EXTENDED FINITE ELEMENT METHOD FOR FRACTURE ANALYSIS

OF STRUCTURES 1ST EDITION SOHEIL MOHAMMADI -

COMPLETE PDF

[Link]
element-method-for-fracture-analysis-of-structures-1st-
edition-soheil-mohammadi/

★★★★★
4.8 out of 5.0 (85 reviews)

COMPLETE PDF

[Link]
 Extended Finite Element Method for Fracture Analysis of
Structures 1st Edition Soheil Mohammadi

EBOOK

Available Formats

■ PDF eBook Study Guide Ebook

EXCLUSIVE 2025 ACADEMIC EDITION – LIMITED RELEASE

Available Instantly Access Library

Related Study Categories

Finite Element Analysis of Structures through Unified

Formulation 1st Edition Erasmo Carrera

[Link]
structures-through-unified-formulation-1st-edition-erasmo-carrera/

Finite element analysis Method Verification and Validation

Second Edition Barna A. Szabó

[Link]
verification-and-validation-second-edition-barna-a-szabo/

The Finite Element Method for Electromagnetic Modeling 1st

Edition Gerard Meunier

[Link]
electromagnetic-modeling-1st-edition-gerard-meunier/

Introductory Finite Element Method 1st Edition Chandrakant

S. Desai

[Link]
method-1st-edition-chandrakant-s-desai/
Essentials of the finite element method for mechanical and
structural engineers 1st Edition Pavlou

[Link]
method-for-mechanical-and-structural-engineers-1st-edition-pavlou/

The Finite Element Method for Fluid Dynamics Sixth Edition

O. C. Zienkiewicz

[Link]
dynamics-sixth-edition-o-c-zienkiewicz/

The finite element method for elliptic problems 2nd

Edition Philippe G. Ciarlet

[Link]
elliptic-problems-2nd-edition-philippe-g-ciarlet/

An Introduction to the Finite Element Method FEM for

Differential Equations 1st Edition Mohammad Asadzadeh

[Link]
element-method-fem-for-differential-equations-1st-edition-mohammad-
asadzadeh/

Concepts and Applications of Finite Element Analysis 4th

Edition Cook

[Link]
element-analysis-4th-edition-cook/
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

EXTENDED FINITE ELEMENT METHOD

for Fracture Analysis of Structures

Soheil Mohammadi
School of Civil Engineering
University of Tehran
Tehran, Iran

i
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35


C 2008 by Soheil Mohammadi
Published by Blackwell Publishing Ltd

Editorial offices:
Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK
Tel: +44 (0) 1865 776868
Blackwell Publishing Inc., 350 Main Street, Malden, MA 02148-5020, USA
Tel: +1 781 388 8250
Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia
Tel: +61 (0)3 8359 1011

The right of the Author to be identified as the Author of this Work has been asserted in
accordance with the UK Copyright, Designs, and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, except as permitted by the UK Copyright, Designs, and Patents Act
1988, without the prior permission of the publisher.

Designations used by companies to distinguish their products are often claimed as trade-
marks. All brand names and product names used in this book are trade names, service
marks, trademarks or registered trademarks of their respective owners. The Publisher is not
associated with any product or vendor mentioned in this book.

This publication is designed to provide accurate and authoritative information in regard to

the subject matter covered. It is sold on the understanding that the Publisher is not engaged in
rendering professional services. If professional advice or other expert assistance is required,
the services of a competent professional should be sought.

First published 2008 by Blackwell Publishing Ltd

ISBN: 978-1-4051-7060-4

Library of Congress Cataloging-in-Publication Data

Mohammadi, Soheil.
Extended finite element method for fracture analysis of structures / Soheil Mohammadi.
p. cm.
Includes bibliographical references and index.
ISBN-13: 978-1-4051-7060-4 (hardback : alk. paper)
ISBN-10: 1-4051-7060-3 (hardback : alk. paper)
1. Fracture mechanics. 2. Finite element method. I. Title.
TA409.M65 2007
624.1 76–dc22
2007018717

A catalogue record for this title is available from the British Library

Typeset by Soheil Mohammadi

Printed and bound in Singapore
by Utopia Press Pte Ltd

The publisherís policy is to use permanent paper from mills that operate a sustainable
forestry policy, and which has been manufactured from pulp processed using acid-free and
elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper
and cover board used have met acceptable environmental accreditation standards.

For further information on Blackwell Publishing, visit our website:

[Link]/construction

ii
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Contents

Dedication viii
Preface ix
Nomenclature xi

Chapter 1 Introduction
1.1 ANALYSIS OF STRUCTURES 1
1.2 ANALYSIS OF DISCONTINUITIES 2
1.3 FRACTURE MECHANICS 3
1.4 CRACK MODELLING 3
1.4.1 Local and non-local models 4
1.4.2 Smeared crack model 4
1.4.3 Discrete inter-element crack 6
1.4.4 Discrete cracked element 6
1.4.5 Singular elements 6
1.4.6 Enriched elements 7
1.5 ALTERNATIVE TECHNIQUES 7
1.6 A REVIEW OF XFEM APPLICATIONS 7
1.6.1 General aspects of XFEM 7
1.6.2 Localisation and fracture 8
1.6.3 Composites 9
1.6.4 Contact 9
1.6.5 Dynamics 9
1.6.6 Large deformation/shells 10
1.6.7 Multiscale 10
1.6.8 Multiphase/solidification 10
1.7 SCOPE OF THE BOOK 11

Chapter 2 Fracture Mechanics, a Review

2.1 INTRODUCTION 13
2.2 BASICS OF ELASTICITY 14
2.2.1 Stress–strain relations 14
2.2.2 Airy stress function 17
2.2.3 Complex stress functions 18
2.3 BASICS OF LEFM 19
2.3.1 Fracture mechanics 19
2.3.2 Circular hole 20
2.3.3 Elliptical hole 21
2.3.4 Westergaard analysis of a sharp crack 22

iii
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

iv Contents

2.4 STRESS INTENSITY FACTOR, K 24

2.4.1 Definition of the stress intensity factor 24
2.4.2 Examples of stress intensity factors for LEFM 28
2.4.3 Griffith theories of strength and energy 29
2.4.4 Brittle material 30
2.4.5 Quasi-brittle material 32
2.4.6 Crack stability 32
2.4.7 Fixed grip versus fixed load 33
2.4.8 Mixed mode crack propagation 34
2.5 SOLUTION PROCEDURES FOR K AND G 38
2.5.1 Displacement extrapolation/correlation method 38
2.5.2 Mode I energy release rate 39
2.5.3 Mode I stiffness derivative/virtual crack model 40
2.5.4 Two virtual crack extensions for mixed mode cases 41
2.5.5 Single virtual crack extension based on displacement decomposition 42
2.5.6 Quarter point singular elements 43
2.6 ELASTOPLASTIC FRACTURE MECHANICS (EPFM) 47
2.6.1 Plastic zone 47
2.6.2 Crack tip opening displacements (CTOD) 50
2.6.3 J integral 51
2.6.4 Plastic crack tip fields 54
2.6.5 Generalisation of J 55
2.7 NUMERICAL METHODS BASED ON THE J INTEGRAL 56
2.7.1 Nodal solution 56
2.7.2 General finite element solution 57
2.7.3 Equivalent domain integral (EDI) method 59
2.7.4 Interaction integral method 59

Chapter 3 Extended Finite Element Method for Isotropic Problems

3.1 INTRODUCTION 61
3.2 A REVIEW OF XFEM DEVELOPMENT 61
3.3 BASICS OF FEM 65
3.3.1 Isoparametric finite elements, a short review 65
3.3.2 Finite element solutions for fracture mechanics 67
3.4 PARTITION OF UNITY 68
3.5 ENRICHMENT 69
3.5.1 Intrinsic enrichment 70
3.5.2 Extrinsic enrichment 71
3.5.3 Partition of unity finite element method 72
3.5.4 Generalised finite element method 73
3.5.5 Extended finite element method 73
3.5.6 Hp-clouds enrichment 73
3.5.7 Generalisation of the PU enrichment 74
3.5.8 Transition from standard to enriched approximation 74
3.6 ISOTROPIC XFEM 76
3.6.1 Basic XFEM approximation 76
3.6.2 Signed distance function 78
3.6.3 Modelling strong discontinuous fields 78
3.6.4 Modelling weak discontinuous fields 85
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Contents v

3.6.5 Plastic enrichment 86

3.6.6 Selection of nodes for discontinuity enrichment 87
3.6.7 Modelling the crack 90
3.7 DISCRETIZATION AND INTEGRATION 90
3.7.1 Governing equation 90
3.7.2 XFEM discretization 91
3.7.3 Element partitioning and numerical integration 95
3.7.4 Crack intersection 96
3.8 TRACKING MOVING BOUNDARIES 97
3.8.1 Level set method 98
3.8.2 Fast marching method 105
3.8.3 Ordered upwind method 108
3.9 NUMERICAL SIMULATIONS 109
3.9.1 A tensile plate with a central crack 109
3.9.2 Double edge cracks 111
3.9.3 Double internal collinear cracks 112
3.9.4 A central crack in an infinite plate 114
3.9.5 An edge crack in a finite plate 115

Chapter 4 XFEM for Orthotropic Problems

4.1 INTRODUCTION 117
4.2 ANISOTROPIC ELASTICITY 118
4.2.1 Elasticity solution 118
4.2.2 Anisotropic stress functions 120
4.2.3 Orthotropic mixed mode problems 121
4.2.4 Energy release rate and stress intensity factor for anisotropic
materials 122
4.2.5 Anisotropic singular elements 123
4.3 ANALYTICAL SOLUTIONS FOR NEAR CRACK TIP 124
4.3.1 Near crack tip displacement field (class I) 124
4.3.2 Near crack tip displacement field (class II) 128
4.3.3 Unified near crack tip displacement field (both classes) 131
4.4 ANISOTROPIC XFEM 134
4.4.1 Governing equation 134
4.4.2 XFEM discretization 135
4.4.3 SIF calculations 137
4.5 NUMERICAL SIMULATIONS 140
4.5.1 Plate with a crack parallel to material axis of orthotropy 142
4.5.2 Edge crack with several orientations of the axes of orthotropy 144
4.5.3 Single edge notched tensile specimen with crack inclination 147
4.5.4 Central slanted crack 150
4.5.5 An inclined centre crack in a disk subjected to point loads 156
4.5.6 A crack between orthotropic and isotropic materials subjected to
tensile tractions 160

Chapter 5 XFEM for Cohesive Cracks

5.1 INTRODUCTION 163
5.2 COHESIVE CRACKS 164
5.2.1 Cohesive crack models 166
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

vi Contents

5.2.2 Numerical models for cohesive cracks 168

5.2.3 Crack propagation criteria 172
5.2.4 Snap-back behaviour 173
5.2.5 Griffith criterion for cohesive crack 174
5.2.6 Cohesive crack model 175
5.3 XFEM FOR COHESIVE CRACKS 176
5.3.1 Enrichment functions 176
5.3.2 Governing equations 179
5.3.3 XFEM discretization 181
5.4 NUMERICAL SIMULATIONS 183
5.4.1 Mixed mode bending beam 183
5.4.2 Four point bending beam 184
5.4.3 Double cantilever beam 187

