Fundamentals of Compressible

Fluid Mechanics

Genick Bar–Meir, Ph. D.

1107 16th Ave S. E.
Minneapolis, MN 55414-2411
email: “[email protected]”

Copyright © 2007, 2006, 2005, and 2004 by Genick Bar-Meir

See the file copying.fdl or copyright.tex for copying conditions.
Version (0.4.8.5 January 13, 2009)
‘We are like dwarfs sitting on the shoulders of giants”

from The Metalogicon by John in 1159

 CONTENTS

Nomenclature xv
Feb-21-2007 version . . . . . . . . . . . . . . . . . . . . . . . . . . xix
Jan-16-2007 version . . . . . . . . . . . . . . . . . . . . . . . . . . . xix
Dec-04-2006 version . . . . . . . . . . . . . . . . . . . . . . . . . . xx
GNU Free Documentation License . . . . . . . . . . . . . . . . . . . . . . xxiii
1. APPLICABILITY AND DEFINITIONS . . . . . . . . . . . . . . . . xxiv
2. VERBATIM COPYING . . . . . . . . . . . . . . . . . . . . . . . . xxv
3. COPYING IN QUANTITY . . . . . . . . . . . . . . . . . . . . . . . xxvi
4. MODIFICATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvi
5. COMBINING DOCUMENTS . . . . . . . . . . . . . . . . . . . . . xxviii
6. COLLECTIONS OF DOCUMENTS . . . . . . . . . . . . . . . . . xxix
7. AGGREGATION WITH INDEPENDENT WORKS . . . . . . . . . xxix
8. TRANSLATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix
9. TERMINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix
10. FUTURE REVISIONS OF THIS LICENSE . . . . . . . . . . . . . xxx
ADDENDUM: How to use this License for your documents . . . . . . xxx
How to contribute to this book . . . . . . . . . . . . . . . . . . . . . . . . xxxi
Credits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxi
John Martones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxi
Grigory Toker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxii
Ralph Menikoff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxii
Domitien Rataaforret . . . . . . . . . . . . . . . . . . . . . . . . . . xxxii
Gary Settles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxii
Your name here . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxii
Typo corrections and other ”minor” contributions . . . . . . . . . . . xxxiii
Version 0.4.8 Jan. 23, 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . xliii

iii
iv CONTENTS

Version 0.4.3 Sep. 15, 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . xliii

Version 0.4.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xliv
Version 0.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xlv
Version 0.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xlv
Version 0.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . li
Version 0.4.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lii
Version 0.4.1.7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lii
Speed of Sound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvi
Stagnation effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvi
Nozzle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvi
Normal Shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvi
Isothermal Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvi
Fanno Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvii
Rayleigh Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvii
Evacuation and filling semi rigid Chambers . . . . . . . . . . . . . . lvii
Evacuating and filling chambers under external forces . . . . . . . . lvii
Oblique Shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvii
Prandtl–Meyer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvii
Transient problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . lvii

1 Introduction 1
1.1 What is Compressible Flow ? . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Why Compressible Flow is Important? . . . . . . . . . . . . . . . . . 2
1.3 Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3.1 Early Developments . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.2 The shock wave puzzle . . . . . . . . . . . . . . . . . . . . . 5
1.3.3 Choking Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3.4 External flow . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3.5 Filling and Evacuating Gaseous Chambers . . . . . . . . . . 15
1.3.6 Biographies of Major Figures . . . . . . . . . . . . . . . . . . 15

2 Review of Thermodynamics 25
2.1 Basic Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3 Fundamentals of Basic Fluid Mechanics 33

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2 Fluid Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.3 Control Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.4 Reynold’s Transport Theorem . . . . . . . . . . . . . . . . . . . . . . 33

4 Speed of Sound 35
4.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.3 Speed of sound in ideal and perfect gases . . . . . . . . . . . . . . . 37
4.4 Speed of Sound in Real Gas . . . . . . . . . . . . . . . . . . . . . . 39
CONTENTS v

4.5 Speed of Sound in Almost Incompressible Liquid . . . . . . . . . . . 43

4.6 Speed of Sound in Solids . . . . . . . . . . . . . . . . . . . . . . . . 44
4.7 Sound Speed in Two Phase Medium . . . . . . . . . . . . . . . . . . 45

5 Isentropic Flow 49
5.1 Stagnation State for Ideal Gas Model . . . . . . . . . . . . . . . . . . 49
5.1.1 General Relationship . . . . . . . . . . . . . . . . . . . . . . . 49
5.1.2 Relationships for Small Mach Number . . . . . . . . . . . . . 52
5.2 Isentropic Converging-Diverging Flow in Cross Section . . . . . . . . 53
5.2.1 The Properties in the Adiabatic Nozzle . . . . . . . . . . . . . 54
5.2.2 Isentropic Flow Examples . . . . . . . . . . . . . . . . . . . . 58
5.2.3 Mass Flow Rate (Number) . . . . . . . . . . . . . . . . . . . 61
5.3 Isentropic Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.3.1 Isentropic Isothermal Flow Nozzle . . . . . . . . . . . . . . . 72
5.3.2 General Relationship . . . . . . . . . . . . . . . . . . . . . . . 72
5.4 The Impulse Function . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.4.1 Impulse in Isentropic Adiabatic Nozzle . . . . . . . . . . . . 79
5.4.2 The Impulse Function in Isothermal Nozzle . . . . . . . . . . 82
5.5 Isothermal Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.6 The effects of Real Gases . . . . . . . . . . . . . . . . . . . . . . . . 84

6 Normal Shock 89
6.1 Solution of the Governing Equations . . . . . . . . . . . . . . . . . . 92
6.1.1 Informal Model . . . . . . . . . . . . . . . . . . . . . . . . . . 92
6.1.2 Formal Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
6.1.3 Prandtl’s Condition . . . . . . . . . . . . . . . . . . . . . . . . 96
6.2 Operating Equations and Analysis . . . . . . . . . . . . . . . . . . . 97
6.2.1 The Limitations of the Shock Wave . . . . . . . . . . . . . . . 98
6.2.2 Small Perturbation Solution . . . . . . . . . . . . . . . . . . . 98
6.2.3 Shock Thickness . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.2.4 Shock or Wave Drag . . . . . . . . . . . . . . . . . . . . . . . 99
6.3 The Moving Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.3.1 Shock or Wave Drag Result from a Moving Shock . . . . . . 103
6.3.2 Shock Result from a Sudden and Complete Stop . . . . . . . 105
6.3.3 Moving Shock into Stationary Medium (Suddenly Open Valve) 108
6.3.4 Partially Open Valve . . . . . . . . . . . . . . . . . . . . . . . 117
6.3.5 Partially Closed Valve . . . . . . . . . . . . . . . . . . . . . . 118
6.3.6 Worked–out Examples for Shock Dynamics . . . . . . . . . . 119
6.4 Shock Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.5 Shock with Real Gases . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.6 Shock in Wet Steam . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.7 Normal Shock in Ducts . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.8 More Examples for Moving Shocks . . . . . . . . . . . . . . . . . . . 129
6.9 Tables of Normal Shocks, k = 1.4 Ideal Gas . . . . . . . . . . . . . . 132
vi CONTENTS

7 Normal Shock in Variable Duct Areas 139

7.1 Nozzle efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
7.2 Diffuser Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

8 Nozzle Flow With External Forces 151

8.1 Isentropic Nozzle (Q = 0) . . . . . . . . . . . . . . . . . . . . . . . . 152
8.2 Isothermal Nozzle (T = constant) . . . . . . . . . . . . . . . . . . . 154

9 Isothermal Flow 155

9.1 The Control Volume Analysis/Governing equations . . . . . . . . . . 156
9.2 Dimensionless Representation . . . . . . . . . . . . . . . . . . . . . 156
9.3 The Entrance Limitation of Supersonic Branch . . . . . . . . . . . . 161
9.4 Comparison with Incompressible Flow . . . . . . . . . . . . . . . . . 162
9.5 Supersonic Branch . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
9.6 Figures and Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
9.7 Isothermal Flow Examples . . . . . . . . . . . . . . . . . . . . . . . . 165
9.8 Unchoked situations in Fanno Flow . . . . . . . . . . . . . . . . . . . 170

10 Fanno Flow 175

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
10.2 Fanno Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
10.3 Non–Dimensionalization of the Equations . . . . . . . . . . . . . . . 177
10.4 The Mechanics and Why the Flow is Choked? . . . . . . . . . . . . . 180
10.5 The Working Equations . . . . . . . . . . . . . . . . . . . . . . . . . 181
10.6 Examples of Fanno Flow . . . . . . . . . . . . . . . . . . . . . . . . . 185
10.7 Supersonic Branch . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
10.8 Maximum Length for the Supersonic Flow . . . . . . . . . . . . . . . 190
10.9 Working Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
10.9.1 Variations of The Tube Length ( 4fDL ) Effects . . . . . . . . . . 192
10.9.2 The Pressure Ratio, P P1 , effects . . . . . . . . . . . . . . .
2
. . 197
10.9.3 Entrance Mach number, M1 , effects . . . . . . . . . . . . . . 199
10.10Practical Examples for Subsonic Flow . . . . . . . . . . . . . . . . . 206
10.10.1Subsonic Fanno Flow for Given 4fDL and Pressure Ratio . . . 206
10.10.2Subsonic Fanno Flow for a Given M1 and Pressure Ratio . . 208
10.11The Approximation of the Fanno Flow by Isothermal Flow . . . . . . 211
10.12More Examples of Fanno Flow . . . . . . . . . . . . . . . . . . . . . 211
10.13The Table for Fanno Flow . . . . . . . . . . . . . . . . . . . . . . . . 213
10.14Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

11 Rayleigh Flow 217

11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
11.2 Governing Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
11.3 Rayleigh Flow Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
11.4 Examples For Rayleigh Flow . . . . . . . . . . . . . . . . . . . . . . 223
CONTENTS vii

12 Evacuating SemiRigid Chambers 231

12.1 Governing Equations and Assumptions . . . . . . . . . . . . . . . . 232
12.2 General Model and Non-dimensioned . . . . . . . . . . . . . . . . . 234
12.2.1 Isentropic Process . . . . . . . . . . . . . . . . . . . . . . . . 236
12.2.2 Isothermal Process in The Chamber . . . . . . . . . . . . . . 236
12.2.3 A Note on the Entrance Mach number . . . . . . . . . . . . . 236
12.3 Rigid Tank with Nozzle . . . . . . . . . . . . . . . . . . . . . . . . . . 237
12.3.1 Adiabatic Isentropic Nozzle Attached . . . . . . . . . . . . . . 237
12.3.2 Isothermal Nozzle Attached . . . . . . . . . . . . . . . . . . . 239
12.4 Rapid evacuating of a rigid tank . . . . . . . . . . . . . . . . . . . . 239
12.4.1 With Fanno Flow . . . . . . . . . . . . . . . . . . . . . . . . . 239
12.4.2 Filling Process . . . . . . . . . . . . . . . . . . . . . . . . . . 241
12.4.3 The Isothermal Process . . . . . . . . . . . . . . . . . . . . . 242
12.4.4 Simple Semi Rigid Chamber . . . . . . . . . . . . . . . . . . 243
12.4.5 The “Simple” General Case . . . . . . . . . . . . . . . . . . . 243
12.5 Advance Topics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245

13 Evacuating under External Volume Control 247

13.1 General Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
13.1.1 Rapid Process . . . . . . . . . . . . . . . . . . . . . . . . . . 248
13.1.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
13.1.3 Direct Connection . . . . . . . . . . . . . . . . . . . . . . . . 251
13.2 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252

14 Oblique Shock 255

14.1 Preface to Oblique Shock . . . . . . . . . . . . . . . . . . . . . . . . 255
14.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
14.2.1 Introduction to Oblique Shock . . . . . . . . . . . . . . . . . . 256
14.2.2 Introduction to Prandtl–Meyer Function . . . . . . . . . . . . 256
14.2.3 Introduction to Zero Inclination . . . . . . . . . . . . . . . . . 257
14.3 Oblique Shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
14.4 Solution of Mach Angle . . . . . . . . . . . . . . . . . . . . . . . . . 260
14.4.1 Upstream Mach Number, M1 , and Deflection Angle, δ . . . . 260
14.4.2 When No Oblique Shock Exist or When D > 0 . . . . . . . . 263
14.4.3 Upstream Mach Number, M1 , and Shock Angle, θ . . . . . . 271
14.4.4 Given Two Angles, δ and θ . . . . . . . . . . . . . . . . . . . 273
14.4.5 Flow in a Semi–2D Shape . . . . . . . . . . . . . . . . . . . . 274
14.4.6 Small δ “Weak Oblique shock” . . . . . . . . . . . . . . . . . 276
14.4.7 Close and Far Views of the Oblique Shock . . . . . . . . . . 277
14.4.8 Maximum Value of Oblique shock . . . . . . . . . . . . . . . . 277
14.5 Detached Shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
14.5.1 Issues Related to the Maximum Deflection Angle . . . . . . . 279
14.5.2 Oblique Shock Examples . . . . . . . . . . . . . . . . . . . . 281
14.5.3 Application of Oblique Shock . . . . . . . . . . . . . . . . . . 283
14.5.4 Optimization of Suction Section Design . . . . . . . . . . . . 294
viii CONTENTS

14.5.5 Retouch of Shock or Wave Drag . . . . . . . . . . . . . . . . 294

14.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
14.7 Appendix: Oblique Shock Stability Analysis . . . . . . . . . . . . . . 296

15 Prandtl-Meyer Function 299

15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
15.2 Geometrical Explanation . . . . . . . . . . . . . . . . . . . . . . . . . 300
15.2.1 Alternative Approach to Governing Equations . . . . . . . . . 301
15.2.2 Comparison And Limitations between the Two Approaches . 305
15.3 The Maximum Turning Angle . . . . . . . . . . . . . . . . . . . . . . 305
15.4 The Working Equations for the Prandtl-Meyer Function . . . . . . . . 306
15.5 d’Alembert’s Paradox . . . . . . . . . . . . . . . . . . . . . . . . . . 306
15.6 Flat Body with an Angle of Attack . . . . . . . . . . . . . . . . . . . . 308
15.7 Examples For Prandtl–Meyer Function . . . . . . . . . . . . . . . . 308
15.8 Combination of the Oblique Shock and Isentropic Expansion . . . . 311

A Computer Program 315

A.1 About the Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
A.2 Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
A.3 Program listings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

Index 319
Subjects Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
Authors Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
 LIST OF FIGURES

1.1 The shock as a connection of Fanno and Rayleigh lines . . . . . . . 7

1.2 The schematic of deLavel’s turbine . . . . . . . . . . . . . . . . . . . 9
1.3 The measured pressure in a nozzle . . . . . . . . . . . . . . . . . . 11
1.4 Flow rate as a function of the back pressure . . . . . . . . . . . . . . 12
1.5 Portrait of Galileo Galilei . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.6 Photo of Ernest Mach . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.7 The photo of thebullet in a supersonic flow not taken in a wind tunnel 17
1.8 Photo of Lord Rayleigh . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.9 Portrait of Rankine . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.10 The photo of Gino Fanno approximately in 1950 . . . . . . . . . . . 20
1.11 Photo of Prandtl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.12 The photo of Ernst Rudolf George Eckert with the author’s family . . 22

4.1 A very slow moving piston in a still gas . . . . . . . . . . . . . . . . . 36

4.2 Stationary sound wave and gas moves relative to the pulse. . . . . . 36
4.3 The Compressibility Chart . . . . . . . . . . . . . . . . . . . . . . . . 40

5.1 Flow thorough a converging diverging nozzle . . . . . . . . . . . . . 49

5.2 Perfect gas flows through a tube . . . . . . . . . . . . . . . . . . . . 51
5.3 The stagnation properties as a function of the Mach number, k = 1.4 52
5.4 Control volume inside a converging-diverging nozzle. . . . . . . . . . 54
5.5 The relationship between the cross section and the Mach number . 58
5.6 Various ratios as a function of Mach number for isothermal Nozzle . 75
5.7 The comparison of nozzle flow . . . . . . . . . . . . . . . . . . . . . 76
5.8 Comparison of the pressure and temperature drop (two scales) . . . 77
5.9 Schematic to explain the significances of the Impulse function . . . . 80
5.10 Schematic of a flow thorough a nozzle example (5.8) . . . . . . . . . 81

ix
x LIST OF FIGURES

6.1 A shock wave inside a tube . . . . . . . . . . . . . . . . . . . . . . . 89

6.2 The intersection of Fanno flow and Rayleigh flow . . . . . . . . . . . 91
6.3 The Mexit and P0 as a function Mupstream . . . . . . . . . . . . . . . 95
6.4 The ratios of the static properties of the two sides of the shock. . . . 97
6.5 The shock drag diagram . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.6 Comparison between stationary shock and moving shock . . . . . . 101
6.7 The shock drag diagram for moving shock. . . . . . . . . . . . . . . 103
6.8 The diagram for the common explanation for shock drag. . . . . . . . 104
6.9 Comparison between a stationary shock and a moving shock in a stationary medium
6.10 Comparison between a stationary shock and a moving shock in a stationary medium
6.11 The moving shock a result of a sudden stop . . . . . . . . . . . . . . 107
6.12 A shock as a result of a sudden Opening . . . . . . . . . . . . . . . 108
6.13 The number of iterations to achieve convergence. . . . . . . . . . . . 109
6.14 Schematic of showing the piston pushing air. . . . . . . . . . . . . . 111
6.15 Time the pressure at the nozzle for the French problem. . . . . . . . 113
6.16 Max Mach number as a function of k. . . . . . . . . . . . . . . . . . 113
6.17 Time the pressure at the nozzle for the French problem. . . . . . . . 117
6.18 Moving shock as a result of valve opening . . . . . . . . . . . . . . . 117
6.19 The results of the partial opening of the valve. . . . . . . . . . . . . . 118
6.20 A shock as a result of partially a valve closing . . . . . . . . . . . . . 119
6.21 Schematic of a piston pushing air in a tube. . . . . . . . . . . . . . . 122
6.22 Figure for Example (6.10) . . . . . . . . . . . . . . . . . . . . . . . . 124
6.23 The shock tube schematic with a pressure ”diagram.” . . . . . . . . . 125
6.24 Figure for Example (6.13) . . . . . . . . . . . . . . . . . . . . . . . . 129
6.25 The results for Example (6.13) . . . . . . . . . . . . . . . . . . . . . 130
6.26 Figure for example (6.13) . . . . . . . . . . . . . . . . . . . . . . . . 130
6.27 The results for Example (6.13) . . . . . . . . . . . . . . . . . . . . . 131

7.1 The flow in the nozzle with different back pressures. . . . . . . . . . 139
7.2 A nozzle with normal shock . . . . . . . . . . . . . . . . . . . . . . . 140
7.3 Description to clarify the definition of diffuser efficiency . . . . . . . . 146
7.4 Schematic of a supersonic tunnel example(7.3) . . . . . . . . . . . . 146

9.1 Control volume for isothermal flow . . . . . . . . . . . . . . . . . . . 155

9.2 Working relationships for isothermal flow . . . . . . . . . . . . . . . . 161
9.3 The entrance Mach for isothermal flow for 4fDL . . . . . . . . . . . . 172

10.1 Control volume of the gas flow in a constant cross section . . . . . . 175
10.2 Various parameters in Fanno flow as a function of Mach number . . 184
10.3 Schematic of Example (10.1) . . . . . . . . . . . . . . . . . . . . . . 185
10.4 The schematic of Example (10.2) . . . . . . . . . . . . . . . . . . . . 186
10.5 The maximum length as a function of specific heat, k . . . . . . . . . 191
10.6 The effects of increase of 4fDL on the Fanno line . . . . . . . . . . . 192
10.7 The development properties in of converging nozzle . . . . . . . . . 193
10.8 Min and ṁ as a function of the 4fDL . . . . . . . . . . . . . . . . . . . 194
LIST OF FIGURES xi

10.9 M1 as a function M2 for various 4fDL . . . . . . . . . . . . . . . . . . 195

10.10M1 as a function M2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
10.11The pressure distribution as a function of 4fDL for a short 4fDL . . . . 198
10.12The pressure distribution as a function of 4fDL for a long 4fDL . . . . 199
10.13The effects of pressure variations on Mach number profile . . . . . . 200
10.14Mach number as a function of 4fDL when the total 4fDL = 0.3 . . . . . 201
10.15Schematic of a “long” tube in supersonic branch . . . . . . . . . . . 202
10.16The extra tube length as a function of the shock location . . . . . . . 202
10.17The maximum entrance Mach number as a function of 4fDL . . . . . 203
10.18Unchoked flow calculations showing the hypothetical “full” tub when choked206
10.19The results of the algorithm showing the conversion rate. . . . . . . 208
10.20Solution to a missing diameter . . . . . . . . . . . . . . . . . . . . . 210
10.21M1 as a function of 4fDL comparison with Isothermal Flow . . . . . . 212
10.22“Moody” diagram on the name Moody who netscape H. Rouse work to claim as his own. In this section the turbulent are

11.1 The control volume of Rayleigh Flow . . . . . . . . . . . . . . . . . . 217

11.2 The temperature entropy diagram for Rayleigh line . . . . . . . . . . 219
11.3 The basic functions of Rayleigh Flow (k=1.4) . . . . . . . . . . . . . 224
11.4 Schematic of the combustion chamber. . . . . . . . . . . . . . . . . 228

12.1 The two different classifications of models . . . . . . . . . . . . . . . 231

12.2 A schematic of two possible . . . . . . . . . . . . . . . . . . . . . . . 232
12.3 A schematic of the control volumes used in this model . . . . . . . . 232
12.4 The pressure assumptions in the chamber and tube entrance . . . . 233
12.5 The reduced time as a function of the modified reduced pressure . . 241
12.6 The reduced time as a function of the modified reduced pressure . . 242

13.1 The control volume of the “Cylinder”. . . . . . . . . . . . . . . . . . . 248

13.2 The pressure ratio as a function of the dimensionless time . . . . . . 253
13.3 P̄ as a function of t̄ for choked condition . . . . . . . . . . . . . . . . 254
13.4 The pressure ratio as a function of the dimensionless time . . . . . 254

14.1 A view of a normal shock as a limited case for oblique shock . . . . 255
14.2 The oblique shock or Prandtl–Meyer function regions . . . . . . . . . 256
14.3 A typical oblique shock schematic . . . . . . . . . . . . . . . . . . . 257
14.4 Flow around spherically blunted 30◦ cone-cylinder . . . . . . . . . . 263
14.5 The different views of a large inclination angle . . . . . . . . . . . . . 264
14.6 The three different Mach numbers . . . . . . . . . . . . . . . . . . . 265
14.7 The various coefficients of three different Mach numbers . . . . . . . 269
14.8 The “imaginary” Mach waves at zero inclination. . . . . . . . . . . . 270
14.9 The D, shock angle, and My for M1 = 3 . . . . . . . . . . . . . . . . 271
14.10The possible range of solutions . . . . . . . . . . . . . . . . . . . . . 273
14.11Two Dimensional Wedge . . . . . . . . . . . . . . . . . . . . . . . . . 274
14.12Schematic of finite wedge with zero angle of attack. . . . . . . . . . 275
14.13/; A local and a far view of the oblique shock. . . . . . . . . . . . . . 277
xii LIST OF FIGURES

14.14The schematic for a round–tip bullet in a supersonic flow. . . . . . . 278

14.15The schematic for a symmetrical suction section with Mach reflection. 279
14.16 The “detached” shock in a complicated configuration . . . . . . . . 280
14.17 Oblique shock around a cone . . . . . . . . . . . . . . . . . . . . . 281
14.18 Maximum values of the properties in an oblique shock . . . . . . . 282
14.19 Two variations of inlet suction for supersonic flow. . . . . . . . . . . 283
14.20 Schematic for Example (14.5). . . . . . . . . . . . . . . . . . . . . . 283
14.21 Schematic for Example (14.6). . . . . . . . . . . . . . . . . . . . . . 285
14.22 Schematic of two angles turn with two weak shocks. . . . . . . . . 285
14.23Revisiting of shock drag diagram for the oblique shock. . . . . . . . . 294
14.24 Typical examples of unstable and stable situations. . . . . . . . . . 296
14.25The schematic of stability analysis for oblique shock. . . . . . . . . . 297

