corresponds to the real
conditions, when the active physical processes in the simulation area (upstream),
largely determine the value of the unknown quantities at the border (downstream),
and not vice versa. It is natural to assume that for developed flux of plasma from the
core of the inductor, the temperature at the exit of plasma torch depends mainly on
the temperature of plasma flow in inductor (upstream), and the approximate
boundary condition at the exit of the plasma torch little effect on the result of the
decision of the equation of energy balance in the upstream, i.e., in the zone of
inductor. A famous computational scheme (method) upstream (Sect. 6.2.1) in the
solution of problems of gas dynamics is based on this real picture. A discrete
analogue for the points adjacent to the exit is obtained by integrating the equation of
balance energy in the control volume 2 (Fig. 7.31) with this assumption. As a result
we have a discrete analogue (7.36) at aE ¼ 0. This form of the boundary conditions
at the exit using a computational scheme “upstream” is highly effective, it gives the
discrete analogue for points on this border and it will be used in the future.
The discrete analogue for a point located on the axis of symmetry at the exit of
the plasma torch is obtained similarly (for the control volume 3 at Fig. 7.31) using
two boundary conditions for the points on the axis of symmetry and for the points at
the exit of the plasma torch, so the discrete analogue (7.36) for this point have
coefficients aS ¼ aE ¼ 0.
2. The boundary conditions and the corresponding discrete analogues of the
momentum equations. The discrete analogue of the two-dimensional
momentum equation with for vz (6.25) for common interior points in a cylin-
drical coordinate system, obtained above (Sect. 6.1.2) has the form (6.28)

aM vzM ¼ aE vzE þ aW vzW þ aN vzN þ aS vzS þ bzM þ ðpw
pe ÞDSr ; ð7:37Þ

where the coefficients aM ; aE ; aW ; aN ; aS ; bzM ; DSr are determined using the

Eq. (6.61).
At the entrance of the plasma torch (z = 0), the value vz , m/s is determined using

3
a given mass consumption of gas Q, l/min as vz0 ¼ Q10 60pR2 , where R—the radius of
the input channel of the plasma torch, m.
The discrete analogue for the points, adjacent to the axis of symmetry (the
control volume
 1 Fig. 7.32). Boundary conditions on the axis of the plasma torch
@vz
vr ¼ 0 and @r ¼ 0 mean that the convective and viscous flows through the verge
s of the control volume 1 equal to zero, i.e., for these points we have the discrete
analogue (7.37) at aS ¼ 0.
The discrete analogue for the points, located at the exit of the plasma torch.
Similarly to the boundary conditions for T, the boundary condition vz at the exit of
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 245

Fig. 7.32 The grid and the control volumes for the boundary points in the decision of the
momentum equation for vz in the RF plasma torches

the plasma torch is obtained with condition of a steady (@v @z ¼ 0) and the developed
z

flows of the plasma stream. A discrete analogue for these points is obtained by
integrating the momentum Eq. (6.25) for vz at the control volume 3 (Fig. 7.32). As
@vz @vz
@z ¼ 0 at the exit, and viscous forces flz l @z , there is a loss of momentum (the
viscous flow) on the axis z is equal to zero. The established character of the gas flow
at the exit allows to receive the speed vze ¼ vzM . This assumption about the
boundary condition allows in obtaining the discrete analogue for these points that
does not include information about the plasma velocity at the exit of the plasma
torch. It has the same form as (7.37) at aE ¼ 0.
The discrete analogue for a point at the exit and on the axis of symmetry (control
volume 2, Fig. 7.32). For two boundary conditions on the symmetry axis and at the
exit it can be obtained that, for a point located on the axis of symmetry at the exit of
plasma torch, the discrete analogue (7.37) is using aE ¼ as ¼ 0.
The discrete analogue for the points adjacent to the wall of plasma torch, is
obtained integrating the momentum equation (6.25) for control volume 5 (Fig. 7.32)
using the boundary conditions for the points on the wall of plasma torch:
vz ¼ vr ¼ 0. Gas moves without “slipping”, clinging to the wall. If we take the
structure of the grid for determination of vz such that the point on the right boundary
is shifted by half a step inside the walls, you can simply set the speed at this
boundary (Fig. 7.32) vz ¼ 0, and the discrete analogue for points of the control
volume 5 will have the form (7.37).
246 7 The RF Plasma Torches

The discrete analogue to a point adjacent to the wall and to the exit, it is also
obtained by integration of the momentum equation (6.25) for the control volume 4
(Fig. 7.32) using the same boundary conditions at the exit and on the wall. As in the
previous case, it is very easy to get the discrete analogue the grid has the boundary
points not on the wall, but inside it (Fig. 7.32). At the same time vz ¼ 0 and the
discrete analogue has the form (7.37) with aE ¼ 0.
Similarly the discrete analogue for the boundary nodal points in the decision of
the momentum equation for the component of velocity vr can be obtained. The
discrete analogue of two-dimensional momentum equation for vr (6.26) at the
common internal points in a cylindrical coordinate system, obtained in Sect. 6.1.2,
has the form (6.28)

aM vrM ¼ aE vrE þ aW vrW þ aN vrN þ aS vrS þ brM þ ðps
pn ÞDSz : ð7:38Þ

The coefficients aM ; aE ; aW ; aN ; aS ; aM ; brM are determined using the formula

(6.61).
On the wall and on the axis of symmetry the velocity vr is equal to zero:
vr jr¼0 ¼ vr jr¼R ¼ 0 using the conditions of the cylindrical axial symmetry and
sticking of the gas to the wall, respectively.
The discrete analogue for the points adjacent to the entrance is obtained by
integrating of the momentum equation (6.26) for the control volume 1 (Fig. 7.33).

Fig. 7.33 The grid and the control volume of the boundary points in the solution of the
momentum equation for vr in the RF plasma torches
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 247

In integrating using the boundary conditions at the entrance: vr = 0 (for axial gas
supply) and @v@z ¼ 0, the convective and viscous flows through the verge w of the
r

control volume 1 are equal to zero, i.e., the discrete analogue for these points has
the form (7.38) with aW ¼ 0.
The discrete analogue for the points at the exit is obtained by integrating the
momentum equation (6.26) for half of the control volume 2 (Fig. 7.33). Here the
boundary condition at the exit @v
@z ¼ 0 as a result of the steady flow of plasma, in the
r

end we have the discrete analogue in the form (7.38) with aE ¼ 0.

3. The boundary conditions and the corresponding discrete analogues of the
continuity equation for p. The pressure at the exit of the plasma torch is
considered to be known and for most of the RF plasma torches with an open
channel is equal to atmospheric pressure. Therefore, the boundary points of the
grid for determination of pressure are directly on the exit of the plasma torch
(Fig. 7.34). At the same time, the values of the inlet pressure on the wall and on
the axis of symmetry are usually unknown, but the values of the velocity
components on these boundaries are known. Therefore, the grid for determi-
nation of pressure compared with the grid for definition of vz and vr is shifted
inward from the boundaries on a half-step, that makes it possible to determine
accurately the velocity components vz and vr on the verges of the control volume
(Fig. 7.34).

Fig. 7.34 The grid and the control volume of the boundary points in the solution of the pressure
p in the RF plasma torches
248 7 The RF Plasma Torches

Due to this structure of the shifted grid the special discrete analogue for pressure
using the continuity equation and through the components of the velocity and vr
using the above methodology (Sect. 6.1.2) is obtained. General discrete analogue
for the continuity equation (6.38) is

aM pM ¼ aE pE þ aW pW þ aN pN þ aS pS þ bM ; ð7:39Þ

where bM ¼ ðqw^vzw
qe^vze Þ ðrn
rs Þ2ðrn þ rs Þ þ ðrs qs^vrs
rn qn^vrn Þðze
zw Þ; The coef-
ficients aM ; aE ; aW ; aN ; aS are determined by (6.36–6.38).
The main advantage of the shifted grid is obtaining of the discrete analogue for
points, adjacent to the exit, the axis of symmetry and to the wall, i.e., at the borders
where the pressure is unknown. In Fig. 7.34 five types of control volumes for the
boundary points for solving of the continuity equation and obtaining of the discrete
analogue for pressure can be identified. As it is shown earlier (Sect. 6.1.2), the
discrete analogue for these points is obtained by integrating of the continuity
equation for the control volume. Using the boundary conditions the continuity
equation for vz and vr at the entrance (z ¼ 0; vz ¼ vz0 ), on the wall (r ¼ R; vr ¼ vrn
for the plasma torch with a porous wall the velocity at the wall is not equal zero)
and at the axis of symmetry (r ¼ 0; vr ¼ 0) we have to discrete analogue for the
boundary points of continuity equation in a general form (7.39) with the following
coefficients:
For the points, located at the entrance of the plasma torch (control volume 1 in
Fig. 7.34.

ðrn
rs Þðrn þ rs Þ
aW ¼ 0; bM ¼ ðqw vz0
qe^vze Þ þ ðrs qs^vrs
rn qn^vrn Þze ; ð7:40Þ
2

For points adjacent to the axis of symmetry (control volume 4)

ðrn
rs Þðrn þ rs Þ
aS ¼ 0; bM ¼ ðqw^vzw
qe^vze Þ
rn qn^vrn ðze
zw Þ; ð7:41Þ
2

For points adjacent to the wall of the plasma torch (control volume 5)

aN ¼ 0;
ð7:42Þ
bM ¼ ðqw^vzw
qe^vze Þ ðrn
rs Þ2ðrn þ rs Þ þ ðrs qs^vrs
rn qn vrn Þðze
zw Þ;

For a point on the axis of symmetry near the entrance of the plasma torch
(control volume 2)

rn2
aW ¼ aS ¼ 0; bM ¼ ðqw vz0
qe^vze Þ
rn qn^vrn ze ; ð7:43Þ
2

For the point at the entrance of the wall of plasma torch (control volume 3)
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 249

8
< aW ¼ aN ¼ 0;
ð7:44Þ
: bM ¼ ðq vz0
q ^vze Þ ðrn
rs Þðrn þ rs Þ þ ðrs q ^vrs
rn q vrn Þze :
w e s n
2

Thus, using the method of control volume and the structure of shifted grids the
complete systems of discrete analogue of mathematical model of equilibrium
plasma in the RF plasma torches were prepared. The obtained discrete analogues
represent a closed system of discrete equations, that include internal and external
communication of various physical processes and obtained from the boundary
conditions, the regime of operation and the features of construction of the RF
plasma torches, so they can the most accurately simulate the physical processes and
be a reliable tool for the study of the characteristics of plasma.

7.3.2 Analysis of the Simulation Results

Construction of the grids for calculations As it is shown above (Fig. 6.5a), for
simulation of the RF plasma torches four shifted grids are used: [z1, r1]—for
determination of the vector potential A, [z1, r2]—for defining of the velocity vz and
temperature T, [z2, r1]—for obtaining of the speed vr, and [z2, r2]—for the deter-
mination of the pressure p. Depending on the control volume the transfer coefficients
and thermodynamic properties can be found by approximating the value of tem-
perature on the different grids. The coordinates of the grid [z1, r1] are given, and the
coordinates of the grid [z2, r2] can be determined in the middle of the grid coor-
dinates [z1, r1], respectively. Grids for simulation are built taking into account the
peculiarities of the temperature field and the flow of the RF plasma. Constructing the
grids it is also necessary to take into account the geometric characteristics of the
plasma torch, zone of active release of energy and current flow, the zone of active
interaction between the field and plasma, and others. All active processes in the RF
discharge occur in the area of inductor and in the skin layer, so the grid steps in these
areas should not be large. Grids are built in the zones, and where sharp temperature
and speed gradients are expected and a smaller step of the grid should be chosen, and
vice versa. For example, in the area between the skin layer and the wall of the plasma
torch it should take the steps Δr = 0.3–0.5 mm compared with the average Δr = 1–
2.5 mm in the remaining zones. In some cases, for further input of chemicals and
processed particles into the discharge the entrance of plasma torch can be located
close to the inductor. The sharp temperature and velocity gradients in this zone are
expected. Therefore it is necessary to take a step Δz = 0.1–0.5 mm in this area,
compared with Δz = 2–4 mm in the zone of inductor and Δz = 5–8 mm in the zone
of the jet of plasma torch. Thus, for example, at the geometric characteristics of the
channel of the RF plasma torch (5 MHz) (Table 7.3; Fig. 7.14) by the number of
grid points at the z-axis is obtained and it should be equal to 40–50 and at the axis
r = 30–40, and the number of unknown variables of the equation will be 1200–2000.
250 7 The RF Plasma Torches

The selection of the step of the grid and the determination of the coefficients of
internal feedback for various equations are carried out taking into account the region
of stability (Figs. 6.17, 6.18, 6.19), that are obtained as a result of numerous
experiments on simulation of plasma processes.
Note that using of a uniform grid (Δz = const, Δr = const) the process of solving
will be more stable, but in this case the number of nodes and respectively the
number of discrete equations will be considerably large.
Analysis of the simulation results The RF plasma torch (5 MHz) and similar
plasma torches (Table 7.3) with three streams are used in practice for spectral
analysis and plasma decomposition of chemical compounds. They work in pure
argon in a mixture of Ar and N2 and in other molecular gases. Some of the results
and analysis of researches of electromagnetic fields in the RF plasma torch (5 MHz)

Fig. 7.35 The temperature and flow fields of Argon in the RF plasma torch (5 MHz): ѱ—current
function; (Q1 = 0, Q2 = 20 l/min, Q3 = 40 l/min)
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 251

Fig. 7.36 The temperature of Argon in the RF plasma torch (5 MHz)

Fig. 7.37 The temperatures of Argon T1–T8 in the sections z = 4, 10, 16, 26, 36, 48, 80, and
120 mm, respectively

is shown above (Sect. 7.2.3), we present here the rest of the simulation results of an
equilibrium plasma [39, 40]. For comparison with data of the field calculation
(Sect. 7.2.3) and in simulation of nonequilibrium plasma (Sect. 7.5.2) the results are
shown in the same cross-sections: z1 = 4 mm, z2 = 10 mm, z3 = 16 mm,
z4 = 26 mm, z5 = 36 mm, z6 = 48 mm, z7 = 80 mm, z8 = 120 mm.
Figures 7.35, 7.36 and 7.37 shows the fields of temperature and velocity of
argon flow of plasma in the RF plasma torch (5 MHz). Comparing the data on the
temperature (Fig. 7.35), and released power (Figs. 7.26, 7.27), we see that the
maximum plasma temperature is not on the axis of the plasma torch, and a zone
close to the region of maximum power allocation, and is Tmax = 9500–10,500 K.
The temperature of argon jet at the exit of plasma torch reaches 8000–9000 K.
These temperatures of argon plasma in the RF plasma torches are confirmed by
252 7 The RF Plasma Torches

experimental data in most works [1, 18]. Figure 7.35 also shows a distinct area of
vortex flow of gas into the upper part of the inductor. Small plasmoid is constantly
held inside the vortex, and the plasma temperature reaches a maximum in the zone
of Tmax = 10,000–10,500 K.
Starting from the second round of inductor, gas actively penetrates into plasma
(Fig. 7.35). According to the data of the fields of temperature and flow, we see that
the cold gas does not penetrate into the discharge but the external layer of hot gas at
a sufficiently high temperature T = 5000–8000 K. By the results of the calculation
of the field the component of electromagnetic force Fr and the released power,
P = σE2 (Figs. 7.26, 7.29), reach a maximum at the center of inductor and decrease
rapidly with the increasing distance from it, so inside the inductor discharge as it is
compressed and has an average temperature T = 9000–9500 K.
At the exit of plasma torch, where there is practically no energy release and
electromagnetic force, there is expansion of the flow with decreasing temperature
(Fig. 7.35) to T = 8000–9000 K due to diffusion and thermal conductivity.

Fig. 7.38 The field of the velocity components vz of Argon in the RF plasma torch (5 MHz)

Fig. 7.39 The field of the velocity components vr of Argon in the RF plasma torch (5 MHz)
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 253

Fig. 7.40 The experimental and calculated data of Argon in the RF plasma torch (5 MHz): a the
distribution of the measured particle velocity silicon vs50 (ds = 50 mm), vs100 (ds = 100 mm) [41]
and the estimated plasma velocity vz, at z = 120 mm; b distribution measured Texp [39, 42] and the
estimated T plasma temperature in the section z = 54 mm

Expansion flow in some cases leads to overheating of the walls of the plasma torch,
so in practice the introduction of shielding gas is carried out by using the plasma
torch exit gas stream oncoming.
Distribution components speed vz and vr of the plasma flux are shown in
Figs. 7.38 and 7.39. The speed of plasma reaches a maximum at the exit of the
plasma torch and is vzmax = 32 m/s and a maximum speed of reverse plasma flow
vzmin = −10 m/s and is equal to the velocity component vr in this zone.
The results of measuring the velocity of the silicon particles with diameter of 50–
100 µm in the jet of plasma torch (z = 120 mm) and the temperature of plasma in
cross section z = 54 mm, and a comparison with the simulation result in these
sections are shown at Fig. 7.40. The measurement results are taken from [39, 41,
42], experiments were carried out at the same conditions for simulation of the RF
plasma torch (5 MHz). To measure the velocity of the particles using the laser
method of registration, and to measure the temperature of plasma the method of the
absolute intensity with the integral Abel transformation were used.
Processing data of other regimes at the constant voltage on inductor shows that
varying amount of plasma-forming and protective gas is 20–100 l/min does not
significantly change the temperature field and the plasma flow. This confirms our
earlier conclusion (Sect. 7.1) that movement of the plasma flow depends mainly on
the electromagnetic force and power density in the discharge.
For analysis of the stream of plasma flow the distribution of specific gas con-
sumption ρvz and ρvr. are presented in Figs. 7.41 and 7.42. We see that gas
not « blows » into the discharge, excessive amounts of cold gas flowing out of the
254 7 The RF Plasma Torches

Fig. 7.41 Distribution of specific consumption ρvz of Argon in the RF plasma torch (5 MHz)

Fig. 7.42 Distribution of the specific gas consumption ρvr of Argon in the RF plasma torch
(5 MHz)

plasma core (in the zone near the wall), forming a natural gas thermal protection
layer.
As expected, the electromagnetic force Fr (Fig. 7.29) compresses the plasma
flow, thus creating the maximum pressure in the center of inductor (Fig. 7.43). At
the entrance a small pressure is created by the counter flow of plasma-forming gas
flow and reverse vortex of plasma. The pressure at the entrance zone of the pro-
tective gas is enough high to provide a large gas consumption. Since the second
spiral inductor, active penetration of gas into plasma by an electromagnetic force
leads to the formation parietal under pressure zone (Figs. 7.35, 7.43). This explains
the number of cases detected in the parietal vortex flow (aft vortex), exciting
additional mass of gas from the exit of the plasma torch in the event of inadequate
supply of plasma gas
The resulting pressure field in the channel of plasma torch allows determination
of the static pressure gradient in the discharge (Fig. 7.44). As it was previously
noted (Sect. 7.1) the movement of gas in the RF discharge takes place under the
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 255

Fig. 7.43 Field of the static pressure Δp = p − pout inside the RF plasma torch (5 MHz)

Fig. 7.44 Gradient of the static pressure
@p @p8

@r =
@r in the sections z = 4, 10, 16, 26, 36, 48, 80,
1

and 120 mm, respectively, of Argon in the RF plasma torch (5 MHz)

influence of the sum of two forces
@p@r þ Fr ; our calculations show that it is only
3–8 % of the forces Fr (Fig. 7.29).
We have seen that for the calculation of plasma the gradient of static pressure
cannot only be taken into account or to equate it to the electromagnetic force Fr.
Created field of static pressure in plasma acts as a negative feedback loop that
prevents to the excessive penetration of cold gas into plasma, and provides the
ability to hold the high temperature of discharge, creating the favorable conditions
for the flow of current, the release of energy and maintenance of discharge.
Thus, the plasma parameters depend on close cooperation between the various
physical processes inside the discharge. The discharge in the plasma torches as it is
self-regulating and “chooses” the best variant of its sustainable implementation with
minimal energy consumption. Some researchers, such as M. Steinbeck, based on the
analysis of the thermal heating in the Arc call it as a principle of minimum energy
256 7 The RF Plasma Torches

Table 7.5 Some Parameter Frequency in inductor (MHz)

thermodynamic parameters of
0.44 1.76 5.00 5.28
the RF plasma torches
Tmax (K) 10,800 10,900 10,700 10,600
vzmax (m/s) 128.4 72.8 30.9 27.6
vzmin (m/s) −177.5 −87.7 −7.5 −11.8
vrmax (m/s) 40.6 43.2 13.3 1.35
vrmin (m/s) −40.8 −9.98 −7.7 −4.5
Δpmax (Pa) 156.9 67.0 21.9 15.8
Δpmin (Pa) −212.7 −27.7 −9.1 −7.2
Tmax maximum temperature of plasma in the discharge, vzmax,
vzmin, vrmax, vrmin minimum and maximum values of the velocity
components, respectively, Δpmax, Δpmin difference of maximum
and minimum pressures in the channel compared with the
pressure at the exit of the plasma torch

consumption, it is the result of cooperation of closed physical processes in the

discharge, that are implemented at the expense of fundamental conservation laws.
Some thermodynamic parameters obtained as a result of simulation of other types
of the RF plasma torches (p = 1 atm), they are shown in Table 7.5.
For comparison with the data in Fig. 7.35 the simulation results of the RF
plasma torch (5.28 MHz, p = 1 atm) are shown in Fig. 7.45. We see that the
temperature field and the flow of plasma, the maximum temperature in the area of
inductor, the reverse vortex flow remain practically the same as in the case of the
RF plasma torch (5 MHz). Such data can be considered typical for Argon in the RF
plasma torches of the atmosphere pressure.
Checking the accuracy of the results and the implementation of conservation
laws To the check of the accuracy of the solution of each equation we can be
determine their residual function R—principle of conservation in differential form.
At sufficiently exact solution the value of this function R should be close to zero.
The continuity equation is often chosen as the test, as it requires only a simple
calculation of residual function by the equation

@ ðqvz Þ 1 @
R ¼ divðq~
vÞ ¼ þ ðrqvr Þ: ð7:45Þ
@z r @r
In order to establish the accuracy of the simulation of different processes the
compliance of the conservation laws in the equations in integral form for each
variable can also be checked. For example, knowing a predetermined amount of gas
Q it can be verified the law of mass conservation in integral form

ZR
Qq
w¼ ¼ qvz rdr: ð7:46Þ
2p
0
7.3 Simulation of the Equilibrium Plasma in the RF Plasma Torches 257

Fig. 7.45 The temperature and flow fields of Argon in the RF plasma torch (5.28 MHz)
(Q = 60 l/min, U1 = 6 kV, I1 = 155 A, P2 = 30 kW)

The right side of (7.46) is the function of current on the wall of the plasma torch
(R = R5) is determined from the resulting velocity and temperature fields. Since the
field of temperature and speed, in fact, are the result of the combined action of all
closed physical processes of mathematical model, and the observance of equality
(7.46) indicates the proper and correct simulation results. Figure 7.46 shows the
distribution of the current function in three dimensions. The obtained value of the
current function at the wall of plasma torch is equal ψAr(R5) = 0.219 g/s and cor-
responds to the total amount of gas at the entrance to the plasma torch
QAr = 60 l/min, that indicates the accurate result of our solutions.
258 7 The RF Plasma Torches

Fig. 7.46 Distribution of the current function ψ (flow field) in the channel of the RF plasma torch
(5 MHz) (QAr = 0 + 20 40 l/min)

7.4 Disturbance of the Thermal and Ionization

Equilibrium in Plasma, Caused by the Gases
Movement

The assumption of the thermal and ionization equilibrium in plasma allows the
creation of a suitable model for its investigation, although a number of significant
events occurring in plasma cannot be explained. Disturbance of detailed equilib-
rium in plasma can be caused by different mechanisms of transfer of mass, energy,
momentum, and particles (convection, heat conduction, radiation, bipolar diffusion,
and others) from the depth of discharge to the periphery, and for fast plasma flow
convection may be the dominant mechanism. Compared to thermal imbalance that
is detected and confirmed by both experimentally and theoretically, the number of
works devoted to the analysis of nonequilibrium plasma ionization is small. This is
mainly due to the extremely complex processes of interaction of plasma particles
(energy exchange between particles, momentum exchange, and transformation of
one type of particles into particles of another sort, radiative processes, the presence
of external fields, etc.). In such a situation it is important to identify the main factor
that disorders thermal and ionization equilibrium of plasma and to determine its
impact and the degree of disturbances of these equilibriums. Such analysis can be
done using a relatively simple model of plasma.
One of the first studies about the thermal equilibrium disturbances caused by the
movement of gas through the two-temperature model is [43]. The task of heating it
in a half-heated plasma space is considered for analyze of the thermal equilibrium
of gas: cold gas with an initial temperature Ta0 penetrates into the half-plasma space
with electron temperature Te at the given gas consumption qvz and pressure p. It is
believed that the plasma is in a state of partial local thermal equilibrium (PLTE),
i.e., Maxwell (1.2), Boltzmann (1.4) and Saha (1.9) equations are valid. The
one-dimensional equation of the energy balance was decided for atom–ion com-
ponent in the form of
7.4 Disturbance of the Thermal and Ionization Equilibrium in Plasma … 259

dTa
qvz cp ¼ bðTe
Ta Þ; ð7:47Þ
dz
qffiffiffiffiffiffiffi
Here b ¼ 32 k 2m
ma ne
e
pme ðna Qea þ ni Qei Þ—coefficient of energy exchange between
8kTe

electrons and atom–ion components.

The model does not take into account the energy flux due to thermal conduction,
especially in the initial section where the gas is actively heated and high thermal
gradient ddTza and hence a considerable flow of heat by conduction kai ddTza should be
expected. In the case of small consumptions qvz , corresponding to the real condi-
tions of the RF plasma torches, it can lead to inaccurate results. This statement of
the problem makes it possible to eliminate the need of the decision of the
momentum equation, the continuity equation and Maxwell equations and to restrict
with the decision of the simplified equation of energy balance for atom–ion com-
ponent in order to determine the degree of the gas heating.
On the basis of solving the problem of gas heating in half-space of plasma the
first results of the analysis of disturbances of thermal ionization equilibrium of the
moving gas were obtained in [44]. It turned out that the disturbance of the ion-
ization equilibrium of plasma can significantly influence to the speed of heating of
atomic–ion components, i.e., there is a definite connection between the thermal and
ionization processes. Our task is to determine this connection.
In the nonequilibrium model of heating gas in a half-heated space of plasma
includes the following equation [44]:
Equation of the energy balance for atom–ion component

 rffiffiffiffiffiffiffiffiffi
dðqvz Ta Þ d 2ma kai dTa 3 2me 8kTe
¼ þ k ne ðna Qea þ ni Qei ÞðTe
Ta Þ;
dz dz 5k dz 2 ma pme
ð7:48Þ

Equation of the ionization equilibrium in one-dimensional form


2 
d d dne n n2
ðqvz ne Þ ¼ qDamb þ qkr ne na e
e
dz dz dz na na

 ð7:49Þ
dne dq
þ ne vz
Damb :
dz dz

Here ne ; na —the concentration of electrons and atoms in the two-temperature
ionization equilibrium, calculated on the basis of the law of mass action by the Saha
(4.1), or Potapov equation (4.4), and Dalton law of partial pressures (4.6). Plasma is
considered to be quasi-neutral (ni  ne ), singly and doubly ionized, atomic, and the
charged particles are the result of two opposite processes—impact (stepped) ion-
ization and triple (impact-radiative) recombination.
260 7 The RF Plasma Torches

Fig. 7.47 The grid and control volumes for solving of one-dimensional equation of the balance
energy and equation of ionization equilibrium in the problem of gas heating

These data of the effective sections Qea ; Qei and the ways of defining the transfer
coefficients and thermodynamic properties of two-temperature plasma are given in
Sect. 2.2, 4.1–4.4.
Total discrete analogue for solving the equation of the energy balance for atomic
and ionic components and equation of ionization equilibrium (Fig. 7.47) can be
obtained by solving the one-dimensional GDE (Sect. 6.1.1), taking into account the
linearization of the source function (6.55):

aM UM ¼ aE UE þ aW UW þ bM ; ð7:50Þ

where

aE ¼ De AðPe Þ
0; aW ¼ Dw AðPw Þ þ Fw
0
aM ¼ aE þ aW
Sp ðze
zw Þ
0; bM ¼ Sc ðze
zw Þ;

for equation of the energy balance

2ma kai
U  Ta ; C ¼ ;
5kffiffiffiffiffiffiffiffiffi
r
3 2me 8kTe
S¼ k ne ðna Qea þ ni Qei ÞðTe
Ta Þ ¼ Sc þ Sp Ta ;
2 ma pme

for equation of ionization equilibrium

7.4 Disturbance of the Thermal and Ionization Equilibrium in Plasma … 261

U  ne ; C ¼ qDamb ;


2 2
 
n n dvz dne dq
S ¼ q kr ne na e
e
ne
Damb ¼ Sc þ Sp ne :
na na dz dz dz

Sustaining the stability of computational procedures In a joint decision of the

two Eqs. (7.48) and (7.49) for the discrete analogue (7.50) the circuit with internal
feedback equation of the energy balance for atom–ion components can be installed
by the method described in Sects. 6.3.2, 6.3.3. In particular, the linearization of the
source function for equation of the energy balance can be carried out by the algo-
rithm (6.56), i.e., according to its absolute value. However, at the linearization of the
source function the equation of ionization equilibrium it is necessary to consider the
following factors: the temperature difference of electron and atomic–ion components
varies as the heating of atoms and ions in the range of DT ¼ Te
Ta = 5000–
20,000 K, so the electron density in this area also varies in a wide range of several
orders of magnitude. In this case, rapid convergence is important to determine the
initial concentration of electrons that correspond to the range of possible changes in
the electron and ion atomic components temperature. The most convenient way to
get the initial electron density from the initial field of temperature of electron and
atom–ion components is in ionization equilibrium (4.1) or (4.4).
This method is convenient to define the initial field of the electron density,
possesses, however, with a serious disadvantage. In the first iteration, we have
near-zero source function in equation of ionization equilibrium ne ¼ ne ;
na ¼ na ; S  0, so that the internal feedback coefficient for the first iteration in
equation of ionization equilibrium is very small (Sp  0) and at the beginning the
computational process will be carried out virtually using the open circuit without
internal feedback. Practice has shown that this is enough to cause a divergence of
solutions of equation of ionization equilibrium. So a special way of linearization
should be found for the source function in equation of ionization equilibrium and the
factor of internal feedback should be set.
The effectiveness of internal feedback loop in the stability of solutions of dif-
ferential equations is not violated if the source function for equation of ionization
equilibrium (7.49) is linearized as

Sp ¼ bqkr n2e ; Sc ¼ S
Sp ne : ð7:51Þ

Here, the coefficient of internal feedback b

1, and is subject for possible regu-
lation. An important result of this method of installation of internal feedback loop is
nonzero coefficient Sp even for the first iteration, when the source function for
equation of ionization equilibrium S  0, i.e., internal feedback loop is always
established and continues to operate effectively. It should be noted that with the
(7.51) it was possible to ensure the stability of the computational procedure at the
joint decision of equation of the energy balance (7.48) and equation of ionization
equilibrium (7.49).
262 7 The RF Plasma Torches

Boundary condition and the discrete analogue for the points on the boundary
are obtained by integrating the corresponding equation for half of the control
volume (Fig. 7.47). In the integration it should take into account the following
boundary conditions: z ¼ z0 , Ta ¼ Ta0 , but the concentration of electrons in z ¼ z0
 
is unknown; for steady-state process in the end (z ¼ zn
1 ), ddTza  ¼ ddnze  ¼ 0.
zn
1 zn
1
In the case when the values of variables on the boundary are unknown, it can be
assumed that the value of Φ at the surface is completely determined by an internal
process. It is justified for the processes that are actively occurring not at the border
but inside the study area. This boundary condition is completely determined by the
internal process (like a computer scheme “upstream” in Sect. 6.2.1). In view of this
boundary condition the discrete analogue for equation of ionization equilibrium at
the starting point (z ¼ z0 ¼ 0, Fig. 7.47) has the form
aM UM ¼ aE UE þ bM ; ð7:52Þ

where
aE ¼ De AðPe Þ
0; aM ¼ aE
Sp ze
0; bM ¼ Sc ze

For the last point (z ¼ zn
1 ¼ 0, Fig. 7.47) the discrete analogue of equation of
the balance energy and equation of ionization equilibrium is

aM UM ¼ aW UW þ bM ; ð7:53Þ

Here

aW ¼ Dw AðPw Þ þ Fw
0; aM ¼ aW
Sp ðzM
zw Þ
0; bM ¼ Sc ðzM
zw Þ:

The decision procedure of equation of balance energy and equation of

ionization equilibrium is as follows.
1. Take the initial temperature field of atomic and ionic components Ta ¼ Ta0 and
using (4.1) or (4.4) determine the initial field of the electron density ne0 .
2. Calculate the coefficients of discrete analoueg of equation of ionization equi-
librium (7.49).
3. Solve the system of the discrete analogue for equation of ionization equilibrium
(7.50) and a new field of electron density ne is obtained.
4. Calculate the coefficients of the discrete analogue (7.48).
5. Solve the system of the discrete analogue for equation of the balance energy
(7.50) and a new temperature field Ta is obtained.
6. Check on convergence and repeat the procedure from position 2, until the full
convergence.

Analysis of the obtained results Analysis of disturbances of thermal and ion-

ization equilibrium caused by the movement of gases is carried out at the example
of Argon plasma in two stages in order to further possible comparing of the results:
7.4 Disturbance of the Thermal and Ionization Equilibrium in Plasma … 263

(1) Analysis of disturbance of thermal equilibrium with maintaining of the ion-

ization equilibrium. In this case, the equation of balance energy for only atom–
ion component is solved (7.48), the concentration of particles is determined by
(4.4).
(2) Analysis of disturbance of thermal equilibrium using disturbances of ioniza-
tion equilibrium. The energy balance equation of the atom–ion component
(7.48) can be solved with equation of ionization equilibrium (7.49), and the
equilibrium values ne ; na are defined using (4.4).
Data on the specific consumption of Argon were chosen for the conditions in RF
and Arc plasma torches close to real: qvz ¼ 0.05; 0.1; 0.2; 0.3; 0.5 g/(cm2 c). The
values of cold gas temperature are set Ta0 ¼ 300
9000 K, the electron tempera-
ture Te ¼ 9000
20; 000 K and the pressure p = 105 Pa.
The results of calculations for the ionization equilibrium are shown in Fig. 7.48,
and its disturbance is in Fig. 7.49. The comparison of these figures shows that the

Fig. 7.48 The temperature

for the atom–ion component
(a), electron density (b) and
the gas velocity (c) at the
ionization equilibrium and
Te = 10,000 K: doted line
qmz ¼ 0:05 g=ðcm2 sÞ, dash
line with double doted line
qmz ¼ 0; 1 g=ðcm2 sÞ, dash
line qmz ¼ 0; 2 g=ðcm2 sÞ,
dash line with single doted
line qmz ¼ 0; 3 g=ðcm2 sÞ,
solid line qmz ¼ 0; 5 g=ðcm2 sÞ
264 7 The RF Plasma Torches

Fig. 7.49 The temperature

for the atom–ion component
(a), the electron density
(b) and the gas velocity (c) at
the disturbance of ionization
equilibrium and
Te = 10,000 K: notation is the
same as in Fig. 7.48

movement of gas can lead to the substantial disturbance of the plasma ionization
equilibrium.
When the electron temperature Te  9000 K (Fig. 7.50), the cold gas is practi-
cally not heated (Ta \1000 K), and when Te ¼ 10;000 K and above (Figs. 7.48,
7.49a) there is its efficient heating. This means that in Argon plasma at atmospheric
pressure it is necessary in some way to maintain the electron temperature above
10,000 K, and only in this conditions they are able to effectively heat the cold gas
for its use as a tool of thermal effects. The electron temperature in the RF plasma
torch in area near the wall Te  8000 K, it should be expected that in this area there
will be a thermal imbalance. Therefore, in the future it is better to use a model for
the study of nonequilibrium plasma in the RF discharge, especially for plasma
torches of small and medium power.
7.4 Disturbance of the Thermal and Ionization Equilibrium in Plasma … 265

Fig. 7.50 The change of the

temperature of atom–ion
components along the z-axis
in disturbance of ionization
equilibrium and Te = 9000 K,
Ta0 = 300 K: notations are
the same as in Fig. 7.48

Fig. 7.51 The change of the

temperature of atom–ion
component along the z-axis in
disturbance of ionization
equilibrium and
Te = 13,000 K, Ta0 = 300 K:
notations are the same as in
Fig. 7.48

When the electron temperature Te

13;000 K (Fig. 7.51) the atom–ion com-
ponents are heated up quickly, that leads to a leveling of their temperature and
electron temperature, so for most of Arc plasma torches with medium and large
powers at atmospheric pressure (Te
13;000 K) there is a thermal equilibrium, and
for study of these plasma torches it is sufficient to use the model of equilibrium
plasma.
Gas with temperature Ta ¼ 2500
3000 K begins effectively heating. The data
for the cases when the gas with such a temperature can be heated even at
Te ¼ 9000 K are shown in Fig. 7.52. From this it follows that the thermal efficiency
of various plasma processes can be substantially increased if the plasma-forming
gas is preheated to the indicated temperature. For other gases, at a certain tem-
perature of electrons and pressure the temperature of the gas can be determined
using this model, in this conditions the atom–ion components are rapidly heated and
maintain the state of thermal equilibrium of discharge.

Fig. 7.52 The temperature of

atom–ion components in
disturbance of ionization
equilibrium and Te = 9000 K,
Ta0 = 2500 K: notations are
the same as in Fig. 7.48
266 7 The RF Plasma Torches

The simulation results for the RF plasma torch at atmospheric pressure

(Figs. 7.35, 7.45) show that the cold gas does not penetrate into plasma and the hot
gas with a temperature of Ta  7000
8000 K penetrates. In this case, the gas
heats up quickly and state of thermal equilibrium is established in the core of
discharge.
Data on changes in the electron density at the gas penetration in two cases
(Figs. 7.48b, 7.49b) allow us to trace how composition of plasma changes in the
disturbance of ionization equilibrium. Thermal mechanism of plasma acceleration is
clearly revealed from the results of the calculation of plasma velocity—gas at the
penetration into it is simultaneously heated and accelerated (Figs. 7.48c, 7.49c).
It should be noted that these results indicate only the convenience in use of such
a model for the qualitative analysis of disturbance of the thermal and ionization
equilibrium caused by the movement of gas. In the real conditions of working of the
RF and Arc plasma torches the gas flow is rarely a one-dimensional, in most cases it
is two-dimensional and idealized conditions when cold gas (Ta  300 K) penetrates
to the space of the heated electrons (Te [ 10;000 K), is virtually nonexistent.
Numerous experimental and theoretical data suggest that in natural conditions the
cold gas with such temperature cannot penetrate into plasma, and if to feed it into
plasma forcibly, for the dense plasma at atmospheric pressure, the temperature of
electrons in such a cold channel will be not higher than several thousand degrees of
Kelvin.
The electron temperature and specific consumption of gas do not remain con-
stant as the gas penetration into the discharge. For example, for the RF plasma torch
at atmospheric pressure the temperature of electrons in the internal layer of dis-
charge Te = 7000–8000 K, and in central and medium layers Te = 9000–11,000.
As it is described above, the calculation results depend strongly on the consumption
of gas qvz and temperature of electrons Te , so the field of gas heating in the
discharges of real plasma torches may be different.
Thus, the reduced one-dimensional model allows quickly and qualitatively
determination of the heating degree of heavy components of gas and analyze the
disturbance of thermal and ionization equilibrium in plasma, but for the study of
real processes in the RF discharges the full two-dimensional model of nonequi-
librium plasma should be used.

7.5 Simulation of the Nonequilibrium Plasma in the RF

Plasma Torches

As it was noted in analysis of disturbances of the thermal ionization equilibrium

plasma, the presence of various convection, diffusion and radiative fluxes of mass,
energy, momentum and particles in plasma can disrupt its detailed equilibrium and
put it in nonequilibrium as thermal and ionization state. For the study of
nonequilibrium plasmas in the RF discharge its two-temperature model can be used
(Sect. 1.3.), that includes: the equation of balance energy for electron gas (6.43), the
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 267

equation of balance energy for atom–ion component (6.44), the momentum equa-
tion of gas (6.25), (6.26) the continuity equation of gas (6.33), the equation if
ionization equilibrium (6.47), electromagnetic Maxwell equation (7.1) transformed
into the equation of the vector potential (7.8), the system of boundary conditions,
system of equations and for determination of composition, thermodynamic prop-
erties and transport coefficients of two-temperature plasma (Sect. 4.1–4.4).

7.5.1 Discrete Analogues of the Mathematical Model

The equations of balance energy of the electron gas and atom–ion component, the
momentum equations, the continuity equation for all the gas, that are included in the
model of nonequilibrium plasma can be written in the general form of GDE (6.2)
with the corresponding coefficients (Table 6.1). Formation of the discrete analogues
of these equations (Sect. 6.1.2, 6.1.3) is obtained on the basis of an algorithm of
solving the GDE (Sect. 6.1.1), taking into account the construction of internal
feedback loop (Sect. 6.3.2, 6.3.3).
Because of the similarity in structure, the equations of balance energy of the
atom–ion component, the electron gas, and the equation of ionization equilibrium
will be solved on the same grid [z1, r2] (Fig. 6.5a), as for solving the equation of
balance energy for the model of equilibrium plasma. Structure of the shifted grid for
solving of the momentum equationы and the continuity equation is presented above
(Sect. 6.1.2, Figs. 6.5, a, 6.6–6.8). The boundary conditions and the corresponding
discrete analogue of the momentum equations and the continuity equation were
obtained above (Sect. 7.3.1). Boundary conditions and the discrete analogue of
equation of balance energy for atom–ion gas (6.44) for internal and boundary points
have the same form as the equation of the balance energy in the model of equi-
librium plasma (6.39).
For the maximum use of boundary conditions the equation for the vector
potential (7.8) is solved using the grid [z1, r1] (Fig. 6.5a), its discrete analogue and
the boundary condition, the equations for calculation of the field and the electrical
parameters of the RF plasma torches are given above (Sect. 7.2.3).
Thus, for the complete formation of the systems of discrete analogues and matrix
of coefficients in two-temperature model of nonequilibrium plasma of the RF
plasma torches it is necessary to define the boundary conditions and the corre-
sponding discrete analogues for the boundary points of the equation of balance
energy of the electron gas and the equation of ionization equilibrium.
1. On the axis of plasma torch using the condition of cylindrical symmetry we
have
@Te
@r ¼ @n
@r ¼ 0.
e

Note that at the solving the equation of energy balance in the model of equilibrium
plasma (6.39) is taken into account such a boundary condition as on the axis of plasma
torch. Therefore, if the equation of energy balance for electron gas and the equation of
268 7 The RF Plasma Torches

ionization equilibrium are solved using the same grid [z1, r2], as for the equation of
energy balance using the model equilibrium plasma (Fig. 6.5a), the discrete ana-
logues in equation of energy balance for electron gas and the equation of ionization
equilibrium for points adjacent to the axis of symmetry, have the same form as for the
equation of balance energy in model of equilibrium plasma (Sect. 7.3.1).
2. Definition of the boundary conditions for the electron gas on the wall of the
plasma torch is more complex, as the total recording of the boundary conditions
for electrons requires consideration of transport phenomena in the regions near
the wall of plasma torch, taking into account the interaction of electrons with
material on the surface of the wall. This problem is quite complex and difficult
to achieve because of lack of information about the processes of interaction of
electrons with the material of the wall in plasma torch in terms of the external
field and the processes occurring in plasma. Therefore, the boundary condition
at the wall of plasma torch is usually defined using simple models of the
interaction of electrons with cooled walls of plasma torch [43, 45–52]. In most
cases, these are the one-dimensional balance equations for energy, current and
particles of plasma in the boundary layer of gas.

Model of the collisionless layer In the model of collisionless layer it is assumed

that the electrons with a kinetic energy me v2e =2 at the distance from the wall that are
equal to the mean free path l for particles do not have the collisions with other
particles. Because of the high mobility electrons overcome this layer faster than the
ions and charge the wall negatively. As a result is a potential drop uR occurs in the
layer with thickness of the order of Debye radius rD—in the layer of charge sep-
aration, that prevents the flow of electrons and accelerates the flow of ions, and also
leads to the equalization of ion and electron currents to the wall. Ononly those
electrons can access to the wall of plasma torch whose kinetic energy is greater or
mv2
equal to the potential energy of barrier 2 e
euR . In this case, the current density of
electrons je , ions ji and energy flux qe to the wall can be determined as
Z1

1 euR
je ¼
e ve fM;e ðve Þdve ¼
eneve exp
; ð7:54Þ
pffiffiffiffiffiffi 4 kTe
2eu R
me

Z1
1
ji ¼ e vi fM;i ðvi Þdvi ¼ enivi ; ð7:55Þ
4
0

Z1
me v2e je
qe ¼
ve fM;e ðve Þdve ¼
ðeuR þ 2kTe Þ; ð7:56Þ
pffiffiffiffiffiffi 2 e
2eu R
me
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 269

where ne ; ni —the concentration of electrons and ions at the boundary of the col-
lision layer; fM;e ; fM;i —the Maxwell distribution functions (1.2) for electron and ion
qffiffiffiffiffiffiffi qffiffiffiffiffiffi qffiffiffiffiffiffiffi
velocity; ve ¼ 8kT pme , 
e
pmi ¼
vi ¼ 8kT i
pma —the average thermal velocity of elec-
8kTa

trons and ions; me ; mi ; ma —the mass of electron, ion and atom, respectively; e—
the charge of electron.
Potential uR is determined by the condition of the balance of current densities to
the nonconductive wall je þ ji ¼ 0; taking into account (7.54) and (7.55) we have



ve kTe Te ma
euR ¼ kTe ln ¼ ln : ð7:57Þ
vi 2 Ta me

Conditions for the electron density and temperature (ne , Te ) on the border
between collision and collisionless areas can be determined from the balance of
flows of particles and energy for the condition of the quasi-neutrality on this border.
Due to the small radius of Debye these conditions at this boundary are valid to the
wall of plasma torch: ne ¼ neR ,Te ¼ TeR .
It is assumed that the flow of electrons moving to the wall comes from the area
of the collisions due to the bipolar diffusion of charged particles, so from the
condition of the balance of particles through this boundary we obtain the following
boundary condition for the electron density

 rffiffiffiffiffiffiffiffiffiffiffi
@ne neRvi kTaR

DambR ¼ ¼ neR : ð7:58Þ
@r R 4 2pma

In determination of the electron temperature at the wall it is assumed that the

cooling of electron gas occurs due to diffusion of high-energy electrons that can
overcome the potential barrier in the layer of charge separation. In this case, the
total energy flux density in view of transfer of the ionization energy that is released
at the recombination at the wall, will be

jeR
qe ¼
ðeuR þ 2kTeR þ Ei Þ: ð7:59Þ
e

At the same time, this energy from the collision area is supplied with the dif-
fusion flux of electrons and due to the thermal conductivity, while


@Te 5 jeR
qe ¼
ke
kTe þ Ei : ð7:60Þ
@r 2 e

The energy balance on the border of the collision and collisionless regions (7.59)
(7.60) is
270 7 The RF Plasma Torches



@Te kTeR jeR

ke ¼
euR :
@r R 2 e

Taking into account (7.55), (7.57) and condition of the balance in current
densities at the boundary of two areas we obtain the boundary condition for the
electron temperature in the form



  rffiffiffiffiffiffiffiffiffiffiffi
@Te kTeR TeR ma kTaR

ke ¼ ln
1 neR : ð7:61Þ
@r R 2 TaR me 2pma

Relation (7.61) is consistent with the data of [45], where the heat flux to the wall
using its catalytic properties was calculated:



 
@Te 1
k nevi 1
b vi 3

ke ¼
ð1
bÞ ln kTe þ b kTe :
@r R 1
b 4 2 ve 2

Here k; b—the coefficients of reflection from the wall for electrons and heavy
particles. As it is shown in calculations [45], the change in reflection coefficients
from 0 to 0.99 does not affect the integral characteristics and axial values Te0 ; ne0 ,
on the axis, but close to the wall, it may have a significant impact.
Model of elastic collisions In the conditions of strong energy exchange between
atom and ion components and electron gas it can be offered that all the energy of
electric field near the wall is spent on heating of the heavy plasma components in
elastic collisions with electrons. The energy balance near the wall of the plasma
torch can be written as

re E 2 r¼R ¼ bðTe
Ta Þjr¼R ; ð7:62Þ
qffiffiffiffiffiffiffi
Here b ¼ 32 k 2m
ma ne
e
pme ðna Qea þ ni Qei Þ—the coefficient of energy exchange of
8kTe

electrons with atoms and ions (Sect. 1.3).

Since the electric intensity E on the wall of plasma torch and the temperature of
the heavy particles Ta ¼ TaR are known, then from (7.62) we can obtain the
equations for the boundary temperature

re E 2  e2 m a E 2
TeR ¼ TaR þ ¼ TaR þ hqffiffiffiffiffiffiffiffi i2 : ð7:63Þ
b r¼R
pme ðna Qea þ ni Qei Þ
8kTeR
3k

Calculation of the power transmitted to the heavy particles by electrons (Fig. 4.27)
through elastic collisions showed that at low electron temperature near the wall of
plasma torch (Te < 7000 K), they almost cannot transfer the energy to atoms and ions
due to the low concentrations of electrons and effective interaction cross section. In
this case, the assumption about the heating of the heavy components of plasma due to
electron impact in accordance with (7.62) is no longer justified. In particular, in the
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 271

region near the wall of the plasma torch (electrons in this region have relatively little
energy) atoms and ions are heated not due to elastic collisions with electrons, but as a
result of other mechanisms of heat transfer such as conduction and convection.
Model of adiabatic wall The simplest condition for determination of temperature
and concentration of electrons on the wall is the assumption of adiabatic walls of
plasma torch for the electron gas



@Te @ne
¼ ¼ 0; ð7:64Þ
@r R @r R

i.e., energy that electrons have when they reach the wall is the same that they have
in reflection from it (there is no conductive heat flux for electrons). The presence of
collisionless shielding layer with a negative potential on the wall makes this con-
dition more reasonable.
As a result of numerical computations using these models it is found that differ-
ences TeR near the cold walls of the channel of electric Arc are quite large. However,
this does not affect to the characteristics of plasma distribution over the cross-section
of the channel. The deviations of calculation results for the different models are
already at r/R < 0.9 and less than 3 %. It is noted in [53, 54]. At the heating of the
walls of the channel or decreasing the arc current the differences in the values TeR ,
found using different models reduce. This is due to the fact that for environments such
as plasma, the boundary values depend on the dynamics of the processes in the
internal areas, but not vice versa, so a small error in the determination of the boundary
conditions can not have a significantly influence to the development of processes in
plasma. Therefore, in a two-temperature plasma model, the simpler boundary con-
dition can be used for determination of TeR on the wall of plasma torch.
3. In the case where the entrance of plasma torch (z = 0) is at the long distance
from the discharge, the boundary condition for the electron gas at the entrance
of plasma torch can be determined using a simple model (7.64) or take the
equilibrium value Te ¼ Ta , ne ¼ ne0 (ne0 is found by the Saha equation (4.1)).
From the condition of steady state at the exit of plasma torch the boundary
condition for the electron gas in this area will be:

@Te @ne
¼ ¼ 0: ð7:65Þ
@z @z

Note that the decision of equation of balance energy for atom–ion gas (6.44) and
the equation of balance energy for equilibrium plasma (6.39) such as the boundary
condition at the exit of the plasma torch is taken into account. If the equation of
balance energy of the electron gas and the equation of ionization equilibrium is
resolved on the same grid [z1, r2] as the equation of balance energy of equilibrium
plasma (Fig. 6.5a), then the discrete analogue of the equation of balance energy of
272 7 The RF Plasma Torches

electron gas and the equation of ionization equilibrium for the points adjacent to the
exit, have the same form that the equation of balance energy of equilibrium plasma
(Sect. 7.3.1).

7.5.2 Comparative Analysis of the Simulation Results

For comparison with the results of simulation of equilibrium plasma (Sect. 7.3.2)
perform the simulation of nonequilibrium plasma in the RF plasma torch (5 MHz)
with the same regime of operation (with two streams of argon gas Q1 = 0,
Q2 = 20 l/min, Q3 = 40 l/min) [55, 56]. The obtained results are also listed for
sections z = 4, 10, 16, 26, 36, 48, 80 and 120 mm.
Figure 7.53 presents the distribution of the electron and atom–ion temperature in
these sections. Inside the discharge (r < 15 mm) there is equality of temperatures
for electrons and heavy particles (Te = Ta). Divergence of temperatures takes place
only in the external layer (r > 15 mm) of the discharge. In the layer near the wall
the temperature difference can be up to ΔT = Te − Ta = 5000–6000 K, that indi-
cates the existence of a zone of thermal nonequilibrium in the RF plasma torch at
atmospheric pressure.
Comparison of the results also shows that the temperature of electrons and heavy
particles inside the discharge (Fig. 7.53) does not differ from the temperature
obtained by the model of equilibrium plasma (Figs. 7.36, 7.37). High temperature
(Te > 9500 K) and electron density (ne > 1016 cm−3) in the core layer of the plasma
can effectively transfer energy to atoms and ions through elastic collisions and
maintain plasma in thermal equilibrium.
Figures 7.54 and 7.55 shows the distribution of the densities for electrons and
atoms in argon plasma in the different sections, and in Fig. 7.56—the distribution of
frequencies at the elastic collisions meai of electrons with atoms and ions. These data
allow us to determine the rate of energy exchange between electrons and heavy
particles and analyze the heating of gas in discharge.
In the near-wall zone ne < 1015 cm−3 (Fig. 7.54) and the frequency of collisions
meai < 1010 s−1 (Fig. 7.56),—the electrons cannot effectively heat the atoms and
ions in elastic collisions (Fig. 7.57), plasma is in the state of thermal nonequilib-
rium (Fig. 7.53).
Balance of power in the channel of plasma torch allows determination of the
ratio of energy exchange in the discharge. Figure 7.57 shows the distribution of the
total specific power density P = σE2, received by electrons from the electromag-
netic field. The electrons through the elastic collisions give the power Pai to atoms
and ions (Fig. 7.58), and some of this power comes free from plasma by radiation
Urad (Fig. 7.59).
Accounting of nonequilibrium
 ionization
in this model allows us to find the part
of the power PI ¼ div Ei ne
52 kTe na ~ v spent on uncompensated ionization that
occurs mainly in the areas of active ionization (Fig. 7.60), flow of the current
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 273

Fig. 7.53 The electron Te–Te8 and atom–ion Ta1–Ta8 temperatures in the cross-sections z = 4, 10,
16, 26, 36, 48, 80, and 120 mm, respectively

(Figs. 7.24, 7.25) and the allocation of capacity (Figs. 7.26, 7.27, 7.57) in the skin
layer of plasma.
The distribution of the specific heat capacity of the electron gas heating is shown
in Figs. 7.61 and 7.62. The balance of power in the central section of inductor.
An important characteristic of the state of ionization equilibrium in plasma is
shown in Fig. 7.63 using the source function of equation of ionization equilibrium
Sne ¼ ki ne na
kr n2e ni —the resulting source of production (birth) and loss (dead) of
charged particles in the reactions of ionization and triple recombination.
As it was expected, the area of ionization disequilibrium in plasma is in the field
of active current and maximum energy release. In the remained part of the channel
in the RF plasma torch Sne ’ 0 and plasma is in ionization equilibrium. The charged
274 7 The RF Plasma Torches

Fig. 7.54 The electron concentration lgne1–lgne8 in the sections z = 4, 10, 16, 26, 36, 48, 80, and
120 mm, respectively

Fig. 7.55 The atom concentration lgna1–lgna8 in the sections z = 4, 10, 16, 26, 36, 48, 80, and
120 mm, respectively

particles appear mainly in the area of current flow and allocation of power in
plasma, but due to convection and diffusion, they are carried to other parts of the
discharge, where they recombine.
Thus, using the two-temperature model a complete picture of gas heating in the
RF discharge is obtained. The area of thermal (area near the wall) and ionization
(area of flow current and energy release) nonequilibrium plasma are clearly defined,
as well as the area of the active heating of the gas in the channel of the RF plasma
torches.
Basically, the RF discharge at atmospheric pressure is thermal equilibrium,
except for the area near the wall where the gap of the electron temperature and the
temperature of heavy components may reach several thousand degrees of Kelvin.
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 275

Fig. 7.56 The frequency of elastic collisions of electrons with atoms and ions lgνeai1–lgνeai8 in the
sections z = 4, 10, 16, 26, 36, 48, 80, and 120 mm, respectively

Fig. 7.57 The specific power P ¼ re E 2 obtained by electrons from electromagnetic field

Fig. 7.58 The distribution of specific heat power of the heavy components from the electron gas
Pai ¼ bðTe
Ta Þ
276 7 The RF Plasma Torches

Fig. 7.59 The specific power of radiation Urad

Fig. 7.60 The specific power PI spent on uncompensated ionization

Fig. 7.61 The specific power of the heating of electron gas Pe ¼ rE 2
Pai
Urad
PI
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 277

Fig. 7.62 The specific power in the central section of inductor z4 = 26 mm

Fig. 7.63 The source of production and loss of charged particles Sne

However, the temperature and the electron density in this zone are low, and the
thermal nonequilibrium is almost no effect on the discharge parameters. This means
that we can use the approximate models to determine the boundary conditions of
the electron gas on the wall of plasma torch. The coincidence of the simulation
results with the use of two models (one- and two-temperature) in the study of the
RF discharge at atmospheric pressure makes it possible in the future to solve the
technological problems to use a simple model of equilibrium plasma.
The RF plasma torch with a cold channel As it was noted above the RF dis-
charge due to the technological properties, such as purity of plasma that is not
contaminated by materials of electrodes, the relatively large spatial stream, a vir-
tually unlimited lifetime in any gaseous medium is widely used in plasma chemistry
and processing of material.
278 7 The RF Plasma Torches

One of the most common methods of introducing the powder material and
chemicals for the processing into the channel of the RF plasma torch is the input
passing through a central gas flow (Fig. 7.64). There is also a method of the
entrance of particles and chemical into the plasma jet from a side of the plasma
torch. However, this method has a disadvantage: the particles fly towards the cold
stream of shielding gas in the zone of the plasma jet expansion, that itself prevents
the penetration of particles into plasma. At the same time, the low temperature of
plasma jet (T = 7000–8000 K) cannot provide the efficient heating of particles and
materials.
In this case the first method of the entrance of particles through the central tube it
is preferable, because using it the particles easier penetrate to plasma and rapidly
fall into the high temperature zone of discharge at the center of inductor. In order to
overcome this reverse flow of the gas flow in inductor (Fig. 7.35) it is necessary to

Fig. 7.64 The atom–ion temperature in the RF plasma torch (5 MHz) with a cold channel
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 279

give a certain velocity to particles, i.e., feed a corresponding amount of transport

gas through the central channel. In the RF plasma torches for chemical analysis and
decomposition of harmful substances such a channel is also used to enter the
chemical reagents into the plasma core (Fig. 7.64).
The transport gas together with particles or chemicals, as their penetration into
plasma is heated and simultaneously cools plasma, thereby forming a cold channel
in the center of plasma torch. Heating rate and efficiency of chemical reactions and
processing in the RF discharge is directly dependent on the temperature, gas flow
rate in a channel and discharge parameters. Therefore, the definition parameters of
the discharge in the event of such a channel are an important practical task simu-
lation of the RF plasma torches.
For comparison with the results obtained above, simulation of the RF plasma
torch (5 MHz) was carried out with the same initial data (Table 7.3), but with
additional stream of transport argon gas (Q1 = 8 l/min) supplied to the entrance by
the central tube with diameter of 3 mm [57, 58]. As in the case of the RF plasma
torch (5 MHz) with two flows, we present results in the same cross-sections: z = 4,
10, 16, 26, 36, 48, 80, and 120 mm.
The atom–ion temperature field Ta is shown in Fig. 7.64, and the distribution of
Te and Ta in the same sections is shown in Fig. 7.65. At these pictures we can see a
gap of the temperatures Te and Ta, and formation of cold channel in the center of RF
plasma torch. In addition to the near-wall zone there is another zone of thermal
nonequilibrium—the zone of cold channel. Here, there is also some discrepancy
between the temperatures of electron and atom–ion gases. Disturbance of thermal
equilibrium, caused by the movement of gases, was obtained on the basis of
one-dimensional problem of gas heating in the half-space of plasma (Sect. 7.4).
However, unlike to this case in cold channel the electron temperature is not
specified, but it is determined by solving the equation of the energy balance for the
electron gas.
From these results we can say that the electron temperature in a cold channel is
not higher than 3000 K, and the temperature of the atom–ion gas is lower. Cold
channel is located on the axis of symmetry, the electrons, atoms and ions inside a
channel are heated mostly due to mixing with the other layers of more heated
plasma. On the axis of discharge electrons also cannot get the energy from electric
field due to zero intensity of this field.
The data presented in Figs. 7.64 and 7.65 also allow the determination of the
zone of thermal equilibrium in discharge as compared with the data of the RF
plasma torch (5 MHz) with two streams (Figs. 7.35, 7.53) it is smaller and is biased
towards the wall of the plasma torch. When the cold channel appears, the plasma
flow as it is displaced to the wall, and a picture of plasma flow is stored mainly in
the core of discharge, including the reverse vortex plasma flow (Fig. 7.66) in the
upper zone of inductor. However, due to the displacement of plasma to the wall the
coupling coefficient of inductor with plasma, current density and the allocated
power increase. As a result, we obtain the total allocated power in plasma
P2 = 14.6 kW, the total current in plasma I2 = 407 A (compared with
280 7 The RF Plasma Torches

Fig. 7.65 The electron Te1–Te8 and atom–ion Ta1–Ta8 temperatures in the sections z = 4, 10, 16,
26, 36, 48, 80, and 120 mm, respectively of the RF plasma torch (5 MHz) with a cold channel

P2 = 13.3 kW and I2 = 400 A in the case of two gas streams at the same voltages
of inductor U1 = 4.4 kV and a current in inductor I1 = 120 A).
The field of the gas flow is shown in Fig. 7.66, and the axial gas flow velocity
vz—in Fig. 7.67. The distribution of the specific consumption of the gas flow is
shown in Fig. 7.68. The big values of specific consumption (ρvz = 3 g/(cm2 s)) in
the center of plasma torch should be noted. The above analysis (Sect. 7.4) shows
that even the high-temperature electrons (Te = 10,000–11,000 K) cannot provide a
heating of such amount of cold gas and the discharge stay cold.
From the comparison of the obtained results it follows that there are large
differences between the parameters and areas of the flow (Figs. 7.35, 7.66), speeds
(Figs. 7.38, 7.67) and specific consumptions (Figs. 7.41, 7.68) of plasma flows
7.5 Simulation of the Nonequilibrium Plasma in the RF Plasma Torches 281

Fig. 7.66 The field of gas flow in the RF plasma torch (5 MHz) with a cold channel

Fig. 7.67 The axial gas flow velocity vz in the RF plasma torch (5 MHz) with the cold channel
282 7 The RF Plasma Torches

Fig. 7.68 The specific consumption of plasma flow ρvz in the RF plasma torch (5 MHz) with a
cold channel

without the central transport gas and with it. We see that with advent of the cold
channel parameters of the RF discharge rarely change, it should be considered using
RF plasma torches for the treatment of particles and plasma chemistry.