Chapter 6 New Frontiers

6.1 INTRODUCTION 189
6.2 INTERFACE CRACKS 190
6.2.1 Elasticity solution for isotropic bimaterial interface 190
6.2.2 Stability of interface cracks 191
6.2.3 XFEM approximation for interface cracks 193
6.3 CONTACT 194
6.3.1 Numerical models for a contact problem 194
6.3.2 XFEM modelling of a contact problem 195
6.4 DYNAMIC FRACTURE 198
6.4.1 Dynamic crack propagation by XFEM 198
6.4.2 Dynamic LEFM 199
6.4.3 Dynamic orthotropic LEFM 200
6.4.4 Basic formulation of dynamic XFEM 202
6.4.5 XFEM discretization 203
6.4.6 Time integration 205
6.4.7 Time finite element method 206
6.4.8 Time extended finite element method 207
6.5 MULTISCALE XFEM 207
6.5.1 Basic formulation 207
6.5.2 The zoom technique 208
6.5.3 Homogenisation based techniques 209
6.5.4 XFEM discretization 210
6.6 MULTIPHASE XFEM 211
6.6.1 Basic formulation 211
6.6.2 XFEM approximation 212
6.6.3 Two-phase fluid flow 214
6.6.4 XFEM approximation 215

Chapter 7 XFEM Flow

7.1 INTRODUCTION 219
7.2 AVAILABLE OPEN-SOURCE XFEM 220
7.3. FINITE ELEMENT ANALYSIS 220
7.3.1 Defining the model 220
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Contents vii

7.3.2 Creating the finite element mesh 220

7.3.3 Linear elastic analysis 221
7.3.4 Large deformation 221
7.3.5 Nonlinear (elastoplastic) analysis 222
7.3.6 Material constitutive matrix 223
7.4 XFEM 223
7.4.1 Front tracking 223
7.4.2 Enrichment detection 225
7.4.3 Enrichment functions 226
7.4.4 Ramp (transition) functions 229
7.4.5 Evaluation of the B matrix 229
7.5 NUMERICAL INTEGRATION 229
7.5.1 Sub-quads 230
7.5.2 Sub-triangles 230
7.6 SOLVER 231
7.6.1 XFEM degrees of freedom 231
7.6.2 Time integration 231
7.6.3 Simultaneous equations solver 232
7.6.4 Crack length control 232
7.7 POST-PROCESSING 233
7.7.1 Stress intensity factor 233
7.7.2 Crack growth 233
7.7.3 Other applications 234
7.8 CONFIGURATION UPDATE 234

References 235
Index 249
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

To Mansoureh

viii
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Preface

‘I am always obliged to a person who has taught me a single word.’

Progressive failure/fracture analysis of structures has been an active research topic for
the past two decades. Historically, it has been addressed either within the framework
of continuum computational plasticity and damage mechanics, or the discontinuous
approach of fracture mechanics. The present form of linear elastic fracture mechanics
(LEFM), with its roots a century old has since been successfully applied to various
classical crack and defect problems. Nevertheless, it remains relatively limited to simple
geometries and loading conditions, unless coupled with a powerful numerical tool such
as the finite element method and meshless approaches.
The finite element method (FEM) has undoubtedly become the most popular and
powerful analytical tool for studying a wide range of engineering and physical prob-
lems. Several general purpose finite element codes are now available and concepts of
FEM are usually offered by all engineering departments in the form of postgraduate
and even undergraduate courses. Singular elements, adaptive finite element procedures,
and combined finite/discrete element methodologies have substantially contributed to
the development and accuracy of fracture analysis of structures. Despite all achieve-
ments, the continuum basis of FEM remained a source of relative disadvantage for
discontinuous fracture mechanics. After a few decades, a major breakthrough seems
to have been made by the fundamental idea of partition of unity and in the form of the
eXtended Finite Element Method (XFEM).
This book has been prepared primarily to introduce the concepts of the newly
developed extended finite element method for fracture analysis of structures. An at-
tempt has also been made to discuss the essential features of XFEM for other related
engineering applications. The book can be divided into four parts. The first part is ded-
icated to the basic concepts and fundamental formulations of fracture mechanics. It
covers discussions on classical problems of LEFM and their extension to elastoplastic
fracture mechanics (EPFM). Issues related to the standard finite element modelling
of fracture mechanics and the basics of popular singular finite elements are reviewed
briefly.
The second part, which constitutes most of the book, is devoted to a detailed dis-
cussion on various aspects of XFEM. It begins by discussing fundamentals of partition
of unity and basics of XFEM formulation in Chapter 3. Effects of various enrichment
functions, such as crack tip, Heaviside and weak discontinuity enrichment functions are
also investigated. Two commonly used level set and fast marching methods for track-
ing moving boundaries are explained before the chapter is concluded by examining a
number of classical problems of fracture mechanics. The next chapter deals with the
orthotropic fracture mechanics as an extension of XFEM for ever growing applications

ix
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

x Preface

of composite materials. A different set of enrichment functions for orthotropic media

is presented, followed by a number of simulations of benchmark orthotropic problems.
Chapter 5, devoted to simulation of cohesive cracks by XFEM, provides theoretical
bases for cohesive crack models in fracture mechanics, classical FEM and XFEM.
The snap-back response and the concept of critical crack path are studied by solving a
number of classical cohesive crack problems.
The third part of the book (Chapter 6) provides basic information on new frontiers
of application of XFEM. It begins with discussions on interface cracking, which include
classical solutions from fracture mechanics and XFEM approximation. Application of
XFEM for solving contact problems is explained and numerical issues are addressed.
The important subject of dynamic fracture is then discussed by introducing classical
formulations of fracture mechanics and the recently developed idea of time–space
discretization by XFEM. New extensions of XFEM for very complex applications of
multiscale and multiphase problems are explained briefly.
The final chapter explains a number of simple instructions, step-by-step proce-
dures and algorithms for implementing an efficient XFEM. These simple guidelines, in
combination with freely available XFEM source codes, can be used to further advance
the existing XFEM capabilities.
This book is the result of an infinite number of brilliant research works in the
field of computational mechanics for many years all over the world. I have tried to
appropriately acknowledge the achievements of corresponding authors within the text,
relevant figures, tables and formulae. I am much indebted to their outstanding research
works and any unintentional shortcoming in sufficiently acknowledging them is sin-
cerely regretted. Perhaps such a title should have become available earlier by one of
the pioneers of the method, i.e. Professor T. Belytschko, a shining star in the universe
of computational mechanics, Dr J. Dolbow, Dr N. Moës, Dr N. Sukumar and possibly
others who introduced, contributed and developed most of the techniques.
I would like to extend my acknowledgement to Blackwell Publishing Limited,
for facilitating the publication of the first book on XFEM; in particular N. Warnock-
Smith, J. Burden, L. Alexander, A. Cohen and A. Hallam for helping me throughout
the work. Also, I would like to express my sincere gratitude to my long-time friend,
Professor A.R. Khoei, with whom I have had many discussions on various subjects of
computational mechanics, including XFEM. Also my special thanks go to my students:
Mr A. Asadpoure, to whom I owe most of Chapter 4, Mr S.H. Ebrahimi for solving
isotropic examples in Chapter 3 and Mr A. Forghani for providing some of the results
in Chapter 5.
This book has been completed on the eve of the new Persian year; a ‘temporal
interface’ between winter and spring, and an indication of the beginning of a blooming
season for XFEM, I hope.
Finally, I would like to express my gratitude to my family for their love, under-
standing and never-ending support. I have spent many hours on writing this book; hours
that could have been devoted to my wife and little Sogol: the spring flowers that inspire
the life.

Soheil Mohammadi
Tehran, Iran
Spring 2007
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Nomenclature

α Curvilinear coordinate
αc Load factor for cohesion
α f , αs Thermal diffusivity of fluid and solid phases
β Curvilinear coordinate
γs Surface energy density
p
γse , γs Elastic and plastic surface energies
γx y Engineering shear strain
δ Plastic crack tip zone
δ Variation of a function
δ(ξ ) Dirac delta function
δi j Kronecker delta function
ε Strain tensor
ε f , εc Strain field at fine and coarse scales
εij Strain components
ε̄ij Dimensionless angular geometric function
εijaux Auxiliary strain components
εv Kinetic mobility coefficient
εyld Yield strain
η Local curvilinear (mapping) coordinate system
θ Crack propagation angle with respect to initial crack
θ Angular polar coordinate
κ, κ  Material parameters
λ Lame modulus
λ Eigenvalue of the characteristic equation
μ Shear modulus
ν, νij Isotropic and orthotropic Poisson’s ratios
ξ Local curvilinear (mapping) coordinate system
ξ(x) Distance function
ρ Radius of curvature
ρ Density
ρ f , ρc Density of fine and coarse scales
ρint Curvature of the propagating interface
σ Stress tensor
σ f , σc Stress field at fine and coarse scales
σg Stress tensor at a Gauss point
tip
σt Normal tensile stress at crack tip
σ0 Applied normal traction

xi
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

xii Nomenclature

σcr Critical stress for cracking

σij Stress components
σ̄ij Dimensionless angular geometric function
σijaux Auxiliary stress components
σn Stress component normal to an interface
σn Stress component at time step n
σyld Yield stress
τ Deviatoric stress
τ0 Applied tangential traction
τc Cohesive shear traction
τn Time functions
τn Deviatoric stress tensor at time step n
φ(x) Level set function
φ(z) Complex stress function
φs (z) Stress function for shear problem
ϕ Angle of orthotropic axes
ϕ Phase angle for interface fracture
χ (x) Enrichment function for weak discontinuities
χ (z) Stress function
ψ (x) Enrichment function
ψ (z) Stress function
ω Oscillation index
 Boundary
c Crack boundary
t Traction (natural) boundary
u Displacement (essential) boundary
 Finite variation of a function
 Coefficient matrix
Ξ Homogenisation/average operator
 Potential energy
 j (x) Moving least squares shape functions
(x) Stress function
 Domain
 f , c Fine and course scale domains
 f , s Fluid and solid domains
pu Domain associated with the partition of unity

a Crack length/half length

a Semi-major axis of ellipse
ab, af Backward and forward indexes in fast marching method
ah Heaviside enrichment degrees of freedom
ai Enrichment degrees of freedom
ak Enrichment degrees of freedom
A∗ Area associated with the domain J integral
b Width of a plate
b Semi-minor axis of ellipse
bi Crack tip enrichment degrees of freedom
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Nomenclature xiii