15.1 The definition of the angle for the Prandtl–Meyer function. . . . . . . 299
15.2 The angles of the Mach line triangle . . . . . . . . . . . . . . . . . . 299
15.3 The schematic of the turning flow. . . . . . . . . . . . . . . . . . . . 300
15.4 The mathematical coordinate description . . . . . . . . . . . . . . . 301
15.5 Prandtl-Meyer function after the maximum angle . . . . . . . . . . . 306
15.7 Diamond shape for supersonic d’Alembert’s Paradox . . . . . . . . . 306
15.6 The angle as a function of the Mach number . . . . . . . . . . . . . 307
15.8 The definition of the angle for the Prandtl–Meyer function. . . . . . . 308
15.9 The schematic of Example 15.1 . . . . . . . . . . . . . . . . . . . . . 308
15.10 The schematic for the reversed question of example (15.2) . . . . 310
15.11Schematic of the nozzle and Prandtle–Meyer expansion. . . . . . . . 312

A.1 Schematic diagram that explains the structure of the program . . . . 318
 LIST OF TABLES

1 Books Under Potto Project . . . . . . . . . . . . . . . . . . . . . . . . xxxix

1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xl

2.1 Properties of Various Ideal Gases [300K] . . . . . . . . . . . . . . . 30

4.1 Water speed of sound from different sources . . . . . . . . . . . . . 43

4.2 Liquids speed of sound . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.3 Solids speed of sound . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5.1 Fliegner’s number a function of Mach number . . . . . . . . . . . . . 66

5.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.2 Isentropic Table k = 1.4 . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.3 Isothermal Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.3 Isothermal Table (continue) . . . . . . . . . . . . . . . . . . . . . . . 83

6.1 The shock wave table for k = 1.4 . . . . . . . . . . . . . . . . . . . . 132

6.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
6.2 Table for a Reflective Shock suddenly closed valve . . . . . . . . . . 133
6.2 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.3 Table for shock suddenly opened valve (k=1.4) . . . . . . . . . . . . 134
6.3 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.3 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.4 Table for shock from a suddenly opened valve (k=1.3) . . . . . . . . 136
6.4 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

9.1 The Isothermal Flow basic parameters . . . . . . . . . . . . . . . . 165

9.2 The flow parameters for unchoked flow . . . . . . . . . . . . . . . . 170

xiii
xiv LIST OF TABLES

9.2 The flow parameters for unchoked flow (continue) . . . . . . . . . . 171

10.1 Fanno Flow Standard basic Table . . . . . . . . . . . . . . . . . . . 213

10.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

11.1 Rayleigh Flow k=1.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . 222

11.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

14.1 Table of maximum values of the oblique Shock k=1.4 . . . . . . . . 277

14.1 continue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
 NOMENCLATURE

R̄ Universal gas constant, see equation (2.26), page 29

` Units length., see equation (2.1), page 25

ρ Density of the fluid, see equation (4.1), page 36

B bulk modulus, see equation (4.35), page 43

Bf Body force, see equation (2.9), page 27

c Speed of sound, see equation (4.1), page 36

Cp Specific pressure heat, see equation (2.23), page 29

Cv Specific volume heat, see equation (2.22), page 29

E Young’s modulus, see equation (4.37), page 44

EU Internal energy, see equation (2.3), page 26

Eu Internal Energy per unit mass, see equation (2.6), page 26

Ei System energy at state i, see equation (2.2), page 26

H Enthalpy, see equation (2.18), page 28

h Specific enthalpy, see equation (2.18), page 28

k the ratio of the specific heats, see equation (2.24), page 29

M Mach number, see equation (5.8), page 50

n The poletropic coefficient, see equation (4.32), page 42

xv
xvi LIST OF TABLES

P Pressure, see equation (4.3), page 36

q Energy per unit mass, see equation (2.6), page 26
Q12 The energy transfered to the system between state 1 and state 2, see equa-
tion (2.2), page 26
R Specific gas constant, see equation (2.27), page 30

Rmix The universal gas constant for mixture, see equation (4.48), page 46

S Entropy of the system, see equation (2.13), page 28

t Time, see equation (4.15), page 39

U velocity , see equation (2.4), page 26

w Work per unit mass, see equation (2.6), page 26
W12 The work done by the system between state 1 and state 2, see equa-
tion (2.2), page 26

z The compressibility factor, see equation (4.19), page 39

 The Book Change Log

Version 0.4.8.5rc
On 31st December 2008 (3.3M pp. 380)
• Add Gary Settles’s color image in wedge shock and an example.
• Improve the wrap figure issue to oblique shock.
• Add Moody diagram to Fanno flow.
• English corrections to the oblique shock chapter.

Version 0.4.8.4
On 7th October 2008 (3.2M pp. 376)
• More work on the nomenclature issue.
• Important equations and useful equations issues inserted.
• Expand the discussion on the friction factor in isothermal and fanno flow.

Version 0.4.8.3
On 17th September 2008 (3.1M pp. 369)
• Started the nomenclature issue so far only the thermodynamics chapter.
• Started the important equations and useful equations issue.
• Add the introduction to thermodynamics chapter.

xvii
xviii LIST OF TABLES

• Add the discussion on the friction factor in isothermal and fanno flow.

Version 0.4.8.2
On 25th January 2008 (3.1M pp. 353)
• Add several additions to the isentropic flow, normal shock,

• Rayleigh Flow.

• Improve some examples.

• More changes to the script to generate separate chapters sections.

• Add new macros to work better so that php and pdf version will be similar.

• More English revisions.

Version 0.4.8
November-05-2007
• Add the new unchoked subsonic Fanno Flow section which include the “un-
known” diameter question.

• Shock (Wave) drag explanation with example.

• Some examples were add and fixing other examples (small perturbations of
oblique shock).

• Minor English revisions.

Version 0.4.4.3pr1
July-10-2007
• Improvement of the pdf version provide links.

Version 0.4.4.2a
July-4-2007 version
• Major English revisions in Rayleigh Flow Chapter.

• Continue the improvement of the HTML version (imageonly issues).

• Minor content changes and addition of an example.

LIST OF TABLES xix

Version 0.4.4.2
May-22-2007 version
• Major English revisions.

• Continue the improvement of the HTML version.

• Minor content change and addition of an example.

Version 0.4.4.1
Feb-21-2007 version
• Include the indexes subjects and authors.

• Continue the improve the HTML version.

• solve problems with some of the figures location (float problems)

• Improve some spelling and grammar.

• Minor content change and addition of an example.

• The main change is the inclusion of the indexes (subject and authors). There
were some additions to the content which include an example. The ”naughty
professor’s questions” section isn’t completed and is waiting for interface
of Potto-GDC to be finished (engine is finished, hopefully next two weeks).
Some grammar and misspelling corrections were added.
Now include a script that append a title page to every pdf fraction of the book
(it was fun to solve this one). Continue to insert the changes (log) to every
source file (latex) of the book when applicable. This change allows to follow
the progression of the book. Most the tables now have the double formatting
one for the html and one for the hard copies.

Version 0.4.4pr1
Jan-16-2007 version
• Major modifications of the source to improve the HTML version.

• Add the naughty professor’s questions in the isentropic chapter.

• Some grammar and miss spelling corrections.

xx LIST OF TABLES

Version 0.4.3.2rc1
Dec-04-2006 version
• Add new algorithm for Fanno Flow calculation of the shock location in the
supersonic flow for given fld (exceeding Max) and M1 (see the example).

• Minor addition in the Sound and History chapters.

• Add analytical expression for Mach number results of piston movement.

Version 0.4.3.1rc4 aka 0.4.3.1

Nov-10-2006 aka Roy Tate’s version
For this release (the vast majority) of the grammatical corrections are due to Roy
Tate

• Grammatical corrections through the history chapter and part of the sound
chapter.

• Very minor addition in the Isothermal chapter about supersonic branch.

Version 0.4.3.1rc3
Oct-30-2006
• Add the solutions to last three examples in Chapter Normal Shock in variable
area.

• Improve the discussion about partial open and close moving shock dynamics
i.e. high speed running into slower velocity

• Clean other tables and figure and layout.

Version 0.4.3rc2
Oct-20-2006
• Clean up of the isentropic and sound chapters

• Add discussion about partial open and close moving shock dynamics i.e.
high speed running into slower velocity.

• Add the partial moving shock figures (never published before)

LIST OF TABLES xxi

Version 0.4.3rc1
Sep-20-2006
• Change the book’s format to 6x9 from letter paper
• Clean up of the isentropic chapter.
• Add the shock tube section
• Generalize the discussion of the the moving shock (not including the change
in the specific heat (material))
• Add the Impulse Function for Isothermal Nozzle section
• Improve the discussion of the Fliegner’s equation
• Add the moving shock table (never published before)

Version 0.4.1.9 (aka 0.4.1.9rc2)

May-22-2006
• Added the Impulse Function
• Add two examples.
• Clean some discussions issues .

Version 0.4.1.9rc1
May-17-2006
• Added mathematical description of Prandtl-Meyer’s Function
• Fixed several examples in oblique shock chapter
• Add three examples.
• Clean some discussions issues .

Version 0.4.1.8 aka Version 0.4.1.8rc3

May-03-2006
• Added Chapman’s function
• Fixed several examples in oblique shock chapter
• Add two examples.
• Clean some discussions issues .
xxii LIST OF TABLES

Version 0.4.1.8rc2
Apr-11-2006
• Added the Maximum Deflection Mach number’s equation
• Added several examples to oblique shock
 Notice of Copyright For This
Document:

This document published Modified FDL. The change of the license is to prevent
from situations where the author has to buy his own book. The Potto Project Li-
cense isn’t long apply to this document and associated docoments.

GNU Free Documentation License

The modification is that under section 3 “copying in quantity” should be add in the
end.
”If you print more than 200 copies, you are required that you furnish the author with
two (2) copies of the printed book.”

Version 1.2, November 2002

Copyright ©2000,2001,2002 Free Software Foundation, Inc.

51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA

Everyone is permitted to copy and distribute verbatim copies of this license

document, but changing it is not allowed.

Preamble

The purpose of this License is to make a manual, textbook, or other

functional and useful document ”free” in the sense of freedom: to assure everyone
the effective freedom to copy and redistribute it, with or without modifying it, either

xxiii
xxiv LIST OF TABLES

commercially or noncommercially. Secondarily, this License preserves for the au-

thor and publisher a way to get credit for their work, while not being considered
responsible for modifications made by others.
This License is a kind of ”copyleft”, which means that derivative works
of the document must themselves be free in the same sense. It complements the
GNU General Public License, which is a copyleft license designed for free software.
We have designed this License in order to use it for manuals for free
software, because free software needs free documentation: a free program should
come with manuals providing the same freedoms that the software does. But this
License is not limited to software manuals; it can be used for any textual work,
regardless of subject matter or whether it is published as a printed book. We
recommend this License principally for works whose purpose is instruction or ref-
erence.

1. APPLICABILITY AND DEFINITIONS

This License applies to any manual or other work, in any medium, that
contains a notice placed by the copyright holder saying it can be distributed under
the terms of this License. Such a notice grants a world-wide, royalty-free license,
unlimited in duration, to use that work under the conditions stated herein. The
”Document”, below, refers to any such manual or work. Any member of the public
is a licensee, and is addressed as ”you”. You accept the license if you copy,
modify or distribute the work in a way requiring permission under copyright law.
A ”Modified Version” of the Document means any work containing the
Document or a portion of it, either copied verbatim, or with modifications and/or
translated into another language.
A ”Secondary Section” is a named appendix or a front-matter section
of the Document that deals exclusively with the relationship of the publishers or
authors of the Document to the Document’s overall subject (or to related matters)
and contains nothing that could fall directly within that overall subject. (Thus, if
the Document is in part a textbook of mathematics, a Secondary Section may not
explain any mathematics.) The relationship could be a matter of historical connec-
tion with the subject or with related matters, or of legal, commercial, philosophical,
ethical or political position regarding them.
The ”Invariant Sections” are certain Secondary Sections whose titles
are designated, as being those of Invariant Sections, in the notice that says that
the Document is released under this License. If a section does not fit the above
definition of Secondary then it is not allowed to be designated as Invariant. The
Document may contain zero Invariant Sections. If the Document does not identify
any Invariant Sections then there are none.
The ”Cover Texts” are certain short passages of text that are listed, as
Front-Cover Texts or Back-Cover Texts, in the notice that says that the Document
is released under this License. A Front-Cover Text may be at most 5 words, and a
Back-Cover Text may be at most 25 words.
GNU FREE DOCUMENTATION LICENSE xxv

A ”Transparent” copy of the Document means a machine-readable

copy, represented in a format whose specification is available to the general public,
that is suitable for revising the document straightforwardly with generic text editors
or (for images composed of pixels) generic paint programs or (for drawings) some
widely available drawing editor, and that is suitable for input to text formatters or
for automatic translation to a variety of formats suitable for input to text formatters.
A copy made in an otherwise Transparent file format whose markup, or absence
of markup, has been arranged to thwart or discourage subsequent modification
by readers is not Transparent. An image format is not Transparent if used for any
substantial amount of text. A copy that is not ”Transparent” is called ”Opaque”.
Examples of suitable formats for Transparent copies include plain ASCII
without markup, Texinfo input format, LaTeX input format, SGML or XML using a
publicly available DTD, and standard-conforming simple HTML, PostScript or PDF
designed for human modification. Examples of transparent image formats include
PNG, XCF and JPG. Opaque formats include proprietary formats that can be read
and edited only by proprietary word processors, SGML or XML for which the DTD
and/or processing tools are not generally available, and the machine-generated
HTML, PostScript or PDF produced by some word processors for output purposes
only.
The ”Title Page” means, for a printed book, the title page itself, plus
such following pages as are needed to hold, legibly, the material this License re-
quires to appear in the title page. For works in formats which do not have any title
page as such, ”Title Page” means the text near the most prominent appearance of
the work’s title, preceding the beginning of the body of the text.
A section ”Entitled XYZ” means a named subunit of the Document
whose title either is precisely XYZ or contains XYZ in parentheses following text
that translates XYZ in another language. (Here XYZ stands for a specific section
name mentioned below, such as ”Acknowledgements”, ”Dedications”, ”En-
dorsements”, or ”History”.) To ”Preserve the Title” of such a section when you
modify the Document means that it remains a section ”Entitled XYZ” according to
this definition.
The Document may include Warranty Disclaimers next to the notice which
states that this License applies to the Document. These Warranty Disclaimers are
considered to be included by reference in this License, but only as regards dis-
claiming warranties: any other implication that these Warranty Disclaimers may
have is void and has no effect on the meaning of this License.

2. VERBATIM COPYING
You may copy and distribute the Document in any medium, either com-
mercially or noncommercially, provided that this License, the copyright notices, and
the license notice saying this License applies to the Document are reproduced in
all copies, and that you add no other conditions whatsoever to those of this Li-
cense. You may not use technical measures to obstruct or control the reading or
further copying of the copies you make or distribute. However, you may accept
xxvi LIST OF TABLES

compensation in exchange for copies. If you distribute a large enough number of

copies you must also follow the conditions in section 3.
You may also lend copies, under the same conditions stated above, and
you may publicly display copies.

3. COPYING IN QUANTITY
If you publish printed copies (or copies in media that commonly have
printed covers) of the Document, numbering more than 100, and the Document’s
license notice requires Cover Texts, you must enclose the copies in covers that
carry, clearly and legibly, all these Cover Texts: Front-Cover Texts on the front
cover, and Back-Cover Texts on the back cover. Both covers must also clearly
and legibly identify you as the publisher of these copies. The front cover must
present the full title with all words of the title equally prominent and visible. You
may add other material on the covers in addition. Copying with changes limited to
the covers, as long as they preserve the title of the Document and satisfy these
conditions, can be treated as verbatim copying in other respects.
If the required texts for either cover are too voluminous to fit legibly, you
should put the first ones listed (as many as fit reasonably) on the actual cover, and
continue the rest onto adjacent pages.
If you publish or distribute Opaque copies of the Document number-
ing more than 100, you must either include a machine-readable Transparent copy
along with each Opaque copy, or state in or with each Opaque copy a computer-
network location from which the general network-using public has access to down-
load using public-standard network protocols a complete Transparent copy of the
Document, free of added material. If you use the latter option, you must take rea-
sonably prudent steps, when you begin distribution of Opaque copies in quantity, to
ensure that this Transparent copy will remain thus accessible at the stated location
until at least one year after the last time you distribute an Opaque copy (directly or
through your agents or retailers) of that edition to the public.
It is requested, but not required, that you contact the authors of the Doc-
ument well before redistributing any large number of copies, to give them a chance
to provide you with an updated version of the Document.

4. MODIFICATIONS
You may copy and distribute a Modified Version of the Document under
the conditions of sections 2 and 3 above, provided that you release the Modified
Version under precisely this License, with the Modified Version filling the role of
the Document, thus licensing distribution and modification of the Modified Version
to whoever possesses a copy of it. In addition, you must do these things in the
Modified Version:

A. Use in the Title Page (and on the covers, if any) a title distinct from that of
the Document, and from those of previous versions (which should, if there
were any, be listed in the History section of the Document). You may use the
GNU FREE DOCUMENTATION LICENSE xxvii

same title as a previous version if the original publisher of that version gives
permission.
B. List on the Title Page, as authors, one or more persons or entities responsible
for authorship of the modifications in the Modified Version, together with at
least five of the principal authors of the Document (all of its principal authors,
if it has fewer than five), unless they release you from this requirement.
C. State on the Title page the name of the publisher of the Modified Version, as
the publisher.
D. Preserve all the copyright notices of the Document.
E. Add an appropriate copyright notice for your modifications adjacent to the
other copyright notices.
F. Include, immediately after the copyright notices, a license notice giving the
public permission to use the Modified Version under the terms of this License,
in the form shown in the Addendum below.
G. Preserve in that license notice the full lists of Invariant Sections and required
Cover Texts given in the Document’s license notice.
H. Include an unaltered copy of this License.
I. Preserve the section Entitled ”History”, Preserve its Title, and add to it an
item stating at least the title, year, new authors, and publisher of the Modified
Version as given on the Title Page. If there is no section Entitled ”History”
in the Document, create one stating the title, year, authors, and publisher of
the Document as given on its Title Page, then add an item describing the
Modified Version as stated in the previous sentence.
J. Preserve the network location, if any, given in the Document for public access
to a Transparent copy of the Document, and likewise the network locations
given in the Document for previous versions it was based on. These may be
placed in the ”History” section. You may omit a network location for a work
that was published at least four years before the Document itself, or if the
original publisher of the version it refers to gives permission.
K. For any section Entitled ”Acknowledgements” or ”Dedications”, Preserve the
Title of the section, and preserve in the section all the substance and tone of
each of the contributor acknowledgements and/or dedications given therein.
L. Preserve all the Invariant Sections of the Document, unaltered in their text
and in their titles. Section numbers or the equivalent are not considered part
of the section titles.
M. Delete any section Entitled ”Endorsements”. Such a section may not be in-
cluded in the Modified Version.
xxviii LIST OF TABLES

N. Do not retitle any existing section to be Entitled ”Endorsements” or to conflict

in title with any Invariant Section.

O. Preserve any Warranty Disclaimers.

If the Modified Version includes new front-matter sections or appendices

that qualify as Secondary Sections and contain no material copied from the Docu-
ment, you may at your option designate some or all of these sections as invariant.
To do this, add their titles to the list of Invariant Sections in the Modified Version’s
license notice. These titles must be distinct from any other section titles.
You may add a section Entitled ”Endorsements”, provided it contains
nothing but endorsements of your Modified Version by various parties–for example,
statements of peer review or that the text has been approved by an organization
as the authoritative definition of a standard.
You may add a passage of up to five words as a Front-Cover Text, and a
passage of up to 25 words as a Back-Cover Text, to the end of the list of Cover Texts
in the Modified Version. Only one passage of Front-Cover Text and one of Back-
Cover Text may be added by (or through arrangements made by) any one entity. If
the Document already includes a cover text for the same cover, previously added
by you or by arrangement made by the same entity you are acting on behalf of, you
may not add another; but you may replace the old one, on explicit permission from
the previous publisher that added the old one.
The author(s) and publisher(s) of the Document do not by this License
give permission to use their names for publicity for or to assert or imply endorse-
ment of any Modified Version.

5. COMBINING DOCUMENTS
You may combine the Document with other documents released under
this License, under the terms defined in section 4 above for modified versions,
provided that you include in the combination all of the Invariant Sections of all
of the original documents, unmodified, and list them all as Invariant Sections of
your combined work in its license notice, and that you preserve all their Warranty
Disclaimers.
The combined work need only contain one copy of this License, and
multiple identical Invariant Sections may be replaced with a single copy. If there
are multiple Invariant Sections with the same name but different contents, make
the title of each such section unique by adding at the end of it, in parentheses, the
name of the original author or publisher of that section if known, or else a unique
number. Make the same adjustment to the section titles in the list of Invariant
Sections in the license notice of the combined work.
In the combination, you must combine any sections Entitled ”History” in
the various original documents, forming one section Entitled ”History”; likewise
combine any sections Entitled ”Acknowledgements”, and any sections Entitled
”Dedications”. You must delete all sections Entitled ”Endorsements”.
GNU FREE DOCUMENTATION LICENSE xxix

6. COLLECTIONS OF DOCUMENTS
You may make a collection consisting of the Document and other doc-
uments released under this License, and replace the individual copies of this Li-
cense in the various documents with a single copy that is included in the collection,
provided that you follow the rules of this License for verbatim copying of each of
the documents in all other respects.
You may extract a single document from such a collection, and distribute
it individually under this License, provided you insert a copy of this License into
the extracted document, and follow this License in all other respects regarding
verbatim copying of that document.

7. AGGREGATION WITH INDEPENDENT WORKS

A compilation of the Document or its derivatives with other separate and
independent documents or works, in or on a volume of a storage or distribution
medium, is called an ”aggregate” if the copyright resulting from the compilation
is not used to limit the legal rights of the compilation’s users beyond what the
individual works permit. When the Document is included in an aggregate, this Li-
cense does not apply to the other works in the aggregate which are not themselves
derivative works of the Document.
If the Cover Text requirement of section 3 is applicable to these copies of
the Document, then if the Document is less than one half of the entire aggregate,
the Document’s Cover Texts may be placed on covers that bracket the Document
within the aggregate, or the electronic equivalent of covers if the Document is in
electronic form. Otherwise they must appear on printed covers that bracket the
whole aggregate.

8. TRANSLATION
Translation is considered a kind of modification, so you may distribute
translations of the Document under the terms of section 4. Replacing Invariant
Sections with translations requires special permission from their copyright holders,
but you may include translations of some or all Invariant Sections in addition to
the original versions of these Invariant Sections. You may include a translation of
this License, and all the license notices in the Document, and any Warranty Dis-
claimers, provided that you also include the original English version of this License
and the original versions of those notices and disclaimers. In case of a disagree-
ment between the translation and the original version of this License or a notice or
disclaimer, the original version will prevail.
If a section in the Document is Entitled ”Acknowledgements”, ”Dedica-
tions”, or ”History”, the requirement (section 4) to Preserve its Title (section 1) will
typically require changing the actual title.

9. TERMINATION
xxx LIST OF TABLES

You may not copy, modify, sublicense, or distribute the Document except
as expressly provided for under this License. Any other attempt to copy, modify,
sublicense or distribute the Document is void, and will automatically terminate your
rights under this License. However, parties who have received copies, or rights,
from you under this License will not have their licenses terminated so long as such
parties remain in full compliance.
10. FUTURE REVISIONS OF THIS LICENSE
The Free Software Foundation may publish new, revised versions of the
GNU Free Documentation License from time to time. Such new versions will be
similar in spirit to the present version, but may differ in detail to address new prob-
lems or concerns. See http://www.gnu.org/copyleft/.
Each version of the License is given a distinguishing version number. If
the Document specifies that a particular numbered version of this License ”or any
later version” applies to it, you have the option of following the terms and conditions
either of that specified version or of any later version that has been published (not
as a draft) by the Free Software Foundation. If the Document does not specify a
version number of this License, you may choose any version ever published (not
as a draft) by the Free Software Foundation.
ADDENDUM: How to use this License for your documents
To use this License in a document you have written, include a copy of
the License in the document and put the following copyright and license notices
just after the title page:

Copyright ©YEAR YOUR NAME. Permission is granted to copy, dis-

tribute and/or modify this document under the terms of the GNU Free
Documentation License, Version 1.2 or any later version published by
the Free Software Foundation; with no Invariant Sections, no Front-
Cover Texts, and no Back-Cover Texts. A copy of the license is included
in the section entitled ”GNU Free Documentation License”.