References

1. Dresvin S.V. Low-temperature plasma generators // Encyclopedia of Low-Temperature

Plasma / Ed. by V.E. Fortov. M.: Nauka, 2000. V. II. PP 280–328.
2. Dresvin S.V., Donskoy A.B., Goldfarb V.M. Determination of conductivity of RF discharge
in Argon // J. Technical Physics. 1965. V. 35. No. 9. PP. 1646–1653.
3. Dundas P.D. Induction plasma heating // Report NASA. 1969. No. 11487.
4. Vermeulen R.S., Lee Bowden, Vieram P.N. Regulation of the moving state of the plasma
using RF electromagnetic fields // Rocket technical and Astronautics. 1967. V. 5. No. 12.
PP. 251–260.
5. Dymschits B.M., Koretsky J.P. Experimental study of inductive discharge // J. Technical
Physics. 1964. V. 5. No. 19. PP. 1677–1679.
6. Jonston P.D. Determination of temperature in a radiofrequency discharge using a reversal
technique // Brit. J. Appl. Phys. 1968. Ser. 2. No. 1. PP. 479–484.
7. Molinet F. Иccлeдoвaниe pacпpeдeлeния элeктpoннoй тeмпepaтypы в apгoнoвoй плaзмe,
вoзбyждaeмoй BЧ-гeнepaтopoм // C.r. Acad. Sc. 1966. Vol. 262. No. 21. PP. 1377–1380.
Study of the electron temperature distribution in an argon plasma excited by RF generator
8. Goldfarb V.M., Goikhman V.H. Characteristics and possible use of spectroscopic
high-frequency discharge at atmospheric pressure // J. Plasma chemistry reactions and
processes. 1968. V. 8. No. 2. PP. 193–196.
9. Determination of the temperature in the steady high-frequency induction discharge / R.E.
Rovinskii, V.A. Gruzdev, V.M. Gutenmakher et al. // J. Thermal physics of High
Temperatures. 1967. V. 5. No. 4. PP. 557–561.
10. Goldfarb V.M., Goikhman V.H., Dresvin S.V. Characteristics and possible use of
spectroscopic high-frequency discharge at atmospheric pressure // Proc. Colloq. Spectrosc.
Hungary. 1967. PP. 751–760.
11. Eckert H.U., Kelly F.L., Olsen H.N. Spectroscopic observation on induction-coupled plasma
flames in air and argon // J. Appl. Phys. 1968. Vol. 39. No. 3. PP. 1846–1852.
References 283

12. A study of RF plasma flame of Argon burner / V.M. Goldfarb, Donskoy A.V., Dresvin S.V.
et al. // J. Thermal physics of High Temperatures. 1967. V. 5. No. 4. PP. 549–555.
13. Apsit A.R., Goikhman V.H. Preparation and study of pulsed RF inductive discharge at
atmospheric pressure // J. Technical Physics. 1970. V. 5. No. 7. PP. 1551–1560.
14. Dresvin S.V., Klubnikin V.S. The study of nonequilibrium in a stream of Argon plasma RF
discharge at atmospheric pressure // J. Thermal physics of High Temperatures. 1971. V. 9.
No. 3. PP. 475–480.
15. Biberman L.M., Vorob’ev V.S., Yakubov I.T. Low-temperature plasma with nonequilibrium
ionization // Successes of physical sciences. 1979. V. 128. No. 2. pp. 233–271.
16. Velikhov E.P., Kovalev A.S., Rakhimov A.T. Physical phenomena in gas-discharge plasma.
M .: Nauka, 1987.
17. Goldfarb V.M., Ilyina E.V. Non-equilibrium low-temperature plasma sources. The popula-
tions of the levels and spectral diagnostics // Proceedings of the Academy of Sciences of the
USSR. 1975. No. 3. Vol. 1. pp. 28–38.
18. Goikhman V.H., Goldfarb V.M. RF thermal discharge. // J. Plasma chemical reactions and
processes / Ed. L.S. Polak. M.: Science. 1977. PP. 232–278.
19. Reboux J. Ingeniers et techniciens. 1963. No. 166. P. 109.
20. Chase J.D. Magnetic Pinch Effect in the Thermal RFI Plasma // J. Appl. Phys. 1969. Vol. 40.
No. 1. PP. 318–325.
21. Chase J.D. Theoretical and Experimental Investigation of Pressure and Flow in Induction
Plasmas // J. Appl. Phys. 1971. Vol. 42. No. 12. PP. 4870–4879.
22. Klubnikin V.S. Thermal and gas-dynamic characteristics of the RF discharge in a stream of
Argon // J. Thermal physics of High Temperatures. 1975. V. 13. No. 3. PP. 473–474.
23. Sorokin L.M. Theory electric Arc in a forced heat exchange. M.: Nauka. 1977.
24. Sorokin L.M. // J. Physics and chemistry of material treatments. 1980. PP. 32–34.
25. Raiser Y.P. // Successes of physical sciences. 1969. V. 99. No.. 4. P. 545.
26. Soshnikov V.N., Trekhov E.S. On the theory of RF vortex high pressure discharge //
J. Thermal physics of High Temperatures. 1966. V. 4. No. 2. P. 166.
27. Dresvin S.V. // Abstracts of the VII Conf. by generators of Low-temperature plasma.
Alma-Ata 1977.
28. Dresvin S.V., El-Mikati H. // J. Thermal physics of High Temperatures. 1977. No. 2.
29. Thomson J.J. // Philos. Mag. 1926. Vol. 2. P. 674.
30. Thomson J.J. // Philos. Mag. 1927. Vol. 4. P. 1128.
31. Fomenko A.A., Trekhov E.S. Questions of Low-temperature plasma physics. Minsk: Science
and Technology, 1970. PP. 195–198.
32. Eckert H.U. Measurement of the Magnetic Field: Distribution in a Thermal Induction Plasma
// J. Appl. Phys. 1971. Vol. 42. No.. 8. P. 3108–3113.
33. Eckert H.U. Dual Magnetic Probe Systems for Phase Measurement in Thermal Induction
Plasma // J. Appl. Phys. 1972. Vol. 43. No. 6. P. 2707–2713.
34. Nguyen Quoc S. 2D-Electromagnetic field calculation on the RF plasma torch // Proc. of the
Inter. Sem. on Heating by Internal Sources. Padua. Italy, 2001. P. 609–615.
35. Mostaghimi J., Boulos M. Two-Dimensional electromagnetic field effects in induction plasma
modeling // J. Plasma Chem. Plasma Process. 1989. Vol. 9. P. 25.
36. L.R. Neiman, K.S. Demirchyan. Theoretical Foundations of Electrical Engineering. L.:
Energoizdat, 1981. V. 2.
37. Korn G., Korn T. Handbook of mathematics for scientists and engineers. M.: Science. 1984.
38. Kalantarov P.L., Zeitlin L.A. Calculation of inductances: Handbook. L.: Energoatomisdat
1986.
39. RF multiflux plasmatron study. Experimental measurement and modeling / D. Morvan,
J. Erin, S. Magnaval et al. // Proc. of the XII Inter. Sym. on Plasma Chemistry —
Minneapolis. USA, 1995. PP. 17430–1748.
40. Modelling of the RF multiflux plasmatron / D. Morvan, J. Erin, S. Magnaval et al. // Proc. of
the IV European Conference on Thermal plasma Processes. Athens. Greece, 1996. PP.712–
717.
284 7 The RF Plasma Torches

41. In-flight measurement of particle size and velocity in a radio frequency plasma torch / E.
Franke, F. Krayem, D. Morvan et al. // Proc. of the V European Conference on Thermal
plasma Processes. St. Petersburg, 1998. PP. 347–353.
42. Experimental investigation of impurity evaporation from powders injected in an RF plasma
torch // D. Morvan, J. Amouroux, S. Magnaval et al. // Proc. of the IV European Conference
on Thermal plasma Processes. Athens. Greece, 1996. PP. 713–720.
43. Physics and technology of Low-temperature plasma / Ed. by S.V. Dresvin. M.: Atomizdat.
1972.
44. Dresvin S., Amouroux J., Nguen Quoc Shi. Analisis of deviation from thermal and ionization
equilibrium in an Argon plasma flow // J. High temperature Material Processes. 1997. Vol. 1.
No. 3. PP. 369–381.
45. Clarke C.J., Inkropera F.P. Thermochemical nonequilibrium in a stabilized Argon Arc plasma
// J. Rocket and technical complex. 1972. V. 10. No. 1. PP. 19–21.
46. The calculation of the initial section of the channel characteristics of the electric Arc on the
basis of two-temperature plasma model / V.M. Lelevkin, E.P. Pakhomov, V.F. Semenov, V.S.
Engelsht // J. Thermal physics of High Temperatures. 1986. V. 24. No. 3. PP. 587–593.
47. Mathematical modeling of electric Arc / Ed. V.S. Engelsht. Frunze: Ilim. 1983.
48. Asinovsky E.I., Pakhomov E.P. An analysis of the temperature field in a cylindrically
symmetric pole of electric Arc // J. Thermal physics of High Temperatures. 1968. V. 6. No. 2.
PP. 333–336.
49. Litvinov I.I., Lyumkis E.D., Philipov S.S. Nonequilibrium model of a strongly radiating
electric discharge in an inert gas // Preprint No. 135. M. 1976. P. 72.
50. Gogosov V.V., Shelchkova I.N. The conclusion of the boundary conditions for concentration,
velocity and temperature components of a partially ionized plasma taking into account the
capacity of the parietal falls // USSR Academy of Sciences. Ser. Fluid Mechanics. 1974.
No. 5. PP. 76–88.
51. Gogosov V.V. Shelchkova I.N. Examples of simple boundary conditions for the electron and
ion densities and temperatures of electrons and heavy particles // USSR Academy of Sciences.
Ser. Fluid Mechanics. 1976. No. 6. PP. 169–172.
52. Nazarenko I.P., Panevin I.G. Calculation canal stabilized Arcs based on radiative transfer and
nonequilibrium plasma // Electric Arc theory in conditions of forced heat. Nauka.
Novosibirsk. 1977. PP. 61–67.
53. Alievsky M.Y., V.M. Zhdanov. The transport equations for a many-component
non-isothermal plasma // J. Appl. Phys. 1963. No. 5. pp. 11–17
54. Dresvin S.V. Fundamentals of theory and design of RF plasma torches. L.: Energoatomizdat
1991.
55. Modelling of the RF multiflux plasmatron / D. Morvan, J. Amouroux, S. Magnaval et al. //
J. High Tem. Mat. Proc. 1998. Vol. 1. No. 3. P. 78–85.
56. Caculation of RF plasma torch parameters by means of nonequilibrium model of Ar Plasma /
S.V. Dresvin, S. Nguyen-Kuok, D. Ivanov, J. Amouroux // Proc. of the VI European
Conference on Thermal plasma Processes. Strasbourg, 2000. P. 257–267.
57. Calculation of temperature and flow in RF plasma torch. Influence of cold channel produced
by axial gas injection / S. Nguyen-Kuok, S. Dresvin, J. Amouroux, D. Morvan // Proc. of the
V European Conference on Thermal plasma Processes. St. Petersburg, 1998. P. 169–175.
58. Modeling of coupled motion, heat and mass transfer, electromagnetic and atomic hydrogen in
a Ar — H2 thermal RF plasma / Ph. Mandin, D. Morvan, J. Amouroux et al. // Proc. of the
XIV Inter. Sym. on Plasma Chemistry. Prague, 1999. Vol. 1. P. 415–420.
Chapter 8
The Arc Plasma Torches

8.1 The Features of the Arc Plasma Torches

Arc plasma torches are one of the most common types of discharge with medium
and high pressures and are widely used in engineering and technology. The dif-
ferences of Arc plasma torches from other types of discharge are the high density
(j = 102–107 A/cm2) and low cathode voltage drop. Thus, if the cathode voltage
drop in the glow discharge is usually Uk ≥ 100 V, in Arc discharge, it is only
Uk = 10–40 V. This is due to the fact that effective mechanisms of thermo-emission
of electrons from the cathode of the Arc provides a current close to the total current
of the Arc, and it is not necessary to increase it due to high Uk.
The created devices for controlling the gas flow in the electric Arc—the Arc
plasma torches allow to obtain the plasma flow with temperature T = (10–
50) × 103 K—and the velocity of the flow v = 10–104 m/s in inert and molecular
gases at the atmospheric pressure. The large number of types and constructions of
electric Arc plasma torches is developed that are intended for different purposes.
They are used for cutting, welding and melting of metals, covering with
heat-resistant coatings, heating of the gas in the aerodynamic tubes and chemical
reactors, heating of the gas in the direct reduction of metals from ores, the inten-
sification of blast and open-hearth processes. The Arc plasma torches are also used
for thermal drilling of hard rocks, in spectrometric study of properties of the gas, to
obtain a stable standard of brightness, as the jet thrusters for spacecraft and others.
The Arc plasma torches can be divided into two great classes, fundamentally
different from each other. First class—the plasma torches with electric Arc of the
direct action, where the electric Arc lights and closed on technological object—
metal (Fig. 8.1). The second class—Arc plasma torch, generating a plasma jet,
where the electric Arc is in the channel of the plasma torch (cathode and anode
inside the channel) and technological effect—is achieved indirectly through expo-
sure of the Arc to the plasma jet (Fig. 8.2).

© Springer International Publishing Switzerland 2017 285

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_8
286 8 The Arc Plasma Torches

Fig. 8.1 The Arc plasma torches with the direct action [1]: a, b, e, f with the compressed Arc in
the channel (cutting, spraying and planing); c, d with open Arc (welding, melting)

By the way of exposure to the Arc and its stabilization in the space the plasma
torches of the first class are divided into (1) plasma torches with the Arc stabilized
by the gas flow in a space (welding and melting of metals); (2) plasma torches with
the Arc, compressed in the channel and stabilized by the gas flow (plasma Arc
cutting of metals).
8.1 The Features of the Arc Plasma Torches 287
 288 8 The Arc Plasma Torches

b Fig. 8.2 Arc plasma torches, producing the plasma jet [1]: a with axial gas stabilization, b with
axial fixation of the Arc length by the inter-electrode insert (IEI), c with fixation Arc length by
ledge, d double chamber with vortex stabilization, self-setting Arc length, the hollow cathode,
e double chamber with magnetic vortex stabilization, self-setting Arc length, f with magnetic
vortex stabilization and fixation of the Arc length by IEI and the ledge, g with vortex stabilization,
distributed gas supply and fixing the Arc length by IEI, h coaxial with magnetic vortex
stabilization, i three phase battery, j three phase battery with the vortex stabilization, k with the
imposition of the Arc and axial stabilization

Plasma torches of both the first and the second class may be classified using the
following main features (Fig. 8.2):
1. The method of space stabilization of the Arc in the channel of plasma torch [gas
vortex (d–j), gas with straight-jet (a–c, k), magnetic (e, f, h)];
2. The type of cathode [tungsten (a–c, j, k), thermo-chemical (e–g), hollow copper
(d, i)];
3. The methods for fixing of the Arc length in the channel [with the inter-electrode
inserts (b, f, g), with fixation by the ledge (c, f)].
4. The degree of reduction of the electrodes erosion in the area of the cathode and
anode spots.
5. The electric current [with DC (a–h, k) or AC (i, j) current].
The main part of the Arc plasma torch is the electric Arc between the electrodes.
The numerous physical processes take place in the electric Arc: electromagnetic,
thermal, gas dynamics, radiation, ionization and recombination, and others. These
processes are coupled and their characteristics have the spatial distribution and
change at the changes of operation regime. Contingently, the electric Arcs can be
divided into several areas and each of them has the specific properties.
1. Near-electrode areas, rather rigidly connected to the electrodes. Their longitu-
dinal dimensions vary little with increase or decrease of the gap. The
nonequilibrium state of the plasma and the existence of space charge are
characterized for this zone.
2. The transitional areas adjacent to areas near the electrodes. Here the physical
processes are two-dimensional, the cross-section and other properties vary in the
axial direction. There is a close relationship between heat, gas-dynamic, and
electrical parameters of the discharge.
3. The Arc with the cylindrical symmetry, with the axial uniformity of physical
properties and phenomena. Changes of the properties of discharge in its
cross-section are almost exclusively determined by the processes of radial
transport.
4. The jet of the Arc plasma stream output caused by interaction with the sur-
rounding cold gas medium and the treated material. As a result of this inter-
action, the temperature and velocity of the Arc jet gradually decrease to the
temperature and velocity of the environment.
8.1 The Features of the Arc Plasma Torches 289

Fig. 8.3 Current–voltage characteristics of the electric Arc (a) and the characteristic distributions
of electric potential along the length of the Arc (b): 1 decreasing characteristic, 2 characteristic
with the growing region, 3 the distribution with the positive decrease of anode, 4 the same with
negative decrease, I near-cathode region, II the Arc column, III near-anode region

The current–voltage characteristics of electric Arc in many cases are decreasing.

However, for the Arc in the channel where the growth of the diameter of the plasma
column is limited by the walls at the increase of current, the growing area can
appear at the current–voltage characteristic (Fig. 8.3a). The choice of a method for
stabilizing of the discharge, of the power supply, the efficiency of the Arc instal-
lation and other properties of the Arc are determined using the current–voltage
characteristic.
Near-electrode processes of the electric Arc are emerged as an independent
branch of plasma physics. In addition to clarification of the mechanisms of electric,
heat and mass transfer, important practical problem of the increase of the service
life for the electrode was solved here. And the perspectives for simulation of electric
discharges from electrode to electrode are emerged only recently. Conditions in the
near-electrode regions are signally different from the conditions in plasma column,
and three zones can be distinguished at the study of the processes in electric Arc:
the Arc column, near-cathode, and near-anode regions (Fig. 8.3b).
The electric current in the near-cathode region is created mainly due to the
motion of electrons emitted from the cathode surface and due to the ions coming
from the Arc column. In the near-anode region, the electric current mainly occurs at
the moving of electrons from the Arc column to anode. The length of the
near-electrode zones is very small and is in the order of a few mean free paths of the
particles. Consequently, the intensity of electric field in the near-electrodes areas
should be very large. For the Arc at the atmospheric pressure it is estimated at
approximately *104–106 V/cm near the cathode surface.
At the Arc column, the intensity of electric field is largely dependent on the
diameter of the discharge channel, current, type and regime of the gas flow and
other conditions and has a much smaller value compared with the value at the
electrodes *(5–20 V/cm). Some larger value of the intensity of electric field can be
obtained by holding of the Arc column using the additional magnetic field. The Arc
290 8 The Arc Plasma Torches

column is constricted and a strong constriction of the Arc is observed near the
electrode; and the diameter of the Arc near the cathode is typically less than near the
anode. The physical processes, leading to the narrowing of the Arc near the surface
of electrodes, are associated with condition near the electrode and with character-
istics of the Arc. Constriction of the Arc column has the thermal nature and is
linked with the challenge of thermal energy from the central zone to the periphery
due to heat conductivity and convection. At a sufficient current of the intrinsic
magnetic field of the Arc it reaches a large value, leading to a magnetic compression
of the Arc column.
The theory of electric columns was developed in a phased accounting of the
conditions and processes where the used approximation is valid. Further develop-
ment of the theory is mainly due to various external factors that are taken into
account: the longitudinal and transverse blowing of the Arc discharge, interaction of
the Arc with electric and magnetic fields, taking into account the impact of the
near-electrode processes and walls of the channel, and others.
The series of monographs, summaries and reviews such as the famous book of
Finkelberg and Mekker [2] are devoted to the theory of electric Arcs. It provides a
detailed qualitative description of physical processes in the electric Arc, the basic
theoretical models; the methods of diagnostic and experimental data are presented.
The monograph of Granovsky [3] electrodynamics of gas is described in details.
The second part of this book published by a group of authors [4] includes the
information about the classification of the types of discharges and currents. Written
by experimentalists and theoreticians, it includes data on virtually all types of gas
discharges and their classification. An important review of the analytical methods
and results of researches of the Arc discharges is given in the book edited by Polak
[5]. In addition to the problems of the plasma chemistry the new results in the
theory of electric Arcs and RF-discharges in plasma are described in it, they
stimulate the researches, the theoretical description and numerical modeling of
processes in electric Arc. The physical ideas in the theory of open free Arcs, their
movement and stability are described in the monographs of Anderson [6] and
Novikov [7].
The papers [8–10], theory and experimental description of the electric Arcs
edited by Zhukov M. and the monographs [11, 12] are devoted to the theory and
experimental description of the electric Arcs in the conditions of forced heat
exchange. The authors, in addition to the presentation of original results, summed
up the researches in this area. Stated in monographs, [11, 12] made it possible to
identify the directions for development of the theory of electric Arc plasma gen-
erators. The described problems of the transfer of radiation can be found in [13, 14].
The work of Alexandrov and Ruhadze [15] is devoted to the theory of high-current
pulsed Arc discharge. The criteria approach to the description of generators
low-temperature plasma is considered in the works [8, 16, 17]. The communication
of electric Arc with other types of discharges and a single theoretical approach to
the description of a wide range of processes in gas discharges are considered by
Raiser [18].
8.1 The Features of the Arc Plasma Torches 291

The theory of thermal plasma of the cylindrical Arc column and extended
electric Arcs is well-developed now, where the pattern of flow and heating of the
gas by the electric Arc is presented on the basis of qualitative, channel, integrated,
self-similar, and numerical methods for the solving, and the characteristics of
plasma are determined depending on the external adjustable parameters.
The basis for most of the theoretical studies of the electric Arcs is assumption of
the existence of local thermodynamic equilibrium (LTE) of plasma. However, the
application of the equilibrium plasma model for description of the processes in
electric Arc is not always justified. The experiments [19, 20] show that, in the Arc
plasma, there are spatial areas where the equality of temperatures of electrons and
heavy particles (atoms, ions, and molecules) is violated, the Maxwell velocity
distribution and Boltzmann distribution over the excited levels are not installed, and
the condition of ionization equilibrium is not fulfilled. Near the electrodes, in the
initial section of the Arc and in the peripheral regions of the discharge, the
nonequilibrium of plasma is the most significant, and the taking into account of
deviations from LTE becomes necessary.
In connection with the development of a new branch of industry—plasma
chemistry is often used for modeling of nonequilibrium plasma and modeling of
plasma and chemical reactions in a channel of the plasma torch. It is often necessary
to determine the velocity of the chemical reactions in a predetermined plasma flow,
high energy flows, the degree of heating of the chemical components, the overall
efficiency of the plasma-chemical process, etc.
Another application of the model of nonequilibrium plasma is connected with
the investigation of electrode processes in the Arcs. The relevance of researches of
such processes is related not only with the checking of physical hypotheses, with
the expansion of the circle of tasks, but also with determination of the ways for
improve of the efficiency of the practical use of Arc plasma torches. It is known that
the investigations of near-electrode processes using the model of equilibrium
plasma cannot give a satisfactory result. Because of the complexity and the mutual
influence of physical processes and the difficulties of their mathematical descrip-
tion, taking into account nonequilibrium, to a first approximation, it can be per-
formed using the model of two-temperature plasma [21]. The mathematical
formulation of this approach for the initial area of the Arc using the approximation
of boundary layer proposed at first in [22]. In [23], two-temperature approximation
for the three-fluid plasma model is proposed, but in this paper the goal is actually
stated. In [24], the method for calculation of the initial part of the electric Arc is
presented; it is based on the deviation of the plasma composition from the equi-
librium at the assumption that the temperatures of all the components of plasma are
equal. Attempt of calculation of the initial part of the Arc on the basis of
two-temperature plasma model is carried out in [25, 26], here a relatively simple
system of equations using the assumption of ionization equilibrium in plasma is
used. The development of this approach is reflected in the work [27], wherein the
framework of assumption of two-temperature layer the calculation of initial part of
the Arc in different hydrodynamic regimes of the flow of the plasma-forming gas is
made. Attempt to take into account the thermo-chemical nonequilibrium of plasma
292 8 The Arc Plasma Torches

takes place in [28–31], here the analysis of two-temperature plasma flow in the
transition region of the channel and asymptotic Arc is carried out.
Further development of the theoretical method for research of the Arc is asso-
ciated with the use of models of equilibrium and nonequilibrium plasma for
description of the dynamics of gas heating in the Arc, for the study of phenomena in
the near-electrode, near the wall areas and interaction of the plasma flows with the
disperse material. Theoretical methods should also be used to study the curved Arc,
Arc of alternating voltage and high-voltage discharge, for description of the kinetic
processes in the moving nonequilibrium plasma, especially molecular, turbulence
and interaction of the kinetic processes with the turbulence, for solving of the
unsteady problem and problem with violation of the axial symmetry [32].

8.2 The Arc Plasma Torches for Cutting and Spraying

Arc Plasma Torches for Cutting of Metals Using of the compressed plasma Arc
for cutting of the refractory metals and alloys is one of the broad areas of the
practical application of low-temperature plasma. The electric Arc, compressed in
the channel of the plasma torch and stabilized by the gas flow, has the strong
thermal and gas-dynamic effect on the anode (detail). The abilities of the com-
pressed electric Arc for intensively melting of the metal anode (detail) and actively
blow out the melt using a high-speed gas-dynamic flow are in the basis of plasma
Arc cutting of metals. Currently the plasma cutting is one of the most effective
methods of thermal cutting of metal. Its quality of the cut is only slightly inferior to
the laser cutting, but exceeds the latest in economic indexes and broader oppor-
tunities, especially relating to the maximum thickness for the cutting of metals.
The installation for plasma cutting Maxigraph DD/3.5/4.0/4.5/5.0/5.5 mm of the
company KOIKE (Japan) is shown in Fig. 8.4 has the following characteristics: the
maximum number of burners—6; the maximum velocity of cutting for parts of steel

Fig. 8.4 The installation for plasma cutting Maxigraph (a) with device of the numerical control
(b), and plasma burner (c) of the company KOIKE (Japan)
8.2 The Arc Plasma Torches for Cutting and Spraying 293

St3 with the thickness 10 mm–6000 mm/min; the effective cutting width—2.6, 3,
1, 3.6, 4.1 and 4.6 mm; effective cutting length—the length of the rail minus
1.75 m; the library of standard shapes for cutting—60; the system of numerical
control (CNC)—Hybrid 400/D–180i FS; the format of the control system EIA
(US); external interface RS–232C; plasma cutting system—max-200/HPR 130/260,
Super 200/400.
The characteristics of installations for plasma cutting of metals from the series
APR of the company “Power Engineering” are shown in Table 8.1.
High velocity of plasma and high thermal flow in anode spot of the electric Arc
is achieved by the forced compression of electric Arc in the channel of small
diameter (d = 1–5 mm) and its stabilization it in space using straight-jet or vortex
gas flow (air, nitrogen, oxygen, etc.). Below as an example there are the specifi-
cations of one of the most effective plasma torches for cut RPM-6, produced serially
with the installation of OPR-6. The construction of such plasma torch (plasma cut)
and the scheme of process are shown in Fig. 8.5.
There are some technical features of the use of compressed Arc for cutting of
metals. They include, first of all, the correct choice of power, current, the diameter
of nozzle and conformity of their thickness and velocity of movement of the cut
detail. The parameters of developed regimes of plasma Arc cutting of stainless steel
X10CrNiTi18-10 (Euro analogous) using plasma torch RPM-6 are shown in
Table 8.2. There are the values of optimum cutting velocity and the width of
resulting cut. Increase of the thickness of cut metal makes it necessary to increase
the current of the cutting Arc. However, current and cutting velocity is limited by
resistance of the nozzle in the plasma torch.
Big values of the dynamic pressure ρv2/2 and the velocity v of the gas should be
noted. The electric Arc for cutting the velocity of the flow plays a very important
role in the process of heat exchange, and at the destruction of the layer of the liquid
metal. At velocities of the flow of several kilometers per second, obviously, not
only the active melt takes place, but also mechanical action of the jet on the molten
metal. The compressed Arc simultaneously melts and blows the molten metal from
the surface of the cut. The effect of pressure p on the regime of plasma cutting and
its strong dependence on the current gas flow and the diameter of the nozzle
determine significant differences in the dynamic pressures in plasma torches of
various constructions. This, in turn, explains the difference in technical parameters
of plasma cuts of many companies. Successfully selected configuration and the
diameter of the nozzle, the proper organization of the gas vortex (there is no
recommendations in this area) can significantly increase the velocity and the
dynamic pressure of the Arc and, therefore, at the same current I to provide the
great velocity and better quality of cutting.
 294

Table 8.1 Characteristics of installations for plasma cutting of series APR

Characteristic APR-40 APR-60 APR-90 APR-91 APR-140 APR-150
Supply voltage, V 220 380 380 380 380 380
Consumption power, kW 4 16 20 14 25 29
Cutting current, A 25 (35) 40 (60) 60 (90) 60 (100) 70 (130) 50 (100, 150)
Duty cycle, % 100 (60) 70 (60) 70 (60) 80 (70) 70 (60) 100 (90, 85)
The air pressure, MPa 4–5 5–6 5–6 5–6 5–6 5–6
Air consumption, l/min 100 170 180 180 210 210
Thickness of cut material, mm
Maximum 8 18 30 35 45 50
High-quality cut 6 10–12 20–25 25–30 30–35 35–40
Mass, kg 28 60 90 100 125 170
Dimensions 320 × 400 × 625 535 × 320 × 625 535 × 320 × 625 550 × 420 × 820 550 × 420 × 820 640 × 420 × 820
(length × width × height),
mm
8 The Arc Plasma Torches
8.2 The Arc Plasma Torches for Cutting and Spraying 295

Fig. 8.5 General view of the

plasma torch RPM-6 and
scheme of the process of
metal cutting using the
compressed Arc: 1 cathode, 2
sleeve, 3 insulators, 4 union
nut, 5 housing, 6 nozzle, 7
power supply, 8 anode-detail

Table 8.2 Recommended parameters of regimes of plasma torch operating RPM-6 (nitrogen) in
the plasma Arc cutting of stainless steel of different thickness [1]
Thickness of cut Diameter of Current of Cutting Width of cut,
metal, mm channel, mm cutting Arc, A velocity, mm
mm/s Up Bottom
20 2.8 320 1700 5.5 2.8
40 4.0 500 900 6.5 3.5
60 4.8 610 540 7.5 4.5
80 5.2 650 420 8.5 7.0
100 5.6 710 300 10.5 8.5
120 6.0 740 250 11.5 9.5

Arc Plasma Torches and Technology of Plasma Spray of the Coatings Arc
plasma torches for plasma spraying are designed for the covering with the
corrosion-resistant, wear-resistant, sealing, friction, insulation, and other special
coatings on the surfaces of details. Coating often consists of refractory oxides,
296 8 The Arc Plasma Torches

carbides, nitrides, borides, composite materials. It is applied by spraying of wire

and powder materials, and there is no oxidation of the surface of detail and the
sprayed substance. Widespread use of plasma spraying is in heat engineering and
metallurgy. The turbine blades, molds, autoclaves, screws, bearings and others are
reduced using it.
The plasma coating is in the forming on the surface of details (products, con-
structions) of the layer of the powder particles that have a certain margin of thermal
and kinetic energy obtained as a result of interaction with the jet of Arc plasma. The
temperature of plasma jet can reach 5000–20,000 K, and the velocity of the flow
200–2000 m/s. In the plasma jet, the powder particles melt and get the velocity 50–
800 m/s. The velocity of powder particles depends on their size, density of the
material, the current of the Arc, nature and consumption of plasma-forming gas.
Advantages of the plasma spraying method are follows.
1. Preparation of coatings from most of materials, melting without decomposition,
and restrictions on the melting point.
2. Use for the formation of Arc plasma jet for gases of various types: inert (argon,
helium), reducing (hydrogen) and oxidant (air, nitrogen) and ammonia, natural
gas, water vapor that in combination with the use of cameras with protective
medium (vacuum) or protective nozzles allows adjustment of characteristics of
the medium, where the powder particles are heated and move.
3. Flexible regulation of electric and gas regimes of plasma torches operation,
including the process of coating of materials that allows controlling of electrical
properties of sputtered particles and the conditions for formation of the coating.
DC Arc plasma torches, operating in inert, neutral and oxygen containing
mediums, are the most widely used in industry. By way for supplying of the
plasma-forming gas there are plasma torches with tangential and axial feeds of gas.
The plasma torches of indirect action are widely used for plasma spraying of
coverings of powders.
Today in the world, there is a large number of constructions of Arc plasma
torches for spraying, working on direct current, and that have the different con-
figurations of the discharge channel, the nozzle, anode and cathode, the method of
input of powder mixture, construction, materials and methods of construction of
electrodes, the principles of the Arc stabilization. The linear (axial) scheme of
plasma torches with traverse of plasma-forming gas are most widely used
(Fig. 8.6). Electric Arc discharge burns between the cylindrical anode and the
cathode with smooth or stepped channel, and in the latter case, the expansion of the
8.2 The Arc Plasma Torches for Cutting and Spraying 297

Fig. 8.6 Schematic view of the channels of Arc plasma torches for spraying of different
constructions [33]
298 8 The Arc Plasma Torches

channel at the output is used for a gas-dynamic fixation of anode spots of discharge.
The front part of the nozzle is usually made in the form of the confuser with a
smooth transition into the cylindrical channel. At the output in some cases, the
nozzle has the stepped and tapered extension. The Laval nozzle is used to create a
supersonic flow. Increased service life of the nozzle that is usually made of copper,
is achieved through the use of tungsten inserts. Cathode is made of tungsten doped
with thorium, barium, lanthanum or other materials, reduction of the work function
of the cathode. The distance between the anode and the cathode, and the alignment
of them significantly affect the stability of the work and parameters of the plasma
jet, and the strict centering of the electrodes is recommended, especially at the axial
stabilization of the Arc. The flow or closed cooling systems are usually used for
cooling of electrodes. The separate cooling of the anode and cathode can be used
and it allows a quick change of electrodes without switching off the cooling system.
The spraying material in the form of powder with size of particles 40–100 μm,
are injected into the plasma jet into the region beyond the anode spot (Fig. 8.6a–c)
or outlet nozzle exit (Fig. 8.6d–g). This allows to split the combustion zone of the
Arc and the heating of the powder and to avoid their mutual influence that greatly
improves the reliability of the plasma torch. Entrance of the powder into the flow
through one hole, as it is done in most of the plasma torches, is simple, but is
characterized by uneven loading of the plasma jet. Significantly better results are
provided at the entrance of the powder into the flow through a few holes, it
increases the efficiency of the process.
For Arc plasma torches in Fig. 8.6, the simplicity of construction is character-
ized, but the small Arc length and, consequently, the low voltage of discharge
makes it necessary to increase the current of Arc for increase of the power that
negative affects the life time of electrodes. The plasma torches with inter-electrode
inserts have no this disadvantage, where the increase of capacity is due to increase
of the voltage in Arc. Various modifications of the plasma torch of the type PN with
sectional inter-electrode insert (Fig. 8.7) are developed (they have from 2–6 sec-
tions). Application of inserts allows to length the Arc, to decrease the current,
saving the power and, consequently, reduce the erosion of electrodes and improve
the stability of the work.
As the plasma-forming gas in the plasma torches of PN-type with tungsten rod
cathode the inert gases are widely used, and with the hafnium or zirconium cath-
odes—the air. The place and the angle of entrance of the powder relative to the axis
of the channel change depending on the type of powder and granulation.
Performance of the coating of aluminum oxide particles with diameter of 50 μm is
10 kg/h.
8.3 Simulation of the Arc Plasma Torches in the Channel 299

Fig. 8.7 Schematic view of the Arc plasma torches of PN-type [33]: a PN-1; b PN-2; c PN-6;
d PN-14 m, PN-15, PN-16; e PN-20, PN-21, PN-22; f PN-23; g plasma spraying installation:
I = 100–250 A, U = 80–150 V, P = 10–30 kW, QAr = 18–75 l/min, QH2 O ¼ 1012 l/min,
pH2 O ¼ 6 atm, the efficiently of powder (Ni–Cr–B–Si) with diameter 40 μm up to 5 kg/h;
G plasma gas supply, P entrance of the powder, dn diameter of the nozzle, d, l diameter and length
of the Arc channel

8.3 Simulation of the Arc Plasma Torches in the Channel

8.3.1 Models of the Arc Plasma Torches in the Channel

As it was noted, a feature of Arc plasma torches for cutting and spraying is maintain
of steady burning of plasma flow between the cathode and the anode in a cylindrical
channel, the length of them substantially exceeds its diameter. From all areas of the
300 8 The Arc Plasma Torches

Fig. 8.8 Scheme of the electric Arc in a long channel

electric Arc it usually takes most of the discharge and has important technological
applications (Fig. 8.8). Research of the Arc without near-electrode regions allows
creation of the simple and effective method to determine the parameters on the basis
of the solution of the equation of energy balance, as well as get a picture of plasma
moving and the structure of the electromagnetic field of the Arc in the channel.
Below we consider the model of the channel for the study of the Arc with the
following features:
Plasma stays in the state of local thermodynamic equilibrium (LTE) and is con-
sidered as the continuous current-carrying environments;
The radiation of plasma is optically thin, and may be described in the volume
approximation;
The radius of the channel is much smaller than its length;
The motion of gas flow in the Arc channel is one-dimensional;
The electromagnetic field in the Arc is one-dimensional;
There are no external magnetic fields;
All processes in the Arc channel are stationary.

Model of the Steady Part of the Arc in a Long Channel For a long Arc in the
channel with radius much smaller than its length d  1 the flow can be considered
one-dimensional (vr ¼ vu ¼ 0). Heating of the gas occurs very fast, and constant
along
the channel
 temperature field is set in the Arc, so there is a large established
@T
part @z ¼ 0 , that takes up almost all the space of the Arc, except for the initial
part. At the steady part it can be assumed that the balance of convective flows is
equal to zero and the withdrawal of energy from the Arc is made by thermal
conductivity in the radial direction. In this case, the calculated model of steady-state
part of the long Arc includes a one-dimensional equation of energy balance for
obtaining of temperature and one-dimensional Maxwell’s equations for determi-
nation of electromagnetic field.
The one-dimensional equation of energy balance has the form:
8.3 Simulation of the Arc Plasma Torches in the Channel 301

 
1d dT
rE  Urad
2
¼ rk ð8:1Þ
r dr dr

The one-dimensional intensity of electric field ðEr ¼ Eu ¼ 0; E ¼ Ez Þ, current

density and specific power are calculated as

I
E¼ ; jz ¼ rE; P ¼ rE 2 ð8:2Þ
RR
2p rrdr
0

The one-dimensional intensity of magnetic field ðHz ¼ Hr ¼ 0; H ¼ Hu Þ:

Zr
E
Hu ¼ rrdr ð8:3Þ
r
0

The electric and magnetic fields are one-dimensional and the only one compo-
nent of electromagnetic force, that appears as a result of interaction of the Arc
current with its own magnetic field (Fz = Fφ = 0):

Zr
rE 2
Fr ¼ l0 jz Hu ¼ l0 rrdr ð8:4Þ
r
0

The decision of the equation of energy balance (8.1) is carried out using the
method of the control volume (Sect. 6.1.1). The balance of powers and the discrete
analogue are obtained by integrating of the equation of balance energy in the
control volume (Fig. 8.9).

Zrn   Zrn
1d dT  2 
rk rdr þ rE  Urad rdr ¼ 0 ð8:5Þ
r dr dr
rs rs

Fig. 8.9 Grid and the control volumes for solving of the equation of energy balance for the steady
part of the Arc in a long channel
302 8 The Arc Plasma Torches

After integrating we will have

   
dT dT  
rk  rk þ 0:5 rE 2  Urad M ðrn  rs Þðrn þ rs Þ ¼ 0 ðÞ
dr n dr s

Here, the derivative of temperature in the radial direction on the verge of the
control volume n, s can be approximated as:
   
dT TN  TM dT T M  TS
¼ ; ¼ ðÞ
dr n rN  rM dr s rM  rS

Substituting Eqs. (**) in (*), we obtain the discrete analogue of the equation of
energy balance for the Arc in the channel at steady part for the internal points

aM TM ¼ aN TN þ aS TS þ bM

where
8
< aN ¼ rNrnr
kn
M
[ 0; aS ¼ rMrsr
ks
S
[ 0;
aM ¼ aN þ aS  0:5SpM ðrn  rs Þðrn þ rs Þ [ 0; ð8:6Þ
:
bM ¼ 0:5ScM ðrn  rs Þðrn þ rs Þ; SM ¼ ðrE2  Urad ÞM :

Coefficients SpM ; ScM are determined by (6.56). For efficient use of the boundary

 
d T
conditions on the axis of symmetry dr  ¼ 0 and on the wall (T ¼ TR , where
r¼0
TR —the temperature on the wall) construct the computational grid so that the first
point is located on half a step from the axis of symmetry (rM ¼ Dr=2), and the latter
—on the wall (Fig. 8.9). Taking into account the boundary conditions on the axis of
symmetry for the first control volume the flow due to heat conduction through the
verge s is equal to zero. Consequently, discrete analogue of the equation of energy
balance for the first point has the form (8.6) at as ¼ 0. For the latest point on the
wall, we have T ¼ TR .
We give the algorithm of solving.
1. Build the calculated grid lines (Fig. 8.9) with constant or variable step Dr, and
take the values of convergence criteria ε and internal feedback coefficient β for
the solving of equation of balance energy. For example, for an Arc with
I ¼ 100 A, R ¼ 2 mm, we have Dr ¼ 0:1 mm, e ¼ 103 , b ¼ 0:1.
2. Take some initial field of temperature (providing conductivity) in the channel of
the Arc, for example, for rapid convergence it can be a constant value
T = 10,000 K, except for the boundary condition at the wall TR ¼ 500 K.
3. Calculate the intensity of electric field, specified power using (8.2) and deter-
mine the matrixes for calculation of the coefficients of the discrete analogue for
the equation of balance energy (8.6) for interior and boundary points (Fig. 8.9).
8.3 Simulation of the Arc Plasma Torches in the Channel 303

4. Solve the system of the discrete analogue for the equation of balance energy
(8.6) using iterative or direct sweep method in order to find a new field
temperature.
5. Check the condition of convergence, if it is not fulfilled, repeat the decision
from the 3 before the full convergence.
6. Determine the results of calculation of the field using the Eqs. (8.2–8.4).
The results of calculation and experimental data for the steady part of the Arc in
the channels with diameter d = 4; 10 mm are shown in Fig. 8.10. Although not
taking into account of the removal of energy by the convective flow in this model
some inflated gas temperature is obtained in the area near the wall, but in an Arc
there is a good coincidence of the calculation result with the experimental data that
suggests the applicability of this model for calculation of the steady-state area of the
Arc in the channel.
In addition, this model can be successfully used for determination of temperature
in the zone of conductivity of the initial section (after the near-cathode region) of the
Arc and as an initial approximation at the full simulation of the Arc in the channel.
Model of the Long Arc in the Channel The model of steady part of the Arc (8.1)
allows to quickly get the basic parameters of the Arc in the channel, but does not
allow the determination of the parameters of the initial part, in the flow in the area
near the wall where the heat removal from the depths of the Arc is not only by
conductive, but also the convective mechanisms.
For a long channel of Arc with axial gas supply (Fig. 8.8) the radial and tan-
gential velocities are assumed equal to zero (vr ¼ vu ¼ 0). Specific consumption
through the cross-section of the channel of the Arc is given or is determined by
taking into account a given gas consumption G, kg/s, or Q, m3/s:
qvz ðrÞ ¼ pR
G
2 ¼ pR2 . Note that the specific consumption qvz ðrÞ may not remain
qQ

constant over the section, the Arc always operates as a magnetic pump, and only a
certain amount of gas enters into the discharge for heating and the remaining part
(80–90 %) is moving along the wall of the channel forming the most natural
thermal protection layer. Thus, the channel can be divided into two zones with
different specific consumption: the discharge and the zone near the wall. The exact
distribution qvz ðrÞ can be obtained by the calculation of the field of the gas flow and
solving the momentum and continuity equations of gas.
The assumption of one-dimensional motion of gas in a long narrow channel
allows determination of the axial velocity component of gas from the condition of
the specific consumption through the cross-section.
In contrast to the model of steady part of the Arc, the equation of balance energy
for the model of the long Arc in the channel is:
   
@T @ @T 1@ @T
qcp vz ¼ k þ rk þ rE2  Urad ð8:7Þ
@z @z @z r @r @r

This model includes the equations for determination of the transfer coefficients
and thermodynamic functions of approximation functions of equilibrium plasma or
304 8 The Arc Plasma Torches

Fig. 8.10 Temperature and specified power of Ar-Arc in the channel d = 4 mm (a), and
d = 10 mm (b): lines calculated data, T1, P1—I = 100 A, T2, P2—I = 150 A, T3, P3—I = 200 A;
points—the experiments for d = 4 mm, I = 200 A [34]

approximation equations of their temperature dependences: σ(T), λ(T), cp(T), ρ(T),

Urad(T). Calculation of the parameters of the electromagnetic field is carried out
according to Eqs. (8.2–8.4).
8.3 Simulation of the Arc Plasma Torches in the Channel 305

Such a model of the Arc in the channel is used in many theoretical studies, for
example [9–11].
The boundary condition: in initial (zero) section (z = 0), a one-dimensional
model of energy and current balance can be used (8.1), (8.2) for determination of
the temperature field in the zone of conduction at the zero section, but it is nec-
essary to set or find the radius of the zone of conduction Rk in this section. Radius
Rk can be obtained by assuming that the current flowing through the section of the
Arc I, and the resulting distribution of the intensity of the field along the length of
the Arc has a decreasing character (due to thermal expansion of the Arc in the
channel). Out of the zone of conduction (r > Rk) the temperature of the cold gas is
known: T = 300 K.
It is believed that at the wall (r = R) the gas is in thermal equilibrium with
plasma, the temperature of the wall (taking into account heating of the wall from the
plasma flow due to radiation and heat conductivity) is T = 300–1000 K.
On the axis, we have @T @r ¼ 0 from the condition of the cylindrical axial
symmetry.
At the exit (at the end of a long cylindrical channel), where all active processes
are considered steady, a simple boundary condition @T @z ¼ 0 can be taken, and the
temperature is equal to the temperature upstream [application of the scheme «up-
stream» (Sect. 6.2.1)].
The decision of the equation of balance energy (8.7), together with the calculation
of the intensity of electric field for the Arc in the channel is carried out using the
method of control volume for solving of general differential equation (Sect. 6.1.1).
First, we transform the equation of balance energy (8.7) in the general differential
equation for the control volume. Because of the smallness of the control volume it
can be assumed that the specific heat of plasma is constant (cp ¼ const) inside each
of the control volume. The law of mass conservation for the one-dimensional plasma
flow in the channel has the form @ðqv zÞ
@z ¼ 0, and qvz ¼ const along the z axis. Thus,
the equation of energy balance (8.7) can be written as:
   
@ðqvz TÞ @ k @T 1@ k @T 1 2 
¼ þ r þ rE  Urad ð8:8Þ
@z @z cp @z r @r cp @r cp

The procedure for integrating of the equation of balance energy (8.8) is similar to
the procedure for the integration of general differential Eq. (6.2) (Sect. 6.1.1), and
U ¼ T; C ¼ ckp ; S ¼ c1p ðrE2  Urad Þ ¼ Sc þ Sp T, qvr ¼ 0. The discrete analogue of
the equation of energy balance (8.8) for internal points (Fig. 6.5a) is:

aM TM ¼ aE TE þ aW TW þ aN TN þ aS TS þ bM
306 8 The Arc Plasma Torches

where
8
> aE ¼ De AðPee Þ ¼ De AðjPee jÞ þ maxðFe ; 0Þ  0;
>
>
> aW ¼ Dw AðPew Þ þ Fw ¼ Dw AðjPew jÞ þ maxðFw ; 0Þ  0;
<
aN ¼ Dn AðPen Þ ¼ Dn  0; aS ¼ Ds AðPes Þ þ Fs ¼ Ds  0; ð8:9Þ
>
>
>
> a M ¼ aE þ aW þ aN þ aS  SpM DV [ 0; bM ¼ ScM DV;
:
DV ¼ ðze  zw Þ ðrn rs Þ2ðrn þ rs Þ :

Function AðPeÞ
h that characterize
i the computational scheme (Sect. 6.1.1) is
5
AðPeÞ ¼ max 0; ð1  0:1PeÞ . The coefficients SpM ; ScM are determined using
(6.56).
 
 SM 
SpM ¼  ; ScM ¼ SM  SpM TM ð8:10Þ
bT TM 
h i
Here SM ¼ 1
cp ðrE 2  Urad Þ ; bT —coefficient of internal feedback (Sect. 6.3.2).
M
For simulation of the long Arc in the channel, we use the similarity of boundary
conditions of Arc and RF-discharge in the channel and we use the higher developed
(Sect. 6.1.2) structure of the shifted chess grids (Fig. 6.5a). And the discrete ana-
logue for the boundary points are obtained in the same way and have the same form
as in the case of RF-plasma torch (Sect. 7.3.1).
Algorithm for solving of the equation of balance energy for this model is similar
to the above algorithm for solving of the model of established part of the Arc. Note
some features for modeling of the long Arc in the channel:
1. First, for the solution the initial profile of the temperature field in the channel of
the Arc should be specified. The intensity of electric field is calculated using the
Eq. (8.2) and the specified power rE 2 are sensitive for the temperature distri-
bution, and the stability of solution of the problem depends on a predetermined
initial temperature field in the Arc channel and on the selection of the coefficient
of internal feedback bT . The one-dimensional model of the steady part of the
Arc can be used again for obtaining of the initial field of the temperature [(8.1),
(8.2)]. This method usually gives the initial field of temperature sufficiently
close to the solution of the equation of balance energy (8.7), because it satisfies
to the one-dimensional balance of energy and of the Arc current. Using the
obtained initial temperature field can ensure a smooth change of the solution and
fast convergence of the problem.
2. The coefficient of internal feedback bT is determined based on the following
reasoning: if the solution is carried out using a coarse mesh or using the initial
temperature field that is significantly different from the result of the final
decision the lower value of the coefficient bT to provide more effect of feedback
should be taken into account. For example, for argon Arc in the channel with
diameter of 6 mm and a length of 140 mm with the current of 150 A in the
8.3 Simulation of the Arc Plasma Torches in the Channel 307

selection of the grid with the steps Δz = Δr = 0.05–0.1 mm at the initial part of
the Arc and Δz = 0.1–5 mm, Δr = 0.1–0.2 mm in the other zone of the Arc for
sustainable solution the coefficient βT = 0.1–0.5 is sufficient. However, for the
same Arc in the selection of the grid with steps Δz = Δr = 0.2–0.5 mm at the
initial part of the Arc and Δz = 0.5–10 mm, Δr = 0.2–0.5 mm in the other zone
of the Arc the value of coefficient bT \ 0:01 for obtaining of the sustainable
solutions. The number of iterations is also different: for the same conditions the
lower values of bT are able to provide the greater stability of the computational
process, but obtaining of the convergence of solutions will be slower, i.e., the
number of iterations will be greater, respectively.

8.3.2 Analysis of the Simulation Results

The results of modeling of different Arcs are different to each other in the tech-
nological purpose, structure, and the regime of operation. The input data of several
types of Arcs in the channel are presented in Table 8.3. They are selected in
accordance with the characteristic data of the serial industrial models of the Arc
plasma torches RPM-6 [21], PV-25 [35], PN-6, PRD-9 [36]. These and other types
of Arcs in the channel in practice are often used for cutting, plasma spraying,
spheronization of the particles, as well as for plasma-chemical reactions and
treatment of the surfaces of materials. Most of these technological processes occurs
out of the nozzle of the plasma torch, in the plasma jet, the definition of the Arc
parameters, including temperature and velocity at the exit of plasma torches, is of
great practical importance.