B Matrix of derivatives of shape functions

Bh Matrix of derivatives of final shape functions
Bc B matrix for coarse scale
Bf B matrix for fine scale
Bri Strain–displacement matrix (derivatives of shape functions)
Biu Strain–displacement matrix (derivatives of shape functions)
Bia Matrix of derivatives of enrichment (Heaviside) of shape functions
Bib Matrix of derivatives of enrichment (crack tip) of shape functions
c Constant parameter
c Size of crack tip contour for J integral
cij Material constants
cR Rayleigh speed
c f , cs Specific heat for fluid and solid phases
C Material constitutive matrix
d Distance
d/dt Time derivative
D Material modulus matrix
Dc , D f Material modulus in coarse and fine scales
Dloc Localisation modulus
D/Dt Material time derivative
Dxb , Dxf Backward and forward finite difference approximations
E, E i Isotropic and orthotropic Young’s modulus
E Material parameter
ft Uniaxial tensile strength
f (r ) Radial function
f Nodal force vector
f ri Nodal force components (classic and enriched)

fb Body force vector

ft External traction vector
fc Cohesive crack traction vector
f coh Cohesive nodal force vector
f ext External force vector
f uint Internal nodal force vector due to external loading
f aint Internal nodal force vector due to cohesive force
Fli (x) Crack tip enrichment functions
g Applied gravitational body force
g(θ) Angular function for a crack tip kink problem
g j (θ ) Orthotropic crack tip enrichment functions
G Shear modulus
G Fracture energy release rate
G1, G2 Mode I and II fracture energy release rates
dyn
GI Dynamic mode I fracture energy release
H (ξ) Heaviside function
Hl Latent heat
i Complex number, i 2 = −1
J Jacobian matrix
J J integral
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

xiv Nomenclature

J act Actual J integral

J aux Auxiliary J integral
Jk Mode k contour integral J
k0 Dimensionless constant for the power hardening law
k0 , k1 , k2 , k3 , k4 Constant coefficients
ki Conductivity coefficient for phase i
ks , k f Thermal conductivity for solid and fluid phases
kn , kt Normal/tangential interface properties
K Stiffness matrix
Khom Homogenised stiffness matrix
Krijs Stiffness matrix components
K Stress intensity factor
KC Critical stress intensity factor
K eq Equivalent mixed mode stress intensity factor
K I , K II , K III Mode I , II and III stress intensity factors
K̄ I , K̄ II Normalized mode I and mode II stress intensity factors
K Iaux , K IIaux Auxiliary mode I and mode II stress intensity factors
K Ic , K IIc Critical mode I and mode II stress intensity factors
K Icohesion Cohesive mode I stress intensity factor
K Icrack Crack mode I stress intensity factor
dyn
KI Dynamic mode I stress intensity factor
le Characteristic length
lc Characteristic length for crack propagaion
m Number of enrichment functions
mti Number of nodes to be enriched by crack tip enrichment functions
mf Number of crack tip enrichment functions
Mj Mach number
M Interaction integral
M0 Total mass
Mi Lumped mass component
Mij Mass matrix component
n Power number for the plastic model (Section 2.6.4)
ng Number of Gauss points
n̄ Number of nodes within each moving least squares support domain
np Number of independent domains of partition of unity
nn Number of nodes in a finite element
n Normal vector
nint Normal vector to an internal interface
Nj Matrix of shape functions
Nj Shape function
N̄ j New set of generalised finite element method shape functions
p(x) Basis function
p Hydrostatic pressure
p̄ Predefined hydrostatic pressure
Pi Loading condition i
q Arbitrary smoothing function
q Heat flux
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Nomenclature xv

qi Nodal values of the arbitrary smoothing function

Q Input heat to system
Qij Matrix of homogenous anisotropic solids
r Radial distance/coordinate
rg Radial distance of a Gauss point from crack tip
rp Crack tip plastic zone
R Ramp function
s f , ss Heat source for fluid and solid phases
Sa Set of accepted nodes
Sc Set of candidate nodes
Sd Set of distant nodes
t Time
t Traction
tint Surface traction along internal boundary
T Temperature
T f , Ts Temperature of fluid and solid phases
Ti (t) Time shape functions
Tm Melting/fusion temperature
u Displacement vector
û Local symmetric displacement vector
u̇ Velocity vector
ū Prescribed displacement
u̇¯ Prescribed velocity
ü Acceleration vector
uiaux Auxiliary displacement field
ucoh Displacement field obtained from crack surface tractions
uenr Enriched displacement field
uFE Classical finite element displacement field
uh (x) Approximated displacement field
ū j Nodal displacement vector
ũ j Transformed displacement
up Periodic displacement
ux , u y xand y displacement component
U 1, U 2 Symmetric and antisymmetric crack tip displacements
Uk Kinetic energy
Us Strain energy
p
Use , Us Elastic and plastic strain energies
U Surface energy
v Velocity vector
v̄ Prescribed velocity
vn Normal interface speed
v1 , v2 Longitudinal and shear wave velocities
V(t) Vector of approximated velocity degrees of freedom
w Crack opening
wc Critical crack opening
W External work
W aux Auxiliary work
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

xvi Nomenclature

W coh Virtual work of cohesive force

W ext Virtual work of external loading
Wg Gauss weight factor
Wgr Radial weight function at a Gauss point g
W int Internal virtual work
WM Interaction work
Ws Strain energy
Wt (t) Time weight function
x Position vector
xc , x f Position vector for coarse and fine scales
x Position of projection point on an interface
x1 , x2 Two-dimensional coordinate system
x Local crack tip coordinate axes
y Local crack tip coordinate axes
z = x + iy Complex variable
z̄ = x − i y Conjugate complex variable
zi Complex parameters

f  , f  First and second derivative of a function

f¯, f¯ First and second integrals of a function

∇ = ∂ ∂x Nabla operator
 Jump operator across an interface

BEM Boundary Element Method

CBS Characteristic Based Split
COD Crack Opening Displacement
CTOD Crack Tip Opening Displacement
DCT Displacement Correlation Technique
DOF Degree Of Freedom
EDI Equivalent Domain Integral
EFG Element-Free Galerkin
ELM Equilibrium On Line
EPFM Elastic Plastic Fracture Mechanics
FDM Finite Difference Method
FE Finite Element
FEM Finite Element Method
FMM Fast Marching Method
FPM Finite Point Method
FPZ Fracture Process Zone
GFEM Generalised Finite Element Method
HRR Hutchinson–Rice–Rosengren
LEFM Linear Elastic Fracture Mechanics
LSM Level Set Method
MCC Modified Crack Closure
MEPU Multiscale Enrichment Partition of Unity
MLPG Meshless Local Petrov–Galerkin
MLS Moving Least Squares
 P1: PBUPrinter: Yet to Come
BLUK134-FM Mohammadi August 21, 2007 18:35

Nomenclature xvii

NURBS Non-Uniform Rational B-Spline

OUM Ordered Upwind Method
PU Partition of Unity
PUFEM Partition of Unity Finite Element Method
RKPM Reproducing Kernel Particle Method
SAR Statically Admissible stress Recovery
SIF Stress Intensity Factor
SPH Smoothed Particle Hydrodynamics
TFEM Time Finite Element Method
TXFEM Time eXtended Finite Element Method
WLS Weighted Least Squares
XFEM, X-FEM eXtended Finite Element Method
 Chapter 1

Introduction

1.1 ANALYSIS OF STRUCTURES

The finite element method (FEM) has undoubtedly become the most popular and
powerful analytical tool for studying the behaviour of a wide range of engineering and
physical problems. Several general purpose finite element softwares have been developed,
verified and calibrated over the years and are now available to almost anyone who asks
and pays for them. Furthermore, concepts of FEM are usually offered by all engineering
departments in the form of postgraduate and even undergraduate courses.
One of the important applications of FEM is the analysis of crack propagation
problems. Fundamentals of the present form of the linear elastic fracture mechanics
(LEFM) came to the existence practically in naval laboratories during the First World
War. Since then, LEFM has been successfully applied to various classical crack and
defect problems, but remained relatively limited to simple geometries and loading
conditions.
Introduction and fast development of the finite element method drastically changed
the extent of application of LEFM. FEM virtually had no limitation in solving complex
geometries and loading conditions, and soon it was extended to nonlinear materials and
large deformation problems (Zienkiewicz et al. 2005). As a result, LEFM could now rely
on a powerful analytical tool in order to determine its fundamental concepts and
governing criteria such as the crack energy release rate and the stress intensity factor for
any complex problem. General LEFM stability criteria could then be used to assess the
stability/propagation of an existing crack.
Application of FEM into linear elastic fracture mechanics and its extension to elastic
plastic fracture mechanics (EPFM) has now expanded to almost all crack problems.
Parametric studies and experimental observations have even resulted in the introduction
of new design codes for containing a stable crack. However, the essence of analyses
remained almost unchanged: LEFM basic concepts combined with classical continuum
based FEM techniques through smeared or discrete crack models.
After a few decades, a major breakthrough seemed to be evolving in the fundamental
idea of partition of unity and in the form of the eXtended Finite Element Method (X-FEM
or XFEM).

Trim added in PDF - Aptara

 2 Extended Finite Element Method

1.2 ANALYSIS OF DISCONTINUITIES

Progressive failure/fracture analysis of structures has been an active subject of research

for many years. Historically, it was addressed either within the framework of continuum
mechanics, including computational plasticity and damage mechanics, or the
discontinuous approach of fracture mechanics (Owen and Hinton 1980).
These methods, however, are applied to fundamentally different classes of failure
problems. While the theory of plasticity and damage mechanics are basically designed for
problems where the displacement field and usually the strain field remain continuous
everywhere (continuous problems), fracture mechanics is essentially formulated to deal
with strong discontinuities (cracks) where both the displacement and strain fields are
discontinuous across a crack surface (Fig. 1.1) (Mohammadi 2003).

Figure 1.1 Different categories of continuities.

In practice, fracture mechanics is also used for weak discontinuity problems, and both
damage mechanics and the theory of plasticity have been modified and adapted for
failure/fracture analysis of structures with strong discontinuities. It is, therefore, difficult
to distinguish between the practical engineering applications exclusively associated with
each class of analytical methods.
Inclusion of some basic concepts from fracture mechanics such as non-local models,
energy release rate, softening models in combination with adaptive remeshing techniques
have allowed for successful simulations of crack problems with a certain level of
accuracy.

Trim added in PDF - Aptara

 Introduction 3

1.3 FRACTURE MECHANICS

Fundamental concepts of fracture mechanics can be traced back to the late nineteenth and
early twentieth centuries. Both experimental observations and theoretical elasticity helped
to create the fundamental aspects of the theory of fracture mechanics. Major differences
between the theoretical prediction of tensile strength in brittle materials and the
experimentally measured one was explained by the assumption of existing minute flaws
and defects; predicting drastic changes in the distribution of the stress field around each
flaw, regardless of its actual size.
Introduction of the fundamental concepts of stress intensity factor, energy release rate,
etc. changed the way a crack problem used to be analysed. Theoretical studies proved that
even for the case of a small tiny circular hole inside an infinite tensile plate, a tensile
stress concentration factor of 3 is predictable at a point adjacent to the hole, and in
addition, generation of a compressive stress field for the infinite tensile plate is also
anticipated.
Global (non-local) energy based methods were gradually developed and solutions for
classical problems were also obtained. Energy based methods allowed the classical
fracture mechanics to be extended to nonlinear problems. Introduction of the J integral
was a major breakthrough that allowed powerful numerical methods such as the finite
element method to be efficiently used for determining the necessary fields and variables.
Future developments benefited greatly from this joint approach; basic formulations from
fracture mechanics to assess the stability of cracks, and the analytical tool from the finite
element method to allow simulation of problems with arbitrary geometries, boundary
conditions and loadings.

1.4 CRACK MODELLING

Various methods have been developed over the years for simulation of the problems
involved with creation and propagation of cracks. Analytical, semi-analytical and
numerical approaches, such as the boundary integral method, the boundary element
method, the finite element method and recently a number of meshless methods, have been
successfully used for modelling cracks; each one provides advantages and drawbacks in
handling certain parts of the simulation. Although the same concepts can be more or less
applied to many numerical methods, the emphasis in this book is only put on the finite
element method as a basis for its extension to the extended finite element method.
Crack simulation in the finite element method has been performed by a number of
methods. They include the continuous smeared crack model and several discontinuous
approaches such as the discrete inter-element crack model, the discrete crack model and
the discrete element based model. Recently a new class has been proposed that simulates
the singular nature of discrete models within a geometrically continuous mesh of finite
elements. The extended finite element method has emerged from this class of problems,
and is based on the concept of partition of unity for enriching the classical finite element
approximation to include the effects of singular or discontinuous fields around a crack.