If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts,

replace the ”with...Texts.” line with this:

with the Invariant Sections being LIST THEIR TITLES, with the Front-
Cover Texts being LIST, and with the Back-Cover Texts being LIST.

If you have Invariant Sections without Cover Texts, or some other com-
bination of the three, merge those two alternatives to suit the situation.
If your document contains nontrivial examples of program code, we rec-
ommend releasing these examples in parallel under your choice of free software
license, such as the GNU General Public License, to permit their use in free soft-
ware.
 CONTRIBUTOR LIST

How to contribute to this book

As a copylefted work, this book is open to revision and expansion by any interested
parties. The only ”catch” is that credit must be given where credit is due. This is a
copyrighted work: it is not in the public domain!
If you wish to cite portions of this book in a work of your own, you must
follow the same guidelines as for any other GDL copyrighted work.

Credits
All entries arranged in alphabetical order of surname. Major contributions are listed
by individual name with some detail on the nature of the contribution(s), date, con-
tact info, etc. Minor contributions (typo corrections, etc.) are listed by name only for
reasons of brevity. Please understand that when I classify a contribution as ”minor,”
it is in no way inferior to the effort or value of a ”major” contribution, just smaller in
the sense of less text changed. Any and all contributions are gratefully accepted. I
am indebted to all those who have given freely of their own knowledge, time, and
resources to make this a better book!

• Date(s) of contribution(s): 2004 to present

• Nature of contribution: Original author.

• Contact at: [email protected]

John Martones
• Date(s) of contribution(s): June 2005

xxxi
xxxii LIST OF TABLES

• Nature of contribution: HTML formatting, some error corrections.

Grigory Toker
• Date(s) of contribution(s): August 2005

• Nature of contribution: Provided pictures of the oblique shock for oblique

shock chapter.

Ralph Menikoff
• Date(s) of contribution(s): July 2005

• Nature of contribution: Some discussions about the solution to oblique

shock and about the Maximum Deflection of the oblique shock.

Domitien Rataaforret
• Date(s) of contribution(s): Oct 2006

• Nature of contribution: Some discussions about the French problem and

help with the new wrapImg command.

Gary Settles
• Date(s) of contribution(s): Dec 2008

• Nature of contribution: Four images for oblique shock two dimensional, and
cone flow.

Your name here

• Date(s) of contribution(s): Month and year of contribution

• Nature of contribution: Insert text here, describing how you contributed to

the book.

• Contact at: my [email protected]

CREDITS xxxiii

Typo corrections and other ”minor” contributions

• H. Gohrah, Ph. D., September 2005, some LaTeX issues.
• Roy Tate November 2006, Suggestions on improving English and grammar.
• Nancy Cohen 2006, Suggestions on improving English and style for various
issues.
• Irene Tan 2006, proof reading many chapters and for various other issues.
xxxiv LIST OF TABLES
 About This Author

Genick Bar-Meir holds a Ph.D. in Mechanical Engineering from University of Min-

nesota and a Master in Fluid Mechanics from Tel Aviv University. Dr. Bar-Meir was
the last student of the late Dr. R.G.E. Eckert. Much of his time has been spend
doing research in the field of heat and mass transfer (related to renewal energy
issues) and this includes fluid mechanics related to manufacturing processes and
design. Currently, he spends time writing books (there are already three very pop-
ular books) and softwares for the POTTO project (see Potto Prologue). The author
enjoys to encourage his students to understand the material beyond the basic re-
quirements of exams.
In his early part of his professional life, Bar-Meir was mainly interested in
elegant models whether they have or not a practical applicability. Now, this author’s
views had changed and the virtue of the practical part of any model becomes the
essential part of his ideas, books and software.
He developed models for Mass Transfer in high concentration that be-
came a building blocks for many other models. These models are based on an-
alytical solution to a family of equations1 . As the change in the view occurred,
Bar-Meir developed models that explained several manufacturing processes such
the rapid evacuation of gas from containers, the critical piston velocity in a par-
tially filled chamber (related to hydraulic jump), application of supply and demand
to rapid change power system and etc. All the models have practical applicability.
These models have been extended by several research groups (needless to say
with large research grants). For example, the Spanish Comision Interministerial
provides grants TAP97-0489 and PB98-0007, and the CICYT and the European
Commission provides 1FD97-2333 grants for minor aspects of that models. More-
over, the author’s models were used in numerical works, in GM, British industry,
Spain, and Canada.
1 Where the mathematicians were able only to prove that the solution exists.

xxxv
xxxvi LIST OF TABLES

In the area of compressible flow, it was commonly believed and taught

that there is only weak and strong shock and it is continue by Prandtl–Meyer func-
tion. Bar–Meir discovered the analytical solution for oblique shock and showed that
there is a quiet buffer between the oblique shock and Prandtl–Meyer. He also build
analytical solution to several moving shock cases. He described and categorized
the filling and evacuating of chamber by compressible fluid in which he also found
analytical solutions to cases where the working fluid was ideal gas. The common
explanation to Prandtl–Meyer function shows that flow can turn in a sharp cor-
ner. Engineers have constructed design that based on this conclusion. Bar-Meir
demonstrated that common Prandtl–Meyer explanation violates the conservation
of mass and therefor the turn must be around a finite radius. The author’s explana-
tions on missing diameter and other issues in fanno flow and ““naughty professor’s
question”” are used in the industry.
In his book “Basics of Fluid Mechanics”, Bar-Meir demonstrated that flu-
ids must have wavy surface when the materials flow together. All the previous
models for the flooding phenomenon did not have a physical explanation to the
dryness. He built a model to explain the flooding problem (two phase flow) based
on the physics. He also constructed and explained many new categories for two
flow regimes.
The author lives with his wife and three children. A past project of his was
building a four stories house, practically from scratch. While he writes his programs
and does other computer chores, he often feels clueless about computers and
programing. While he is known to look like he knows about many things, the author
just know to learn quickly. The author spent years working on the sea (ships) as a
engine sea officer but now the author prefers to remain on solid ground.
 Prologue For The POTTO Project

This books series was born out of frustrations in two respects. The first issue is
the enormous price of college textbooks. It is unacceptable that the price of the
college books will be over $150 per book (over 10 hours of work for an average
student in The United States).
The second issue that prompted the writing of this book is the fact that
we as the public have to deal with a corrupted judicial system. As individuals we
have to obey the law, particularly the copyright law with the “infinite2 ” time with the
copyright holders. However, when applied to “small” individuals who are not able
to hire a large legal firm, judges simply manufacture facts to make the little guy
lose and pay for the defense of his work. On one hand, the corrupted court system
defends the “big” guys and on the other hand, punishes the small “entrepreneur”
who tries to defend his or her work. It has become very clear to the author and
founder of the POTTO Project that this situation must be stopped. Hence, the
creation of the POTTO Project. As R. Kook, one of this author’s sages, said instead
of whining about arrogance and incorrectness, one should increase wisdom. This
project is to increase wisdom and humility.
The POTTO Project has far greater goals than simply correcting an abu-
sive Judicial system or simply exposing abusive judges. It is apparent that writing
textbooks especially for college students as a cooperation, like an open source,
is a new idea3 . Writing a book in the technical field is not the same as writing a
novel. The writing of a technical book is really a collection of information and prac-
tice. There is always someone who can add to the book. The study of technical
2 After the last decision of the Supreme Court in the case of Eldred v. Ashcroff (see

http://cyber.law.harvard.edu/openlaw/eldredvashcroft for more information) copyrights prac-

tically remain indefinitely with the holder (not the creator).
3 In some sense one can view the encyclopedia Wikipedia as an open content project (see

http://en.wikipedia.org/wiki/Main Page). The wikipedia is an excellent collection of articles which

are written by various individuals.

xxxvii
xxxviii LIST OF TABLES

material isn’t only done by having to memorize the material, but also by coming to
understand and be able to solve related problems. The author has not found any
technique that is more useful for this purpose than practicing the solving of prob-
lems and exercises. One can be successful when one solves as many problems
as possible. To reach this possibility the collective book idea was created/adapted.
While one can be as creative as possible, there are always others who can see
new aspects of or add to the material. The collective material is much richer than
any single person can create by himself.
The following example explains this point: The army ant is a kind of
carnivorous ant that lives and hunts in the tropics, hunting animals that are even
up to a hundred kilograms in weight. The secret of the ants’ power lies in their
collective intelligence. While a single ant is not intelligent enough to attack and hunt
large prey, the collective power of their networking creates an extremely powerful
intelligence to carry out this attack4 . When an insect which is blind can be so
powerful by networking, So can we in creating textbooks by this powerful tool.
Why would someone volunteer to be an author or organizer of such a
book? This is the first question the undersigned was asked. The answer varies
from individual to individual. It is hoped that because of the open nature of these
books, they will become the most popular books and the most read books in their
respected field. For example, the books on compressible flow and die casting be-
came the most popular books in their respective area. In a way, the popularity of
the books should be one of the incentives for potential contributors. The desire
to be an author of a well–known book (at least in his/her profession) will convince
some to put forth the effort. For some authors, the reason is the pure fun of writing
and organizing educational material. Experience has shown that in explaining to
others any given subject, one also begins to better understand the material. Thus,
contributing to these books will help one to understand the material better. For
others, the writing of or contributing to this kind of books will serve as a social
function. The social function can have at least two components. One component
is to come to know and socialize with many in the profession. For others the social
part is as simple as a desire to reduce the price of college textbooks, especially
for family members or relatives and those students lacking funds. For some con-
tributors/authors, in the course of their teaching they have found that the textbook
they were using contains sections that can be improved or that are not as good as
their own notes. In these cases, they now have an opportunity to put their notes
to use for others. Whatever the reasons, the undersigned believes that personal
intentions are appropriate and are the author’s/organizer’s private affair.
If a contributor of a section in such a book can be easily identified, then
that contributor will be the copyright holder of that specific section (even within
question/answer sections). The book’s contributor’s names could be written by
their sections. It is not just for experts to contribute, but also students who hap-
pened to be doing their homework. The student’s contributions can be done by
4 see also in Franks, Nigel R.; ”Army Ants: A Collective Intelligence,” American Scientist, 77:139,

1989 (see for information http://www.ex.ac.uk/bugclub/raiders.html)

CREDITS xxxix

adding a question and perhaps the solution. Thus, this method is expected to
accelerate the creation of these high quality books.
These books are written in a similar manner to the open source software
process. Someone has to write the skeleton and hopefully others will add “flesh
and skin.” In this process, chapters or sections can be added after the skeleton has
been written. It is also hoped that others will contribute to the question and answer
sections in the book. But more than that, other books contain data5 which can be
typeset in LATEX. These data (tables, graphs and etc.) can be redone by anyone
who has the time to do it. Thus, the contributions to books can be done by many
who are not experts. Additionally, contributions can be made from any part of the
world by those who wish to translate the book.
It is hoped that the books will be error-free. Nevertheless, some errors
are possible and expected. Even if not complete, better discussions or better ex-
planations are all welcome to these books. These books are intended to be “con-
tinuous” in the sense that there will be someone who will maintain and improve the
books with time (the organizer(s)).
These books should be considered more as a project than to fit the tradi-
tional definition of “plain” books. Thus, the traditional role of author will be replaced
by an organizer who will be the one to compile the book. The organizer of the book
in some instances will be the main author of the work, while in other cases only
the gate keeper. This may merely be the person who decides what will go into the
book and what will not (gate keeper). Unlike a regular book, these works will have
a version number because they are alive and continuously evolving.
The undersigned of this document intends to be the organizer–author–
coordinator of the projects in the following areas:

Table -1: Books under development in Potto project.

DownLoads

Availability
Number
ss

Project Remarks Version for

re
og

Name Public
Pr

Download
Compressible Flow beta 0.4.8.4 4 120,000
Die Casting alpha 0.1 4 60,000
Dynamics NSY 0.0.0 6 -
Fluid Mechanics alpha 0.1.8 4 15,000
Heat Transfer NSY Based 0.0.0 6 -
on
Eckert
Mechanics NSY 0.0.0 6 -
Open Channel NSY 0.0.0 6 -
Flow
5 Data are not copyrighted.
xl LIST OF TABLES
Table -1: Books under development in Potto project. (continue)

DownLoads
Availability

Number
ss
Project Remarks Version for

re
og
Name Public

Pr
Download
Statics early first 0.0.1 6 -
alpha chapter
Strength of Material NSY 0.0.0 6 -
Thermodynamics early 0.0.01 6 -
alpha
Two/Multi phases NSY Tel- 0.0.0 6 -
flow Aviv’notes

NSY = Not Started Yet

The meaning of the progress is as:
• The Alpha Stage is when some of the chapters are already in a rough draft;
• in Beta Stage is when all or almost all of the chapters have been written and
are at least in a draft stage;
• in Gamma Stage is when all the chapters are written and some of the chap-
ters are in a mature form; and
• the Advanced Stage is when all of the basic material is written and all that is
left are aspects that are active, advanced topics, and special cases.
The mature stage of a chapter is when all or nearly all the sections are in a mature
stage and have a mature bibliography as well as numerous examples for every
section. The mature stage of a section is when all of the topics in the section
are written, and all of the examples and data (tables, figures, etc.) are already pre-
sented. While some terms are defined in a relatively clear fashion, other definitions
give merely a hint on the status. But such a thing is hard to define and should be
enough for this stage.
The idea that a book can be created as a project has mushroomed from
the open source software concept, but it has roots in the way science progresses.
However, traditionally books have been improved by the same author(s), a process
in which books have a new version every a few years. There are book(s) that
have continued after their author passed away, i.e., the Boundary Layer Theory
originated6 by Hermann Schlichting but continues to this day. However, projects
such as the Linux Documentation project demonstrated that books can be written
as the cooperative effort of many individuals, many of whom volunteered to help.
6 Originally authored by Dr. Schlichting, who passed way some years ago. A new version is created

every several years.

CREDITS xli

Writing a textbook is comprised of many aspects, which include the ac-

tual writing of the text, writing examples, creating diagrams and figures, and writing
the LATEX macros7 which will put the text into an attractive format. These chores can
be done independently from each other and by more than one individual. Again,
because of the open nature of this project, pieces of material and data can be used
by different books.

7 One can only expect that open source and readable format will be used for this project. But more

than that, only LATEX, and perhaps troff, have the ability to produce the quality that one expects for these
writings. The text processes, especially LATEX, are the only ones which have a cross platform ability to
produce macros and a uniform feel and quality. Word processors, such as OpenOffice, Abiword, and
Microsoft Word software, are not appropriate for these projects. Further, any text that is produced by
Microsoft and kept in “Microsoft” format are against the spirit of this project In that they force spending
money on Microsoft software.
xlii LIST OF TABLES
 Prologue For This Book

Version 0.4.8 Jan. 23, 2008

It is more than a year ago, when the previous this section was modified. Many
things have changed, and more people got involved. It nice to know that over
70,000 copies have been download from over 130 countries. It is more pleasant
to find that this book is used in many universities around the world, also in many
institutes like NASA (a tip from Dr. Farassat, NASA ”to educate their “young scien-
tist, and engineers”) and others. Looking back, it must be realized that while, this
book is the best in many areas, like oblique shock, moving shock, fanno flow, etc
there are missing some sections, like methods of characteristics, and the introduc-
tory sections (fluid mechanics, and thermodynamics). Potto–GDC is much more
mature and it is changing from “advance look up” to a real gas dynamics calculator
(for example, calculation of unchoked Fanno Flow). Today Potto–GDC has the only
capability to produce the oblique shock figure. Potto-GDC is becoming the major
educational educational tool in gas dynamics. To kill two birds in one stone, one,
continuous requests from many and, two, fill the introductory section on fluid me-
chanics in this book this area is major efforts in the next few months for creating
the version 0.2 of the “Basic of Fluid Mechanics” are underway.

Version 0.4.3 Sep. 15, 2006

The title of this section is change to reflect that it moved to beginning of the book.
While it moves earlier but the name was not changed. Dr. Menikoff pointed to this
inconsistency, and the author is apologizing for this omission.
Several sections were add to this book with many new ideas for example
on the moving shock tables. However, this author cannot add all the things that he
was asked and want to the book in instant fashion. For example, one of the reader

xliii
xliv LIST OF TABLES

ask why not one of the example of oblique shock was not turn into the explanation
of von Neumann paradox. The author was asked by a former client why he didn’t
insert his improved tank filling and evacuating models (the addition of the energy
equation instead of isentropic model). While all these requests are important, the
time is limited and they will be inserted as time permitted.
The moving shock issues are not completed and more work is needed
also in the shock tube. Nevertheless, the ideas of moving shock will reduced the
work for many student of compressible flow. For example solving homework prob-
lem from other text books became either just two mouse clicks away or just looking
at that the tables in this book. I also got request from a India to write the interface
for Microsoft. I am sorry will not be entertaining work for non Linux/Unix systems,
especially for Microsoft. If one want to use the software engine it is okay and
permitted by the license of this work.
The download to this mount is over 25,000.

Version 0.4.2
It was surprising to find that over 14,000 downloaded and is encouraging to receive
over 200 thank you eMail (only one from U.S.A./Arizona) and some other reactions.
This textbook has sections which are cutting edge research8 .
The additions of this version focus mainly on the oblique shock and re-
lated issues as results of questions and reactions on this topic. However, most
readers reached to www.potto.org by searching for either terms “Rayleigh flow”
(107) and “Fanno flow” ((93). If the total combined variation search of terms
“Fanno” and “Rayleigh” (mostly through google) is accounted, it reaches to about
30% (2011). This indicates that these topics are highly is demanded and not many
concerned with the shock phenomena as this author believed and expected. Thus,
most additions of the next version will be concentrated on Fanno flow and Rayleigh
flow. The only exception is the addition to Taylor–Maccoll flow (axisymmetricale
conical flow) in Prandtl–Meyer function (currently in a note form).
Furthermore, the questions that appear on the net will guide this author
on what is really need to be in a compressible flow book. At this time, several
questions were about compressibility factor and two phase flow in Fanno flow and
other kind of flow models. The other questions that appeared related two phase
and connecting several chambers to each other. Also, an individual asked whether
this author intended to write about the unsteady section, and hopefully it will be
near future.

8 A reader asked this author to examine a paper on Triple Shock Entropy Theorem and Its Conse-

quences by Le Roy F. Henderson and Ralph Menikoff. This led to comparison between maximum to
ideal gas model to more general model.
VERSION 0.4 xlv

Version 0.4
Since the last version (0.3) several individuals sent me remarks and suggestions.
In the introductory chapter, extensive description of the compressible flow history
was written. In the chapter on speed of sound, the two phase aspects were added.
The isothermal nozzle was combined with the isentropic chapter. Some examples
were added to the normal shock chapter. The fifth chapter deals now with normal
shock in variable area ducts. The sixth chapter deals with external forces fields.
The chapter about oblique shock was added and it contains the analytical solution.
At this stage, the connection between Prandtl–Meyer flow and oblique is an note
form. The a brief chapter on Prandtl–Meyer flow was added.

Version 0.3
In the traditional class of compressible flow it is assumed that the students will be
aerospace engineers or dealing mostly with construction of airplanes and turbo-
machinery. This premise should not be assumed. This assumption drives students
from other fields away from this knowledge. This knowledge should be spread to
other fields because it needed there as well. This “rejection” is especially true when
students feel that they have to go through a “shock wave” in their understanding.
This book is the second book in the series of POTTO project books.
POTTO project books are open content textbooks. The reason the topic of Com-
pressible Flow was chosen, while relatively simple topics like fundamentals of
strength of material were delayed, is because of the realization that manufacture
engineering simply lacks fundamental knowledge in this area and thus produces
faulty designs and understanding of major processes. Unfortunately, the under-
signed observed that many researchers who are dealing with manufacturing pro-
cesses are lack of understanding about fluid mechanics in general but particularly
in relationship to compressible flow. In fact one of the reasons that many manufac-
turing jobs are moving to other countries is because of the lack of understanding
of fluid mechanics in general and compressible in particular. For example, the lack
of competitive advantage moves many of the die casting operations to off shore9 .
It is clear that an understanding of Compressible Flow is very important for areas
that traditionally have ignored the knowledge of this topic10 .
As many instructors can recall from their time as undergraduates, there
were classes during which most students had a period of confusion, and then
later, when the dust settled, almost suddenly things became clear. This situation
is typical also for Compressible Flow classes, especially for external compressible
flow (e.g. flow around a wing, etc.). This book offers a more balanced emphasis
which focuses more on internal compressible flow than the traditional classes. The
9 Please read the undersigned’s book “Fundamentals of Die Casting Design,” which demonstrates
how ridiculous design and research can be.
10 The fundamental misunderstanding of choking results in poor models (research) in the area of die

casting, which in turn results in many bankrupt companies and the movement of the die casting industry
to offshore.
xlvi LIST OF TABLES

internal flow topics seem to be common for the “traditional” students and students
from other fields, e.g., manufacturing engineering.
This book is written in the spirit of my adviser and mentor E.R.G. Eckert.
Who, aside from his research activity, wrote the book that brought a revolution in
the heat transfer field of education. Up to Eckert’s book, the study of heat transfer
was without any dimensional analysis. He wrote his book because he realized that
the dimensional analysis utilized by him and his adviser (for the post doc), Ernst
Schmidt, and their colleagues, must be taught in engineering classes. His book
met strong criticism in which some called to burn his book. Today, however, there
is no known place in world that does not teach according to Eckert’s doctrine. It is
assumed that the same kind of individuals who criticized Eckert’s work will criticize
this work. This criticism will not change the future or the success of the ideas in
this work. As a wise person says “don’t tell me that it is wrong, show me what is
wrong”; this is the only reply. With all the above, it must be emphasized that this
book will not revolutionize the field even though considerable new materials that
have never been published are included. Instead, it will provide a new emphasis
and new angle to Gas Dynamics.
Compressible flow is essentially different from incompressible flow in
mainly two respects: discontinuity (shock wave) and choked flow. The other is-
sues, while important, are not that crucial to the understanding of the unique phe-
nomena of compressible flow. These unique issues of compressible flow are to
be emphasized and shown. Their applicability to real world processes is to be
demonstrated11 .
The book is organized into several chapters which, as a traditional text-
book, deals with a basic introduction of thermodynamics concepts (under construc-
tion). The second chapter deals with speed of sound. The third chapter provides
the first example of choked flow (isentropic flow in a variable area). The fourth
chapter deals with a simple case of discontinuity (a simple shock wave in a noz-
zle). The next chapter is dealing with isothermal flow with and without external
forces (the moving of the choking point), again under construction. The next three
chapters are dealing with three models of choked flow: Isothermal flow12 , Fanno
flow and Rayleigh flow. First, the Isothermal flow is introduced because of the rel-
ative ease of the analytical treatment. Isothermal flow provides useful tools for the
pipe systems design. These chapters are presented almost independently. Every
chapter can be “ripped” out and printed independently. The topics of filling and
evacuating of gaseous chambers are presented, normally missed from traditional
textbooks. There are two advanced topics which included here: oblique shock
wave, and properties change effects (ideal gases and real gases) (under construc-
tion). In the oblique shock, for the first time analytical solution is presented, which
is excellent tool to explain the strong, weak and unrealistic shocks. The chapter on
one-dimensional unsteady state, is currently under construction.
The last chapter deals with the computer program, Gas Dynamics Cal-
11 If you have better and different examples or presentations you are welcome to submit them.
12 It is suggested to referred to this model as Shapiro flow
VERSION 0.3 xlvii

culator (CDC-POTTO). The program design and how to use the program are de-
scribed (briefly).
Discussions on the flow around bodies (wing, etc), and Prandtl–Meyer
expansion will be included only after the gamma version unless someone will pro-
vide discussion(s) (a skeleton) on these topics.
It is hoped that this book will serve the purposes that was envisioned
for the book. It is further hoped that others will contribute to this book and find
additional use for this book and enclosed software.
xlviii LIST OF TABLES
 How This Book Was Written

This book started because I needed an explanation for manufacturing engineers.