Table 8.3 The input data, parameters of the regime and the technological purpose of different
Arcs in the channel
Diameter of Arc Consumption and Length of Technological purpose
the channel d, current type of gas Q, the Arc l, mm
mm I, A l/min
3 50–200 7.5–56, air, N2, 50 Cutting
Ar
4 50–250 7.5–56, air, N2, 50 Cutting
Ar
6 50–300 7.5, 15, 25, 46, 140 Spray, spheroidization
Ar
10 50–300 7.5, 15, 25, 46, 140 Spray, spheroidization,
Ar treatment of the surfaces,
implementation of
plasma-chemical reactions
20 80–160 11, 22, 45, Ar 45 Plasma-chemical reactions,
spheroidization, spray,
treatment of the surfaces
308 8 The Arc Plasma Torches

The Arc for the cutting of metal is usually characterized by the small diameter of
the channel (0.4 ≤ d ≤ 4 mm), high-current compressed flow of plasma with a high
velocity of the flow (up to several kilometers per second), high temperature and the
specific power of the Joule heating. These conditions are necessary for the effective
destruction of the metal and at the same time making of the smallest seams and the
high speed of cutting.
Arcs for the plasma spraying and for the plasma-chemical reactions and, in
contrast, have a larger diameter of the channel (d > 5 mm) for creation of the
plasma flow with a large cross-section, and favorable conditions for its interaction
with material. The processed is often entered into the plasma jet through the exit of
the channel of the plasma torch, so it does not disturb the flow of plasma and does
not pollute the electrodes (Figs. 8.6, 8.7).
The Intensity of Electric Field in the Arcs The intensity of electric field char-
acterizes the local allocation of energy along the length of the Arc. In the electric
Arcs with the longitudinal stabilization, the intensity of electric field has the
maximum value close to the input electrode and is reduced to a certain limit along
the length of the Arc (Figs. 8.11, 8.12, 8.13). This describes the output of the Arc to
the established zone, where the parameters of plasma remains practically constant
along the length of the Arc. In general, a large number of parameters affects to the
intensity of electric field, however, there are two main ones: the Arc current I and
the diameter of the channel d. The data on the changes of the intensity of electric
field depending on these variables are presented in Figs. 8.11, 8.12 and 8.13.
For comparison of the results obtained in Fig. 8.14 the experimental data [36]
are presented, that have similar conditions with the calculations. These data cor-
respond to the electric Arc that is burning in a cylindrical Arc channel with

Fig. 8.11 The intensity of electric field along the length of the Arc in the channel d = 3 mm
(a) and d = 4 mm (b): E1, E5 intensities of electric field at I = 100 A; E2, E6 the same at I = 150
A; E3, E7 the same at I = 200 A; E4, E8 the same at I = 300 A; gas consumption QAr = 37.5 l/min
8.3 Simulation of the Arc Plasma Torches in the Channel 309

Fig. 8.12 The intensity of the electric field along the length of the Arc in the channel with
d = 6 mm (a) and d = 10 mm (b): notation is the same as in Fig. 8.11

Fig. 8.13 The intensity of electric field along the length of the Arc in the channel d = 6 mm
(a) and d = 10 mm (b) at I = 150 A: E1, E5 intensities of electric field at QAr = 7.5 l/min; E2, E6
the same at QAr = 15 l/min; E3, E7 the same at QAr = 25 l/min; E4, E8 the same at QAr = 46 l/min

sectioned inter-electrode insert with axial feed of argon. The intensity of electric
field was determined by differentiating of the potential distributions along the length
of the Arc, measured by the sections of the Arc channel relative to the cathode. The
difference between the theoretical (Figs. 8.12, 8.13) and the experimental data
(Fig. 8.14) is 5–15 %.
A constant value of the intensity of electric field in the steady zone (Figs. 8.11,
8.12, 8.13, 8.14) shows the linear dependence of the potential along the length of
the Arc, and confirms that the parameters of the Arc in the channel remains
310 8 The Arc Plasma Torches

Fig. 8.14 The experimental data of the intensity of the electric field along the length of the Arc
[36] at a flow rate of Argon of 25 l/min (a–c) and a current of 150 A (d–f): a, d d = 6 mm; b,
e d = 10 mm; c, f d = 12 mm; 1 I = 100 A; 2 I = 150 A; 3 I = 200 A; 4 I = 300 A; 5
QAr = 9.4 l/min; 6 QAr = 15 l/min; 7 QAr = 46 l/min

practically constant along its length. These results justify the use of the above
described model for calculation of the Arc at the steady zone.
For the Arc in the channel the intensity of electric field at the steady area mainly
depends on the current of the Arc and to a lesser degree depends on the gas
consumption, the similar result was obtained in [36]. The data on the intensity of
the electric field at the steady zone in dependence on the Arc current channels of
various diameters are presented in Fig. 8.15. The results of calculation differ from
the results of experiments, not more than 10 % that corresponds to the error of
experimental methods for determination of E.
8.3 Simulation of the Arc Plasma Torches in the Channel 311

Fig. 8.15 The intensity of

electric field on the steady
part of the Arc: the gas
consumption
QAr = 37.5 l/min. lines
without points calculations;
lines with points experimental
data [37–40]

Note that the value of the intensity of electric field in the initial section depends
on the set temperature distribution in the initial section, the radius of the conduction
zone and the current of the Arc. Using the one-dimensional model (8.1), (8.2), the
temperature field and the radius of the conductive zone in the initial section taking
into account the falling characteristics of the intensity of electric field and the
temperature along the length of the Arc channel (with the expansion of the Arc) can
be determined. This technique allows the simple definition the boundary condition
in the initial section of the Arc and to eliminate some of the errors associated with
the assumption of one-dimensional motion of gas and fields. It was used for
selection of the temperature field in the first cross-sectional area of the cathode.
The Current–Voltage Characteristics of the Arc The most important charac-
teristic is the drop of the voltage in the Arc, depending on the current of the Arc.
Taking into account the voltage drop in the electrode areas of these features allow to
define a coherent operation of the arc plasma torch with a power source and set the
range of stable operation of the generator in the regulation of its parameters. The
lack of data on the local parameters of the Arc leads to the necessary of the use of
published data on the approximate current–voltage characteristics of different Arcs
depending on the current of the Arc, the diameter of the channel and the con-
sumption of the gas [8, 36, 41].
Simulation of the Arc allows obtaining of the local distributions of the param-
eters and to establish their relationship with the external parameters of the plasma
torch using the current–voltage characteristic. For the long Arc in the channel
voltage drop in the Arc can be greater than in the near-electrode regions. In the
model of the long Arc in the channel the simultaneous solving of equation of the
312 8 The Arc Plasma Torches

Fig. 8.16 Current–voltage

characteristics of the Arc
column in the channels of
different diameters: U1, U2—
d = 3; 4 mm, l = 50 mm; U3,
U4—d = 6; 10 mm,
l = 140 mm; the gas
consumption
QAr = 37.5 l/min

energy balance of the Arc and the intensity of the electric field allows the deter-
Rl
mination of the voltage drop in the Arc using the equation U ¼ EðzÞdz—where
0
l—the length of the Arc. The current–voltage characteristics of the Arc produced in
channels of different diameters are presented in Fig. 8.16.
These obtained results indicate that for the Arcs in the channel of the small
diameter (d < 4 mm) at the currents I < 100 A the current–voltage characteristic
can have a small decreasing zone. In the other area of the currents (I > 100 A), and
for the Arcs of large diameter (d > 4 mm) the current–voltage characteristic has the
increasing form (Fig. 8.16). This can be explained using the nonlinear relationship
between the voltage and the current of the Arc:

Zl
I dz
U¼ :
2p RR
0 rðTÞrdr
0

In the zone of low current and at its increase the Arc is rapidly expanding due to
the sharp increase in the size of the conductive zone, leading to a possible reduction
of the voltage drop in the Arc. However, from a certain value of the current of the
Arc due to the presence of the wall of plasma torch, as well as a streamlined flow of
gas and power losses due to the radiation and heat conduction in the radial direc-
tion, the growth of the conductive area of the Arc is constrained. So, the current–
voltage characteristic of the Arc has a growing character.
The Current Density and Specified Power in the Arc Column For practical, the
most important are the parameters of the Arc in the steady zone, because the
technological process of the Arc are implemented using these parameters. Thus, the
thermal effect of the Arc to the treated materials is determined by the heat flux and
momentum, i.e., by temperature and velocity of the flow of plasma at the exit. The
results of calculation of the main parameters at the exits of different channels of the
8.3 Simulation of the Arc Plasma Torches in the Channel 313

Arcs: distribution of current densities and specified powers are presented in

Fig. 8.17. Total power from the current of the Arc is calculated based on the
received local parameters and presented in Fig. 8.18. After reviewing the formula
P = UI, we have a good agreement with the data of Fig. 8.18, indicating that the
main Arc voltage drop in its long column and accuracy of simulation results.
The Fields of Temperatures and Velocities of Plasma Very pressed Arcs can
have the best thermal and dynamic effects on the materials. So, for the cutting of
metals due to the narrow channel of the Arc the value of current density reaches
2000–6000 A/cm2, and the value of specific power 40–300 kW/cm3 (Fig. 8.17a, b).
Such values of these parameters provide the high temperature and the velocity of the
gas flow at the exit of the Arc channel. The temperature of plasma at the exit of the
channel of the Arc for the cutting of metals is (20–40) × 103 K, and the velocity of
plasma can reach 1300–6000 m/s, that creates very favorable conditions for the
cutting of metal (Fig. 8.19a, b).
At the increase in diameter of the Arc channel to 6 mm and 10 mm, there is a
sharp decrease of the current density (400–1800 A/cm2) and a specific power of the
Arc (3–28 kW/cm3) (Fig. 8.17c, d), that leads to a rare reduction of the temperature
(12–24) × 103 K and the velocity of plasma flow (60–800 m/s) at the exit of the
channel (Fig. 8.19b, c). However, a large spatial extent and high-enthalpy of
plasma flow with such characteristics at the exit of the channel create more
favorable conditions for such technological processes such as spheroidization of
refractory particles, spraying and treatment of the surfaces, as well as the suitability
of the environment for the implementation of various plasma-chemical reactions.
The temperature and velocity of plasma at the exit of the channel depending on
the Arc currents and the diameter of channel are presented in Fig. 8.19, and the
changes of temperature at the axis along the length of the Arc are shown in Fig. 8.20.
The maximum temperature of the Arc is achieved on the axis in the initial section
and at the move away from it rapidly decreases to the value of temperature at the
steady zone of the Arc. From these data, it is also possible to determine the length of
the initial, transient and steady zones of the Arc. Comparing the data of the tem-
perature field and the specified powers of two models (Figs. 8.10, 8.17, 8.19) and
comparing them with results of some experiments, we see a good their match. The
fields of temperature and specified power obtained using the model of steady zone of
the Arc are more flat, and they are wider at a lower maximum values at the axis
compared with the fields obtained using the model of the long Arc in the channel.
This not so big difference is explained by the relatively small role of convection and
thermal conductivity in the axial direction in the energy balance of the electric Arc.
Thus, on the basis of the model of equilibrium plasma of the Arc in the channel
model the main parameters of a number of electric Arcs used in practice are
obtained: they are used for cutting, spraying, in plasma chemistry and for
spheroidization of particles. Using these data we can determine the appropriate
flows of the heat at the exit, the power losses due to radiation and heat conductivity
to the wall of plasma torch, the power carried away by the flow of plasma to the jet,
the efficiency of the installation and effectiveness of the technological process.
314 8 The Arc Plasma Torches

Fig. 8.17 Distribution of

current density and specified
power along the radius of the
Arc at the steady zone of the
Arc in the channels of
different diameters:
a d = 3 mm; b d = 4 mm;
c d = 6 mm; d d = 10 mm;
j1, j5, P1, P5 the current
densities and specific powers
at I = 100 A; j2, j6, P2, P6 the
same at I = 150 A; j3, j7, P3,
P7 the same at I = 200 A; j4,
j8, P4, P8 the same at I = 300
A; gas consumption
QAr = 37.5 l/min
8.3 Simulation of the Arc Plasma Torches in the Channel 315

Fig. 8.18 Dependence of the

total power from the current
of the Arc in the channel with
diameters d = 3; 4 mm
(l = 50 mm, P1 and P2) and
d = 6; 10 mm (l = 140 mm,
P3 and P4): points—the
experimental data [35]: filled
square d = 3 mm; filled circle
d = 4 mm; gas consumption
QAr = 37.5 l/min

Using these values the connection can also establish between such local parameters
of the electric Arc as the temperature field and velocity field, intensity of the electric
field and current density with external parameters of the Arc, such as the current and
the diameter of the Arc in the channel, consumption of the plasma-forming gas, the
length, and current–voltage characteristic of the Arc and others. The data about
these links are extremely important in the choice of a coherent and stable operation
of the Arc with the power supply, as well as for solving of the problem of direct
control of the electric Arc.

8.4 Two-Dimensional Electromagnetic Problem

for the Arc Plasma Torches

Calculation of the electromagnetic field in the Arc plasma torches is often carried
out using the one-dimensional model with assumption that in the Arc there is only
one component of the intensity of electric field Ez (Er = Eφ = 0) and one compo-
nent of the intensity of magnetic field Hφ (Hz = Hr = 0). Physically, this is possible
only for the Arc with the cross-section constant along the length of the flow of
current, or in the long compressed Arc in the channel.
In many cases, this assumption of one-dimensional electric field is not entirely
justified, especially for the free-stabilized Arcs, uncompressed in the channel, for
short and curved Arcs, for the Arcs with high power and active expansion of the
conductive zone. It cannot be applied even for Arcs in the channel at the calculation
of the initial and transition zones and the zone of attachment of the Arc to electrodes.
In particular, the change of the conductive zone of the electric Arc can lead to a
second component of the intensity of electric field Er, especially in the zones where
the Arc is actively growing or shrinking. For example, in the zones of the cathode
and anode anchor in the initial and transition zones of the Arc, very often the radial
component of the intensity of electric field can be comparable with axial (Er > Ez)
316 8 The Arc Plasma Torches

Fig. 8.19 The distribution of

the temperature and velocities
of plasma at the steady zone
of the Arc in the channel: a—
d = 3 mm; b—d = 4 mm;
c d = 6 mm; d—d = 10 mm;
T1, T5, v1, v5 the temperature
and velocities at I = 100 A;
T2, T6, v2, v6 the same at
I = 150 A; T3, T7, v3, v7 the
same at I = 200 A; T4, T8, v4,
v8 the same at I = 300 A;
filled square experimental
data [34], d = 4 mm,
I = 200 A; gas consumption
QAr = 37.5 l/min
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 317

Fig. 8.20 Axial temperature distribution along the length of the plasma channel: a—d = 3 mm
(T01, T02, T03, T04), d = 4 mm (T05, T06, T07, T08); b—d = 6 mm (T01, T02, T03, T04), d = 10 mm
(T05, T06, T07, T08); T01, T05—I = 100 A; T02, T06—I = 150 A; T03, T07—I = 200 A; T04, T08—
I = 300 A; gas consumption QAr = 37.5 l/min

in its value. Moreover, in these zones, the most active processes of the electric Arc
very often occur and their characteristics are largely determined by these processes.
All of this suggests the need for the calculation of a two-dimensional electro-
magnetic field at the simulation of electric Arcs.
Two-dimensional electromagnetic problem presupposes the existence of two
components of electric Arc Ez, Er (Eφ = 0) and one component of the magnetic field
Hφ (Hz = Hr = 0). For its productions we can use the Maxwell’s system of equa-
tions for description of the field in the form of the following two equations:
318 8 The Arc Plasma Torches

!
rotH ¼ ~j; ð8:11Þ
!
! @B
rot E ¼  ; ð8:11aÞ
@t
!
Here ~j ¼ r E —the vector of the current density of conduction in the Arc,
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
jz ¼ rEz , jr ¼ rEr —the components of the current density; E ¼ Er2 þ Ez2 —
module of the intensity of electric field. For a DC Arc the variation of the vector of
!
magnetic induction in time is equal to zero, @ B ¼ 0 and the Eq. (8.11a) has the
@t
form:
!
rot E ¼ 0: ð8:12Þ

In a cylindrical coordinate system, the two-dimensional Eqs. (8.11) (8.12) have

the following form:

1@  
rHu ¼ rEz ; ð8:13Þ
r @r
@Hu
 ¼ rEr ; ð8:14Þ
@z
@Er @Ez
 ¼ 0: ð8:15Þ
@z @r

For solving of the system of Eqs. (8.13–8.15), there are two effective methods—
using the electric potential V and using the function of electric current Ф. The main
advantage of these methods is the reduction of the number of equations to a single
equation with for V and Ф that greatly simplifies the solution of the problem.
Consider the features of these two methods.

8.4.1 Solution Using the Electric Potential

To reduce the number of Eqs. (8.13–8.15) and bringing them to a single equation the
electric potential V can be used, that is connected with the intensity of electric field:
!
E ¼ gradV. This is possible because for every scalar potential V from the
Eq. (8.12) the equality rotðgradVÞ ¼ 0 is always satisfied. Consequently, the com-
ponents of the intensity of electric and magnetic fields can be expressed using V:
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 319

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2  2ffi
@V @V @V @V
Ez ¼  ; Er ¼  ; E¼ þ ; ð8:16Þ
@z @r @z @r

Zr
1 @V
Hu ¼  rr dr: ð8:16aÞ
r @z
0

And the components of current density, expressed using V have the form:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
   ffi
@V @V @V 2 @V 2
jz ¼ r ; jr ¼ r ; j¼ r þ r : ð8:17Þ
@z @r @z @r

! !
On the other hand, for any continuous vector H we always have divðrot H Þ ¼ 0,
therefore, from the Eq. (8.11) a well-known law of continuity in the current form
div~j ¼ 0 can be obtained. This equation is written in the two-dimensional cylin-
drical coordinate system:

@jz 1 @
þ ðrjr Þ ¼ 0: ð8:18Þ
@z r @r

Substituting (8.17) into (8.18), we obtain a single equation describing the field
using the electric potential V:
   
@ @V 1@ @V
r þ rr ¼ 0: ð8:19Þ
@z @z r @r @r

The solution of (8.19) for the electrical potential is carried out using the method
of control volume (Sect. 6.1.1). The differential Eq. (8.19) expresses the law of the
field conservation in electric Arc for the infinitely small volume, so, as it is shown
in (Sect. 6.1.1) to carry out the law of conservation of the field in the finite control
volume Eq. (8.19) should be integrated over the control volume (Fig. 8.21).

Zrn Zze   Zze Zrn  

@ @V 1@ @V
r dzrdr þ rr rdrdz ¼ 0: ð8:20Þ
@z @z r @r @r
rs zw zw rs

Under the integral of each member of Eq. (8.20) there is the first derivative of
the same order, so we can accurately produce the integration of (8.20) at least once,
and in the end we get:
320 8 The Arc Plasma Torches

Fig. 8.21 Control volume

and location of the grid points
[z1, r1] (Fig. 8.22) for the
solution of two-dimensional
problem of electromagnetic
Arc

Zrn   Zrn   Zze   Zze  

@V @V @V @V
r rdr  r rdr þ rr dz  rr dz ¼ 0:
@z e @z w @r n @r s
rs rs zw zw

ð8:21Þ

Note that the integration is carried out along each verge of the control volume
(Fig. 8.21). For the further integration of (8.21) assume that integrands take its
averages in the points e, w, n, s, and the first derivatives are approximated using the
finite differences:
8   
>
> @V VE  VM @V VM  VW
< r @z ¼ re zE  zM ; r @z ¼ rw zM  zW ;
>
e w
    ð8:22Þ
>
> @V V  V @V VM  VS
>
: r ¼ rn
N M
; r ¼ rs :
@r n rN  rM @r s rM  rS

Substituting Eqs. (8.22) into (8.21), we obtain the discrete analogue for
Eq. (8.19) for the interior points:

aM VM ¼ aE VE þ aW VW þ aN VN þ aS VS ;

where
8 re ðrn rs Þðrn þ rs Þ rw ðrn rs Þðrn þ rs Þ
< aE ¼ 2ðzW zM Þ [ 0; aW ¼ 2ðzM zW Þ
> [ 0;
rn rn ðze zw Þ rs rs ðze zw Þ
a ¼ rN rM [ 0; aS ¼ rM rS [ 0; ð8:23Þ
>
: N
aM ¼ aE þ aW þ aN þ aS [ 0:
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 321

Note that the coefficients of the discrete analogue (8.23) are positive and satisfy
to the criterion of stability (6.49).
Example of the mesh for solving the two-dimensional electromagnetic problem
of open Arc is shown in Fig. 8.22.
The Boundary Condition of the Equation for Electric Potential We believe that
the electric potential of anode is zero (Va = 0). Then the potential of all other points
of the electric Arc is the relative potential compared with anode.
On the axis of the symmetry we have that the radial component of the current
density is equal to zero using the condition
 of the cylindrical axial symmetry.
Taking into account (8.17) we have: @V  ¼ 0. At the integration of the equation
@r r¼0
for electric potential (8.19) in the half of the control volume 1 (Fig. 8.22), this
boundary condition leads to the fact that the coefficient of the discrete analogue
(8.23) aS ¼ 0 (taking into account the boundary condition for V on the axis, the
Eq. (8.19) can be solved using the displaced grid [z1, r2] (Fig. 6.5b)]. The discrete
analogue for the points on the axis of symmetry is:

aM VM ¼ aE VE þ aW VW þ aN VN ;

where

aN ¼ rn rrnNðr
ze zw Þ
r r2 r r2
aE ¼ 2ðzEezn M Þ [ 0; aW ¼ 2ðzMwzn W Þ [ 0; M
[ 0; ð8:24Þ
aM ¼ aE þ aW þ aN [ 0:

On the wall (r = R) for the Arc in the channel or on the free boundary
 of the
open Arc that is located well away from the Arc it can be put that @V  ¼ 0 or
@r r¼R
Vjr¼1 ¼ 0 for open Arc.
At the cathode, the potential V should be negative and can be defined using the
predetermined value of the Arc current I. If we take into account that the constant
and one-dimensional intensity of the electric field Ez at the cathode can be deter-
mined using the predetermined Arc current, then using (8.16) for V we have the
following boundary condition on the cathode:

@V  I
¼ Ez jz¼0 ¼ ; ð8:25Þ
@z z¼0 RRc
2p rrdr
0

where Rc—the radius of the cathode spot.

322 8 The Arc Plasma Torches

Fig. 8.22 The grid and the control volume for solving the two-dimensional electromagnetic
problem for the open Arc

8.4.2 Solution Using the Function of the Electric Current

At the solving of the two-dimensional problem using the electrical potential the
boundary condition for the gradient of electric potential at the cathode (8.25) is not
very convenient for obtaining of the appropriate discrete analogue. Usually, the
approximation of the gradient V leads to an additional term in the discrete analogue
for points on the cathode that is not desirable. On the other hand, the approach
approximation of the gradient @V@z in a narrow zone of the cathode spot with a high
temperature gradient can lead to an inaccurate definition of V on the cathode border
and affect to the result of calculation of the field in the Arc zone. In this regard,
consider the second method for solving of the two-dimensional system of elec-
tromagnetic Eqs. (8.13–8.15) using the function of electric current Φ, in this case
the boundary conditions is easily implemented and does not affect to the accuracy
of the solution.
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 323

The function of an electric current express the law of total current for the
cross-section of the Arc and can be used for the reduction of the number of
equations that are necessary for solving of the two-dimensional electromagnetic
problem of the electric Arc. The function Φ is also an excellent feature for the
analysis of the field of the current flow in the electric Arc and one of the main
parameters in the study and design of Arc plasma torches. In a cylindrical coor-
dinate system, the function of the electric current Φ can be expressed using the
intensity of the magnetic field Hφ:

U ¼ rHu : ð8:26Þ

Accordingly, the component of the intensity of the magnetic field can be

determined using the function Φ:

Hu ¼ U=r: ð8:27Þ

Substituting (8.27) into (8.13) and into (8.14), we obtain new equations for
calculation of the components of the current density, expressed using the function
Φ:

1 @U 1 @U
jz ¼ rEz ¼ ; jr ¼ rEr ¼  : ð8:28Þ
r @r r @z

In this case it is possible to express the components of the electric field Ez and Er
using the function Φ:

1 @U 1 @U
Ez ¼ ; Er ¼  : ð8:29Þ
rr @r rr @z

Substituting (8.29) into (8.15), we obtain a new equation for description of the
field using the function of the electric current Φ, that is symmetric for the
derivatives for z and r.
   
@ 1 @U @ 1 @U
þ ¼0 ð8:30Þ
@z rr @z @r rr @r

Thus, using function of electric current Φ we can also reduce the system of field
Eqs. (8.13–8.15) to a single equation.
Discrete Analogue for the Function of the Electric Current The numerical solu-
tion of the Eq. (8.30) can also be produced using the method of the control volume
(Sect. 6.1.1). For maximize effective of the boundary conditions the Eq. (8.30) should
be solved using the grid [z1, r1] (Fig. 8.22), border points of it are located directly on
the cathode (z = 0), at the axis of symmetry (r = 0), on the wall or at the free boundary
(r = R).
324 8 The Arc Plasma Torches

To implement the law of the conservation of the field in the finite control volume
it is necessary to integrate Eq. (8.30) over the control volume (Fig. 8.21).

Zrn Zze   Zze Zrn  

@ 1 @U @ 1 @U
dzrdr þ rdrdz ¼ 0:
@z rr @z @r rr @r
rs zw zw rs

We obtain the following equation after the first integrating:

Zrn   Zrn   Zze  

1 @U 1 @U 1 @U rn þ rs
dr  dr þ dz
r @z e r @z w rr @r n 2
rs rs zw
Zze  
1 @U rn þ rs
 dz ¼ 0:
rr @r s 2
zw

For the further solving the following finite-difference equations of approxima-

tion are used:
   
@U UE  UM @U UM  UW @U UN  UM @U UM  US
¼ ; ¼ ; ¼ ; ¼ :
@z e zE  zM @z w zM  zW @r n rN  rM @r s rM  rS

Substituting these formulas into the last equation, we get the discrete analogue
for Φ in the form:

aM UM ¼ aE UE þ aW UW þ aN UN þ aS US ;

where
8 rn rs rn rs
< aE ¼ re ðzE zM Þ [ 0; aW ¼ rw ðzM zW Þ [ 0;
ðrn þ rs Þðze zw Þ ðrn þ rs Þðze zw Þ ð8:31Þ
: aN ¼ 2rn rn ðrN rM Þ [ 0; aS ¼ 2rs rs ðrM rS Þ [ 0;
aM ¼ aE þ aW þ aN þ aS [ 0:

All the coefficients of the discrete analogue of (8.31) are positive and satisfy the
stability criterion (6.49).
The Boundary Conditions of the Equation for the Functions of Electric
Current On the axis of the plasma torch Φ|r=0 = 0 (8.26).
For the cathode the existence of only one component of the electric and magnetic
fields Ez ðrÞ ¼ const and Hφ, we have from (8.28):
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 325

Rr
Zr I rrdr
0
UðrÞjz¼0 ¼ rEz rdr ¼ ;
RRc
0 2p rrdr
0

Here, Rc—the radius of the conductive zone of the Arc on the cathode (the radius
of the cathode spot).
On the wall of the Arc channel or at the free boundary of open Arc (r = R),
where there is no the conduction current, and from (8.28) we have
RR
UðR; zÞ ¼ rEz rdr ¼ 2pI
¼ const. This function of the electric current should be
0
obtained out of the conductive zone of the Arc, and it is an excellent criterion for
testing of the accuracy of field calculation.
We believe that at the anode there is only one component of the intensity of the
@U
electric field Ez ðrÞ ¼ const, Er ¼ 0. From (8.29) we obtain @z z¼L ¼ 0. The discrete
analogue for the points on the anode is obtained by integrating Eq. (8.30) for the
half of the control volume 2 (Fig. 8.22).
Taking into account the boundary conditions at the anode, the discrete analogue
for these points is:

aM UM ¼ aW UW þ aN UN þ aS US ;

where
(
aW ¼ rw ðrznMrzs W Þ [ 0; aN ¼ ðr2r
n þ rs ÞðzM zw Þ
n rn ðrN rM Þ
[ 0;
ð8:32Þ
aS ¼ ðr2r
n þ rs ÞðzM zw Þ
s rs ðrM rS Þ
[ 0; aM ¼ aW þ aN þ aS [ 0:

Equations for Calculation of the Two-Dimensional Field of the Electric Arc

After solving of the equation for the electric potential V (8.19) or the function of
electric current Φ (8.30), the obtained field of V or Φ can be used for determination
of the parameters of the field of the electric Arc and their distribution using the
following equations:
1. The components of the intensity of the electric field are determined using
Eqs. (8.16) or (8.29).
2. The intensity of the magnetic field is determined using Eqs. (8.16a) or (8.27).
3. The components of the current density are determined using Eqs. (8.17) or
(8.28).
4. Specific allocated power of the Arc is determined as:
326 8 The Arc Plasma Torches

" 2  2 # "   2 #
@V @V 1 @U 2 @U
Pðr; zÞ ¼ rE ¼ r
2
þ ¼ 2 þ : ð8:33Þ
@z @r rr @r @z

5. The components of the electromagnetic force are determined as:

Rr 9
Fz ðr; zÞ ¼ l0 rEr Hu ¼ l0 rr @V rr @V U @U >
>
@r @z dr ¼ l0 r2 @z ; =
0
Rr ð8:34Þ
U @U > >
Fr ðr; zÞ ¼ l0 rEz Hu ¼ l0 rr @V
@z rr @V
@z dr ¼ l0 r2 @r : ;
0

6. The total power allocated in the Arc is determined as:

Z l ZR Z l ZR " 2  2 #
@V @V
P ¼ 2p rE rdrdz ¼ 2p
2
r þ rdrdz
@z @r
0 0 0 0
" ð8:35Þ
Z l ZR 2  2 #
1 @U @U
¼ 2p þ rdrdz;
rr 2 @r @z
0 0

Here l and R—length and radius of the Arc.

7. Strictly speaking, from the formulation of the two-dimensional problem of
electromagnetic Arc the voltage drop at the Arc has the distribution along the
radius of the Arc. However, as in the one-dimensional problem we take the total
value of the voltage drop as a voltage drop on the axis of the Arc. It is defined as:

Zl Zl  
1 @U 
U¼ Ez ð0; zÞdz ¼ dz: ð8:36Þ
rr @r r¼0
0 0

By solving Eqs. (8.13–8.15) for the electric potential V the absolute value in the
center of the cathode spot is a voltage drop in the Arc (the electrical potential at
the anode is taken equal to 0).
8. On the axis of the Arc plasma torch from the condition of the cylindrical axial
symmetry we have the following relations: Er ¼ 0, jr ¼ rEr ¼ 0,
@U @U
@z ¼ @r ¼ Hu ¼ 0, U ¼ 0, Fz ¼ Fr ¼ 0.
9. From Eq. (8.15) we have @E @Er
@r ¼ @z . Out of the conductive zone of the Arc we
z

have: Er ¼ Ez ¼ 0, U ¼ 2p
I
, Fz ¼ Fr ¼ jz ¼ jr ¼ P ¼ 0.
Thus, using the electric potential V or the function of the electric current Φ we
are able to reduce the number of equations to be solved to one equation for V or Φ.
As a result of the mathematical decision appropriate discrete analogue is obtained in
the form (8.23) or (8.31). In each case the boundary conditions and the corre-
sponding discrete analogues for points at the borders are considered. We can also
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 327

obtain the necessary algorithms for determination of the two-dimensional electro-

magnetic field of the electric Arc. As an example, we use the proposed method-
ology for calculating of two-dimensional electromagnetic field of open Arc.

8.4.3 Calculation of the Electromagnetic Field of the Arc

Plasma Torches

The electromagnetic field of the Arc plasma torches depends on many factors, but it
largely depends on the current, diameter and the length of the channel of discharge,
type and regime of the gas flow and other conditions. For the calculation of elec-
tromagnetic field of the Arc it is necessary to know the temperature field. The
temperature field is determined from the solving of the equation of the balance
energy that in turn can be solved if the electromagnetic field, the velocity field of
the Arc and the pressure field the composition of the plasma are known for the
determination of the properties of plasma, etc. Thus, the calculation of the elec-
tromagnetic field of the Arc should be carried out with the calculation of the
temperature field, velocity, pressure, composition, transport coefficients and ther-
modynamic functions of plasma, etc., i.e., the calculation of the electromagnetic
field is the part of a complex the task of mathematical modeling of plasma processes
in electric Arc. The found results of the calculation of the fields, so are correct,
because they satisfy not only the Maxwell’s equations, but also all the laws of
conservation: of energy, momentum, mass and others.
At this stage, we believe that the temperature field of the Arc is already set
correctly (it was actually found from the joint result of simulation of all the physical
processes in electric Arc), and determine the corresponding intensities of electro-
magnetic field. Here, according to the above described method of determining the
intensity of the two-dimensional electromagnetic field of the electric Arc using the
function of the electric current Φ we analyze the results of calculation of the field of
open Arc for welding using the current I = 90 A and 160 A, the plasma-forming gas
is Argon, the gas consumption—QAr = 20 l/min, the diameter of the nozzle dn =
14 mm, the length of the Arc l = 18 mm. Open Arc between vertically arranged
cathode and the anode has a cylindrical axial symmetry. The scheme of such Arc,
and obtained field of a function of the electric current are shown in Fig. 8.23.
From the obtained field of the electric current (Fig. 8.23), it can be seen that at
the initial and transition areas the free open Arc is actively expanding. A similar
pattern (narrowing of the electric Arc) is observed at the anode, that in some case is
determined by the form of attachment of the Arc to the anode (contracted or
diffusion). In the zone where there is uniform distribution of the function of the
electric current along the radius, we have the zero value of the current density of the
Arc according to the Eq. (8.28). On this basis, the radius of the zone of the current
flow can be determined using the cross-sections of the electric Arc. In this case, for
the current I = 90 A this area of the Arc is bounded with the function of electric
328 8 The Arc Plasma Torches

Fig. 8.23 The fields of temperature and the function of electric current of the welding Arc
(QAr = 20 l/min, I = 90 A)

current Ф = 14.3 A (Fig. 8.23). If we compare the field of the current flow with a
field of the Arc temperature (Fig. 8.23), then for Argon plasma the boundary of the
current flow roughly coincides with the line of the isotherm T = 7000 K, at this
temperature the electrical conductivity σ * 3–4 Ω−1 cm−1. These conclusions are
also confirmed in [8, 32].
The maximum value of the intensity of the electric field Ez = 22.9 and
23.64 V/cm is achieved in the initial section of the Arc at the regimes of operation
with the currents I = 90 and 160 A, respectively. At the cathode the conductive
zone is minimal, but the value of the intensity of the electric field is maximal. With
the expansion of the Arc from the cathode to the anode the maximal value on the
axis reduces, but the zone of the electric field is extended (Fig. 8.28) for keeping
unchanged current in the cross-section of the Arc. The maximal value of the radial
component of the electric field (Er = 7.2 V/cm) is achieved in the zone of active
extension of the Arc current near the cathode (Fig. 8.25).
From the comparison of the obtained distributions Ez and Er (Figs. 8.24, 8.25) it
can be seen that Er can be up to * 30 % of Ez and more in the initial and
transitional areas of the free open Arc. This means that the components of the
current density jr and the electromagnetic force Fz and will also be significant in
comparison with the basic components jz and Fr. So, the pattern of current flow
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 329

Fig. 8.24 The distribution of the intensity of the electric field in Argon welding Arc at I = 90 A
(a) and its changes on the axis Ez1, Ez2 (b): Ez1 the intensity of the electric field at I = 90 A, Ez2 the
same at I = 160 A

Fig. 8.25 The changes of the radial components of the intensity of the electric field in Argon
welding Arc at I = 90 A in the sections z = 1, 2, 3, 5, 9, 14, 17 and 18 mm, respectively
330 8 The Arc Plasma Torches

Fig. 8.26 The changes of the intensity of the magnetic field Hφ1–Hφ8 in Argon welding Arc at
I = 90 A (a) and 160 A (b) in the sections z = 1, 2, 3, 5, 9, 14, 17 and 18 mm, respectively

Fig. 8.27 The changes of the radial components of the current density jr1–jr8 in Argon welding
Arc at I = 90 A in the sections z = 1, 2, 3, 5, 9, 14, 17 and 18 mm, respectively
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 331

(Fig. 8.23), the released energy (Fig. 8.29), and the distribution of electromagnetic
force (8.30), that directly affect to the velocity of heating and acceleration of plasma
in the open free Arc will be significantly distorted if we neglect the radial

Fig. 8.28 The distribution of the axial component of the current density in Argon welding Arc
I = 90 A (a) and the changes of jz1–jz8 at I = 90 A (b) and 160 A (c) in the sections z = 1, 2, 3, 5,
9, 14, 17 and 18 mm, respectively
332 8 The Arc Plasma Torches

Fig. 8.29 The distribution of specific power in the Argon welding Arc at I = 90A (a) and the
changes P1–P8 at I = 90 A (b) and I = 160 A (c) in the sections z = 1, 2, 3, 5, 9, 14, 17 and
18 mm, respectively

component of the intensity of the electric field that proves the necessity of a
two-dimensional setting of the problem of the field calculation in electric Arc.
8.4 Two-Dimensional Electromagnetic Problem for the Arc Plasma Torches 333

Fig. 8.30 The distribution of radial components of the electromagnetic force Fr1–Fr8 in the Argon
welding Arc at I = 90 A (a) and I = 160 A (b) in the sections z = 1, 2, 3, 5, 9, 14, 17 and 18 mm,
respectively

Fig. 8.31 The distribution of the axial component of the electromagnetic force Fz in the Argon
welding Arc at I = 90 A
334 8 The Arc Plasma Torches

The distribution of the components of the intensity of the magnetic field Hφ is

presented in Fig. 8.26. Maximum of Hφ is achieved in a zone close to the boundary
of the current flow and the component reduces out of the conductive zone by law
Hu ¼ 2prI
. The fields of the components of the current density of the Arc jr and jz are
presented in Figs. 8.27 and 8.28, the field of the specific power P—in Fig. 8.29,
and the distribution of the components of electromagnetic force Fr and Fz—in
Figs. 8.30 and 8.31, respectively.
Pay attention to the nature of the changes, the absolute values of the ratio of
these quantities for the open Arc in comparison with the corresponding data for
RF-plasma torch (Sect. 7.2.3). Due to the high current density, high specific power
and the electromagnetic force the temperature and the velocity of plasma flow in the
open Arcs are usually much higher than in RF-plasma torches.

8.5 The Free Burning Arc Plasma Torches

8.5.1 The Arc Plasma Torches for Welding and Melting

of Metals

The most important technical applications of the open free burning Arc are welding
and melting of metals. Unlike the Arc in the channel, the conductive area of it is
severely restricted by the walls of the channel and the open free Arc is stabilized
due to the flowing stream of the gas and there is the opportunity to expand its
conductive area that leads to relatively low current density and released power. The
temperature and velocity of plasma in the free Arc are typically lower than the
temperature and velocity of the Arc in the channel and are T ≈ (10–20) × 103 K
and several tens of meters per second [8, 9]. Lack of the wall of the channel greatly
simplifies the construction of the plasma torch and creates favorable conditions for
welding and melting of metals. The disadvantages of the free Arc include the low
efficiency associated with the loss of energy to the environment.
The Arc Plasma Torches for Welding of Metals Currently, up to 80 % of
technologies for processing of metals using the electric Arc—is welding [42]. This
is the most spread process of the heat treatment of metals can be divided into three
groups depending on the method of welding (Fig. 8.32):
1. Welding using the consumable electrode in inert gas (MIG—Metal Inert Gas).
This process is characterized in that the Arc burns between a consumable anode
(reversed polarity) and the base metal, and the weld is formed from the anode
material. Typically, this method is used for automatic welding, where the
plasma torch is moved at a constant speed and supply of cathode material is
performed automatically (Fig. 8.32a).
2. Welding using the nonconsumable electrode in inert gas (TIG—Tungsten Inert
Gas). In this way, the welding Arc is similar to Arc for cutting between the
8.5 The Free Burning Arc Plasma Torches 335

Fig. 8.32 Schemes of the Arc welding methods: a using the consumable electrode, b using the
nonconsumable electrode, c the same using the metal additive; 1 nozzle, 2 electrode, 3 welded
products, 4 plasma–forming gas, 5 metal additive

nonconsumable (tungsten) cathode and anode of the welded metal. The Arc in
this regime significantly differs from the Arc for cutting. This difference is due
to the fact that in the welding of metals the melting of the metal is only required
and spattering of it is unacceptable, so the welding Arc has lower speeds (30–
150 m/s) and lower temperatures T ≈ (10–15) × 103 K. At the welding using
the nonconsumable electrode (Fig. 8.32b, c) weld seam is formed from the
molten metal (with a small thickness of the metal) or from the metal additives,
that is fed into the welding area (with a large thickness of the welded metal).
This welding method can be applied for welding of metals that cannot be welded
by the other means (titanium, aluminum).
3. Welding of metals using the submerged consumable electrode.
Similar to the process of cutting of metals the efficiency of welding is largely
determined by the speed of the process. At the same time the requirements for the
regime of the Arc are tougher because of the necessity of forming the weld of
specified quality that is determined not only by its geometry, but also by the
chemical composition of the weld metal. In all developed countries in the recent
decades the number of equipment and devices for plasma welding has been created.
The Arc Plasma Torches for Melting of Metals Melting Arc plasma torches are
plasma torches with the Arc of direct action. One of their spots is out of the molten
metal. The heat released in the anode spot and to a lesser extent, the radiation
energy of the Arc is also useful.
Currently, there are the industrial examples of the plasma torches operating at the
currents lower than 10 kA. Some data on plasma torches for melting of metals are
given in Table 8.4 (the cathode is made of thoriated or lantaniumed tungsten).
The DC Arc plasma torches of several thousand of amperes are used for the
melting. The voltage of the Arc depends on the composition, consumption of the
gas and its length in the range of the voltage is 30–150 V and higher. Very often,
for increase the power of the furnace for melting of the metal in one bath several
plasma torches operates in parallel. Efficiency of plasma torches during the
336 8 The Arc Plasma Torches

Table 8.4 Parameters of Arc plasma torches for melting of metals [1]
Arc current I, A Diameter of the cathode dk, mm Consumption of Argon G, g/s
800 5–6 1
2000 10–12 2
4000 18–20 3
5000 24–25 4
10,000 48–50 6

surfacing reaches 85 %. Plasma furnaces are designed primarily for melting of the
steels, but can also be used for melting non-ferrous metals and their alloys. Their
main advantage over conventional Arc steel-melting furnaces is the absence of
graphite electrodes that eliminates carbonization of metal and allows smelting of
low carbon steels.
Another important advantage of these furnaces is the opportunity to seal the
furnace space. For graphite electrodes it was not possible to create a sealed input
device into the furnace, in contrast with it the sealing of metal water-cooled
housings of plasma torches have no the great difficulties. This fact allows you to
maintain any predetermined atmosphere in the furnace space in the process of
melting (neutral, oxidizing or reducing), and change it for a given program. Last not
least important feature of the plasma furnaces is a significant reduction in pollution
of the environment, and improving of the working health. Plasma furnaces almost
operate silently, unlike the Arc ones. On the other hand, the cost of melting is
higher than for the Arc, due to the consumption of Argon and large power con-
sumption due to the greater number of water-cooled units and higher temperatures
for the furnace lining. Therefore, the use of plasma furnaces is most advisable in the
smelting of high-strength steels and alloys hen the reduction of the consumption of
alloying elements by reducing of the burn has the determining value.
Currently, plasma furnaces are being built that in the form of furnace space are
like the Arc steel-melting furnaces (Fig. 8.33). At the using of a single plasma torch
strengthen it is stayed vertically in the roof (Fig. 8.33a). Several plasma torches can
be positioned vertically or obliquely in the side walls (Fig. 8.33b).
As an example, the main there are parameters of industrial plasma furnaces with
a capacity of 5, 10 and 30 t that are in operation at plants in Russia and Germany.
All furnaces are equipped with a DC Arc plasma torches, and the melting space is
sealed so that the overpressure is hold on in it during the melting process.
The furnace with capacity of 5 t has one plasma torch, introduced vertically
through the roof. Power of the plasma torch is 3500 kW. Energy consumption for
melting is about 700 kWh/t. Argon consumption is 200–250 l/min. The lifetime is
over 70 melting and the hearth—150–160 heats. The furnace with capacity of 10 t is
equipped with three Arc plasma torches, entered through the side walls of the furnace.
Plasma torches are mounted on special mechanisms that allow axial and angular
move them. Each plasma torch is powered by a separate power supply thyristor with a
wide control range. Operating current of the plasma torches is 6 kA at the voltage of
200–600 V. The real power introduced into the furnace during the melting process 7–
8.5 The Free Burning Arc Plasma Torches 337

Fig. 8.33 Scheme of the plasma furnace with ceramic crucible [1]: a a plasma torch in the roof;
b the plasma torches mounted in the walls: 1 housing, 2 lining, 3 an arch, 4 plasma torch, 5 bottom
electrode

8.5 MW, the melting speed—7–10 t/h, the energy consumption—650 kWh/t, the
consumption of argon—400–500 l/min.
The largest plasma furnace is a 30 t furnace of Russian OKB—1556. Four
plasma torches are introduced obliquely through the side walls using the hydraulic
actuators they can be moved along the axis and change the angle of the incline.
Power of DC plasma torches is carried by three thyristor converters, so that both
three plasma torch operates simultaneously. At the operating, any of the plasma
torches can be turned off, and the other can be turned on to speed up the melting
process. Maximum current of the power supply is 10 kA at the voltage of 150–
660 V, introduced during the melting capacity of 12–15 MW, the specific energy
consumption for melting is 625 kWh/t, the consumption of Argon—1000 l/min. It
would be expected that with the advent of more resistant refractory (ZrO2-based)
technical and economic parameters of plasma furnaces the characteristics of the
furnaces will improve at a faster rate than for conventional Arcs.

8.5.2 Algorithms of the Mathematical Models

One of the works devoted to the theoretical study of the opened electric Arcs is a
work of Lovke J., where the results of the of electric Arcs calculation at the low
currents are presented, taking into account the natural convection [43] and the
338 8 The Arc Plasma Torches

high-current welding Arc [44]. Magnet-gas-dynamic equations with variables

«pressure—velocity» were solved by the method of establishment. It was assumed
that the intensity of electric field is constant in each cross-section of the electric Arc
(one-dimensional of the field). Note that in [43] the plasma temperature of 3000 K
was set on the anode surface. In the approach of local thermodynamic equilibrium
that is used in this work, gas conductivity at this temperature is equal to zero, that
leads to the infinite values of the intensity of electric field and the break of the Arc.
Therefore, this statement of the problem in a single—temperature approximation is
incorrect. In [30], a mathematical model of a stationary high-current electric Arc is
constructed. The characteristics of the Arc with conical cathode, depending on the
distance between electrodes and current, are described. The calculated temperature
fields have the quite well agreement with the experimental data.
The authors of [32, 45–47] performed a numerical analysis of the influence of
the geometry and the size of electrode, the velocities of coaxial blowing and the
influence of type of plasma—forming gas on the characteristics of an open Arc. The
decision was carried out in the variables «vortex—flow function». However, in
these works the pressure gradient along the radius in the Arc is taken equal to the
radial electromagnetic force that leads to the overestimate field of the pressure in the
Arc and to elimination of the role of acceleration in plasma by the electromagnetic
force. In the studied electric devices used in these works, cathode is a rod electrode
with a conical or a hemispherical shape of the end and a composite cylindrical with
a flat end and a coaxial blowing. The anode in all these cases is an infinite flat plate.
The physical variables «pressure—velocity» and on the basis of two-temperature
plasma model in [31] analyzed the Argon open free Arc for welding of metals is
carried out. The field of electron and ion-atom temperatures are calculated as well as
the velocity field of plasma, all the electrical parameters of the welding Arc current
with I = 75; 100 A with the argon consumption Q = 10–20 l/min.
As it is noted in Sects. 1.2 and 1.3, equilibrium and nonequilibrium plasma
models are widely used for simulation of plasma processes in the electric Arc. Here,
on the basis of these models the simulation of open free Arc and the comparative
analysis of simulation results using the examples of the Arcs for welding and
melting of metals will be carried out. In contrast to the simplified model of the Arc
in the channel (Sect. 8.3.1) the used models take into account the two-dimensional
electromagnetic field of the Arc (Sect. 8.4) and the motion of gas, i.e., full
two-dimensional field momentum and continuity equations of gas included in the
decision process. Note that the technique outlined here can also be successfully
used for the study of the Arc in the channel, especially to its initial, transitional
zones and in the zone of the Arc attachment taking into account the boundary
conditions at the wall.
The full two-dimensional model of equilibrium plasma in Arc plasma torches in
a cylindrical coordinate system includes the following equation: equation of the
balance energy (1.25), the momentum Eq. (1.26, 1.27), the continuity equation of
all the gas flow (1.29), the Maxwell’s Eqs. (1.31), that can be solved using the
equation for the electric potential V (8.19) or a function of electric current Ф (8.30).
In Sect. 8.4, the algorithms for solving of these equations and their boundary
8.5 The Free Burning Arc Plasma Torches 339

conditions corresponding to the discrete analogue for internal and boundary points
are and equations for calculation of the parameters of the field of the electric Arc are
presented. For the two-temperature model, instead of the equation of the energy
balance for all the gas (1.25) two separate equations of the energy balance for
electrons (1.34) and heavy components (1.39) are used. In the case of ionization
equilibrium the equation of ionization equilibrium is included into the model (1.42).
The general algorithms for solving of these equations are given in Sect. 6.1. The
issue of sustainability and organization of the computational procedures are con-
sidered in Sects. 6.2 and 6.3. In Sect. 6.3.3, presented the discrete analogues for
various equations related to the construction of additional internal feedback loop.
These models of plasma also include the equations for determination of the
composition, the transport coefficients and thermodynamic functions of plasma
Sects. 4.1, 4.2, 4.3 and 4.4. In many cases, the coefficients of the transfer and the
thermodynamic functions of equilibrium plasma (especially at atmospheric pres-
sure) are known and are often specified as a table depending on the temperature and
pressure of the gas. The determination of them using the set value of the temper-
ature and the pressure are reduced to normal approximation problem, for example,
using the linear or cubic spline.
For the complete formation of two-dimensional systems of discrete analogues
for equilibrium and nonequilibrium models of the electric Arc we can determine
only the boundary conditions and the corresponding discrete analogues of the
boundary points of the equation of the balance energy in the model of equilibrium
plasma, the momentum equation, the continuity equation for all the gas, the
equation of the balance energy for electron gas and heavy components in the model
of nonequilibrium plasma and equation of ionization equilibrium.
Boundary Conditions At the cathode, the temperature and concentration of
plasma should be determined from the experiment or the calculation of
near-cathode processes. However, the complexity of calculation of the
near-electrode processes and the complexity of experiments limits these approa-
ches. In this case, the spot of cathode can be excluded from the general review of
the Arc and use the simple approximate models for determination of the plasma
temperature in the zero section. Note that at the cathode the plasma velocity is
relatively small, convection does not play a significant role in the balance of power,
and the heat extraction is mainly carried out due to the thermal conductivity in
radial direction. It is also assumed that movement of the gas in the near-electrode
regions does not violate the conditions of the ionization equilibrium, that the
concentration of particles can be determined using the Saha equation (4.1). Thus,
we have, respectively, the following approximate models for determination of the
boundary conditions on the cathode surface:
340 8 The Arc Plasma Torches

(a) One-dimensional model of equilibrium plasma

 
1d dT I
rE2  Urad ¼  rk ; E¼ ; ð8:37Þ
r dr dr RRc
2p rrdr
0

(b) One-dimensional model of nonequilibrium plasma

 
1d dTe
rE 2  Urad  bðTe  T Þ ¼  rke ;
r dr dr
 
1d dT I ð8:38Þ
bðTe  T Þ ¼  rkai ; E¼ ;
r dr dr R
Rc
2p rrdr
0

where Rc—the radius of the cathode spot is equal to the radius of the optical
emission of the gas in the zero section at the temperatures of TR, TeR TaR; TR,
TeR TaR—the gas temperature in equilibrium plasma (8.37), electrons and atoms
(8.38) at the border of optical radiation of the conductive cathode zone. For
example, for argon TR = 6500–7000 K, TeR = 7000–7500 K, TaR = 300–
500 K. The other notations are entered similar to the case of equilibrium and
nonequilibrium models (Sects. 1.2, 1.3).
Approximate models (8.37) and (8.38) can be used for determination of the
temperature in the conductive area in the initial section of the Arc. The main
advantage of such models is the guarantee of implementation of the energy balances
and the electromagnetic field in the zero section.
Outside the conductive zone temperature of the cold gas is equal to T = 300 K.
At the cathode, the plasma velocity is zero, and outside the conductive zone is
considered that the gas is uniformly applied on the inner channel of the Arc nozzle
and the velocity is determined using the predetermined gas consumption G, kg/s or
Q, m3/s using the following equation:

G Q
vz ¼  ¼  ; vr ¼ 0; ð8:39Þ
pq R2 Rk
2 p R  R2k
2

where Rk and R—radiuses of the cathode and the nozzle of the Arc, respectively.
On the axis of symmetry from the condition of the cylindrical axial symmetry we
have:

@T @Te @ne @vz

¼ ¼ ¼ ¼ vr ¼ 0 ð8:40Þ
@r @r @r @r

At the anode, the gas velocity is zero. For determination of the temperature at the
anode models (8.37) and (8.38) as the analogy can be used with the set of anode
spot radius Ra instead of the cathode spot Rc. Unlike the cathode the anode is
8.5 The Free Burning Arc Plasma Torches 341

usually downstream and the near-electrode processes on the anode are much less
active (in comparing with cathode) occur, and therefore to determine the boundary
conditions at the anode a simple model of steady-state process at the exit can be
used, i.e., the target variable gradients are equal to zero:

@T @Te @ne
¼ ¼ ¼ vz ¼ vr ¼ 0 ð8:41Þ
@z @z @z

To validate these models various boundary conditions were calculated using

(8.37), (8.38) and (8.41) for the open free Arc with currents of 90 and 160 A. The
discrepancy between the temperature at the anode (downstream) are not more than
10 %, and the resulting temperature field in the Arc (upstream) practically do not
differ from each other.
To simulate the open free Arc border of the free boundary should be chosen at
the sufficient distance from the Arc. If the anode is a flat surface and the distance
between the electrodes is small, the plasma flow of the Arc, striking with anode can
spread along long sides. In this case, the zone of interaction between the Arc and
the environment can occupy a large space, here the temperature and velocity of the
plasma flow gradually reduced until the temperature and the velocity of the envi-
ronment. Therefore, the selected area of calculation should be sufficient for simu-
lation of plasma processes in the Arc and the establishment of the zone of
interaction between the Arc and environment. For the majority of electric free Arcs
such a free boundary can be selected in the region of three–five radius of the Arc.
As in the case of the boundary conditions at the exit (downstream) of the channel
of RF-plasma torch (Sect. 7.3.1), on the free boundary all the processes can be
considered established, the changes of the temperature and the velocity at the
normal to the boundary are non-essential, i.e.:

@T @Te @ne @vz @vr

¼ ¼ ¼ ¼ ¼0 ð8:42Þ
@r @r @r @r @r

For the free Arc the pressure on the free surface is known for the Arc at the
atmospheric pressure at this boundary p = 1 atm.
At some borders of the free Arc not all the unknown variables can be prede-
termined, for example the gas pressure at the cathode, at the anode and at the axis of
symmetry are unknown. At these boundaries only the values of the velocity
components or their normal gradients are known. Therefore, the determination of
the pressure on these boundaries can be carried out using the Shifted Chess Grid
and the techniques, described above in Sect. 6.1.2. On the other hand, the points for
determination of the target variable are desirable to be placed directly on the bor-
ders, where their values are known. This is done for obtaining of the maximum use
of this information on the boundary conditions.
Using the considered boundary conditions the following structure of Shifted
Chess Grids for modeling of plasma processes in the open free Arc is proposed
(Figs. 6.5b, 8.34, 8.35)
342 8 The Arc Plasma Torches

Fig. 8.34 The shifted chess grid and the control volumes for obtaining of the discrete analogues
of the equation of the energy balance, equation of ionization equilibrium and the momentum
equation for open free Arc: filled circle points for determination of the temperature and density,
down arrow the same for the component of velocity vz, right arrow the same for component of
velocity vr, cross line the same for the pressure p

Electromagnetic equation for the function of electric current Ф is solved on the

grid [z1, r1], its origin of coordinates begins on the axis of symmetry and at the
cathode surface. Grid takes the area between the cathode and the anode, the
boundary points are on both electrodes and on the axis of symmetry;
Equation of the balance energy and equation of ionization equilibrium are
resolved on the grid [z1, r2], shifted compared with the grid [z1, r1] for a half step
8.5 The Free Burning Arc Plasma Torches 343

Fig. 8.35 The shifted chess grid and the control volumes for obtaining of the discrete analogues
of the continuity equation for the pressure of the open free Arc: the notations are the same as in
Fig. 8.34

to the side of the free boundary. The boundary points of this grid are placed directly
on the cathode, the anode and on the free boundary;
The momentum equation of plasma for vz is solved using the grid [z1, r2];
The momentum equation of plasma for vr is solved on the grid [z2, r1], shifted
compared with the net [z1, r1] for a half step toward the anode, the boundary points
are located on the axis of symmetry and on the free boundary;
The continuity equation is solved and the pressure field is defined using the mesh
[z2, r2], shifted toward the free surface and the anode, respectively, for a half step
344 8 The Arc Plasma Torches

relative to the grid [z1, r1]. The boundary points of this grid are located directly on
the free surface, the pressure is usually known here.
Discrete Analogues for the Points on the Borders Because of the displacement
the points for definition of the target variable cannot be on located on the border,
but they are adjacent to it, so for these points it is necessary to use the appropriate
discrete analogues using the boundary conditions. The procedure for obtaining
discrete analogues of boundary points of the free Arc in many respects similar to the
procedure discussed above (Sects. 7.3.1, 8.4.1, 8.4.2) and we are presenting here
only received discrete analogues.
1. Discrete analogues of equation of the balance energy, equation of ionization
equilibrium and the momentum equation. Equation of the balance energy,
equation of ionization equilibrium and the momentum equation for vz are solved
at the same grid (Fig. 8.34) and have some similarity in boundary conditions,
they have same type of the discrete analogues for points on the boundary.
The discrete analogues for the internal points (Fig. 8.34) of the equation of the
balance energy, equation of ionization equilibrium and the momentum equation for
vz, in accordance with the described in Sect. 6.1 decisions with internal feedback
(Sect. 6.3.2, 6.3.3), have the form (6.61) (6.34). Where the discrete analogue (6.61)
is used for equation of the balance energy for the models of equilibrium and
nonequilibrium plasma. As a result of the offset of the grid for the solution of the
equation of the balance energy, equation of ionization equilibrium and the
momentum equation there are seven types of the special boundary points (control
volumes 1–7 on Fig. 8.34).
(a) The discrete analogues of the equation of the balance energy, equation of
ionization equilibrium and the momentum equation for vz for points of the
control volume 1 (Fig. 8.34). For these points, the boundary conditions (8.40)
are used. Therefore, the contribution of the flows through the left verge s of the
control volume 1 into the total balance is equal to zero. It follows that the
discrete analogue of the equation of the balance energy, equation of ionization
equilibrium and the momentum equation for vz for these points has the same
form (6.61) and (6.34) with AS = 0.
(b) The discrete analogues of the equation of the balance energy, equation of
ionization equilibrium and the momentum equation for points of the control
volume 2 or 4. At the obtaining the discrete analogues of the equation of the
balance energy, equation of ionization equilibrium and the momentum equa-
tion for vz for the points on the free border we use the approximate boundary
condition (8.42), and a predetermined pressure on this boundary p. Integration
of the equation of the energy balance, equation of ionization equilibrium and
the momentum equation for vz is only in the half of the control volume 2, and
the momentum equation for vr—in the control volume 4. Taking into account
the boundary condition on the free boundary the discrete analogues for these
points have the same form (6.61), (6.34), at AN = 0.
8.5 The Free Burning Arc Plasma Torches 345

(c) The discrete analogues of the momentum equation for vr for points of the
control volume 3 or 5. At the entrance and at the anode there are the similar
boundary conditions vr ¼ @v @z ¼ 0. Therefore, for the points at the entrance the
r

contribution of flows into the momentum balances through the upper verge
w of the control volume 3 (Fig. 8.34) is equal to zero. Accordingly, for these
points the discrete analogue has the form (6.34) at AW = 0. Similarly, for the
points adjacent to the anode, AE = 0.
(d) The discrete analogues of the momentum equation for vr for points of the
control volume 6 or 7. For these points there are two types of boundary
conditions for the points adjacent to the free boundary (@v @r ¼ 0) and to the
r

points adjacent to the entrance or the anode (vr ¼ @v @z ¼ 0). Therefore, the
r

contributions of the flows through the verges w and n for the control volume 6
and flows through the verges e and n for the control volume 7 are not taking
into account at the obtaining of the discrete analogue. And for these points the
discrete analogue has the form (6.34) at AW = AN = 0 for points of the control
volume 6 and at AE = AN = 0 for the point of the control volume 7.

2. The discrete analogues of the continuity equation for the pressure p.

According to the decision of the continuity equation for the pressure (Sect. 6.1.2)
its discrete analogue for the internal points (Fig. 8.35) has the form (6.38). Using
the shifted chess grid for the solution of the continuity equation for the open free
Arc, there are five types of special boundary points (the control volume 1–5 Fig. 8.
35), their discrete analogue is different from the discrete analogue for the internal
points because of the boundary conditions.
(a) The discrete analogues of the continuity equation for points of the control
volume 1. The feature of the points of the control volume 1 is the given
boundary condition for the velocity vz on the up verge w. In accordance with
the general procedure for the obtaining of the discrete analogue of the con-
tinuity equation (Sect. 6.1.2) the discrete analogue for these points has the
same form (6.38) with the following coefficients:

aW ¼ 0;
ðrn  rs Þðrn þ rs Þ
bM ¼ ðqw vzw  qe^vze Þ þ ðrs qs^vrs  rn qn^vrn Þðze  zw Þ:
2

From these expressions it is seen that using the shifted grid it is not necessary
to determine the unknown pressure at the point W, and well-known velocity at
the entrance vzw is included into the source coefficient bM instead of ^vzw .
(b) The discrete analogues of the continuity equation for points of the control
volume 2. The boundary condition for points of the control volume 2 is the
zero radial component of the velocity on the axis of symmetry, vrs = 0.
346 8 The Arc Plasma Torches

Therefore, the discrete analogue for these points has the same form (6.38)
with the following coefficients:

rn2
aS ¼ 0; bM ¼ ðqw^vzw  qe^vze Þ  rn qn^vrn ðze  zw Þ:
2

(c) The discrete analogues of the continuity equation for points of the control
volume 3. The boundary condition at the anode vze = 0. The discrete analogue
for these points has the same form (6.38) and:

ðrn  rs Þðrn þ rs Þ
aE ¼ 0; bM ¼ qw^vzw þ ðrs qs^vrs  rn qn^vrn Þðze  zw Þ:
2

d) The discrete analogues of the continuity equation for points of the control
volume 4 and 5. For the points 4 and 5 there are two boundary conditions: on
the axis of symmetry (vrs = 0) and at the cathode (vzw = 0 or it is set) or at the
anode (vze = 0). So, the discrete analogue for these points has the form (6.38)
with the following coefficients:
For points of the control volume 4

rn2
aW ¼ aS ¼ 0; bM ¼ ðqw vzw  qe^vze Þ  rn qn^vrn ðze  zw Þ;
2

For points of the control volume 5

rn2
aE ¼ aS ¼ 0; bM ¼ qw^vzw  rn qn^vrn ðze  zw Þ:
2

Thus, taking into account the features of the shifted grids and the boundary
conditions we have all the discrete analogues of equation of the balance energy,
equation of ionization equilibrium, the momentum equation and the continuity
equation for the open free Arc. For internal and boundary points, they have the total
view, respectively (6.61), (6.34) and (6.38) with the change of the coefficients for
the boundary points.
Using the similar techniques, the discrete analogues of different equations for
points on the boundary of the electric Arc in the channel and others can be easily
obtained. The shifted grids presented in Fig. 6.5 also can be successfully used.