Trim added in PDF - Aptara

 4 Extended Finite Element Method

1.4.1 Local and non-local models

Early attempts to simulate crack problems by the finite element method adopted a simple
plasticity based FEM. The algorithm was to check the stress state at the integration points
against a material strength criterion, similar to hardening plasticity problems except for a
negative hardening modulus to account for the softening effects of cracking. The
behaviour of a point was only affected by its own stress state (point 1 in Fig. 1.2).
However, it was soon realised that such local results may become mesh dependent
and unreliable. The conclusion was that the cracking could not solely be regarded as a
local point-wise stress based criterion, and non-local models had to be adopted.

Figure 1.2 Local and non-local evaluation of stress state.

Then, non-local models were proposed to avoid mesh dependency of the plasticity
based solutions for simulating crack problems (Bazant and Planas 1997). To clarify the
basic idea, consider the simple case of point 2 in Fig. 1.2, where the fracture behaviour of
each point is determined from a non-local criterion expressed in terms of the state
variables at that point and a number of surrounding points.

1.4.2 Smeared crack model

The smeared crack model has been frequently used in the finite element simulation of
fracture and crack propagation problems. Rather than trying to geometrically model a
crack, the smeared crack model simulates the mechanical effects of the crack in terms of
stiffness or strength reduction. It is in fact a continuous approach for a
discontinuous/singular problem. In this model, the discontinuity caused by a discrete
crack within an element is simulated by a distributed (smeared) equivalent field over the
entire domain of the element, as depicted in Fig. 1.3a (Owen and Hinton 1980). The main
advantage of the method is that it does not require any local or global remeshing in the
process of crack propagation.

Trim added in PDF - Aptara

 Introduction 5

Figure 1.3 Different models for simulation of a crack.

Trim added in PDF - Aptara

 6 Extended Finite Element Method

1.4.3 Discrete inter-element crack

In this approach, existing cracks are simply defined along the finite element edges (Fig.
1.3b). The strong discontinuity is then automatically assumed in the displacement field
across the crack. However, it cannot account for the singular field around the crack tip,
unless special singular finite elements are used. The model is extremely simple for
predefined existing crack paths along the element edges, however, it becomes rather
difficult for modelling general crack propagation paths as it requires a remeshing of the
model. It also dramatically increases the risk of mesh dependency.

1.4.4 Discrete cracked element

This model is an improvement on the inter-element discrete crack model, as it allows for
cracks to be defined or propagated inside the finite elements. Fig. 1.3c illustrates a model
in which the crack path is through the middle of a finite element. A local remeshing
technique combined with adaptivity methods have to be adopted to create a new mesh by
splitting the cracked element and dividing adjacent elements to ensure compatibility of the
neighbouring finite elements. Adaptivity techniques are applied to compute the state
variables within the newly created elements from the state variables of their parent
elements.
A class of combined finite/discrete element procedures have been successfully
developed in the past decade for simulation of progressive fracturing due to impact and
explosive loadings. They may also take into account the effects of post-cracking
interactions, including fully nonlinear frictional behaviour (Mohammadi 2003).
The method is clearly very expensive as it requires time-consuming algorithms for
cracking, remeshing and contact detection/interactions every time a new crack or body is
created or each time a new potential contact is anticipated. A remedy is to avoid
remeshing of the whole model by gradual local remeshing and updating techniques
according to the advancement of cracks.

1.4.5 Singular elements

In a major development, singular finite elements were developed for simulating crack tip
singular fields. They provide major advantages; the model is simply constructed by
moving the nearby midside nodes to the quarter points – absolutely no other changes in
the finite element formulation are required (Fig. 1.3d). The use of these elements has
considerably upgraded the level of accuracy obtained by the finite element method for
simulation of crack tip fields (Owen and Fawkes 1983). Prior to the development of
XFEM, singular elements have been the most popular approach for fracture analysis of
structures.
Singular elements, however, lacked the capability of modelling discontinuity across a
crack path. As a result, they had to be used with one of the other preceding methods to
simulate the crack path.

Trim added in PDF - Aptara

 Introduction 7

1.4.6 Enriched elements

In this model, the singular or discontinuous displacement field within a finite element is
simulated by a special set of enriched shape functions that allow for accurate
approximation of the displacement field. Fig. 1.3e illustrates a model in which the crack
path is through the middle of a finite element. Presence of the crack is not geometrically
modelled and the mesh does not need to conform to the crack path. Additional enrichment
approximation is added to the classical finite element model to account for the effects of a
crack or discontinuity (Moës et al. 1999).
The main advantage of the method is that it does not require any remeshing in the
process of crack propagation. By advancement of the crack tip location or any change in
its path due to loading conditions, the method automatically determines the elements
around the crack path/tip and generates necessary enrichment functions for the associated
finite elements or nodal points accordingly.
Locations of initial cracks or potential propagation paths do not affect the way the
initial finite element model is constructed. Multiple cracking and intersecting cracks can
be similarly simulated by the same finite element mesh with comparable levels of
accuracy.

1.5 ALTERNATIVE TECHNIQUES

The finite element method has been widely used for fracture analysis of structures for
many years. Its earlier disadvantages have been avoided by the development of new ideas
and techniques and it has now become a mature powerful approach for the analysis of
many engineering and physical problems. Nevertheless, alternative methods, such as
various classes of meshless methods, are increasingly being adopted.
Meshless methods have developed significantly in the last decade. The element-free
Galerkin method (EFG) (Belytschko et al. 1994), meshless local Petrov–Galerkin
(MLPG) (Atluri and Shen 2002), smoothed particle hydrodynamics (SPH) (Belytschko et
al. 1996), finite point method (FPM) (Onate et al. 1995), reproducing kernel particle
method (RKPM) (Liu et al. 1996), HP-clouds (Duarte and Oden 1995), equilibrium on
line method (ELM) (Sadeghirad and Mohammadi 2007) and many other meshless
methods have been used for failure and fracture analysis of structures. Discussion on these
techniques is out of the scope of this book. Nevertheless, they share similar ideas and
many parts of the present discussion, methodology and formulation on enrichment
techniques can be extended to meshless methods and other numerical techniques.

1.6 A REVIEW OF XFEM APPLICATIONS

1.6.1 General aspects of XFEM

The basic ideas and the mathematical foundation of the partition of unity finite element
method (PUFEM) were discussed by Melenk and Babuska (1996) and Duarte and Oden

Trim added in PDF - Aptara

Other documents randomly have
different content
say about the Globe player what Polonius said about the player in
Hamlet:
Look, whe’r he has not turn’d his colour, and
has tears in’s eyes. Prithee no more!
[II, ii, 542-543]
At the peak of his passion he might well have fitted Hamlet’s
description of the player who
in a dream of passion,
Could force his soul so to his own conceit
That, from her working, all his visage wann’d,
Tears in his eyes, distraction in’s aspect,
A broken voice, and his whole function suiting
With forms to his conceit.
[II, ii, 578-583]
To this type of ceremonious acting, the heart of which was
overwhelming passion intensively portrayed, neither the adjective
formal nor natural applies. I suggest that we accept the inevitable
adjective and call it romantic acting, but romantic acting understood
in the finest sense before decadence and extravagance set in. The
Globe company brought this art to perfection.
 Chapter Five