Apparently many manufacturing engineers and even some researchers in manu-
facturing engineering were lack of understanding about fluid mechanics in particu-
larly about compressible flow. Therefore, I wrote to myself some notes and I con-
verted one of the note to a chapter in my first book, “Fundamentals Of Die Casting
Design.” Later, I realized that people need down to earth book about compressible
flow and this book was born.
The free/open content of the book was created because the realization
that open content accelerated the creation of books and reaction to the corruption
of the court implementing the copyright law by manufacturing facts and laws. It
was farther extended by the allegation of free market and yet the academic educa-
tion cost is sky rocketing without a real reason and real competition. There is no
reason why a textbook which cost at the very most 10$ to publish/produce to cost
about 150 dollars. If a community will pull together, the best books can be created.
Anyone can be part of it. For example, even my 10 years old son, Eliezer made me
change the chapter on isothermal flow. He made me realized that the common ap-
proach to supersonic branch of isothermal as non–existent is the wrong approach.
It should be included because this section provides the explanation and direction
on what Fanno flow model will approach if heat transfer is taken into account13 .
I realized that books in compressible flow are written in a form that is hard
for non fluid mechanic engineer to understand. Therefore, this book is designed
to be in such form that is easy to understand. I wrote notes and asked myself
what materials should be included in such a book so when I provide consultation
to a company, I do not need to explain the fundamentals. Therefore, there are
some chapters in this book which are original materials never published before.
The presentation of some of the chapters is different from other books. The book
13 Still in untyped note form.

xlix
l LIST OF TABLES

does not provide the old style graphical solution methods yet provide the graphical
explanation of things.
Of course, this book was written on Linux (MicrosoftLess book). This
book was written using the vim editor for editing (sorry never was able to be com-
fortable with emacs). The graphics were done by TGIF, the best graphic program
that this author experienced so far. The old figures where done by grap (part the
old Troff). Unfortunately, I did not have any access to grap and switched to Grace.
Grace is a problematic program. Finally, the gle is replacing the old grace. So far, it
seems much better choice and from version 0.4.8 all will be done using GLE. The
spell checking was done by gaspell, a program that cannot be used on new system
and I had to keep my old Linux to make it work14 . I hope someone will write a new
spell check so I can switch to a new system.
The figure in cover page was created by Michael Petschauer, graphic
designer, and is open/free content copyright by him ( happy [email protected]).

14 If you would like to to help me to write a new spell check user interface, please contact me.
 About Gas Dynamics Calculator

Gas Dynamic Calculator, (Potto–GDC) was created to generate various tables for
the book either at end the chapters or for the exercises. This calculator was given
to several individuals and they found Potto–GDC to be very useful. So, I decided
to include Potto–GDC to the book.
Initially, the Potto-GDC was many small programs for specific tasks. For
example, the stagnation table was one such program. Later, the code became a
new program to find the root of something between the values of the tables e.g.
finding parameters for a given 4fDL . At that stage, the program changed to contain
a primitive interface to provide parameters to carry out the proper calculations. Yet,
then, every flow model was a different program.
When it become cumbersome to handle several programs, the author
utilized the object oriented feature of C++ and assigned functions to the common
tasks to a base class and the specific applications to the derived classes. Later,
a need to intermediate stage of tube flow model (the PipeFlow class) was created
and new classes were created.
The graphical interface was created only after the engine was written.
The graphical interface was written to provide a filter for the unfamiliar user. It also
remove the need to recompile the code every time.

Version 0.5
In this version the main point was on the bugs fixing but also add the results can
be shown in a HTML code. In fanno flow, many problems of unchoked Fanno flow
now possible to solve (by one click).

li
lii LIST OF TABLES

Version 0.4.3
This version add several features among them is the shock dynamics calculations
with the iteration info. The last feature is good for homework either for the students
or the instructors.

Version 0.4.1.7
Version 4.1.7 had several bug fixes and add two angle calculations to the oblique
shock. Change the logtable to tabular environment for short tables.
 Preface
‘‘In the beginning, the POTTO project was without
form, and void; and emptiness was upon the face of the
bits and files. And the Fingers of the Author moved
upon the face of the keyboard. And the Author said,
Let there be words, and there were words.’’ 15 .

This book, Fundamentals of Compressible Flow, describes the funda-

mentals of compressible flow phenomena for engineers and others. This book is
designed to replace the book(s) or instructor’s notes for the compressible flow in
(mostly) undergraduate classes for engineering/science students. It is hoped that
the book could be used as a reference book for people who have at least some
knowledge of the basics of fundamental fluid mechanics, and basic science such
as calculus, physics, etc. It is hoped that the computer program enclosed in the
book will take on a life of its own and develop into an open content or source
project.
The structure of this book is such that many of the chapters could be
usable independently. For example, if you need information about, say, Fanno
flow, you can read just chapter 10. I hope this makes the book easier to use as a
reference manual. However, this manuscript is first and foremost a textbook, and
secondly a reference manual only as a lucky coincidence.
I have tried to describe why the theories are the way they are, rather than
just listing “seven easy steps” for each task. This means that a lot of information
is presented which is not necessary for everyone. These explanations have been
marked as such and can be skipped.16 Reading everything will, naturally, increase
your understanding of the fundamentals of compressible fluid flow.
This book is written and maintained on a volunteer basis. Like all vol-
unteer work, there is a limit on how much effort I was able to put into the book
and its organization. Moreover, due to the fact that English is my third language
and time limitations, the explanations are not as good as if I had a few years to
perfect them. Nevertheless, I believe professionals working in many engineering
15 Tothe power and glory of the mighty God. This book is only to explain his power.
16 Atthe present, the book is not well organized. You have to remember that this book is a work in
progress.

liii
liv LIST OF TABLES

fields will benefit from this information. This book contains many original models,
and explanations never published before.
I have left some issues which have unsatisfactory explanations in the
book, marked with a Mata mark. I hope to improve or to add to these areas in the
near future. Furthermore, I hope that many others will participate of this project and
will contribute to this book (even small contributions such as providing examples or
editing mistakes are needed).
I have tried to make this text of the highest quality possible and am in-
terested in your comments and ideas on how to make it better. Incorrect language,
errors, ideas for new areas to cover, rewritten sections, more fundamental material,
more mathematics (or less mathematics); I am interested in it all. If you want to be
involved in the editing, graphic design, or proofreading, please drop me a line. You
may contact me via Email at “[email protected]”.
Naturally, this book contains material that never was published before.
This material never went through a peer review. While peer review and publication
in a professional publication is excellent idea in theory. In practice, this process
leaves a large room to blockage of novel ideas and plagiarism. If you would like
be “peer reviews” or critic to my new ideas please send me your idea(s). Even
reaction/comments from individuals like David Marshall17
Several people have helped me with this book, directly or indirectly. I
would like to especially thank to my adviser, Dr. E. R. G. Eckert, whose work was
the inspiration for this book. I also would like to thank Amy Ross for her advice
ideas, and assistance.
The symbol META was added to provide typographical conventions to
blurb as needed. This is mostly for the author’s purposes and also for your amuse-
ment. There are also notes in the margin, but those are solely for the author’s pur-
poses, ignore them please. They will be removed gradually as the version number
advances.
I encourage anyone with a penchant for writing, editing, graphic ability,
LATEX knowledge, and material knowledge and a desire to provide open content
textbooks and to improve them to join me in this project. If you have Internet e-mail
access, you can contact me at “[email protected]”.

17 Dr. Marshall wrote to this author that the author should review other people work before he write

any thing new (well, literature review is always good?). Over ten individuals wrote me about this letter.
I am asking from everyone to assume that his reaction was innocent one. While his comment looks like
unpleasant reaction, it brought or cause the expansion the oblique shock chapter. However, other email
that imply that someone will take care of this author aren’t appreciated.
 To Do List and Road Map

This book is not complete and probably never will be completed. There will always
new problems to add or to polish the explanations or include more new materials.
Also issues that associated with the book like the software has to be improved. It
is hoped the changes in TEX and LATEX related to this book in future will be min-
imal and minor. It is hoped that the style file will be converged to the final form
rapidly. Nevertheless, there are specific issues which are on the “table” and they
are described herein.
At this stage, several chapters are missing. The effects of the deviations
from the ideal gas model on the properties should be included. Further topics
related to non-ideal gas such as steam and various freons are in the process of
being added to this book especially in relationship to Fanno flow.
One of the virtue of this book lay in the fact that it contains a software that
is extensible. For example, the Fanno module can be extended to include effects
of real gases. This part will be incorporated in the future hopefully with the help of
others.
Specific missing parts from every chapters are discussed below. These
omissions, mistakes, approach problems are sometime appears in the book under
the Meta simple like this

Meta
sample this part.

Meta End
Questions/problems appear as a marginal note. On occasions a footnote was
used to point out for a need of improvement. You are always welcome to add a
new material: problem, question, illustration or photo of experiment. Material can

lv
lvi LIST OF TABLES

be further illuminate. Additional material can be provided to give a different angle

on the issue at hand.

Speed of Sound
Discussion about the movement in medium with variation in speed of sound. This
concept in relation of the wind tunnel and atmosphere with varied density and
temperature.
Mixed gases and liquids.
More problems in relationship to two phase.
Speed of sound in wet steam.

Stagnation effects
Extend the applicability with examples
Cp as a function of temperature (deviation of ideal gas model)
“real gas”’ like water vapor
History – on the teaching (for example when the concept of stagnation was first
taught.

Nozzle
The effect of external forces (add problems).
Real gases effects (only temperature effects)
Flow with “tabulated gases” calculations
Phase change and two phase flow (multi choking points) effects (after 1.0 version).
The dimensional analysis of the flow when the flow can be considered as isother-
mal.
The combined effects of isentropic nozzle with heat transfer (especially with rela-
tionship to the program.).

Normal Shock
Extend the partially (open/close) moving shock theory.
Provide more examples on the previous topic
Shock in real gases like water vapor
Shock in (partially) two phase gases like air with dust particles

Isothermal Flow
Classification of Problems
Comparison of results with Fanno flow
Pipes Network calculations.
VERSION 0.4.1.7 lvii

Fanno Flow
More examples: various categories
Some improvement on the software (clean up)
Real gas effects (compressible factor)
Tabulated gas

Rayleigh Flow
To mature the chapter: discussion on the “dark” corners of this model.
Provide discussion on variations of the effecting parameters.
Examples: provide categorization

Evacuation and filling semi rigid Chambers

To construct the Rayleigh flow in the tube (thermal chocking)
Energy equation (non isentropic process)
Examples classifications
Software (converting the FORTRAN program to c++)

Evacuating and filling chambers under external forces

Comparison with chemical reaction case
Energy equation (non isentropic process)
Examples
Software transformation from FORTRAN to c++. The FORTRAN version will not
be included.

Oblique Shock
Add application to design problems
Real Gas effects

Prandtl–Meyer
The limitations (Prandtl-Meyer). Application
Marcell–Taylor (from the notes)
Examples

Transient problem
lviii LIST OF TABLES
 CHAPTER 1
Introduction

1.1 What is Compressible Flow ?

This book deals with an introduction1 to the flow of compressible substances (gases).
The main difference between compressible flow and almost incompressible flow is
not the fact that compressibility has to be considered. Rather, the difference is in
two phenomena that do not exist in incompressible flow2 . The first phenomenon
is the very sharp discontinuity (jump) in the flow in properties. The second phe-
nomenon is the choking of the flow. Choking is when downstream variations don’t
effect the flow3 . Though choking occurs in certain pipe flows in astronomy, there
also are situations of choking in general (external) flow4 . Choking is referred to as
the situation where downstream conditions, which are beyond a critical value(s),
doesn’t affect the flow.
The shock wave and choking are not intuitive for most people. However,
one has to realize that intuition is really a condition where one uses his past expe-
riences to predict other situations. Here one has to learn to use his intuition as a
tool for future use. Thus, not only aeronautic engineers, but other engineers, and
even manufacturing engineers will be able use this “intuition” in design and even
research.
1 This book gradually sliding to include more material that isn’t so introductory. But attempt is made

to present the material in introductory level.

2 It can be argued that in open channel flow there is a hydraulic jump (discontinuity) and in some

ranges no effect of downstream conditions on the flow. However, the uniqueness of the phenomena
in the gas dynamics provides spectacular situations of a limited length (see Fanno model) and thermal
choking, etc. Further, there is no equivalent to oblique shock wave. Thus, this richness is unique to gas
dynamics.
3 The thermal choking is somewhat different but similarity exists.
4 This book is intended for engineers and therefore a discussion about astronomical conditions isn’t

presented.

1
2 CHAPTER 1. INTRODUCTION

1.2 Why Compressible Flow is Important?

Compressible flow appears in many natural and many technological processes.
Compressible flow deals with more than air, including steam, natural gas, nitrogen
and helium, etc. For instance, the flow of natural gas in a pipe system, a common
method of heating in the u.s., should be considered a compressible flow. These
processes include the flow of gas in the exhaust system of an internal combustion
engine, and also gas turbine, a problem that led to the Fanno flow model. The
above flows that were mentioned are called internal flows. Compressible flow also
includes flow around bodies such as the wings of an airplane, and is considered
an external flow.
These processes include situations not expected to have a compressible
flow, such as manufacturing process such as the die casting, injection molding.
The die casting process is a process in which liquid metal, mostly aluminum, is
injected into a mold to obtain a near final shape. The air is displaced by the liquid
metal in a very rapid manner, in a matter of milliseconds, therefore the compress-
ibility has to be taken into account.
Clearly, Aero Engineers are not the only ones who have to deal with some
aspect of compressible flow. For manufacturing engineers there are many situa-
tions where the compressibility or compressible flow understating is essential for
adequate design. For instance, the control engineers who are using pneumatic
systems use compressed substances. The cooling of some manufacturing sys-
tems and design of refrigeration systems also utilizes compressed air flow knowl-
edge. Some aspects of these systems require consideration of the unique phe-
nomena of compressible flow.
Traditionally, most gas dynamics (compressible flow) classes deal mostly
with shock waves and external flow and briefly teach Fanno flows and Rayleigh
flows (two kind of choking flows). There are very few courses that deal with isother-
mal flow. In fact, many books on compressible flow ignore the isothermal flow5 .
In this book, a greater emphasis is on the internal flow. This doesn’t in any
way meant that the important topics such as shock wave and oblique shock wave
should be neglected. This book contains several chapters which deal with external
flow as well.

1.3 Historical Background

In writing this book it became clear that there is more unknown and unwritten about
the history of compressible fluid than known. While there are excellent books about
the history of fluid mechanics (hydraulic) see for example book by Rouse6 . There
are numerous sources dealing with the history of flight and airplanes (aeronau-

5 Any search on the web on classes of compressible flow will show this fact and the undersigned can

testify that this was true in his first class as a student of compressible flow.
6 Hunter Rouse and Simon Inc, History of Hydraulics (Iowa City: Institute of Hydraulic Research,

1957)
1.3. HISTORICAL BACKGROUND 3

tic)7 . Aeronautics is an overlapping part of compressible flow, however these two

fields are different. For example, the Fanno flow and isothermal flow, which are
the core of gas dynamics, are not part of aerodynamics. Possible reasons for the
lack of written documentation are one, a large part of this knowledge is relatively
new, and two, for many early contributors this topic was a side issue. In fact, only
one contributor of the three main models of internal compressible flow (Isothermal,
Fanno, Rayleigh) was described by any text book. This was Lord Rayleigh, for
whom the Rayleigh flow was named. The other two models were, to the under-
signed, unknown. Furthermore, this author did not find any reference to isothermal
flow model earlier to Shapiro’s book. There is no book8 that describes the history
of these models. For instance, the question, who was Fanno, and when did he
live, could not be answered by any of the undersigned’s colleagues in University of
Minnesota or elsewhere.
At this stage there are more questions about the history of compressible
flow needing to be answered. Sometimes, these questions will appear in a section
with a title but without text or with only a little text. Sometimes, they will appear in a
footnote like this9 For example, it is obvious that Shapiro published the erroneous
conclusion that all the chocking occurred at M = 1 in his article which contradicts
his isothermal model. Additional example, who was the first to “conclude” the “all”
the chocking occurs at M = 1? Is it Shapiro?
Originally, there was no idea that there are special effects and phenomena
of compressible flow. Some researchers even have suggested that compressibility
can be “swallowed” into the ideal flow (Euler’s equation’s flow is sometimes re-
ferred to as ideal flow). Even before Prandtl’s idea of boundary layer appeared, the
significant and importance of compressibility emerged.
In the first half of nineteen century there was little realization that the com-
pressibility is important because there were very little applications (if any) that re-
quired the understanding of this phenomenon. As there were no motivations to
investigate the shock wave or choked flow both were treated as the same, taking
compressible flow as if it were incompressible flow.
It must be noted that researchers were interested in the speed of sound
even long before applications and knowledge could demand any utilization. The
research and interest in the speed of sound was a purely academic interest. The
early application in which compressibility has a major effect was with fire arms. The
technological improvements in fire arms led to a gun capable of shooting bullets
at speeds approaching to the speed of sound. Thus, researchers were aware that
the speed of sound is some kind of limit.
In the second half of the nineteen century, Mach and Fliegner “stumbled”
over the shock wave and choking, respectively. Mach observed shock and Fliegner
7 Anderson, J. D., Jr. 1997. A History of Aerodynamics: And Its Impact on Flying Machines, Cam-

bridge University Press, Cambridge, England.

8 The only remark found about Fanno flow that it was taken from the Fanno Master thesis by his

adviser. Here is a challenge: find any book describing the history of the Fanno model.
9 Who developed the isothermal model? The research so far leads to Shapiro. Perhaps this flow

should be named after the Shapiro. Is there any earlier reference to this model?
4 CHAPTER 1. INTRODUCTION

measured the choking but theoretical science did not provide explanation for it (or
was award that there is an explanation for it.).
In the twentieth century the flight industry became the pushing force. Un-
derstandably, aerospace engineering played a significant role in the development
of this knowledge. Giants like Prandtl and his students like Van Karman , as well
as others like Shapiro , dominated the field. During that time, the modern ba-
sic classes became “solidified.” Contributions by researchers and educators from
other fields were not as dominant and significant, so almost all text books in this
field are written from an aerodynamic prospective.

1.3.1 Early Developments

The compressible flow is a subset of fluid mechanics/hydraulics and therefore the

knowledge development followed the understanding of incompressible flow. Early
contributors were motivated from a purely intellectual curiosity, while most later
contributions were driven by necessity. As a result, for a long time the question of
the speed of sound was bounced around.

Speed of Sound

The idea that there is a speed of sound and that it can be measured is a major
achievement. A possible explanation to this discovery lies in the fact that mother
nature exhibits in every thunder storm the difference between the speed of light
and the speed of sound. There is no clear evidence as to who came up with this
concept, but some attribute it to Galileo Galilei: 166x. Galileo, an Italian scientist,
was one of the earliest contributors to our understanding of sound. Dealing with
the difference between the two speeds (light, sound) was a major part of Galileo’s
work. However, once there was a realization that sound can be measured, people
found that sound travels in different speeds through different mediums. The early
approach to the speed of sound was by the measuring of the speed of sound.
Other milestones in the speed of sound understanding development were
by Leonardo Da Vinci, who discovered that sound travels in waves (1500). Marin
Mersenne was the first to measure the speed of sound in air (1640). Robert Boyle
discovered that sound waves must travel in a medium (1660) and this lead to
the concept that sound is a pressure change. Newton was the first to formulate
a relationship between the speed of sound in gases by relating the density and
compressibility in a medium (by assuming isothermal process). Newton’s equation
is missing the heat ratio, k√(late 1660’s). Maxwell was the first to derive the speed
of sound for gas as c = kRT √ from particles (statistical) mechanics. Therefore
some referred to coefficient k as Maxwell’s coefficient.
1.3. HISTORICAL BACKGROUND 5

1.3.2 The shock wave puzzle

Here is where the politics of science was a major obstacle to achieving an ad-
vancement10 . not giving the due credit to Rouse. In the early 18xx, conservation
of energy was a concept that was applied only to mechanical energy. On the other
side, a different group of scientists dealt with calorimetry (internal energy). It was
easier to publish articles about the second law of thermodynamics than to convince
anyone of the first law of thermodynamics. Neither of these groups would agree to
“merge” or “relinquish” control of their “territory” to the other. It took about a century
to establish the first law11 .
At first, Poisson found a “solution” to the Euler’s equations with certain
boundary conditions which required discontinuity12 which had obtained an implicit
form in 1808. Poisson showed that solutions could approach a discontinuity by
using conservation of mass and momentum. He had then correctly derived the
jump conditions that discontinuous solutions must satisfy. Later, Challis had no-
ticed contradictions concerning some solutions of the equations of compressible
gas dynamics13 . Again the “jumping” conditions were redeveloped by two different
researchers independently: Stokes and Riemann. Riemann, in his 1860 the-
sis, was not sure whether or not discontinuity is only a mathematical creature or
a real creature. Stokes in 1848 retreated from his work and wrote an apology on
his “mistake.”14 Stokes was convinced by Lord Rayleigh and Lord Kelvin that he
was mistaken on the grounds that energy is conserved (not realizing the concept
of internal energy).
At this stage some experimental evidence was needed. Ernst Mach stud-
ied several fields in physics and also studied philosophy. He was mostly interested
in experimental physics. The major breakthrough in the understanding of com-
pressible flow came when Ernest Mach “stumbled” over the discontinuity. It is
widely believed that Mach had done his research as purely intellectual research.
His research centered on optic aspects which lead him to study acoustic and there-
fore supersonic flow (high speed, since no Mach number was known at that time).
10 Amazingly, science is full of many stories of conflicts and disputes. Aside from the conflicts of

scientists with the Catholic Church and Muslim religion, perhaps the most famous is that of Newton’s
netscaping (stealing and embracing) Leibniz[’s] invention of calculus. There are even conflicts from
not giving enough credit, like Moody Even the undersigned encountered individuals who have tried
to ride on his work. The other kind of problem is “hijacking” by a sector. Even on this subject, the
Aeronautic sector “took over” gas dynamics as did the emphasis on mathematics like perturbations
methods or asymptotic expansions instead on the physical phenomena. Major material like Fanno flow
isn’t taught in many classes, while many of the mathematical techniques are currently practiced. So,
these problems are more common than one might be expected.
11 This recognition of the first law is today the most “obvious” for engineering students. Yet for many it

was still debatable up to the middle of the nineteen century.

12 Siméon Denis Poisson, French mathematician, 1781-1840 worked in Paris, France. ”M’emoire sur

la th’eorie du son,” J. Ec. Polytech. 14 (1808), 319-392. From Classic Papers in Shock Compression
Science, 3-65, High-press. Shock Compression Condens. Matter, Springer, New York, 1998.
13 James Challis, English Astronomer, 1803-1882. worked at Cambridge, England UK. ”On the veloc-

ity of sound,” Philos. Mag. XXXII (1848), 494-499

14 Stokes George Gabriel Sir, Mathematical and Physical Papers, Reprinted from the original journals

and transactions, with additional notes by the author. Cambridge, University Press, 1880-1905.
6 CHAPTER 1. INTRODUCTION

However, it is logical to believe that his interest had risen due to the need to achieve
powerful/long–distance shooting rifles/guns. At that time many inventions dealt
with machine guns which were able to shoot more bullets per minute. At the time,
one anecdotal story suggests a way to make money by inventing a better killing
machine for the Europeans. While the machine gun turned out to be a good killing
machine, defense techniques started to appear such as sand bags. A need for
bullets that could travel faster to overcome these obstacles was created. There-
fore, Mach’s paper from 1876 deals with the flow around bullets. Nevertheless, no
known15 equations or explanations resulted from these experiments.
Mach used his knowledge in Optics to study the flow around bullets. What
makes Mach’s achievement all the more remarkable was the technique he used to
take the historic photograph: He employed an innovative approach called the shad-
owgraph. He was the first to photograph the shock wave. In his paper discussing
”Photographische Fixierung der durch Projektile in der Luft eingeleiten Vorgange”
he showed a picture of a shock wave (see Figure 1.7). He utilized the variations
of the air density to clearly show shock line at the front of the bullet. Mach had
good understanding of the fundamentals of supersonic flow and the effects on
bullet movement (supersonic flow). Mach’s paper from 1876 demonstrated shock
wave (discontinuity) and suggested the importance of the ratio of the velocity to the
speed of sound. He also observed the existence of a conical shock wave (oblique
shock wave).
Mach’s contributions can be summarized as providing an experimental
proof to discontinuity. He further showed that the discontinuity occurs at M = 1 and
realized that the velocity ratio (Mach number), and not the velocity, is the impor-
tant parameter in the study of the compressible flow. Thus, he brought confidence
to the theoreticians to publish their studies. While Mach proved shock wave and
oblique shock wave existence, he was not able to analyze it (neither was he aware
of Poisson’s work or the works of others.).
Back to the pencil and paper, the jump conditions were redeveloped and
now named after Rankine16 and Hugoniot17 . Rankine and Hugoniot, redeveloped
independently the equation that governs the relationship of the shock wave. Shock
was assumed to be one dimensional and mass, momentum, and energy equa-
tions18 lead to a solution which ties the upstream and downstream properties.
What they could not prove or find was that shock occurs only when upstream is
15 The words “no known” refer to the undersigned. It is possible that some insight was developed but

none of the documents that were reviewed revealed it to the undersigned.