8.5.3 Comparative Analysis of the Simulation Results

The results of simulation of typical plasma processes of open free Arc for welding
and melting of metals are presented here.
8.5 The Free Burning Arc Plasma Torches 347

The open welding Arc with the operating range of currents I = 50–300 A and a
length l = 5–18 mm have the radius of nozzle R = 5–7 mm, the radius of the
cathode spot Rc = 0.3–0.6 mm, the radius of the conductive zone in the zero section
is selected for the radius of the radiation R1 = 1.3–1.6 mm, respectively, the radius
of the modeling zone R2 = 30 mm, the radius of result presentation area
R3 = 20 mm. The consumption of Argon QAr = 10–40 l/min. At the simulation of
the welding Arc with a nominal current I = 90 A and 160 A the main obtained
results are: the temperature field, the field of the plasma flow, the field of velocities,
the field of the static pressure, distribution of densities of particle, etc. The results of
the calculation of electromagnetic field of these Arcs and analyze of them were
presented in Sect. 8.4.3.
The boundary conditions for the temperature at the cathode and at the anode are
obtained using the one-dimensional model (8.37) and (8.38) at the predetermined
radiuses of the conductive zones at the cathode and anode. The boundary condition
for the electromagnetic field is defined like in Sects. 8.4.1 and 8.4.2. The shifted
grids, boundary conditions for the velocity, pressure and corresponding discrete
analogues are presented in Sect. 8.5.2. The initial temperature in the Arc is equal to
10,000 K. The initial components of the velocity, the function of electric current
and the electric potential are equal to zero, the atmospheric pressure is taken as the
initial one.
1. The results of the simulation of the Arc on the basis of equilibrium model.
The obtained fields of temperature and the gas flow in welding Arc with dif-
ferent regimes of the operation are presented in Figs. 8.36, 8.37 and 8.38. From
the figures it is seen that the open free Arc has a conical shape with a maximum
temperature in the zero section: at the regime with I = 90 A, QAr = 20 l/min,
T = 18,500 K; at the regime with I = 160 A, QAr = 20 l/min, T = 21,500 K.
The total specific power in Arc at these two regimes, respectively, P = 1.45 kW
and 3.2 kW. The voltage drop in the Arc U = 16.1 and 20.3 V, respectively.
Rr
The distribution of the current function (gas-dynamic) w ¼ qvz rdr gives us the
0
picture of the plasma flow in the open free Arc (Figs. 8.36a, 8.37a, 8.38). Thus, the
main part of the plasma-forming gas does not penetrate into plasma, and wraps
around it, forming an external stabilizing gas flow.
The used in the model concept of a free boundary allows the determination of the
degree of interaction between the plasma flow and the environment. The fields of
the gas flows are presented in figures and the outside vortex of the gas can be seen
there. The streamlined flow of plasma-forming gas of the Arc interacts with envi-
ronment, forming the external vortex flow. Thus, the free Arc is stabilized under not
one, but two external flows: streamlined flow of the plasma-forming gas and the
vortex flow formed by the interaction of the streamlined stream with the environ-
ment. The data with the values of the stream function also allow the determination
of the amount of plasma-forming gas entrapped into the Arc.
348 8 The Arc Plasma Torches

Fig. 8.36 The fields of temperature and gas flow in the welding Arc at I = 90 A, QAr = 20 l/min
(a) and the temperature changes T1–T8 in the sections z = 1, 2, 3, 5, 9, 14, 17 and 18 mm,
respectively (b)

For the comparison of the data at the different regimes at Fig. 8.38 there are the
fields of temperature and the gas flow for the same Arc with I = 160 A,
QAr = 40 l/min. The larger flow of the plasma-forming gas creates the additional
8.5 The Free Burning Arc Plasma Torches 349

Fig. 8.37 The fields of temperature and gas flow in the welding Arc at I = 160 A, QAr = 20 l/min
(a) and the current function changes ψ1–ψ8 in the sections z = 1, 2, 3, 5, 9, 14, 17 and 18 mm,
respectively (b)

pressure in the cathode zone and changes the shape of the temperature field in the
Arc. However, the maximum values of the temperature and velocity of plasma on
the axis of symmetry are virtually unchanged. The total released power in the Arc
and the nature of the gas flow does not change. Consequently, the external
plasma-forming flow of plasma plays only the role of stabilizing flow and the
350 8 The Arc Plasma Torches

Fig. 8.38 The fields of temperature and gas flow in the welding Arc I = 160 A, QAr = 40 l/min

dynamics of the flow and heating of the gas in the Arc are determined by internal
plasma processes.
In Fig. 8.39 there is the field of the plasma velocity, and at Fig. 8.40—the
distributions of specific gas consumption. The results of simulation also show that,
although the absolute value vr is small compared with vz (Fig. 8.39), the specific
consumptions qvz and qvr are comparable (Fig. 8.40), it proves that it is necessary
to solve the two-dimensional problem of gas flow in the open free Arc.
Field of the static (excess) pressure in the Arc is shown in Fig. 8.41. The excess
pressure on the axis of the Arc is small (compared with the compressed Arc in the
channel) and does not affect the properties of plasma. However, the created field of
the pressure gradient is comparable with electromagnetic force in the Arc
(Figs. 8.30, 8.31, 8.42) and the gradient of static pressure can not excluded from the
equations of the gas motion. The created pressure in the Arc prevents to the
excessive penetration of gas into plasma, and provides the high level of temperature
and conductivity of the Arc, contributing to steady and sustainable heating of the
plasma flow in the Arc. Under the influence of two opposing forces—the radial
component of the electromagnetic force Fr and the gradient of the static pressure
@p=@r, a small amount of hot gas with a temperature T = 7000–8000 K pene-
trates into the Arc for heating (5–10 % Q) (Figs. 8.36, 8.37 and 8.38).
8.5 The Free Burning Arc Plasma Torches 351

Fig. 8.39 The fields of plasma velocity components in the welding Arc at I = 90 A, (a, b) and
160 A (c) (QAr = 20 l/min)

The approximate method

fordetermination of the pressure in the Arc using the
@p
electromagnetic force @r ¼ Fr is popular in literature, for example [8, 9, 21, 26,
32, 45, 46]. This estimate may not lead to the large error of the calculation of the
absolute value of the pressure. However,  @p @r ¼ Fr , the result force that leads to
the plasma acceleration in the radial direction is equal to zero, that greatly reduces
the radial velocity vr and the amount of gas, entrained from the outside into the Arc.
352 8 The Arc Plasma Torches

Fig. 8.40 The distribution of the specific gas consumption in the welding Arc at I = 90 A,
QAr = 20 l/min (a), and the changes qvz1  qvz8 (b) and qvr1  qvr8 (c) in the respective sections
z = 1, 2, 3, 5, 9, 14, 17 and 18 mm
8.5 The Free Burning Arc Plasma Torches 353

Fig. 8.41 Field of the static (excess) pressure in the welding Arc at I = 90 A, QAr = 20 l/min
(a) and the changes Δp1–Δp8 at I = 90 A (b) and 160 A (c) in the respective sections z = 1, 2, 3, 5,
9, 14, 17 and 18 mm

It should lead to a significant error in the calculation of the field of the flow and
velocity of plasma. Furthermore, some authors ignore the gradient of static pressure
that contrary, leads to overestimated value of the forces of gas acceleration flow,
distortion of the fields of flow and gas velocity.
354 8 The Arc Plasma Torches

Fig. 8.42 Fields of the gradients of pressure in welding Arc at I = 90 A, QAr = 20 l/min

In fact, the component of electromagnetic force Fr, causing the moving of the
gas into plasma simultaneously helps the creation of the large gradient of static
pressure in the Arc that prevents the penetration of an excessive amount of gas into
the Arc, and does not lead to cooling down of plasma. As a result, at the thermal
influence the Arc maintains itself its high temperature and the steady-state of
combustion. In this case, the gradient of the static pressure acts as internal feedback,
and together with the electromagnetic force automatically adjusts the required
amount of cold gas entrained into plasma and heat plasma to the high temperature.
So it becomes clear why an excessive amount of plasma gas does not penetrate into
the plasma bunch and wraps around it and have the function of the plasma flow
stabilization (Fig. 8.40). The simulation results show that the sum (difference) of
these forces (Fig. 8.43) is equal to 3–5 % of the value of the force Fr (Fig. 8.30a)
that is responsible for the acceleration of the gas in the radial direction and the
amount of gas entrained into plasma.
From the standpoint of gas-dynamic the Arc operates like the magnetic pump,
especially in the near-cathode zone, where there is the maximum current density,
specific power and of electromagnetic force (Figs. 8.27, 8.29, 8.30, 8.31).
According to the law of mass conservation the amount of gas entrained into the
near-cathode area is equal to the amount of heated gas, accelerated in the Arc.
8.5 The Free Burning Arc Plasma Torches 355

Fig. 8.43 The sum of the forces ( @p @p8

@r þ Fr1 )–( @r þ Fr8 ) in open free Arc at I = 90 A,
1

QAr = 20 l/min in the respective sections z = 1, 2, 3, 5, 9, 14, 17 and 18 mm

Therefore, the accuracy of the modeling of the dynamic regime in the Arc depends
on the accuracy of the determination of the total force  @p @r þ Fr in the cathode
zone.
Note also that near the axis of symmetry the gradient of static pressure  @p @r can
be even greater than electromagnetic force −Fr that leads to lockup of the gas flow
in the radial direction to zero velocity at the axis. In this zone, the force Fr is
sufficiently small, and the pressure generated mainly due to reciprocal movement of
gas in the radial direction (Fig. 8.43).
It should be noted that often in practice, for analysis of the gas flow in the Arc
the one-dimensional formulation of the problem of the electromagnetic field is used,
i.e., axial component of the electromagnetic force (Fz = 0) is ignored and it is
assumed that the acceleration of plasma in the axial direction is carried out due to
the gradient of static pressure  @p
@z . The obtained results of the simulation of the Arc
taking into account the two-dimensional electromagnetic field (Sect. 8.4.3) show
that the absolute values of the gradient of the axial component of force of static
pressure  @p@z (Fig. 8.42a) and electromagnetic force Fz (Fig. 8.31) are comparable.
Neglecting of Fz leads to a significant reduction of the plasma velocity and to
inaccurate result of calculation of the plasma and the field of its course; thus once
again we prove the need to define a second component of the electromagnetic force
Fz in the simulation of plasma and two-dimensional setting of electromagnetic
problem of the electric Arc (Sect. 8.4).
There is the analogy in motion and heating of the gas in the electric Arc and in
RF-plasma torch (Sect. 7.3.2), note that the mechanism of the penetration of gas
into the plasma flow and its heating are the same in both cases.
About the Role of the Boundary Condition at the Anode The boundary con-
dition for the temperature at the anode can be determined by the one-dimensional
356 8 The Arc Plasma Torches

model of the balances of energy and field (8.37) or using the simple condition of the
steady process at the exit (8.41): @T@z ¼ 0. This condition allows us to obtain an
approximate discrete analogue of the equation of the energy balance for the points
on the anode (on the boundary) from the total differential Eq. (6.61) with coefficient
aE ¼ 0 (Fig. 8.34).
The anode is usually located at the bottom of the stream, and the temperature of
plasma at the anode is largely determined by the active processes that take place
upstream in the Arc and its own processes at the anode. Therefore, whatever the
approximate boundary condition was not taken at the anode, it did not significantly
affect the parameters of the Arc located upstream. To analyze the effect of different
kinds of boundary conditions on the results of simulation of the internal field the
temperature field and the flow of the gas in the Arc at I = 90 A, QAr = 20 l/min are
shown in Fig. 8.44, they are obtained using the boundary condition (8.41) at the
anode. With this boundary condition the obtained temperature field and the gas flow
near the anode are more evenly than in the case of the boundary condition (8.37),
because the near-anode processes and the zone of the binding are excluded from the
consideration. At a small distance (2–3 mm) from the anode the field of temperature
and the gas flow are almost the same in both cases (Figs. 8.36a, 8.44), i.e., and all
other parameters of the Arc are the same.

Fig. 8.44 The fields of the temperature and the gas flow in the welding Arc at I = 90 A,
QAr = 20 l/min, obtained using the boundary conditions @T
@z ¼ 0 at the anode
8.5 The Free Burning Arc Plasma Torches 357

The numerical verification also showed that at using of the boundary condition
@T
@z ¼ 0 at the anode, for the calculation of this zone the law of conservation of mass
ðdivðq~
vÞ ¼ 0Þ better takes place.
2. The simulation results of the welding Arc on the basis of the
two-temperature plasma model. The simulation results of welding Arc,
derived from the two-temperature plasma model will be compared with the data
obtained by the equilibrium plasma model. Terms simulation of the Argon
welding Arc identical in both cases, and the results are given in the same
cross-sections: z = 1, 2, 3, 5, 9, 14, 17 and 18 mm.
The temperature field of the electron gas and the function of electric current are
shown in Fig. 8.45, and the fields of temperature of heavy components and the gas
flow are shown in Fig. 8.46. Note that these results are almost identical to the data
that are obtained on the basics of the model of equilibrium plasma (Figs. 8.23,
8.36a). The temperature distribution of the electron gas and heavy components in
different sections of the Arc are shown in Fig. 8.47. For determination of the
properties of plasma there are the distributions of concentration of electrons and
atoms in different sections of the Arc in Fig. 8.48.

Fig. 8.45 The fields of the electron gas temperature and the function of the electric current in the
welding Argon Arc at I = 90 A, QAr = 20 l/min
358 8 The Arc Plasma Torches

Fig. 8.46 The fields of temperature of the heavy components and the gas flow in the Argon
welding Arc at I = 90 A, QAr = 20 l/min

In the central zone of Arc due to high electron temperature (Te > 10,000 K),
high current density and electron density (ne > 1016 cm−3) the higher frequency of
collisions between electrons and heavy particles (ν > 1011 s−1) is provided. Due to
this high frequency of collisions of electrons can efficiently transfer their energy to
the atoms and ions; this process leads to the thermal equilibrium among the groups
of particles. Such a channel of energy exchange running operates efficiently in the
dense plasma of the Arc, it ensures the stable combustion, and in the central part of
the Arc the thermal equilibrium of plasma is observed.
In the external zone of the Arc due to the losses of energy into environment, the
electron temperature is less than in the axial zone (Te < 9000 K), the concentration
of electrons is relatively low (ne < 1015 cm−3), and the frequency of collision of
electrons with atoms and ions ν < 1010 s−1. At these conditions, electrons cannot
give all the obtained energy to atoms and ions the gas stays in the thermal
nonequilibrium. The difference of temperatures between electrons and heavy
components in this zone is 3000–4000 K (Fig. 8.47). However, due to the low
energy of electrons and heavy components in this layer the thermal nonequilibrium
cannot have a significant effect to the motion, heating of the gas and the other
parameters of the Arc, that is confirmed by the coincidence of the basic data of two
models. Therefore, for modeling of the majority of technological processes using
8.5 The Free Burning Arc Plasma Torches 359

Fig. 8.47 The distribution of temperatures of electron gas Te1–Te8 and heavy components Ta1–Ta8
in the Argon welding Arc I = 90 A, QAr = 20 l/min in the respective sections z = 1, 2, 3, 5, 9, 14,
17 and 18 mm

the Arc at atmospheric pressure, including welding and melting of metals, it is

enough to use the model of plasma equilibrium. Apparently, nonequilibrium state
significantly affects the properties of plasma flow in the case of the small Arc
current density (I=d  100 A/cm) and at the reduction of the pressure (p  1atm).
An important result is the balance of power in the Arc. The distribution of
specific power in the different sections of the welding Arc is shown in Fig. 8.49: the
power P ¼ rE 2 obtained by electrons from the electric field; power Pai ¼
bðTe  Ta Þ transmitted by electrons to heavy particles at the elastic collisions; the
power losses due to radiation Urad; power Pe ¼ rE2  Pai  Urad of the electron gas
heating, most of this power
 spent to heat  conduction, convection and uncompen-
sated ionization PI ¼ div Ei ne  52 kTe na ~
v. Thus, by these data we can to define the
360 8 The Arc Plasma Torches

Fig. 8.48 The distribution of the concentration of electrons lg ne1  lg ne4 and atoms lg na1  lg na4
in the Argon welding Arc I = 90 A, QAr = 20 l/min in the respective sections z = 1, 2, 3 and
5 mm

Fig. 8.49 The balance of power in the Argon welding Arc I = 90 A, QAr = 20 l/min in the
sections z = 1 and 2 mm, respectively

role of each component in the power balance of the electric Arc; for example, the
role of radiation in the balance of energy increases with increasing temperature, etc.
3. Results of the simulation of the plasma Arc for melting of the steels. The
results of the modeling of free Arc in the chamber for melting of the steels with
following data: the current I = 1000 A, the consumption of argon gas
QAr = 90 l/min, the diameter of the tungsten (1.5 % ThO2) cathode dc = 10
mm, the size of the channel of plasma-forming gas Δr = 1.5 mm, the distance
between the cathode and the liquid metal bath l = 150 mm.
8.5 The Free Burning Arc Plasma Torches 361

Fig. 8.50 The fields of the temperature and the flow of the gas in the chamber of the Argon
plasma Arc for the melting of the steels

These obtained data of the fields of the temperature, the gas flow, the velocities
and the specific consumptions of the Argon gas (Figs. 8.50, 8.51); the distribution of
the intensity of electromagnetic field and specific power (Fig. 8.52), the components
of the electromagnetic force Fr and the excess static pressure Δp (Fig. 8.53) and
other parameters of the electric Arc allows the determination of the heating degree of
the gas, to analyze the efficiency of the plasma torch, to determine the thermal
efficiency of the gas heating, to estimate the heat flow into the anode, etc. The main
application of the obtained results and the developed method is to use them for the
theoretical study of plasma processes in the electric Arc, for design of the Arc plasma
torches, for determination of the optimal and stable regime of the operation with the
362 8 The Arc Plasma Torches

Fig. 8.51 The distributions of temperature T1–T8 (a), the components of velocity vz1 vz8 (b) and
the specific consumptions qvz1 qvz8 (c) in the respective sections z = 2.5, 10, 20, 50, 80, 110, 140
and 150 mm of the chamber of the Argon plasma Arc for melting of the steels
8.5 The Free Burning Arc Plasma Torches 363

Fig. 8.52 The distributions of the intensity of electric field E on the axis (a), the intensities of
magnetic field H1–H8 (b) and the specific power P1–P8 (c) in the respective sections z = 2.5, 10,
20, 50, 80, 110, 140 and 150 mm of the chamber of the Argon plasma Arc for the melting of the
steels
364 8 The Arc Plasma Torches

Fig. 8.53 The distribution of the components of the electromagnetic force Fr1–Fr8 (a) and excess
static pressure Δp1–Δp8 (b) in the respective sections z = 2.5, 10, 20, 50, 80, 110, 140 and
150 mm of the chamber of the Argon plasma Arc for melting of the steels

power supply, for establishment of the links between the internal and external
parameters of the Arc, that is important for the design of Arc plasma torches and
control of their parameters.

References

1. Dresvin S.V. Low-temperature plasma generators // Encyclopedia of Low-Temperature

Plasma / Ed. by V.E. Fortov. M.: Nauka, 2000. V. II. PP 280–328.
2. Finkelberg V., Mekker G. Electric Arcs and thermal plasma. M.: Publishing House of Foreign
Literature, 1961.
3. Granovsky V.L. Electric current in the gas. M.: Gostechizdat 1952.
References 365

4. Granovsky V.L. Electric current in the gas. Continuous current. M.: Nauka, 1971.
5. Modelling and calculation methods of physical and chemical processes in Low temperature
plasma / Ed. L.S. Polak. M.: Nauka, 1974.
6. Anderson J.A. Transport phenomena in thermal plasma. M.: Energy 1972.
7. Novikov O.J. The stability of the electric Arc. L.: Energy 1978.
8. Zhukov M.F., Koroteev A.S., Yuryukov B.A. Applied Dynamics of thermal plasma.
Novosibirsk: Nauka. 1975.
9. Zhukov M.F., Smolyakov V.J., Yuryukov B.A. Electric Arc gas heaters (plasma torches).
Novosibirsk: Nauka. 1973.
10. The theory of an electrical Arc in a forced heat transfer / Ed. Zhukov M.F. Novosibirsk:
Nauka. 1977.
11. The theory of thermal plasma Arc. Part 1: Mathematical Methods Plasma / M.F. Zhukov, B.A.
Yuryukov, V.S. Engelsht et al. Novosibirsk. Nauka. 1987.
12. The theory of thermal plasma Arc. Part 2: Non-stationary processes and radiation heat transfer
in thermal plasmas / M.F. Zhukov, B.P. Devyatov, O.J. Novikov et al. Novosibirsk. Nauka.
1987.
13. Zeldovich J.B., Raiser Y.P. Physics of Shock Waves and High-Temperature Hydrodynamic
Phenomena. M.: Nauka. 1966.
14. The generation of electric Arc plasma streams / Ed. P.E. Nakoryakov. Novosibirsk: Nauka.
1987.
15. Alexandrov A.F., Ruhadze A.A. The physics of high-current electric-discharge light sources.
M.: Atomizdat. 1976.
16. Dautov G.Y., Dzyuba V.L., Carp I.N. Plasma torches with stabilized electric Arcs. Kiev:
Naukova Dumka. 1984.
17. Heat transfer in the Arc gas heater / A.G. Shashkoff, L. Krejci, V.I. Krylovich et al. M.:
Energy. 1974.
18. Y.P. Raiser. Physics of the gas discharge. M.: Nauka, 1987.
19. Kolesnikov V.N. The Arc discharge in inert gases // Physical Optics (Proc. of Phys. Inst of
Acad. Sc. of USSR.). M.: Publishing PhIAN, 1964. V. 30. pp. 66–157.
20. Asinovsky E.I., Pakhomov E.P. An analysis of the temperature field in a cylindrically
symmetric pole of electric Arc // J. Thermal physics of High Temperatures. 1968. V. 6. No 2.
PP. 333–336.
21. Physics and technology of Low-temperature plasma / Ed. by S.V. Dresvin. M.: Atomizdat.
1972.
22. Incropera F.P. Procedures for modeling laminar cascade Arc behavior // IEEE Transaction on
Plasma Science. 1973. Vol. Ps-I, No 3. P. 3–9.
23. Panasenko L.N. The two-temperature plasma model Arc // Study of plasma devices. Minsk.
1978. PP. 23–29.
24. Averin I.B., Sinyarev G.B. The method of calculating the electric Arc blown laminar gas flow
in a cylindrical channel with chemical nonequilibrium // Proc. of the VII Conf. by generators
of low-temperature plasma. Alma-Ata. 1977. No 2. PP. 23–26.
25. Dresvin S.V. The two-temperature plasma model under steady purge gas through the plasma
torch // J. Technical Physics Letters. 1973. No 4. PP. 3–12.
26. Voropaev A.A., Dresvin S.V. The two-temperature model of laminar Arc in a forced purge
gas through the plasma torch // J. Thermal physics of High Temperatures. 1973. T. 11. No 2.
PP. 333–341.
27. Donskoy A.V., Klubnikin V.S., Salangin A.A. Two-temperature simulation of Argon plasma
in the channel // J. Technical Physics. 1985. V. 55. No 11. PP. 2124–2128.
28. Clarke C.J., Inkropera F.P. Thermochemical nonequilibrium in a stabilized Argon Arc plasma
// J. Rocket and technical complex. 1972. V. 10. No 1. PP. 19–21.
29. Haidar J. Nonequilibrium modeling of transferred Arcs // J. Appl. Phys. 1999. Vol. 32.
P. 263–272.
30. Hsu K.C., Etemadi K., Pfender E. Study of the free-burning high-intensity arcs //
J. App. Phys. 1983. Vol. 54. No 3. P. 1293–1301.
366 8 The Arc Plasma Torches

31. Calculations of a free-burning electric ARC by nonequilibrium model / J. Amouroux, S.

Dresvin, S. Nguyen-Kuok, A.V. Krylov // Proc. of the VI European Conference on Thermal
plasma Processes. Strasbourg, 2000. P. 289–294.
32. The theory of an electrical Arc column / V.S. Engelsht, V.S. Gurovich, G.A. Desyatkov et al.
Novosibirsk: Science. 1990. V.1: The Low-temperature plasma.
33. Kudinov V.V. Bobrov G.V. Coating by spraying: Theory, technology and equipment. M.:
Metallurgy. 1992.
34. Research and optimization of plasma metal cutting / Y.Y. Kiselev, I.A. Protsuk et al.
Kishinev: Shtiintsa. 1981.
35. Kiselev Y.Y. Energy Processes of Air plasma cutting of metals. Kishinev: Shtiintsa, 1980.
36. Donskoy A.V., Klubnikin V.S. Electrical plasma processes and installation engineering. L.:
Engineering, 1979.
37. Schulz-Gulde E. // Paper of Institute of atom and molecular physics.- Hannover. Germany,
1997.
38. AIAA Paper. 1966. No 66(189). P. 1.
39. Donskoy A.V., Klubnikin V.S. // J. of Engineering Physics. 1972. V. 22. No 6.
40. Asinovsky E.I., Kirillin A.V. Encyclopedia of Low-Temperature Plasma: / Ed. Fortov V.E.
V. II. M.: Nauka. 2000.
41. A study of the plasma torch with the rising current-voltage characteristics / A.S. Anshakov,
M.F. Zhukov et al. // News of Sib. Dep. USSR Academy of Sciences. Ser. Tech. Sciences.
1970. No 8. Vol. 2, pp. 3–11.
42. Theory of welding processes // Ed. V.V. Frolov. M.: Higher School, 1988.
43. Lowke J.J. Calculated properties of vertical arcs stabilized by natural convection //
J. App. Phys. 1979. Vol. 50. No 1. P. 147–157.
44. Kovitya P., Lowke J.J. Two-dimensional calculation of welding arc in argon // Urania, I. I. W.
Doc. 1982. Vol. 53. No 11. P. 20–24.
45. Mathematical modeling of electric Arc / Ed. V.S. Engelsht. Frunze: Ilim. 1983.
46. Zhaynakov A., Engelsht V.S. Mathematical modeling of open high-current arc // Generate arc
plasma streams / Ed. M.F. Zhukov. Novosibirsk: Institute of thermal physics of Sib.
Dep. USSR Academy of Sciences. 1987. PP. 323–342.
47. Mechev V.S., Eroshenko L.E. Axial distribution of electric Arc temperature in Argon //
Automatic Welding. 1975. No 6. PP. 14–17.
Chapter 9
Electrode Processes in the Arc Plasma
Torches

In the previous chapter the investigation of the Arc plasma torches (Arc column in
the channel and in the free Arc) was represented based on the model of equilibrium
and nonequilibrium plasmas without considering the near electrode processes. In
these models, the boundary conditions for different parameters in the original
section of the Arc column were determined approximately using the balance of
energy and electromagnetic field (8.37, 8.38) without taking into account the near
electrode processes. However, the near electrode processes and processes in the Arc
column are closely linked: they determine the number of electrons, emitted from the
cathode surface, the dynamics of gas heating in the electric Arc, current, and power
in the cathode region and the Arc column, and others. Therefore, the correct study
of the Arc plasma torches requires an integrated approach joint with consideration
of the physical processes in the electrode regions and in the Arc column.
On the other hand, the effective use of the electric Arc in technology requires the
creation of Arc plasma torches with a long service life that is mainly determined by
the service life of the electrode system. Results of the study of electrode processes,
in addition to identifying the physical picture of the near electrode phenomena helps
to solve the problem of increasing the resource of the installation. In practice, the
experimental study of near electrode processes, there are a number of difficulties
related primarily to the fast physical processes, the presence of the evaporating
electrode material, and zone of the charge separation and others. All of this
determines the necessity of using the theoretical methods for the study of electrode
processes of the electric Arc, especially for Arc plasma torches, operating at
atmospheric pressure in inert and molecular gases.
Theoretical study of electrode processes is often carried out by means of
one-dimensional two-layer model [1–3], that is permissible at length of the elec-
trode region, much less than its radial dimensions. Based on this model, the most
important parameters of the Arc can be found, such as current density and heat
fluxes to the electrodes, the power allocated at them, the voltage drop in the
electrode regions, the electric field strength, and the temperature of electron and the
heavy components of plasma at the cathode and anode. These values can then be
© Springer International Publishing Switzerland 2017 367
S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_9
368 9 Electrode Processes in the Arc Plasma Torches

used to determine the extent of the electrodes erosion, their share of the work, as
well as boundary conditions for the solution of the equation of the energy balance
of the Arc column.
It should be noted that despite the numerous studies devoted to the development
of models and calculation of electrode processes, the results are far from perfect,
many important factors are not taken into account or are difficult to analyze, and
only recently important data and suggestions on the construction of a closed model
of electrode processes appeared. In the description of electrode processes, we will
use mainly on the two-layer model with these latest proposals and try as much as
possible reflect the physical meaning and the role of each process in the model.
Much attention will also be paid to the development of methods of modeling of
electrode processes, procedures of the solutions and analysis of the electrodes
operating in extreme cases.
In practice, at the operating of the Arc plasma torches in inert gases, Nitrogen,
Hydrogen, and mixtures the thermo-emission Arc cathodes are commonly used the
tungsten-based ones [1]. In air, oxygen, nitrogen, and carbon-containing gases
cathodes based on metals (Zirconium, Hafnium, Molybdenum, etc.) are used and
they form the film of the compounds (oxides, nitrides, carbides) in interaction with
plasma forming gases, they have high emission properties and thermal stability. The
data obtained as a result of experimental and theoretical studies show that near
electrode processes represent a group of closely related with each other physical
processes occurring simultaneously in the near electrode region, in the body of the
electrode and on the surface. Changes that occur in any of these three areas will
automatically lead to changes in the other. Therefore, at the study of Arc cathodes it
is necessary to consider the complex collaborative processes in the area of the near
electrode, on the surface and in the material of the electrode.

9.1 Model of the Cathode Processes

Description of the Near Cathode Processes The main function of the near
cathode region is to provide an enabling environment transport of charged particles
between the cathode and the Arc column. A simplified picture of the phenomena in
the near cathode region of the Arc plasma torches can be presented on the basis of
two-layer model (Fig. 9.1).
The first layer, if we count from the cathode surface, is called the space charge
layer. It has a length of the order of the Debye radius and less than the mean free
path of electrons and ions lei . For example, in Argon plasma at atmospheric pressure
over a temperature range Te ¼ ð10  40Þ  103 K, Debye length is rD ¼ 0:05
0:02 µm, lei ¼ 4  1 µm.
In this layer, due to the relatively large proportion of the ion current and the low
mobility of ions excess positive space charge forms that causes the potential jump at
9.1 Model of the Cathode Processes 369

Fig. 9.1 The scheme of the electric Arc

the cathode surface (cathode potential drop). The role of the potential jump near the
surface of the cathode is reduced: to accelerate the ions moving to the cathode; to
ensure a sufficiently high electron emission due to the increased temperature of the
surface bombarded by ions; to create at the surface of the cathode of a strong
electric field, reducing the electron work function and thus facilitates the issuance;
to the acceleration of the electrons emitted by the cathode to the energy levels
needed to provide intensive generation of charged particles in the cathode region.
This is caused by the need to create demand (in terms of energy and current
balances) density of the ion current flowing to the cathode. Through this process,
the proportion of the ion current in the cathode region is much larger than in the
discharge column [1, 4].
The second layer, that separates the space charge layer and the Arc column, is
called ionization layer. In this layer, the condition of quasi-neutrality and intensive
generation of charged particles is performed and intensive generation of charged
particles occurs due to the energy acquired by the electrons in the first layer. From
the second layer to the first one not only ions move, but also the reverse electrons.
Because of the inhibitory effect of the field only a small fraction of the electrons,
that have sufficient energy to overcome the barrier, reaches the cathode surface.
Ions that are accelerated by the electric field reach the surface of the cathode and
that ions are neutralized by electrons of the cathode metal. The surface emitted
electrons and atoms move from the cathode. Atoms, having a first layer without
collision, fall into the second one, reducing the degree of ionization [1, 3, 4]. At the
same time, the electrons accelerated in the first layer, and reached to the second one,
increase gradually the degree of ionization and the concentration of charged par-
ticles. A value of the concentration is close to the equilibrium at the border with
plasma Arc column.
In the second layer there is a redistribution of an electron current. The results of
our calculation shows that the percentage of the electron current Se ¼ je =j increases
from Se  0.6–0.9 at the cathode surface to Se = µe/(µe + µi) in the discharge
column (µe, µi—the mobility of electrons and ions respectively) [5]. As has been
said, the large value of the ion current, supported by the processes occurring in the
370 9 Electrode Processes in the Arc Plasma Torches

second layer, provides formation of the positive charges excess and the potential
jump in the first layer. Thus, the processes in both two layers are closely related,
and the calculation of the near cathode processes should be carried out jointly for
both layers.
Sum of the emission currents of ions and electrons equals the current discharge
and supply of energy to the cathode emitting surface may be implemented due to
ion bombardment, the reverse electrons flow, the plasma radiation, the neutraliza-
tion of the ions on the cathode surface, Joule heat, etc. Energy removal from the
cathode is carried out basically due to the emission cooling, heat conductivity,
atomic evaporation, and thermal radiation from the cathode surface.
Consider the case of the cathode with a completely smooth surface with trans-
verse dimension much larger than the mean free paths of the plasma particles. That
allows the creation of near cathode region model in the one-dimensional
approximation.
In general, a rigorous theoretical study of the near cathode region can be based
on the solution of the Boltzmann kinetic equations for each component of the
plasma (9.1) [6].
!
Ze~
E
~
vs  r r þ  rv fs ðr; vs Þ ¼ Sts ð9:1Þ
ms

where fs—the distribution function of the sth plasma component, s = a, e, i; ~ vs ,

ms—the velocity and mass of the sth plasma component; Z—the charge number of
the sth plasma component; e—the electron charge; ~ E—the intensity of electric field;
Sts—collision term which takes into account both elastic and inelastic collisions
between the particles of sort s with each other and with particles of other varieties.
Kinetic Boltzmann equation is written in the stationary form (9.1), since in all
cases the real-time characteristic of the Arc parameters (current, voltage, gas
pressure) are much longer than the plasma relaxation time. The exact solution of
Eq. (9.1) causes serious difficulties due to the difficulty in determining the collision
term Sts. In practice, the Eq. (9.1) is usually solved by approximate methods
(Chap. 5). Currently, a two-layer model is most relevant and proven for study of
near cathode processes. Here is a description of the model using the method of
integral balances, which allows description of processes in the layer of space
charge, in the ionization layer and on the cathode surface by means of integral
equations on the basis of the current and energy balance.
The Layer of Space Charge In the space charge region with the size of the order
of the Debye radius virtually there are not collisions of particles, and the flow of
charged particles through this area can be considered constant. The potential dis-
tribution in the layer is described by the Poisson equation:
9.1 Model of the Cathode Processes 371

d2 V
e0 ¼ eðne  Zni Þ; ð9:2Þ
dz2

where V—the electrostatic potential associated with the electric field intensity as
E ¼  ddVz , the z-axis is perpendicular to the surface of the cathode (Fig. 9.1);
Z—the charge number of the ions; e0—dielectric permittivity in vacuum.
The solution of this Eq. (9.2) can be obtained by direct integration, if the
velocity distribution of the electrons and ions on the boundary is known. At the
moment there is only a rigorous solution for the case when ions moving dominate
the collisions with the atoms with change their charge [7]. It is usually assumed that
the distribution of electrons in the space charge region is described by the
Boltzmann equation. Zero electrostatic potential is selected at the boundary of space
charge layer, and the ion density is determined by solving the Boltzmann equation
for ions (9.1). As a result, the electric field intensity near the cathode surface can be
found from the well-known McCone equation:
rffiffiffiffiffiffi rffiffiffiffiffiffiffiffi 
4 me mi pffiffiffiffiffiffi
Ec2 ¼ ji  je Ud ; ð9:3Þ
e0 2e Zme

where ji, je—current density of ions and electrons; mi, me—mass of the ions and
electrons, respectively; Ud—the voltage drop in the space charge layer.
Equation (9.3) does not take into account the fluctuations of the electric field
associated with the discrete of charges of the ions. In [8] it is shown that in
conditions of atmospheric pressure Arc (j  105 A/cm2, Ek < 107 V/cm), the
number of additionally emitted electrons by the cathode due to the fluctuations of
the electric field per ion is much less than unit.
The presence of the potential barrier at the metal-vacuum boundary prevents the
output electrons from the metal. So it is necessary to give energy (work function)
for electron to overcome this barrier. There are various types of emission (it
depends on the nature of energy supply): thermionic, secondary electronic due to
ion, photons, and excited atoms bombardment. Due to quantum mechanical effects
arising from the action of the field, there is a possibility of electron tunneling
through a potential barrier (field emission). Consider the current and energy balance
on the surface of refractory Arc cathodes at different mechanisms of electron
emission.
The Current Balance on the Cathode Surface When thermionic emission first of
all it is necessary to carry heat to the cathode, due to which there occupier of the
energy states of the atom lying above the Fermi level occurs. Some of these states
have energy greater than the height of the potential barrier, and electrons from these
levels can leave the cathode surface. Thermionic emission current density is
described by the Richardson–Dushman equation [1, 3, 7]:
372 9 Electrode Processes in the Arc Plasma Torches

 
eu  DðeuÞ
jem ¼ ATc2 exp  ; ð9:4Þ
kTc

where A—Richardson constant, which is determined by experiment; Tc—temper-

ature of the emitting surface of the cathode; eu—the work function of the cathode
material; k—Boltzmann constant; D(eu)—lowering the work function due to the
strong electric field on the surface of the cathode—Schottky effect,
rffiffiffiffiffiffiffiffiffi
eEc
DðeuÞ ¼ e : ð9:5Þ
4pe0
Schottky effect becomes significant when electric field intensity is
Ec  104 V/cm. Work function eu depends on many factors: the chemical com-
position and structure of the material, surface conditions, etc. Existing theoretical
methods do not allow reliably determination its relevance for the particular mate-
rials and working conditions. The main source of information is an experiment.
Data on values of work function of different materials and characteristics of the
cathodes can be found, for example, in the [1–3, 9].
As an example, Fig. 9.2 shows the dependence of the thermionic current density
on temperature and work function. The lowest work function eu = 1.0–1.6 eV
oxide cathodes have widely used in other electronic devices. They are made of
alkaline metal oxide layer (e.g., an oxide sintered body containing BaO and SrO),
deposited on the surface of metal wire or ribbon. Along with barium–strontium
oxide cathodes other types of cathodes, including pure tungsten ones are used in
vacuum devices.
On cold cathodes at very high intensity of electric fields the potential barrier is
narrow and permeable to electrons and due to the tunnel effect field electron
emission mechanism (cold emission) is realized. Relation between the field emis-
sion current density (A/cm2) and the electric field intensity (V/cm) is determined by
the Fowler–Nordheim equation [10]:

Fig. 9.2 The thermionic

currents dependence of the
temperature and the work
function eu of the Arc
cathodes [3]: I Tungsten
(impregnated), II Tungsten
(activated), III
thermochemical, IV Tungsten
(pure), Tm, Tb the melting and
boiling temperature points of
Tungsten
9.1 Model of the Cathode Processes 373

 
Ec u3=2
jem ¼ 1:54  106 exp 6:8  107 : ð9:6Þ
u Ec

Therefore, the field electron emission (9.6) becomes significant only when
Ec > 107 V/cm. The analysis of experimental data on refractory Arc cathodes and
the results of further calculation show that when implemented current densities at
the cathode Arc plasma torches j = 102–104 A/cm2, the electric field intensity
Ec < 107 V/cm. The analysis in [3, 7], also shows that, in the Arc of the secondary
electron emission due to ion bombardment, and photoemission excited atoms can
be neglected. Therefore, in practice for thermionic cathodes are usually limited to
only emission is described by the Richardson–Dushman (9.4).
Velocity of the ions at the boundary of the space charge layer and the ionization
layer vis is determined by the Bohm criterion [1, 7]:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vis ¼ k ðTi þ ZTe Þ=mi : ð9:7Þ

The ion current density at the cathode is

ji ¼ Zenis vis : ð9:8Þ

Ion concentration at the boundary of the space charge region nis can be identified
by the correlation equation [7], obtained from the analysis of the results of
numerical calculation of the multicomponent equations for the ionization layer with
a constant temperature of electrons, atoms, and ions.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
0:8 kTi
nis ¼ ni1 ; a¼ ; ð9:9Þ
2þa mi Di1 kr n2i1

where ni1 —equilibrium ion density at the boundary of the ionization layer and the
Arc column; Di1 ; kr —the diffusion coefficients of the ions and the three-body
recombination at the border of the ionization layer and the Arc column (Sect. 4.4).
To determine the density of the reverse electron current jec, diffusing from the
quasi-neutral plasma to the cathode, it can be assumed that the velocity of the
electrons at the boundary of the space charge region is close to the average velocity
at the electron temperature Te [1, 3, 7] and their Maxwell distribution, then
 
enes
ve eUd
jec ¼ exp  ; ð9:10Þ
4 kTe
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where
ve ¼ 8kTe =pme —the average thermal velocity of the electrons;
nes ¼ Znis —the electron density at the boundary of the space charge region.
374 9 Electrode Processes in the Arc Plasma Torches

Full current density at the cathode is

j ¼ ji þ jem  jec ¼ ji þ je : ð9:11Þ

The Energy Balance at the Cathode Surface Analysis of the energy balance at
the cathode surface plays a key role in the study of the cathode and cathode
processes, because it allows to evaluate the thermal state of the cathode and to
determine its service life. Heating of the cathode is mainly carried out by the flow of
ions, accelerated in the zone of the cathode potential drop. At the same time, the
thermionic emission current determines the cooling of the cathode, and electrons
are emitted ions are involved in the formation of the space charge determining ion
flux to the cathode, and the electric field. Thus, closely interrelated processes take
place in the cathode region and on the cathode surface. For drawing of the energy
balance knowledge of the conditions in the area bind to the surface discharge and
the space charge zone requires, it depends on many factors including the material,
geometry, the cooling conditions, the cathode current level, gas type, gas pressure,
and others.
The energy flux density caused by ion bombardment of the cathode surface can
be obtained by assuming a Maxwell ion velocity distribution in the space charge
layer [7]:

ji
qci ¼ ½kð2Ti þ ZTe =2Þ þ ZeUd : ð9:12Þ
Ze

On the boundary of the space charge layer flow density of kinetic energy of fast
m v2
electrons with kinetic energy e2 e  eUd can be obtained by integrating with
Maxwell function and equal to jece ð2kTe þ eUd Þ. By the time when electron reach the
surface of the cathode its kinetic energy decreases by eUd . Consequently, the
kinetic energy flux density brought by reverse electrons to cathode is:

jec
qec ¼ ð2kTe þ eueff Þ; ð9:13Þ
e

where eueff ¼ eu  DðeuÞ—effective emission work function.

The kinetic energy flux density of the electrons leaving the cathode surface can
be estimated by semi-Maxwell velocity distribution with a surface temperature Tc
[7] and it will be equal to 2jem kTc =e. Therefore:

jem
qem ¼ ð2kTc þ eueff Þ: ð9:14Þ
e
9.1 Model of the Cathode Processes 375

Similarly, the energy flux density of atoms leaving the cathode as a result of the
neutralization of the ions coming from the plasma will be equal to:

2kTc
qa ¼ ji : ð9:15Þ
Ze

Energy released at the cathode as a result of the neutralization of the ions is

Ei  Zeueff , Ei—ionization energy. Accordingly, the flow of energy, evolved at the
cathode as a result of neutralization of ions, equal to:

ji
qii ¼ ðEi  Zeueff Þ: ð9:16Þ
Ze

Energy supply due to the emission of plasma is:

Urad Rl
qpR ¼ ; ð9:17Þ
2ð R þ l Þ

Urad—average volume coefficient of isotropic plasma radiation, it depends on the

type, pressure and temperature of the plasma; l, R—the length and radius of the Arc
column. For long Arcs l > R we have qpR  U2rad R.
The energy flux density of thermal radiation from the surface of the cathode:

qcR ¼ erTc4 ; ð9:18Þ

where r—Stefan–Boltzmann constant; e—the degree of blackness of the cathode

surface, depending on the type of material, surface conditions, temperature level,
and it is determined experimentally.
The energy flux density with drawn during evaporation surface of the cathode:

_
qm ¼ hm; ð9:19Þ

_
where h—specific heat of evaporation of the cathode material; m—specific mass
evaporation rate of the cathode discharge zone bindings.
Experimental data on the speed evaporation of the material in a binding dis-
charge are absent. For estimate we use the results obtained under normal conditions.
Our calculations show that for a tungsten cathode at a temperature in the spot
T = 3900 K, m_ ¼ 103 g/(cm2 s), h = 9600 J/g, qm = 9.6 W/cm2, which is
approximately two orders less than the energy flux due to ion bombardment.
Furthermore, evaporated tungsten atoms can be ionized in the cathode zone, and
return in the form of ions to the cathode.
The energy flux density withdrawn into the body of the cathode by heat con-
ductivity is:
376 9 Electrode Processes in the Arc Plasma Torches

dTk
qk ¼ k ; ð9:20Þ
dz

where k—thermal conductivity of the cathode material.

Thus, the steady-state operation of the cathode when the temperature of the
cathode and plasma does not change during this period of time, the energy balance
on the surface of the cathode can be written as:
   
ji ZTe
q ¼ qk ¼ k 2Ti þ  2Tc þ ZeUd þ Ei  Zeueff
Ze 2
ð9:21Þ
jec jem
þ ð2kTe þ eueff Þ  ð2kTc þ eueff Þ  hm_ þ qpR  qcR :
e e

Evaluations of author and other authors data [1, 3, 7, 8] show that the contri-
bution to the output power flow on the surface of the cathode Arc plasma torches
such streams qpR ; qcR as well as streams of emission and the secondary electron
emission due to ion bombardment (because of the low strength of the field at the
cathode) can be neglected.
It should be noted that when determining the internal temperature of the cathode
body, Eq. (9.21) can be used as boundary conditions for solving the heat con-
duction equation of the cathode.
Ionization Layer The ionization layer is quasi-neutral region, adjacent to the space
charge layer and the Arc column (Fig. 9.1). In this layer, there is intense ionization
and redistribution of the proportion of electronic and ionic currents. The electrons
emitted by the cathode and accelerated in the first layer supply necessary energy for
this process.
The thickness of the ionization layer is of the order of recombination length,
which is defined as the scale at which the rate of change of the concentration of
charged particles due to ambipolar diffusion is comparable to the rate of ionization
and recombination. For a dense plasma in the Arc discharge of high-pressure main
mechanism of decay of charged particles is a triple (three-particle impact-radiation)
recombination with the participation of a third body—the electron [11]. In this case,
qffiffiffiffiffiffiffi
the length of the recombination can be estimated as d ¼ Dkramb n2 , where Damb—the
e

ambipolar diffusion coefficient; kr—the coefficient of triple recombination [5,

12–18]. For the characteristic temperature Te = 10,000–40,000 K, d = 0.1–1 mm.
In some parts inside the ionization layer the fraction of the ion current layer is the
same as in the space charge region. At some distance from the ionization region
(Fig. 9.1) ion current is avalanche reduced to a very small value. Since the flow of
ions leaving, the ionization layer in the space charge region is much greater than the
flow coming from the layer of thermal heating, ionization in the ionization layer
9.1 Model of the Cathode Processes 377

must prevail over recombination, so a deviation from the ionization equilibrium

takes place ðn_ e [ 0Þ.
For a complete description of the processes in the ionization layer it is necessary
to aware of the following phenomena: the movement and interaction of the parti-
cles, the relaxation of the electron distribution function, ionization, and recombi-
nation. Direct solution of the kinetic equations in this case is a difficult task.
Therefore, in the theoretical studies used simplified models to give a mathematical
description of a set of these processes. One of the following two models of the
ionization layer is currently used as a rule.
The first model is based on the integral balance equations. In this model the
ionization layer is assumed infinitely thin [1–3, 5]. This approximation is justified in
cases where the length of the ionization of atoms by electron impact is significantly
less than the mean free path of ions in the plasma. This condition is valid only when
electron temperature in the ionization layer is high.
The second model can be used in those cases where the length of the ionization
layer is much larger than the mean free path. This model is a system of differential
equations describing the processes in the ionization layer. To get this information,
method of integral–differential kinetic equation to a system of equations for the
moments of the distribution function [1] is used, the plasma is considered to be in a
state of partial local thermodynamic equilibrium, hold the Maxwell–Boltzmann–
Saha equations (1.2, 1.4, 1.9). As is known, the first moments of the distribution
functions are normal gas-dynamic variables: the concentration, velocity, tempera-
ture of the particles. Instead of the Poisson equation in the second layer the
quasi-neutrality condition is used ðne ¼ Zni Þ.
Gas-dynamic equations based on nonequilibrium plasma model shown in
Sect. 1.3. General procedure for the solution of these equations is discussed in
Sect. 6.1, and their joint solution and simulation results for the free Arc column
listed in Sects. 8.5.2 and 8.5.3. The use of gas-dynamic model allows us to model
simultaneously all the processes in the ionization layer and the Arc column. In this
case, there is no need to separate the ionization layer (including a layer of thermal
heating) from the main Arc column. The main advantage of this model is the ability
to set all the parameters of the distribution of the length and radius of the ionization
layer, which is not less important. However, the lack of the second model is the
need for the joint solution of equations and review processes in the Arc column.
In this case, to calculate the ionization layer first model including the method of
integral balance is widely used. Its main provisions are represented lower.
The method of integral balances allows description of processes in the ionization
layer through a series of simple integral relations based on the balance of energy
flows in this layer:
1. The energy flux density contributed to the emissive layer by electrons accel-
erated in the space charge layer in accordance with the foregoing is
e ð2kTc þ eUd Þ.
jem
378 9 Electrode Processes in the Arc Plasma Torches

2. Work of the electric field on the electrons moving inside the layer is equal to the
voltage drop in the layer Ui, multiplied by the average density of the electron
current in the layer. According to the work [7], the electron current density is
equal to half the sum of the electron current density at the boundary of the space
charge layer, which is equal to je ¼ jem  jec , and the electron current density at
the outer edge of the ionization layer, which in the first approximation is
j ¼ ji þ je . Thus, the work of the electric field on the electrons moving inside the
layer is jem j2ec þ j Ui , where Ui ¼ kTe e ln nne1
es
—the voltage drop in the ionization
layer [7]; ne∞ and nes—the concentration of electrons in the outer border of the
Arc column and the inner boundary of the space charge layer, respectively.
3. The energy flux density of the electrons introduced into the space charge layer
by electrons leaving the ionization layer jeec ð2kTe þ eUd Þ.
4. The energy flux density of the electrons introduced into the column of the Arc
discharge in plasma by electrons leaving the ionization layer, it is 3:2jkTe =e.
Factor 3.2 is made of coefficient 5/2, taking into account the transfer of enthalpy
due to electric current adjusted thermal diffusion coefficient which is designed
for highly ionized plasma in [7].
5. Electron energy loss in the ionization of atoms equal to jZe i Ei
; losses in elastic
collisions with heavy particles inside the layer is neglected.
Thus, the Energy balance equation for the ionization layer is as follows:

jem  jec þ j ji E i
jem ð2kTc þ eUD Þ þ eUi ¼ jec ð2kTe þ eUd Þ þ 3:2jkTe þ :
2 Z

Excluding the total current density j and substituting in the expression for the
voltage drop in the layer Ui, we obtain the final form of the energy balance equation
in the ionization layer:
     
ne1 ne1
jem eUd þ 2kTc þ kTe ln  3:2 ¼ jec eUd þ kTe ln  1:2
nes nes
  
Ei ne1
þ ji þ kTe 3:2  0:5 ln ;
Z nes
ð9:22Þ

Taking into account the equality Uc = Ud + Ui, we have:

 
Ei
jem ðeUc þ 2kTc  3:2kTe Þ ¼ jec ðeUc  1:2kTe Þ þ ji þ 3:2kTe  0:5eUi :
Z
ð9:23Þ
9.1 Model of the Cathode Processes 379

The remainder of the near cathode region that is external to the ionization layer is a
layer of thermal heating (Fig. 9.1). The thermal heating layer, thermal nonequi-
librium state is maintained as long as the temperature of the electrons and heavy
particles are different.

9.2 Calculation of the Cathode Processes

The Procedure for Calculating the Cathode Region The above Eqs. (9.3–9.23)
allow us to calculate a one-dimensional current transport and energy through the
cathode region of the Arc. The transfer mode is specified by two parameters: the
first—the temperature of the cathode surface Tc, and the second (it should be
asked), for example, the potential difference cathode region Uc or electron plasma
temperature Te. All other parameters describing the transfer current and energy can
be determined by solving the respective equations of the model.
The calculation procedure is as follows, for example: parameters Tc and Te as input
factors and we can solve the system of equations of ionization equilibrium for a
given Te, the temperature of the heavy particles in the plasma and pressure. From
Eq. (9.9), we obtain the concentration of electrons and ions at the boundary of the
space charge layer and the ionization (nes = Znis), and from (9.7) and (9.8), the ion
current density ji in the space charge layer. The Eq. (9.23) can be solved together
with the Eq. (9.10) and the Richardson–Dushman equation (9.4) for define the
voltage drop in the space charge layer Ud, the reverse current density jec and the
current density of electrons thermionic jem. It should be emphasized that in this
procedure, the decision should be a joint decision procedure McCone equa-
tion (9.3), the formulas of work effectively (9.5) and Richardson–Dushman equa-
tion (9.4). Finally, we give the current densities and all kinds of heat fluxes to the
cathode, and their total values are given by (9.11), (9.21), respectively.
This solution corresponds to a one point of the current–voltage characteristics
(CVC) of the cathode region. For other values of Tc and Te should obtain the
following points CVC, etc. The obtained data allow us to identify the integral
characteristics of the cathode region, such as the power dissipated in the cathode,
the rate of evaporation of the cathode material and the radius of the cathode spot as
a function of the Arc current.
An analysis of the calculation results with the experimental data of other authors
[1, 2, 19] (Fig. 9.3) allows us to conclude the approximate constancy of the current
density in the cathode spot when changing the Arc current. Persistence several
variables including total current density along the radius of the cathode spot allows
to determine the cross section Ss and the radius Rs of the cathode spot in certain of
the Arc current and its density:
380 9 Electrode Processes in the Arc Plasma Torches

Fig. 9.3 Dependence of

current density tungsten
cathode from the current [19]:
dc = 5 mm—diameter of the
cathode, lc = 60 mm—length
of the cathode output portion:
1 argon, 2 nitrogen; dc = 5–
8 mm, lc = 0: 3 argon, 4
nitrogen, 5 hydrogen

sffiffiffiffi
I I
Ss ¼ ; Rs ¼ : ð9:24Þ
j pj

To calculate the Arc current is given in the region of 50–1500 A, and the current
density is taken for the characteristic mode of the thermionic cathode of the electric
Arc at atmospheric pressure. For example, for the tungsten cathode W + Ar:
Tc = 4200 K, Te = 40,000 K; for thoriated tungsten cathode W − ThO2 + Ar
(1.5 % ThO2): Tc = 3200 K, Te = 40,000 K.
Power to the cathode, respectively, equal to

I
P ¼ qSs ¼ q : ð9:25Þ
j

Volt equivalent of the heat flux to the cathode is defined as:

V ¼ q=j: ð9:26Þ

Results of Calculation of the Cathode Region of an Electric Arc We present the

results of calculations cathode region of pure tungsten cathode W and thoriated
tungsten cathode W(1.5 % ThO2), typical for the majority of Arc plasma torches
operating at atmospheric pressure in Argon. In the first case, the Richardson
coefficient A = 6.02  105 A/(m K)2, the work function eu = 4.55 eV, the melt-
ing temperature Tm = 3700 K. For the second case, A = 3.00  104 A/(m K)2,
9.2 Calculation of the Cathode Processes 381

eu = 2.63 eV, the melting temperature Tm = 2920 K [3]. Table 9.1 gives the
values of the work function of a tungsten cathode with the addition of various oxide
additives.
On the outer boundary of the cathode region at atmospheric pressure and a
relatively high degree of ionization and electron temperature (Te = 20,000–
40,000 K) the Argon plasma temperature of the ions, atoms, and electrons are the
same (Te = Ti = Ta). For a pure tungsten cathode W operating temperature range of
the cathode surface is selected in accordance with the experimental data of [1, 3, 19]
Tc = 3000–4400 K. For thoriated tungsten cathode W(1.5 % ThO2): Tc = 2000–
3200 K.
The dependence of the electric field intensity at the cathode surface temperature
of the cathode and the plasma temperature is shown in Figs. 9.4 and 9.5. The
electric field intensity decreases with increasing temperature of the cathode; this is
due to the fact that as the temperature increases thermionic cathode of the cathode.
As a result, for a given current of maintaining the Arc does not need a large
amplification of the electric field. Similarly, an increase in the plasma temperature at
the outer boundary of the cathode region of the electric field decreases, which is
consistent with the physical process of the thermionic cathode.