THE STAGING

I. STAGE ILLUSION AT THE GLOBE PLAYHOUSE

Staging, like acting, is an art of illusion, but its illusion, unlike that of
acting, deals not with being but with time and space. In the
manipulation of time, it has long been recognized that Shakespeare
is a master. An oft-cited example of his mastery occurs in the guard
scene in Othello (II, iii). During the course of the action a night is
made to pass. At the beginning of the scene, the time is not yet “ten
o’ the clock” (15). At the conclusion, Iago remarks, “By th’ mass, ’tis
morning!” (384). In the midst of the alarum, Othello speaks of night
and Iago agrees that Cassio should see Desdemona “betimes in the
morning” (335). Here, as elsewhere, Shakespeare creates his own
illusion of time corresponding neither to actual chronology nor to
agreed convention, but solely to narrative demands.
It has also been generally recognized that Shakespeare may utilize
more than one time scheme within a single play. For example, after
Edmund has shown “Edgar’s” letter to his father, the Duke of
Gloucester, he assures him that he will seek out Edgar as quickly as
he can,
convey the business as I shall find means, and
acquaint you withal.
[I, ii, 109-111]
In Act II, scene i, three scenes later, he expedites his plot,
presumably without delay, for the action picks up where it had left
off. In the intervening scenes, however, Lear spends sufficient time
at Goneril’s castle for her to complain to the Steward, “By day and
night, he wrongs me!” (I, iii, 3). Certainly the spectator is to suppose
that a good portion of a month has gone by.
Through a kind of illusion the author accelerates or decelerates
the passage of time to fit the needs of his narrative. Thus, the time
sequence varies during the course of the play. In some scenes time
is extended, in others highly contracted. Antony is told, only a
moment after the mob, which he has stirred to fury, rushes out to
revenge Caesar’s death, that Brutus and Cassius have fled before
this same mob. The reference point, manifestly, is not the length of
time that the events would require in actuality, or a fixed standard of
time, such as the twenty-four-hour neoclassical day, or a symbolic
dimension, such as the morality time scheme of man’s life on earth,
but the duration of time required to tell the story. This narrative
ordering of time, moreover, has a parallel in a similar narrative
ordering of space.
Simultaneous staging illustrates the operation of such ordering of
space. By simultaneous staging is meant, in this instance, the
practice of mounting more than one setting on stage at the same
time so that during one scene the setting for another is already
present. The degree to which it was employed by the popular
companies is a matter of controversy.
In 1924 E. K. Chambers endeavored to distinguish between
simultaneous staging in the private theaters and sequential staging
in the public playhouses. But Professor George Reynolds has shown
that at the Red Bull, some of the time at least, simultaneous staging
was practiced. Later studies by George Kernodle and C. Walter
Hodges have supported his position. In writing about simultaneous
staging Reynolds, as well as Kernodle and Hodges, refers to the
disposition of properties only. Reynolds argues that properties from
one scene were occasionally left on-stage during the playing of
another. Or he suggests that tents or shops, utilized much like the
mansions of the medieval stage, were erected on-stage. He cites the
tents scene in Richard III (V, iii), where both Richard’s and
Richmond’s tents occupy the stage, as evidence that “theaters
permitted violation of realistic distance and the use of simultaneous
settings.” Instances of such simultaneity, although not abundant, do
occur among the Shakespearean Globe plays.
The disguised Kent is placed in stocks before Gloucester’s castle
where he is to remain all night (II, ii). The Quarto specifies that at
the end of a soliloquy he “sleeps.” A soliloquy by Edgar follows. After
Edgar’s exit, with the coming of morning, Lear arrives. Editors
frequently treat the sleep and Edgar’s exit as the conclusions of
separate scenes, thus marking Edgar’s soliloquy Act II, scene iii, and
the scene commencing with Lear’s arrival, Act II, scene iv. However,
neither the Folio nor the Quarto texts have any divisions at these
points, although the Folio text is otherwise divided. John C. Adams,
in his proposed staging of King Lear, suggests that the “inner stage”
curtain was closed while Kent sleeps in order to allow Edgar to
deliver his soliloquy, and then reopened for the next scene. But the
direction “sleeps” indicates that this was not the case. Edgar merely
entered while Kent slept in the stocks. Whether he was supposed to
be in the same part of the castle yard or another part does not much
matter. In this instance an imaginative expansion of space occurs
and he “does not” see Kent.
A similar instance occurs in As You Like It. While Amiens and
Jaques are singing in the Forest of Arden, a banquet is brought out.
Seeing the uncovered dishes, Amiens says,
Sirs, cover the while; the Duke will drink under
this tree.
[II, v, 32-33]
After they sing some more, Jaques announces that he will go off to
sleep and Amiens replies:
And I’ll go seek the Duke. His banquet is prepar’d.
[64-65]
These definite exit lines spoken by Amiens, as well as those spoken
by the Duke at the end of Act II, scene vii (where he is careful to
have Adam supported off stage), indicate that discovery of the
banquet is not intended in either scene. Between the setting and
partaking of the banquet, there intervenes the scene in which
Orlando and Adam enter the forest fainting from want of food. Here
is demonstration of the blending of general localization with
simultaneous staging.
However, such simultaneous staging did not set the style for an
entire play. Nowhere is there evidence that mansions or properties
were left on-stage throughout an entire play. Nor is this surprising. It
is apparent by now that scenic materials appeared infrequently on
the Globe stage. Therefore, if there were conventions of spatial
order, they involved not merely the physical elements of staging but
more especially the organic elements, namely, the actors.
A nonrealistic ordering of space becomes necessary when the
demands of a dramatic story create a disparity between the actual
dimensions of the stage and the spatial dimensions of the action.
Utilizing the theatrical conventions of the age, illusion masks this
disparity. Such illusion is a product of two factors: the extension
and/or compression of space and the juxtaposition of actors and
properties.
As in the case of temporal illusion, Elizabethan spatial illusion does
not obey a fixed proportion between stage and reality. It employs
neither the unity of place nor the cosmic range of medieval drama.
Between property and actor and between actor and actor, space
assumes whatever dimension the narrative requires. This is true not
only of the compression of space, that is, how closely characters
stand to one another, but of their dramatic relationship, that is, the
quality of that proximity.
To illustrate how the Elizabethans employed narrative space
relationships between actors, I turn to a striking, and, as far as I am
aware, hitherto unnoticed instance of compression in one of the
Globe plays, Pericles.
 In the first scene of the play Pericles seeks the hand of the
Daughter of Antiochus. To win her, he must successfully answer a
riddle. To fail, as many princes before him have done, means death.
After the Daughter appears before him in all her regal beauty,
Pericles receives the text of the riddle which he reads aloud. Almost
immediately he fathoms the meaning: Antiochus and his daughter
have committed incest. Pericles expresses this revelation in an aside,
in the midst of which he addresses the Daughter directly.
Y’are a fair viol, and your sense the strings:
Who, finger’d to make man his lawful music,
Would draw heaven down, and all the gods, to hearken;
But being play’d upon before your time,
Hell only danceth at so harsh a chime.
Good sooth, I care not for you.
[I, i, 81-86]
We might assume that, since the character speaks an aside, the
actor was standing some distance from the Daughter in order to give
the illusion that he is not overheard. But the next line, which
Antiochus addresses to Pericles, shows that Pericles was actually
next to the Daughter.
Ant. Prince Pericles, touch not, upon thy life,
For that’s an article within our law,
As dangerous as the rest. Your time’s expir’d.
Either expound now, or receive your sentence.
[I, i, 87-90]
Apparently, Pericles in his aside gestures toward the Daughter on the
line, “Good sooth, I care not for you.” Antiochus misinterprets the
meaning of the gesture and warns Pericles not to touch his daughter.
Thus, instead of speaking from afar, Pericles delivers the aside in the
midst of the other actors.
In analyzing the aside as a dramatic device, writers have accepted
the convention but rejected a conventional delivery by suggesting
that in performance the platform stage enabled the actor to render it
realistically. Not only this scene in Pericles, but equally significant
instances of spatial compression contradict this theory. Many asides
give the actor neither time nor motivation for creating verisimilitude.
When Othello meets Desdemona, after Iago has awakened the
“green-eyed monster” within him, he is struggling to hide his
conviction of her guilt. Desdemona greets him.
Des. How is’t with you, my lord?
Oth. Well, my good lady. O, hardness to dissemble!
How do you, Desdemona?
[III, iv, 33-35]
Today the actor mutters the aside, “O, hardness to dissemble,” turns
away, or in some other manner endeavors to give plausibility to the
convention that Desdemona does not hear the remark. In final
desperation, he may cut the line. The study of asides below shows
that these were not the methods employed at the Globe.
Naturally, the high degree of spatial compression among the
players caused a change in the quality of their relationships. When
one actor comes closer to another than realistic action plausibly
admits, as in the scene in Pericles, he destroys illusion, if it is one of
reality, or he creates a new illusion, if it is one of convention. By
standing near the defiled princess while he unravels the mystery, the
actor of Pericles can convey his horror with maximum effectiveness,
and by speaking his aside near her while he paints a word picture of
her outer beauty and inner pollution, he can project his revulsion at
her foul proximity. The Globe players, in the staging of asides, did
not think in terms of creating an illusion of actuality but of relating
the crucial elements of the narrative to each other. Within such a
frame of reference the dilemma, folly, or scheme which gives rise to
an aside is demonstrated more lucidly and more dramatically than it
could be within a realistic frame of reference. What is true of the
aside is equally true of observations, disguises, concealments,
parleys, and other theatrical devices.
The conventions governing grouping of actors also governed the
sequence of actions. From scene to scene, and within scenes, space
had a fluidity which was accommodated to the narration.
Generalization of locale required such fluidity, for locale was as broad
or as narrow as occasion demanded. The picturing of locale, we
must remember, was not accomplished with scenery. Nor was
passage from one locale to another accomplished through physical
changes in the stage façade, as some scholars have insisted.
According to various views, the drawing of a curtain or a shift from
one part of the stage to another or from one mansion to another
was a conventional means of conveying a change of place to the
audience. All these views assume in common that the establishment
of space was dependent upon clues of a physical sort.
The application as well as the refutation of such an assumption
can be illustrated in the assassination scene of Julius Caesar, which
begins in the streets of Rome and moves to the Capitol (III, i).
Ronald Watkins would express this sequence in a change in the
stage itself. To mark the moment when the scene shifts into the
Capitol, he would open the “inner stage” curtain to reveal a state for
Caesar.[1] The possibility that the street and the Capitol were
situated in the same imaginary area is never explored although there
is no instance in a Globe play where a shift takes place like that
which Watkins predicates. Before examining this scene in detail, it
might be well to turn to another Globe scene which is unqualified
evidence against Watkins’ method of staging.
In scene x of Miseries of Enforced Marriage, it will be recalled,
Butler has convinced Ilford, Bartley, and Wentloe that he can provide
them with rich wives. Appointing a time to introduce them to their
“brides-to-be,” he arranges to meet first Ilford and then the other
two “at the sign of the Wolfe against Gold-smiths row” (Sig. G1v).
After these rakes depart, Butler soliloquizes upon the punishment
that he will inflict upon them for their villainy. At the conclusion of
this brief soliloquy, he does not exit. Instead, Thomas and John
Scarborrow enter.
But. O, are you come. And fit as I appointed.
[Sig. G2r]

He bids them wait while he sets up the plot for Ilford. The scene
with Ilford is played in continuous fashion. There is no indication
that the scene has shifted to any other part of the stage, for Ilford
observes the Scarborrows from a window. When Wentloe and
Bartley appear, Wentloe points out the sign of the Wolfe. Through
dialogue, the audience is made aware that a change of locale has
occurred without either a clearing of the stage or a shift in area.
Furthermore, the appearance of the sign suggests one of three
possibilities: the sign was visible throughout the scene, thus creating
a type of simultaneous setting; it was not employed physically and
thus Wentloe’s line is imaginative; or it was placed in position during
the course of the scene. In any one of these instances the change of
scene did not depend upon any change in the form or size of the
stage space.
To return to Julius Caesar. It is possible to carry out the staging of
the scene as Watkins suggests. But there is no instance in the Globe
plays which clearly shows this to be Globe practice. A scene in The
Devil’s Charter (II, i, Sig. E1r) contains a similar scene of procession,
this time to a papal state. In the other stage directions of the play,
Barnes has carefully indicated when the enclosure was employed,
even within a scene, so that his failure to mention it in a stage
direction for this scene argues against its use. In that event the state
must have been thrust out. This method would serve equally well in
Julius Caesar with the result that both street and Capitol would be
simultaneously presented.
Essentially the stage was a fluid area that could represent
whatever the author wished without the necessity for him to indicate
a change in stage location. The actors did not regard the stage as a
place but as a platform from which to project a story, and therefore
they were unconscious of the discrepancy between real and dramatic
space. How far behind Malvolio were the “box tree” and his
tormentors? How far from Brutus and Cassius are Caesar and Antony
when Caesar sneers at Cassius’ “lean and hungry look”? Is the eye
meant to take in both parties at once? In performing these scenes,
the Globe players probably concentrated on making the observation
of Malvolio and the scornful characterization of Cassius dramatically
effective. That this frequently necessitated the substitution of
imaginary for real distance must have passed unobserved both by
the players and the audience.
Space, though flexible, was not amorphous. Principles of order in
staging existed independently of the stage façade and machinery. As
in Elizabethan graphic art during this period, the principles were
simple and derivative. The “primitive” art of the medieval period had
been suppressed by Henry VIII. No vital growth in a secular art
appeared to take its place. Save for some painters who created
original and masterly miniatures, among them the master Nicholas
Hilliard, the Elizabethans failed to develop a school of graphic art
and thus resorted to foreign artists or imitators. It is not surprising,
therefore, that the stage which developed at this period was simple
in composition and imitative in adornment. Massive and symmetrical,
not easily varied in its fundamental appearance, its boards served
any scene.
Evidence for fixing stage positions is scanty at best. The text of a
drama, unless it is accompanied by detailed stage directions, does
not contain the kind of evidence needed. Unfortunately, no one at
the Globe thought of preparing a regiebuch. Furthermore, methods
of rehearsal indicate that the pictorial arrangement of the actors
received little attention. Considering the history of the Elizabethan
acting company and the conditions of its repertory, it is not unlikely
that traditional patterns of arrangement were retained and repeated.
Novelty in the stage picture is a characteristic of the director’s
theater, not of the stock company’s repertory. But, though the
evidence for stage composition is scanty, what evidence there is is
consistent.
The simplest order in art is symmetrical balance. It is this type of
composition which one observes in the Globe plays from time to
time. At a banquet in The Devil’s Charter, Act V, scene iv, Pope
Alexander enters with three cardinals and three soldiers. The stage
direction reads,
The Pope taketh his place, three Cardinals on one side and [three]
captaines on the other. [Sig. L1r]
Poisoned at this banquet by the Devil, Alexander rushes to his study,
Alexander unbraced betwixt two Cardinalls in his study looking upon a
booke, whilst a groome draweth the Curtaine. [Sig. L3r]