16 William John Macquorn Rankine, Scottish engineer, 1820-1872. He worked in Glasgow, Scotland

UK. ”On the thermodynamic theory of waves of finite longitudinal disturbance,” Philos. Trans. 160
(1870), part II, 277-288. Classic papers in shock compression science, 133-147, High-press. Shock
Compression Condens. Matter, Springer, New York, 1998
17 Pierre Henri Hugoniot, French engineer, 1851-1887. ”Sur la propagation du mouvement dans les

corps et sp’ecialement dans les gaz parfaits, I, II” J. Ec. Polytech. 57 (1887), 3-97, 58 (1889), 1-125.
Classic papers in shock compression science, 161-243, 245-358, High-press. Shock Compression
Condens. Matter, Springer, New York, 1998
18 Today it is well established that shock has three dimensions but small sections can be treated as

one dimensional.
1.3. HISTORICAL BACKGROUND 7

supersonic, i.e., direction of the flow. Later, others expanded Rankine-Hugoniot’s

conditions to a more general form19 .
Here, the second law has been around for over 40 years and yet the sig-
nificance of it was not was well established. Thus, it took over 50 years for Prandtl
to arrive at and to demonstrate that the shock has only one direction20 . Today
this equation/condition is known as Prandtl’s equation or condition (1908). In fact
Prandtl is the one who introduced the name of Rankine-Hugoniot’s conditions not
aware of the earlier developments of this condition. Theodor Meyer (Prandtl’s
student) derived the conditions for oblique shock in 190821 as a byproduct of the
expansion work.
It was probably later
that Stodola (Fanno’s adviser)
realized that the shock is the
intersection of the Fanno line
with the Rayleigh line. Yet,
the supersonic branch is miss-
ing from his understanding (see
Figure (1.1)). In fact, Stodola
suggested the graphical solu-
tion utilizing the Fanno line.
The fact that the condi-
tions and direction were known
did not bring the solution to
the equations. The “last nail”
of understanding was put by
Landau, a Jewish scientist who
worked in Moscow University in
the 1960’s during the Commu-
nist regimes. A solution was Fig. -1.1: The shock as a connection of Fanno and
Rayleigh lines after Stodola, Steam and Gas
found by Landau & Lifshitz
Turbine
and expanded by Kolosnitsyn &
Stanyukovich (1984).
Since early in the 1950s the analytical relationships between the oblique
shock, deflection angle, shock angle, and Mach number was described as impossi-
ble to obtain. There were until recently (version 0.3 of this book) several equations
that tied various properties/quantities for example, the relationship between up-
stream Mach number and the angles. The first full analytical solution connecting
the angles with upstream Mach number was published in this book version 0.3.
The probable reason that analytical solution was not published because the claim

19 Toadd discussion about the general relationships.

20 Some view the work of G. I. Taylor from England as the proof (of course utilizing the second law)
21 Theodor Meyer in Mitteil. üb. Forsch-Arb. Berlin, 1908, No. 62, page 62.
8 CHAPTER 1. INTRODUCTION

in the famous report of NACA 1135 that explicit analytical solution isn’t possible22 .

The question whether the oblique shock is stable or which root is stable
was daunting since the early discovery that there are more than one possible so-
lution. It is amazing that early research concluded that only the weak solution is
possible or stable as opposed to the reality. The first that attempt this question
where in 1931 by Epstein24 . His analysis was based on Hamilton’s principle when
he ignore the boundary condition. The results of that analysis was that strong
shock is unstable. The researchers understood that flow after a strong shock was
governed by elliptic equation while the flow after a weak shock was governed by
hyperbolic equations. This difference probably results in not recognizing that The
boundary conditions play an important role in the stability of the shock25 . In fact
analysis based on Hamilton’s principle isn’t suitable for stability because entropy
creation was recognized 1955 by Herivel26 .
Carrier27 was first to recognize that strong and weak shocks stable. If fact
the confusion on this issue was persistent until now. Even all books that were pub-
lished recently claimed that no strong shock was ever observed in flow around cone
(Taylor–Maccoll flow). In fact, even this author sinned in this erroneous conclusion.
The real question isn’t if they exist rather under what conditions these shocks exist
which was suggested by Courant and Friedrichs in their book “Supersonic Flow
and Shock Waves,” published by Interscience Publishers, Inc. New York, 1948, p.
317.
The effect of real gases was investigated very early since steam was used
move turbines. In general the mathematical treatment was left to numerical in-
vestigation and there is relatively very little known on the difference between ideal
gas model and real gas. For example, recently, Henderson and Menikoff28 dealt
22 Since writing this book, several individuals point out that a solution was found in book “Analytical

Fluid Dynamics” by Emanuel, George, second edition, December 2000 (US$ 124.90). That solution
is based on a transformation of sin θ to tan β. It is interesting that transformation result in one of root
being negative. While the actual solution all the roots are real and positive for the attached shock. The
presentation was missing the condition for the detachment or point where the model collapse. But more
surprisingly, similar analysis was published by Briggs, J. “Comment on Calculation of Oblique shock
waves,” AIAA Journal Vol 2, No 5 p. 974, 1963. Hence, Emanuel’s partial solution just redone 36
years work (how many times works have to be redone in this field). In additonal there was additional
publishing of similar works by Mascitti, V.R. and Wolf, T. 23 In a way, part of analysis of this book is also
redoing old work. Yet, what is new in this work is completeness of all the three roots and the analytical
condition for detached shock and breaking of the model.
24 Epstein, P. S., “On the air resistance of Projectiles,” Proceedings of the National Academy of Sci-

ence, Vol. 17, 1931, pp. 532-547.

25 In study this issue this author realized only after examining a colleague experimental Picture (14.4)

that it was clear that the Normal shock along with strong shock and weak shock “live” together peacefully
and in stable conditions.
26 Herivel, J. F., “The Derivation of The Equations of Motion On an Ideal Fluid by Hamilton’s Principle,,”

Proceedings of the Cambridge philosophical society, Vol. 51, Pt. 2, 1955, pp. 344-349.
27 Carrier, G.F., “On the Stability of the supersonic Flows Past as a Wedge,” Quarterly of Applied

Mathematics, Vol. 6, 1949, pp. 367–378.

28 Henderson and Menikoff, ”Triple Shock Entropy Theorem,” Journal of Fluid Mechanics 366 (1998)

pp. 179–210.
1.3. HISTORICAL BACKGROUND 9

with only the procedure to find the maximum of oblique shock, but no comparison
between real gases and ideal gas is offered there.
The moving shock and shock tube were study even before World War
Two. The realization that in most cases the moving shock can be analyzed as
steady state since it approaches semi steady state can be traced early of 1940’s.
Up to this version 0.4.3 of this book (as far it is known, this book is first to publish
this tables), trial and error method was the only method to solve this problem.
Only after the dimensionless presentation of the problem and the construction of
the moving shock table the problem became trivial. Later, an explicit analytical
solution for shock a head of piston movement (special case of open valve) was
originally published in this book for the first time.

1.3.3 Choking Flow

The choking problem is almost unique
to gas dynamics and has many differ-
ent forms. Choking wasn’t clearly to
be observed, even when researcher
stumbled over it. No one was look-
ing for or expecting the choking to oc-
cur, and when it was found the sig-
nificance of the choking phenomenon
was not clear. The first experimental
choking phenomenon was discovered
by Fliegner’s experiments which were
conducted some time in the middle of
186x29 on air flow through a converg-
ing nozzle. As a result deLavel’s noz-
zle was invented by Carl Gustaf Patrik Fig. -1.2: The schematic of deLavel’s turbine af-
de Laval in 1882 and first successful ter Stodola, Steam and Gas Turbine
operation by another inventor (Curtis)
1896 used in steam turbine. Yet, there was no realization that the flow is choked
just that the flow moves faster than speed of sound.
The introduction of the steam engine and other thermodynamics cycles led
to the choking problem. The problem was introduced because people wanted to
increase the output of the Engine by increasing the flames (larger heat transfer or
larger energy) which failed, leading to the study and development of Rayleigh flow.
According the thermodynamics theory (various cycles) the larger heat supply for
a given temperature difference (larger higher temperature) the larger the output,
but after a certain point it did matter (because the steam was choked). The first to

29 Fliegner Schweizer Bauztg., Vol 31 1898, p. 68–72. The theoretical first work on this issue was

done by Zeuner, “Theorie die Turbinen,” Leipzig 1899, page 268 f.

10 CHAPTER 1. INTRODUCTION

discover (try to explain) the choking phenomenon was Rayleigh30 .

After the introduction of the deLavel’s converging–diverging nozzle theo-

retical work was started by Zeuner31 . Later continue by Prandtl’s group32 start-
ing 1904. In 1908 Meyer has extend this work to make two dimensional calcula-
tions33 . Experimental work by Parenty34 and others measured the pressure along
the converging-diverging nozzle.

It was commonly believed35 that the choking

√ occurs only at M = 1. The
first one to analyzed that choking occurs at 1/ k for isothermal flow was Shapiro
(195x). It is so strange that a giant like Shapiro did not realize his model on isother-
mal contradict his conclusion from his own famous paper. Later Romer at el ex-
tended it to isothermal variable area flow (1955). In this book, this author adapts
E.R.G. Ecert’s idea of dimensionless parameters control which determines where
the reality lay between the two extremes. Recently this concept was proposed (not
explicitly) by Dutton and Converdill (1997)36 . Namely, in many cases the reality is
somewhere between the adiabatic and the isothermal flow. The actual results will
be determined by the modified Eckert number to which model they are closer.

30 Rayleigh was the first to develop the model that bears his name. It is likely that others had noticed

that flow is choked, but did not produce any model or conduct successful experimental work.
31 Zeuner, “Theorie der Turbinen, Leipzig 1899 page 268 f.
32 Some of the publications were not named after Prandtl but rather by his students like Meyer,

Theodor. In the literature appeared reference to article by Lorenz in the Physik Zeitshr., as if in 1904.
Perhaps, there are also other works that this author did not come crossed.
33 Meyer, Th., Über zweidimensionals Bewegungsvordange eines Gases, Dissertation 1907, er-

schienen in den Mitteilungen über Forsch.-Arb. Ing.-Wes. heft 62, Berlin 1908.
34 Parenty, Comptes R. Paris, Vol. 113, 116, 119; Ann. Chim. Phys. Vol. 8. 8 1896, Vol 12, 1897.
35 The personal experience of this undersigned shows that even instructors of Gas Dynamics are not

aware that the chocking occurs at different Mach number and depends on the model.
36 These researchers demonstrate results between two extremes and actually proposed this idea.

However, that the presentation here suggests that topic should be presented case between two ex-
tremes.
1.3. HISTORICAL BACKGROUND 11

Nozzle Flow
The first “wind tunnel” was not
a tunnel but a rotating arm at-
tached at the center. At the
end of the arm was the ob-
ject that was under observation
and study. The arm’s circular
motion could reach a velocity
above the speed of sound at its
end. Yet, in 1904 the Wright
brothers demonstrated that re-
sults from the wind tunnel and
spinning arm are different, due
to the circular motion. As a
result, the spinning arm was
no longer used in testing. Be-
tween the turn of the century Fig. -1.3: The measured pressure in a nozzle taken
and 1947-48, when the first su- from Stodola 1927 Steam and Gas Turbines
personic wind tunnel was built,
several models that explained choking at the throat have been built.

A different reason to study the converging-diverging nozzle was the Venturi me-
ter which was used in measuring the flow rate of gases. Bendemann 37 carried
experiments to study the accuracy of these flow meters and he measured and re-
found that the flow reaches a critical value (pressure ratio of 0.545) that creates
the maximum flow rate.

There are two main models or extremes that describe the flow in the nozzle:
isothermal and adiabatic.

37 Bendemann Mitteil über Forschungsarbeiten, Berlin, 1907, No. 37.

12 CHAPTER 1. INTRODUCTION

Nozzle flow
Romer et al38 analyzed the
isothermal flow in a nozzle.
It is remarkable that√chok-
ing was found as 1/ k as
opposed to one (1). In gen-
eral when the model is as-
sumed to be isothermal√ the
choking occurs at 1/ k.
The concept that the chok-
ing point can move from the
throat introduced by39 a re-
searcher unknown to this
author. It is very interesting
that the isothermal nozzle
was proposed by Romer
at el 1955 (who was be-
hind the adviser or the stu-
dent?). These researchers Fig. -1.4: Flow rate as a function of the back pressure taken
were the first ones to real- after Stodola 1927 Steam and Gas Turbines
√
ized that choking can occurs at different Mach number (1/ k other then the
isothermal pipe.

Rayleigh Flow
Rayleigh was probably40 , the first to suggest a model for frictionless flow with a
constant heat transfer. Rayleigh’s work was during the time when it was debatable
as to whether there are two forms of energies (mechanical, thermal), even though
Watt and others found and proved that they are the same. Therefore, Rayleigh
looked at flow without mechanical energy transfer (friction) but only thermal energy
transfer. In Rayleigh flow, the material reaches choking point due to heat transfer,
hence term “thermally choked” is used; no additional flow can occur.

Fanno Flow
The most important model in compressible flow was suggested by Gino Fanno in
his Master’s thesis (1904). The model bears his name. Yet, according to Dr. Rudolf
39 Romer, I Carl Jr., and Ali Bulent Cambel, “Analysis of Isothermal Variable Area Flow,” Aircraft Eng.

vol. 27 no 322, p. 398 December 1955.

39 This undersign didn’t find the actual trace to the source of proposing this effect. However, some

astronomy books showing this effect in a dimensional form without mentioning the original researcher.
In dimensionless form, this phenomenon produces a dimensionless number similar to Ozer number
and therefor the name Ozer number adapted in this book.
40 As most of the history research has shown, there is also a possibility that someone found it earlier.

For example, Simeon–Denis Poisson was the first one to realize the shock wave possibility.
1.3. HISTORICAL BACKGROUND 13

Mumenthaler from UTH University, no copy of the thesis can be found in the orig-
inal University and perhaps only in the personal custody of the Fanno family41 .
Fanno attributes the main pressure reduction to friction. Thus, flow that is domi-
nantly adiabatic could be simplified and analyzed. The friction factor is the main
component in the analysis as Darcy f 42 had already proposed in 1845. The arrival
of the Moody diagram, which built on Hunter Rouse’s (194x) work made Darcy–
Weisbach’s equation universally useful. Without the existence of the friction factor
data, the Fanno model wasn’t able to produce a prediction useful for the indus-
try. Additionally an understating of the supersonic branch of the flow was unknown
(The idea of shock in tube was not raised at that time.). Shapiro organized all the
material in a coherent way and made this model useful.

Meta
Did Fanno realize that the flow is choked? It appears at least in Stodola’s
book that choking was understood in 1927 and even earlier. The choking
was assumed only to be in the subsonic flow. But because the actual Fanno’s
thesis is not available, the question cannot be answered yet. When was
Gas Dynamics (compressible flow) as a separate class started? Did the
explanation for the combination of diverging-converging nuzzle with tube for
Fanno flow first appeared in Shapiro’s book?

Meta End
Isothermal Flow

The earliest reference to isothermal flow

√ was found in Shapiro’s Book. The model
suggests that the choking occurs at 1/ k and it appears that Shapiro was the first
one to realize this difference compared
√ to the other models. In reality, the flow is
choked somewhere between 1/ k to one for cases that are between Fanno (adi-
abatic) and isothermal flow. This fact was evident in industrial applications where
the expectation of the choking is at Mach one, but can be explained by choking at
a lower Mach number. No experimental evidence, known by the undersigned, was
ever produced to verify this finding.

1.3.4 External flow

When the flow over an external body is about .8 Mach or more the flow must
be considered to be a compressible flow. However at a Mach number above 0.8
(relative of velocity of the body to upstream velocity) a local Mach number (local
velocity) can reach M = 1. At that stage, a shock wave occurs which increases
the resistance. The Navier-Stokes equations which describe the flow (or even
41 This material is very important and someone should find it and make it available to researchers.
42 Fanning f based radius is only one quarter of the Darcy f which is based on diameter
14 CHAPTER 1. INTRODUCTION

Euler equations) were considered unsolvable during the mid 18xx because of the
high complexity. This problem led to two consequences. Theoreticians tried to
simplify the equations and arrive at approximate solutions representing specific
cases. Examples of such work are Hermann von Helmholtz’s concept of vortex
filaments (1858), Lanchester’s concept of circulatory flow (1894), and the Kutta-
Joukowski circulation theory of lift (1906). Practitioners like the Wright brothers
relied upon experimentation to figure out what theory could not yet tell them.
Ludwig Prandtl in 1904 explained the two most important causes of drag by intro-
ducing the boundary layer theory. Prandtl’s boundary layer theory allowed various
simplifications of the Navier-Stokes equations. Prandtl worked on calculating the
effect of induced drag on lift. He introduced the lifting line theory, which was pub-
lished in 1918-1919 and enabled accurate calculations of induced drag and its
effect on lift43 .
During World War I, Prandtl created his thin–airfoil theory that enabled the calcula-
tion of lift for thin, cambered airfoils. He later contributed to the Prandtl-Glauert rule
for subsonic airflow that describes the compressibility effects of air at high speeds.
Prandtl’s student, Von Karman reduced the equations for supersonic flow into a
single equation.
After the First World War aviation became important and in the 1920s a push of
research focused on what was called the compressibility problem. Airplanes could
not yet fly fast, but the propellers (which are also airfoils) did exceed the speed of
sound, especially at the propeller tips, thus exhibiting inefficiency. Frank Caldwell
and Elisha Fales demonstrated in 1918 that at a critical speed (later renamed the
critical Mach number) airfoils suffered dramatic increases in drag and decreases
in lift. Later, Briggs and Dryden showed that the problem was related to the shock
wave. Meanwhile in Germany, one of Prandtl’s assistants, J. Ackeret, simplified
the shock equations so that they became easy to use. After World War Two, the
research had continued and some technical solutions were found. Some of the
solutions lead to tedious calculations which lead to the creation of Computational
Fluid Dynamics (CFD). Today these methods of perturbations and asymptotic are
hardly used in wing calculations44 . That is the “dinosaur45 ” reason that even today
some instructors are teaching mostly the perturbations and asymptotic methods in
Gas Dynamics classes.
More information on external flow can be found in , John D. Anderson’s Book “His-
tory of Aerodynamics and Its Impact on Flying Machines,” Cambridge University
Press, 1997
43 The English call this theory the Lanchester-Prandtl theory. This is because the English Astronomer

Frederick Lanchester published the foundation for Prandtl’s theory in his 1907 book Aerodynamics.
However, Prandtl claimed that he was not aware of Lanchester’s model when he had begun his work
in 1911. This claim seems reasonable in the light that Prandtl was not ware of earlier works when he
named erroneously the conditions for the shock wave. See for the full story in the shock section.
44 This undersigned is aware of only one case that these methods were really used to calculations of

wing.
45 It is like teaching using slide ruler in today school. By the way, slide rule is sold for about 7.5$ on

the net. Yet, there is no reason to teach it in a regular school.

1.3. HISTORICAL BACKGROUND 15

1.3.5 Filling and Evacuating Gaseous Chambers

It is remarkable that there were so few contributions made in the area of a filling
or evacuation gaseous chamber. The earlier work dealing with this issue was by
Giffen, 1940, and was republished by Owczarek, J. A., the model and solution to
the nozzle attached to chamber issue in his book “Fundamentals of Gas Dynam-
ics.”46 . He also extended the model to include the unchoked case. Later several
researchers mostly from the University in Illinois extended this work to isothermal
nozzle (choked and unchoked).

The simplest model of nozzle, is not sufficient in many cases and a connection by
a tube (rather just nozzle or orifice) is more appropriated. Since World War II con-
siderable works have been carried out in this area but with very little progress47 . In
1993 the first reasonable models for forced volume were published by the under-
signed. Later, that model was extended by several research groups, The analyti-
cal solution for forced volume and the “balloon” problem (airbag’s problem) model
were published first in this book (version 0.35) in 2005. The classification of fill-
ing or evacuating the chamber as external control and internal control (mostly by
pressure) was described in version 0.3 of this book by this author.

1.3.6 Biographies of Major Figures

In this section a short summary of major figures that influenced the field of gas
dynamics is present. There are many figures that should be included and a biased
selection was required. Much information can be obtained from other resources,
such as the Internet. In this section there is no originality and none should be
expected.

46 InternationalTextbook Co., Scranton, Pennsylvania, 1964.

47 Infact, the emergence of the CFD gave the illusion that there are solutions at hand, not realizing
that garbage in is garbage out, i.e., the model has to be based on scientific principles and not detached
from reality. As anecdotal story explaining the lack of progress, in die casting conference there was
a discussion and presentation on which turbulence model is suitable for a complete still liquid. Other
“strange” models can be found in the undersigned’s book “Fundamentals of Die Casting Design.
16 CHAPTER 1. INTRODUCTION

Galileo Galilei

Galileo was born in Pisa, Italy on Febru-

ary 15, 1564 to musician Vincenzo
Galilei and Giulia degli Ammannati. The
oldest of six children, Galileo moved
with his family in early 1570 to Florence.
Galileo started his studying at the Uni-
versity of Pisa in 1581. He then became
a professor of mathematics at the Uni-
versity of Padua in 1592. During the
time after his study, he made numerous
discoveries such as that of the pendu-
lum clock, (1602). Galileo also proved
that objects fell with the same velocity
regardless of their size.

Fig. -1.5: Portrait of Galileo Galilei

Galileo had a relationship with Marina Gamba (they never married) who lived and
worked in his house in Padua, where she bore him three children. However, this
relationship did not last and Marina married Giovanni Bartoluzzi and Galileo’s son,
Vincenzio, joined him in Florence (1613).

Galileo invented many mechanical devices such as the pump and the telescope
(1609). His telescopes helped him make many astronomic observations which
proved the Copernican system. Galileo’s observations got him into trouble with
the Catholic Church, however, because of his noble ancestry, the church was not
harsh with him. Galileo was convicted after publishing his book Dialogue, and he
was put under house arrest for the remainder of his life. Galileo died in 1642 in his
home outside of Florence.
1.3. HISTORICAL BACKGROUND 17

Ernest Mach (1838-1916)

Ernst Mach was born in 1838 in
Chrlice (now part of Brno), when
Czechia was still a part of the
Austro–Hungary empire. Johann,
Mach’s father, was a high school
teacher who taught Ernst at home
until he was 14, when he stud-
ied in Kromeriz Gymnasium, be-
fore he entered the university of Vi-
enna were he studies mathematics,
physics and philosophy. He grad-
uated from Vienna in 1860. There
Mach wrote his thesis ”On Electrical
Discharge and Induction.” Mach was Fig. -1.6: Photo of Ernest Mach
interested also in physiology of sen-
sory perception.
At first he received a professorship position at Graz in mathematics (1864) and
was then offered a position as a professor of surgery at the university of Salzburg,
but he declined. He then turned to physics, and in 1867 he received a position in
the Technical University in Prague48 where he taught experimental physics for the
next 28 years.
Mach was also a great thinker/philosopher and influenced the theory of relativity
dealing with frame of reference. In 1863, Ernest Mach (1836 - 1916) published
Die Machanik in which he formalized this argument. Later, Einstein was greatly
influenced by it, and in 1918, he named it Mach’s Principle. This was one of the
primary sources of inspiration for Einstein’s theory of General Relativity.
Mach’s revolutionary exper-
iment demonstrated the ex-
istence of the shock wave
as shown in Figure 1.7.
It is amazing that Mach
was able to photograph
the phenomenon using the
spinning arm technique (no
wind tunnel was available
at that time and most def-
initely nothing that could Fig. -1.7: The Photo of the bullet in a supersonic flow that
take a photo at supersonic Mach made. Note it was not taken in a wind tunnel
speeds. His experiments
required exact timing. He was not able to attach the camera to the arm and utilize
the remote control (not existent at that time). Mach’s shadowgraph technique and
48 It is interesting to point out that Prague provided us two of the top influential researchers[:] E. Mach

and E.R.G. Eckert.

18 CHAPTER 1. INTRODUCTION

a related method called Schlieren Photography are still used today.