Table 9.1 The work Oxide eu (eV) Oxide eu (eV)

function of tungsten oxide
with a mass content of 1– ThO2 3.00 Gd2O3 3.96
1.5 % [3] La2O3 2.96 Yb2O3 4.05
Y2O3 3.30 ZrO2 3.97
Ce2O3 3.40 HfO2 3.84
Nd2O3 3.88 – –

Fig. 9.4 The electric field intensity at the cathode depending on the surface temperature of the
cathode at different temperatures of the plasma: a W + Ar; b W − ThO2 + Ar
382 9 Electrode Processes in the Arc Plasma Torches

Fig. 9.5 The electric field intensity at the cathode surface, depending on the temperature of the
plasma at different values of the surface temperature of the cathode: a W + Ar; b W − ThO2 + Ar

Because of the improved thermionic thoriated tungsten cathode has a value of

the electric field as a pure tungsten cathode at a much lower temperature of the
cathode (Tc = 2000–3200 K compared to Tc = 3000–4400 K).
In the working temperature range of the cathode (Tc = 3000–4400 K and
Tc = 2000–3200 K, respectively, for W + Ar and W − ThO2 + Ar) and in a wide
range of plasma temperatures (Te = 20,000–40,000 K) value of the electric field
intensity E ¼ 105  2  106 \107 V/cm. This confirms the assumption of low
current field emission (9.6) and secondary electron emission at the electric Arc
(photoemission can be neglected).
The voltage drop in the cathode region is an important parameter in the selection
of the cathode unit. Voltage drops of the cathode surface temperature in the space
charge layer Ud, in the ionization layer Ui and in the cathode region Uc are shown in
Fig. 9.6. The conventional wisdom is that the voltage drop in the cathode region is
mainly in the space charge layer due to the high electric field intensity at the
cathode. However, the results of our calculations show that this is true only at
relatively low temperatures of the cathode and to the cathode temperature near the
melting temperature the voltage drop in the ionization layer can be more than the
voltage drop in the space charge layer. When the cathode temperature is near the
melting temperature, i.e., close to the operating point of the thermionic cathode,
high current thermionic emission does not need a large voltage drop in the space
charge layer to enhance the current (Fig. 9.6). The voltage drop in the ionization
layer mainly depends on the properties of plasma, and therefore Fig. 9.6, it has a
constant value.
The dependence of the voltage drop in the cathode region of the plasma tem-
perature is shown in Fig. 9.7. With increasing temperature, the plasma obtained a
slight increase in the voltage drop in the ionization layer and the cathode,
9.2 Calculation of the Cathode Processes 383

Fig. 9.6 The voltage drop in the space charge layer Ud, in the ionization layer Ui and cathode
region Uc depending on the temperature of the cathode surface: a W + Ar; b W − ThO2 + Ar

Fig. 9.7 The voltage drop in the cathode region, depending on the electron temperature for
different values of the cathode surface temperature: a W + Ar; b W − ThO2 + Ar

respectively, in the field. We see that in both cases the thermionic cathode voltage
drop in the cathode region of the order of *10–30 V, which coincides with the
experimental data [1, 3].
Experimental data on the reduction of the work function of the cathode is
practically absent, this value is mainly determined only by calculation. Figure 9.8
shows the dependence of the work reduction cathode on its surface temperature at
different values of the electron temperature, and Fig. 9.9 shows the dependence of
the work reduction cathode on the electron temperature. The average value
DðeuÞ * 0.2–0.3 eV, it is about 10 % of the work function. However, both values
384 9 Electrode Processes in the Arc Plasma Torches

Fig. 9.8 Reducing the work function of the cathode, depending on the surface temperature of the
cathode at different temperatures of the electrons: a W + Ar; b W − ThO2 + Ar

Fig. 9.9 Reducing the work function of the cathode, depending on the temperature of the
electrons for different cathode surface temperatures: a W + Ar; b W − ThO2 + Ar

included in the calculation of the exponent of thermionic emission current density,

so that the accuracy of calculation of the current density and thermionic heat flux
depends on the accuracy determination of value DðeuÞ: Since DðeuÞ directly
proportional to the electric field at the cathode (9.5), the variation depending DðeuÞ
of the temperatures of the cathode and the electrons (Fig. 9.9) is the same as that of
the electric field intensity (Figs. 9.4, 9.5).
The most important parameter in the calculation of the cathode region is the
current density on the cathode surface. The balance of currents in dependence on
the temperature of the cathode surface shown in Fig. 9.10, data from our calculation
9.2 Calculation of the Cathode Processes 385

Fig. 9.10 Current density balance on the surface of the cathode, depending on the surface
temperature of the cathode at a plasma temperature Te = 40,000 K: a W + Ar; b W − ThO2 + Ar

are practically identical with those of other authors in Fig. 9.2. In the working
temperature range of both cathodes the current density is *102–104 A/cm2. We
also see that in the working range of the temperature of the cathode surface and the
electron, thermionic emission current is larger proportion in comparison with the
current of reverse electron and ion current. In the area of the cold cathode (for
W + Ar, Tc < 3000 K and for W − ThO2 + Ar, Tc < 2000 K) thermionic emission
current and the current of reverse electrons are small, and the ion bombardment of
the surface of the cathode current prevails, due to the high electric field at the
cathode (Fig. 9.4) and the voltage drop (Fig. 9.6), and plays a major role in the
formation of the Arc current; for j < 100 A/cm2 cathode is gradually sinking into
the cold emission region, where it is necessary to take into account of the field
emission and secondary emission.
The total current density at the cathode depending on the electron temperature
(for different temperatures of the cathode surface) is shown in Fig. 9.11. For the
same surface temperature of the cathode current density decreases with increasing
the plasma temperature due to increased current density reverse electrons. From the
data in Figs. 9.10 and 9.11 we can see that the current density at the cathode
depends largely on the temperature of the cathode and is less dependent on the
temperature of the plasma.
The proportion of electronic (thermionic emission and reverse electrons) and the
ion current as a function of the temperature of the cathode surface for different
values of the electron temperature is shown in Fig. 9.12. We see that a large
proportion of the ion current at the cathode ðSi ¼ ji =j [ 10 %Þ is more than in the
Arc column, it is the result of the work of most of Arc plasma torches atmospheric
pressure. How to show the data on the current balance (Fig. 9.10) and the pro-
portion of electron and ion currents (Fig. 9.12) on the cathode surface, for the
region of low temperature (cold cathode and cold emission), the proportion of ion
386 9 Electrode Processes in the Arc Plasma Torches

Fig. 9.11 The total current density on the cathode as a function of electron temperature, at
different temperatures of the cathode surface: a W + Ar; b W − ThO2 + Ar

Fig. 9.12 Proportion of the electronic and ionic current as a function of the cathode surface
temperature for different values of plasma temperature: a W + Ar; b W − ThO2 + Ar

current even prevails over the electron current. At the operating temperature region
of thermionic cathodes, ion current can be up to 30–50 % of the total current at the
cathode.
Changing the proportion of the currents depending on the electron temperature
(Te = 20,000–40,000 K) does not exceed 20 %, and the higher the electron tem-
perature, the greater the proportion of electron current, and good thermionic
emission of electrons from the cathode surface may be provided by a high tem-
perature of the cathode (close to the temperature melting) and the electrons
9.2 Calculation of the Cathode Processes 387

(Te = 20,000–40,000 K). This is the condition of most thermo-emission cathode

Arc plasma torches operating at atmospheric pressure for various technological
applications (plasma cutting and melting of metals, plasma spraying and hardening
of machine parts, plasma treatment of surfaces and materials, plasma waste dis-
posal, etc.).
The energy balance at the cathode surface, depending on the temperature of the
cathode is shown in Fig. 9.13. The energy fluxes determined from the current
density and the nature of the change of these two values, depending on the tem-
perature of the cathode and the plasma is the same. We note an important role of the
heat flux of reverse electron in the energy balance at the cathode. The calculation
result shows that the operating temperature range of the cathode at the electric Arc,
when the temperature of the cathode is more than the melting point of the material,
the energy of the reverse electron can be more, than the energy of the ion bom-
bardment energy and thermionic emission (Fig. 9.13). This is so due to the fact that
the introduction of “hot” electrons from the plasma to the lattice of the cathode
material recovered their work function (eueff) and kinetic energy (2kTe). As a result,
the total flow of heating the surface of the cathode may be larger than the flux of ion
bombardment (Fig. 9.13).
Additional information can be obtained from analysis of the energy balance of
the cathode depending on the electron temperature of the plasma (Fig. 9.14). These
calculations show that the energy flux of atoms leaving the surface of the cathode,
an order less than the other fluxes and in the further it can be ignored. The total

Fig. 9.13 The energy balance on the cathode surface, depending on the temperature of the
cathode at the electron temperature Te = 40,000 K: qem —thermionic flux from cathode; qec —the
energy flux density of reverse electrons to the cathode; qi —the energy flux density of ions to the
cathode, including the energy of recombination energy carried atoms by neutralization of the ions
on the cathode surface: a W + Ar; b W − ThO2 + Ar
388 9 Electrode Processes in the Arc Plasma Torches

Fig. 9.14 Energy balance at the cathode, depending on the electron temperature: qci —the energy
flux density of ion bombardment to the cathode; qii —the energy flux density resulting from the
neutralization of the ions on the cathode surface; qa —the energy flux density of atoms leaving the
surface of the cathode; a W + Ar, Tc = 4000 K; b W − ThO2 + Ar, Tc = 3200 K

energy flux to the cathode of electric Arc weakly depends on the electron tem-
perature of the plasma and is about 5–10 kW/cm2, which coincides with the
experimental data of [1, 3]. The results of our calculations show that the total
energy flux to the cathode is largely dependent on the temperature of the cathode Tc
and to a lesser extent on the electron temperature Te of the plasma. For the majority
of thermionic cathode Arc plasma torches in the operating temperature range of the
cathode and the plasma total energy flux to the cathode is q = 1–10 kW/cm2.
Parameter of the cathode spot is also the result of calculating the cathode region
and is great practical importance for the selection and design of the cathode. Volt
equivalent of the energy flux to the cathode, depending on the temperatures of the
cathode and the electrons is shown in Fig. 9.15. At the melting and boiling points of
the cathode material temperature the volt equivalent is order *10–1 V.
Figure 9.16 shows the dependence of the power to the cathode and the radius of
the cathode spot in the current range of electric Arc (I = 50–1500 A). Good
agreement between calculated and experimental data leads to the conclusion about
the correctness of the model and the results of calculation and the linear dependence
of the power to the cathode of the Arc current allows the use of equivalent volt
rapid assessment capacity in the cathode.
To determine the extent of erosion of the cathode in Fig. 9.17 presents the data
of the specific mass evaporation rate and heat flux of different cathode materials
depending on its temperature.
9.2 Calculation of the Cathode Processes 389

Fig. 9.15 Volt equivalent of the energy flux to the cathode, depending on the temperatures of the
cathode (a) (with Te = 40,000 K) and the electrons (b) (with Tc = 3200 and 4000 K)

Fig. 9.16 Power transmitted to the cathode (a) and the radius of the cathode spot (b), depending
on the Arc current: lines the calculation data; point the experiments; 1 lc = 0 [20]; 2 lc = 0, 3
lc = 0.6 cm, 4 lc = 1.2 cm, 5 lc = 6 cm [19]

Tungsten cathode is usually composed of tungsten with a small proportion of

different impurities. For this reason, the Richardson’s coefficient A of the Eq. (9.4)
has a different meaning, but in practice this factor is determined experimentally by
measuring the current density of the Arc. We present the results of calculations of
the current and energy balances in the tungsten cathode region with the
390 9 Electrode Processes in the Arc Plasma Torches

Fig. 9.17 Specific mass evaporation rate and heat flux at the cathode material evaporates
depending on its temperature: 1—W + Ar; 2—Cu + Ar

Richardson’s coefficient A = 2  106 A/(m K)2 and thoriated tungsten cathode

with A = 1.6  105 A/(mK)2 of Arc plasma torches operating at atmospheric
pressure with Argon (Figs. 9.18, 9.19).

Fig. 9.18 The current balance on the surface of the cathode depending on the surface temperature
of the cathode at a plasma temperature Te = 40,000 K: a W + Ar; b W − ThO2 + Ar
9.3 Thermal Problem of Heating the Cathode 391

Fig. 9.19 The energy balance on the cathode surface, depending on the temperature of the
cathode at the plasma temperature Te = 40,000 K: the notation is the same as in Fig. 9.13

9.3 Thermal Problem of Heating the Cathode

Phenomena inside the cathode play a significant role in the formation of its thermal
state. To determine the thermal state of the cathode the energy equation, should be
solved, that in the stationary case and a cylindrical axis-symmetry coordinate
system has the following form:
   
@ @T 1@ @T
 k  rk ¼ rE 2 ; ð9:27Þ
@z @z r @r @r

where k, r—thermal conductivity and electrical conductivity coefficients of the

cathode material, depending on temperature of the cathode; E—the electric field
intensity.
Distribution of the specific power released in the form of Joule heating and
current density inside the cathode can be found using equations for the electric
potential V (8.19).
   
@ @V 1@ @V
r þ rr ¼ 0;
@z @z r @r @r
and
" 2  2 # s
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
  ffi
@V @V @V 2 @V 2
S ¼ rE ¼ r
2
þ ; j¼ r þ r : ð9:28Þ
@z @r @z @r

The boundary conditions for the energy Eq. (9.27) and the electric potential
(9.28) have the following form:
392 9 Electrode Processes in the Arc Plasma Torches

(a) At the hot end of the cathode (z = 0):

@T q; r  Rcs ;
k ¼
@z z¼0 qcR ; Rcs \r  Rc :
(
@V I
2 ; r  Rcs ;
r ¼ pRcs
@z z¼0 0; Rcs \r  Rc :

Here Rcs—the radius of the cathode spot; Rc—radius of the cathode; q, qcR —the
total flux in the cathode spot and the radiation flux from the cathode are
determined by the Eqs. (9.21) and (9.18), respectively.
Because of the complex picture of the near electrode processes more reliable
boundary condition for the Eq. (9.27) is the experimental temperature distri-
bution on the surface of the cathode: Tc ðrÞ ¼ Tc ðrÞexp ;
(b) On the cold end (z = L): T = 300 K, V = 0;
(c) On the axis of symmetry of the cathode (r = 0): @T @V
@r ¼ @r ¼ 0;
@T @V
(d) The side surface of the (r = Rc): @r ¼ @r ¼ 0. For a water-cooled cathode
temperature and electric potential on the cathode surface can be defined or
determined experimentally.
A discrete analogue of the Eq. (9.28) given in Sect. 8.4.1. For the solution of the
energy Eq. (9.27) it is possible to use the well-known method of control volume,
described in Sect. 6.1. With the above boundary condition the Eqs. (9.27, 9.28) can
be solved on the grid line [z1, r2], shifted by a half step along the radius relative to
the grid [z1, r1] (Fig. 9.20). The counting of grid coordinates [z1, r1] is from the
center on the surface of the cathode (z10 = r10 = 0).
Power balance of the control volume and its discrete analog can be obtained by
integrating an Eq. (9.27) around all of the each control volume (Fig. 9.20). Where

Fig. 9.20 Grid lines and

control volume for the
solution of the energy
equation of the cathode
9.3 Thermal Problem of Heating the Cathode 393

the source function is constant inside control volume DV ¼ ðze zw Þðrn r
2
s Þðrn þ rs Þ
and it
is a linear function of temperature S ¼ Sc þ Sp T, and the coefficients of Sc and Sp
are defined like (6.56). We have:

Zrn Zze   Zze Zrn   Zze Zrn

@ @T 1@ @T
 k dzrdr  rk rdrdz ¼ Srdrdz
@z @z r @r @r
rs zw zw rs zw rs

The integral of each term on the left side of this equation contains the first
derivative of the same order, so you can accurately produce its integration at least
once, we finally obtain:

Zrn   
  Zze    
@T@T @T @T
  k
k rdr  rk  rk dz
e @z@z w @r n @r s
rs   z w

¼ ScM þ SpM TM DV:

In order to implement the integration of the last equation admits that her inte-
grands take the average values in the points of e, w, n, s, respectively, and the first
derivatives are approximated as:
(   
TE TM TM TW
k @T @T
@z e ¼ ke zE zM ; k @z w ¼ kw zM zW ;
 @T   
k @r n ¼ kn TrNN T @T TM TS
rM ; k @r s ¼ ks rM rS :
M

Substituting these expressions in the last integral equation, we obtain a general

discrete analog for cathode energy Eq. (9.27) in the form:

aM TM ¼ aE TE þ aW TW þ aN TN þ aS TS þ bM ; ð9:29Þ

where
8
ke ðrn rs Þðrn þ rs Þ
>
< aE ¼ 2ðzE zM Þ [ 0; aW ¼ kw ðr2nðr s Þðrn þ rs Þ
zM zW Þ [ 0;
kn rn ðze zw Þ ks rs ðze zw Þ
> a ¼ rN rM [ 0; aS ¼ rM rS [ 0;
: N
aM ¼ aE þ aW þ aN þ aS  SpM DV [ 0; bM ¼ ScM DV:

Note that all the discrete analog coefficients of (9.29) are positive and satisfy the
convergence criterion (6.49). So for example, the temperature distribution inside the
body of tungsten is shown in Fig. 9.21.
394 9 Electrode Processes in the Arc Plasma Torches

Fig. 9.21 The temperature

distribution inside the
tungsten cathode
(dcs = 6 mm, dc = 10 mm)

Diffusion of Additive Material to the Surface of the Cathode Introduction of

activating additives (Table 9.1), reducing the work function, leads to a decrease in
the required operating temperature of the cathode. However, the rapid diffusion of
the additive material to the surface of the cathode and thus of its concentration
within a short period of time causes an increase in the work function and corre-
spondingly increase the temperature of the cathode surface.
In this regard, the main objective of increasing the cathode performance and
service life of the electric Arc is the reduction in the rate of release of the activating
additive outside and to ensure effective recharge its surface layers from the deep
layers of the cathode material [1].
In theoretical terms, this problem can be reduced to the study processes of
diffusion additives within the body of the cathode, output it to the surface and
evaporation from it. The corresponding diffusion equation of additives from the
depths of the cathode is:
   
@n @ @n 1@ @n
¼ D þ rD ; ð9:30Þ
@t @z @z r @r @r

where n—the concentration of the additive in the cathode; D—diffusion coefficient

of the additive.
Equation (9.30) has the general form of nonstationary heat conduction equation
and can be solved on the grid [z1, r2] (Fig. 9.20). A discrete analog of this equation
for a fully implicit scheme [21] is as follows:

aM n M ¼ aE n E þ a W n W þ a N n N þ a S n S þ b M ;
8
>
> a ¼ De ðr2n ðr s Þðrn þ rs Þ
[ 0; aW ¼ Dw ðr2nðr s Þðrn þ rs Þ
[ 0;
< E zE zM Þ zM zW Þ
ð9:31Þ
Dn rn ðze zw Þ Ds rs ðze zw Þ
> aN ¼ rN rM [ 0; aS ¼ rM rS [ 0;
>
:
aM ¼ aE þ aW þ aN þ aS þ DV Dt [ 0; bM ¼ DV Dt nM :
0
9.3 Thermal Problem of Heating the Cathode 395

where DV ¼ ðze zw Þðrn r

2
s Þðrn þ rs Þ
—the control volume; Dt—time step; n0M —the value
of the concentration of the additive in the node M at the previous moment;
nM ; nE ; nW ; nN ; nS —additive concentration in the respective nodes of the control
volume (Fig. 9.20) in the given time.

9.4 Model of the Anode Processes

In the cathode region more than 99 % of the electrons that are incident on the
surface of the anode, comes to the potential drop of the arc column and serve as
charge carriers in this area. This fact is illustrated by the value of specific erosion
(average 5  10‒7 g/C) of copper water-cooled cylindrical anode used in the Arc
plasma torches. Even if we assume that all atoms that are evaporated from the anode
surface are fully once ionized, the ion current will be a few one of thousandths of
the total current [20]. Thus, the Arc area immediately adjacent to the anode and the
charges are transferred only by electrons, resulting in the formation in this region of
the negative space charge. It is this uncompensated charge that determines the
anode potential drop. It usually decreases with increasing current and at the tran-
sition to the currents in the hundreds and thousands of amperes is 5–1 V [22]. The
value of the anode potential drop is generally determined by experiment, the data
about it can be found in [20].
Near the surface of the anode separation of the electron temperature and the
temperature of ions and neutral particles is observed and, consequently, the breach
of thermodynamic equilibrium occurs. If it is assumed that plasma is in a state of
PLTE, the equation of energy balance for the anode can be written as follows:
 
je 5kTe
qa ¼ eUa þ eueff þ þ qpR þ qc þ qT  qaR  qm : ð9:32Þ
e 2

In this case, as for the energy balance at the cathode, the below relations will be
shown.
1. The energy flux of electrons that was brought to the surface of the anode,
 
je 5kTe
qea ¼ eUa þ eueff þ ;
e 2

where the first term je Ua corresponds to the kinetic energy obtained by electrons
in the near anode voltage drop, the second je ueff —the potential energy of
electrons arriving at the anode surface and neutralizing the positive ions of metal
in the entrance to the lattice eueff ¼ eu  DðeuÞ—effective work function,
qffiffiffiffiffiffiffi
DðeuÞ ¼ e 4pe eEa
0
—reduction of the work function due to the electric field (Ea)
396 9 Electrode Processes in the Arc Plasma Torches

on anode surface—the Schottky effect; third term determines the enthalpy of the
electron, that corresponds to the temperature of the plasma on the boundary of
the near anode potential drop and the energy acquired by electrons in motion
inside the area. Last energy, reflecting the contribution of the thermal energy of
the electrons, is somewhat different from 32 kTe , and electrons cannot save the
Maxwell distribution in the area of near anode potential drop [20].
2. The energy flux of radiation from the plasma to the surface of the anode
qpR ¼ 2UðRradþRllÞ, where R, l—the average radius and Arc length.
3. The energy flux of radiation from the surface of the anode qaR ¼ erTa4 , where e—
the degree of blackness of the anode; r—the Stefan–Boltzmann constant; Ta—
the surface temperature of the anode.
4. The energy flux brought to the surface of the anode by plasma convection
qc ¼ qvz Hp St, where qvz—specific mass flow of plasma; St—the Stanton’s
number; vz—the plasma velocity at the distance of the mean free path of the
anode surface; Hp—the plasma enthalpy of its temperature at the anode surface.
5. The energy flux brought to the surface of the anode due to heat conduction can
be expressed in the case of large temperature gradient in the boundary layer
through the difference of enthalpies [20]:

ka Hp  Ha
qT ¼ ;
cpa Dz

where Ha ; ka ; cpa —the enthalpy, thermal conductivity and specific heat of

plasma, respectively at a temperature of the surface anode; Dz—the average
length of near anode layer.
_ where h—
6. The energy flux used to evaporate the material of the anode qm ¼ hm,
_
the specific heat of vaporization of the anode material; m—the specific mass
evaporation rate of the anode material.
One of the most important technological parameters using Arc plasma torches is
the power transmitted to anode. To find this parameter, it is necessary to integrate
the energy flux to the anode (9.32) for the cross section of anode. An approximate
equation for this power is:
 
5kTe
Pa ¼ I Ua þ u þ þ PpR þ Pc þ PT  PaR  Pm : ð9:33Þ
2e

If the contribution of the last terms in (9.33) is small, then we have a linear
relationship between the power transmitted to the anode and the current Arc.
Representation about the power transmitted to the anode at atmospheric pressure
can be obtained from the experimental data given in Fig. 9.22. Most details about the
power are represented in [19, 20, 23]. The results of these experiments [19, 23, 24]
have showed that the power slightly depends on the electrode material and the anode
is a merely passive collector of current. Volt equivalent of the heat flux to the anode
9.4 Model of the Anode Processes 397

Fig. 9.22 The power

transmitted to the anode [19]:
1 in the air; 2 in the nitrogen;
3 in the argon (contracted
binding); 4 the same (diffuse
binding)

is not dependent on the current and in diffuse attachment of the Arc in argon is
5.85 V (Fig. 9.22, point 4). In the [25] some the less value of volt equivalent for the
energy flux through the anode spot to the copper electrode (5 V) is indicated at the
range of the current I = 10–200 A.
In the contracted binding the volt equivalent of power transferred to anode
increases to 7.35 V (Fig. 9.22, point 3). The current density at the anode in a diffuse
attachment of the Arc to the anode j  103 A/cm2, and at the transition to the
contracted mode it increases to j  104–105 A/cm2 [19]. Experiments [19] show
that in nitrogen and in the air the anode binding is always contracted and volt
equivalent of heat flux approximately 12.5 V (Fig. 9.22, point 1, 2). The energy
flux density in this case reaches 105–106 W/cm2. Thus, for the model of anode
region the role of the last terms in Eq. (9.33) is small and it can be neglected in the
calculation of energy fluxes toward the anode.

References

1. Near-electrode processes in an Arc discharge / M.F. Zhukov, N.P. Kozlov, A.V. Pustogarov
et al. Novosibirsk: Science. 1982.
2. Thermochemical cathodes / M.F. Zhukov, A.V. Pustogarov et al. Novosibirsk: Science. 1985.
3. Gordeev V.F., Pustogarov A.V. Thermionic cathodes Arc. M.: Energoatomizdat. 1988.
4. The Encyclopedia of Low-Temperature Plasma / Ed. V.E. Fortov. V. I. M.: Nauka, 2000.
5. M. Mitchner, C. Kruger. Partially ionized gases. M.: Mir, 1976.
398 9 Electrode Processes in the Arc Plasma Torches

6. Kozlov N.P., Khvesyuk V.I. On the theory of cathode processes Arcs // J. Technical physics.
1971. Part I: pp. 2135–2141; Part II: pp. 2142–2150.
7. Belinov M.S., Marotta A. A model of the cathode region of atmospheric pressure Arcs //
J. App. Phys. 1995. Vol. 28. PP. 1869–1882.
8. Mojzes B.J., Nemchinskii V.A. On the issue of abnormally high emission of non-evaporable
hot cathode Arc discharge // J. Technical physics. 1974. V. 14. PP. 2539–2547.
9. Gavryushenko B.S., Pustogarov A.V. Research electrodes of plasma torches // Near-electrode
processes and erosion of electrodes of plasma torches / Ed. M.F. Zhukov. Novosibirsk:
Institute of thermal physics. 1977. PP. 95–122.
10. Ecker G. Questions of vacuum Arc theory // Vacuum Arc: Trans. from English. / Ed.
J. Laffetri. M.: Mir. 1982. PP. 267–384.
11. Alievsky M.Y., V.M. Zhdanov. The transport equations for a many-component
non-isothermal plasma // J. Appl. Phys. 1963. № 5. pp. 11–17.
12. Hinnov E., Hirchberg J.G. Electron-Ion recombination in dense plasmas // Phys. Rev. 1962.
Vol. 126. P. 795.
13. Atomic and Molecular Processes / Ed. by D. Bates. M.: Mir, 1964.
14. Bates D.R., Kingston A.E. Collisional-radiative recombination at low temperatures and
densities // Proc. Phys. Soc. 1963. Vol. 83. P. 43.
15. Biberman L.M., Vorobiev V.M., Yakubov N.T. Nonequilibrium Low-temperature plasma IV
ionization and recombination functions // J. Thermal physics of High Temperatures. 1969. V.
7. № 4. PP. 593–603.
16. Aleskovsky Y.M. Study volume recombination in a Cesium plasma // J. Experimental
technical physics. 1963. V. 17. P. 570.
17. Cool T.A., Zukoski E.E. Recombination, ionization and nonequilibrium electrical conduc-
tivity in seeded plasmas // Phys. Fluids. 1966. Vol. 9. P. 780.
18. Desai S.V. Corcoran W.H. Recombination of electrons and ions in an atmospheric argon
plasma // J. Quant. Spectrosc. Radiat. Transfer. 1969. Vol. 9. PP. 1371–1386.
19. Near-electrode processes and erosion of electrodes of plasma torches // Collected Articles /
Ed. M.F. Zhukov. Novosibirsk: Institute of thermal physics. 1977.
20. Zhukov M.F., Koroteev A.S., Yuryukov B.A. Applied Dynamics of thermal plasma.
Novosibirsk: Nauka. 1975.
21. Patankar S. Numerical methods for solving problems of heat transfer and fluid dynamics: M .:
Energoatomizdat 1984.
22. Finkelberg V., Mekker G. Electric Arcs and thermal plasma. M.: Publishing House of Foreign
Literature, 1961.
23. A study of thermal, electric characteristics of plasma and erosion of the anode /
Jeenbaev J. et al. // News of Sib. Dep. USSR Academy of Sciences. Ser. Tech.
Dandaron B., Urbak E.K. Investigation of the thermal regime of the rod tungsten cathode
// Abs. V Conf. by generators of Low-temperature plasma. Novosibirsk. 1972. Part II. PP. 40–
43.
24. Sciences. 1973. № 3. Vol. 1. PP. 3–6.
25. Bykhovsky D.G. Plasma cutting. L.: Engineering. 1972.
Chapter 10
Heat Exchange and Movement of the Solid
Particles in the Plasma

The high-energy plasma flow is widely used for the treatment of particles and for
spraying. Elucidation of the physical pattern of heating and movement of particles
in plasma, as well as determination of the technological parameters of the process,
such as the thermal efficiency, the degree of heating, temperature and velocity of
particles in plasma, the optimal heating regime can be realized based on a study of
heat transfer and movement of solid particles, including their phase conversion in
the plasma flow. This problem is particularly important in connection with the
development of a new branch of plasma chemistry that uses a plasma stream for
recycling and disposal of chemical, medical and industrial waste, chemicals warfare
and toxic substances. Recently a unique device are designed to solve such problems
on the basis of RF and arc plasma torches with hundreds kW of power, working
with such mediums as air, methane and propane. Study of heat transfer and
movement of solid (dust) particles in plasma is also important to study the aero-
dynamics of supersonic gas-dispersion streams of plasma, and determination of the
structure and development of atmospheres of many galaxies, and to simulate the
conditions of entry of spacecrafts and satellites systems in the dense atmosphere of
the Earth.
Description of the processes of heat exchange and movement of solid particles in
plasma using the existing concepts is possible in principle in the preparation of the
complete system of equations for conservation of mass, energy and momentum,
using the chemical reactions and phase transitions. However, at the solving of such
a system of equations for even more simple than plasma objects there are serious
difficulties. Therefore, the choice of the mathematical formulation of the task should
be determined as the results of preliminary analysis of the role of individual factors
and the complexity of calculations from that ate ultimately very important at the use
of mathematical models. It should be noted that many important results can be
obtained with a model estimate of the role of individual factors in greatly simplified

© Springer International Publishing Switzerland 2017 399

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_10
400 10 Heat Exchange and Movement of the Solid Particles in the Plasma

conditions, particularly if the processes are considers that are limited only by
heating and melting of materials, and there is no the aim to obtain a product of a
given particle size and composition during a complex physic-chemical conversion
of feedstock.
Currently, most mathematical models of plasma processes with dispersed phase
is based on the one-dimensional approximation, describing changes of the internal
temperature field of a spherical particle in an isothermal non-gradient plasma flow,
or non-gradient heating of particles in the flow with variable parameters. In many
cases, such a limited statement of the task is justified and allows obtaining of
information sufficient for proper selection of the values of the basic technological
parameters. However, using the high-power plasma flow and a working area with a
large size, as well as in the processing of poly-dispersion materials for industrial
production both internal and external parameters for the raw material particles
should be noted. The consideration of the kinetics of the product formation pro-
cesses requires also our greater attention. All of these things are connected with the
inevitable complexity of the mathematical description.
In such complex processes as receiving of the powders and their compounds
with the specified quality in restoring of the dispersed materials in plasma flow, it is
difficult to predict the result. The estimation of the role of individual factors can be
found only in the course of researches. The most important of these parameters are:
the composition and properties of the initial gas and the dispersed phases of the
products; initial temperatures and velocities, as well as consumptions of the gas and
material; the structure of plasma flux; method of the entrance and size of the particle
of material; possible chemical reactions in the system; conditions of the heat
exchange and assumption of the flux in the surrounding space. For the solving of
the technological task of the particles treatment in plasma it is necessary to deter-
mine: the efficiency of the technological process; the quality of the processed
product; optimal regimes for processing of material and operation of the installa-
tion; the maximum mass factor for the processing of particles at the given
parameters of the plasma installation and others.
To construct a model that takes into account the interaction between plasma and
the groups of particles, and the creation of the technique of their modeling, it is
necessary to consider the following tasks: the movement and heating of the particles
in given fields of temperature and velocity of plasma in view of phase transfor-
mations during melting and evaporation; movement and heating of plasma loaded
by the flow of particles and their interaction. In the theoretical study the funda-
mental equations of the model for the calculation of heating and motion of particles
in the plasma flow are non-gradient and (or) the gradient equation of the particles
heating; the equations of motion of particles; equation of heating, taking into
account the phase transformation at melting and evaporation of the particles.
Example of calculation is performed for the plasma treatment of MgO particles in
the air RF plasma torch (1.76 MHz–60 kW).
10.1 The Basic Approaches to the Description of the Heat Exchange … 401

10.1 The Basic Approaches to the Description of the Heat

Exchange and Movement of the Particles
in the Plasma

10.1.1 The Criteria of Heat Transfer and Gas Dynamics

of Particles

The heating of particles in the plasma flow is mainly determined by the active heat
exchange and is estimated by the Nusselt number (Nu), that is the ratio of con-
vective to conductive heat transfer across (normal to) the boundary. Because of the
lack of direct experimental data of the heating of small free-moving particles in
plasma flow in the range of values of Nu there are some difficulties. The equations
of Rants–Marshall are mostly used for spherical particles in [1, 2] where the
determining is usually the temperature of plasma flow.

Nu ¼ 2 þ 0:6 Re0:5 Pr 0:33 ; ð10:1Þ

where Re, Pr—the numbers of Reynolds and Prandtl.

The temperature difference between plasma and surface of particles in plasma
flow is often essential, and the conditions of heat transfer depend on the parameters
of the boundary layer on the surface of particles, in expression (10.1), it was

0:2
q l
proposed to introduce a parameter q p lp , here qp ; lp —the density and vis-
ps ps

cosity of plasma; qps ; lps —the density and viscosity of plasma at the surface
temperature of particles. Here and after index «s» corresponds to the parameters of
particles, index «p» corresponds to the parameters of plasma, index «ps»—to
parameters of plasma at the surface temperature of particles. Experimental data of
the heating of assigned spherical particles with diameter of 0.15–1 mm in plasma
k
jets [3] confirm the opportunity to use then Eq. (10.1), using kpsp in the form:

!0:2
kps qp lp
Nu ¼ 2 þ 0:6Re0:5 0:33
ps Prps : ð10:2Þ
kp qps lps

kps
Using of kp clarifies the possibility of reducing of Nu to the level less than two,
not only for irregular shaped bodies but also for spherical particles at high tem-
perature gradients.
The Reynolds number is the ratio of the intensities of convective and viscous
flow is defined as:
 
qp vp  vs ds
Re ¼ ; ð10:3Þ
lp
402 10 Heat Exchange and Movement of the Solid Particles in the Plasma

where vp ; vs —the velocity of plasma and particles, respectively; ds —diameter of

particles.
The Prandtl number as the ratio of the intensity of the conductive and viscous
flows is calculated by equation
lp cpp
Pr ¼ : ð10:4Þ
kp

There are more than two dozen of equations for determination of the aerody-
namic resistance Cd in depending on Re, obtained by calculation and experimen-
tally (mainly for enshrined particles larger than 1 mm). Most equations for the
calculation of Cd(Re) is:

24
Cd ðReÞ ¼ þ AReB ;
Re

where the coefficients A and B depend on the specific conditions and are determined
experimentally.
Many factors influence to the ratio Cd, a number of papers [1–7] is devoted to
analyze of these factors. The following equation for calculation of Cd(Re) for the
sphere (the so-called “standard” curve of resistance), obtained in the cases of a
steady isothermal flow. It has been well studied and confirmed by numerous
experiments for a wide range of Re numbers.

16:6
Cd ðReÞ ¼ þ 0:2: ð10:5Þ
Re0:75

The number of Bio characterize the intensity of heat exchange of particles with a
plasma flow

ads
Bi ¼ ; ð10:6Þ
2ks

where a—coefficient of heat transfer of particles; ks ; ds —the thermal conductivity

and diameter of the particle.
The number of Fourier for heat homochronicity is defined as

4ks t
Fo ¼ ; ð10:7Þ
qs cps ds2

where t—time.
The other dependences can be found in the references about the heat transfer
[5–7]. In the phase changes in the processes of melting and evaporation
the dependence of heat transfer of particles in the plasma flow should change. In
[1, 2, 8], there are the equations for calculation of Nu and Cd for evaporation of
particle, but their use is hampered due to the lack of reliable experimental data.
10.1 The Basic Approaches to the Description of the Heat Exchange … 403

10.1.2 The Non-gradient Heating and Movement

of the Particles in the Plasma

Model of non-gradient heating (Fig. 10.1) can be used to study of the heating of
small particles ðds \50 lm) with high thermal conductivity ks , i.e., when the
number of Bio (10.6), that characterize the intensity of heat exchange with plasma
flow particles, is much less than one [1, 2]. Similar conditions may take place at the
low heat transfer coefficient a. In this case the temperature gradient of the particle is
small and it is believed that the particle in the whole volume has the same tem-
perature. For the model calculation of heating the spherical shape of particles is
usually assumed (Fig. 10.1).
The equation of non-gradient heating of particle is:

dTs 6   
qs cps ¼ a Tp  Ts  erTs4 þ Urad ; ð10:8Þ
dt ds

where qs ; cps —the density and specific heat of particles, respectively; Tp —the
temperature of plasma at the external edge of the particle; Ts —the temperature of
the particles; e—the degree of blackness of particles; r—the constant of Stefan–
Boltzmann; Urad —the specific power of the plasma radiation.
The total heat flux that is going to the heating of particles,

  ds
qP ¼ a Tp  Ts þ Urad  erTs4 ;
6
 
where a Tp  Ts ¼ q is the heat flux coming from plasma to the heating of par-
ticles by convection and thermal conductivity; Urad d6s ¼ qrad —the heat flux from the
plasma emission to the surface of particle; erTs4 ¼ qrs —the heat flux of the
emission from the surface of particle. The total flux of energy to the surface of

Fig. 10.1 The scheme of

non-gradient heating of
particle in plasma
404 10 Heat Exchange and Movement of the Solid Particles in the Plasma

particle determines the velocity of changes of its temperature that is expressed by

the left side of Eq. (10.8).
The heat transfer coefficient depends on many factors and is determined by the
dynamics of heat transfer between plasma and particle. In most of the works it is
calculated by the Nusselt number (10.2) as:

kps
a ¼ Nu : ð10:9Þ
ds

Analytical Solution of the Equation the Non-gradient Heating Analytical

solution of Eq. (10.8) exists only at all its constant coefficients and can be justified
for a sufficiently small period of time when all the parameters of particles and
plasma can be considered permanent. For simplicity we assume that the contribu-
tion of the radiation flux into the total flux that is going into the heating of particles
is insignificant and it can be neglected. Then the analytic solution of Eq. (10.8) with
the initial condition Ts jt¼0 ¼ Ts0 has the form
 
Ts ðtÞ ¼ Tp 1  ekt þ Ts0 ekt ; ð10:10Þ

where the time constant k ¼ q c6aps ds .

s
The total time of heating of particle to the melting temperature Ts1 is calculated
using the equation

1 Tp  Ts0 q cps ds2 Tp  Ts0

t0 ¼ ln ¼ s ln : ð10:11Þ
k Tp  Ts1 6Nukps Tp  Ts1

After the reaching of the melting temperature of heating the energy due to a
classical representation is consumed only the destruction of the molecular structures
and the temperature of particle remain constant. Thus, the general equation of
non-gradient heating of particle (10.8), taking into account the melting phase is as:

dTs 6    6L1 dms

qs cps ¼ a Tp  Ts  erTs4 þ 3 ; ð10:12Þ
dt ds pds dt

pd 3
where L1—the specific heat of melting of the material of particles; ms ¼ qs 6s —the
6L1 dms
mass of the solid core of particle; pd 3 —the specific power expended to the
s dt
melting of particles.

After the reaching of the temperature of melting ddTts ¼ 0 (10.12) is trans-
formed into an equation describing the change in the diameter of the solid core of
particle over the time:
10.1 The Basic Approaches to the Description of the Heat Exchange … 405

 
dds a Tp  Ts1  erTs1
4
¼ 2 ¼ C1 :
dt qs L1

At the constant transfer coefficients of plasma and particles with diameter of

solid core of particle in dependence on time is defined as

ds ¼ C1 t þ ds0 ; ð10:13Þ

where ds0—the diameter of particle at the beginning of melting.

Time of the full melting, respectively, is calculated using the equation

ds0 q d2 L
t1 ¼  ¼   s s0 1 : ð10:14Þ
C1 2 Nukps Tp  Ts1  erTs1
4

A total time required for a full heating and melting of particles is:
"
#
2
qs ds0 cps Tp  Ts0 L1
t ¼ t0 þ t1 ¼ ln þ   : ð10:15Þ
2 3Nukps Tp  Ts1 Nukps Tp  Ts1  erTs1
4

Calculation of non-gradient heating of particles on the basis of this simplified

heat balance equation, despite its approximation, allows in some cases without
cumbersome calculations to make estimation of the velocity of heating of particles
and to carry out the selection of the optimal regime of the technological process.
In real cases, when solid particles move inside the plasma flow with sharply
varying parameters, the assumption of the constancy of the parameters of plasma
and particles in the analytical solution of (10.10, 10.11) and (10.13–10.15) is not
justified. At the same time, a numerical method for the solution of heating of
particles and for obtaining of heat flow going into the heating of particle should be
used, it is necessary to determine the coordinates of location of the particle and the
speed of its movement in flight, i.e., it is necessary to solve the equation of motion
of particles.
Movement of the Solid Particles in Plasma Particles moving at the high speed
plasma flow, are not only actively heated, but are quickly accelerated. Different
forces acts to the particle in the plasma flow: the force of aerodynamic resistance Fa,
the Magnus force FM, the force of thermophoresis Ft, gravity Fg, Archimedes force
FA, the force of the apparent mass Fam, the force of Basse FB, electrostatic forces,
the force of light pressure and others [1]. The results of calculation of these forces
for typical plasma jets of RF and Arc plasma torches are shown in Table 10.1.
The data in Table 10.1 show, the basic forces that determine the motion of
particles in plasma flow are the force of aerodynamic resistance and the force of
gravity, and the force of thermophoresis should be considered only for very small
 406
10

Table 10.1 The results of the calculation of forces acting on particles of Al2O3 in jets of RF and Arc plasma torches
Parameters Argon Air
Tp = 10,000 K, vp = 20 m/s Tp = 13,000 K, vp = 400 m/s Tp = 7000 K, vp = 20 m/s Tp = 9000 K, vp = 400 m/s
ds (lm) 1 10 100 1 10 100 1 10 100 1 10 100
Fa (N) 6  10−8 6  10−7 6  10−6 8.3  10−7 7.5  10−6 5.6  10−5 4  10−8 4  10−7 4  10−6 1.1  10−5 10−5 8  10−5
−18 −15 −12 −7 −14 −11 −18 −15 −12 −17 −14
FA (N) 6  10 6  10 6  10 8  10 8  10 8  10 4  10 4  10 4  10 5  10 5  10 5  10−11
−14 −11 −8 −14 −11 −8 −14 −11 −8 −14 −11
Fg (N) 2  10 2  10 2  10 2  10 2  10 2  10 2  10 2  10 2  10 2  10 2  10 2  10−8
−10 −9 −8 −8 −6 −5 −10 −9 −8 −8 −8
Ft (N) 8  10 2  10 2  10 5.7  10 5.7  10 5.7  10 9  10 10 10 1.3  10 6  10 6  10−7
−12 −10 −8 −10 −8 −6 −12 −10 −8 −10 −8
FM (N) 4  10 4  10 4  10 7  10 7  10 7  10 2  10 2  10 2  10 7  10 7  10 7  10−6
−17 −14 −11 −15 −12 −9 −17 −14 −11 −15 −14
Fam (N) 2  10 2  10 2  10 2  10 2  10 2  10 10 10 10 10 10 10−9
−12 −10 −8 −10 −8 −6 −12 −10 −8 −10 −8
FB (N) 9  10 9  10 9  10 2  10 2  10 2  10 6  10 6  10 6  10 2  10 2  10 2  10−6
Heat Exchange and Movement of the Solid Particles in the Plasma
10.1 The Basic Approaches to the Description of the Heat Exchange … 407

particles (ds < 1 lm). Other forces are much smaller in absolute values, and can be
neglected in simulation of motion of particles in the plasma stream.
Two-dimensional equations of motion of particles in the plasma flow recorded in a
cylindrical coordinate system are:
8
< q dvsz ¼ 0:75C qp ðvpz vsz Þjvpz vsz j þ q g ;
s dt dz s z
ds
ð10:16Þ
q ð v v Þ
: q dvsr ¼ 0:75C p pr sr pr sr :j v v j
s dt dr ds

Here vsz ; vsr ; vpz ; vpr —the axial and radial components of the velocity of particle
and plasma, respectively; Cdz ; Cdr —the coefficients of aerodynamic resistance in
two directions.

10.1.3 The Gradient Heating Particles in the Plasma

The assumption of the absence of a temperature gradient inside the particle is true
only in the region of small number of Bio (Bi < 0.1) and for sufficiently large
values of the Fourier number (Fo  1). In this case, the kinetics of heating of
particles depends weakly on heat-inertial properties and is almost completely
determined by the conditions of the external heat transfer.
In those cases when the intensity of heating and speed of passage (numbers of
Bi, Fo) define a substantial temperature gradient inside the particle, for calculation
of its temperature in dependence of time the following equation should be solved, it
is written in a spherical coordinate system

@Ts ðr; tÞ 1 @ @Ts ðr; tÞ

qs cps ¼ 2 r 2 ks : ð10:17Þ
@t r @r @r

The initial condition Ts ðr; 0Þ ¼ Ts0 and the boundary conditions of the third kind
are accepted.

@Ts ð0; tÞ @Ts ðR; tÞ  

¼ 0; ks ¼ a Tp  Ts ðR; tÞ þ Urad R=3  erTs4 ðR; tÞ; ð10:18Þ
@r @r
 
where R ¼ d2s —the radius of particle; a Tp  Ts ðR; tÞ —the heat flow (convective
and conductive) from plasma to the surface of particles; Urad R=3—the heat flux
from plasma radiation coming to the particle; erTs4 ðR; tÞ—the heat flux from the
radiation emitted by the surface of particles.
The Numerical Solution of the Equation of Gradient Heating of Particles
Equation of the gradient heating (10.17) is the equation of transient heat conduc-
tion, the solution of it can be obtained using the method of control volume.
408 10 Heat Exchange and Movement of the Solid Particles in the Plasma

The energy balance inside each of the control volume (Fig. 10.2) can be
obtained by integrating of the Eq. (10.17) for the control volume and time:

Zre tZþ Dt tZþ Dt Zre

@Ts 2 1 @ 2 @Ts
qs cps dtr dr ¼ r k s r 2 drdt: ð10:19Þ
@t r 2 @r @r
rw t t rw

In the integration of the Eq. (10.19), the following assumptions are used: inside
the control volume (rw, re) and at the time interval (t, t + Dt) the value qs cps takes a
constant value corresponding to the central point M and at time t; the temperature of
particle in the transient member @Ts@tðr;tÞ is constant, as the central point
M corresponding for the interval (rw, re).
If the left and right side of the Eq. (10.19) is once integrated (for convenience we
denote Ts  T), we have:

    tZþ Dt tZþ Dt
qs cps re3  rw3 @TM TE  TM
M
dt ¼ re2 kse dt
3 @t rE  rM
t t
tZþ Dt
T M  TW
 rw2 ksw dt: ð10:20Þ
rM  rW
t

We use the following approximations for the change in time from t to t + Dt of

R Dt
tþ  
the temperatures TM ; TE ; TW : TM dt ¼ e TM þ ð1  eÞTM0 Dt, where e—a
t
certain weight coefficient, that varies from 0 to 1. Here and below we use the
notations TM, TE, TW as the temperature at the time t + Dt, and TM0 ; TE0 ; TW0 —at the

Fig. 10.2 Grid and the control volumes for the solving of the equation of gradient heating of
particles
10.1 The Basic Approaches to the Description of the Heat Exchange … 409

time t. Using the similar relations for the integral TE, TW from t to t + Dt, from the
Eq. (10.20) we find:
     
qs cps re3  rw3   TE  TM TM  TW
M
TM  TM ¼ e re kse
0 2
 rw ksw
2
3Dt rE  rM rM  rW
 
T 0
 T 0
TM0  TW0
þ ð1  eÞ re kse
2 E M
 rw ksw
2
:
rE  rM rM  rW

The result is a discrete analogue of the gradient heating of particles (10.17) in the
form:
   
aM TM ¼ aE eTE þ ð1  eÞTE0 þ aW e TW þ ð1  eÞTW0
 
þ a0M  ð1  eÞðaE þ aW Þ TM0 ; ð10:21Þ

where
8
< aE ¼ re2 kse [ 0; aW ¼ rw2 ksw [ 0;
rE rM rM rW
: a0 ¼ ðqs cps ÞM ðre3 rw3 Þ [ 0; a ¼ eða þ a Þ þ a0 [ 0:
M 3Dt M E W M

For certain values of the weigh coefficients e have known schemes for solving of
the transient equation of heat conduction. In particular, e = 0 we get the explicit
scheme for e = 0.5—implicit scheme of Crank–Nicolson for e = 1—a fully implicit
scheme. The basic principle of a fully implicit scheme is that in the step of time the
temperature is assumed to the new value of TM. Thus, the thermal conductivity
should be recalculated using TM in the iterative process in the same way as in the
solution of the stationary case. The analysis of these schemes in [9] shows that the
sustainable solution of discrete analogues is guaranteed using only a fully implicit
scheme, i.e., if e = 1, and this scheme is taken for solving of our task. The discrete
analogue for a fully implicit scheme have the form:

aM TM ¼ aE TE þ aW TW þ a0M TM0 ; ð10:22Þ

where
8
< aE ¼ re2 kse [ 0; aW ¼ rw2 ksw [ 0;
rE rM rM rW
: a0 ¼ ðqs cps ÞM ðre3 rw3 Þ [ 0; a ¼ a þ a þ a0 [ 0:
M 3Dt M E W M

Note that if Dt ! 1 from (10.22) the discrete analogue of the stationary

equation of heat conductivity can be obtained.
410 10 Heat Exchange and Movement of the Solid Particles in the Plasma

Using the boundary conditions (10.18), the discrete analogue can be obtained for
the boundary points near the center and the surface of particles.
For a point near the center of the particle r = r20 (Fig. 10.2):

aM TM ¼ aE TE þ a0M TM
0
; ð10:23Þ

Here, from the boundary condition in the center of the particle the discrete
analogue aW = 0.
For a point near the surface of the particle r = r2n−2 (Fig. 10.2):

aM TM ¼ aW TW þ a0M TM
0
þ qR R 2 ; ð10:24Þ
 
where qR ¼ a Tp  Ts ðR; tÞ  er Ts4 ðR; tÞ þ Urad R=3—the total heat flux at the
surface of the particle.

10.1.4 Evaporation of Particles in the Plasma

Evaporation of particles in the plasma flow is an important technological process

that should be studied in many problems of plasma chemistry. In particular,
evaporation of particles is an indispensable condition for obtaining of the materials
with small disperse, such as by direct evaporation, and then using of chemical
reactions. The mechanism of evaporation processes of crystalline solids in sufficient
considered in details [10–13]. It can be presented as step process consisting in
transition of atoms and molecules through the different states at the surface of the
border crystal–vapor and through the adsorption layer into the vapor phase. In the
case of liquid, taking into account its intermediate phase state between solid and
gaseous two mechanisms of evaporation are possible—the direct and step through
the adsorption layer. The mechanism of evaporation is mainly determined by the
activation energy and at high heat fluxes particle vaporization starts before reaching
of the boiling point. The resulting vapors create the steam jacket around the particle,
consisting of a mixture of heavy vapor molecules and gas molecules. As a result,
the process has to be stabilized by the heat balance. Thermal energy supplied from
plasma to the particle, will be consumed to its heating, evaporation and heating of
the vapor to the temperature of plasma and in the steady regime—to the evaporation
and heating at a constant temperature of particles.
According to J. Stefan, the evaporation process can be represented the diffusion
of vapor from the particle surface, above this surface there is a saturated vapor, into
the environment gas. The processes of heat and mass transfer in the plasma flow are
similar to each other; to the heat transfer due to the molecular thermal conductivity
corresponds to molecular diffusion, convection heat transfer—convective diffusion.
Thus, the theoretical and experimental results obtained in the study of heat transfer
processes and methods of their investigation can be directly applied to the processes
of diffusion, and vice versa.
10.1 The Basic Approaches to the Description of the Heat Exchange … 411

Basic Equations and Criteria In case of simultaneous heating, melting and

evaporation of the particle the equation of non-gradient heating can be written as:

dTs 6    6L1 dms

qs cps ¼ a Tp  Ts  erTs4  L2 j þ 3 ; ð10:25Þ
dt ds pds dt

where L1, L2—the specific heat of melting and evaporation of the material of
pd 3
particles, respectively; ms ¼ qs 6s —the mass of the solid core of particles; j—the
density of the flux from the surface of the vaporized material.
The density of vapor stream from the surface of vaporized material, i.e., the
quantity of substances transported through the unit of surface area per unit of time,
is often expressed by Fick’s law [14]:
– in a stationary medium:

~j ¼ DgradC; ð10:26Þ

– in a mobile environment:

~j ¼ DgradC þ~
vC; ð10:26aÞ

where C—the concentration of the evaporated substance; D—the diffusion coeffi-

cient; ~
v—the difference between velocity of medium and of particle.
The differential equation of diffusion in a stationary medium has the form

@C
¼ div ðDgradCÞ þ q; ð10:27Þ
@t

where q—density of sources of the substance, i.e., the quantity of the substance
formed by chemical reactions in the unit of volume per the unit of time.
The diffusion equation in the mobile environment, respectively, has the form:

@C
þ div ðC~
vÞ ¼ div ðDgradCÞ þ q: ð10:27aÞ
@t

In the cases when the resulting amount of substance can be taken as q = 0 and
for the stationary processes @C
@t ¼ 0, the diffusion Eq. (10.27a) can be written as:

vÞ ¼ divðDgradCÞ:
divðC~ ð10:28Þ

The velocity of evaporation of particles I, i.e., the quantity of substance m,

carried through the surface with area S per unit of time is defined as:
412 10 Heat Exchange and Movement of the Solid Particles in the Plasma

dm
I¼ ¼ jS: ð10:29Þ
dt

Solution of the problem of particles evaporation is reduced to the determination

of the evaporation velocity I or the flux density j. Depending on the size of particles,
pressure, and composition of the medium the evaporation can be at diffusion (the
Knudsen number Kn < 1), free-molecular (Kn > 1) or transient regime [1, 14].
Quasi-stationary Evaporation of Particles Fixed in Relation to the
Environment In the case where the particle is immobile in the environment, and
its radius r0 is much higher than the mean free path of the vapor molecules l, the
vapor concentration at the surface of the particle is equal to the equilibrium con-
centration C0. This equation of Maxwell [14] is valid:

I ¼ I0 ¼ 4pr0 DðC0  C1 Þ; ð10:30Þ

Here C = C0 at r = r0, C = C∞ at r = ∞.
In this case, the velocity of particles evaporation is completely determined by the
velocity of steam diffusion into plasma, and we have a purely diffusive regime of
evaporation.
Vapor concentration at a distance r from the center of particle is calculated using
the equation
r0
C ¼ C1 þ ðC0  C1 Þ: ð10:31Þ
r

Particle radius in dependence of time t can be defined as:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2D
r ¼ r02  ðC0  C1 Þt: ð10:32Þ
qs

If we assume that the vapor of substance conforms to the laws of ideal gases, and
express the concentration of vapor C using its partial pressure p : C ¼ pMRT , where
M—the molecular mass of the evaporating substance in a gaseous state; R—the gas
constant; T—the temperature of vapor, the Maxwell equation takes the form of

ðp0  p1 ÞM
I0 ¼ 4pr0 D : ð10:33Þ
RT

Here the vapor pressure p = p0 at r = r0, p = p∞ at r = ∞.

As first pointed out by J. Stefan, near the surface of the evaporating body the
hydrodynamic flow (the Stefan’s flow) of the gaseous medium from the surface
should exist. In this case, the velocity of the particles evaporation can be determined
approximately [1] using the equation
10.1 The Basic Approaches to the Description of the Heat Exchange … 413

p0 þ p1
I ¼ I0 1 þ ; ð10:34Þ
2pp  p0  p1

Here pp—the pressure of the plasma.

Influence of the boundary layer on the velocity of evaporation can be accounted
if we assume that the equation of Fick (10.26) and the ensuing Eq. (10.26a) are
applicable only at a distance of more than D * l from the surface of the particles,
and in the layer near the wall with the thickness D the exchange of molecules of
vapor smoothly occurs in vacuum. The thickness of boundary layer D is the dis-
tance from the particle surface, where evaporated molecules have its first collision
with other molecules, and the vapor density decreases from C0 near the surface of
particle to C∞ in the plasma flow. The velocity of evaporation of the particle in this
case can be defined as

I0
I¼ ; ð10:35Þ
r0 =ðr0 þ DÞ þ D=ðr0 eaÞ
qffiffiffiffiffiffi
Here e ¼ 14 pM —one
8kT
fourth of the average velocity of the vapor molecules;
a—coefficient of vapor condensation.
With increasing ratio l/r0 the diffusion regime of evaporation is gradually
transformed into kinetic and at large l/r0 the diffusion finish to play a role in the
process of evaporation [11].
Quasi-stationary Evaporation of Particles Moving in Relation to the
Environment The greatest practical interest is the process of evaporation of par-
ticles moving relative to plasma by the force of aerodynamic resistance and other
forces.
The accounting of all factors influencing to the evaporation of particles com-
plicated its calculations, very often various criterial dependences of the numbers of
Reydnols Re, Nusselt Nu, Sherwood Sh, Schmidt Sc and others are taken. They take
into account the phase transformations of particles and different from the usual
equations by some correction factors. Thus, the velocity of the particles evaporation
taking into account the boundary layer in the moving flow of plasma is determined
as [1, 7]:


I ¼ I0 1 þ bRe1=2 Sc1=3 ; ð10:36Þ

where b ¼ Dl ; I0 —the velocity of evaporation obtained by Maxwell’s Eq. (10.30)

or (10.33).
Similarly, comparing the models (when the particles are fixed and when they are
addicted by the flow of plasma) for large values of Re and the thickness of the
414 10 Heat Exchange and Movement of the Solid Particles in the Plasma

boundary layer D  r0 , we have the following equation for determination of the

velocity of evaporation of particles moving in relation to the flow of plasma:
 
I0 1 þ bRe1=2 Sc1=3
I¼ : ð10:37Þ
1 þ eað2r0 =Shbl
D
Þ

At very high velocities of the flow, when the thickness of the diffusion boundary
layer is smaller than the mean free path of gas molecules ðD\lÞ, almost all
evaporated molecules fly away by the gas flow, i.e., evaporation will move with the
same speed as vacuum:

I ¼ 4pr02 DC0 ea: ð10:38Þ

For estimate of the velocity of evaporation in the transition regime, the equation
of Sherman [1, 8] can be used.

Id Ik
I¼ : ð10:39Þ
Id þ Ik

Here Id ; Ik —the velocity of evaporation in the diffusion and kinetic regimes.

Experimental Methods for the Study of Evaporation In practice, for deter-

mination of the velocity of evaporation of the material of particles the experimental
data are often used. In the experimental study of evaporation and dissociation of
materials the following methods are most frequently used: the high-temperature
mass spectrometry (combination of the Knudsen method with mass spectrometric
analysis of the products of evaporation); dynamic, such as method of flow; static;
method of boiling points, etc.
The flux density j in view of the evaporated material is determined by measuring
pressure of the saturated vapor at this surface at a known value of temperature. It is
believed that vapor molecules on the surface of the particles are in equilibrium and
have the Maxwell velocity distribution. Then we have the following equation for
calculation of the flux density from evaporation:
p
j ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi : ð10:40Þ
2pRT = M

There is a large number of experimental data for determination of the saturated

vapor pressure for different materials in literature, and they can be used for the
calculation of the flux density from evaporation. However, it should be noted that
the data of the evaporation of particles at high temperatures in many cases are not
available.
10.1 The Basic Approaches to the Description of the Heat Exchange … 415

Experiments [15] show that the process of evaporation of material of particles at

high temperature plasma is often accompanied by complex chemical transforma-
tions, dissociation and condensation of molecules. For example, at the evaporation
of a magnesium oxide MgO the dissociation mainly occurs in two ways:

MgOðsÞ ¼ MgðgÞ þ 0:5O2ðgÞ ;

MgOðsÞ ¼ MgðgÞ þ OðgÞ ;

and in a lesser degree extent by sublimation according to the scheme:

MgOðsÞ ¼ MgOðgÞ ;

Here the solid phase is signified by the index “s”, and index “g”—the gaseous one.
As a result of experimental measurements [15] at the temperature of particles
Ts ¼ 1900 K content of molecules of vapor is obtained on the surface of MgO(s):
Mg—52.27 %, O2—32.91 %, O—13.82 %, MgO—1 %. The velocities of the
chemical reactions in plasma are different, and their mathematical descriptions are
rather cumbersome and time-consuming, so here they are not given here.

10.2 Calculation of the Heating and Movement

of the Particles with the Given Parameters
of the Plasma Flow

10.2.1 Installation for Plasma Treatment of Powder

Materials

The plasma method for processing of refractory material has a number of advan-
tages over mechanical crushing. Firstly, the mechanical crushing method can lead
to further pollution, thereby require the additional chemical cleaning. Secondly,
mechanical crushing to small and ultra-small particles with diameter less than 30–
40 lm is a difficult and sometimes impossible task. Thirdly, the problem of
chemical cleaning of materials and making them spherical cannot be solved by
using this method.
Compared with the traditional method for processing of materials the plasma
method does not have these disadvantages and, furthermore, allows obtaining of
high technical and economical indicators. At present, for the treatment of powder
particles the plasma jets of RF and Arc plasma torches are often used. The jets of
Arc plasma torches have a relatively small diameter, high velocity of flow with a
large temperature gradient. However, the presence of electrodes and its erosion in
some cases does not allow obtaining of the pure product. RF plasma torches with
416 10 Heat Exchange and Movement of the Solid Particles in the Plasma

equal power to the Arc plasma torches allow obtaining of the plasma jet with
greater diameter with and small speed of flow and a large spatial region of uniform
temperature. Small speeds of RF plasma also significantly alleviate the problem of
particles entering into the plasma jet, and provide them to stay longer in the jet in
compare with Arc plasma. Furthermore, the RF plasma, uncontaminated by the
materials of electrode, allows processing and obtaining of the pure powder material.
As an example, the description of RF-installation and plasma technology for
processing of MgO particles, that are widely used in industry. A significant dis-
advantage of currently produced industrial electrical MgO is a large specific surface
of the powder that leads to significant reduction of electrical characteristics of
tubular electric heaters produced using MgO and have very strong requirements on
the chemical composition of the powder, especially the composition of CaO and
Fe2O3.
The proposed process for treatment of the particles is a preparation of melted
spherical grains of MgO in RF plasma jet. Spheroidization of the MgO periclase
particles is produced in air RF plasma with the power of 60 kW (Fig. 10.3).
Installation allows receiving of MgO powder consisting of spherical particles with a
minimum specific surface, high electrical parameters even with high content of
calcium and iron impurities, minimum abrasion and high turnover.