This might be an echo of Richard III’s position between the Bishops

as he receives the Lord Mayor’s embassy from London. A more
dramatic use of symmetry can be found in the finale of Miseries of
Enforced Marriage. At the last moment Scarborrow repents his wild
courses. Surrounded by the brothers and sister he has ruined, the
wife and children he has neglected, and the uncle he has abused, he
is deeply shamed.
Harke how their words like Bullets shoot me thorow
And tel mee I have undone em, this side might say,
We are in want and you are the cause of it,
This points at me, yare shame unto your house,
This tung saies nothing, but her lookes do tell,
Shees married but as those that live in hel.
[Sig. K4r. My italics]
The demonstratives indicate brothers and sister on one hand, the
uncle on the other, and his wife next to him.
This type of symmetry can be seen in Shakespearean plays also.
At one point in Antony and Cleopatra Antony’s soldiers, while on
watch, hear the subterranean music which signifies, according to
one of them, that “the god Hercules, whom Antony lov’d,/Now
leaves him.” For the setting of the watch occurs the stage direction,
“They place themselves in every corner of the stage” (IV, iii, 7).
What arrangement could be simpler? In the same play there is
another example. Antony and Caesar are to meet to settle their
dispute (II, ii). The scene opens with Lepidus urging Enobarbus to
“entreat your captain/To soft and gentle speech.” Then the two
monarchs of the world enter from opposite sides of the stage. I
quote at length to make the balance clear.
Lep. Here comes
the noble Antony.
(Enter Antony and Ventidius.)
 Eno. And yonder, Caesar.
(Enter Caesar, Maecenas, and Agrippa.)
Ant. If we compose well here, to Parthia.
Hark, Ventidius.
Cae. I do not know,
Maecenas. Ask Agrippa.
Lep. Noble friends,
That which combin’d us was most great, and let not
A leaner action rend us. What’s amiss,
May it be gently heard. When we debate
Our trivial difference loud, we do commit
Murther in healing wounds. Then, noble partners,
The rather for I earnestly beseech,
Touch you the sourest points with sweetest terms,
Nor curstness grow to th’ matter.
Ant. ’Tis spoken well,
Were we before our armies, and to fight,
I should do thus.
(Flourish)
Cae. Welcome to Rome.
Ant. Thank you.
Cae. Sit.
Ant. Sit, sir.
Cae. Nay then.
[13-28]
And they sit, to discuss their grievances. From the entrance to the
final seating, the scene and dispositions are balanced. At the end of
this episode there is a formal symmetrical grouping: Caesar seated
with his two supporters in attendance facing Antony with his two
supporters in attendance. Between them, mediating the matter, is
Lepidus.
Throughout the Shakespearean Globe plays instances of this sort
can be found, not only in the arrangement of the actors but also in
the writing of the scenes. An extended example of verbal symmetry
occurs in As You Like It, where Rosalind vows to marry Phebe if she
marries any woman (IV, ii, 90-118). Often these symmetrical
arrangements are taken for granted because they seem dramatic
and do not disturb the flow of narrative. Yet occasionally we can
discern dramatic logic sacrificed for symmetrical arrangement. This
“failing” can be more graphically observed in the buildings of the
period and, therefore, I digress for a moment. A feature of the great
houses built as show places during the Tudor age was the adherence
to symmetrically balanced design. Usually a central structure would
be flanked by more or less elaborately developed ells or wings, as at
Wollaton Hall, Hatfield House, Charlton House, or Hardwick Hall. The
main hall was in the center, naturally, and the quarters of the
noblemen were in one wing. In the other wing the buttery, scullery,
or otherwise menial part of the household was located. In both
Wollaton and Hardwick Halls, the kitchen or scullery occupies the
front chamber of only one wing to balance the opposite lordly wing.
[2] From a functional point of view in planning, the symmetrical
arrangement did not satisfy the living accommodations of the Tudor
household. But from a visual point of view, it represented a dignity
and order that relatively unsophisticated builders could create.
Despite the obvious waste in space, the visual need determined the
structural design.
This tendency can be observed on the stage. I have already cited
the scene in Twelfth Night, when Malvolio “returns” Olivia’s ring to
Viola. Olivia had sent him to run after the “peevish” boy to tell him
that she would not take “his” ring (I, v, 318-323). We should
suppose that, in order to catch the boy, Malvolio would have
followed Viola on the stage. Yet the stage direction clearly specifies
that they enter “severally,” that is, from opposite sides of the stage.
The entrance is symmetrical but not logical.
The general thesis for symmetrical staging that I have advanced
must be qualified in two respects. First, the reliance upon
symmetrical arrangement was probably stronger in the earlier than
in the later period. The plays themselves change from a more
formal, balanced arrangement of speeches to a more colloquial,
asymmetrical arrangement. The balanced dirges of the various
queens in Richard III (IV, iv) and the measured laments of Blanch in
King John (III, i, 326 ff.) begin to disappear. However, they do not
wholly vanish during the Globe period.
 Secondly, the principles of composition may not be readily
perceived in scenes involving only a few characters. Therefore, in the
Globe plays symmetry as an element of staging can be best studied
in group scenes, for it is a simple way to arrange groups of actors.
The nature of Elizabethan dramatic material made simple balance
not only the most feasible but also the most meaningful method of
composition.

II. STAGE GROUPING AT THE GLOBE PLAYHOUSE

In considering grouping on the Elizabethan stage, we should keep

in mind the basic conditions of production. During its periods of
rehearsal the Globe company was actively engaged in daily
performance. Within two weeks customarily, the actors had to learn
extensive parts and mount a multiscene play. In a certain proportion
of these scenes many characters appeared on stage. Once presented
the play was not repeated for some days. Furthermore, the stage on
which the actors played had poor sightlines. The only area from
which they could be seen by virtually all members of the audience
was at the center of the platform in front of the pillars, at the very
place where DeWitt’s Swan drawing shows a scene in progress.
Although most scenes in the Shakespearean Globe plays require
five people or less on the stage at any one time, there are still quite
a number of scenes or sections of scenes in which more than five
people appear. In the fifteen plays of Shakespeare in the Globe
repertory, I count one hundred and sixty-six such scenes or
episodes, or an average of more than ten in each play. The lowest
proportion is 14 per cent in Twelfth Night, the highest 61 per cent in
Coriolanus. Generally 20 to 30 per cent of a play consists of what I
term “group” scenes or episodes.[3]
In terms of the problems of staging, these group scenes fall into
four distinct categories. More than half of the group scenes, eighty-
eight, fall into category one. These are scenes in which though there
are actually five or fewer speaking characters on the stage, the
addition of one or more mute supernumeraries increases the size of
the group to six or more Almost all of these mute supers fall into one
of several distinct generic types, easily recognizable and probably
conventionally portrayed. The most frequently recurring types are
soldiers in thirty scenes; attendants and servants in twenty-three
scenes; and noblemen of one sort or another in twenty-one scenes.
A small but important type consists of the crowds in Julius Caesar
and Coriolanus. The rest of the supers come from various
miscellaneous classes, such as ladies, musicians, sailors, and so on.
It is probable that the stock-in-trade of the hired men and gatherers
was a standardized portrayal of such types. Problems in grouping
must have been solved as readily. The prevailing types, soldiers,
attendants, and noblemen, contain in their ranks and duties the
rationale for their positions upon the stage. Implied in the
relationship of servant to master or nobleman to king is an attitude
of service expressed in a characteristic manner. That this pattern
was representative of Globe plays as a whole is borne out by the
examination of the non-Shakespearean and non-Jonsonian plays in
the Globe repertory. Although in proportion there are fewer group
scenes in the non-Shakespearean plays than in the Shakespearean,
in the separation into types of group scenes, the same divisions are
evident.
A second category consists of group scenes which require more
than five actors with speaking roles on-stage at one time. This
numbers twenty-two. However, though there are more than five
characters on-stage, no more than five of them are active. In effect,
the others become mute observers, functioning much as the
nobleman, soldier, or attendant type. For example, in the debate
upon Grecian policy in Troilus and Cressida, Act I, scene iii, 1-212,
three people speak: Agamemnon, Nestor, and Ulysses. During the
utterance of these 212 lines neither Diomedes nor Menelaus speaks,
although they are present throughout. In this scene they are mere
supers. Donalbain is on-stage throughout Macbeth, Act I, scenes ii
and iv, but he does not speak. He, too, functions as a mute
nobleman. Once Lear faces Goneril and Regan before Gloucester’s
castle (II, iv, 129-298), Kent and the Fool, who have been prominent
hitherto, drop into the background as mute attendants. This
practice, not of subordinating characters but of reducing them to
ciphers, facilitated the handling of large groups of characters. That
this was the technique of the poet is evident when one considers
those scenes where characters, who have every reason to be active,
fail to respond to events in which they are immediately involved.
When the Duke reveals to Isabella the brother whom she thought
dead (Measure for Measure, V, i, 495-498), we might expect Isabella
to say something, but she does not. Or when Cleopatra beats the
messenger who brings the report of Antony’s marriage to Octavia
(II, v), we might expect the otherwise talkative Iras or Alexas to say
something, but Charmian alone intervenes. In that scene the others
play mute supers.
The third, and second most numerous, category of group scenes
requires more than five active characters on-stage at one time
excluding mute supers. There are forty-six instances of such scenes.
What distinguishes them as a class is that all of them represent
some type of situation which demands ceremonious grouping.
Among others there are banquet scenes, single combats, council
sessions, trials, parleys, processions, and greetings. In all of them
the formal character is marked, attention is directed to one focal
point, and the arrangement of the action is often symmetrical and
ceremonial (see Appendix C, chart ii).
It is apparent from categories one, two, and three that in the case
of 156 of the 166 group scenes, the organizing principle is ceremony
or duty. Movement and arrangement, though formal, are not
artificial. Rather, they reflect circumstances of Elizabethan life. In the
group scenes the personage of greatest prestige is usually the one
who directs the action and to whom the other characters relate
themselves. The importance of this organizing principle is
demonstrated by considering the plays of domestic life, such as The
Merry Wives of Windsor. Without a ranking figure, another system of
grouping had to be developed. In such a play, an object of ridicule,
accusation, or pity serves as the focal point, as in the final scene of
Merry Wives.
 In the ceremonious scenes it happens sometimes that the focal
figure is not a major character in the play, yet as the person of
highest rank he is the one to whom all the characters address
themselves. This is clearly the situation in Othello (I, iii), where
Brabantio accuses Othello before the Duke and Senate. It is the
Duke whom Othello answers.
Where no single figure serves as the point of reference in the
grouping, a center of activity invariably does. The wrestling scene in
As You Like It (I, iii) or the duel in Hamlet (V, ii) are examples of this
kind of organization. Another method is the processional. Most
processions pass over the stage with or without halting for brief
speeches. Occasionally the procession might combine a focus of
both activity and a central figure, as in Julius Caesar (I, ii). In some
instances characters on-stage describe or discuss members of the
procession (All’s Well, III, v; Pericles, II, ii; Troilus and Cressida, I,
ii). Given the free passage of the stage and a point of observation
when needed, these scenes offer little problem in staging. In fact,
the regular recurrence and similar arrangement of these scenes
suggest the influence of standardized staging.
Even where more than five characters are active in the course of a
group scene, more than five are rarely active during extended
portions of the scene. The finale of As You Like It will serve as a
succinct and relatively typical example. The scene opens with
Orlando and Duke Senior briefly discussing Ganymede (1-4).
Rosalind enters, still disguised, to make certain that the mutual
pledges of marriage hold. She asks each interested person in turn
for confirmation (5-25). Five speak, all but Phebe answering Rosalind
with one line. She speaks two. Orlando and the Duke return to the
discussion of Ganymede (26-34). Touchstone enters and engages in
conversation with Jaques and the Duke while Rosalind has a chance,
off-stage, to change into her maidenly garments (35-113). Hymen
appears, leading in Rosalind; the pledges are finally confirmed in
single-line refrains. Hymen blesses the marriages. Five speak (114-
156). The Second Brother enters to tell the story of Duke Frederick’s
conversion. He is welcomed by the Duke only. In fact, his brothers,
Orlando and Oliver, never speak to him. Jaques, in his own fashion,
blesses each marriage. Three speak (157-204). Rosalind delivers the
Epilogue. In scenes of this pattern there is no need for all the
characters to be seen at all times. Instead, the actors could come
forward when needed, to play where they could be heard and seen
by everyone. At the conclusion of such a portion of the scene, as
when Touchstone and Jaques finish speaking, the unneeded
characters could retire to the rear until called for once again.
That this was indeed the practice is illustrated by the Globe play,
Every Man Out of His Humour. Jonson’s stage directions in Act II,
scene iii, show that when Sordido and Fungoso are not needed, they
“with-draw to the other part of the stage” and that when Puntarvolo
has completed one part of his action he “falls in with Sordido, and
his Sonne” while other action is in progress.
The last category of group scenes contains as did those already
enumerated, more than five characters excluding supers. However,
these scenes do not have a formal arrangement. Thus, the method
of grouping these scenes is not quite so rigidly set as that of the
previous category. Of this sort there are relatively few examples,
only ten or about 6 per cent of the group scenes. Some of these
verge on a formal arrangement without fully realizing it. The scene
of choosing a husband in All’s Well (II, iii) is a unique example in
Shakespeare, although in the pattern of the writing there is a
symmetry which tends to give the scene a schematic quality. The
farewell scene in Antony and Cleopatra (III, ii) and the arrest scene
in Twelfth Night (III, iv) also approach formality. What distinguishes
these scenes from the rest of the formal group scenes is merely the
degree of ceremony.
The presence of formal patterns in stage grouping enabled the
Globe company to present large-cast plays with a minimum of
rehearsal. The presence of sub-scenes within the larger scene
enabled the essential action to be brought forward and viewed. Such
a practice naturally reduced the importance of the stage façade as a
frame for the stage picture, for the attending figures remained in the
background, near the tiring house, and the active characters came
forward to the front of the stage where they could be seen in the
round. Nothing hindered the operation of such a stage procedure,
for more than 80 per cent of all Globe scenes required no stage
machinery or properties whatsoever. Everything favored it. The
platform stage was not a gargantuan apron before a modern
proscenium. It was the stage and the group scenes were played to
make full use of its expanse and flexibility.
I have devoted this much attention to Elizabethan stage illusion
and the group scenes in order to show that there were theatrical
practices in operation which did not depend upon the stage façade
or machinery. Yet the scholar of Elizabethan staging invariably
approaches the subject by first considering the function of the stage
and its properties in identifying the location of scenes. E. K.
Chambers categorizes scenes according to what setting they need.
Even Reynolds, who understands the necessity for considering scene
situations rather than stage locations, uses the latter to determine
the arrangement of his book. The result of such an approach has
been that a drama, which in production relied almost wholly upon
the voice and movement of the actor, has been studied in terms of
its settings, its least pertinent part. When a modern character enters
a scene, he enters a definitely indicated place. The audience or
readers are made very conscious of that place, its odors, its
atmosphere, its effect upon the characters. But in the Elizabethan
drama, particularly in the Shakespearean, a character enters not into
a place but to another character. Where he enters is of secondary
importance—to whom he enters or with whom he enters is of
primary interest.
Coordinately, the continuity of action from scene to scene was
independent of the stage façade. This conclusion is a logical
corollary of the evidence offered in Chapter Three. The enclosure,
used for discovery or concealment, is introduced sometimes within
scenes, sometimes with scenes, but not for the purpose of providing
flow from scene to scene, as we saw. Neither the above nor the hell
below ever serve the function of enabling one scene to follow
another. Properties, even though they serve conventional uses,
appear too infrequently and too irregularly to afford a means of
scene connection. Consequently, these conclusions have led me to
draw up five premises covering continuity in staging.
(1) The mention of place in the dialogue does not necessarily
mean that either a part of the stage façade or a property is
employed. Only actual use of the stage area or property confirms its
employment or appearance on stage. (2) A new scene does not have
to be played in a different part of the stage from the previous one.
This premise is closely connected with the idea that (3) a change of
location in the narrative is not necessarily accompanied by a change
in location on stage. Most scholars have recognized that the exit of
one character and the entrance of another from a different door is
enough to signify a change of location. Although this is generally
true, there are exceptions even in these cases, for examples of
scenes exist where a change of location is effected without the
clearing of the stage (Julius Caesar, III, i; Miseries of Enforced
Marriage, scene x; Measure for Measure, III, i-ii; London Prodigal,
D3r-E1v). (4) No regular system of scene alternation occurs.
Brödmeier’s simple theory of alternation, one scene in front of a
curtain and one scene behind, has been discarded by scholars long
ago. But more elaborate systems of alternation, employing the
“inner” and “upper stages,” are still advanced. Examples are
available for examination in Watkins’ book and Reynolds’
reconstruction of Troilus and Cressida. (5) Evidence for the use of
the enclosure in one scene of a play does not mean that the
enclosure was used in other scenes for which there is no evidence.
Many years ago Ashley Thorndike advocated the opposite premise.
“Clear evidence of the curtained inner stage in one scene of a play
must be taken as a presumptive evidence that it was used in others,”
he wrote.[4] Thorndike’s presumption has been liberally interpreted
by students of staging. Perhaps the absence of additional mention of
the enclosure is the clearest proof of its limited use. After all, when
the total evidence for a curtained space is gathered together, the
bulk is fairly slim in comparison to the vast number of scenes which
contain no such mention. Of the 519 scenes in the Globe plays,
sixteen of them show fairly strong evidence of being partly placed in
the enclosure. This is about 3 per cent of the total. Perhaps the texts
of the non-Shakespearean plays offered by the Globe company
reflect a truer percentage. Of their 182 scenes, twelve show
evidence of enclosure use, or about 6½ per cent.[5] In either case
the total percentage is low.
These premises arise from my conviction that the part which the
stage façade played in the presentation of the plays has been greatly
overestimated. Visually, the façade was always the formal
background, but in the overwhelming number of cases the action
took place before it, not within it. Instead of looking to the façade
for the organizing principles of staging, it might be better to look to
the patterns of the scenes themselves.