Yet, Mach’s contributions to supersonic flow were not limited to experimental meth-
ods alone. Mach understood the basic characteristics of external supersonic flow
where the most important variable affecting the flow is the ratio of the speed of the
flow49 (U) relative to the speed of sound (c). Mach was the first to note the transi-
tion that occurs when the ratio U/c goes from being less than 1 to greater than 1.
The name Mach Number (M) was coined by J. Ackeret (Prandtl’s student) in 1932
in honor of Mach.

John William Strutt (Lord Rayleigh)

A researcher with a wide interest, started stud-
ies in compressible flow mostly from a math-
ematical approach. At that time there wasn’t
the realization that the flow could be choked. It
seems that Rayleigh was the first who realized
that flow with chemical reactions (heat transfer)
can be choked.
Lord Rayleigh was a British physicist born near
Maldon, Essex, on November 12, 1842. In 1861
he entered Trinity College at Cambridge, where
he commenced reading mathematics. His ex-
ceptional abilities soon enabled him to overtake
his colleagues. He graduated in the Mathe-
matical Tripos in 1865 as Senior Wrangler and
Smith’s Prizeman. In 1866 he obtained a fellow-
ship at Trinity which he held until 1871, the year
of his marriage. He served for six years as the Fig. -1.8: Photo of Lord Rayleigh
president of the government committee on ex-
plosives, and from 1896 to 1919 he acted as Scientific Adviser to Trinity House.
He was Lord Lieutenant of Essex from 1892 to 1901.
Lord Rayleigh’s first research was mainly mathematical, concerning optics and vi-
brating systems, but his later work ranged over almost the whole field of physics,
covering sound, wave theory, color vision, electrodynamics, electromagnetism,
light scattering, flow of liquids, hydrodynamics, density of gases, viscosity, capillar-
ity, elasticity, and photography. Rayleigh’s later work was concentrated on electric
and magnetic problems. Rayleigh was considered to be an excellent instructor. His
Theory of Sound was published in two volumes during 1877-1878, and his other
extensive studies are reported in his Scientific Papers, six volumes issued during
1889-1920. Rayleigh was also a contributer to the Encyclopedia Britannica. He
published 446 papers which, reprinted in his collected works, clearly show his ca-
pacity for understanding everything just a little more deeply than anyone else. He
49 Mach dealt with only air, but it is reasonable to assume that he understood that this ratio was applied

to other gases.
1.3. HISTORICAL BACKGROUND 19

intervened in debates of the House of Lords only on rare occasions, never allowing
politics to interfere with science. Lord Rayleigh, a Chancellor of Cambridge Uni-
versity, was a Justice of the Peace and the recipient of honorary science and law
degrees. He was a Fellow of the Royal Society (1873) and served as Secretary
from 1885 to 1896, and as President from 1905 to 1908. He received the Nobel
Prize in 1904. Lord Rayleigh died on June 30, 1919, at Witham, Essex.

In 1871 he married Evelyn, sister of the future prime minister, the Earl of Balfour
(of the famous Balfour declaration of the Jewish state). They had three sons, the
eldest of whom was to become a professor of physics at the Imperial College of
Science and Technology, London.

As a successor to James Clerk Maxwell, he was head of the Cavendish Laboratory

at Cambridge from 1879-1884, and in 1887 became Professor of Natural Philos-
ophy at the Royal Institute of Great Britain. Rayleigh died on June 30, 1919 at
Witham, Essex.

William John Macquorn Rankine

William John Macquorn Rankine (July 2,

1820 - December 24, 1872) was a Scottish
engineer and physicist. He was a founding
contributor to the science of thermodynam-
ics (Rankine Cycle). Rankine developed a
theory of the steam engine. His steam en-
gine manuals were used for many decades.
Rankine was well rounded interested beside
the energy field he was also interested in
civil engineering, strength of materials, and
naval engineering in which he was involved
in applying scientific principles to building
ships.
Rankine was born in Edinburgh to British
Army lieutenant David Rankine and Barbara Fig. -1.9: Portrait of Rankine
Grahame, Rankine. Rankine never married, and his only brother and parents died
before him.
20 CHAPTER 1. INTRODUCTION

Gino Girolamo Fanno

Fanno a Jewish Engineer was born on

November 18, 1888. He studied in a tech-
nical institute in Venice and graduated with
very high grades as a mechanical engi-
neer. Fanno was not as lucky as his
brother, who was able to get into academia.
Faced with anti–semitism, Fanno left Italy for
Zurich, Switzerland in 1900 to attend gradu-
ate school for his master’s degree. In this
new place he was able to pose as a Ro-
man Catholic, even though for short time
he went to live in a Jewish home, Isaak
Baruch Weil’s family. As were many Jews
at that time, Fanno was fluent in several lan-
guages including Italian, English, German, Fig. -1.10: The photo of Gino Fanno ap-
and French. He likely had a good knowl- proximately in 1950
edge of Yiddish and possibly some Hebrew.
Consequently, he did not have a problem studying in a different language. In July
1904 he received his diploma (master). When one of Professor Stodola’s assis-
tants attended military service this temporary position was offered to Fanno. “Why
didn’t a talented guy like Fanno keep or obtain a position in academia after he pub-
lished his model?” The answer is tied to the fact that somehow rumors about his
roots began to surface. Additionally, the fact that his model was not a “smashing50
success” did not help.
Later Fanno had to go back to Italy to find a job in industry. Fanno turned out to be
a good engineer and he later obtained a management position. He married, and
like his brother, Marco, was childless. He obtained a Ph.D. from Regian Istituto
Superiore d’Ingegneria di Genova. However, on February 1939 Fanno was de-
graded (denounced) and he lost his Ph.D. (is this the first case in history) because
his of his Jewish nationality51 . During the War (WWII), he had to be under house
arrest to avoid being sent to the “vacation camps.” To further camouflage himself,
Fanno converted to Catholicism. Apparently, Fanno had a cache of old Italian cur-
rency (which was apparently still highly acceptable) which helped him and his wife
survive the war. After the war, Fanno was only able to work in agriculture and agri-
cultural engineering. Fanno passed way in 1960 without world recognition for his
model.
Fanno’s older brother, mentioned earlier Marco Fanno is a famous economist who
later developed fundamentals of the supply and demand theory.

50 Missingdata about friction factor

51 Insome places, the ridicules claims that Jews persecuted only because their religion. Clearly,
Fanno was not part of the Jewish religion (see his picture) only his nationality was Jewish.
1.3. HISTORICAL BACKGROUND 21

Ludwig Prandtl
Perhaps Prandtl’s greatest achievement
was his ability to produce so many great
scientists. It is mind boggling to look at
the long list of those who were his stu-
dents and colleagues. There is no one
who educated as many great scientists
as Prandtl. Prandtl changed the field of
fluid mechanics and is called the mod-
ern father of fluid mechanics because of
his introduction of boundary layer, tur-
bulence mixing theories etc.
Ludwig Prandtl was born in Freising,
Bavaria, in 1874. His father was a pro-
fessor of engineering and his mother
suffered from a lengthy illness. As a re-
sult, the young Ludwig spent more time
with his father which made him inter-
ested in his father’s physics and ma-
chinery books. This upbringing fostered Fig. -1.11: Photo of Prandtl
the young Prandtl’s interest in science
and experimentation.
Prandtl started his studies at the age of 20 in Munich, Germany and he graduated
at the age of 26 with a Ph.D. Interestingly, his Ph.D. was focused on solid mechan-
ics. His interest changed when, in his first job, he was required to design factory
equipment that involved problems related to the field of fluid mechanics (a suction
device). Later he sought and found a job as a professor of mechanics at a tech-
nical school in Hannover, Germany (1901). During this time Prandtl developed his
boundary layer theory and studied supersonic fluid flows through nozzles. In 1904,
he presented the revolutionary paper “Flussigkeitsbewegung Bei Sehr Kleiner Rei-
bung” (Fluid Flow in Very Little Friction), the paper which describes his boundary
layer theory.
His 1904 paper raised Prandtl’s prestige. He became the director of the Institute for
Technical Physics at the University of Göttingen. He developed the Prandtl-Glauert
rule for subsonic airflow. Prandtl, with his student Theodor Meyer, developed the
first theory for calculating the properties of shock and expansion waves in super-
sonic flow in 1908 (two chapters in this book). As a byproduct they produced the
theory for oblique shock. In 1925 Prandtl became the director of the Kaiser Wil-
helm Institute for Flow Investigation at Göttingen. By the 1930s, he was known
worldwide as the leader in the science of fluid dynamics. Prandtl also contributed
to research in many areas, such as meteorology and structural mechanics.
Ludwig Prandtl worked at Göttingen until his death on August 15, 1953. His work
and achievements in fluid dynamics resulted in equations that simplified under-
standing, and many are still used today. Therefore many referred to him as the
22 CHAPTER 1. INTRODUCTION

father of modern fluid mechanics. Ludwig Prandtl died in Göttingen, Germany on

August 15th 1953.
Prandtl’s other contributions include: the introduction of the Prandtl number in fluid
mechanics, airfoils and wing theory (including theories of aerodynamic interfer-
ence, wing-fuselage, wing-propeller, biplane, etc); fundamental studies in the wind
tunnel, high speed flow (correction formula for subsonic compressible flows), the-
ory of turbulence. His name is linked to the following:
• Prandtl number (heat transfer problems)
• Prandtl-Glauert compressibility correction
• Prandtl’s boundary layer equation
• Prandtl’s lifting line theory
• Prandtl’s law of friction for smooth pipes
• Prandtl-Meyer expansion fans (supersonic flow)
• Prandtl’s Mixing Length Concept (theory of turbulence)

E.R.G. Eckert
Eckert was born in 1904 in Prague,
where he studied at the German Insti-
tute of Technology. During World War
II, he developed methods for jet engine
turbine blade cooling at a research lab-
oratory in Prague. He emigrated to the
United States after the war, and served
as a consultant to the U.S. Air Force
and the National Advisory Committee
for Aeronautics before coming to Min-
nesota.
Eckert developed the understanding of Fig. -1.12: The photo of Ernst Rudolf George
heat dissipation in relation to kinetic en- Eckert with the author’s family
ergy, especially in compressible flow.
Hence, the dimensionless group has
been designated as the Eckert number, which is associated with the Mach number.
Schlichting suggested this dimensionless group in honor of Eckert. In addition to
being named to the National Academy of Engineering in 1970, He authored more
than 500 articles and received several medals for his contributions to science. His
book ”Introduction to the Transfer of Heat and Mass,” published in 1937, is still
considered a fundamental text in the field.
Eckert was an excellent mentor to many researchers (including this author), and
he had a reputation for being warm and kindly. He was also a leading figure in
bringing together engineering in the East and West during the Cold War years.
1.3. HISTORICAL BACKGROUND 23

Ascher Shapiro
MIT Professor Ascher Shapiro52 , the Eckert equivalent for the compressible flow,
was instrumental in using his two volume book “The Dynamics of Thermodynamics
of the Compressible Fluid Flow,” to transform the gas dynamics field to a coherent
text material for engineers. Furthermore, Shapiro’s knowledge of fluid mechanics
enabled him to “sew” the missing parts of the Fanno line with Moody’s diagram
to create the most useful model in compressible flow. While Shapiro viewed gas
dynamics mostly through aeronautic eyes, The undersigned believes that Shapiro
was the first one to propose an isothermal flow model that is not part of the aero-
nautic field. Therefore it is proposed to call this model Shapiro’s Flow.
In his first 25 years Shapiro focused primarily on power production, high-speed
flight, turbomachinery and propulsion by jet engines and rockets. Unfortunately for
the field of Gas Dynamics, Shapiro moved to the field of biomedical engineering
where he was able to pioneer new work. Shapiro was instrumental in the treatment
of blood clots, asthma, emphysema and glaucoma.
Shapiro grew up in New York City and received his S.B. in 1938 and the Sc.D. (It
is M.I.T.’s equivalent of a Ph.D. degree) in 1946 in mechanical engineering from
MIT. He was assistant professor in 1943, three years before receiving his Sc.D.
In 1965 he become the head of the Department of Mechanical Engineering until
1974. Shapiro spent most of his active years at MIT. Ascher Shapiro passed way
in November 2004.

52 Parts taken from Sasha Brown, MIT

24 CHAPTER 1. INTRODUCTION
 CHAPTER 2
Review of Thermodynamics

In this chapter, a review of several definitions of common thermodynamics terms is

presented. This introduction is provided to bring the student back to current place
with the material.

2.1 Basic Definitions

The following basic definitions are common to thermodynamics and will be used in
this book.
Work
In mechanics, the work was defined as
Z Z
mechanical work = F • d` = P dV (2.1)

This definition can be expanded to include two issues. The first issue that must be
addressed, that work done on the surroundings by the system boundaries similarly
is positive. Two, there is a transfer of energy so that its effect can cause work. It
must be noted that electrical current is a work while heat transfer isn’t.
System
This term will be used in this book and it is defined as a continuous (at least par-
tially) fixed quantity of matter. The dimensions of this material can be changed. In
this definition, it is assumed that the system speed is significantly lower than that
of the speed of light. So, the mass can be assumed constant even though the true
conservation law applied to the combination of mass energy (see Einstein’s law).
In fact for almost all engineering purpose this law is reduced to two separate laws
of mass conservation and energy conservation.

25
26 CHAPTER 2. REVIEW OF THERMODYNAMICS

Our system can receive energy, work, etc as long the mass remain constant the
definition is not broken.
Thermodynamics First Law
This law refers to conservation of energy in a non accelerating system. Since all
the systems can be calculated in a non accelerating systems, the conservation is
applied to all systems. The statement describing the law is the following.

Q12 − W12 = E2 − E1 (2.2)

The system energy is a state property. From the first law it directly implies that for
process without heat transfer (adiabatic process) the following is true

W12 = E1 − E2 (2.3)

Interesting results of equation (2.3) is that the way the work is done and/or interme-
diate states are irrelevant to final results. There are several definitions/separations
of the kind of works and they include kinetic energy, potential energy (gravity),
chemical potential, and electrical energy, etc. The internal energy is the energy that
depends on the other properties of the system. For example for pure/homogeneous
and simple gases it depends on two properties like temperature and pressure. The
internal energy is denoted in this book as EU and it will be treated as a state prop-
erty.
The potential energy of the system is depended on the body force. A common
body force is the gravity. For such body force, the potential energy is mgz where
g is the gravity force (acceleration), m is the mass and the z is the vertical height
from a datum. The kinetic energy is

mU 2
K.E. = (2.4)
2

Thus the energy equation can be written as

mU1 2 mU2 2
+ mgz1 + EU 1 + Q = + mgz2 + EU 2 + W
2 2 (2.5)

For the unit mass of the system equation (2.5) is transformed into

U1 2 U2 2
+ gz1 + Eu 1 + q = + gz2 + Eu 2 + w
2 2 (2.6)

where q is the energy per unit mass and w is the work per unit mass. The “new”
internal energy, Eu , is the internal energy per unit mass.
2.1. BASIC DEFINITIONS 27

Since the above equations are true between arbitrary points, choosing any point in
time will make it correct. Thus differentiating the energy equation with respect to
time yields the rate of change energy equation. The rate of change of the energy
transfer is
DQ
= Q̇ (2.7)
Dt
In the same manner, the work change rate transfered through the boundaries of
the system is
DW
= Ẇ (2.8)
Dt
Since the system is with a fixed mass, the rate energy equation is
D EU DU D Bf z
Q̇ − Ẇ = + mU +m (2.9)
Dt Dt Dt
For the case were the body force, Bf , is constant with time like in the case of
gravity equation (2.9) reduced to

D EU DU Dz
Q̇ − Ẇ = + mU + mg
Dt Dt Dt (2.10)

The time derivative operator, D/Dt is used instead of the common notation be-
cause it referred to system property derivative.
Thermodynamics Second Law
There are several definitions of the second law. No matter which definition is used
to describe the second law it will end in a mathematical form. The most common
mathematical form is Clausius inequality which state that
I
δQ
≥0 (2.11)
T
The integration symbol with the circle represent integral of cycle (therefor circle)
in with system return to the same condition. If there is no lost, it is referred as a
reversible process and the inequality change to equality.
I
δQ
=0 (2.12)
T
The last integral can go though several states. These states are independent of
the path the system goes through. Hence, the integral is independent of the path.
This observation leads to the definition of entropy and designated as S and the
derivative of entropy is
µ ¶
δQ
ds ≡ (2.13)
T rev
28 CHAPTER 2. REVIEW OF THERMODYNAMICS

Performing integration between two states results in

Z 2µ ¶ Z 2
δQ
S2 − S1 = = dS (2.14)
1 T rev 1

One of the conclusions that can be drawn from this analysis is for reversible and
adiabatic process dS = 0. Thus, the process in which it is reversible and adiabatic,
the entropy remains constant and referred to as isentropic process. It can be noted
that there is a possibility that a process can be irreversible and the right amount of
heat transfer to have zero change entropy change. Thus, the reverse conclusion
that zero change of entropy leads to reversible process, isn’t correct.
For reversible process equation (2.12) can be written as
δQ = T dS (2.15)
and the work that the system is doing on the surroundings is
δW = P dV (2.16)
Substituting equations (2.15) (2.16) into (2.10) results in
T dS = d EU + P dV (2.17)

Even though the derivation of the above equations were done assuming that there
is no change of kinetic or potential energy, it still remail valid for all situations. Fur-
thermore, it can be shown that it is valid for reversible and irreversible processes.
Enthalpy

It is a common practice to define a new property, which is the combination of al-

ready defined properties, the enthalpy of the system.
H = EU + P V (2.18)
The specific enthalpy is enthalpy per unit mass and denoted as, h.
Or in a differential form as
dH = dEU + dP V + P dV (2.19)

Combining equations (2.18) the (2.17) yields

T dS = dH − V dP (2.20)
For isentropic process, equation (2.17) is reduced to dH = V dP . The equation
(2.17) in mass unit is
dP
T ds = du + P dv = dh − (2.21)
ρ
2.1. BASIC DEFINITIONS 29

when the density enters through the relationship of ρ = 1/v.

Specific Heats
The change of internal energy and enthalpy requires new definitions. The first
change of the internal energy and it is defined as the following
µ ¶
∂Eu
Cv ≡
∂T
(2.22)

And since the change of the enthalpy involve some kind of work is defined as

µ ¶
∂h
Cp ≡
∂T
(2.23)

The ratio between the specific pressure heat and the specific volume heat is called
the ratio of the specific heat and it is denoted as, k.

Cp
k≡ (2.24)
Cv

For solid, the ratio of the specific heats is almost 1 and therefore the difference
between them is almost zero. Commonly the difference for solid is ignored and
both are assumed to be the same and therefore referred as C. This approximation
less strong for liquid but not by that much and in most cases it applied to the
calculations. The ratio the specific heat of gases is larger than one.
Equation of state
Equation of state is a relation between state variables. Normally the relationship of
temperature, pressure, and specific volume define the equation of state for gases.
The simplest equation of state referred to as ideal gas. and it is defined as

P = ρRT (2.25)

Application of Avogadro’s law, that ”all gases at the same pressures and tempera-
tures have the same number of molecules per unit of volume,” allows the calculation
of a “universal gas constant.” This constant to match the standard units results in
kj
R̄ = 8.3145 (2.26)
kmol K

Thus, the specific gas can be calculate as

R̄
R= (2.27)
M
30 CHAPTER 2. REVIEW OF THERMODYNAMICS

Table -2.1: Properties of Various Ideal Gases [300K]

h i h i h i
Gas Chemical Molecular R kj Cv kj
CP kj
k
Formula Weight KgK KgK KgK

Air - 28.970 0.28700 1.0035 0.7165 1.400

Argon Ar 39.948 0.20813 0.5203 0.3122 1.400
Butane C4 H10 58.124 0.14304 1.7164 1.5734 1.091
Carbon CO2 44.01 0.18892 0.8418 0.6529 1.289
Dioxide
Carbon CO 28.01 0.29683 1.0413 0.7445 1.400
Monoxide

Ethane C2 H6 30.07 0.27650 1.7662 1.4897 1.186

Ethylene C2 H4 28.054 0.29637 1.5482 1.2518 1.237
Helium He 4.003 2.07703 5.1926 3.1156 1.667
Hydrogen H2 2.016 4.12418 14.2091 10.0849 1.409
Methane CH4 16.04 0.51835 2.2537 1.7354 1.299
Neon Ne 20.183 0.41195 1.0299 0.6179 1.667

Nitrogen N2 28.013 0.29680 1.0416 0.7448 1.400

Octane C8 H18 114.230 0.07279 1.7113 1.6385 1.044
Oxygen O2 31.999 0.25983 0.9216 0.6618 1.393
Propane C3 H8 44.097 0.18855 1.6794 1.4909 1.327
Steam H2 O 18.015 0.48152 1.8723 1.4108 1.327

The specific constants for select gas at 300K is provided in table 2.1.
From equation (2.25) of state for perfect gas it follows

d(P v) = RdT (2.28)

For perfect gas

dh = dEu + d(P v) = dEu + d(RT ) = f (T ) (only) (2.29)

From the definition of enthalpy it follows that

d(P v) = dh − dEu (2.30)

Utilizing equation (2.28) and subsisting into equation (2.30) and dividing by dT
2.1. BASIC DEFINITIONS 31

yields

Cp − Cv = R (2.31)

This relationship is valid only for ideal/perfect gases.

The ratio of the specific heats can be expressed in several forms as

R
Cv =
k−1 (2.32)

kR
Cp =
k−1 (2.33)

The specific heat ratio, k value ranges from unity to about 1.667. These values
depend on the molecular degrees of freedom (more explanation can be obtained
in Van Wylen “F. of Classical thermodynamics.” The values of several gases can
be approximated as ideal gas and are provided in Table (2.1).
The entropy for ideal gas can be simplified as the following
Z 2µ ¶
dh dP
s2 − s1 = − (2.34)
1 T ρT

Using the identities developed so far one can find that

Z 2 Z 2
dT R dP T2 P2
s2 − s1 = Cp − = Cp ln − R ln (2.35)
1 T 1 P T1 P1

Or using specific heat ratio equation (2.35) transformed into

s2 − s1 k T2 P2
= ln − ln (2.36)
R k − 1 T1 P1
For isentropic process, ∆s = 0, the following is obtained
µ ¶ k−1
T2 P2 k
ln = ln (2.37)
T1 P1

There are several famous identities that results from equation (2.37) as

µ ¶ k−1 µ ¶k−1
T2 P2 k
P2
= =
T1 P1 P1
(2.38)
32 CHAPTER 2. REVIEW OF THERMODYNAMICS

The ideal gas model is a simplified version of the real behavior of real gas. The real
gas has a correction factor to account for the deviations from the ideal gas model.
This correction factor referred as the compressibility factor and defined as

PV
Z= (2.39)
RT
 CHAPTER 3

Fundamentals of Basic Fluid

Mechanics

3.1 Introduction
This chapter is a review of the fundamentals that the student is expected to know.
The basic principles are related to the basic conservation principle. Several terms
will be reviewed such as stream lines. In addition the basic Bernoulli’s equation
will be derived for incompressible flow and later for compressible flow. Several
application of the fluid mechanics will demonstrated. This material is not covered
in the history chapter.

3.2 Fluid Properties

3.3 Control Volume
3.4 Reynold’s Transport Theorem
For simplification the discussion will be focused on one dimensional control volume
and it will be generalzed later. The flow through a stream tube is assumed to be
one-dimensional so that there isn’t any flow except at the tube opening. At the
initial time the mass that was in the tube was m0 . The mass after a very short time
of dt is dm. For simplicity, it is assumed the control volume is a fixed boundary.
The flow on the right through the opening and on the left is assumed to enter the
stream tube while the flow is assumed to leave the stream tube.

33
34 CHAPTER 3. FUNDAMENTALS OF BASIC FLUID MECHANICS

Supposed that the fluid has a property η

¶
dNs Ns (t0 + ∆t) − Ns (t0 )
= lim (3.1)
dt ∆t→0 ∆t
 CHAPTER 4
Speed of Sound

4.1 Motivation
In traditional compressible flow classes there is very little discussion about the
speed of sound outside the ideal gas. The author thinks that this approach has
many shortcomings. In a recent consultation an engineer1 design a industrial sys-
tem that contains converting diverging nozzle with filter to remove small particles
from air. The engineer was well aware of the calculation of the nozzle. Thus, the
engineer was able to predict that was a chocking point. Yet, the engineer was
not ware of the effect of particles on the speed of sound. Hence, the actual flow
rate was only half of his prediction. As it will shown in this chapter, the particles
can, in some situations, reduces the speed of sound by almost as half. With the
“new” knowledge from the consultation the calculations were within the range of
acceptable results.
The above situation is not unique in the industry. It should be expected
that engineers know how to manage this situation of non pure substances (like
clean air). The fact that the engineer knows about the chocking is great but it is not
enough for today’s sophisticated industry2 . In this chapter an introductory discus-
sion is given about different situations which can appear the industry in regards to
speed of sound.