Fig. 10.3 Air RF plasma installation (1.76 MHz–60 kW) and a melting chamber for the
processing of MgO periclase particles in a stream with the distributed entrance of particles at the
perimeter of plasma torch
10.2 Calculation of the Heating and Movement of the Particles … 417

The jet of air plasma coming out of RF plasma torch with metal cut water-cooled
chamber, consisting of some of sections, assembled with clamps on the housing of
the discharge chamber (Fig. 10.3). The discharge chamber is placed in a quartz
cover and induction plasma is generated in a channel by a RF inductor. There is a
straight-jet air supply in the plasma torch. The plasma forming air is supplied
through the central tube and the thermal protection gas is formed by a coaxial quartz
tube and gas-former. Thus, the air forms a thermally protective layer of the moving
cold gas between plasma and the walls of plasma torch. This air supply significantly
increases the efficiency of the plasma torch and makes its work stable. Basic
parameters of RF installation for processing of particles are shown below:

Main features of the feed medium:

Voltage (V) 3  380
Frequency (Hz) 50
Input power (kW) 90
Power of the oscillation circuit (kW) 60
Operating frequency (MHz) 1.76 ± 0.044
The plasma-forming and transport gas Argon, air
Air consumption (m3/h) 1.8–3.5
Cooling water consumption (m3/h), no less than 4.5
Productivity (kg/h), for the particles with size of 100 lm *8

It was found that the plasma installation, working using a once-side scheme of
entrance of materials into plasma, it is impractical to use in industrial environments
due to low technical and economic factors. The consumption of the power in its use
is 33–59 kW h/kg of the finished product at a performance of 1.1–2.6 kg/h. In this
case, it was decided to change the scheme for entrance of MgO particles from the
once-side to the reverse with the spouted layer (Fig. 10.3). Powder particles of
MgO with diameter 100–200 lm are fed downwards to the plasma jet directed
upwards. The difference of this method of entrance of particles compared with input
by once-side or horizontal scheme is that the particles flying toward the plasma jet,
is inhibited there and stops at a certain point and then starts to move in the opposite
direction and is come into the melting chamber (Fig. 10.3). The scheme of RF
installation for the thermal treatment of the particles is shown in Fig. 10.4.
Such organization of “plasma fountain” significantly increases the residence
time of the particles in plasma compared with once-side or horizontal scheme, that
allows melting of even the larger particles. Due to greater residence time of the
particles in a plasma jet they more efficiently extract energy of the jet for its heating.
Thus, at the fixed power of the jet per unit of time more particles can be melted and
the best performance of the process can be achieved.
The electrical MgO produced by Bogdanovich’s refractory plant was taken for
the researches. For all the investigated MgO a number of analyzes was made:
characteristics of chemical and grain compositions, liquid absorption, fluidity,
418 10 Heat Exchange and Movement of the Solid Particles in the Plasma

Fig. 10.4 Scheme of the RF installation for the thermal treatment of the particles: 1 RF generator;
2 RF plasma torch; 3 melting chamber; 4 compressor making the necessary consumption of
plasma-forming and transport gas G (the air); 5 rotameters unit for control the consumption of
water W and air G; 6 cylinder with Argon for ignition of the plasma torch; 7 the unit for entrance of
initial material feed (feeder); 8 pre-filter; 9 exit to the main filter and ventilation system; 10
collecting hopper of processed particles

density after tapping; the specific, volume resistivity of insulation and leakage
current at voltage U = 500 V are measured for experienced electric heaters.
The results showed that treatment with MgO periclase counter submitting it to
the plasma allows to obtain a finished product with high electrical insulating
properties at the level of the best periclase of foreign companies and technical and
economic parameters are in several times higher than in one-side feed of powder
into the plasma jet. Basic parameters of periclase particles before treatment and after
it are shown in Tables 10.2, 10.3 and 10.4.
The sufficient degree of spheroidization of the periclase grains was reached at
their treatment in this installation (80 ± 10) %. The physical and chemical
parameters of obtained periclase MgO are better than material PPE-1 M and the
achieved performance 70 % spheroidization of 8–3 kg/h for MgO fractions with
diameter 100–200 lm.

10.2.2 Solution for Equations of the Heating

and Movement Particles

The equations of motion and non-gradient heating of particles is a system of

nonlinear ordinary differential equations and can be written in the Cauchy form as
follows [16, 17]:
10.2 Calculation of the Heating and Movement of the Particles … 419

Table 10.2 Influence of plasma treatment on the chemical composition of MgO periclase
particles
Class of MgO Material Chemical composition (%)
periclase particles CaO SiO2 Al2O3 Fe2O3 C MgO ppp
Third Original 1.68 1.55 0.31 0.11 0.079 96.11 0.16
First and second After plasma 1.62 1.70 0.05 0.07 0.031 96.47 0.06
treatment

Table 10.3 Changes of the electrical-isolating characteristics of the MgO particles after plasma
treatment
Class of MgO Material Thermal load of heating element (W/cm2)
periclase particles 6 8 10
R (MOm) I (mA) R (MOm) I (mA) R (MOm) I
(mA)
Third Original 4.70 0.35 0.47 1.20 0.10 3.32
First and second After plasma 26.7 0.20 2.70 0.53 0.53 1.31
treatment

Table 10.4 Change of specification of MgO periclase particles after plasma treatment
Class of MgO Material Liquid Dielectric Fluidity Density
periclase particles absorption strength (g/s) (g/cm3)
(%) (kV/mm)
Third Original 2.5 0.8 0.6 230
First and second After plasma 1.5 1.5 0.8 251
treatment

8 h i
> dTs ¼ 1 6 aT  T   erT 4  þ U ;
>
> qs cps ds p s s rad
< dt
dvsz ¼ 0:75C qp ðvpz vsz Þjvpz vsz j þ g ; ð10:41Þ
>
> dt dz qs d s z
>
: dvsr qp ðvpr vsr Þjvpr vsr j
dt ¼ 0:75Cdr q ds s
:

To solve the system of Eq. (10.41), the following initial boundary conditions are
taken: t = 0, Ts(0) = Ts0, vsz(0) = vsz0, vsr(0) = vsr0, z(0) = z0, r(0) = r0.
In order to obtain the coefficients of (10.41) depending on time it is necessary to
determine the coordinates of the particle by the equations

Zt Zt
zðtÞ ¼ vsz ðtÞdt; rðtÞ ¼ vsr ðtÞdt: ð10:42Þ
0 0
420 10 Heat Exchange and Movement of the Solid Particles in the Plasma

The task of Cauchy reduces to integrating the system of differential equations in

the general form:

dYj  
¼ Fj x; Yj ; ð10:43Þ
dx

where Yj—vector of unknown variables; Fj—vector of the right part of the system
of differential equations; x—independent variable; j = 1−N—serial number of the
equation; N—the number of equations.
The solution of system (10.43) with the initial conditions x = x0, Yj(x0) = Yj0
leads to finding the dependencies, passing through the points, defined by these
conditions. The order of the numerical integration defines the order of solution
method. The method of Runge–Kutta of fourth order is the most widely-used
method for solving of the system (10.43) with a constant integration step h = const.
Its advantages are the high accuracy (error of the method R * h5) and reduced risk
of occurrence of instability of the solution [18]. Algorithm realization of Runge–
Kutta method is the cyclic calculations of Yj,i+1 to every i + 1—step using the
following equations:
8    
< K1j ¼ hFj xi ; Yji ; K2j ¼ hFj xi þ  0:5h; Yji þ0:5K1j ; 
K3j ¼ hFj xi þ 0:5h;
 Yji þ 0:5K2j ; K4j ¼
 hFj xi þ 0:5h; Yji þ K3j ; ð10:44Þ
:
Yj;i þ 1 ¼ Yji þ 16 K1j þ 2K2j þ 2K3j þ K4j :

In the case of gradient heating two tasks should be solved at the same time (step
time): first—the solution of equations of motion of a particle of the Cauchy task
with the definition of its position and the heat flux going into the heating of
particles; second—solution of the equation of gradient heating (10.17) with the
boundary condition (10.18) for finding the temperature distribution inside the
particle at any time. Discrete analogue for equation of the gradient heating are
obtained in the form (10.21–10.24). The following are the results of calculation of
motion and heating of the particles in a plasma flow in cases of gradient and
non-gradient heating of particles.
The Result of the Calculation of the Movement and the Heating of the Particles
in a Given Plasma Flow To calculate the movement and the heating of the par-
ticles in a given flow of plasma it is necessary to define firstly all its parameters. The
developed above procedure (Chaps. 6 and 7) the simulation of air plasma in RF
plasma torch (1.76 MHz–60 kW) is carried out, the main parameters of it are given
in Tables 7.3 and 7.4, and the temperature field and the plasma flow are shown in
Fig. 10.5.
The obtained field of temperature and flow (speed) at the simulation of Aair
plasma in RF plasma torch (1.76 MHz–60 kW) are used to study the motion and
heating of MgO periclase particles. Initial data for the particles of MgO are:
diameter ds = 100, 150, 200 lm; density qs = 3580 kg/m3; the melting temperature
Ts1 = 2900–3100 K [15, 19] in dependence of MgO periclase chemical
10.2 Calculation of the Heating and Movement of the Particles … 421

Fig. 10.5 The trajectories of the particles MgO periclase (ds = 100, 150, 200 lm; vsz0 = −4 m/s,
vsr0 = −1 m/s) in the air RF plasma torch (1.76 MHz–60 kW)

composition; initial components of velocity vsz0 = −4 m/s, vsr0 = −1 m/s; the initial
temperature Ts0 = 300 K. In this case the coordinate of input point z0 = 540 mm
(40 mm from the edge of the plasma torch), r0 = 30 mm.
In the calculation all the local parameters of the particles were determined in a
given plasma flow: temperature and velocity of particles, the heat flux going into
heating of the particles, the temperature distribution inside the particles, the flight
time, the trajectory of motion; melting and evaporation points et al. The obtained
data allow find the optimal parameters depending on the particular material. In
order to optimize the field of temperature and the velocity of plasma processing are
analyzed, the size of particles, that should be melted at the end of plasma jet. In the
decision such criteria of heat exchange are also received: Re, Nu, Cd, and others, in
422 10 Heat Exchange and Movement of the Solid Particles in the Plasma

dependence on the coordinates of particles that makes it possible with the help of
the developed technique to compare the calculated and experimental data and
determine the criteria dependencies of heat exchange.
The reverse entrance of particles into plasma stream increases the processing
time. The obtained trajectories of particles (Fig. 10.5) show that at the same initial
conditions, the larger particles due to inertia penetrates to a greater depth of the
plasma flow and obtain greater heat flow going into the heating (Fig. 10.6). The
intersection of the trajectories of particles with a cold layer of gas at the end of the
flight may lead to a negative value of the heat flux of particles (i.e., to their cooling).
The results of calculation for the gradient and non-gradient heating differ quite
significantly. For example, for the MgO particles ds = 100 lm the temperature at
the end of the flight at the gradient heating reaches only 2250 K and at non-gradient
heating it is 3050 K (Figs. 10.6, 10.7 and 10.8).
The rapid heating of the particle in the calculation of the non-gradient heating
occurs because at such heating the time delay of the heat transfer inside the particles
is equal to zero. Consequently, for particles of MgO larger than 100 lm it is
necessary to determine the temperature gradient inside it. At the same time, at
simultaneous heating, melting and evaporation of the particles it is necessary to get
the temperature gradient inside the particles in dependence on the time and solve
the task of gradient heating. The smaller particles have the high specific surface,
and heat up and evaporate rapidly. Therefore, to increase the thermal efficiency of
the process and installation it is better to use smaller particles for processing. The
maximum heat flux going into heating of the particles takes place in the center of
the plasma jet due to the large difference in temperature and velocities between the
flow of plasma and particles (Figs. 10.5, 10.6, 10.7, 10.8, 10.9 and 10.10).
Therefore, the definition of conditions for entrance of particles into the plasma zone
is an important task for modeling of the motion of particles. Using the developed
technique for modeling of motion and heating of the particles and the calculation
program the input data can be adjusted such as the angle of the particles delivery,
place of delivery and the initial velocity in order to get the maximum heating of the
particles. Losses of energy by radiation of particle is much less than the heat flow
coming to its heating by plasma, so in the further calculations the energy losses of
particle by radiation are negligible.

Fig. 10.6 Change of the total

heat flux going to the heating
of particles MgO in the air RF
plasma torch (1.76 MHz–
60 kW): 1 ds = 100 lm, 2
ds = 150 lm, 3 ds = 200 lm
10.2 Calculation of the Heating and Movement of the Particles … 423

Fig. 10.7 Change of heating temperature (non-gradient heating) of particles MgO in the air RF
plasma torch (1.76 MHz–60 kW): notations are the same as in Fig. 10.6

Fig. 10.8 Temperature distribution inside the particles MgO (gradient heating) of different
diameter at different times during the motion in the air RF plasma torch (1.76 MHz–60 kW):
a t = 15 ms, b t = 25 ms, c t = 35 ms, d t = 45 ms; the other notations are the same as in
Fig. 10.6
424 10 Heat Exchange and Movement of the Solid Particles in the Plasma

Fig. 10.9 Velocity of MgO

particles in the air RF plasma
torch (1.76 MHz–60 kW): the
notation is the same as in
Fig. 10.6

Fig. 10.10 Flight time MgO

particles in the air RF plasma
torch (1.76 MHz–60 kW): the
notation is the same as in
Fig. 10.6

In the future, this technique can be used to determine the optimal regimes for
treatment of the particles. The parameters of this regime are: maximum heat flux
going into heating of the particles, their presence in the plasma jet, the point of
melting and evaporation of the particles and the size of the plasma jet for their
processing. However, a large number of particles in the plasma flow the properties
of plasma dramatically change and this model of motion and heating of the particles
in plasma at given characteristics of plasma does not allow determination of the real
parameters of a two-phase medium: plasma and particulate materials. Therefore, it
is necessary to put and solve the task of taking into account the interaction of the
particles with the plasma flow.
10.3 The Model of Plasma, Loaded the Flow of Solid Particles 425

10.3 The Model of Plasma, Loaded the Flow of Solid

Particles

In order to account the interaction of particles with plasma flow and to determine
the parameters for the processing of the dispersed raw materials it is necessary to
build a model of plasma, loaded the flow of solid particles. Imagine the model [16,
20], that allows not only calculation of the velocity of acceleration and heating of
dispersed particles in plasma, but also to consider the reverse effect of groups of
particles to the parameters of plasma—cooling down and braking of the latter. The
method that will be used for solving of the proposed model can be called the
method of quasi-single particles (like a single particle), because the behavior of a
large number of solid particles in plasma will be considered without taking into
account of their mutual influence (collisions, screening, etc.). Each particle is placed
in the reporting layer interacts with plasma, “unaware” the presence of other par-
ticles. This model allows us to describe the interaction of the flow of solid particles
with plasma and from a technological point of view makes it possible to determine
the important parameters of particle processing—limiting mass consumption of
material and efficiency of the process and the installation as a whole, to choose the
best parameters for obtaining of product with the required quality.
The model of plasma, loaded by the flow of particles includes the following
equations: equations of momentum and heating of the particles; equation of
momentum of plasma in view of the loss of momentum at the particles acceleration;
equation of the energy balance in plasma taking into account the losses of power to
the heating of the particles; continuity equations for all the gas and particulate flow;
the system of equations for the description of the electromagnetic field of plasma;
the system of equations for determination of the composition and thermal properties
of plasma and particles.
The system of equations for the description of electromagnetic field of RF and
Arc plasma torches and its solutions are discussed above (Sects. 7.2, 7.3 and 8.3).
Methods for determination of the composition and properties of plasma is given in
Sects. 4.1, 4.2 and 4.3. In many cases, the thermal properties of plasma particles are
known in advance and are shown in the tables in dependence on the temperature, so
the problem of obtaining of the properties of plasma particles at a certain tem-
perature is reduced to the approximation task. Below we present the basic equations
of the model of two flows of plasma particles in a two-dimensional cylindrical
coordinate system.
We assume that all the particles at this point and at this time have the same
average temperatures and velocities and do not interact to each other. Thus, it is
possible to apply the equations of heating and of motion of a single particle to
determine the parameters (temperature, velocity) of all the flow of particles. Also
note that the proposed model is valid only for the plasma flow that has a much
higher spatial resolution compared with the processed particles. Further in equa-
tions for simplicity we omit all the indices «p», referring to plasma.
426 10 Heat Exchange and Movement of the Solid Particles in the Plasma

1. Equation of heating of the particles

dTs 6   
qs cps ¼ a Tp  Ts  erTs4 þ Urad : ð10:45Þ
dt ds

2. The momentum equations of particles

dvsz qðvz  vsz Þjvz  vsz j

qs ¼ 0:75Cdz þ qs g z ; ð10:46Þ
dt ds

dvsr qðvr  vsr Þjvr  vsr j

qs ¼ 0:75Cdr : ð10:47Þ
dt ds

3. The equation of balance energy in plasma taking into account the loss of power
to the heating of the particles:
 

@T @T @ @T 1@ @T
q cp v z þ vr ¼ k þ rk þ r E 2  Urad
@z @r @z @z r @r @r
6 
 aðT  Ts Þ  erTs4 Gsn : ð10:48Þ
ds

4. The momentum equations of the plasma flow, taking into account the loss of
momentum due to the acceleration of particles.

@ 1@ @p
ð q vz vz Þ þ ðrq vr vz Þ ¼  þ FBz þ Flz þ q gz
@z r @r @z
qðvz  vsz Þjvz  vsz j
 0:75Cdz Gsn ; ð10:49Þ
ds

@ 1@ @p
ð q vz vr Þ þ ðrq vr vr Þ ¼  þ FBr þ Flr
@z r @r @r
qðvr  vsr Þjvr  vsr j
 0:75Cdr Gsn ; ð10:50Þ
ds

where FBz ; FBr —the components of the electromagnetic force; Flz ; Flr —
components of viscous forces; Gsn—the loading coefficient of the material.
The coefficient Gsn is defined as the volume ratio between the processed material
and plasma (the volume occupied by the particles in the unit of plasma volume). In
practice, Gsn is not often given as input data, but the mass consumption Gs of
material, that should be treated in the plasma flow (the performance of processing of
the material). Let the processed particle with density qs are fed and uniformly
distributed in the plasma flow with an average diameter D and the average velocity
in the initial section vs0. The relationship between Gsn and Gs is defined as:
10.3 The Model of Plasma, Loaded the Flow of Solid Particles 427

Gs
Gsn ¼ : ð10:51Þ
900qs pD2 vs0

Note that Gsn is a dimensionless coefficient. Relevant units of the following

values: Gs—kg/h, qs—kg/m3, D—m, vs0—m/s.
In the equation of balance energy of plasma (10.48) a new member appears,
taking into account the losses of power for heating of the particles. Similarly, in the
equations of motion of the plasma flow (10.49, 10.50) the members
0:75Cdz qðvz vszdÞsjvz vsz j Gsn and 0:75Cdr qðvr vsrdÞsjvr vsr j Gsn correspond to the losses of
momentum in due to the particle acceleration in two directions, z and r.
The solution of Eqs. (10.45–10.47) using the method of Runge–Kutta is pre-
sented above (Sect. 10.2.2), method for solving of the momentum equations and the
equation of energy balance in plasma (10.48–10.50) is given in Sect. 6.1. The
sequential procedure for solving of the equations, analysis of stability and the
organization of computing procedure are given in Sects. 6.2, 6.3.
Thus, the joint solution of the Eqs. (10.45–10.50) allows the determination of
not only the parameters of plasma and particles but also the effects of cooling and
braking of plasma at a given load coefficient of material. By adjusting the coeffi-
cient Gsn, we can find the limiting material consumption Gs, the efficiency and
effectiveness of the process.
Normally, the limit mass consumption of material Gs for this plasma unit must
ensure the achievement of the melting temperature of the particles Ts1 at the end of
the plasma flow and the preservation of high temperature of the plasma flow in the
discharge (T > 7000–8000 K).
The mass consumption of the material is the most important parameter in the
design of plasma installations for the processing of particles, and the determination
of its efficiency, and the product cost. The dependence of the heating temperature of

Fig. 10.11 The temperature

of the heating of particles in
dependence on the load
coefficient of MgO periclase
at their processing in the air
RF plasma torch (1.76 MHz–
60 kW)
428 10 Heat Exchange and Movement of the Solid Particles in the Plasma

Fig. 10.12 Changes of the

gas velocity (z = 800 mm) in
the air RF plasma torch
(1.76 MHz–60 kW) at the
maximum loading of MgO
periclase (Gs = 8, 6, 3 kg/h,
for the fractions ds = 100,
150, 200 lm, respectively): 1
without loading of particles, 2
ds = 200 lm, 3 ds = 150 lm,
4 ds = 100 lm

Fig. 10.13 Change of the gas

temperature (z = 800 mm) in
the air RF plasma torch
(1.76 MHz–60 kW) at the
maximum loading of MgO
periclase (Gs = 8, 6, 3 kg/h,
respectively, for the fractions
ds = 100, 150, 200 lm): the
notation is the same as in
Fig. 10.12

particles MgO of different fractions depending on the load coefficient of the

material Gs is shown in Fig. 10.11. From this figure it is possible to find the limiting
consumption, providing the melting point of the particles (Gs = 8, 6, 3 kg/h
respectively to fractions with ds = 100, 150, 200 lm). With increasing particle size
the limited mass consumption decreases. This is because the smaller particles have
a larger surface area and are heated faster than larger particles. Figure 10.12 and
10.13 shows the changes of velocities and temperatures of the gas (at the end of the
plasma jet) at the maximum load of MgO periclase particles. Comparing these data
with the data for the flow without downloading of particles, the degree of cooling
and braking of the plasma jet can be found.
Using the results of calculation of motion of particles and its heating, as well as
analyzing the criteria of similarity for a heat transfer and gas dynamics the flying of
10.3 The Model of Plasma, Loaded the Flow of Solid Particles 429

Fig. 10.14 The air plasma jet of RF plasma torch (0.44 MHz–1000 kW) at the testing of the
frontal part of the metal body (a) and the field of temperature and the flow of air plasma in its
channel (Q = 6000 l/min, U1 = 9.45 kV, I1 = 1200 A, Pstr = 400 kW) (b)

Fig. 10.15 Change of the

density of the heat flux going
into the heating of the front
part of the metal body with
diameter d = 50 mm in the
flow of the air RF plasma
torch (0.44 MHz–1000 kW):
z—the distance from the edge

bodies in the plasma flow can also be simulated. For example, the heat flux was
determined, that heat the front part of the metal body with a diameter d = 50 mm in
the jet of the air RF plasma torch (0.44 MHz–1000 kW). The main results of the
simulation are shown in Figs. 10.14 and 10.15.
430 10 Heat Exchange and Movement of the Solid Particles in the Plasma

References

1. Tsvetkov Y.V., Panfilov S.A. Low-temperature plasma in the recovery process. M.: Nauka.
1980.
2. Zhukov M.F., Solonenko O.P. High dusty powder material in the jet treatment processes.
Novosibirsk: Institute of thermal physics of Sib. Dep. USSR Academy of Sciences, 1990.
3. Coating Plasma / V.V. Kudinov, P.Y. Pekshev et al. M.: Nauka. 1990.
4. Donskoy A.V., Klubnikin V.S. Electrical plasma processes and installation engineering. —
Leningrad: Engineering, 1979.
5. Gorbis E.R. Heat transfer and hydrodynamics dispersed through streams. M.: Energy. 1970.
6. Eckert E.R., Drake R.M. The theory of heat and mass transfer / Trans. from English. Ed. A.V.
Lykov. M.: Gosenergoizdat. 1961.
7. Deitch M.E., Philipov G.A. Gas dynamics of two-phase media. M.: Energoizdat. 1981.
8. Suris A.L. Plasma-chemical processes and devices. M.: Chemistry, 1989.
9. Patankar S. Numerical methods for solving problems of heat transfer and fluid dynamics. —
M.: Energoatomizdat 1984.
10. Hirs D., Pound G. Evaporation and condensation. — M.: Metallurgy, 1966.
11. Fuchs N.A. Evaporation and growth of droplets in a gaseous medium. — M.:
Publishing USSR Academy of Sciences, 1958.
12. Nesis E.I. Boiling liquids. — M.: Nauka. 1973.
13. Knakke O., Stransky I.N. Evaporation mechanism // The success of Metal Physics. — M.:
Metallurgy, 1966. V. 3. PP. 222–282.
14. Frank-Kamenetskiy D.A. Diffusion and Heat Transfer in Chemical Kinetics. — M.: Nauka.
1987.
15. Kazenas E.K., Tsvetkov Yu.V. Evaporation oxides. — M.: Nauka, 1997.
16. Dresvin S.V., Panfilov S.A., Nguyen-Kuok Shi. Calculation of heating solids and the granules
in the plasma Arcs // Proc. Ros. Scientific and Technical. Conf. “Advanced technology
materials handling processes.” SPb. 1995. PP. 5–11.
17. Study of the metallurgical silicon powder treatment by thermal RF plasma / S. Magnaval,
D. Morvan, J. Amouroux et al. // Proc. of the IV European Conference on Thermal plasma
Processes. Athens. Greece, 1996. P. 453–459.
18. Deaconov V.P. Handbook of algorithms and programs in the language BASIC for the
personal computer. — M.: Nauka, 1989.
19. Krzyzanowski R.E., Stern Z.Y. Thermal properties of nonmetallic materials (oxides)
directory. — Leningrad: Energy. 1973.
20. Nguyen-Kuok Shi. Computer simulation of the flow of solid particles with plasma: PhD
Thesis. — Leningrad: St. Petersburg State Technical University, 1992.
Chapter 11
The Features of the Experimental Methods
and Automated Diagnostic Systems of RF
and Arc Plasma Torches

11.1 The Experimental Setup and Automated Diagnostic

System for the Study of the RF and Arc Plasma
Torches

The experimental setup and diagnostic system for the study of RF plasma torches
for spectral analysis are presented in Fig. 11.1. The RF generator has the operating
frequency of 27.12 MHz and power *7 kW. The plasma-forming gas is Argon.
The RF generator allows the adjustment of the voltage applied to the inductor, and
the launch of the plasma torch (plasma ignition) using a high-voltage discharge.
The operating parameters of generator are: the grid current—200 mA, the anode
current—0.7 A, voltage at the anode—2.4–3 kW.
Channel of RF plasma torch is a cylindrical quartz tube with internal diameter of
26 mm. The plasma forming gas is supplied through the central channel 19
(Fig. 11.1) with the internal diameter of 6 mm, the protective gas (Argon) is
introduced tangentially through the holes in the bottom side of the quartz tube of
plasma torch 24. The water-cooled copper induction coil 1 with two–three windings
provides the passage of the high-frequency current from RF generator (27.12 MHz)
on its internal side. The scheme of RF plasma torch, type of RF discharge, the
geometric dimensions of inductor with a quartz tube are shown in Figs. 11.2, 11.3
and 11.4.
The used in operating RF plasma torch allows obtaining plasma with maximum
electron temperature (10–11) 103 K, the electron concentration of the order
(1–3) 1016 cm‒3, and degree of ionization about 12–13 %. One of the features of RF
plasma is nonuniform change in the radial distribution of power density caused by
the skin effect in RF discharge (Sect. 7.2.3). The most significant specific energy is
carried in the external annular area of RF discharge. Therefore, the temperature and
the electron density in this region may exceed their values on the discharge axis.
The RF plasma emission spectrum consists of continuous emission (continuum) and
radiation in atomic and ionic lines.

© Springer International Publishing Switzerland 2017 431

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7_11
432 11 The Features of the Experimental Methods and Automated …

Fig. 11.1 Scheme of the experimental setup for diagnostic of RF plasma torches: 1, 2 inductor
and the body of RF plasma torch; 3 RF generator, 4 three-channel fiber optic spectrometer, 5
collimation lens, 6 reuter, 7 optical table, 8 fiber optic cable, 9 fiber optic splitter, 10 computer, 11
waveguide; 12 switch to turn on the air cooling of RF generator, 13 switch for the cathode filament
in a lamp of RF generator, 14 button for the power supply to anode, 15 the same to remove the
voltage, 16 button for applying of a high voltage to “ignite” RF discharge, 17 switch to turn on the
electronic unit for control of RF discharge, 18 the block devices for measuring of the net current,
current and voltage at the anode (lamp), 19 quartz axial channel, 20 rotameter of annular flow of
protective gas, 21 rotameter of axial flow of the plasma gas, 22 nipple for gas exhaust from plasma
torch to atmosphere, 23 valve for cooling water supply to inductor, 24 the quartz channel of
plasma torch, 25 the handle of fine adjustment of the anode voltage

For the determination of point for the spectral measurement the end of the optical
cable 8, directed to plasma, was installed in the desired position by means of a laser.
To do this, the laser is applied to the second end of the optical cable, and a laser
beam determines the desired point for diagnostic in RF or Arc plasma torches.
The automated system of the spectral measurements for obtaining complex
spectral characteristics of plasma radiation in different zones of RF and Arc plasma
torches determine the plasma parameters (concentration of excited particles, elec-
tron density, temperature, etc.). The system of spectral measurements include
spectral devices of high class of accuracy, the receivers of optical radiation with a
wide spectral range of sensitivity, means of transmission, registration, and pro-
cessing of information. In a first variant the system consists of the diffraction
spectrograph DFS-452 and multichannel optical spectrum analyzers MOAS-2, that
include two photo-head with a photodiode CCD rulers Toshiba TDS1250A
11.1 The Experimental Setup and Automated Diagnostic System … 433

Fig. 11.2 Scheme of RF

plasma torch

Fig. 11.3 Argon RF plasma

torch 27.12 MHz
434 11 The Features of the Experimental Methods and Automated …

Fig. 11.4 The geometrical

dimensions of inductor with a
quartz tube
Ø5

14
3

Ø26

Ø50

(CCD charge-coupled device), and printed circuit board with the ISA interface for
connection with computer. Photo-heads are fixed on the plate corresponding to the
shape and size of the cassette of spectrograph that is set to the place of cassette.
Photo-heads are connected with the interface board by a cable. Each photodiode
ruler contains 3724 light-sensitive elements with length of 8 μm and height of
200 μm (the total length of the line is 30 mm). The size of the spectral sensitivity
range is 180–1000 nm. The time of signal accumulation is given by period of the
survey lines and can be varied between 80 and 20,000 ms. The 12-bit
analog-to-digital converter is included in each photo-head, so only digital signals,
that provide a sufficiently high noise interference protection of the system, are
transmitted to computer by an interface cable. For operating with the system
MOAS-2 a special program is used.
The spectrograph DFS-452 allows working in all over the range of sensitivity of
rulers. In the work with a diffraction grating with 600 lines per 1 mm the inverse
linear dispersion is 1.6 nm/mm or 0.013 nm per cell of ruler and in the work with a
diffraction grating with 1200 lines per 1 mm—respectively, 0.8 nm/mm and
0.0065 nm cell of ruler.
One of the rulers is disposed along the direction of dispersion of the spectro-
graph and serves for detection of the spectrum, and the other—perpendicularly—is
used for the determination of intensity depending on the height of the gap at the
predetermined wavelength.
All received data are transmitted to the personal computer. Program for oper-
ating with the system MOAS-2 allows us to survey any of the photodiode rulers or
both simultaneously with arbitrarily set accumulation time once or cyclically. It also
allows display signal from any ruler or two signals from both of the rulers simul-
taneously. Thereafter, data are written to the disk for further processing.
11.1 The Experimental Setup and Automated Diagnostic System … 435

Fig. 11.5 Scheme of the automated system of spectral measurements for Arc plasma torches: 1
alignment laser, 2 cathode, 3 camera with two quartz optical windows, 4 anode, 5 rotating mirror,
6 lamp with hollow cathode LT-2, 7 band tungsten lamp, 8 condenser, 9 Dove’s prism, 10 split of
spectrograph, 11 spectrograph, 12 drum of wavelengths, 13 CCD ruler, 14 computer

Scheme of automated system for the spectral measurements is shown in

Fig. 11.5. The general view of the stands and automated system for the study of free
arc and RF plasma torches is represented in Fig. 11.6.

11.1.1 Adjusting Elements in Optical System

The purpose of adjustment is the location of selected area of plasma on the optical
axis (z = z0, r = 0) and quartz condenser with a focal length f = 150 mm, providing
the projection of increased sharp image on the middle of spectrograph slit with
height H. Alignment is implemented with a laser in the visible range using the
selected diffraction grating (600 or 1200 lines per mm) with open slit Δint = 200–
400 μm in the following sequence:
1. In the absence of condenser the hit of laser beam to the middle of diffraction
grating in spectrograph is provided. This is achieved by visual observation from
the cassette part of DFS-452 of the movement of light spot on the surface of
diffraction grating (as the moving of laser);
2. Using the adjustment screws of the Arc chamber the z coordinate of the Arc is
approximately hold up on the optical axis, defined by the laser beam;
3. According to the equation of thin lens 1f ¼ 1a þ 1b the position of condenser on the
optical axis is selected, it provides the selected increase k = b/a.
The next adjustable element is the gas-discharge lamp with a hollow cathode
LT-2, emitting the spectrum of atoms CuI in the wavelength range 200–600 nm and
the spectrum of atoms NeI in the range 600–900 nm. This lamp is used to perform
two important tasks of spectral studies:
1. For the calibration of the wavelength scales of spectrograph, i.e., for establishing
communication between the wavelengths on the limb of drum kl in DFS-452
(approximately corresponding to the wavelength in the middle of the focal plane
436 11 The Features of the Experimental Methods and Automated …

Fig. 11.6 Experimental setup and automated diagnostic system of free Arc plasma torch (a) and
RF plasma torch (b)

of device) and a wavelength at the CCD ruler. This calibration is carried out by
the registration of spectrum of the lamp LT-2 emission at a fixed wavelength kl
and its decryption using a database on the position of the radiating lines of
copper atoms CuI (the example of such a calibration is shown in Fig. 11.7).
2. For measuring the hardware function of the used optical system. The hardware
functions is formed by the basic elements of spectrograph—an entrance slit and
the diffraction grating—and is included in the monitoring contour of the spectral
lines in the form of a convolution of physical contour of the observed lines and
11.1 The Experimental Setup and Automated Diagnostic System … 437

Fig. 11.7 The identification of the emission spectrum for copper CuI

hardware functions. A feature of radiation for the lamps with hollow cathode is a
small physical width of the spectral lines of CuI and NeI, due to the small Stark
component (ne  1012 cm‒3, dkSt  0:01 Å), and the Doppler width of the lines
(Ta  0:05 eV  Te ; dkD  0:04 Å). Therefore, registration of the spectrum of
this lamp with a narrow entrance slit of spectrograph Dent D (Dent —opening of the
split, mm; D—the reverse linear dispersion, Å/mm), less than the expected
physical line width of ArI and ArII in Argon plasma d1=2 : Dent D\d1=2  0.2–
2 Å allows obtaining of the hardware function of the measuring system in the
entire range of wavelength 250–950 nm. Chosen spectrograph slit width should
be coordinated with the spectral resolution of the system. In this case, the res-
olution is equal to the interval, falling at one photosensitive element of CCD ruler
(one pixel with width Dpix ¼ 8 lm) Dk1200 ¼ 8  103  8 ¼ 0:064 Å and
Dk600 ¼ 8  103  16 ¼ 0:13 Å for diffraction gratings with 1200 and 600
lines per mm, respectively.
For full use of the resolution of system for registration of narrow spectral lines
the disclosure of spectrograph slit should be chosen smaller than the pixel size
Dent \Dpix ¼ 8 lm, and for the reliable determination of the hardware function of
the measuring system the contours of the spectral lines of CuI with high ratio of
signal to noise should be obtained.
The value of the hardware function is installed as the mean arithmetic value of
the half-widths of several measured line contours:
PN
d1=2
dhf ¼ 1
ð11:1Þ
N

By analyzing the spectrum for CuI in the wavelength range 320–570 nm, it was
found that hardware function dhf ¼ 0:2 Å.
438 11 The Features of the Experimental Methods and Automated …

Table 11.1 Features of the channels in spectrometer Avaspect of the company Avantes
Characteristics First channel Second channel Third channel
The wavelength range (nm) 195–469 455–593 487–1100
Resolution (nm), not less 0.18 0.12 0.3
Type of diffraction grating 1200 1800 600
(lines per mm)

Similarly, the standard source of brightness is adjusted—it can be the tungsten

band lamp with a graduated dependence of temperature at the surface of tungsten
from the direct current, flowing through the tungsten spiral of lamp. At this adjust-
ment the mirror and the condenser, as in the case of a lamp LT-2, provide a projection
of the sharp image of tungsten spiral to the entrance split of spectrograph DFS-452.
Adjustment of the spectrum analyzer MOAS-2 is to provide a contact of the laser
line on the horizontal CCD ruler. This is carried out (when the system MOAS-2 is
switched on) by moving the cassette part of spectrograph by the switch toggle “up–
down” for moving the laser signal that is getting on the CCD ruler (at the minimum
split of DFS–452 Dpix  10 lm). To implement this procedure, it is necessary to use
the calibration characteristic of the system connecting the values of wavelength at
the limb of spectrograph with a wavelength at the center of the horizontal CCD
ruler. This curve depends on the chosen diffraction grating (1200 or 600 lines per
mm) and on the orientation of the system MOAS-2 that allows two positions in the
focal plane of DFS-452: using the horizontal ruler in the long-wave (right) or
short-wave (left) part of the spectrum by turning the system to angle 180°.
Instead of DFS-452 the spectrometer Avaspect of the company Avantes,
equipped with three channels can be also used. The characteristics of the channels
are shown in Table 11.1.
For all the channels of spectrometer the necessary spectrometer calibration was
made that provided the absolute values of the radiation intensity. The spectrometer
was connected to computer using the USB cable and working with it was carried
out using the software «Avasoft». It allows checking of changes in spectrum or in
its parts at the real time, obtaining the values of intensity or wavelength for the
individual points and record the resulting data in formats suitable for further
decoding the spectrum and the determination of plasma parameters.

11.2 The Basic Methods of the Plasma Diagnostics

The main task for the diagnostics of plasma is to determine the local values of the
following parameters:
• concentrations of atoms na , electrons
 ne , and ions ni ; 
• concentration of excited atoms na Ea and ions ni Ei ;
11.2 The Basic Methods of the Plasma Diagnostics 439

• temperature of electrons Te and heavy particles Ta ;

• distribution function of the electron energy fe ðeÞ;
• intensity of electric E and magnetic fields H;
• Volt–ampere characteristics of the Arc and others.
Establishing relationships between the plasma parameters with its spectral
characteristics in a mathematical sense is an inverse problem, the accuracy of its
solutions is often linked to the right of the model for the description of study
plasma. The main source of information for us will be the absolute and relative
intensities of the spectral lines and continuum.
For RF and Arc plasma torches a model of local thermodynamic equilibrium
(LTE) can be used, and the Maxwell, Saha, and Boltzmann equations are valid. In
the spectral diagnostics, the following basic methods can be used:
1. Determination of the electron temperature from analysis of the relative distri-
butions of excited atoms and ions as a function of the excitation energy. The
range of changes in the excitation energy of atoms and ions should be
approximately equal to kTe and more.
2. Calculation of temperature of the electrons from the decay of continuum
intensity at short wavelengths (λ ≤ 300 nm). In this case, obtained dependences
of the absolute intensity of continuum from the radiation frequency are
approximated using the model of local thermodynamic equilibrium: ε(ν) * exp
(−hν/kTe).
3. Determination of the concentration of electrons using the Stark width of the
lines for atoms and ions. This method does not require any models describing
plasma—the necessary condition for its application is the predominant role of
electrons (and ions) in the broadening of the studied spectral lines, as well as
commensurate (or exceed) of the Stark width of investigated line with its
hardware broadening. The movement of particles in dense plasma and associ-
ated with it change in the distance between them lead to rapid changes in the
electric field around each of emitting particles. As a result, the energy levels of
atoms (ions) are displaced by different distances and it leads to broadening of the
spectral lines in the emission spectrums of plasma. The Stark broadening allows
to estimate the concentration of charged particles in plasma.
The studied plasma in RF and Arc plasma torches even in the axial region maybe
not in a state of full thermodynamic equilibrium primarily due to the fact that
induction and Arc discharges are the “open” systems. The exhaust of radiation and
heat (due to gradients of parameters), and the presence of an external electric field
leads plasma to the state different from the state of full thermodynamic equilibrium.
However, in areas with significant concentrations of electrons (these particles have
“lion’s share” of the energy of the field E, try to maintain a state of thermodynamic
equilibrium) the conditions of local thermodynamic equilibrium can be ensured. In
such conditions, the following basic equations are valid: normalized per unit dis-
tribution function of the electron energy is described by Maxwell equation;
behavior of excited atoms and ions is described using the Boltzmann equation at the
440 11 The Features of the Experimental Methods and Automated …

electron temperature; the ionization equilibrium in plasma is described by Saha

equation at the ionization temperature, equal to the electron temperature.
Numerous calculations and experiments [1–3] show that in the case of spatially
inhomogeneous plasma in the areas of highest values of electron density the
Maxwell distribution of electrons fe ðeÞ  fMe ðeÞ inside the core of function (ε = 0–
7 eV) can be expected. When the degree of plasma ionization ai ¼ na nþe ni [ 102
the population of the upper excited states of atoms and ions are often close to the
Boltzmann values and ionization equilibrium is described using the Saha equation.
Method of Relative Intensities for the Determination of Electron
Temperature The electron temperature is one of the main plasma parameters. Its
value is necessary to calculate the composition of plasma and its transfer coeffi-
cients, for the determination of electrical, magnetic, energy, and thermal charac-
teristics of the discharge. Furthermore, the temperature of electrons allows us also
to find the interaction cross sections of particles in plasma, its thermodynamic
properties, and many fundamental physical parameters, such as ionization and
recombination coefficients, thermal diffusion coefficients, and others.
According to the literature, for example, linear and continuous emission of
Argon plasma at atmospheric pressure in the studied range of temperature and
wavelength is optically thin. Therefore, the intensity of the spectral line, W/(cm3
sr), at the transition of atom (ion) from the kth level to the ith level is expressed by

Z1
1
Iki ¼ hmki Aki nk ¼ iki ðkÞdk; ð11:2Þ
4p
0

where mki —the frequency of the line center ðmki ¼ mo ¼ c=k0 Þ, Aki —the probability
of spontaneous transition from the kth level into the ith level (the first Einstein
coefficient); nk—population of emitting (upper) level of transition. Equation (11.2)
is accurate and then

ZL
s0 ¼ k0 dl  1 ð11:3Þ
0

Here s0 is the optical thickness of plasma layer with depth L, it should be small. In
(11.3) k0—absorption coefficient at the center of the studied line.
Thus, the ratio (11.2) allows us to analyze the populations of excited states of
atoms and ions with different excitation energies Ek, Ei. In the context of local
thermodynamic equilibrium the Boltzmann law is valid,
 it binds the concentration
of particles of the excited levels nk
gk ¼ ngii exp  EkkTE
e
i
; where gk , gi —the statistical
11.2 The Basic Methods of the Plasma Diagnostics 441

weight of the particles of kth and ith levels. Therefore, the analysis of experimental
curves ngkk ¼ f ðEk Þ allows the determination of electron temperature using:

• relative intensity of two lines in this class (ArI or ArII):

 
Iki Ek  Em
kTe ¼ f ¼  : ð11:4Þ
Imn ln IIkimnmmnmkiAAmnki ggkm

• relative intensities of lines with different multiplicity of ionization (Saha and

Boltzmann equations are valid):
 
Ikii 3=2 1 Ema  Eki  Iia
a
¼ kTe exp  : ð11:5Þ
Imn ne kTe

In this case, for the determination of electron temperature knowledge of their

concentration ne is required.
The analysis of dependences ngkk ðEk Þ for atomic and ionic lines is the most
informative. It is possibly covering the greatest variation range of the excitation
energy for the particles, because the determining error for Te type from the
expression (11.4) dTe DE Te
k
is reduced with the increase of this range. In the same
direction, it operates a statistical set of values nk =gk as much as possible the number
of excited states. Evaluation is both a method of checking the correctness of the
Boltzmann law to communicate the populations of the levels of particles.
The Abel Transformation for the Determination of Radial Distribution of
Plasma Parameters Abel transformation is the standard procedure of transition
from the chordates intensities to the radial ones if there is axial symmetry of the
emitting object. The relationship between the intensities is described using the Abel
integral equation.
In the case of the light source with cylindrical or spherical symmetry the
observed intensities can be transformed into the radial distribution of plasma
intensity. Let the z-axis to be the axis of symmetry for the cylindrical discharge. The
x-axis of the circular plasma disk is aimed to the observer I (Fig. 11.8). The
observer I is looking along the line, parallel to the x-axis and passing at the distance
y from the center. The local emissivity of plasma at a distance r from the axis is
noted as ε(r). The observer I sees the projection [integral I(y)] as axial-symmetric
function ε(r) along the direction of observation. The function I(y) corresponds to the
radiation intensity integrated over x at the distance y from the plane xz. Then the
function I(y) is linked to ε(r) by the integral Abel transformation:
pffiffiffiffiffiffiffiffiffi
ZR0 y
2 2
ZR0
eðrÞrdr
IðyÞ ¼ 2 eðrÞdx ¼ 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi; ð11:6Þ
r 2  y2
0 y
442 11 The Features of the Experimental Methods and Automated …

Fig. 11.8 The partition of the

plasma column into the
cylinders

where R0—radius, ε(r) = 0. Note that the variable r is associated with x and y as
r2 = x2 + y2, if integrating it does not change the sign, then the integrand is an even
function.
Transformation of Eq. (11.6) gives equation for the required value ε(r):

ZR0
1 dIðyÞ dy
eðrÞ ¼  pffiffiffiffiffiffiffiffiffiffiffiffiffiffi: ð11:7Þ
p dy r  y2
2
r

Equation (11.7) is back of Eq. (11.6), it is used to restore the function ε(r) using
its projection I(y).
Included in the limits of integration radius R0 is the desired optical radius—it
corresponds to the vanishing of the experimental intensity I(y). As it is adopted, we
provide the calculation using (11.7) in relative coordinates r/R0, where r—the
current radius.
Carrying out the Abel transformation allows obtaining the relative radial dis-
tributions from the chordates distributions that are registered during the experiment.
However, radial distribution of plasma parameters is of the most interest. Thus,
there is the problem of transition from the radial distributions of the relative
intensities of radiation to the radial distribution of the main parameters of plasma.
The basis for the solution of this problem is the following relationship for the
absolute intensity of the continuous radiation, containing single and double ions:

  expðhm=Te Þ
eff ¼ const ne ðTe Þniþ ðTe Þ þ 4ne ðTe Þniþ þ ðTe Þ pffiffiffiffiffi : ð11:8Þ
Te

Equilibrium plasma composition ne ðTe Þ; niþ ðTe Þ; niþ þ ðTe Þ can be obtained
from the system of Saha equations. If we know the temperature of electrons in the
axial zone Te0 (from the spectrometric measurements) and the equilibrium
11.2 The Basic Methods of the Plasma Diagnostics 443

ðTe Þ
ff
composition of plasma, it is possible to build a dependence eeff ðT e0 Þ
, compare it with
the existing radial relative intensity distribution of radiation, received at wavelength
corresponding to the continuum, we can obtain the dependence Te(r). Further, the
known distribution Te(r) and the composition of plasma equilibrium allow the
building of distributions ne ðrÞ; niþ ðrÞ; niþ þ ðre Þ; na ðrÞ.
The Method for Determination of the Electron Density in Plasma The electron
density can be found from the absolute intensity of the continuum at any
(nonequilibrium) plasma state. The contour of spectral lines at the emission of
low-temperature plasma contains information about the density of electrons
(broadening of line due to the quadratic Stark effect), and temperature of the heavy
particles (Doppler effect due to thermal motion of the studied atoms or ions). As it
follows from the data in [4], in the area with the density of electrons in 1016–1017
cm‒3 the Stark width of atomic lines is 0.1–3 Å, i.e., it lies in the area of available
spectral resolution, and the width of ion lines, typically, in 2–3 times more narrow.
As for the Doppler width of the lines ArI and ArII, it is in the temperature range
qffiffiffiffiffiffiffiffiffiffiffiffiffi
(10–20) 103 K even if k0 ¼ 800 nm and dkD ¼ 2 kc0 2kT M ln 2  0:15 Å or less than
the hardware functions of the measuring system.
Therefore, the practical work with the contours of the spectral lines—to analyze
the mixed (Voigt) contours of lines, the Lorentz component of them corresponds to
the Stark broadening of lines, and the Gaussian component is the sum of the
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Doppler effect and hardware broadening: dkG ¼ dk2D þ dk2hf .
Thus, right selection of specific spectral lines of atoms and ions can successfully
help to solve two problems of the spectral diagnostics:
• Measuring Te using the analysis of dependence ngkk ¼ f ðEk Þ;
• Measuring ne using the Lorentz component of the Voigt profiles of the spectral
lines.
One of the main spectral characteristics that it is necessary to be obtained from
the recorded spectrums is width of the spectral lines at the half of its height. It is
very important to identify this characteristic correctly. Hardware circuit of the
recording system has a Gaussian form:
"  #
I0 k  k0 2
IG ðkÞ ¼ pffiffiffi 0 exp  ; ð11:9Þ
pdkG dk0G

where λ—wavelength; I0—intensity of radiation in the center of line; λ0—wave-

pffiffiffiffiffiffiffi
length at the center of the line; dkG ¼ 2 ln 2 dk0G —the line width at the half of
height.
Lorentz contour describes the summary influence of several mechanisms of
broadening and several widths: the natural width, the width of the Van der Waals
δλV (turbulence of the radiating level by the neighboring excited states, not optically
444 11 The Features of the Experimental Methods and Automated …

connected with the ground state of atom or ion), and the width of the resonance
broadening δλr, due to the interaction with the ground state, or bound states (for
example, the first excited state). In the conditions of low-temperature plasma at
atmospheric pressure only Stark broadening mechanism plays an important role.
The broadening of the contour due to the quadratic Stark effect has the Lorentz
form:

I0 dkL
IL ðkÞ ¼ ; ð11:10Þ
2p ðk  k0 Þ2 þ ðdkL =2Þ2

where dkL —the line width at the half of height for the Lorentz contour.
Thus, the studied experiment lines are described by two contours—the Gaussian
and the Lorentz. The resulting line is the superposition of these contours and looks
as the Voigt contour:

Zþ 1
a expðy2 Þ
IF ðkÞ ¼ I0 dy; ð11:11Þ
p ðu  yÞ2 þ a2
1

where IF(λ)—the intensity of Voigt contour in dependence of the wavelength;

u ¼ kk
dk0
0
; a ¼ 2dk
dkL
0 .
G G
This expression using the Voigt contour in the future is a function of wavelength
that is defined using four parameters: the intensity of radiation I0 and the wave-
length λ0 in the center of spectral line, the Gaussian dkG , and the Lorentz dkL
half-widths IF ðkÞ ¼ f ðk; I0 ; k0 ; dkG ; dkL Þ.
The unknown quantity is the half-width of the Stark broadening that should be
determined from the available experimental Voigt contour. For this the half-width
dkF of the Voigt contour is determined by the direct measurement. After the
determination of this value and if the value dkG is known, the value of the
half-width of the Stark broadening using the expression shown in Fig. 11.9 [5] can
be found.
The method for the determination of electrons concentration is based on the
comparison of known tabular data on the Stark broadening constants [6] with
the line width at the half of height for the Lorentz contour, obtained in experiment.
The concentration is calculated from the equation

ne ðdSt Þ
e ðdkL Þ ¼
nexp dkL ; ð11:12Þ
dSt

where dSt —the Stark width at the half of height of the spectral line ne ðdSt Þ—
electron density that is corresponding to the Stark width at the half of height, taken
from the reference data for a given temperature of plasma; dkL —the experimental
11.2 The Basic Methods of the Plasma Diagnostics 445

Fig. 11.9 The dependence of

dkL dkF
2dkG from 2dkG

Stark width at the half of height of the spectral line. The obtained data on the
electron density can be compared with the results of solving the Saha equation.
Explanation of Superposed Lines Because each line represents the Voigt contour,
then the superposition of several spectral lines of the intensity of the overall circuit
I(λ) will consist of the intensities of all lines and continuum emission intensity:
X
IðkÞ ¼ SðkÞ þ IFk ðkÞ; ð11:13Þ
k

where k—the notation of the spectral line.

Using the dependence of intensity for each of the four Voigt contours from four
parameters (I0, λ0, dkL , dkG ) the expression for the function of total contour con-
sisting of P lines, can be written as
 
I k; I01 ; k10 ; dk1G ; dk1L ; I02 ; k20 ; dk2G ; dk2L ; . . .; I0P ; kP0 ; dkPG ; dkPL ; SðkÞ ¼ SðkÞ
   
þ IF k; k10 ; I01 ; dk1G ; dk1L þ IF k; I02 ; k20 ; dk2G ; dk2L þ


ð11:14Þ
 P P P 
þ IF k; I0 ; k0 ; dkG ; dkL ; P

where the superscript of parameters I0, λ0, dkL , dkG denotes the number of spectral
line, and registered contour is the described function by the functions of the
intensity of continuum and the Voigt contour s, each of them is defined by four
parameters I0, λ0, dkL , dkG .
The spectrometer presents the data about spectrum in the following format
hkexp exp exp
i i  hIi i, where ki —the wavelength corresponding to the ith pixel in the
exp
CCD detector, Ii —the intensity of radiation registered by the ith pixel. For the
decryption the recorded spectrum is divided into sections, each of them contains a
446 11 The Features of the Experimental Methods and Automated …

certain number of data points, and for each region containing P lines, the following
system of equations is solved:
 
I kk ; I01 ; k10 ; dk1G ; dk1L ; I02 ; k20 ; dk2G ; dk2L ; . . .; I0P ; kP0 ; dkPG ; dkPL ; S ¼ Ikexp ;
 
I kk þ 1 ; I01 ; k10 ; dk1G ; dk1L ; I02 ; k20 ; dk2G ; dk2L ; . . .; I0P ; kP0 ; dkPG ; dkPL ; S ¼ Ikexp
þ 1;
..
.
 
I kk þ M ; I01 ; k10 ; dk1G ; dk1L ; I02 ; k20 ; dk2G ; dk2L ; . . .; I0P ; kP0 ; dkPG ; dkPL ; S ¼ Ikexp
þ M;
ð11:15Þ
 
where I kk ; I01 ; k10 ; dk1G ; dk1L ; I02 ; k20 ; dk2G ; dk2L ; . . .; I0P ; kP0 ; dkPG ; dkPL ; S —the analytical
record of the Voigt contour; λ—experimental wavelength; k—the number of the
first line of array for M lines, corresponding to a given area, for the substitution to
the system of equations. In the range of decryption of part of the spectrum the
intensity of continuum can be considered constant [function SðkÞ can be replaced
by constant S].
Parameters I0, λ0, dkL , dkG , S are taken as unknown variables. If the number of
Eq. (11.15) (i.e., the experimental points of the plot) is equal to the number of
variables in the system or exceeds it, the system can be solved. However, the exact
solution of the system is not possible to get, because the experimental values of
intensity cannot be exactly equal to the theoretical one due to the influence of
random errors, the noise in the CCD detector, and a number of other factors, the
account for all of them is impossible. However, an approximate solution can be
found, it provides the best agreement of analytical contour with the experimental
one. The special function Minerr of the mathematical package MathCAD is used to
find the approximate solution. It differs from other functions for solutions of
equations that iterate until the solution stops “improving”, and when it occurs,
returns the best result. The decryption of recorded line using three superposed lines
of spectrum is presented in Fig. 11.10.

Fig. 11.10 The result of

decryption for three
superposed lines
11.2 The Basic Methods of the Plasma Diagnostics 447

This algorithm allows theoretical decryption of almost any of the spectrums.

Basically the number of lines for consideration is limited only by the speed of
computer and by the number of data points belonging to each of lines in this area.
At least four points that do not lie on the level of continuum should be matched to
one line, i.e., for superposed lines, the number of points should be equal to 4R,
where R—the number of superposed lines. This should be taken from the type of
Voigt function that consists of four parameters I0, λ0, dkL , dkG which correspond to
the kind of line and one of them S—for her “rise above the x-axis.” Processing all
the spectrum gives the decoded lines with specific parameters of the contours and
lines from the reference data associated with them and contains the necessary
constants for further calculations. Using this information, it is sufficient to substitute
the required data into the equations for the determination of electron temperature by
the method of relative intensities and the concentration of electrons by the Stark
broadening.
Note that the presence of a greater number of points of continuum increase the
accuracy of the finding of theoretical circuit, particularly if there is a sufficiently
large number of points that belong to a single line, there is no need to solve the
problem of determination of its parameters with the help of a system of equations,
because the spline, constructed using the experimental points, describes it quite
significantly.
The Absolute Calibration of the Intensity of Radiation and Registration of the
Spectrum Using the Standard of Brightness For obtaining the absolute values of
intensity of the spectral lines and the continuous radiation of plasma it is necessary
to compare the registered in experiment relative intensities of the radiation with
intensities of the standard source. In the wavelength range of 320–1000 nm such a
source can be the calibrated tungsten band lamp with the known dependence of
brightness temperature in the pyrometric wavelength (λ ≈ 650 nm) from the current
flowing through the band. The tungsten lamp with small dimensions with
Tbr = 2400–2500 K at currents of 3–3.5 A was used in our work.
The calculations of absolute spectral intensity of radiation, W/(cm3 nm sr), are
performed by the relation:

isp
k st k
k ¼
eabs I ; ð11:16Þ
istk k;T R0

where ispk —the recorded in relative units spectral intensity of plasma emission; ik —
st

the intensity of the radiation of standard lamp, expressed in relative units; Ik;T —the
st

spectral intensity of radiation for the standard tungsten band W/(cm2 nm sr); R0—
length of radiator in the direction of observer, cm; k—coefficient, that takes into
account differences in recording of the spectrums for plasma and standard lamp.
st
The dependence Ik;T from the frequency is obtained using the Planck’s law and
the literature data about the effect of temperature and wavelength to the emissivity
of tungsten:
448 11 The Features of the Experimental Methods and Automated …

2hc2 ew ðT Þ
st
Ik;T ¼ h  i wk br ; ð11:17Þ
k5 exp kThcbr  1 epyr ðTbr Þ

where ewk ðTbr Þ—the degree of black for the tungsten surface at the temperature Tbr,
and the wavelength λ, ewpyr ðTbr Þ—the same at pyrometric wavelength.
For the correctness of the procedure of absolute calibration of spectrum the
standard source should be placed on the optical axis of our system instead of
discharge, maintaining the position of condenser as the same. It will project the
sharp image of a vertical tungsten band to the entrance slit of spectrograph. The
registration of the spectrum of a standard source should be carried out at the same
spectral ranges as for plasma spectrums. Thus, we ensure the full compliance of
registration schemes for plasma and for the standard source. In the spectrum range
of λ = 200–300 nm, the deuterium lamp can be used with continuous radiation as a
standard source.
The Method of Measuring the Magnetic Field of Plasma Magnetic probe can be
used for measuring intensity of the magnetic field of RF plasma torch. In the
magnetic probe method the induction of the magnetic field is characterized by
electromotive force at the magnetic probe. The electromotive force is recorded
using oscilloscope.
The magnetic probe is a coil of several turns of a conductive material that is
placed into the investigated magnetic field. The flow of the vector of magnetic field
induction for one turn is defined as

U ¼ B? Sc ; ð11:18Þ

where B? —the projection of the vector of magnetic field induction on the normal to
the surface turns; Sc—the square of the windings in the coil.
For coils of identical turns, if we take the magnetic field inside it as uniform,
electromotive force is defined as the product of the changes in time for Ф and the
number of turns N:

dU
e¼ N: ð11:19Þ
dt

If to place a magnetic probe into the region near inductor of the plasma torch, the
induced magnetic field will create electromotive force inside this probe and its
change can be fixed with oscilloscope. The peak value of the magnetic induction
can be found using the equation
e
B¼ ; ð11:20Þ
2pfSeff

where Seff—effective square of the cross section in the coil; f—the frequency of
oscillation of the field. The intensity of magnetic field is defined as
11.2 The Basic Methods of the Plasma Diagnostics 449

B
H¼ : ð11:21Þ
l0 l

In plasma μ = 1,
e
H¼ : ð11:22Þ
8p2 107 fSeff

The experimental value of electromotive force used in the calculation is defined

as the root mean square of the oscilloscopic waveform:
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u Zt2
u
u 1
e¼t e2os dt; ð11:23Þ
t2  t1
t1

where t1, t2—the start and the final time of readings; eos —the instantaneous elec-
tromotive force, obtained from the oscilloscopic waveform.

11.3 Measurements of Plasma Parameters in the RF

and Arc Plasma Torches

Registration of the Emission Spectrums of Argon Plasma Atoms Ar, as well as

other atoms of inert gases (except He), are filled with p-shell (4p6), that is especially
stable. Therefore, for the excitation of one of the p-electrons the high energy of
levels is required (the ionization potential of the Ar atoms EI = 15.76 eV, and the
excitation energy of the lowest excited state E1 = 13.08 eV). Due to the significant
energies of excitation the atom (and ion) spectrum decays into Argon spectrum
lying in the wavelength k\180 nm corresponding to a transition from the excited
to the ground level, and on the spectrum, that lies in the infrared, visible, and
near-ultraviolet ranges and appearing in the transitions from one excited state to
another one [5, 7].
The emission spectrum of Argon plasma at atmospheric pressure in the range of
λ = 400–450 nm (Fig. 11.11) at the temperature 16 × 103 K represents some of
continuous emission spectrums arising in free–free and free–bound transitions of
electrons to the field of ions, and a large number of spectral lines of atoms
(spectrum of ArI) and ions (spectrum of ArII).
The Program of the Automated Processing of Plasma Emission Spectrums
The automated program for the registration and processing of the spectrum of
plasma is carried out in mathematical package MathCAD 14 of the professional
firm Mathsoft. The program is a set of procedures that are common for all the
processed spectrums. It consists of two modules, processing the information
received from two CCD lines (spectrum itself and the radial distribution). The work
450 11 The Features of the Experimental Methods and Automated …

Fig. 11.11 The emission spectrum of Argon plasma in the open free Arc in the range λ = 400–
450 nm at distance from the cathode tip z = 2 mm (I = 80 A, a diffraction grating—600 lines per
mm)

with spectrum chord distribution is processed at first in order to obtain an optical

radius of the cable R0, and to use the conversions of Abel. The value of optical
radius of image is written into the file «R0» in the ASCII symbols and then is used
in a second module of the program for the determination of absolute values of the
intensities of lines.
Processing the spectrum in the program starts with the input of the following
data:
1. Parameters of the experimental spectrum;
2. Spectrum of the standard (tungsten bands);
3. Database for the lines of the studied spectrum;
4. Number of points in continuum;
5. Coordinates of points in continuum;
6. Coordinates of base points are sets of the start and the final wavelengths (in
nanometers) of the studied range;
7. Conditions of the experiment (the slit width and the registration time of the
standard spectrum and plasma).