III. ACTORS’ ENTRANCES UPON THE GLOBE STAGE

At one time, Sir Mark Hunter defined a scene as the action

between clearances of the stage.[6] Since this definition is generally
accepted, we can consider that the scene concludes with the exit of
all characters and commences with the entrance of other characters.
This so-called “law of reentry” operates in the overwhelming
majority of scene changes. It is rare for a character who has left the
stage in one scene to enter immediately in the very next. As C. M.
Haines has pointed out, most of the exceptions occur in battle
scenes. In those instances it is usual for an alarum or excursion to
separate the two scenes. The other exceptions are in large measure
suspect.[7]
Ready analogy to cinematic technique has led a number of
scholars to minimize the scene markings. Emphasis has been placed
on the flow of scene to scene, to the extent that the separation of
scene from scene has had to be made by a shift from one stage area
or mansion to another or by the opening or closing of a curtain.
However, in deemphasizing the contribution of the stage façade to
the continuity of the play, it is necessary to consider that the
pointing of scene divisions was managed by the actors themselves.
Overlapping of the exit and the entrance may not have been the
habit of the Globe company; instead separation and pause may have
been the method. The actors or stage attendants, on occasion, had
to bring out properties. This necessitated a pause, however brief.
Nor need this pause have been reflected in the text. For one
entrance in The Battle of Alcazar the stage direction in the Quarto
reads:
Enter the king of Portugall and his Lords, Lewes de Sylva, and the
Embassadours of Spaine.

In the plot of the play, however, the corresponding direction reads:

Enter: 2 bringing in a chair of state (mr. Hunt): w. Kendall Dab & Harry
enter at one dore: Sebastian: Duke of Avero; Stukeley: 1 Pages: Jeames
Ionas: & Hercules (th) to them at another dore Embassadors of Spaine mr
Iones mr Charles: attendants George and w. Cartwright:[8]

Unfortunately, no similar parallel of stage direction and plot exists for

any of the Globe plays. In these same plots, we may also notice, a
line was drawn across the page to separate one scene from another.
Probably this was done to clarify the sequence of scenes, but it had
the added effect of fixing the scene divisions firmly in the actor’s
mind. Together with the rhyming couplet which concluded so many
scenes, it may have encouraged the insertion of a slight pause
between the scenes.
In Chapter Three I fully examined the character of the scene
endings. The conclusions are relevant at this point although the
evidence need not be reviewed. Seventy-nine per cent of the scene
endings indicate explicitly or implicitly that the actors march off-
stage. About ten and one-half per cent of the scenes end with solo
exits. About the same number of scenes fail to indicate that the
actors actually move out. It is obvious, from this distribution, that at
the ends of scenes the playwright normally provided the actors with
exit lines or movements. These served a double purpose. They
stressed the conclusion of the scene, and they bridged the
movement across the large platform.
The sufficiency of such simple movement to separate scenes is
reflected in what I call split entrances or exits. The split entrance or
exit occurs when characters come together or go apart through
more than one entryway. Entrance of two or more characters “at
several doors” or exit of two or more characters bidding farewell to
one another are split. Of the 644 entrances and exits which begin or
end scenes in the Shakespearean Globe plays, only 12.1 per cent are
split scenes. Even of this low figure only 6.4 per cent are definitely
split scenes, the remaining number including probable cases. Thus
nearly 90 per cent of the scenes merely involve the exit of one actor
or group at one door and the entrance of another actor or group at
another. The split scenes are readily staged, if the third entry
through the center curtain is employed. Thus the burden of
maintaining the continuity and clarifying the story is placed on the
actors—not on the stage.
Shakespeare relies on few methods for opening a scene. In 339
entrances[9] in the Shakespearean Globe plays he employs eight
methods for 88 per cent of the entrances. The most frequent type of
entrance is that of the mid-speech, which accounts for over 40 per
cent of the scene beginnings. In such an entrance two or more
characters come on-stage engaged in a conversation the topic of
which was begun off-stage. This type of entrance is best adapted to
emphasize continuity of action. Among the seven other types is the
processional entrance, 9½ per cent of the total; the inquiry,
soliloquy, and commanding entrance, about 7 per cent each; and
finally the salutation, summoning, and emotional entrances, between
5 and 6 per cent each. In the commanding entrance a character
enters giving a command to someone already on-stage; in the
summoning entrance the character summons someone who is off-
stage, and in the emotional entrance a character enters disturbed by
some emotional experience, as Julius Caesar is after the
tempestuous night (II, ii).
 Except for the processional and salutation entrances, the
entrances plunge the audience into the midst of a new situation or a
more highly developed stage of an earlier situation. In this respect
the evidence would appear to contradict my suggestion that a hiatus
may have defined the scenes. But considered in terms of the stage,
the contradiction is more apparent than real. This can be seen by
turning to the mid-speech entrance, 132 examples of which appear
at the beginning of scenes. A typical example opens Othello.
Roderigo and Iago enter, apparently after Iago has told Roderigo of
Desdemona’s marriage.
Rod. Tush, never tell me! I take it much unkindly
That thou, Iago, who hast had my purse
As if the strings were thine, shouldst know of this.
Iago. ’Sblood, but you will not hear me!
If ever I did dream of such a matter,
Abhor me.
[I, i, 1-6]
But where do the characters begin speaking? At a stage door? The
stage doors on either side of the stage are virtually behind the stage
pillars. No matter how narrow one supposes these pillars to be, and
they cannot be very narrow considering their function of supporting
the heavens and huts, they interfere with action at the stage doors.
Although the exact locations of the doors in the back wall are
uncertain, they must have been behind or nearly behind the pillars if
one allows for the enclosure. Consequently, I doubt that the mid-
speech, which usually provides information vital to the narrative, was
begun at a door, and think it more likely that the characters took
several paces toward the center or forward before speaking. This
action may have provided a hiatus sufficient to mark a new scene.
Presence of such a hiatus is supported by the fact that the mid-
speech entrance seldom occurs within the body of a scene.
Shakespeare uses it almost exclusively to enable the actor to
maintain continuity from scene to scene. For example, in All’s Well
and Measure for Measure, fifteen and ten mid-speech entrances
respectively all occur at the beginnings of scenes.
 However, if the characters entered through the rear curtain, they
could engage in immediate conversation. Entrance of actors through
the enclosure curtains was not unusual, and, in fact, may have
occurred more frequently than we usually assume. For instance, in
The Battle of Alcazar, the Quarto stage direction reads:
Enter the king of Portugall and the Moore, with all theyr traine.