4.2 Introduction
1 Aerospace engineer that alumni of University of Minnesota, Aerospace Department.
2 Pardon, but a joke is must in this situation. A cat is pursuing a mouse and the mouse escape and
hide in the hole. Suddenly, the mouse hear a barking dog and a cat yelling. The mouse go out to
investigate, and cat is catching the mouse. The mouse ask the cat I thought I hear a dog. The cat reply,
yes you right. My teacher was right, one language is not enough today.

35
36 CHAPTER 4. SPEED OF SOUND

The people had recognized for sev-

eral hundred years that sound is sound wave
dU
a variation of pressure. The ears velocity=dU
c

sense the variations by frequency P+dP P

and magnitude which are trans- ρ+dρ ρ

ferred to the brain which translates

to voice. Thus, it raises the ques-
tion: what is the speed of the Fig. -4.1: A very slow moving piston in a still gas
small disturbance travel in a “quiet”
medium. This velocity is referred to as the speed of sound.
To answer this question consider a piston moving from the left to the right
at a relatively small velocity (see Figure 4.1). The information that the piston is
moving passes thorough a single “pressure pulse.” It is assumed that if the velocity
of the piston is infinitesimally small, the pulse will be infinitesimally small. Thus,
the pressure and density can be assumed to be continuous.
In the control volume it is
convenient to look at a control vol-
ume which is attached to a pressure Control volume around
the sound wave

pulse. Applying the mass balance c-dU c

yields P+dP P
ρ+dρ ρ

ρc = (ρ + dρ)(c − dU ) (4.1)

or when the higher term dU dρ is Fig. -4.2: Stationary sound wave and gas moves
neglected yields relative to the pulse.

cdρ
ρdU = cdρ =⇒ dU = (4.2)
ρ
From the energy equation (Bernoulli’s equation), assuming isentropic flow and
neglecting the gravity results

(c − dU )2 − c2 dP
+ =0 (4.3)
2 ρ
neglecting second term (dU 2 ) yield
dP
−cdU + =0 (4.4)
ρ
Substituting the expression for dU from equation (4.2) into equation (4.4) yields
µ ¶
2 dρ dP dP
c = =⇒ c2 = (4.5)
ρ ρ dρ

An expression is needed to represent the right hand side of equation (4.5). For
an ideal gas, P is a function of two independent variables. Here, it is considered
4.3. SPEED OF SOUND IN IDEAL AND PERFECT GASES 37

that P = P (ρ, s) where s is the entropy. The full differential of the pressure can be
expressed as follows:
¯ ¯
∂P ¯¯ ∂P ¯¯
dP = dρ + ds (4.6)
∂ρ ¯s ∂s ¯ρ

In the derivations for the speed of sound it was assumed that the flow is isentropic,
therefore it can be written
¯
dP ∂P ¯¯
= (4.7)
dρ ∂ρ ¯s

Note that the equation (4.5) can be obtained by utilizing the momentum
equation instead of the energy equation.

Example 4.1:
Demonstrate that equation (4.5) can be derived from the momentum equation.

S OLUTION
The momentum equation written for the control volume shown in Figure (4.2) is
P R
F cs
U (ρU dA)
z }| { z }| {
(P + dP ) − P = (ρ + dρ)(c − dU )2 − ρc2 (4.8)

Neglecting all the relative small terms results in

à !
:+∼ 0»»
»» 2
:∼ 0
»
»
2cdU
dP = (ρ + dρ) c −» » dU 2 − ρc2 (4.9)

dP = c2 dρ (4.10)

This yields the same equation as (4.5).

4.3 Speed of sound in ideal and perfect gases

The speed of sound can be obtained easily for the equation of state for an ideal
gas (also perfect gas as a sub set) because of a simple mathematical expression.
The pressure for an ideal gas can be expressed as a simple function of density, ρ,
and a function “molecular structure” or ratio of specific heats, k namely

P = constant × ρk (4.11)
38 CHAPTER 4. SPEED OF SOUND

and hence
P
s z }| {
dP constant × ρk
c= = k × constant × ρk−1 =k×
dρ ρ
P
=k× (4.12)
ρ
Remember that P/ρ is defined for an ideal gas as RT , and equation (4.12) can be
written as
√
c = kRT (4.13)

Example 4.2:
Calculate the speed of sound in water vapor at 20[bar] and 350◦ C, (a) utilizes the
steam table (b) assuming ideal gas.

S OLUTION
The solution can be estimated by using the data from steam table3
s
∆P
c∼ (4.14)
∆ρ
s=constant
h i h i
kg
At 20[bar] and 350◦ C: s = 6.9563 KkJkg ρ = 6.61376 m 3
h i h i
◦ kJ kg
At 18[bar] and 350 C: s = 7.0100 K kg ρ = 6.46956 m3
h i h i
kg
At 18[bar] and 300◦ C: s = 6.8226 KkJkg ρ = 7.13216 m 3

After interpretation of the temperature:

h i h i
kg
At 18[bar] and 335.7◦ C: s ∼ 6.9563 KkJkg ρ ∼ 6.94199 m 3

and substituting into the equation yields

r hmi
200000
c= = 780.5 (4.15)
0.32823 sec

for ideal gas assumption (data taken from Van Wylen and Sontag, Classical Ther-
modynamics, table A 8.)
√ p hmi
c = kRT ∼ 1.327 × 461 × (350 + 273) ∼ 771.5
sec

Note that a better approximation can be done with a steam table, and it
3 This data is taken form Van Wylen and Sontag “Fundamentals of Classical Thermodynamics” 2nd

edition
4.4. SPEED OF SOUND IN REAL GAS 39

Example 4.3:
The temperature in the atmosphere can be assumed to be a linear function of the
height for some distances. What is the time it take for sound to travel from point
“A” to point “B” under this assumption.?

S OLUTION
The temperature is denoted at “A” as TA and temperature in “B” is TB . The distance
between “A” and “B” is denoted as h.
x
T = (TB − TA ) + TA
h
Where the distance x is the variable distance. It should be noted that velocity is
provided as a function of the distance and not the time (another reverse problem).
For an infinitesimal time dt is equal to
dx dx
dt = p =r ³ ´
kRT (x)
kRTA (TBT−T
Ah
A )x
+ 1

integration of the above equation yields

õ ¶ 32 !
2hTA TB
t= √ −1 (4.16)
3 kRTA (TB − TA ) TA

For assumption of constant temperature the time is

h
t= √ (4.17)
kRT̄
Hence the correction factor
r õ ¶ 32 !
tcorrected TA 2 TA TB
= −1 (4.18)
t T̄ 3 (TB − TA ) TA

This correction factor approaches one when TB −→ TA .

4.4 Speed of Sound in Real Gas

The ideal gas model can be improved by introducing the compressibility factor. The
compressibility factor represents the deviation from the ideal gas.
Thus, a real gas equation can be expressed in many cases as

P = zρRT (4.19)

The speed of sound of any gas is provided by equation (4.7). To obtain the ex-
pression for a gas that obeys the law expressed by (4.19) some mathematical ex-
pressions are needed. Recalling from thermodynamics, the Gibbs function (4.20)
40 CHAPTER 4. SPEED OF SOUND

Fig. -4.3: The Compressibility Chart

is used to obtain
dP
T ds = dh − (4.20)
ρ
The definition of pressure specific heat for a pure substance is
µ ¶ µ ¶
∂h ∂s
Cp = =T (4.21)
∂T P ∂T P
The definition of volumetric specific heat for a pure substance is
µ ¶ µ ¶
∂u ∂s
Cv = =T (4.22)
∂T ρ ∂T ρ

From thermodynamics, it can be shown 4

· µ ¶ ¸
∂v
dh = Cp dT + v − T (4.23)
∂T P
4 See Van Wylen p. 372 SI version, perhaps to insert the discussion here.
4.4. SPEED OF SOUND IN REAL GAS 41

The specific volumetric is the inverse of the density as v = zRT /P and thus

µ ¶ à ¡ ¢! µ ¶ µ ¶½ 1
∂v ∂ zRT RT ∂z zR ∂T >
= P
= + ½ (4.24)
∂T ∂T P ∂T P P ∂T ½
P P ½ P

Substituting the equation (4.24) into equation (4.23) results

  v v 
z T
z}|{ µ z}|{
  RT ∂z ¶ zR  
  
dh = Cp dT + v − T  +  dP (4.25)
  P ∂T P P 

Simplifying equation (4.25) to became

· µ ¶ ¸ µ ¶
T v ∂z T ∂z dP
dh = Cp dT − dP = Cp dT − (4.26)
z ∂T P z ∂T P ρ
Utilizing Gibbs equation (4.20)
dh
z }|
µ ¶ { · µ ¶ ¸
T ∂z dP dP dP T ∂z
T ds = Cp dT − − = Cp dT − +1
z ∂T P ρ ρ ρ z ∂T P
zRT
z}|{ · µ ¶ ¸
dP P T ∂z
=Cp dT − +1 (4.27)
P ρ z ∂T P
Letting ds = 0 for isentropic process results in
· µ ¶ ¸
dT dP R ∂z
= z+T (4.28)
T P Cp ∂T P
Equation (4.28) can be integrated by parts. However, it is more convenient to
express dT /T in terms of Cv and dρ/ρ as follows
" µ ¶ #
dT dρ R ∂z
= z+T (4.29)
T ρ Cv ∂T ρ

Equating the right hand side of equations (4.28) and (4.29) results in
" µ ¶ # · µ ¶ ¸
dρ R ∂z dP R ∂z
z+T = z+T (4.30)
ρ Cv ∂T ρ P Cp ∂T P

Rearranging equation (4.30) yields

" ¡ ∂z ¢ #
dρ dP Cv z+T
= ¡∂T
∂z
¢P (4.31)
ρ P Cp z+T ∂T ρ
42 CHAPTER 4. SPEED OF SOUND

If the terms in the braces are constant in the range under interest in this study,
equation (4.31) can be integrated. For short hand writing convenience, n is defined
as
k
z}|{ Ã ¡ ∂z ¢ !
Cp z + T ∂T
n= ¡ ∂z ¢ ρ (4.32)
Cv z + T ∂T P

Note that n approaches k when z → 1 and when z is constant. The integration of

equation (4.31) yields
µ ¶n
ρ1 P1
= (4.33)
ρ2 P2
Equation (4.33) is similar to equation (4.11). What is different in these derivations
is that a relationship between coefficient n and k was established. This relationship
(4.33) isn’t new, and in–fact any thermodynamics book shows this relationship. But
the definition of n in equation (4.32) provides a tool to estimate n. Now, the speed
of sound for a real gas can be obtained in the same manner as for an ideal gas.

dP
= nzRT
dρ
(4.34)

Example 4.4:
Calculate the speed of sound of air at 30◦ C and atmospheric pressure ∼ 1[bar].
The specific heat for air is k = 1.407, n = 1.403, and z = 0.995.
Make the calculation based on the ideal gas model and compare these calculations
to real gas model (compressibility factor). Assume that R = 287[j/kg/K].

S OLUTION
According to the ideal gas model the speed of sound should be
√ √
c = kRT = 1.407 × 287 × 300 ∼ 348.1[m/sec]
For the real gas first coefficient n = 1.403 has
√ √
c = znRT = 1.403 × 0.995times287 × 300 = 346.7[m/sec]

The correction factor for air under normal conditions (atmospheric conditions or
even increased pressure) is minimal on the speed of sound. However, a change
in temperature can have a dramatical change in the speed of sound. For example,
at relative moderate pressure but low temperature common in atmosphere, the
compressibility factor, z = 0.3 and n ∼ 1 which means that speed of sound is only
q
0.3
1.4 about factor of (0.5) to calculated by ideal gas model.
4.5. SPEED OF SOUND IN ALMOST INCOMPRESSIBLE LIQUID 43

4.5 Speed of Sound in Almost Incompressible Liquid

Even liquid normally is assumed to be incompressible in reality has a small and
important compressible aspect. The ratio of the change in the fractional volume
to pressure or compression is referred to as the bulk modulus of the material. For
example, the average bulk modulus for water is 2.2×109 N/m2 . At a depth of about
4,000 meters, the pressure is about 4 × 107 N/m2 . The fractional volume change
is only about 1.8% even under this pressure nevertheless it is a change.
The compressibility of the substance is the reciprocal of the bulk modulus. The
amount of compression of almost all liquids is seen to be very small as given in
Table (4.5). The mathematical definition of bulk modulus as following
dP
B=ρ (4.35)
dρ

In physical terms can be written as

s s
elastic property B
c= =
inertial property ρ
(4.36)

For example for water

s
2.2 × 109 N/m2
c= = 1493m/s
1000kg/m3

This agrees well with the measured speed of sound in water, 1482 m/s at 20◦ C.
Many researchers have looked at this velocity, and for purposes of comparison it is
given in Table (4.5)

Remark reference Value [m/sec]

Fresh Water (20 ◦ C) Cutnell, John D. & Kenneth W. 1492
Johnson. Physics. New York:
Wiley, 1997: 468.
Distilled Water at (25 ◦ C) The World Book Encyclopedia. 1496
Chicago: World Book, 1999. 601
Water distilled Handbook of Chemistry and 1494
Physics. Ohio: Chemical Rubber
Co., 1967-1968:E37

Table -4.1: Water speed of sound from different sources

The effect of impurity and temperature is relatively large, as can be observed from
the equation (4.37). For example, with an increase of 34 degrees from 0◦ C there is
44 CHAPTER 4. SPEED OF SOUND

an increase in the velocity from about 1430 m/sec to about 1546 [m/sec]. According
to Wilson5 , the speed of sound in sea water depends on temperature, salinity, and
hydrostatic pressure.
Wilson’s empirical formula appears as follows:
c(S, T, P ) = c0 + cT + cS + cP + cST P , (4.37)

where c0 = 1449.14[m/sec] is about clean/pure water, cT is a function temperature,

and cS is a function salinity, cP is a function pressure, and cST P is a correction
factor between coupling of the different parameters.

material reference Value [m/sec]

Glycerol 1904
Sea water 25◦ C 1533
Mercury 1450
Kerosene 1324
Methyl alcohol 1143
Carbon tetrachloride 926

Table -4.2: Liquids speed of sound, after Aldred, John, Manual of Sound Recording, Lon-
don: Fountain Press, 1972

In summary, the speed of sound in liquids is about 3 to 5 relative to the speed of

sound in gases.

4.6 Speed of Sound in Solids

The situation with solids is considerably more complicated, with different speeds
in different directions, in different kinds of geometries, and differences between
transverse and longitudinal waves. Nevertheless, the speed of sound in solids is
larger than in liquids and definitely larger than in gases.
Young’s Modulus for a representative value for the bulk modulus for steel is 160
109 N /m2 .
Speed of sound in solid of steel, using a general tabulated value for the bulk mod-
ulus, gives a sound speed for structural steel of

s s
E 160 × 109 N/m2
c= = = 4512m/s (4.38)
ρ 7860Kg/m3

Compared to one tabulated value the example values for stainless steel lays be-
tween the speed for longitudinal and transverse waves.
5 J. Acoust. Soc. Amer., 1960, vol.32, N 10, p. 1357. Wilson’s formula is accepted by the National

Oceanographic Data Center (NODC) USA for computer processing of hydrological information.
4.7. SOUND SPEED IN TWO PHASE MEDIUM 45
material reference Value [m/sec]
Diamond 12000
Pyrex glass 5640
Steel longitudinal wave 5790
Steel transverse shear 3100
Steel longitudinal wave (extensional 5000
wave)
Iron 5130
Aluminum 5100
Brass 4700
Copper 3560
Gold 3240
Lucite 2680
Lead 1322
Rubber 1600

Table -4.3: Solids speed of sound, after Aldred, John, Manual of Sound Recording, Lon-
don:Fountain Press, 1972

4.7 Sound Speed in Two Phase Medium

The gas flow in many industrial situations contains other particles. In actuality,
there could be more than one speed of sound for two phase flow. Indeed there is
double chocking phenomenon in two phase flow. However, for homogeneous and
under certain condition a single velocity can be considered. There can be several
models that approached this problem. For simplicity, it assumed that two materials
are homogeneously mixed. Topic for none homogeneous mixing are beyond the
scope of this book. It further assumed that no heat and mass transfer occurs
between the particles. In that case, three extreme cases suggest themselves:
the flow is mostly gas with drops of the other phase (liquid or solid), about equal
parts of gas and the liquid phase, and liquid with some bubbles. The first case is
analyzed.
The equation of state for the gas can be written as
Pa = ρa RTa (4.39)
The average density can be expressed as
1 ξ 1−ξ
= + (4.40)
ρm ρa ρb
where ξ = ṁṁ is the mass ratio of the materials.
b

For small value of ξ equation (4.40) can be approximated as

ρ
=1+m (4.41)
ρa
46 CHAPTER 4. SPEED OF SOUND
ṁb
where m = ṁ a
is mass flow rate per gas flow rate.
The gas density can be replaced by equation (4.39) and substituted into equation
(4.41)
P R
= T (4.42)
ρ 1+m
A approximation of addition droplets of liquid or dust (solid) results in reduction of
R and yet approximate equation similar to ideal gas was obtained. It must noticed
that m = constant. If the droplets (or the solid particles) can be assumed to have
the same velocity as the gas with no heat transfer or fiction between the particles
isentropic relation can be assumed as
P
= constant (4.43)
ρa k
Assuming that partial pressure of the particles is constant and applying the second
law for the mixture yields
droplets gas
z }| { z }| {
dT dT dP (Cp + mC)dT dP
0 = mC + Cp −R = −R (4.44)
T T P T P
Therefore, the mixture isentropic relationship can be expressed as
γ−1
P γ
= constant (4.45)
T
where
γ−1 R
= (4.46)
γ Cp + mC
Recalling that R = Cp − Cv reduces equation (4.46) into
Cp + mC
γ= (4.47)
Cv + mC
In a way the definition of γ was so chosen that effective specific pressure heat
C +mC
and effective specific volumetric heat are p1+m and Cv1+m +mC
respectively. The
correction factors for the specific heat is not linear.
Since the equations are the same as before hence the familiar equation for speed
of sound can be applied as
p
c = γRmix T
(4.48)
It can be noticed that Rmix and γ are smaller than similar variables in a pure
gas. Hence, this analysis results in lower speed of sound compared to pure gas.
Generally, the velocity of mixtures with large gas component is smaller of the pure
gas. For example, the velocity of sound in slightly wed steam can be about one
third of the pure steam speed of sound.
4.7. SOUND SPEED IN TWO PHASE MEDIUM 47

Meta
For a mixture of two phases, speed of sound can be expressed as

∂P ∂P [f (X)]
c2 = = (4.49)
∂ρ ∂ρ

where X is defined as
s − sf (PB )
X= (4.50)
sf g (PB )

Meta End
48 CHAPTER 4. SPEED OF SOUND
 CHAPTER 5
Isentropic Flow

In this chapter a discussion on a steady

state flow through a smooth and con-
tinuous area flow rate is presented.
A discussion about the flow through
a converging–diverging nozzle is also
part of this chapter. The isentropic flow
PB = P0
models are important because of two
main reasons: One, it provides the infor- P
Subsonic
mation about the trends and important P0
M <1
parameters. Two, the correction factors
can be introduced later to account for Supersonic
deviations from the ideal state. M >1
distance, x

5.1 Stagnation State for Fig. -5.1: Flow of a compressible substance

Ideal Gas Model (gas) through a converging–
diverging nozzle.
5.1.1 General Relationship
It is assumed that the flow is one–
dimensional. Figure (5.1) describes a gas flow through a converging–diverging
nozzle. It has been found that a theoretical state known as the stagnation state
is very useful in simplifying the solution and treatment of the flow. The stagnation
state is a theoretical state in which the flow is brought into a complete motionless
condition in isentropic process without other forces (e.g. gravity force). Several
properties that can be represented by this theoretical process which include tem-
perature, pressure, and density et cetera and denoted by the subscript “0.”

49
50 CHAPTER 5. ISENTROPIC FLOW

First, the stagnation temperature is calculated. The energy conservation

can be written as
U2
h+ = h0 (5.1)
2
Perfect gas is an ideal gas with a constant heat capacity, Cp . For perfect gas
equation (5.1) is simplified into
U2
Cp T + = Cp T0 (5.2)
2
Again it is common to denote T0 as the stagnation temperature. Recalling from
thermodynamic the relationship for perfect gas
R = Cp − Cv (5.3)
and denoting k ≡ Cp ÷ Cv then the thermodynamics relationship obtains the form
kR
Cp = (5.4)
k−1
and where R is a specific constant. Dividing equation (5.2) by (Cp T ) yields
U2 T0
1+ = (5.5)
2Cp T T
Now, substituting c2 = kRT or T = c2 /kR equation (5.5) changes into
kRU 2 T0
1+ 2
= (5.6)
2Cp c T
By utilizing the definition of k by equation (2.24) and inserting it into equation (5.6)
yields
k − 1 U2 T0
1+ 2
= (5.7)
2 c T

It very useful to convert equation (5.6) into a dimensionless form and de-
note Mach number as the ratio of velocity to speed of sound as

U
M≡ (5.8)
c

Inserting the definition of Mach number (5.8) into equation (5.7) reads

T0 k−1 2
=1+ M
T 2 (5.9)
5.1. STAGNATION STATE FOR IDEAL GAS MODEL 51

The usefulness of Mach num- A Q B

ber and equation (5.9) can be demon- T0 T0

strated by this following simple example. P0
ρ0 velocity
P0
ρ0
In this example a gas flows through a
tube (see Figure 5.2) of any shape can
be expressed as a function of only the
stagnation temperature as opposed to Fig. -5.2: Perfect gas flows through a tube
the function of the temperatures and ve-
locities.
The definition of the stagnation state provides the advantage of compact
writing. For example, writing the energy equation for the tube shown in Figure (5.2)
can be reduced to
Q̇ = Cp (T0 B − T0 A )ṁ (5.10)

The ratio of stagnation pressure to the static pressure can be expressed

as the function of the temperature ratio because of the isentropic relationship as
µ ¶ k−1
k µ ¶ k
P0 T0 k − 1 2 k−1
= = 1+ M (5.11)
P T 2

In the same manner the relationship for the density ratio is

µ ¶ k−1
1 µ ¶ 1
ρ0 T0 k − 1 2 k−1
= = 1+ M (5.12)
ρ T 2

A new useful definition is introduced for the case when M = 1 and denoted by
superscript “∗.” The special case of ratio of the star values to stagnation values are
dependent only on the heat ratio as the following:

T∗ c∗ 2 2
= 2 =
T0 c0 k+1
(5.13)

µ ¶ k−1
k
P∗ 2
=
P0 k+1
(5.14)

µ ¶ k−1
1
ρ∗ 2
=
ρ0 k+1
(5.15)
52 CHAPTER 5. ISENTROPIC FLOW

Static Properties As A Function of Mach Number

1

0.9

0.8 P/P0
ρ/ρ0
0.7
T/T0
0.6

0.5

0.4

0.3

0.2

0.1

0
0 1 2 3 4 5 6 7 8 9
Mon Jun 5 17:39:34 2006 Mach number

Fig. -5.3: The stagnation properties as a function of the Mach number, k = 1.4

5.1.2 Relationships for Small Mach Number

Even with today’s computers a simplified method can reduce the tedious work in-
volved in computational work. In particular, the trends can be examined with an-
alytical methods. It further will be used in the book to examine trends in derived
models. It can be noticed that the Mach number involved in the above equations
is in a square power. Hence, if an acceptable error is of about %1 then M < 0.1
provides the desired range. Further, if a higher power is used, much smaller error
results. First it can be noticed that the ratio of temperature to stagnation tempera-
ture, TT0 is provided in power series. Expanding of the equations according to the
binomial expansion of

n(n − 1)x2 n(n − 1)(n − 2)x3

(1 + x)n = 1 + nx + + + ··· (5.16)
2! 3!
will result in the same fashion
P0 (k − 1)M 2 kM 4 2(2 − k)M 6
=1+ + + ··· (5.17)
P 4 8 48
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 53

ρ0 (k − 1)M 2 kM 4 2(2 − k)M 6

=1+ + + ··· (5.18)
ρ 4 8 48

The pressure difference normalized by the velocity (kinetic energy) as correction

factor is
compressibility correction
z }| {
P0 − P M2 (2 − k)M 4
1 =1+ + + ··· (5.19)
2 ρU
2 4 24

From the above equation, it can be observed that the correction factor approaches
zero when M −→ 0 and then equation (5.19) approaches the standard equation
for incompressible flow.
The definition of the star Mach is ratio of the velocity and star speed of
sound at M = 1.

r µ ¶
U∗ k+1 k−1 2
M = ∗ = M 1− M + ··· (5.20)
c 2 4

µ ¶
P0 − P kM 2 M2
= 1+ + ··· (5.21)
P 2 4

µ ¶
ρ0 − ρ M2 kM 2
= 1− + ··· (5.22)
ρ 2 4

The normalized mass rate becomes

s
µ ¶
ṁ kP0 2 M 2 k−1 2
= 1+ M + ··· (5.23)
A RT0 4

The ratio of the area to star area is

µ ¶ 2(k−1)
k+1 µ ¶
A 2 1 k+1 (3 − k)(k + 1) 3
= + M+ M + ··· (5.24)
A∗ k+1 M 4 32

5.2 Isentropic Converging-Diverging Flow in Cross Sec-

tion
54 CHAPTER 5. ISENTROPIC FLOW

The important sub case in this chap-

ter is the flow in a converging–diverging
nozzle. The control volume is shown
T T+dT
ρ ρ+dρ
P P+dP
in Figure (5.4). There are two mod- U U+dU

els that assume variable area flow:

First is isentropic and adiabatic model.
Second is isentropic and isothermal Fig. -5.4: Control volume inside a converging-
model. Clearly, the stagnation tem- diverging nozzle.
perature, T0 , is constant through the
adiabatic flow because there isn’t heat
transfer. Therefore, the stagnation pressure is also constant through the flow be-
cause the flow isentropic. Conversely, in mathematical terms, equation (5.9) and
equation (5.11) are the same. If the right hand side is constant for one variable, it
is constant for the other. In the same argument, the stagnation density is constant
through the flow. Thus, knowing the Mach number or the temperature will provide
all that is needed to find the other properties. The only properties that need to be
connected are the cross section area and the Mach number. Examination of the
relation between properties can then be carried out.