Explanation of the Spectrum At the second stage of processing a comparison of

two lines of the experimental spectrum with lines from a database of spectrums [4]
is carried out. This comparison is carried out by entering the experimental and table
base points. The comparison can be performed separately for two atomic or ionic
lines or in their totality. As a result of decoding all of the experimental lines should
coincide with the tabulated data (Fig. 11.12).
Determination of Absolute Intensities in Emission Spectrum of Plasma The
third step is determination of local intensities, i.e., the transfer of the relative
 11.3
Measurements of Plasma Parameters in the RF and Arc Plasma Torches

Fig. 11.12 Decoded spectrum of Argon plasma in open free Arc in the wavelength range λ = 400–450 nm for z = 2 mm, I = 80 A:
1 experimental spectrum, 2 lines ArI, 3 lines ArII
451
452 11 The Features of the Experimental Methods and Automated …

intensities into absolute units [W/(cm3 nm sr)]. In the program the procedure for
determination of the absolute intensities is carried out automatically. The absolute
intensity of the spectral lines and continuum used tungsten band lamp TRSH-2850,
calibrated with precision (accuracy class of 0.01) of the optical pyrometer EOP-66
in the range of brightness temperatures of 2000–2550 K for pyrometric wavelength.
Determination of Temperature and Electron Density Analysis of the experi-
mental dependences of populations of the kth excited levels of atoms and ions
gk ¼ f ðEk Þ in assumption of quasi-Boltzman distribution of atoms and ions of the
nk

energies Ek, allows us to find the electron temperature from the relative intensities
of the lines in this class (ArI or ArII). Using a large number of registered lines of
ionized argon with known probabilities of transition the temperature is determined
by taking into account the slope of the line lnðnk =gk Þ ¼ f ðEk Þ drawn through the
points corresponding to the experimental populations of the excited states of par-
ticles of the class: kTe ¼ tg1 lnðnk =gk Þ.
Measuring the Parameters of Open Free Arc The main results of the mea-
surement of open free Arc for welding are presented here, data of simulation are
presented above (Sects. 8.4.3, 8.5.3). An example of determination of the electrons
temperature along the lines of one type (ArII) using the method of relative inten-
sities is shown in Fig. 11.13 and the results of determination of the electrons
concentration are shown in Fig. 11.14. The electron density was determined using
the Stark broadening of the lines ArI and ArII that was obtained in program from
the Voigt contours of registered lines. The temperature distribution along the radius
of the free arc, obtained using the Abel transformation is shown in Fig. 11.15.
Measurements of External Characteristics of Open Free Arc Initial parameters
of Argon Arc:
Tungsten cathodes of type VI-15 (30) doped with Y2O3 [1.5 (3) %]; VT-15 (30, 50)
doped with ThO2 [1.5 (3, 5) %];
Internal diameter of nozzle—10–14 mm, the external diameter of the nozzle—15–
19 mm;

Fig. 11.13 Determination of Te using the spectral lines ArII on the axis of the open free Arc
(z = 2 mm)
11.3 Measurements of Plasma Parameters in the RF and Arc Plasma Torches 453

Fig. 11.14 Determination of

ne at the axis of the open free
Arc after processing 60 lines
ArI and ArII (z = 2 mm,
diffraction grating—600 lines
per mm)

Fig. 11.15 The radial

distribution of Te, obtained by
processing the chord
distribution of intensity in
continuum (λ = 420 nm)
using the Abel
transformations in sections
z = 2 mm

Diameter of the tungsten cathode dc = 2–1.5 mm;

Cathode length—83 mm, length of the water-cooled parts—50 mm;
The anode plate made of copper with internal water-cooled size 67 × 67 × 13 mm;
Arc length L = 5–20 mm;
Plasma gas consumption QAr = 20–40 l/min;
Consumption of the cooling water of anode and cathode Gw = 40 g/s;
Initial temperature of cooling water Tc0 ¼ 6:2  C, Ta0 ¼ 5:8  C
Power supply with a range of current adjustment I = 40–160 A and voltage
U = 12–28 V.
The current–voltage characteristics of a free arc of different lengths are shown in
Fig. 11.16.
The power transmitted to electrodes (Fig. 11.17, 11.18) is found using the
calorimetric method. At first, using a thermocouple (thermocouples) the input
temperature of the cooling water of cathode and anode are measured at a prede-
termined consumption, then the power transmitted by electrode is determined using
454 11 The Features of the Experimental Methods and Automated …

Fig. 11.16 The current–

voltage characteristics of open
free Arc for different lengths:
1 L = 5 mm, 2 L = 10 mm; 3
L = 15 mm, 4 L = 18 mm

Fig. 11.17 Dependence of

the power absorbed by the
thoriated cathode from the
Arc current of different
lengths: the notes are the same
as in Fig. 11.16

Fig. 11.18 Dependence of

power absorbed by the copper
anode from the Arc current of
different lengths: the notes are
the same as in Fig. 11.16

the equation: P ¼ cp Gw ðT  T 0 Þ, here cp ¼ 4:18 J=ðg KÞ—the specific heat of the

water; Gw—the consumption of the cooling water of anode and cathode, g/s; T0—
the initial temperature of the cooling water; T—the measured temperature of the
water at the exit of plasma torch.
Determination of Temperature and Density of RF Plasma The temperature and
density are determined using the lines of the degree of ionization of ArI RF plasma
at atmospheric pressure of plasma torch with frequency of 27.12 MHz (Fig. 11.3),
11.3 Measurements of Plasma Parameters in the RF and Arc Plasma Torches 455

Table 11.2 Characteristics of some transitions of Argon atoms ArI (NIST ASD [4]) for the
channel No. 3 of spectrometer Avaspec
Wavelength gk Aki, 10−6 s−1 Excitation energy Experimental integral intensity
(nm) Ek (eV) of the spectral line, relative units
675.2834 9.65 14.7425 165.117
696.5431 19.20 13.3278 1566
703.0251 13.40 14.8388 302.715
706.7218 19.00 13.3022 1837
720.698 7.44 15.0220 168.239
727.2936 5.49 13.3278 1103
735.3293 6.70 14.7805 305.968
737.2118 17.00 14.7570 783.99
738.398 42.40 13.3022 5731
772.4207 35.10 13.3365 10,750

intended for spectral analysis. For the determination of electron temperature using
the method of relative intensities the obtained spectrums ArI with wavelengths
550–1000 nm were used, the information about them has been taken from the
database of NIST [4]. For obtaining the distribution of electrons temperature the
spectrums were recorded at some points of plasma in one section of inductor. The
used spectrometer Avaspect of the company Avantes had the built calibration for
wavelength and absolute intensities that is not allowed to determine the dependence
of its sensitivity from the wavelength and to consider it in the resulting spectrum.
The probabilities of some transitions of the argon atoms, as well as statistical weight
and the energies of the upper levels of these transitions are shown in Table 11.2.
The temperature distribution along the radius for the cross section in the center
of inductor (z = 10 mm), of RF plasma torch (27.12 MHz) is obtained using the
Abel transformation and it is given in Fig. 11.19. Calculated temperature field for
comparison is shown in Fig. 11.20.

Fig. 11.19 The distribution of temperature Te of Argon RF plasma torch (27.12 MHz) in the
central section of inductor (z = 10 mm), resulting in the processing the distribution of intensity for
continuum (λ = 700 nm) using Abel conversions
456 11 The Features of the Experimental Methods and Automated …

Fig. 11.20 Calculated temperature field Te, K in Argon RF plasma torch (27.12 MHz)

For the determination of density of electrons the same spectrums were used for
the electron temperature. The calculated density was compared with its theoretical
value obtained by Saha equations taking into account experimentally determined
temperature. To estimate the density of electrons, for example, the spectral line
λ = 549, 567 nm with strongly expressed Stark broadening (about 0.3 nm) was
used to make it possible to neglect other components and to consider the broad-
ening of half-width in the experiment as a Lorentz one without additional
amendments. As a result, electron density, ne ¼ 2:24  1016 cm‒3 was defined by
the Stark broadening of the lines but the calculated value of electron density for
argon plasma in RF plasma torch (27.12 MHz) ne ¼ 2:11  1016 cm‒3 at the
temperature T = 10,200 K in the center of inductor (z = 10 mm, r = 0 mm).
Measuring of the Intensity of Magnetic Field in the RF Plasma Torch Using
the Magnetic Probe Method The scheme of experimental setup for the determi-
nation of intensity of the magnetic field is shown in Fig. 11.21. Intensity is found
using the magnetic probe that is a coil with a few turns of wire of a refractory
material. If the coil located in the area of inductor under the influence of the induced
magnetic field electromotive force appears, it is registered by oscilloscope, con-
nected to the coil using a shielded cable. In the center of the RF plasma temperature
10 × 103 K exists, the long-term presence of the probe in it is impossible. In this
regard, for its short-term position into the desired point of the plasma torch two
coordinate systems of positioning driven by stepper motors, controlled by a com-
puter, was used. Oscilloscope, recorded the signal from the probe was also controlled
by computer, it allowed to carry out a series of measurements on the path of the
probe to the target point of the plasma torch and back for a short time (less than 2 s).
11.3 Measurements of Plasma Parameters in the RF and Arc Plasma Torches 457

Fig. 11.21 Scheme of the experimental setup for determination of the intensity of magnetic field
in RF plasma torch: 1 PC, 2 digital oscilloscope, 3 LPT controller of stepper motors, 4 RF plasma
torch with a magnetic probe, 5 positioning of probe, 6 step motors

Fig. 11.22 The section of the probe: 1 tungsten spiral, 2 tungsten rod, 3 ceramic insulators, 4
lanthanum screen

Image of bottom part of magnetic probe in section and with removed coil are
shown in Figs. 11.22, 11.23. Screen 4 (Fig. 11.22) is needed to protect the stem
from interference and it covers the stem all over the entire length, where the
influence of the magnetic field is strong. The rod 2 inside the screen is passed
through ceramic cylinder 3, isolating it from the screen. Not having time to warm
up during the experiment to the temperature of destruction, the rod can also take the
heat from the probe.
458 11 The Features of the Experimental Methods and Automated …

Fig. 11.23 The image of the probe with the removed coil: 1 tungsten rod, 2 lanthanum screen

Fig. 11.24 Experimental

(points) and calculated (line)
data of the intensity of the
magnetic field on the axis of
RF plasma torch (27.12 MHz)
at the section z = 10 mm

To estimate the accuracy of the obtained data they were compared with the
theoretical calculation results of the magnetic field of the plasma torch. This
comparison for the section z = 10 mm shown in Fig. 11.24. Experimental data have
the same spread as the theoretical values but they are very different. The result of
this experiment was also the possibility of probe diagnostics of plasma with tem-
perature that is several times greater than the melting temperature of the probe.

References

1. Schulz-Gulde E. // Paper of Institute of atom and molecular physics. — Hannover. Germany,

1997.
2. AIAA Paper. 1966. № 66(189). P. 1.
3. Donskoy A.V., Klubnikin V.S. // J. of Engineering Physics. 1972. V. 22. No.6.
4. NIST Atomic Spectra Database. Lines Data (http://physics.nist.gov/cgi-bin/AtData/main_asd).
5. Sobelman I.I. Introduction to the theory of atomic spectra. — M.: Nauka. 1977.
6. Automated System for plasma emission spectra / E.H. Isakayev, V.F. Chinnov et al. //
Instruments and Experimental Techniques. 2001. № 1. PP. 47–53.
7. Frisch S.E. Optical spectra of atoms. — M.-L.: Fizmatgiz 1963.
Appendix A

The values of the statistical weights sum, affinity energy of the electron and the
ionization energy of the atoms and ions of single, double and triple ionization:
Z—sequence number in the periodic table; A—atomic weight; g-1, g0, g1, g2,
g3—the statistical weight sum of electrons, atoms, and ions of single, double and
triple ionization; I-1—the energy of the electron affinity, eV; I0, I1, I2, I3—ionization
energy of the atoms and ions of single, double and triple ionization, eV.

© Springer International Publishing Switzerland 2017 459

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7
Z A Elements g-1 I-1 g0 I0 g1 I1 g2 I2 g3 I3
460

1 1 H 1 0.754 2 13.599
1 2 D 1 – 2 13.602
1 3 T 1 – 2 13.604
2 3 He 2 0.076 1 24.586 2 54.418
2 4 He 2 – 1 24.588 2 54.418
3 6.94 Li 1 0.61 2 5.392 1 75.641 2 122.45
4 9.01 Be 2 0.30 1 9.323 2 18.211 1 153.90 2 217.72
5 10.81 B 1 – 2 8.298 1 25.155 2 37.931 1 259.38
6 12.01 C 4 1.27 1 11.260 2 24.384 1 47.89 2 64.49
7 14.01 N 5 – 4 14.534 1 29.602 2 47.45 1 77.47
8 16 O 4 1.465 5 13.618 4 35.118 1 54.936 2 77.414
9 19 F 1 3.4 4 17.423 5 34.971 4 62.71 1 87.14
10 20.18 Ne 2 – 1 21.565 4 40.963 5 63.46 4 97.12
11 23 Na 1 0.548 2 5.139 1 47.287 4 71.620 5 98.92
12 24.31 Mg 2 – 1 7.646 2 15.035 1 80.144 4 109.27
13 27 Al 1 0.5 2 5.986 1 18.829 2 28.448 1 119.99
14 28.1 Si 4 1.39 1 8.152 2 16.346 1 33.493 2 45.142
15 31 P 5 0.77 4 10.487 1 19.770 2 30.203 1 51.444
16 32.06 S 4 2.077 5 10.360 4 23.338 1 34.83 2 47.305
17 35.45 Cl 1 3.62 4 12.986 5 23.814 4 39.61 1 53.47
18 39.95 Ar 2 – 1 15.760 4 27.630 5 40.911 4 59.81
19 39.1 К 1 0.501 2 4.341 1 31.63 4 45.81 5 60.91
20 40.08 Ca 4 – 1 6.113 2 11.872 1 50.913 4 67.3
21 45 Sc 5 0.19 4 6.562 3 12.800 4 24.757 1 73.49
22 47.9 Ti 4 0.08 5 6.82 4 13.58 5 27.49 4 43.27
Appendix A

(continued)
(continued)
Z A Elements g-1 I-1 g0 I0 g1 I1 g2 I2 g3 I3
23 51 V 7 0.53 4 6.74 1 14.66 4 29.31 5 46.71
Appendix A

24 52 Cr 6 0.67 7 6.766 6 16.50 1 31.0 4 49.2

25 55 Mn 9 – 6 7.434 7 15.640 6 33.67 1 51.2
26 56 Fe 10 0.16 9 7.902 10 16.188 9 30.65 6 54.8
27 59 Co 9 0.66 10 7.86 9 17.084 10 33.5 9 51.3
28 58.71 Ni 2 1.16 9 7.637 6 18.169 9 35.3 10 54.9
29 63.55 Cu 1 1.23 2 7.726 1 20.293 6 36.84 9 57.4
30 65.37 Zn 2 – 1 9.394 2 17.964 1 39.72 6 59.57
31 69.72 Ga 1 0.5 2 5.999 1 20.515 2 30.7 1 64.2
32 72.59 Ge 4 1.2 1 7.900 2 15.935 1 34.2 2 45.7
33 74.92 As 5 0.8 4 9.789 1 18.59 2 28.4 1 50.1
34 78.96 Se 4 2.0 5 9.752 4 21.16 1 30.82 2 42.95
35 79.9 Br 1 3.37 4 11.814 5 21.81 4 35.90 1 47.3
36 83.8 Kr 2 – 1 14.000 4 24.360 5 36.95 4 52.5
37 85.47 Rb 1 0.468 2 4.177 1 27.290 4 39.2 5 52.6
38 87.62 Sr 4 – 1 5.695 2 11.030 1 42.88 4 56.28
39 88.91 Y 5 0.31 4 6.217 1 12.24 2 20.525 1 60.61
40 91.22 Zr 2 0.43 5 6.837 4 13.13 5 23.1 4 34.419
41 92.91 Nb 7 0.89 2 6.88 1 14.32 4 25.0 5 37.7
42 95.94 Mo 6 0.75 7 7.099 6 16.16 1 27.2 4 46.4
43 [99] Tc 11 1.0 6 7.28 7 15.26 6 29.5 1 43
44 101.1 Ru 10 1.5 11 7.366 10 16.76 9 28.5 6 46.5
45 102.9 Rh 1 1.14 10 7.46 9 18.08 10 31.1 9 45.6
46 106.4 Pd 2 0.56 1 8.336 6 19.43 9 32.9 10 48.8
(continued)
461
(continued)
462

Z A Elements g-1 I-1 g0 I0 g1 I1 g2 I2 g3 I3

47 107.9 Ag 1 1.3 2 7.576 1 21.49 6 34.8 9 52
48 112.4 Cd 2 – 1 8.994 2 16.908 1 37.47 6 55
49 114.8 In 1 0.5 2 5.786 1 18.87 2 28.0 1 57.0
50 118.7 Sn 4 1.25 1 7.344 2 14.632 1 30.50 2 40.74
51 121.8 Sb 5 1.05 4 8.609 1 16.53 2 25.32 1 44.16
52 127.6 Te 4 1.9 5 9.010 18.8 30.6 37.9
53 126.9 I 1 3.07 4 10.451 5 19.131 4 33.0 1 41.7
54 131.3 Xe 2 – 1 12.130 4 20.98 5 31.0 4 45
55 132.9 Cs 1 0.471 2 3.894 1 23.15 4 33.4 5 46
56 137.3 Ba 4 – 1 5.212 2 10.004 1 35.8 4 47
57 138.9 La 9 0.5 4 5.577 5 11.1 4 19.18 1 49.9
58 140.1 Ce 10 – 9 5.539 8 10.8 9 20.20 6 36.76
59 140.9 Pr 9 – 10 5.47 9 10.6 10 21.62 9 38.98
60 144.2 Nd 6 – 9 5.525 8 10.7 3 22.1 10 40.4
61 [141] Pm 1 – 6 5.58 5 10.9 6 22.3 9 41.0
62 150.4 Sm 8 – 1 5.644 2 11.1 1 23.4 6 41.4
63 152 Eu 5 – 8 5.670 9 11.24 8 24.9 1 42.7
64 157.3 Gd 16 – 5 6.150 6 12.1 5 20.6 8 44.0
65 159 Tb 17 – 16 5.864 11.5 16 21.9 13 39.4
66 162.5 Dy 16 – 17 5.939 11.7 17 22.8 16 41.4
67 164.9 Ho 13 – 16 6.022 11.8 16 22.8 17 42.5
68 167.3 Er 8 – 13 6.108 11.9 13 22.7 16 42.7
69 168.9 Tm 1 – 8 6.184 12.1 8 23.7 13 42.7
70 173 Yb 4 – 1 6.254 2 12.18 1 25.05 8 43.6
(continued)
Appendix A
(continued)
Z A Elements g-1 I-1 g0 I0 g1 I1 g2 I2 g3 I3
71 175 Lu 5 – 4 5.426 1 13.9 2 20.96 1 45.25
Appendix A

72 178.5 Hf 4 – 5 6.8 4 14.9 5 23.3 4 33.4

73 180.9 Ta 1 0.32 4 7.89 3 16.2 22.3 33.1
74 183.9 W 6 0.816 1 7.98 2 17.7 24.1 35.4
75 186.2 Re 9 0.15 6 7.88 7 16.6 26 37.7
76 190.2 Os 10 1.4 9 8.73 10 17 25 40
77 192.2 Ir 7 1.57 10 9.05 11 17 27 39
78 195.1 Pt 2 2.13 7 8.96 6 18.56 28.5 41.1
79 197 Au 1 2.31 2 9.226 1 20.5 6 34 43
80 200.6 Hg 2 – 1 10.438 2 18.76 34.2 6 46
81 204.4 Tl 1 0.5 2 6.108 1 20.43 2 29.85 50.7
82 207.2 Pb 4 1.2 1 7.417 2 15.033 1 31.94 2 42.23
83 209 Bi 5 1.0 4 7.286 1 16.7 2 25.56 1 45.3
84 [210] Po 4 1.6 5 8.417 19.4 27.3 38
85 [210] At 1 – 4 9.0 20.1 29.3 41
86 [222] Rn 2 – 1 10.75 21.4 29.4 43.8
87 [223] Fr 1 – 2 4.0 22.5 33.5 43
88 [226] Ra 4 – 1 5.279 2 10.15 34.3 46.4
89 [227] Ac 5 – 4 5.2 1 11.75 2 20 49
90 232 Th – 5 6.1 4 11.9 9 18.3 6 28.7
91 [231] Pa – 6.0 9 9
92 238 U – 6.194 10 11.9 20 10 37
93 [237] Np 1 – 6.266 13 9
94 [244] Pu 8 – 1 6.06 2 6
(continued)
463
(continued)
464

Z A Elements g-1 I-1 g0 I0 g1 I1 g2 I2 g3 I3

95 [243] Am – 8 6.0 9 8 13
96 [247] Cm 16 – 6.02 8 8
97 [247] Bk 17 – 16 6.23 17 13
98 [252] Cf 16 – 17 6.30 16
99 [254] Es 13 – 16 6.42 16 17
100 [257] Fm – 13 6.5 16
101 [257] Md – 8 6.6 13
102 [255] No – 1 6.6 1 8
Appendix A
Appendix B

The bracket integrals for the determination of the viscosity of the monocom-
ponent plasma
 
ð0Þ
 o ð0Þ
 o ð2;2Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ 4X1 ðB:1Þ
1
 
ð0Þ
 o ð1Þ
 o ð2;2Þ ð2;3Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ 7X1  2X1 ðB:2Þ
1
 
ð0Þ
2  o ð2Þ
2  o 63 ð2;2Þ 9 ð2;3Þ 1 ð2;4Þ
S5=2 W1 W1 W1 ; S5=2 W1 W1 W1 ¼ X1  X1 þ X1 ðB:3Þ
1 8 2 2
 
ð0Þ
2  o ð3Þ
2  o 231 ð2;2Þ 99 ð2;3Þ 11 ð2;4Þ
S5=2 W1 W1 W1 ; S5=2 W1 W1 W1 ¼ X  X1 þ X
1 32 1 16 8 1
1 ð2;5Þ
 X1 ðB:4Þ
12
 
ð0Þ
2  o ð4Þ
2  o 3003 ð2;2Þ 429 ð2;3Þ 143 ð2;4Þ
S5=2 W1 W1 W1 ; S5=2 W1 W1 W1 ¼ X  X þ X
1 512 1 64 1 64 1
13 ð2;5Þ 1 ð2;6Þ
 X1 þ X
48 96 1
ðB:5Þ
 
ð1Þ
 o ð1Þ
 o 301 ð2;2Þ ð2;3Þ ð2;4Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X  7X1 þ X1 ðB:6Þ
1 12 1
 
ð1Þ
 o ð2Þ
 o 1365 ð2;2Þ 321 ð2;3Þ 25 ð2;4Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X  X þ X
1 32 1 16 1 8 1
1 ð2;5Þ
 X1
4
ðB:7Þ

© Springer International Publishing Switzerland 2017 465

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7
466 Appendix B

 
ð1Þ
 o ð3Þ
 o 6699 ð2;2Þ 297 ð2;3Þ 385 ð2;4Þ 5 ð2;5Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X  X þ X  X1
1 128 1 8 1 48 1 6
1 ð2;6Þ
þ X
24 1
ðB:8Þ
 
ð1Þ
 o ð4Þ
 o 109;109 ð2;2Þ 52;767 ð2;3Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X1  X þ
1 2048 1024 1
11;869 ð2;4Þ 2405 ð2;5Þ 59 ð2;6Þ 1 ð2;7Þ
þ X1  X þ X  X
768 1152 1 384 1 192 1
ðB:9Þ
 
ð2Þ
 o ð2Þ
 o 25;137 ð2;2Þ 1755 ð2;3Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X1  X
1 256 32 1
381 ð2;4Þ 9 ð2;5Þ 1 ð2;6Þ 1 ð4;4Þ
þ X 1  X1 þ X þ X1
32 8 16 1 2
ðB:10Þ
 
ð2Þ
 o ð3Þ
 o 155;925 ð2;2Þ 56;727 ð2;3Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X1  X1
1 1024 512
3795 ð2;4Þ 249 ð2;5Þ 17 ð2;6Þ
þ X  X þ X
128 1 64 1 64 1
1 ð2;7Þ 11 ð4;4Þ 1 ð4;5Þ
 X1 þ X  X1
96 8 1 4
ðB:11Þ
 
ð2Þ
2  o ð4Þ
2  o 3;072;069 ð2;2Þ 719;433 ð2;3Þ 237;237 ð2;4Þ
S5=2 W1 W1 W1 ; S5=2 W1 W1 W1 ¼ X1  X1 þ X1
1 16;384 4096 4096
4901 ð2;5Þ 891 ð2;6Þ 35 ð2;7Þ
 X þ X  X
512 1 1024 1 768 1
1 ð2;8Þ 143 ð4;4Þ 13 ð4;5Þ 1 ð4;6Þ
þ X þ X  X1 þ X1
768 1 64 1 16 16
ðB:12Þ
 
ð3Þ
 o ð3Þ
 o 1;168;167 ð2;2Þ 245;025 ð2;3Þ 236;005 ð2;4Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X1  X1 þ X1
1 4096 1024 3072
1573 ð2;5Þ 859 ð2;6Þ 11 ð2;7Þ
 X þ X  X
128 1 768 1 192 1
1 ð2;8Þ 1463 ð4;4Þ 11 ð4;5Þ 1 ð4;6Þ
þ X þ X  X1 þ X1
576 1 288 1 8 8
ðB:13Þ
Appendix B 467

 
ð3Þ
 o ð4Þ
 o 26;951;925 ð2;2Þ 13;505;349 ð2;3Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X1  X1
1 65;536 32;768
7;661;797 ð2;4Þ 738;569 ð2;5Þ
þ X1  X
49;152 24;576 1
41;005 ð2;6Þ 4175 ð2;7Þ 85 ð2;8Þ
þ X1  X1 þ X
12;288 18;432 9216 1
1 ð2;9Þ 23;881 ð4;4Þ
 X1 þ X
4608 2304 1
4745 ð4;5Þ 37 ð4;6Þ 1 ð4;7Þ
 X þ X  X1
1152 1 64 1 32
ðB:14Þ
 
ð4Þ
 o ð4Þ
 o 711;736;025 ð2;2Þ 50;053;575 ð2;3Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ X  X1 þ
1 1;048;576 1 65;536
64;969;619 ð2;4Þ 3;630;289 ð2;5Þ
þ X1  X1
196;608 49;152
953;681 ð2;6Þ 29;393 ð2;7Þ 4699 ð2;8Þ
þ X  X þ X
98;304 1 36;864 1 110;592 1
13 ð2;9Þ 1 ð2;10Þ 470;041 ð4;4Þ
 X þ X þ X
9216 1 36;864 1 18;432 1
28;223 ð4;5Þ 5341 ð4;6Þ 13 ð4;7Þ
 X þ X  X1
2304 1 2304 1 64
1 ð4;8Þ 1 ð6;6Þ
þ X þ X1
128 1 96
ðB:15Þ
Appendix C

The bracket integrals for the determination of the viscosity of the multicom-
ponent plasma
 
ð0Þ
 o ð0Þ
 o 80 ð1;1Þ ð2;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 ðM1 M2 Þ1 ¼  X þ 8X12 ðC:1Þ
12 3 12
 
ð0Þ
 o ð1Þ
 o
1
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M12 M2
12
280 ð1;1Þ 112 ð1;2Þ ð2;2Þ ð2;3Þ
¼ X þ X þ 28X12  8X12 ðC:2Þ
3 12 3 12
 
ð0Þ
2  o ð2Þ
2  o
3 1
S5=2 W1 W1 W1 ; S5=2 W2 W2 W2 M1 M2
12 ðC:3Þ
ð1;1Þ ð1;2Þ ð1;3Þ ð2;2Þ ð2;3Þ ð2;4Þ
¼ 210X12 þ 168X12  24X12 þ 63X12  36X12 þ 4X12
 
ð0Þ
 o ð3Þ
 o
4 1
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M1 M2
12
ð1;1Þ ð1;2Þ ð1;3Þ 88 ð1;4Þ ðC:4Þ
¼ 385X12 þ 462X12  132X12 þ X
9 12
231 ð2;2Þ ð2;3Þ ð2;4Þ 4 ð2;5Þ
þ X  99X12 þ 22X12  X12
2 12 3
 
ð0Þ
 o ð4Þ
 o
5 1
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M1 M2
12
5005 ð1;1Þ ð1;2Þ ð1;3Þ
¼ X12 þ 1001X12  429X12
8 ðC:5Þ
572 ð1;4Þ 26 ð1;5Þ 3003 ð2;2Þ
þ X  X12 þ X
9 12 9 16 12
429 ð2;3Þ 143 ð2;4Þ 26 ð2;5Þ 1 ð2;6Þ
 X þ X  X12 þ X12
2 12 2 12 3 3

© Springer International Publishing Switzerland 2017 469

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7
470 Appendix C

 
ð1Þ
 o ð1Þ
 o
2 2 1
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M1 M2
12
1540 ð1;1Þ 784 ð1;2Þ 128 ð1;3Þ ðC:6Þ
¼ X12 þ X  X
3 3 12 3 12
602 ð2;2Þ ð2;3Þ ð2;4Þ ð3;3Þ
þ X  56X12 þ 8X12  16X12
3 12
 
ð1Þ
 o ð2Þ
 o
1 ð1;1Þ ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M13 M22 ¼ 1575X12 þ 1218X12
12
ð1;3Þ ð1;4Þ 1365 ð2;2Þ ð2;3Þ ð2;4Þ ð2;5Þ ð3;3Þ ð3;4Þ
 276X12 þ 24X12 þ X12  321X12 þ 50X12  4X12  72X12 þ 16X12
2
ðC:7Þ
 
ð1Þ
 o ð3Þ
 o

7315 ð1;1Þ 1
ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 X12 þ 4004X12
M14 M22 ¼
12 2
ð1;3Þ 1496 ð1;4Þ 80 ð1;5Þ 6699 ð2;2Þ ð2;3Þ 770 ð2;4Þ 80 ð2;5Þ
 1254X12 þ X12  X12 þ X12  1188X12 þ X  X12
9 9 4 3 12 3
4 ð2;6Þ ð3;3Þ ð3;4Þ ð3;5Þ
þ X12  198X12 þ 88X12  8X12
3
ðC:8Þ
 
ð1Þ
 o ð4Þ
 o
5 2 1 115;115 ð1;1Þ 83;083 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M1 M2 ¼  X12 þ X12
12 16 8
8723 ð1;3Þ 7007 ð1;4Þ 611 ð1;5Þ 22 ð1;6Þ 109;109 ð2;2Þ 52;767 ð2;3Þ
 X12 þ X12  X þ X þ X12  X12
2 9 9 12 9 12 32 16
11;869 ð2;4Þ 2405 ð2;5Þ 59 ð2;6Þ 1 ð2;7Þ ð3;3Þ ð3;4Þ ð3;5Þ 8 ð3;6Þ
þ X12  X þ X12  X12  429X12 þ 286X12  52X12 þ X12
12 18 12 6 3 3
ðC:9Þ
 
ð2Þ
 o ð2Þ
 o
24;255 ð1;1Þ 1
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 X12
M13 M23 ¼
12 4
ð1;2Þ ð1;3Þ ð1;4Þ ð1;5Þ 25;137 ð2;2Þ ð2;3Þ ðC:10Þ
þ 5670X12  1746X12 þ 216X12  12X12 þ X12  1755X12
8
ð2;4Þ ð2;5Þ ð2;6Þ ð3;3Þ ð3;4Þ ð3;5Þ ð4;4Þ
þ 381X12  36X12 þ 2X12  492X12 þ 144X12  16X12 þ 16X12
 
ð2Þ
 o ð3Þ
 o
1 135;135 ð1;1Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M14 M23 ¼ X12
12 8
79;695 ð1;2Þ ð1;3Þ ð1;4Þ ð1;5Þ ð1;6Þ 155;925 ð2;2Þ
þ X12  7722X12 þ 1320X12  106X12 þ 4X12 þ X12
4 16
56;727 ð2;3Þ 3795 ð2;4Þ ð2;5Þ ð2;6Þ 2 ð2;7Þ ð3;3Þ
 X12 þ X12  249X12 þ 17X12  X12  1815X12
8 2 3
2398 ð3;4Þ ð3;5Þ ð3;6Þ ð4;4Þ ð4;5Þ
þ X12  124X12 þ 8X12 þ 88X12  16X12
3
ðC:11Þ
Appendix C 471

 
ð2Þ
 o ð4Þ
 o
1
2;477;475 ð1;1Þ 909;909 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M15 M23 X12 þ ¼ X12
12 64 16
433;719 ð1;3Þ 11;869 ð1;4Þ 2613 ð1;5Þ ð1;6Þ ð1;7Þ 3;072;069 ð2;2Þ
 X12 þ X12  X12 þ 37X12  X12 þ X12
16 2 4 128
719;433 ð2;3Þ 237;237 ð2;4Þ 4901 ð2;5Þ 891 ð2;6Þ 35 ð2;7Þ 1 ð2;8Þ
 X12 þ X12  X12 þ X  X12 þ X12
32 32 4 8 12 6 6
9867 ð3;3Þ 9152 ð3;4Þ 1924 ð3;5Þ ð3;6Þ 8 ð3;7Þ ð4;4Þ ð4;5Þ
 X12 þ X12  X12 þ 64X12  X12 þ 286X12  104X12
2 3 3 3
ð4;6Þ
þ 8X12
ðC:12Þ
 
ð3Þ
 o ð3Þ
 o
4 4 1 875;875 ð1;1Þ 297;297 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M1 M2 ¼  X12 þ X12
12 16 4
69;267 ð1;3Þ 22;264 ð1;4Þ 2425 ð1;5Þ ð1;6Þ 32 ð1;7Þ 1;168;167 ð2;2Þ
 X12 þ X12  X12 þ 44X12  X12 þ X12
2 3 3 27 32
245;025 ð2;3Þ 236;005 ð2;4Þ ð2;5Þ 859 ð2;6Þ 22 ð2;7Þ 2 ð2;8Þ
 X12 þ X12  1573X12 þ X  X12 þ X12
8 24 6 12 3 9
33;737 ð3;3Þ 13;310 ð3;4Þ 8290 ð3;5Þ ð3;6Þ ð3;7Þ 5852 ð4;4Þ ð4;5Þ
 X12 þ X12  X12 þ 88X12  4X12 þ X12  176X12
4 3 9 9
ð4;6Þ 32 ð5;5Þ
þ 16X12  X12
3
ðC:13Þ
 
ð3Þ
 o ð4Þ
 o
1
18;293;275 ð1;1Þ 14;559;545 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M15 M24 X12 þ ¼ X12
12 128 64
3;990;129 ð1;3Þ 1;559;701 ð1;4Þ 107;939 ð1;5Þ 4057 ð1;6Þ 713 ð1;7Þ 7 ð1;8Þ
 X12 þ X12  X12 þ X  X þ X12
32 48 24 12 12 54 12 27
26;951;925 ð2;2Þ 13;505;349 ð2;3Þ 7;661;797 ð2;4Þ 738;569 ð2;5Þ 41;005 ð2;6Þ
þ X12  X12 þ X12  X12 þ X12
256 128 192 96 48
4175 ð2;7Þ 85 ð2;8Þ 1 ð2;9Þ 221;507 ð3;3Þ 655;655 ð3;4Þ 41;873 ð3;5Þ
 X þ X  X12  X12 þ X12  X12
72 12 36 12 18 8 36 9
5386 ð3;6Þ 122 ð3;7Þ 4 ð3;8Þ 23;881 ð4;4Þ 9490 ð4;5Þ ð4;6Þ ð4;7Þ
þ X12  X þ X12 þ X12  X12 þ 148X12  8X12
9 3 12 3 9 9
208 ð5;5Þ 32 ð5;6Þ
 X þ X
3 12 3 12
ðC:14Þ
472 Appendix C

 
ð4Þ
 o ð4Þ
 o
1
428;402;975 ð1;1Þ 47;562;515 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W22 W2 W2 M15 M25 X12 þ ¼ X12
12 1024 64
29;713;255 ð1;3Þ 6;779;773 ð1;4Þ 758;173 ð1;5Þ 27;287 ð1;6Þ 13;213 ð1;7Þ 91 ð1;8Þ
 X12 þ X12  X12 þ X12  X12 þ X
64 48 32 12 108 27 12
5 ð1;9Þ 711;736;025 ð2;2Þ 50;053;575 ð2;3Þ 64;969;619 ð2;4Þ 3;630;289 ð2;5Þ
 X þ X12  X12 þ X12  X12
108 12 2048 128 384 96
953;681 ð2;6Þ 29;393 ð2;7Þ 4699 ð2;8Þ 13 ð2;9Þ 1 ð2;10Þ 1;705;275 ð3;3Þ
þ X12  X12 þ X  X12 þ X12  X12
192 72 216 12 18 72 16
2;879;591 ð3;4Þ 289;133 ð3;5Þ 33;982 ð3;6Þ 3061 ð3;7Þ 52 ð3;8Þ 4 ð3;9Þ
þ X12  X12 þ X12  X12 þ X  X12
36 12 9 9 3 12 9
470;041 ð4;4Þ 56;446 ð4;5Þ 10;682 ð4;6Þ ð4;7Þ ð4;8Þ 5096 ð5;5Þ
þ X12  X12 þ X12  104X12 þ 4X12  X12
36 9 9 9
416 ð5;6Þ 32 ð5;7Þ 16 ð6;6Þ
þ X  X12 þ X
3 12 3 3 12
ðC:15Þ
 
ð0Þ
 o ð0Þ
 o 80 ð1;1Þ
 ð2;2Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ ðM1 M2 ÞX12 þ 8 M22 X12 ðC:16Þ
12 3
 
ð0Þ
 o ð1Þ
 o
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1
12
280
 ð1;1Þ 112
 ð1;2Þ
 ð2;2Þ
 ð2;3Þ
¼ M1 M22 X12  M1 M22 X12 þ 28 M23 X12  8 M23 X12
3 3
ðC:17Þ
 
ð0Þ
 o ð2Þ
 o
 ð1;1Þ
 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ 210 M1 M23 X12  168 M1 M23 X12
12

 ð1;3Þ
 ð2;2Þ
 ð2;3Þ
 ð2;4Þ
þ 24 M1 M23 X12 þ 63 M24 X12  36 M24 X12 þ 4 M24 X12
ðC:18Þ
 
ð0Þ
 o ð3Þ
 o
 ð1;1Þ
 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ 385 M1 M24 X12  462 M1 M24 X12
12


88
 ð1;4Þ 231
5  ð2;2Þ
ð1;3Þ
 ð2;3Þ
þ 132 
M1 M24 M1 M24 X12 þ
X12 M2 X12  99 M25 X12
9 2

 ð2;4Þ 4
 ð2;5Þ
þ 22 M25 X12  M25 X12
3
ðC:19Þ
 
ð0Þ
 o ð4Þ
 o 5005
 ð1;1Þ
 ð1;2Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼
M1 M25 X12  1001 M1 M25 X12
12 8

 ð1;3Þ 572
 ð1;4Þ 26
 ð1;5Þ 3003
6  ð2;2Þ
þ 429 M1 M25 X12  M1 M25 X12 þ M1 M25 X12 þ M2 X12
9 9 16
429
6  ð2;3Þ 143
6  ð2;4Þ 26
6  ð2;5Þ 1
6  ð2;6Þ
 M2 X12 þ M2 X12  M2 X12 þ M2 X12
2 2 3 3
ðC:20Þ
Appendix C 473

   
ð1Þ
 o ð1Þ
 o 1540 4 3 7 ð1;1Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ M1 M2 þ M1 M23 X12
3 11 11
12
 
784
 ð1;2Þ 128
 ð1;3Þ 602 22 2 2 21 4 ð2;2Þ ðC:21Þ
 M1 M2 X12 þ
3
M1 M2 X12 þ
3
M1 M2 þ M2 X12
3 3 3 43 43

 ð2;3Þ
 ð2;4Þ
 ð3;3Þ
 56 M24 X12 þ 8 M24 X12 þ 16 M1 M23 X12
   
ð1Þ
 o ð2Þ
 o 8 3 2 7 ð1;1Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ 1575 M1 M2 þ M1 M24 X12
15 15
  12
8 3 2 21 ð1;2Þ
 ð1;3Þ
 ð1;4Þ
 1218 M1 M2 þ M1 M24 X12 þ 276 M1 M24 X12  24 M1 M24 X12
29 29
   
1365 44 2 3 21 5 ð2;2Þ 44 2 3 63 5 ð2;3Þ
þ M M þ M X12  321 M M þ M X12
2 65 1 2 65 2 107 1 2 107 2

 ð2;4Þ
 ð2;5Þ
 ð3;3Þ
 ð3;4Þ
þ 50 M25 X12  4 M25 X12 þ 72 M1 M24 X12  16 M1 M24 X12
ðC:22Þ
   
ð1Þ
 o ð3Þ
 o 7315 12 3 3 7 ð1;1Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ M1 M2 þ M1 M25 X12
2 19 19
  12
 
6 3 3 7 ð1;2Þ 4 3 3 15 ð1;3Þ
 4004 M1 M2 þ M1 M25 X12 þ 1254 M1 M2 þ M1 M25 X12
13 13 19 19
 
1496
 ð1;4Þ 80
 ð1;5Þ 6699 22 2 4 7 ð2;2Þ
 M1 M25 X12 þ M1 M25 X12 þ M1 M2 þ M26 X12
9 9 4 29 29
   
11 2 4 7 6 ð2;3Þ 770 11 2 4 24 6 ð2;4Þ 80
6  ð2;5Þ
 1188 M1 M2 þ M2 X12 þ M1 M2 þ M2 X12  M2 X12
18 18 3 35 35 3
4
 ð2;6Þ
 ð3;3Þ
 ð3;4Þ
 ð3;5Þ
þ M26 X12 þ 198 M1 M25 X12  88 M1 M25 X12 þ 8 M1 M25 X12
3
ðC:23Þ
   
ð1Þ
 o ð4Þ
 o 115;115 16 3 4 7 ð1;1Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ M1 M2 þ M1 M26 X12
16 23 23
 12  
83;083 48 3 4 35 ð1;2Þ 8723 24 3 4 37 ð1;3Þ
 M1 M2 þ M1 M2 X12 þ
6
M1 M2 þ M1 M26 X12
8 83 83 2 61 61
 
7007 8 3 4 41 ð1;4Þ 611
 ð1;5Þ 22
 ð1;6Þ
 M M þ M1 M2 X12 þ
6
M1 M26 X12  M1 M26 X12
9 49 1 2 49 9 9
   
109;109 88 2 5 21 7 ð2;2Þ 52;767 88 2 5 35 7 ð2;3Þ
þ M1 M2 þ M2 X12  M1 M2 þ M2 X12
32 109 109 16 123 123
   
11;869 44 2 5 39 7 ð2;4Þ 2405 44 2 5 141 7 ð2;5Þ
þ M1 M2 þ M2 X12  M1 M2 þ M2 X12
12 83 83 18 185 185
59
7  ð2;6Þ 1
7  ð2;7Þ
 ð3;3Þ
 ð3;4Þ
þ M2 X12  M2 X12 þ 429 M1 M26 X12  286 M1 M26 X12
6 3

 ð3;5Þ 8
 ð3;6Þ
þ 52 M1 M2 X12  M1 M26 X12
6
3
ðC:24Þ
474 Appendix C

   
ð2Þ
 o ð2Þ
 o 24;255 8 5 48 3 3 21 ð1;1Þ
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ M1 M2 þ M1 M2 þ M1 M25 X12
4 77 77 77
  12
 
8 3 3 7 ð1;2Þ 28 3 3 69 ð1;3Þ
 5670 M1 M2 þ M1 M25 X12 þ 1746 M1 M2 þ M1 M25 X12
15 15 97 97
 

 ð1;4Þ
 ð1;5Þ 25;137 24 4 2 88 2 4 21 6 ð2;2Þ
 216 M1 M25 X12 þ 12 M1 M25 X12 þ M 1 M2 þ M1 M2 þ M2 X12
8 133 133 133
   
44 2 4 21 6 ð2;3Þ 52 2 4 75 6 ð2;4Þ
6  ð2;5Þ
 1755 M M þ M X12 þ 381 M M þ M X12  36 M2 X12
65 1 2 65 2 127 1 2 127 2
 

 ð2;6Þ 14 3 3 27 ð3;3Þ
 ð3;4Þ
þ 2 M26 X12 þ 492 M M þ M1 M25 X12  144 M1 M25 X12 þ
41 1 2 41

 ð3;5Þ
 ð4;4Þ
þ 16 M1 M25 X12 þ 16 M12 M24 X12
ðC:25Þ
  
ð2Þ
 o ð3Þ
 o 135;135 8 5 2 24 3 4
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼ M M þ M M
8 39 1 2 39 1 2
  12 
7 ð1;1Þ 79;695 8 72 3 4 35 ð1;2Þ
þ M1 M26 X12  M15 M22 þ M1 M2 þ M1 M26 X12
39 4 115 115 115
   
20 3 4 19 ð1;3Þ 4 3 4 11 ð1;4Þ
þ 7722 M1 M2 þ M1 M26 X12  1320 M1 M2 þ M1 M26 X12
39 39 15 15
 

 ð1;5Þ
 ð1;6Þ 155;925 24 4 3 44 2 5 7 7 ð2;2Þ
þ 106 M1 M26 X12  4 M1 M26 X12 þ M1 M2 þ M1 M2 þ M2 X12
16 75 75 75
  
56;727 24 4 3 132 2 5 35 7 ð2;3Þ 3795 74 2 5
 M M þ M M þ M X12 þ M M
8 191 1 2 191 1 2 191 2 2 115 1 2
  
41 7 ð2;4Þ 30 2 5 53 7 ð2;5Þ
 ð2;6Þ 2
 ð2;7Þ
þ M X12  249 M M þ M X12 þ 17 M27 X12  M27 X12
115 2 83 1 2 83 2 3
   
924 3 4 891 ð3;3Þ 2398 28 81 ð3;4Þ
þ 1815 M1 M2 þ M1 M26 X12  M13 M24 þ M1 M26 X12
1815 1815 3 109 109

 ð3;5Þ
 ð3;6Þ
 ð4;4Þ
 ð4;5Þ
þ 124 M1 M26 X12  8 M1 M26 X12 þ 88 M12 M25 X12  16 M12 M25 X12
ðC:26Þ
 
ð2Þ
 o ð4Þ
 o
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1
 12

2;477;475 16 5 3 32 3 5 7 ð1;1Þ
¼ M1 M2 þ M1 M2 þ M1 M27 X12
64 55 55 55
 
909;909 16 5 3 64 3 5 21 ð1;2Þ
 M1 M2 þ M1 M2 þ M1 M27 X12
16 101 101 101
 
433;719 16 5 3 208 3 5 113 ð1;3Þ
þ M1 M2 þ M1 M2 þ M1 M27 X12
16 337 337 337
   
11;869 40 3 5 43 ð1;4Þ 2613 16 3 5 51 ð1;5Þ
 M1 M2 þ M1 M27 X12 þ M1 M2 þ M1 M27 X12
2 83 83 4 67 67
 

 ð1;6Þ
 ð1;7Þ 3;072;069 144 4 4 176 2 6 21 8 ð2;2Þ
 37 M1 M27 X12 þ M1 M27 X12 þ M1 M2 þ M1 M2 þ M2 X12
128 341 341 341
Appendix C 475

 
719;433 144 4 4 352 2 6 63 8 ð2;3Þ
 M1 M2 þ M1 M2 þ M2 X12
32 559 559 559
 
237;237 48 4 4 384 2 6 121 8 ð2;4Þ
þ M1 M 2 þ M 1 M2 þ M2 X12
32 553 553 553
   
4901 224 2 6 153 8 ð2;5Þ 891 272 2 6 619 8 ð2;6Þ
 M 1 M2 þ M2 X12 þ M 1 M2 þ M2 X12
4 377 377 8 891 891
 
35
8  ð2;7Þ 1
8  ð2;8Þ 9867 14 3 5 9 ð3;3Þ ðC:27Þ
 M2 X12 þ M2 X12 þ M1 M 2 þ M1 M27 X12
6 6 2 23 23
   
9152 7 3 5 9 ð3;4Þ 1924 7 3 5 30 ð3;5Þ
 M1 M2 þ M1 M27 X12 þ M1 M2 þ M1 M27 X12
3 16 16 3 37 37

 ð3;6Þ 8
 ð3;7Þ
 64 M1 M27 X12 þ M1 M27 X12
3

 ð4;4Þ
 ð4;5Þ
 ð4;6Þ
þ 286 M12 M26 X12  104 M12 M26 X12 þ 8 M12 M26 X12
 
ð3Þ
 o ð3Þ
 o
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1
 12

875;875 64 792 5 3 1188 3 5 231 ð1;1Þ
¼ M 1 M2 þ
7
M1 M2 þ M1 M2 þ M1 M27 X12
16 2275 2275 2275 2275
 
297;297 8 5 3 24 3 5 7 ð1;2Þ
 M1 M2 þ M1 M2 þ M1 M27 X12
4 39 39 39
 
69;267 152 5 3 1320 3 5 627 ð1;3Þ
þ M1 M2 þ M1 M2 þ M1 M27 X12
2 2099 2099 2099
   
22;264 12 3 5 11 ð1;4Þ 2425 676 3 5 1749 ð1;5Þ
 M1 M2 þ M1 M27 X12 þ M1 M2 þ M1 M27 X12
3 23 23 3 2425 2425

 ð1;6Þ 32
 ð1;7Þ
 44 M1 M27 X12 þ M1 M27 X12
 27 
1;168;167 272 6 2 2376 4 4 2178 2 6 231 8 ð2;2Þ
þ M1 M 2 þ M1 M2 þ M 1 M2 þ M2 X12
32 5057 5057 5057 5057
 
245;025 24 4 4 44 2 6 7 8 ð2;3Þ
 M1 M2 þ M1 M2 þ M2 X12
8 75 75 75
 
236;005 2744 4 4 14;652 2 6 4059 8 ð2;4Þ
þ M1 M 2 þ M1 M2 þ M2 X12
24 21;455 21;455 21;455
   
90 2 6 53 8 ð2;5Þ 859 298 2 6 561 8 ð2;6Þ 22 ð2;7Þ
 1573 M 1 M2 þ M2 X12 þ M1 M2 þ M2 X12  M28 X12
143 143 6 859 859 3
 
2 ð2;8Þ 33;737 328 5 3 1848 3 5 891 ð3;3Þ
þ M28 X12 þ M1 M 2 þ M1 M2 þ M1 M27 X12
9 4 3067 3067 3067
   
13;310 28 3 5 27 ð3;4Þ 8290 1076 3 5 3069 ð3;5Þ
 M1 M2 þ M1 M27 X12 þ M1 M2 þ M1 M27 X12
3 55 55 9 4145 4145
 

 ð3;6Þ
 ð3;7Þ 5852 34 4 4 99 2 6 ð4;4Þ
 88 M1 M27 X12 þ 4 M1 M27 X12 þ M1 M 2 þ M1 M2 X12
9 133 133

2 6  ð4;5Þ
2 6  ð4;6Þ 32
3 5  ð5;5Þ
 176 M1 M2 X12 þ 16 M1 M2 X12 þ M1 M2 X12
3
ðC:28Þ
476 Appendix C

 
ð3Þ
 o ð4Þ
 o
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1
 12

18;293;275 256 7 2 1584 5 4 1584 3 6 231 ð1;1Þ
¼ M1 M 2 þ M1 M2 þ M 1 M2 þ M1 M28 X12
128 3655 3655 3655 3655
 
14;559;545 256 4752 5 4 7920 3 6 1617 ð1;2Þ
 M17 M22 þ M1 M2 þ M1 M2 þ M1 M28 X12
64 14;545 14;545 14;545 14;545
 
3;990;129 1744 5 4 5808 3 6 1749 ð1;3Þ
þ M1 M2 þ M 1 M2 þ M1 M28 X12
32 9301 9301 9301
 
1;559;701 688 6864 3 6 3355 ð1;4Þ
 M15 M24 þ M1 M 2 þ M1 M28 X12
48 10;907 10;907 10;907
   
107;939 4288 3 6 4015 ð1;5Þ 4057 1120 3 6 2937 ð1;6Þ
þ M1 M2 þ M1 M28 X12  M1 M 2 þ M1 M28 X12
24 8303 8303 12 4057 4057
713
 ð1;7Þ 7
 ð1;8Þ
þ M1 M28 X12  M1 M28 X12
54  27 
26;951;925 1088 6 3 4752 4 5 2904 2 7 231 9 ð2;2Þ
þ M1 M2 þ M1 M2 þ M1 M2 þ M2 X12
256 8975 8975 8975 8975
 
13;505;349 1088 6 3 14;256 4 5 14;520 2 7 1617 9 ð2;3Þ
 M1 M2 þ M1 M2 þ M1 M2 þ M2 X12
128 31;481 31;481 31;481 31;481
 
7;661;797 15;728 4 5 32;208 2 7 5643 9 ð2;4Þ
þ M M þ M M þ M X12
192 53;579 1 2 53;579 1 2 53;579 2
 
738;569 6224 4 5 38;544 2 7 12;045 9 ð2;5Þ
 M1 M 2 þ M1 M2 þ M2 X12
96 56;813 56;813 56;813
 
41;005 24;472 2 7 16;533 9 ð2;6Þ
þ M M þ M X12
48 41;005 1 2 41;005 2
 
4175 1304 2 7 2871 9 ð2;7Þ 85
9  ð2;8Þ 1
9  ð2;9Þ
 M M þ M X12 þ M X  M X
72 4175 1 2 4175 2 36 2 12 18 2 12
 
221;507 328 5 4 924 3 6 297 ð3;3Þ
þ M M þ M M þ M1 M28 X12
8 1549 1 2 1549 1 2 1549
 
655;655 328 5 4 2772 3 6 1485 ð3;4Þ
 M1 M2 þ M1 M2 þ M1 M28 X12
36 4585 4585 4585
   
41;873 1538 3 6 1683 ð3;5Þ 5386 614 3 6 2079 ð3;6Þ
þ M1 M 2 þ M1 M28 X12  M1 M2 þ M1 M28 X12
9 3221 3221 9 2693 2693
 
122
 ð3;7Þ 4
 ð3;8Þ 23;881 68 4 5 99 2 7 ð4;4Þ
þ M1 M28 X12  M1 M28 X12 þ M1 M2 þ M1 M2 X12
3 3 9 167 167
 
9490 68 4 5 297 2 7 ð4;5Þ
 ð4;6Þ
 M M þ M M X12 þ 148 M12 M27 X12
9 365 1 2 365 1 2

 ð4;7Þ 208
3 6  ð5;5Þ 32
3 6  ð5;6Þ
 8 M12 M27 X12 þ M1 M2 X12  M1 M2 X12
3 3
ðC:29Þ
Appendix C 477

 
ð4Þ
 o ð4Þ
 o 428;402;975
S5=2 W12 W1 W1 ; S5=2 W12 W1 W1 ¼
1024
0 12
1
640 13;312 7 3
B 85;595 M1 M2 þ 85;595 M1 M2
9
C
B C ð1;1Þ
B CX12
@ 41;184 5 5 27;456 3 7 3003 A
þ M1 M2 þ M1 M2 þ M1 M2 9
85;595 85;595 85;595
 
47;562;515 256 7 3 1584 5 5 1584 3 7 231 ð1;2Þ
 M1 M2 þ M1 M2 þ M1 M2 þ M1 M29 X12
64 3655 3655 3655 3655
 
29;713;255 3776 68;016 5 5 113;256 3 7 22;737 ð1;3Þ
þ M17 M23 þ M1 M2 þ M1 M2 þ M1 M29 X12
64 207;785 207;785 207;785 207;785
 
6;779;773 688 5 5 2288 3 7 671 ð1;4Þ
 M M þ M M þ M1 M29 X12
48 3647 1 2 3647 1 2 3647
 
758;173 11;280 5 5 111;488 3 7 52;195 ð1;5Þ
þ M M þ M M þ M1 M29 X12
32 174;963 1 2 174;963 1 2 174;963
 
27;287 1120 3 7 979 ð1;6Þ
 M M þ M1 M29 X12
12 2099 1 2 2099
0 1
3944 3 7
M1 M2 þ C
13;213 B B 13;213 C ð1;7Þ 91
 ð1;8Þ 5
 ð1;9Þ
þ B CX12  M1 M29 X12 þ M1 M29 X12
108 @ 9269 9
A 27 108
M1 M2
13;213
0 1
10;624 8 2 169;728 6 4 370;656 4 6
B M1 M2 þ M1 M2 þ M1 M2 þ C
711;736;025 B 711;025 711;025 711;025 C ð2;2Þ
þ B CX12
2048 @ 151;008 2 8 9009 A
þ M1 M2 þ M2 10
711;025 711;025
 
50;053;575 1088 6 4 4752 4 6 2904 2 8 231 10 ð2;3Þ
 M1 M2 þ M1 M2 þ M1 M2 þ M2 X12
128 8975 8975 8975 8975
 
þ
64;969;619 16;064 6 4 204;464 4 6 209;352 2 8
M1 M2 þ M1 M2 þ M1 M2 þ
24;453 10
M2 X12
ð2;4Þ ðC:30Þ
384 454;333 454;333 454;333 454;333
 
3;630;289 6224 4 6 12;848 2 8 2409 10 ð2;5Þ
 M M þ M M þ M X12
96 21;481 1 2 21;481 1 2 21;481 2
 
953;681 102;480 4 6 636;272 2 8 214;929 10 ð2;6Þ
þ M M þ M M þ M X12
192 953;681 1 2 953;681 1 2 953;681 2
   
29;393 1304 2 8 957 10 ð2;7Þ 4699 1384 2 8 3315 10 ð2;8Þ
 M M þ M X12 þ M M þ M X12
72 2261 1 2 2261 2 216 4699 1 2 4699 2
0 1
368 4264 5 5
M17 M23 þ M1 M2 þ C
13
10  ð2;9Þ 1
10  ð2;10Þ 1;705;275 B B 11;925 11;925 C ð3;3Þ
 M X12 þ M X12 þ B CX12
18 2 72 2 16 @ 6006 3 7 1287 A
M1 M2 þ M1 M29
11;925 11;925
 
2;879;591 328 5 5 924 3 7 297 ð3;4Þ
 M M þ M M þ M1 M29 X12
36 1549 1 2 1549 1 2 1549
 
289;133 4856 5 5 39;988 3 7 21;879 ð3;5Þ
þ M1 M2 þ M1 M2 þ M1 M29 X12
12 66;723 66;723 66;723
   
33;982 614 3 7 693 ð3;6Þ 3061 682 3 7 2379 ð3;7Þ
 M1 M2 þ M1 M29 X12 þ M1 M2 þ M1 M29 X12
9 1307 1307 9 3061 3061
52
 ð3;8Þ 4
 ð3;9Þ
 M1 M29 X12 þ M1 M29 X12 þ
3  9 
470;041 232 6 4 1768 4 6 1287 2 8 ð4;4Þ
þ M1 M2 þ M1 M2 þ M1 M2 X12
36 3287 3287 3287
   
56;446 68 4 6 99 2 8 ð4;5Þ 10;682 1012 4 6 4329 2 8 ð4;6Þ
 M1 M2 þ M1 M2 X12 þ M1 M2 þ M1 M2 X12
9 167 167 9 5341 5341
 