For the same action, the plot reads:

Enter at one dore the Portingall Army with drom & Cullors: Sebastian ...
att another dore Governor of Tanger ... from behind the Curtaines to them
muly mahamet & Calipolis in their Charriott with moores one on each side
& attending young mahamet....

Behind the terse stage direction then, lies a more elaborate entrance
involving the curtain. Although definite evidence for such entrances
does not exist in the Globe plays, there is, on the other hand, no
evidence to exclude such entrances. Moreover, there are several
situations which imply such use. At the conclusion of scene i in
Othello, Brabantio and Roderigo exeunt to seek Othello. At line 160
Brabantio had come out one door, representing his house. At line
184 he and Roderigo go out, certainly not back into the house.
Othello and Iago enter in mid-speech, surely upon the outer stage.
But from where? Not from the door through which Brabantio and
Roderigo just went out. Possibly from the door which only recently
had been the entrance to Brabantio’s house. Probably through the
curtain in the center of the stage. Although the evidence is not
conclusively applicable to the Globe plays, it may be pertinent to
note that in the Roxana drawing, the flap of the curtain is partially
open, and in the frontispiece to The Wits a character is shown
coming through the curtain. In all likelihood, actors regularly entered
through the center curtain, and when they did, they could begin
speaking immediately upon entrance. But when the entrances were
made through a stage door, I suggest that conversation was held
back for the several seconds needed by the actors to move into the
acting area proper and there to mark the beginning of a new scene.
 That a need to focus attention upon an entrance existed is evident
from a consideration of the entrances within the scenes. Many of
these entrances are heralded by some form of announcement or
question, such as “My lady comes,” or “How now?” or “Who comes
here?” Other means of emphasizing entrances were through action,
such as a procession, or through music, such as the horn
announcing Lear (I, iv), or through response to a previous
command, such as Lucius’ report of the Ides of March in Julius
Caesar (II, i). In As You Like It, I count thirty-one intrascene
entrances: twenty-one are announced, one is accompanied by
action, three are responses to a previous command or scheme, and
six are unprepared. In Lear, there are fifty-one intrascene entrances,
of which twenty-five are announced, ten accompanied by action,
three by music, and thirteen unprepared. The unprepared entrances
in Lear are usually unannounced for dramatic purposes. Oswald’s
entering impertinently to Lear (I, iv), Lear’s bearing in the body of
Cordelia (V, iii), and Oswald’s sighting “the proclaimed prize,”
Gloucester, (IV, vi) depend upon suddenness for dramatic effect.
In addition to directing attention to an incoming actor, the
announcement filled an awkward gap. The depth of the stage
caused a dislocation between the actors already on stage and those
coming on-stage. Frequently, the former would be at the front but
the entrant would be at the rear. It was necessary to allow time for
the entrant to come down stage. The full effect of these
announcements was to formalize the entrances and enhance their
ceremonial impression.
Just how conventional the entrance might have been can be seen
by examining a particular group of entrance announcements. About
forty-three entrances in the Shakespearean Globe plays are
accompanied by announcements of greater length than the brief,
“Who’s there?” These announcements run from two lines to sixteen
lines in length. Most of them are short, two to four lines in length,
but a few are longer than ten lines. In each of these instances a
character or characters on-stage describe or comment upon
someone who has just entered. Usually the entrant is aware of the
others, but it is understood that he does not hear the description.
Modern producers often try to cover these awkward entrances by
giving the entrant some motivated business to account for the delay
in speaking. But these scenes are frankly demonstrative, for the
audience is supposed to be aware of both parties. In Hamlet,
Polonius greets Hamlet, Rosencrantz, and Guildenstern. Hamlet,
without answering, says:
Hark you, Guildenstern—and you too—at each ear a hearer!
That great baby you see there is not yet out of his swaddling clouts.
[II, ii, 398-401]
And so forth for another three and one-half lines. Polonius can
“cover up” by waiting upon the prince, or by engaging in character
business, but in essence he becomes an inert object for that period.
The longest delay in an entrance, sixteen lines, occurs in
Coriolanus (V, iii, 19 ff.) when Coriolanus describes the delegation of
Volumnia, Virgilia, young Marcius, and Valeria approaching him. By
no means could it require a speech of that length for the actors to
reach him, no matter from what part of the stage they may have
entered or where he may have been standing. During his speech
they become the visible expression of the inner struggle that he is
about to undergo. If they move, they must move very slowly; if they
stand still, they compose a picture. It is highly unlikely that the
Globe company tried to “naturalize” this entrance by giving the
entrants business or movement which would divert the attention of
the audience from the effect their entrance was having upon
Coriolanus.
Essentially the plays were written to enable the actors to enter
effectively without the aid of the façade, to play intimately near the
audience, and to retire convincingly without loss of attention. When
one takes into account the number of processions, salutations,
commands, summonses, and expressions of duty introduced to
cover and emphasize the entrances, one realizes that continuity from
scene to scene was mannered rather than casual, ceremonious
rather than personal, conventional rather than spontaneous. The
effect was probably not too far removed from the daily social
manner of the Elizabethans, but on stage their natural predilection
for ceremony may have been more fully systematized.

IV. RECURRENT PATTERNS OF STAGING

The patterns of continuity then do not lie in a play’s use of the

stage façade but inhere in a play’s structure. Chapter Two traced the
principal method of Shakespearean storytelling with its apparent
looseness of construction but its actual scheme of central
intensification and narrative finale. Within this framework abounds a
tremendous variety of scenes which seem to defy classification.
Nevertheless, situations and devices do recur in Shakespeare’s plays.
It is to those recurrent devices that I now turn, for an examination
of their patterns provides the best means of envisioning the staging
of Shakespeare’s plays at the Globe.
At one extreme there are those devices, such as the soliloquy,
which are highly conventionalized and frequently employed. At the
other extreme are the situations or episodes which are so
individualized that they seem to rely upon no distinct dramatic
convention, and therefore seem to be “a mirror of nature.” Between
the common theatrical device and the unique dramatic situation exist
the many episodes and devices in Shakespeare which are more or
less formal and which are repeated with greater or lesser frequency
in play after play. Through the reconstruction of the staging of these
recurrent devices and scenes, such as asides, disguises, and so
forth, the practices of the Globe playhouse should become apparent.
I shall first consider the soliloquy, the aside, and the observation
scene. These forms being readily imitable appear throughout the
Globe repertory with frequency. For that reason comparisons in
function and technique are plentiful. Although these devices
compose a brief portion of a play, they contribute to the
development of the action and represent the theatrical method
employed to tell the story.
 The soliloquy is probably the most characteristic theatrical device
of the Elizabethan stage. In the great soliloquies of Hamlet and
Macbeth Shakespeare perfected this form of expression.
Unfortunately, these supreme examples have epitomized the content
and atmosphere of all soliloquies. The result has been injurious both
to the study of literature and the reconstruction of theatrical
conventions.
In tone and character, the soliloquy displays great variation.
Among the 144 soliloquies which I count in the Shakespearean
Globe plays, I distinguish three main subdivisions. All of these
represent some form of conscious thought brought to a point where
it verges on speech. Broadly, the soliloquies can be divided into
those which are essentially emotive in expression, those which are
cerebral, and those which are invocative. The divisions are not hard
and fast, however. The emotional release of Hamlet, after he
castigates himself as a “dull and muddy mettled rascal,” gives way to
rational plotting to ensnare his uncle. For convenience, however, it is
not inaccurate to speak of these three categories. The emotive
soliloquies make up about 40 per cent of the total; the rational,
containing philosophical comments, plotting, and moralizations,
make up about 46 per cent of the total; and the invocative, such as
Lady Macbeth’s call to the spirits of evil, make up about 7 per cent.
These figures are suggestive, not definitive, nor does it matter that
they are so. The important thing to note is that the introspective
soliloquy is rare. Among the emotive soliloquies, there are
expressions of sheer emotion, such as Orlando’s paean of love (As
You Like It, III, ii, 1-10) and Angelo’s cry of remorse (Measure for
Measure, IV, iv, 22-36), Ophelia’s lamentation over Hamlet (Hamlet,
III, i, 158-169), and Thersites’ railings (Troilus and Cressida, V, iv, 1-
18). But there are few examples of the soliloquy of inner conflict, no
more than 5 per cent of all the soliloquies.
Not only in character are the bulk of the soliloquies
nonintrospective, but also in style they are extroverted. Shakespeare
depends a great deal upon apostrophe to sustain the soliloquy. The
character’s address may be directed toward the gods (Pericles, III, i,
1-2: “Thou god of this great vast, rebuke these surges,/Which wash
both heaven and hell”) or to another person not on stage (Antony to
Cleopatra, IV, xiv, 50-52: “I come my queen.... Stay for me./Where
souls do couch on flowers, we’ll hand in hand/And with our sprightly
port make the ghosts gaze”) or to natural forces (Timon, IV, iii, 176-
196, to Mother Earth) or to bodily organs (Claudius in Hamlet, III, iii,
70: “Bow, stubborn knees”). In fact, this form of address may be
directed to anyone or anything. The effect of this literary figure was
to substitute a listener for an absent actor. True, the listener was
imaginative rather than actual, mute rather than responsive. But
instead of directing the soliloquy inward, the apostrophe enabled the
actor to direct it outward.
Other literary forms were also employed toward this end.
Frequently the character makes himself the listener by self-
interrogation. “Am I a coward? Who calls me villain?” asks Hamlet of
himself (II, ii, 598-599). Often the emotive soliloquy is couched in a
series of flat assertions or descriptions or comparisons, all of which
are contained in Hamlet’s soliloquy beginning, “How all occasions do
inform against me” (IV, iv). However, because twentieth-century ears
are acutely sensitive to psychological nuances suggested by a
soliloquy, they very often hear a false echo of inner revelation. Only
a few speeches of admittedly great soliloquies reveal profound
conflicts of the mind (Hamlet, I, ii, 129-159; II, i, 56-89; Macbeth, I,
vii, 1-28; II, i, 33-64; Julius Caesar, II, i, 10-69).
In line with the modern conception of the soliloquies as moments
of the most intimate, intensive personal revelation has arisen the
idea that the very front of the platform stage is the true province of
the soliloquy. Surrounded by the audience, so close that he could
almost touch the spectators, the actor is pictured as unveiling his
soul. But this view of the soliloquy must be questioned. Although
there is no evidence in the Shakespearean Globe plays concerning
the actors’ positions during the delivery of the soliloquies, in the
non-Shakespearean Globe plays there are four instances where
soliloquies are delivered from the enclosure, two each in The Devil’s
Charter and Thomas Lord Cromwell. Whether or not the speaker
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

[Link]