2 8  ð4;7Þ
2 8  ð4;8Þ 5096 10 5 5 39 3 7 ð5;5Þ
 104 M1 M2 X12 þ 4 M1 M2 X12 þ M1 M2 þ M1 M2 X12
9 49 49
416
3 7  ð5;6Þ 32
3 7  ð5;7Þ 16
4 6  ð6;6Þ
 M1 M2 X12 þ M1 M2 X12 þ M1 M2 X12
3 3 3
Appendix D

The bracket integrals to determine the diffusion coefficient, thermal diffusion,

electrical and thermal conductivity of the monocomponent plasma
h  i
ðpÞ
 ðqÞ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ 0; p; q ¼ 0 ðD:1Þ
1
h  i
ð1Þ
 ð1Þ
ð2;2Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ 4X1 ðD:2Þ
1
h  i
ð1Þ
 ð2Þ
ð2;2Þ ð2;3Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ 7X1  2X1 ðD:3Þ
1

h  i 63 ð2;2Þ 9 ð2;3Þ 1 ð2;4Þ

ð1Þ
 ð3Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X1  X1 þ X1 ðD:4Þ
1 8 2 2
h  i 231 ð2;2Þ 99 ð2;3Þ 11 ð2;4Þ
ð1Þ
 ð4Þ
1 ð2;5Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ X1  X1 þ X1  X1
1 32 16 8 12
ðD:5Þ
h  i 3003 ð2;2Þ 429 ð2;3Þ 143 ð2;4Þ
ð1Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X  X þ X
1 512 1 64 1 64 1
13 ð2;5Þ 1 ð2;6Þ
 X1 þ X ðD:6Þ
48 96 1
h  i 77 ð2;2Þ
ð2Þ
 ð2Þ
ð2;3Þ ð2;4Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ X1  7X1 þ X1 ðD:7Þ
1 4
h  i 945 ð2;2Þ 261 ð2;3Þ 25 ð2;4Þ 1 ð2;5Þ
ð2Þ
 ð3Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X  X þ X  X1
1 32 1 16 1 8 1 4
ðD:8Þ
h  i 4389 ð2;2Þ 429 ð2;3Þ 55 ð2;4Þ
ð2Þ
 ð4Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X  X þ X
1 128 1 16 1 8 1
5 ð2;5Þ 1 ð2;6Þ
 X1 þ X ðD:9Þ
6 24 1

© Springer International Publishing Switzerland 2017 479

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7
480 Appendix D

h  i
ð2Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1
1
69;069 ð2;2Þ 35;607 ð2;3Þ 3003 ð2;4Þ 715 ð2;5Þ 59 ð2;6Þ 1 ð2;7Þ
¼ X  X þ X  X þ X  X
2048 1 1024 1 256 1 384 1 384 1 192 1
ðD:10Þ
h  i
ð3Þ
 ð3Þ

S3=2 W12 W1 ; S3=2 W12 W1
1
14;553 ð2;2Þ 1215 ð2;3Þ 313 ð2;4Þ 9 ð2;5Þ 1 ð2;6Þ 1 ð4;4Þ
¼ X1  X1 þ X1  X1 þ X þ X1
256 32 32 8 16 1 6
ðD:11Þ
h  i 81;081 ð2;2Þ 34;155 ð2;3Þ
ð3Þ
 ð4Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X  X1
1 1024 1 512
2717 ð2;4Þ 211 ð2;5Þ 17 ð2;6Þ 1 ð2;7Þ 11 ð4;4Þ 1 ð4;5Þ
þ X1  X1 þ X1  X1 þ X1  X1
128 64 64 96 24 12
ðD:12Þ
h  i 1;486;485 ð2;2Þ 389;961 ð2;3Þ 150;293 ð2;4Þ
ð3Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X1  X1 þ X1
1 16;384 4096 4096
3653 ð2;5Þ 779 ð2;6Þ 35 ð2;7Þ 1 ð2;8Þ 143 ð4;4Þ 13 ð4;5Þ 1 ð4;6Þ
 X þ X  X þ X þ X  X1 þ X
512 1 1024 1 768 1 768 1 192 1 48 48 1
ðD:13Þ
h  i 525;525 ð2;2Þ 127;413 ð2;3Þ 48;323 ð2;4Þ
ð4Þ
 ð4Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X1  X1 þ X
1 4096 1024 1024 1
1155 ð2;5Þ 247 ð2;6Þ 11 ð2;7Þ 1 ð2;8Þ 77 ð4;4Þ 11 ð4;5Þ 1 ð4;6Þ
 X þ X  X þ X þ X  X1 þ X
128 1 256 1 192 1 576 1 48 1 24 24 1
ðD:14Þ
h  i 10;975;965 ð2;2Þ 6;239;805 ð2;3Þ 1;378;663 ð2;4Þ
ð4Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X1  X1 þ X1
1 65;536 32;768 16;384
157;651 ð2;5Þ 10;391 ð2;6Þ 1229 ð2;7Þ 85 ð2;8Þ 1 ð2;9Þ 2431 ð4;4Þ
 X1 þ X  X þ X  X þ X
8192 4096 1 6144 1 9216 1 4608 1 768 1
169 ð4;5Þ 37 ð4;6Þ 1 ð4;7Þ
 X þ X  X1
128 1 192 1 96
ðD:15Þ
h  i 257;041;785 ð2;2Þ 20;383;935 ð2;3Þ 10;212;345 ð2;4Þ
ð5Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1 ¼ X  X1 þ X1
1 1;048;576 1 65;536 65;536
673;803 ð2;5Þ 210;763 ð2;6Þ 2561 ð2;7Þ 1409 ð2;8Þ 13 ð2;9Þ 1 ð2;10Þ
 X þ X  X þ X  X þ X
16;384 1 32;768 1 4096 1 36;864 1 9216 1 36;864 1
15;015 ð4;4Þ 2873 ð4;5Þ 571 ð4;6Þ 13 ð4;7Þ 1 ð4;8Þ 1 ð6;6Þ
þ X  X þ X  X þ X þ X
2048 1 768 1 768 1 192 1 384 1 480 1
ðD:16Þ
Appendix E

The bracket integrals to determine the diffusion coefficient, thermal diffusion,

electrical and thermal conductivity of the multicomponent plasma
h  i  1=2 1=2 1
ð0Þ
 ð0Þ
ð1;1Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ 8X12 ðE:1Þ
12

h  i  3=2 1=2 1
ð0Þ
 ð1Þ
ð1;1Þ ð2;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ 20X12 þ 8X12 ðE:2Þ
12

h  i  5=2 1=2 1
ð0Þ
 ð2Þ
ð1;1Þ ð1;2Þ ð1;3Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ 35X12 þ 28X12  4X12
12
ðE:3Þ
h  i  7=2 1=2 1
ð0Þ
 ð3Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M2
12
105 ð1;1Þ ð1;2Þ ð1;3Þ 4 ð1;4Þ
¼ X þ 63X12  18X12 þ X12 ðE:4Þ
2 12 3
h  i  9=2 1=2 1
ð0Þ
 ð4Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M2
12
ðE:5Þ
1155 ð1;1Þ 231 ð1;2Þ 99 ð1;3Þ 22 ð1;4Þ 1 ð1;5Þ
¼ X þ X  X12 þ X  X12
16 12 2 12 2 3 12 3
h  i  11=2 1=2 1
ð0Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M2
12
3003 ð1;1Þ 3003 ð1;2Þ 429 ð1;3Þ 143 ð1;4Þ 13 ð1;5Þ 1 ð1;6Þ
¼ X þ X  X þ X  X12 þ X
32 12 16 12 4 12 6 12 6 15 12
ðE:6Þ
h  i  3=2 3=2 1
ð1Þ
 ð1Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M2
12
ð1;1Þ ð1;2Þ ð1;3Þ ð2;2Þ
¼ 110X12 þ 40X12  8X12 þ 16X12 ðE:7Þ

© Springer International Publishing Switzerland 2017 481

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7
482 Appendix E

h  i  5=2 3=2 1
ð1Þ
 ð2Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M2
12
ðE:8Þ
595 ð1;1Þ ð1;2Þ ð1;3Þ ð1;4Þ ð2;2Þ ð2;3Þ
¼ X þ 189X12  38X12 þ 4X12 þ 56X12  16X12
2 12
h  i  7=2 3=2 1
ð1Þ
 ð3Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M 2
12
2415 ð1;1Þ ð1;2Þ ð1;3Þ 64 ð1;4Þ 4 ð1;5Þ ð2;2Þ ð2;3Þ ð2;4Þ
¼ X12 þ 588X12  162X12 þ X12  X12 þ 126X12  72X12 þ 8X12
4 3 3
ðE:9Þ
h  i  9=2 3=2 1
ð1Þ
 ð4Þ
33;495 ð1;1Þ 22;407 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M 2 ¼ X12 þ X12
12 32 16
2145 ð1;3Þ 539 ð1;4Þ 49 ð1;5Þ 1 ð1;6Þ ð2;2Þ ð2;3Þ ð2;4Þ 8 ð2;5Þ
 X12 þ X  X12 þ X12 þ 231X12  198X12 þ 44X12  X12
4 6 12 6 3 3
ðE:10Þ
h  i  11=2 3=2 1
ð1Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M2
12
105;105 ð1;1Þ 45;045 ð1;2Þ 22;737 ð1;3Þ 1859 ð1;4Þ 143 ð1;5Þ
¼ X12 þ X12  X12 þ X12  X
64 16 16 6 4 12
7 ð1;6Þ 1 ð1;7Þ 3003 ð2;2Þ ð2;3Þ ð2;4Þ 52 ð2;5Þ 2 ð2;6Þ
þ X12  X12 þ X12  429X12 þ 143X12  X12 þ X12
3 15 8 3 3
ðE:11Þ
h  i  5=2 5=2 1
ð2Þ
 ð2Þ
8505 ð1;1Þ ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ 833X12
12 8
ð1;3Þ ð1;4Þ ð1;5Þ ð2;2Þ ð2;3Þ ð2;4Þ ð3;3Þ
 241X12 þ 28X12  2X12 þ 308X12  112X12 þ 16X12  16X12
ðE:12Þ
h  i  7=2 5=2 1
ð2Þ
 ð3Þ

S3=2 W12 W1 ; S3=2 W22 W2 M1 M2
12
42;735 ð1;1Þ 22;071 ð1;2Þ 2001 ð1;3Þ 499 ð1;4Þ 41 ð1;5Þ
¼ X12 þ X12  X12 þ X  X12
16 8 2 3 12 3
2 ð1;6Þ ð2;2Þ ð2;3Þ ð2;4Þ ð2;5Þ ð3;3Þ ð3;4Þ
þ X12 þ 945X12  522X12 þ 100X12  8X12  72X12 þ 16X12
3
ðE:13Þ
h  i  9=2 5=2 1
ð2Þ
 ð4Þ
705;705 ð1;1Þ 234;927 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ X12
12 128 32
104;973 ð1;3Þ 8437 ð1;4Þ 623 ð1;5Þ 29 ð1;6Þ 1 ð1;7Þ 4389 ð2;2Þ
 X12 þ X  X þ X  X12 þ X12
32 12 12 8 12 6 12 6 2
ð2;3Þ ð2;4Þ 160 ð2;5Þ 8 ð2;6Þ ð3;3Þ ð3;4Þ ð3;5Þ
 1716X12 þ 440X12  X þ X12  198X12 þ 88X12  8X12
3 12 3
ðE:14Þ
Appendix E 483

h  i  11=2 5=2 1
ð2Þ
 ð5Þ
2;567;565 ð1;1Þ 2;129;127 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ X12
12 256 128
579;579 ð1;3Þ 75;933 ð1;4Þ 16;237 ð1;5Þ 655 ð1;6Þ 79 ð1;7Þ 1 ð1;8Þ
 X12 þ X12  X12 þ X  X12 þ X
64 32 48 24 12 60 30 12
69;069 ð2;2Þ 35;607 ð2;3Þ 3003 ð2;4Þ 715 ð2;5Þ 59 ð2;6Þ 2 ð2;7Þ
þ X12  X12 þ X12  X þ X  X12
16 8 2 3 12 3 12 3
ð3;3Þ ð3;4Þ ð3;5Þ 8 ð3;6Þ
 429X12 þ 286X12  52X12 þ X12
3
ðE:15Þ
h  i  7=2 7=2 1
ð3Þ
 ð3Þ
255;255 ð1;1Þ 76;923 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ X12
12 32 8
34;119 ð1;3Þ ð1;4Þ 201 ð1;5Þ ð1;6Þ 2 ð1;7Þ 14;553 ð2;2Þ
 X12 þ 895X12  X þ 6X12  X12 þ X12
8 2 12 9 4
ð2;3Þ ð2;4Þ ð2;5Þ ð2;6Þ ð3;3Þ ð3;4Þ ð3;5Þ 32 ð4;4Þ
 2430X12 þ 626X12  72X12 þ 4X12  444X12 þ 144X12  16X12 þ X
3 12
ðE:16Þ
h  i  9=2 7=2 1
ð3Þ
 ð4Þ
4;879;875 ð1;1Þ 3;516;513 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ X12
12 256 128
919;413 ð1;3Þ 118;107 ð1;4Þ 8353 ð1;5Þ 339 ð1;6Þ 71 ð1;7Þ 1 ð1;8Þ
 X12 þ X12  X þ X  X12 þ X
64 32 16 12 8 12 36 18 12
81;081 ð2;2Þ 34;155 ð2;3Þ ð2;4Þ ð2;5Þ ð2;6Þ 4 ð2;7Þ
þ X12  X12 þ 2717X12  422X12 þ 34X12  X12
8 4 3
ð3;3Þ 2222 ð3;4Þ ð3;5Þ ð3;6Þ 176 ð4;4Þ 32 ð4;5Þ
 1551X12 þ X12  124X12 þ 8X12 þ X  X12
3 3 12 3
ðE:17Þ
h  i  11=2 7=2 1
ð3Þ
 ð5Þ
20;165;145 ð1;1Þ 4;327;323 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ X12
12 512 64
329;043 ð1;3Þ 200;057 ð1;4Þ 34;333 ð1;5Þ 881 ð1;6Þ 613 ð1;7Þ 23 ð1;8Þ
 X12 þ X12  X12 þ X  X þ X
8 16 16 4 12 45 12 45 12
1 ð1;9Þ 1;486;485 ð2;2Þ 389;961 ð2;3Þ 150;293 ð2;4Þ 3653 ð2;5Þ 779 ð2;6Þ
 X12 þ X12  X12 þ X12  X12 þ X
90 64 16 16 2 4 12
35 ð2;7Þ 1 ð2;8Þ 8151 ð3;3Þ 8008 ð3;4Þ 1820 ð3;5Þ ð3;6Þ 8 ð3;7Þ
 X12 þ X12  X12 þ X12  X12 þ 64X12  X12
3 3 2 3 3 3
572 ð4;4Þ 208 ð4;5Þ 16 ð4;6Þ
þ X  X þ X
3 12 3 12 3 12
ðE:18Þ
484 Appendix E

h  i  9=2 9=2 1
ð4Þ
 ð4Þ
105;930;825 ð1;1Þ 10;735;725 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M 2 ¼ X12 þ X12
12 2048 128
6;435;429 ð1;3Þ 483;637 ð1;4Þ 165;827 ð1;5Þ 2123 ð1;6Þ 1189 ð1;7Þ 11 ð1;8Þ
 X12 þ X12  X12 þ X12  X þ X 
128 32 64 8 72 12 18 12
1 ð1;9Þ 525;525 ð2;2Þ 127;413 ð2;3Þ 48;323 ð2;4Þ ð2;5Þ ð2;6Þ
 X12 þ X12  X12 þ X12  2310X12 þ 247X12 
72 16 4 4
44 ð2;7Þ 4 ð2;8Þ 26;169 ð3;3Þ 11;374 ð3;4Þ 2566 ð3;5Þ ð3;6Þ ð3;7Þ
 X12 þ X12  X12 þ X12  X12 þ 88X12  4X12 þ
3 9 4 3 3
1232 ð4;4Þ 352 ð4;5Þ 32 ð4;6Þ 16 ð5;5Þ
þ X12  X þ X  X12
3 3 12 3 12 3
ðE:19Þ
h  i  11=2 9=2 1
ð4Þ
 ð5Þ
489;834;345 ð1;1Þ 452;873;421 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ X12
12 4096 2048
38;424;243 ð1;3Þ 6;633;055 ð1;4Þ 1;330;927 ð1;5Þ 82;491 ð1;6Þ 72;643 ð1;7Þ
 X12 þ X12  X12 þ X12  X12
256 128 128 64 720
199 ð1;8Þ 109 ð1;9Þ 1 ð1;10Þ 10;975;965 ð2;2Þ 6;239;805 ð2;3Þ 1;378;663 ð2;4Þ
þ X  X þ X þ X12  X12 þ X12
40 12 720 12 360 12 128 64 32
157;651 ð2;5Þ 10;391 ð2;6Þ 1229 ð2;7Þ 85 ð2;8Þ 1 ð2;9Þ 160;875 ð3;3Þ
 X12 þ X12  X þ X  X12  X12
16 8 12 12 18 12 9 8
173;173 ð3;4Þ 12;103 ð3;5Þ ð3;6Þ 122 ð3;7Þ 4 ð3;8Þ 4862 ð4;4Þ
þ X12  X12 þ 562X12  X þ X12 þ X12
12 3 3 12 3 3
ð4;5Þ 296 ð4;6Þ 16 ð4;7Þ 104 ð5;5Þ 16 ð5;6Þ
 676X12 þ X  X12  X þ X12
3 12 3 3 12 3
ðE:20Þ
h  i  11=2 11=2 1
ð5Þ
 ð5Þ
2;505;429;927 ð1;1Þ 1;273;569;297 ð1;2Þ
S3=2 W12 W1 ; S3=2 W22 W2 M1 M2 ¼ X12 þ X12
12 8192 2048
958;832;589 ð1;3Þ 23;205;897 ð1;4Þ 10;626;655 ð1;5Þ 1;923;623 ð1;6Þ 181;901 ð1;7Þ
 X12 þ X12  X12 þ X12  X12
2048 128 256 320 320
12;701 ð1;8Þ 2047 ð1;9Þ 13 ð1;10Þ 1 ð1;11Þ 257;041;785 ð2;2Þ
þ X12  X þ X  X þ X12
360 1440 12 360 12 1800 12 1024
20;383;935 ð2;3Þ 10;212;345 ð2;4Þ 673;803 ð2;5Þ 210;763 ð2;6Þ 2561 ð2;7Þ
 X12 þ X12  X12 þ X12  X12
64 64 16 32 4
1409 ð2;8Þ 13 ð2;9Þ 1 ð2;10Þ 1;130;415 ð3;3Þ 232;375 ð3;4Þ 230;789 ð3;5Þ
þ X  X12 þ X  X12 þ X12  X12
36 12 9 36 12 16 4 12
ð3;6Þ ð3;7Þ 52 ð3;8Þ 4 ð3;9Þ 15;015 ð4;4Þ 11;492 ð4;5Þ 2284 ð4;6Þ
þ 3302X12  321X12 þ X12  X12 þ X12  X12 þ X12
3 9 2 3 3
208 ð4;7Þ 8 ð4;8Þ 4108 ð5;5Þ 208 ð5;6Þ 16 ð5;7Þ 32 ð6;6Þ
 X þ X12  X þ X  X12 þ X
3 12 3 15 12 3 12 3 15 12
ðE:21Þ
h  i
ð0Þ
 ð0Þ
ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ 8M2 X12 ðE:22Þ
12
h  i
ð0Þ
 ð1Þ
ð1;1Þ ð1;2Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ 20M22 X12  8M22 X12 ðE:23Þ
12
h  i
ð0Þ
 ð2Þ
ð1;1Þ ð1;2Þ ð1;3Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ 35M23 X12  28M23 X12 þ 4M23 X12 ðE:24Þ
12
Appendix E 485

h  i
ð0Þ
 ð3Þ
105 4 ð1;1Þ ð1;2Þ ð1;3Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M X  63M24 X12 þ 18M24 X12
12 2 2 12
4 ð1;4Þ
 M24 X12
3
ðE:25Þ
h  i
ð0Þ
 ð4Þ

S3=2 W12 W1 ; S3=2 W12 W1
ðE:26Þ
12
1155 5 ð1;1Þ 231 5 ð1;2Þ 99 5 ð1;3Þ 22 5 ð1;4Þ 1 5 ð1;5Þ
¼ M X  M X þ M X  M2 X12 þ M2 X12
16 2 12 2 2 12 2 2 12 3 3
h  i
ð0Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1
12
3003 6 ð1;1Þ 3003 6 ð1;2Þ 429 6 ð1;3Þ 143 6 ð1;4Þ 13 6 ð1;5Þ 1 ð1;6Þ
¼ M X  M X þ M X  M X þ M X  M26 X12
32 2 12 16 2 12 4 2 12 6 2 12 6 2 12 15
ðE:27Þ
h  i
ð1Þ
 ð1Þ

S3=2 W12 W1 ; S3=2 W12 W1
 
12
ðE:28Þ
6 2 5 3 ð1;1Þ ð1;2Þ ð1;3Þ ð2;2Þ
¼ 110 M1 M2 þ M2 X12  40M23 X12 þ 8M23 X12 þ 16M1 M22 X12
11 11
h  
ð1Þ
 ð2Þ
 i 595 12 2 2 5 ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M1 M2 þ M24 X12
12 2 17 17
 
4 5 ð1;2Þ ð1;3Þ ð1;4Þ ð2;2Þ ð2;3Þ
 189 M12 M22 þ M24 X12 þ 38M24 X12  4M24 X12 þ 56M1 M23 X12  16M1 M23 X12
9 9
ðE:29Þ
h i 
ð1Þ
 ð3Þ
 2415 18 2 3 5 5 ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M M þ M X12
12 4 23 1 2 23 2
   
9 2 3 5 5 ð1;2Þ 1 2 3 2 5 ð1;3Þ 64 ð1;4Þ
 588 M M þ M X12 þ 162 M1 M2 þ M2 X12  M25 X12
14 1 2 14 2 3 3 3
4 ð1;5Þ ð2;2Þ ð2;3Þ ð2;4Þ
þ M25 X12 þ 126M1 M24 X12  72M1 M24 X12 þ 8M1 M24 X12
3
ðE:30Þ
h  
ð1Þ
 ð4Þ
 i 33;495 24 2 4 5 6 ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M1 M2 þ M2 X12
12 32 29 29
   
22;407 72 2 4 25 6 ð1;2Þ 2145 36 2 4 29 6 ð1;3Þ
 M1 M2 þ M2 X12 þ M1 M2 þ M2 X12
16 97 97 4 65 65
 
539 12 2 4 37 6 ð1;4Þ 49 6 ð1;5Þ 1 6 ð1;6Þ ð2;2Þ
 M1 M2 þ M2 X12 þ M X  M2 X12 þ 231M1 M25 X12
6 49 49 6 2 12 3
ð2;3Þ ð2;4Þ 8 ð2;5Þ
 198M1 M25 X12 þ 44M1 M25 X12  M1 M25 X12
3
ðE:31Þ
486 Appendix E

h i  
ð1Þ
2  ð5Þ
2  105;105 6 2 5 1 7 ð1;1Þ
S3=2 W1 W1 ; S3=2 W1 W1 ¼ M M þ M X12
12 64 7 1 2 7 2
   
45;045 4 2 5 1 7 ð1;2Þ 22;737 36 2 5 17 7 ð1;3Þ
 M M þ M X12 þ M M þ M X12
16 5 1 2 5 2 16 53 1 2 53 2
   
1859 6 2 5 7 7 ð1;4Þ 143 2 2 5 9 7 ð1;5Þ
 M M þ M X12 þ M M þ M X12
6 13 1 2 13 2 4 11 1 2 11 2
7 ð1;6Þ 1 7 ð1;7Þ 3003 ð2;2Þ ð2;3Þ
 M27 X12 þ M X þ M1 M26 X12  429M1 M26 X12
3 15 2 12 8
ð2;4Þ 52 ð2;5Þ 2 ð2;6Þ
þ 143M1 M26 X12  M1 M26 X12 þ M1 M26 X12
3 3
ðE:32Þ
h  
ð2Þ
 ð2Þ
 i 8505 40 4 168 2 3 35 5 ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M1 M2 þ M1 M2 þ M2 X12
12 8 243 243 243
   
12 2 3 5 5 ð1;2Þ 108 2 3 133 5 ð1;3Þ
 833 M1 M2 þ M2 X12 þ 241 M1 M2 þ M2 X12
17 17 241 241
 
ð1;4Þ ð1;5Þ 4 3 2 7 ð2;2Þ ð2;3Þ
 28M25 X12 þ 2M25 X12 þ 308 M1 M2 þ M1 M24 X12  112M1 M24 X12
11 11
ð2;4Þ ð3;3Þ
þ 16M1 M24 X12 þ 16M12 M23 X12
ðE:33Þ
h  
ð2Þ
 ð3Þ
 i 42;735 120 4 2 252 2 4 35 6 ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M1 M2 þ M1 M2 þ M2 X12
12 16 407 407 407
 
22;071 120 4 2 756 2 4 175 6 ð1;2Þ
 M M þ M M þ M X12
8 1051 1 2 1051 1 2 1051 2
   
2001 450 2 4 217 6 ð1;3Þ 499 198 2 4 301 6 ð1;4Þ
þ M1 M2 þ M2 X12  M1 M2 þ M2 X12
2 667 667 3 499 499
 
41 6 ð1;5Þ 2 6 ð1;6Þ 8 3 3 7 ð2;2Þ
þ M2 X12  M2 X12 þ 945 M1 M2 þ M1 M25 X12
3 3 15 15
 
8 3 3 21 ð2;3Þ ð2;4Þ ð2;5Þ
 522 M1 M2 þ M1 M25 X12 þ 100M1 M25 X12  8M1 M25 X12
29 29
ð3;3Þ ð3;4Þ
þ 72M12 M24 X12  16M12 M24 X12
ðE:34Þ
Appendix E 487

h  
ð2Þ
 ð4Þ
 i 705;705 240 4 3 336 2 5 35 7 ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M1 M2 þ M1 M2 þ M2 X12
12 128 611 611 611
 
234;927 80 4 3 224 2 5 35 7 ð1;2Þ
 M M þ M M þ M X12
32 339 1 2 339 1 2 339 2
 
104;973 240 4 3 2304 2 5 637 7 ð1;3Þ
þ M1 M2 þ M1 M2 þ M2 X12
32 3181 3181 3181
   
8437 480 2 5 287 7 ð1;4Þ 623 208 2 5 415 7 ð1;5Þ
 M1 M2 þ M2 X12 þ M1 M2 þ M2 X12
12 767 767 8 623 623
 
29 ð1;6Þ 1 ð1;7Þ 4389 12 3 4 7 ð2;2Þ
 M27 X12 þ M27 X12 þ M1 M2 þ M1 M26 X12
6 6 2 19 19
   
6 3 4 7 ð2;3Þ 1 3 4 4 ð2;4Þ
 1716 M M þ M1 M2 X12 þ 440 M1 M2 þ M1 M2 X12
6 6
13 1 2 13 5 5
160 ð2;5Þ 8 ð2;6Þ ð3;3Þ ð3;4Þ ð3;5Þ
 M1 M26 X12 þ M1 M26 X12 þ 198M12 M25 X12  88M12 M25 X12 þ 8M12 M25 X12
3 3
ðE:35Þ
h  
ð2Þ
 ð5Þ
 i 2;567;565 80 4 4 84 2 6 7 ð1;1Þ
S3=2 W12 W1 ; S3=2 W12 W1 ¼ M1 M2 þ M1 M2 þ M28 X12
12 256 171 171 171
 
2;129;127 240 4 4 420 2 6 49 8 ð1;2Þ
 M M þ M M þ M X12
128 709 1 2 709 1 2 709 2
 
579;579 240 4 4 936 2 6 175 8 ð1;3Þ
þ M1 M2 þ M1 M2 þ M2 X12
64 1351 1351 1351
 
75;933 80 1128 2 6 385 8 ð1;4Þ
 M M þ
4 4
M M þ M X12
32 1593 1 2 1593 1 2 1593 2
   
16;237 708 2 6 541 8 ð1;5Þ 655 36 2 6 95 8 ð1;6Þ
þ M1 M2 þ M2 X12  M1 M2 þ M2 X12
48 1249 1249 24 131 131
 
79 8 ð1;7Þ 1 8 ð1;8Þ 69;069 16 3 5 7 ð2;2Þ
þ M X  M2 X12 þ M M þ M1 M2 X12
7
60 2 12 30 16 23 1 2 23
   
35;607 48 3 5 35 ð2;3Þ 3003 8 3 5 13 ð2;4Þ
 M1 M2 þ M1 M27 X12 þ M1 M2 þ M1 M27 X12
8 83 83 2 21 21
 
715 8 3 5 47 ð2;5Þ 59 ð2;6Þ 2 ð2;7Þ
 M1 M2 þ M1 M27 X12 þ M1 M27 X12  M1 M27 X12
3 55 55 3 3
ð3;3Þ ð3;4Þ ð3;5Þ 8 ð3;6Þ
þ 429M12 M26 X12  286M12 M26 X12 þ 52M12 M26 X12  M12 M26 X12
3
ðE:36Þ
488 Appendix E

h  i
ð3Þ
 ð3Þ

S3=2 W12 W1 ; S3=2 W12 W1
 12

255;255 112 6 1080 4 3 1134 2 5 105 7 ð1;1Þ
¼ M1 M2 þ M1 M2 þ M 1 M2 þ M2 X12
32 2431 2431 2431 2431
 
76;923 120 4 3 252 2 5 35 7 ð1;2Þ
 M1 M2 þ M1 M2 þ M2 X12
8 407 407 407
 
34;119 440 4 3 2700 2 5 651 7 ð1;3Þ
þ M1 M2 þ M1 M2 þ M2 X12
8 3791 3791 3791
   
594 2 5 301 7 ð1;4Þ 201 26 2 5 41 7 ð1;5Þ
 895 M1 M2 þ M2 X12 þ M1 M2 þ M2 X12 ðE:37Þ
895 895 2 67 67
 
ð1;6Þ 2 ð1;7Þ 14;553 8 5 2 48 3 4 21 ð2;2Þ
 6M27 X12 þ M27 X12 þ M M þ M M þ M1 M26 X12
9 4 77 1 2 77 1 2 77
   
8 3 4 7 ð2;3Þ 176 3 4 450 ð2;4Þ
 2430 M1 M2 þ M1 M26 X12 þ 626 M1 M2 þ M1 M26 X12
15 15 626 626
 
ð2;5Þ ð2;6Þ 120 4 3 324 2 5 ð3;3Þ
 72M1 M26 X12 þ 4M1 M26 X12 þ 444 M1 M2 þ M1 M2 X12
444 444
ð3;4Þ ð3;5Þ 32 3 4 ð4;4Þ
 144M12 M25 X12 þ 16M12 M25 X12 þ M M X
3 1 2 12
h  i
ð3Þ
 ð4Þ

S3=2 W12 W1 ; S3=2 W12 W1
 12

4;879;875 448 6 2 2160 4 4 1512 2 6 105 8 ð1;1Þ
¼ M M þ M M þ M M þ M X12
256 4225 1 2 4225 1 2 4225 1 2 4225 2
 
3;516;513 448 6480 4 4 7560 2 6 735 ð1;2Þ
 M16 M22 þ M1 M2 þ M1 M2 þ M28 X12
128 15;223 15;223 15;223 15;223
 
919;413 2480 4 4 5904 2 6 903 8 ð1;3Þ
þ M1 M2 þ M1 M2 þ M2 X12
64 9287 9287 9287
 
118;107 1040 4 4 7632 2 6 2065 8 ð1;4Þ
 M1 M2 þ M1 M2 þ M2 X12
32 10;737 10;737 10;737
   
8353 5304 2 6 3049 8 ð1;5Þ 339 40 2 6 73 8 ð1;6Þ
þ M1 M2 þ M2 X12  M1 M2 þ M2 X12
16 8353 8353 8 113 113
 
71 ð1;7Þ 1 ð1;8Þ 81;081 8 5 3 24 3 5 7 ð2;2Þ
þ M28 X12  M28 X12 þ M1 M2 þ M1 M2 þ M1 M27 X12
36 18 8 39 39 39
 
34;155 8 72 3 5 35 ð2;3Þ
 M15 M23 þ M1 M2 þ M1 M27 X12
4 115 115 115
   
1364 3 5 1353 ð2;4Þ 52 3 5 159 ð2;5Þ
þ 2717 M1 M2 þ M1 M27 X12  422 M1 M2 þ M1 M27 X12
2717 2717 211 211
 
ð2;6Þ 4 ð2;7Þ 660 4 4 891 2 6 ð3;3Þ
þ 34M1 M27 X12  M1 M27 X12 þ 1551 M1 M2 þ M1 M2 X12
3 1551 1551
 
2222 20 4 4 81 2 6 ð3;4Þ ð3;5Þ ð3;6Þ
 M1 M2 þ M1 M2 X12 þ 124M12 M26 X12  8M12 M26 X12
3 101 101
176 3 5 ð4;4Þ 32 3 5 ð4;5Þ
þ M M X  M1 M2 X12
3 1 2 12 3
ðE:38Þ
Appendix E 489

h  i
ð3Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1
 12

20;165;145 224 6 3 720 4 5 378 2 7 21 ð1;1Þ
¼ M1 M2 þ M1 M2 þ M1 M2 þ M29 X12
512 1343 1343 1343 1343
 
4;327;323 112 6 3 720 4 5 567 2 7 42 ð1;2Þ
 M1 M2 þ M1 M2 þ M1 M2 þ M29 X12
64 1441 1441 1441 1441
 
329;043 112 6 3 2400 4 5 3267 2 7 357 9 ð1;3Þ
þ M1 M2 þ M1 M2 þ M1 M2 þ M2 X12
8 6136 6136 6136 6136
 
200;057 320 4 5 918 2 7 161 9 ð1;4Þ
 M M þ M M þ M X12
16 1399 1 2 1399 1 2 1399 2
 
34;333 200 4 5 1857 2 7 584 9 ð1;5Þ
þ M M þ M M þ M X12
16 2641 1 2 2641 1 2 2641 2
   
881 525 2 7 356 9 ð1;6Þ 613 765 2 7 1687 9 ð1;7Þ
 M1 M2 þ M2 X12 þ M1 M2 þ M2 X12
4 881 881 45 2452 2452
 
23 ð1;8Þ 1 9 ð1;9Þ 1;486;485 16 5 4 32 3 6 7 ð2;2Þ
 M29 X12 þ M2 X12 þ M1 M2 þ M1 M2 þ M1 M28 X12
45 90 64 55 55 55
 
389;961 16 5 4 64 3 6 21 ð2;3Þ
 M1 M2 þ M1 M2 þ M1 M28 X12
16 101 101 101
 
150;293 48 640 3 6 363 ð2;4Þ
þ M15 M24 þ M1 M2 þ M1 M28 X12
16 1051 1051 1051
   
3653 128 3 6 153 ð2;5Þ 779 160 3 6 619 ð2;6Þ
 M1 M2 þ M1 M28 X12 þ M1 M2 þ M1 M28 X12
2 281 281 4 779 779
 
35 ð2;7Þ 1 ð2;8Þ 8151 10 4 5 9 ð3;3Þ
 M1 M28 X12 þ M1 M28 X12 þ M1 M2 þ M12 M27 X12
3 3 2 19 19
   
8008 5 4 5 9 ð3;4Þ 1820 1 4 5 6 2 7 ð3;5Þ
 M1 M2 þ M12 M27 X12 þ M1 M2 þ M1 M2 X12
3 14 14 3 7 7
2 7 ð3;6Þ 8 2 7 ð3;7Þ 572 3 6 ð4;4Þ 208 3 6 ð4;5Þ 16 3 6 ð4;6Þ
 64M1 M2 X12 þ M1 M2 X12 þ M M X  M M X þ M M X
3 3 1 2 12 3 1 2 12 3 1 2 12
ðE:39Þ
h  i
ð4Þ
 ð4Þ
105;930;825
S3=2 W12 W1 ; S3=2 W12 W1 ¼
 12 2048 
1152 8 19;712 6 3 47;520 4 5 22;176 2 7 1155 9 ð1;1Þ
 M1 M2 þ M1 M2 þ M1 M2 þ M1 M2 þ M2 X12
91;715 91;715 91;715 91;715 91;715
 
10;735;725 448 6 3 2160 4 5 1512 2 7 105 9 ð1;2Þ
 M M þ M M þ M M þ M X12
128 4225 1 2 4225 1 2 4225 1 2 4225 2
 
6;435;429 5824 81;840 4 5 97;416 2 7 9933 ð1;3Þ
þ M16 M23 þ M1 M2 þ M1 M2 þ M29 X12
128 195;013 195;013 195;013 195;013
 
483;637 1040 4 5 2544 2 7 413 9 ð1;4Þ
 M1 M2 þ M1 M2 þ M2 X12 þ
32 3997 3997 3997
ðE:40Þ
490 Appendix E

 
165;827 15;600 4 5 116;688 2 7 33;539 9 ð1;5Þ
þ M1 M2 þ M1 M2 þ M2 X12
64 165;827 165;827 165;827
   
2123 120 2 7 73 9 ð1;6Þ 1189 408 2 7 781 9 ð1;7Þ
 M M þ M X12 þ M M þ M X12
8 193 1 2 193 2 72 1189 1 2 1189 2
11 ð1;8Þ 1 9 ð1;9Þ
 M29 X12 þ M2 X12
18  72 
525;525 64 792 5 4 1188 3 6 231 ð2;2Þ
þ M17 M22 þ M1 M2 þ M1 M2 þ M1 M28 X12
16 2275 2275 2275 2275
 
127;413 8 5 4 24 3 6 7 ð2;3Þ
 M M þ M M þ M1 M2 X12
8
4 39 1 2 39 1 2 39
 
48;323 312 5 4 2728 3 6 1353 ð2;4Þ
þ M1 M2 þ M1 M2 þ M1 M28 X12
4 4393 4393 4393
   
52 3 6 53 ð2;5Þ 60 3 6 187 ð2;6Þ
 2310 M1 M2 þ M1 M28 X12 þ 247 M1 M2 þ M1 M28 X12
105 105 247 247
 
44 ð2;7Þ 4 ð2;8Þ 26;169 56 6 3 440 4 5 297 2 7 ð3;3Þ
 M1 M28 X12 þ M1 M28 X12 þ M1 M2 þ M1 M2 þ M1 M2 X12
3 9 4 793 793 793
   
11;374 20 4 5 27 2 7 ð3;4Þ 2566 260 4 5 1023 2 7 ð3;5Þ
 M M þ M M X12 þ M M þ M M X12
3 47 1 2 47 1 2 3 1283 1 2 1283 1 2
 
ð3;6Þ ð3;7Þ 1232 3 5 4 11 3 6 ð4;4Þ
 88M12 M27 X12 þ 4M12 M27 X12 þ M M þ M M X12
3 14 1 2 14 1 2
352 3 6 ð4;5Þ 32 3 6 ð4;6Þ 16 4 5 ð5;5Þ
 M M X þ M M X þ M1 M2 X12
3 1 2 12 3 1 2 12 3
ðE:41Þ
Appendix E 491

h  i
ð4Þ
 ð5Þ
489;834;345
S3=2 W12 W1 ; S3=2 W12 W1 ¼
 12 4096 
1152 8 2 9856 6 4 15;840 4 6 5544 2 8 231 ð1;1Þ
 M1 M2 þ M1 M2 þ M1 M2 þ M1 M2 þ M210 X12
32;623 32;623 32;623 32;623 32;623
 
452;873;421 384 9856 6 4 26;400 4 6 12;936 2 8 693
 M18 M22 þ M1 M2 þ M1 M2 þ M1 M2 þ M210
2048 50;269 50;269 50;269 50;269 50;269
 
ð1;2Þ 38;424;243 8288 6 4 44;880 4 6 33;858 2 8 2541 10 ð1;3Þ
 X12 þ M1 M2 þ M1 M2 þ M 1 M2 þ M2 X12
256 89;567 89;567 89;567 89;567
 
6; 633;055 224 6 4 3696 4 6 4818 2 8 539 10 ð1;4Þ
 M M þ M M þ M M þ M X12
128 9277 1 2 9277 1 2 9277 1 2 9277 2
 
1;330;927 24;400 4 6 66;132 2 8 11;847 10 ð1;5Þ
þ M M þ M M þ M X12
128 102;379 1 2 102;379 1 2 102;379 2
 
82;491 6800 4 6 57;420 2 8 18;271 10 ð1;6Þ
 M M þ M M þ M X12
64 82;491 1 2 82;491 1 2 82;491 2
   
72;643 43;350 2 8 29;293 10 ð1;7Þ 199 190 2 8 407 10 ð1;8Þ
þ M1 M2 þ M2 X12  M1 M2 þ M2 X12
720 72;643 72;643 40 597 597
109 10 ð1;9Þ 1 10 ð1;10Þ
þ M X  M X
720 2 12 360 2 12 
10;975;965 256 7 3 1584 5 5 1584 3 7 231 ð2;2Þ
þ M1 M2 þ M1 M2 þ M1 M2 þ M1 M29 X12
128 3655 3655 3655 3655
 
6;239;805 256 4752 5 5 7920 3 7 1617 ð2;3Þ
 M17 M23 þ M1 M2 þ M1 M2 þ M1 M29 X12
64 14;545 14;545 14;545 14;545
 
1;378;663 1776 5 5 5984 3 7 1881 ð2;4Þ
þ M1 M2 þ M1 M2 þ M1 M29 X12
32 9641 9641 9641
 
157;651 720 7392 3 7 4015 ð2;5Þ
 M15 M25 þ M1 M2 þ M1 M29 X12
16 12;127 12;127 12;127
   
10;391 4880 3 7 5511 ð2;6Þ 1229 272 3 7 957 ð2;7Þ
þ M1 M2 þ M1 M29 X12  M 1 M2 þ M1 M29 X12
8 10;391 10;391 12 1229 1229
 
85 ð2;8Þ 1 ð2;9Þ 160;875 56 6 4 220 4 6 99 2 8 ð3;3Þ
þ M1 M29 X12  M1 M29 X12 þ M1 M2 þ M1 M2 þ M1 M2 X12
18 9 8 375 375 375
 
173;173 56 660 4 6 495 2 8 ð3;4Þ
 M M þ
6 4
M M þ M M X12
12 1211 1 2 1211 1 2 1211 1 2
   
12;103 370 4 6 561 2 8 ð3;5Þ 50 4 6 231 2 8 ð3;6Þ
þ M1 M2 þ M1 M2 X12  562 M1 M2 þ M1 M2 X12
3 931 931 281 281
 
122 2 8 ð3;7Þ 4 2 8 ð3;8Þ 4862 6 5 5 11 3 7 ð4;4Þ
þ M M X  M1 M2 X12 þ M M þ M M X12
3 1 2 12 3 3 17 1 2 17 1 2
 
2 5 5 11 3 7 ð4;5Þ 296 3 7 ð4;6Þ 16 3 7 ð4;7Þ
 676 M M þ M M X12 þ M M X  M1 M2 X12
13 1 2 13 1 2 3 1 2 12 3
104 4 6 ð5;5Þ 16 4 6 ð5;6Þ
þ M M X  M1 M2 X12
3 1 2 12 3
ðE:42Þ
492 Appendix E

h  i
ð5Þ
 ð5Þ

S3=2 W12 W1 ; S3=2 W12 W1
 12
2;505;429;927 2816 74;880 8 3 320;320 6 5 343;200 4 7
¼ M 10 M2 þ M M þ M M þ M M
8192 834;309 1 834;309 1 2 834;309 1 2 834;309 1 2
 
90;090 2 9 3003 ð1;1Þ 1;273;569;297 1152 8 3 9856 6 5
þ M M þ M 11 X12  M M þ M M
834;309 1 2 834;309 2 2048 32;623 1 2 32;623 1 2
 
15;840 4 7 5544 2 9 231 ð1;2Þ 958;832;589 17;280
þ M M þ M M þ M 11 X12 þ M8M3
32;623 1 2 32;623 1 2 32;623 2 2048 2;235;041 1 2

430;976 6 5 1;166;880 4 7 586;872 2 9 33;033 ð1;3Þ
þ M1 M2 þ M1 M 2 þ M1 M2 þ M211 X12
2;235;041 2;235;041 2;235;041 2;235;041
 
23;205;897 1120 6 5 6160 4 7 4818 2 9 385 ð1;4Þ
 M1 M2 þ M1 M 2 þ M1 M2 þ M211 X12
128 12;483 12;483 12;483 12;483
 
10;626;655 19;040 6 5 317;200 4 7 429;858 2 9 51;337 11 ð1;5Þ
þ M 1 M2 þ M1 M 2 þ M1 M2 þ M2 X12
256 817;435 817;435 817;435 817;435
 
1;923;623 34;000 4 7 95;700 2 9 18;271 11 ð1;6Þ
 M1 M2 þ M1 M2 þ M2 X12
320 147;971 147;971 147;971
 
181;901 129;200 4 7 1;127;100 2 9 380;809 11 ð1;7Þ
þ M1 M2 þ M 1 M2 þ M2 X12
320 1;637;109 1;637;109 1;637;109
   
12;701 570 2 9 407 11 ð1;8Þ 2047 630 2 9 1417 11 ð1;9Þ
 M1 M2 þ M2 X12 þ M1 M2 þ M2 X12
360 977 977 1440 2047 2047
13 11 ð1;10Þ 1 ð1;11Þ 257;041;785
 M2 X12 þ M211 X12 þ
360
 1800 1024 
640 13;312 7 4 41;184 5 6 27;456 3 8 3003 ð2;2Þ
 M1 M2 þ
9 2
M1 M2 þ M1 M 2 þ M1 M2 þ M1 M210 X12
85;595 85;595 85;595 85;595 85;595
 
20;383;935 256 7 4 1584 5 6 1584 3 8 231 ð2;3Þ
 M1 M 2 þ M1 M2 þ M 1 M2 þ M1 M210 X12
64 3655 3655 3655 3655

 
10;212;345 1280 7 4 23;088 5 6 38;896 3 8 8151 ð2;4Þ
þ M M þ M M þ M M þ M1 M2 X12
10
64 71;415 1 2 71;415 1 2 71;415 1 2 71;415
 
673;803 720 5 6 2464 3 8 803 ð2;5Þ
 M1 M2 þ M1 M2 þ M1 M210 X12
16 3987 3987 3987
 
210;763 12;240 5 6 126;880 3 8 71;643 ð2;6Þ
þ M1 M2 þ M1 M2 þ M1 M210 X12
32 210;763 210;763 210;763
 
2561 272 3 8 319 ð2;7Þ
 M1 M2 þ M1 M210 X12
4 591 591
 
1409 304 3 8 1105 ð2;8Þ 13 ð2;9Þ 1 ð2;10Þ
þ M1 M2 þ M1 M210 X12  M1 M210 X12 þ M1 M210 X12
36 1409 1409 9 36
 
1;130;415 48 728 6 5 1430 4 7 429 2 9 ð3;3Þ
þ M18 M23 þ M1 M2 þ M1 M2 þ M1 M2 X12
16 2635 2635 2635 2635
 
232;375 56 6 5 220 4 7 99 2 9 ð3;4Þ
 M M þ M M þ M M X12
4 375 1 2 375 1 2 375 1 2
Appendix E 493

 
230;789 840 9620 4 7 7293 2 9 ð3;5Þ
þ M16 M25 þ M1 M2 þ M1 M2 X12
12 17;753 17;753 17;753
   
50 4 7 77 2 9 ð3;6Þ 170 4 7 793 2 9 ð3;7Þ
 3302 M1 M2 þ M1 M2 X12 þ 321 M1 M2 þ M1 M2 X12
127 127 963 963
 
52 2 9 ð3;8Þ 4 2 9 ð3;9Þ 15;015 16 7 4 156 5 6 143 3 8 ð4;4Þ
 M1 M2 X12 þ M1 M2 X12 þ M M þ M M þ M M X12
3 9 2 315 1 2 315 1 2 315 1 2
   
11;492 6 5 6 11 3 8 ð4;5Þ 2284 90 5 6 481 3 8 ð4;6Þ
 M M þ M M X12 þ M M þ M M X12
3 17 1 2 17 1 2 3 571 1 2 571 1 2
 
208 3 8 ð4;7Þ 8 3 8 ð4;8Þ 4108 14 6 5 65 4 7 ð5;5Þ
 M M X þ M1 M2 X12 þ M M þ M M X12
3 1 2 12 3 15 79 1 2 79 1 2
208 4 7 ð5;6Þ 16 4 7 ð5;7Þ 32 5 6 ð6;6Þ
 M M X þ M1 M2 X12 þ M M X
3 1 2 12 3 15 1 2 12
ðE:43Þ
Index

A Boltzmann, 1, 2, 4, 5, 7, 8–11, 18, 25, 44, 102,

Abel, 253, 441, 442, 450, 452, 453, 455 112, 123–125, 132, 134, 145, 154, 241,
Absorption, 8, 36, 37, 39, 42–44, 46, 49, 50, 258, 291, 370–372, 375, 377, 396, 403,
112, 120, 417, 419, 440 439–441
Bremsstrahlung, 36, 37, 42 Bohr, 39, 68, 81
Alexandrov, 290, 365 Bose–Einstein, 6
Allmaras, 32 Bracket, 77, 140–144, 148, 151–154, 157, 158,
ALRG, 36 465, 469, 479, 481
Anderson, 202, 290, 365 Brillouin, 66
Angle, 2, 38, 39, 54, 55, 57–61, 63, 64, 66, 73, Buckingham, 57, 65, 70
76, 85–87, 89–91, 124, 130, 144, 165, 194,
224, 298, 337, 422, 438 C
Scattering, 39, 54, 57, 59–61, 76, 85–87, CAM, 7
89–91, 144 Cathode, 45, 201, 285, 288–290, 295, 296,
Anode, 201, 285, 288, 290, 292, 293, 295, 296, 298, 299, 303, 309, 311, 315, 321–328,
298, 299, 321, 325–328, 334, 335, 338, 334–336, 338–342, 346, 347, 349, 354,
340–343, 345–347, 355–357, 366, 355, 360, 367–376, 379–395, 397, 398,
395–398, 431, 432, 435, 453, 454 432, 435, 437, 450, 452–454
Ansys, 35 Cauchy, 418, 420
Approach, 1, 9, 10, 27, 35, 79, 80, 102, 103, CCD, 432, 434–438, 445, 446, 449
105, 126, 132–134, 143, 145, 149, CFD, 26, 35, 51
153–157, 214, 290, 291, 322, 338, 339, CFX, 35
367, 401 Chamber, 172, 176, 205–207, 209, 215, 288,
Arc plasma torches, 50, 120, 158, 175, 201, 360–364, 416–418, 435
263, 265, 266, 285, 286, 288, 289, 291, Chapman, 102, 104, 125, 132–134, 141–143,
292, 295, 296, 298, 307, 315, 323, 327, 145, 149, 152–158
334–336, 338, 361, 364, 367, 368, 373, Chase, 211, 283
376, 380, 385, 387, 388, 390, 395, 396, Collision, 1, 2, 7–11, 18–20, 23, 25, 29, 36, 39,
399, 405, 406, 415, 416, 425, 431, 432, 47, 53–59, 63, 65–70, 72, 73, 76, 78, 81,
435, 439, 449 91, 102–105, 117–119, 123–125, 128–134,
ASM, 34 138, 144, 151, 269, 271, 272, 275, 358,
359, 369–371, 378, 413, 425
B Coulomb, 36, 39, 53, 56, 57, 59–61, 63, 66,
Barnett, 133, 138 104
Bates, 115, 121, 122, 398 Cowling, 134, 142, 149, 150, 152, 157
Bessel, 76, 78 Crank–Nicolson, 409
Biberman, 50, 115, 116, 121, 283, 398 Cressault, 69, 70
Bio, 402, 403, 407 CRM, 7, 10
BKE, 123–126, 134–136, 143, 153, 154 Cronecker, 14

© Springer International Publishing Switzerland 2017 495

S. Nguyen-Kuok, Theory of Low-Temperature Plasma Physics,
Springer Series on Atomic, Optical, and Plasma Physics 95,
DOI 10.1007/978-3-319-43721-7
496 Index

Cross section, 12, 20, 23, 24, 37–40, 43, 44, H

47, 48, 53, 55, 56, 59–69, 72, 73, 77–91, Hinnov, 115, 116, 122, 398
102, 110, 133, 144, 182, 209, 214, 218, Hirchberg, 115, 116, 122, 398
231, 253, 270, 379, 396, 440, 448, 455 Holtsmark, 49
Current–voltage, 289, 311, 312, 315, 379, 453, Hurwitz, 195
454
CVC, 380 I
IEI, 288
D Impact, 2, 3, 9, 18, 20, 23–25, 34, 49, 54,
Dalton, 6, 17, 25, 95, 259 57–59, 63, 67, 76, 114, 116, 117, 124, 182,
de Broglie, 58, 61, 64 258, 259, 270, 290, 376, 377
Debye, 4, 6, 15, 57, 61, 63, 82, 94, 97, 98, 104, Ionization, 4–6, 9, 10, 16, 18–25, 36, 39,
112, 216, 268, 269, 368, 370 42–44, 50, 51, 59, 67, 68, 93, 94, 96–98,
Delta, 14, 139 102, 106, 112, 114–117, 121, 159,
DES, 35 180–183, 258–269, 271–274, 276, 283,
Devoto, 63, 68, 70, 92, 105, 110, 120, 157, 158 284, 288, 291, 339, 342, 344, 346, 359,
DFS-452, 432, 434, 435, 438 369, 370, 373, 375–379, 381–383, 398,
DNS, 27 431, 440, 441, 449, 454, 459
Doppler, 46, 47, 437, 443 Ismailova, 70, 115, 122
Dove, 435
J
E Joule, 12, 14, 193, 308, 370, 391
EARSM, 34
Einstein, 6, 42, 45, 50, 112, 440 K
Electrode, 3, 205, 214, 277, 288–292, 296, Kirchhoff, 6
298, 300, 308, 309, 311, 315, 334–339, Knudsen, 134, 412, 414
341, 342, 367, 368, 392, 396–398, 415, Kramers, 38, 43, 44, 66
416, 453 Kruger, 50, 63, 70, 120, 397
EMF, 205 Kulik, 70, 145, 155, 157
Emission, 36, 108, 449, 450
Enskog, 102, 104, 125, 129, 131–134, 141, L
143, 145, 149, 153, 155–158 Launder–Sharma, 33
Erosion, 288, 298, 368, 388, 395, 398, 415 Laval, 298
Euler, 62, 68 Legendre, 75, 76, 225
Lennard–Jones, 57, 65
F LES, 34, 35
Fermi, 371 Linearization, 193–200, 260, 261
Fick, 411, 413 Lorentz, 47–49, 443, 444, 456
Finkelberg, 70, 290, 364, 398 LTE, 3, 7–9, 102, 291, 300, 439
Fluent, 35
Fourier, 402, 407 M
Fowler–Nordheim, 372 MathCAD, 446, 449
Furnaces, 336, 337 Maxwell, 1, 2, 6, 8, 9, 12, 14–18, 23, 25, 38,
40, 41, 43, 46, 55, 102, 103, 112, 125, 134,
G 175, 188, 211, 214, 218–220, 241, 258,
Gaunt, 38, 43, 45 259, 267, 269, 291, 300, 317, 327, 338,
Gauss, 46–49, 79, 184, 226, 443, 444 373, 374, 377, 396, 412–414, 439, 440
Gauss–Seidel, 184, 226 McCone, 371, 379
GDE, 159–164, 167–171, 177, 180–184, 186, McWhirter, 8, 50
193, 195, 197, 198, 260, 267 Mekker, 70, 290, 364, 398
Granovsky, 290, 364, 365 Menter, 33, 51
Griem, 8, 50 MERT, 78, 80, 81, 87
Gurevich, 115 MGD, 1, 2
Index 497

MHD, 2, 11 259, 269, 273, 288, 373, 376, 377, 387,

Michner, 63 398, 440
MIG, 334 Reynolds, 27, 29–35, 401
Minerr, 446 RF plasma torches, 16, 205–208, 210, 213,
MOAS-2, 432, 434, 438 214, 216, 218, 222, 227, 228, 230, 232,
Moore, 4, 50 233, 238, 240–243, 245–247, 249, 251,
Morse, 57, 65 256, 259, 266, 267, 274, 279, 282, 284,
415, 431, 432, 435
N Richardson, 371–373, 379, 380, 389, 390
Nagaoka, 227, 233 Richardson–Dushman, 371, 373, 379
Navier–Stokes, 27, 29, 31, 32, 133 RNG, 33
Neumark, 201 RSM, 33, 34
Newman, 78 Ruhadze, 290, 365
NIST, 455, 458 Runge–Kutta, 420, 427
Novikov, 290, 365 Rutherford, 60, 82
Nusselt, 401, 404, 413
S
O Saha, 6, 9, 10, 17, 24, 44, 93, 94, 96, 97, 100,
Oscillator, 5, 10 112, 117, 182, 241, 258, 259, 271, 339,
Oscilloscope, 448, 456, 457 377, 439–441, 445, 456
Scattering, 53, 58, 64, 71, 89, 90
P Schmidt, 413
Peclet, 161, 162, 167 Schottky, 372, 396
Periclase, 416, 418–421, 427, 428 Schrödinger, 71–74
Pfender, 120, 158, 365 SGS, 34
Phelps, 66, 70 Sherwood, 413
Photoionization, 37, 43, 44 Shift, 78
Photo-heads, 434 Shur, 35
Photorecombination, 9, 37, 43, 44 SIMPLE, 173
Pitaevskii, 115 SIMPLER, 173
Planck, 5, 6, 71, 447 Sonine, 132, 138, 140–144, 148, 149, 151,
PLTE, 7–9, 112, 258, 395 152, 157, 158
Poisson, 370, 377 Spalart, 32, 35
Potapov, 94, 96, 97, 100, 117, 120, 259 Spectrograph, 207, 432, 434–438, 448
Prandtl, 401, 402 Spectrometer, 45, 432, 438, 445, 455
Spitzer, 104
R Spot, 106, 201, 288, 293, 298, 321, 322, 325,
Radiation, 3, 6, 7, 8, 10, 12, 19, 20, 21, 23, 24, 326, 335, 339, 340, 347, 375, 379, 388,
36, 37, 39–42, 45, 46, 50, 102, 103, 389, 392, 397, 435
112–114, 120, 188, 190, 196, 241, 258, SST, 32, 33
272, 276, 288, 290, 300, 305, 312, 313, Stabilization, 205–207, 210, 286, 288, 293,
335, 340, 347, 359, 360, 365, 370, 375, 296, 298, 308, 354
376, 392, 396, 403, 404, 407, 422, 431, Stark, 48, 49, 437, 439, 443–445, 447, 452,
432, 437–439, 441–445, 447, 448 456
Raiser, 50, 51, 115, 116, 121, 283, 290, 365 Stefan, 7, 375, 396, 403, 410, 412
Ramseur, 57, 64, 80, 88 Stefan–Boltzmann, 7, 375, 396, 403
Ramseur–Collat, 80 Steinbeck, 255
Ramseur–Townsend, 80, 88 Strelets, 35
RANS, 27, 29, 34, 35
Rants–Marshall, 401 T
Recombination, 9, 16, 19, 21–25, 39, 41, 42, TIG, 334
93, 102, 112, 114–116, 121, 159, 182, 206, Thomson, 24, 214, 283
498 Index

Transfer, 2, 9, 11, 13, 15, 16, 18–21, 25, 26, V

30, 33, 46, 50, 51, 53, 55, 67, 68, 80, 93, Van der Waals, 443
102, 103, 106, 109, 120, 121, 125–129, Voigt, 48, 49, 443–447, 452
132, 133, 148, 172, 183, 184, 186, 188,
190, 193, 202, 205, 218, 241, 249, 258, W
260, 269–272, 284, 289, 290, 303, 339, Weight, 5, 25, 115, 408, 441, 455, 459
358, 365, 378, 379, 398, 399, 401–405, Wilcox, 33, 51
407, 410, 422, 427, 430, 440, 450 Wilson, 8, 50
Transformation, 48, 67, 148, 159, 240, 253,
258, 400, 441, 442, 452, 455 Y
Travin, 35 Yukawa, 79, 82
Tungsten, 288, 298, 334, 335, 360, 368, 372,
375, 380, 381, 382, 389, 390, 393, 394, Z
398, 435, 438, 447, 448, 450, 452, 453, Zhukov, 290, 365, 366, 397, 398, 430
457, 458 Zirconium, 298, 368
Turbulence, 26, 27, 30–35, 51, 292, 443