5.2.1 The Properties in the Adiabatic Nozzle

When there is no external work and heat transfer, the energy equation, reads

dh + U dU = 0 (5.25)

Differentiation of continuity equation, ρAU = ṁ = constant, and dividing by the

continuity equation reads

dρ dA dU
+ + =0 (5.26)
ρ A U

The thermodynamic relationship between the properties can be expressed as

dP
T ds = dh − (5.27)
ρ

For isentropic process ds ≡ 0 and combining equations (5.25) with (5.27) yields

dP
+ U dU = 0 (5.28)
ρ

Differentiation of the equation state (perfect gas), P = ρRT , and dividing the
results by the equation of state (ρRT ) yields

dP dρ dT
= + (5.29)
P ρ T
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 55

Obtaining an expression for dU/U from the mass balance equation (5.26) and
using it in equation (5.28) reads
dU
U
z· }|
¸{
dP dA dρ
− U2 + =0 (5.30)
ρ A ρ

Rearranging equation (5.30) so that the density, ρ, can be replaced by the static
pressure, dP/ρ yields
 1

c2
µ ¶  z}|{ 
dP dA dρ dP 
2 dA dρ dP 
=U 2
+ =U  +  (5.31)
ρ A ρ dP A dP ρ 

Recalling that dP/dρ = c2 and substitute the speed of sound into equation (5.31)
to obtain
" µ ¶2 #
dP U dA
1− = U2 (5.32)
ρ c A

Or in a dimensionless form
dP ¡ ¢ dA
1 − M2 = U2 (5.33)
ρ A

Equation (5.33) is a differential equation for the pressure as a function of the cross
section area. It is convenient to rearrange equation (5.33) to obtain a variables
separation form of

ρU 2 dA
dP = (5.34)
A 1 − M2

The pressure Mach number relationship

Before going further in the mathematical derivation it is worth looking at the phys-
ical meaning of equation (5.34). The term ρU 2 /A is always positive (because all
the three terms can be only positive). Now, it can be observed that dP can be pos-
itive or negative depending on the dA and Mach number. The meaning of the sign
change for the pressure differential is that the pressure can increase or decrease.
It can be observed that the critical Mach number is one. If the Mach number is
larger than one than dP has opposite sign of dA. If Mach number is smaller than
56 CHAPTER 5. ISENTROPIC FLOW

one dP and dA have the same sign. For the subsonic branch M < 1 the term
1/(1 − M 2 ) is positive hence

dA > 0 =⇒ dP > 0
dA < 0 =⇒ dP < 0

From these observations the trends are similar to those in incompressible fluid.
An increase in area results in an increase of the static pressure (converting the
dynamic pressure to a static pressure). Conversely, if the area decreases (as a
function of x) the pressure decreases. Note that the pressure decrease is larger in
compressible flow compared to incompressible flow.
For the supersonic branch M > 1, the phenomenon is different. For M > 1
the term 1/1 − M 2 is negative and change the character of the equation.

dA > 0 ⇒ dP < 0
dA < 0 ⇒ dP > 0

This behavior is opposite to incompressible flow behavior.

For the special case of M = 1 (sonic flow) the value of the term 1 − M 2 = 0
thus mathematically dP → ∞ or dA = 0. Since physically dP can increase only
in a finite amount it must that dA = 0.It must also be noted that when M = 1
occurs only when dA = 0. However, the opposite, not necessarily means that
when dA = 0 that M = 1. In that case, it is possible that dM = 0 thus the diverging
side is in the subsonic branch and the flow isn’t choked.
The relationship between the velocity and the pressure can be observed
from equation (5.28) by solving it for dU .

dP
dU = − (5.35)
PU
From equation (5.35) it is obvious that dU has an opposite sign to dP (since the
term P U is positive). Hence the pressure increases when the velocity decreases
and vice versa.
From the speed of sound, one can observe that the density, ρ, increases
with pressure and vice versa (see equation 5.36).

1
dρ = dP (5.36)
c2
It can be noted that in the derivations of the above equations (5.35 - 5.36), the
equation of state was not used. Thus, the equations are applicable for any gas
(perfect or imperfect gas).
The second law (isentropic relationship) dictates that ds = 0 and from
thermodynamics
dT dP
ds = 0 = Cp −R
T P
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 57

and for perfect gas

dT k − 1 dP
= (5.37)
T k P
Thus, the temperature varies according to the same way that pressure does.
The relationship between the Mach number and the temperature can be
obtained by utilizing the fact that the process is assumed to be adiabatic dT0 = 0.
Differentiation of equation (5.9), the relationship between the temperature and the
stagnation temperature becomes
µ ¶
k−1 2
dT0 = 0 = dT 1 + M + T (k − 1)M dM (5.38)
2
and simplifying equation (5.38) yields
dT (k − 1)M dM
=− (5.39)
T 1 + k−1
2 M
2

Relationship Between the Mach Number and Cross Section Area

The equations used in the solution are energy (5.39), second law (5.37), state
(5.29), mass (5.26)1 . Note, equation (5.33) isn’t the solution but demonstration of
certain properties on the pressure.
The relationship between temperature and the cross section area can be
obtained by utilizing the relationship between the pressure and temperature (5.37)
and the relationship of pressure and cross section area (5.33). First stage equation
(5.39) is combined with equation (5.37) and becomes
(k − 1) dP (k − 1)M dM
=− (5.40)
k P 1 + k−1
2 M
2

Combining equation (5.40) with equation (5.33) yields

ρU 2 dA
1 A 1−M 2 M dM
=− (5.41)
k P 1 + k−1
2 M
2

The following identify, ρU 2 = kM P can be proved as

M2
z}|{ P P
2 U 2 z }| { U 2 z }| {
kM P = k 2 ρRT = k ρRT = ρU 2 (5.42)
c kRT
Using the identity in equation (5.42) changes equation (5.41) into
dA M2 − 1
= ¡ ¢ dM (5.43)
A M 1 + k−1
2 M
2

1 The momentum equation is not used normally in isentropic process, why?

58 CHAPTER 5. ISENTROPIC FLOW

Equation (5.43) is very impor-

tant because it relates the geometry M, A
(area) with the relative velocity (Mach
number). In equation (5.43), the factors ss
¡ k−1 2
¢ croon
M 1 + 2 M and A are positive re- ,
A cti
gardless of the values of M or A. There- se
fore, the only factor that affects relation- dA = 0
ship between the cross area and the dx
dM = 0
2
Mach number is M − 1. For M < 1 dx
the Mach number is varied opposite to M, Much nubmer
the cross section area. In the case of
M > 1 the Mach number increases with
the cross section area and vice versa. x
The special case is when M = 1 which
Fig. -5.5: The relationship between the cross
requires that dA = 0. This condition section and the Mach number on
imposes that internal2 flow has to pass the subsonic branch
a converting–diverging device to obtain
supersonic velocity. This minimum
area is referred to as “throat.”
Again, the opposite conclusion that when dA = 0 implies that M = 1 is
not correct because possibility of dM = 0. In subsonic flow branch, from the
mathematical point of view: on one hand, a decrease of the cross section increases
the velocity and the Mach number, on the other hand, an increase of the cross
section decreases the velocity and Mach number (see Figure (5.5)).

5.2.2 Isentropic Flow Examples

Example 5.1:
Air is allowed to flow from a reservoir with temperature of 21◦ C and with pressure
of 5[MPa] through a tube. It was measured that air mass flow rate is 1[kg/sec]. At
some point on the tube static pressure was measured to be 3[MPa]. Assume that
process is isentropic and neglect the velocity at the reservoir, calculate the Mach
number, velocity, and the cross section area at that point where the static pressure
was measured. Assume that the ratio of specific heat is k = Cp /Cv = 1.4.

S OLUTION
The stagnation conditions at the reservoir will be maintained throughout the tube
because the process is isentropic. Hence the stagnation temperature can be writ-
ten T0 = constant and P0 = constant and both of them are known (the condition at
the reservoir). For the point where the static pressure is known, the Mach number
can be calculated by utilizing the pressure ratio. With the known Mach number,
2 This condition does not impose any restrictions for external flow. In external flow, an object can be

moved in arbitrary speed.

5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 59

the temperature, and velocity can be calculated. Finally, the cross section can be
calculated with all these information.
In the point where the static pressure known
P 3[M P a]
P̄ = = = 0.6
P0 5[M P a]
From Table (5.2) or from Figure (5.3) or utilizing the enclosed program, Potto-GDC,
or simply using the equations shows that

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.88639 0.86420 0.69428 1.0115 0.60000 0.60693 0.53105

With these values the static temperature and the density can be calculated.

T = 0.86420338 × (273 + 21) = 254.076K

ρ0
z }| {
ρ P0 5 × 106 [P a]
ρ= = 0.69428839 × h i
ρ0 RT0 J
287.0 kgK × 294[K]
· ¸
kg
= 41.1416
m3
The velocity at that point is
c
z√ }| { √
U = M kRT = 0.88638317 × 1.4 × 287 × 294 = 304[m/sec]

The tube area can be obtained from the mass conservation as

ṁ
A= = 8.26 × 10−5 [m3 ]
ρU
For a circular tube the diameter is about 1[cm].

Example 5.2:
The Mach number at point A on tube is measured to be M = 23 and the static pres-
sure is 2[Bar]4 . Downstream at point B the pressure was measured to be 1.5[Bar].
Calculate the Mach number at point B under the isentropic flow assumption. Also,
estimate the temperature at point B. Assume that the specific heat ratio k = 1.4
and assume a perfect gas model.

4 Thispressure is about two atmospheres with temperature of 250[K]

4 Well,this question is for academic purposes, there is no known way for the author to directly mea-
sure the Mach number. The best approximation is by using inserted cone for supersonic flow and
measure the oblique shock. Here it is subsonic and this technique is not suitable.
60 CHAPTER 5. ISENTROPIC FLOW

S OLUTION
With the known Mach number at point A all the ratios of the static properties to
total (stagnation) properties can be calculated. Therefore, the stagnation pressure
at point A is known and stagnation temperature can be calculated.
At M = 2 (supersonic flow) the ratios are

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
2.0000 0.55556 0.23005 1.6875 0.12780 0.21567 0.59309

With this information the pressure at point B can be expressed as

from the table

5.2 @ M = 2
z}|{
PA PB PA 2.0
= × = 0.12780453 × = 0.17040604
P0 P0 PB 1.5
The corresponding Mach number for this pressure ratio is 1.8137788 and TB =
0.60315132 PPB0 = 0.17040879. The stagnation temperature can be “bypassed” to
calculate the temperature at point B
M =2 M =1.81..
z}|{ z}|{
T0 TB 1
TB = TA × × = 250[K] × × 0.60315132 ' 271.42[K]
TA T0 0.55555556

Example 5.3:
Gas flows through a converging–diverging duct. At point “A” the cross section area
is 50 [cm2 ] and the Mach number was measured to be 0.4. At point B in the duct
the cross section area is 40 [cm2 ]. Find the Mach number at point B. Assume that
the flow is isentropic and the gas specific heat ratio is 1.4.

S OLUTION
To obtain the Mach number at point B by finding the ratio of the area to the critical
area. This relationship can be obtained by

from the Table 5.2

AB AB AA 40 z }| {
= × ∗ = × 1.59014 = 1.272112
A∗ AA A 50

With the value of A

A∗ from the Table (5.2) or from Potto-GDC two solutions can
B

be obtained. The two possible solutions: the first supersonic M = 1.6265306 and
second subsonic M = 0.53884934. Both solution are possible and acceptable. The
supersonic branch solution is possible only if there where a transition at throat
where M=1.
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 61

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
1.6266 0.65396 0.34585 1.2721 0.22617 0.28772
0.53887 0.94511 0.86838 1.2721 0.82071 1.0440

Example 5.4:
Engineer needs to redesign a syringe for medical applications. The complain in
the syringe is that the syringe is “hard to push.” The engineer analyzes the flow
and conclude that the flow is choke. Upon this fact, what engineer should do with
syringe increase the pushing diameter or decrease the diameter? Explain.

S OLUTION
This problem is a typical to compressible flow in the sense the solution is opposite
the regular intuition. The diameter should be decreased. The pressure in the choke
flow in the syringe is past the critical pressure ratio. Hence, the force is a function
of the cross area of the syringe. So, to decrease the force one should decrease
the area.
End Solution

5.2.3 Mass Flow Rate (Number)

One of the important engineering parameters is the mass flow rate which for ideal
gas is
P
ṁ = ρU A = UA (5.44)
RT
This parameter is studied here, to examine the maximum flow rate and to see what
is the effect of the compressibility on the flow rate. The area ratio as a function of
the Mach number needed to be established, specifically and explicitly the relation-
ship for the chocked flow. The area ratio is defined as the ratio of the cross section
at any point to the throat area (the narrow area). It is convenient to rearrange the
equation (5.44) to be expressed in terms of the stagnation properties as
f (M,k)
r r r z r
}| {
ṁ P P0 U k T0 1 P k P T0
= √ √ = √0 M (5.45)
A P0 kRT R T T0 T0 R P0 T
Expressing the temperature in terms of Mach number in equation (5.45) results
in
µ ¶µ ¶− 2(k−1)
k+1
ṁ kM P0 k−1 2
= √ 1+ M
A kRT0 2
(5.46)
It can be noted that equation (5.46) holds everywhere in the converging-diverging
duct and this statement also true for the throat. The throat area can be denoted
62 CHAPTER 5. ISENTROPIC FLOW

as by A∗ . It can be noticed that at the throat when the flow is chocked or in other
words M = 1 and that the stagnation conditions (i.e. temperature, pressure) do
not change. Hence equation (5.46) obtained the form
Ã√ !µ ¶− 2(k−1)
k+1
ṁ kP0 k−1
= √ 1+ (5.47)
A∗ RT0 2

Since the mass flow rate is constant in the duct, dividing equations (5.47) by
equation (5.46) yields

à ! 2(k−1)
k+1
k−1 2
A 1 1+ 2 M
= k+1
A∗ M 2
(5.48)

Equation (5.48) relates the Mach number at any point to the cross section area
ratio.
The maximum flow rate can be expressed either by taking the derivative of
equation (5.47) in with respect to M and equating to zero. Carrying this calculation
results at M = 1.
µ ¶ r µ ¶− 2(k−1)
k+1
ṁ P0 k k+1
√ = (5.49)
A∗ max T0 R 2

For specific heat ratio, k = 1.4

µ ¶
ṁ P 0.68473
∗
√0 ∼ √ (5.50)
A max T0 R

The maximum flow rate for air (R = 287j/kgK) becomes,

√
ṁ T0
= 0.040418 (5.51)
A∗ P 0
Equation (5.51) is known as Fliegner’s Formula on the name of one of the first en-
gineers who observed experimentally the choking phenomenon. It can be noticed
that Fliengner’s equation can lead to definition of the Fliengner’s Number.
c0
zp }| { z F}|n {
√
ṁ T0 ṁ kRT0 ṁc0 1
∗
=√ =√ √ (5.52)
A P0 ∗
kRA P0 ∗
RA P0 k
The definition of Fliengner’s number (Fn) is
ṁc0
Fn ≡ √ (5.53)
RA∗ P0
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 63

Utilizing Fliengner’s number definition and substituting it into equation (5.47)

results in
µ ¶− 2(k−1)
k+1
k−1 2
F n = kM 1 + M
2
(5.54)

and the maximum point for F n at M = 1 is

µ ¶− 2(k−1)
k+1
k+1
Fn = k (5.55)
2

“Naughty Professor” Problems in Isentropic Flow

To explain the material better some instructors invented problems, which have
mostly academic purpose, (see for example, Shapiro (problem 4.5)). While these
problems have a limit applicability in reality, they have substantial academic value
and therefore presented here. The situation where the mass flow rate per area
given with one of the stagnation properties and one of the static properties, e.g.
P0 and T or T0 and P present difficulty for the calculations. The use of the regu-
lar isentropic Table is not possible because there isn’t variable represent this kind
problems. For this kind of problems a new Table was constructed and present
here5 .
The case of T0 and P
This case considered to be simplest case and will first presented here. Using
energy equation (5.9) and substituting for Mach number M = ṁ/Aρc results in
µ ¶2
T0 k−1 ṁ
=1+ (5.56)
T 2 Aρc

Rearranging equation (5.56) result in

1/kR
p
Rzµ }| ¶{ µ ¶2
z}|{ T k − 1 ṁ
2
T0 ρ = T ρ ρ + (5.57)
c2 2 A

And further Rearranging equation (5.57) transformed it into

µ ¶2
Pρ
2 k−1 ṁ
ρ = + (5.58)
T0 R 2kRT0 A
5 Since version 0.44 of this book.
64 CHAPTER 5. ISENTROPIC FLOW

Equation (5.58) is quadratic equation for density, ρ when all other variables are
known. It is convenient to change it into
µ ¶2
2 Pρ k − 1 ṁ
ρ − − =0 (5.59)
T0 R 2kRT0 A
The only physical solution is when the density is positive and thus the only solution
is
 
v
 uµ ¶2 µ ¶2 
1 P u P k − 1 ṁ  
ρ=  +u +2 (5.60)
u
2  RT0 t RT0 kRT0 A 
| {z }
,→(M →0)→0

For almost incompressible flow the density is reduced and the familiar form of
perfect gas model is seen since stagnation temperature is approaching the static
P
temperature for very small Mach number (ρ = RT 0
). In other words, the terms
for the group over the under–brace approaches zero when the flow rate (Mach
number) is very small.
It is convenient to denote a new dimensionless density as
ρ ρRT0 1
ρ̂ = p = = (5.61)
RT0 P T̄
With this new definition equation (5.60) is transformed into
 s 
µ ¶2
1 (k − 1)RT0 ṁ 
ρ̂ = 1+ 1+2 (5.62)
2 kP 2 A

The dimensionless density now is related to a dimensionless group that is a func-

tion of Fn number and Mach number only! Thus, this dimensionless group is func-
tion of Mach number only.
A∗ P 0
F n2 AP =f (M )
z }| {z }| {
µ ¶2 2
µ ¶2 µ ∗ ¶2 µ ¶2
RT0 ṁ 1 c0 ṁ A P0
= (5.63)
P2 A k P0 2 A∗ A P
Thus,
µ ¶2 µ ¶2
RT0 ṁ F n2 A∗ P0
= (5.64)
P2 A k AP
Hence, the dimensionless density is
 s 
µ ¶2
1 (k − 1)F n2 A∗ P0 
ρ̂ = 1+ 1+2 (5.65)
2 k2 AP
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 65

Again notice that the right hand side of equation (5.65) is only function of Mach
number (well, also the specific heat, k). And the values of AAP ∗P
0
were tabulated
in Table (5.2) and Fn is tabulated in the next Table (5.1). Thus, the problems is
reduced to finding tabulated values.
The case of P0 and T
A similar problem can be described for the case of stagnation pressure, P0 ,
and static temperature, T .
First, it is shown that the dimensionless group is a function of Mach number
only (well, again the specific heat ratio, k also).
µ ¶2 µ ¶2 µ ¶µ ¶2
RT ṁ F n2 A∗ P0 T P0
= (5.66)
P0 2 A k AP T0 P

It can be noticed that

µ ¶µ ¶2
F n2 T P0
= (5.67)
k T0 P

Thus equation (5.66) became

µ ¶2 µ ¶2
RT ṁ A∗ P0
= (5.68)
P0 2 A AP

The right hand side is tabulated in the “regular” isentropic Table such (5.2). This
example shows how a dimensional analysis is used to solve a problems without
actually solving any equations. The actual solution of the equation is left as ex-
ercise (this example under construction). What is the legitimacy of this method?
The explanation simply based the previous experience in which for a given ratio of
area or pressure ratio (etcetera) determines the Mach number. Based on the same
arguments, if it was shown that a group of parameters depends only Mach number
than the Mach is determined by this group.
The method of solution for given these parameters is by calculating the PP0 A
A
∗

and then using the table to find the corresponding Mach number.
The case of ρ0 and T or P
The last case sometimes referred to as the “naughty professor’s question” case
dealt here is when the stagnation density given with the static temperature/pressure.
First, the dimensionless approach is used later analytical method is discussed (un-
der construction).

c0 2
µ ¶2 z }| { µ ¶2 µ ¶2 µ ¶
1 ṁ kRT0 ṁ c0 2 ṁ F n2 P0
= = = (5.69)
Rρ0 P A kRP0 P0 PP0 A kRP0 2 PP0 A k P
66 CHAPTER 5. ISENTROPIC FLOW

The last case dealt here is of the stagnation density with static pressure and the
following is dimensionless group

c0 2
µ ¶2 z }| { µ ¶2 µ ¶2 µ ¶
1 ṁ kRT0 T0 ṁ c0 2 T0 ṁ F n2 T0
= = = (5.70)
Rρ0 2 T A kRP0 2 T A kRP0 2 T A k T

It was hidden in the derivations/explanations of the above analysis didn’t explic-

itly state under what conditions these analysis is correct. Unfortunately, not all the
analysis valid for the same conditions and is as the regular “isentropic” Table, (5.2).
The heat/temperature part is valid for enough adiabatic condition while the pres-
sure condition requires also isentropic process. All the above conditions/situations
require to have the perfect gas model as the equation of state. For example the
first “naughty professor” question is sufficient that process is adiabatic only (T0 , P ,
mass flow rate per area.).

Table -5.1: Fliegner’s number and other parameters as a function of Mach number
³ ´2 ¡ ṁ ¢2 ¡ ṁ ¢2 ¡ ṁ ¢2
P0 A∗ RT0 1 1
M Fn ρ̂ AP P2 A Rρ0 P A Rρ0 2 T A
0.00E+001.400E−06 1.000 0.0 0.0 0.0 0.0
0.050001 0.070106 1.000 0.00747 2.62E−05 0.00352 0.00351
0.10000 0.14084 1.000 0.029920 0.000424 0.014268 0.014197
0.20000 0.28677 1.001 0.12039 0.00707 0.060404 0.059212
0.21000 0.30185 1.001 0.13284 0.00865 0.067111 0.065654
0.22000 0.31703 1.001 0.14592 0.010476 0.074254 0.072487
0.23000 0.33233 1.002 0.15963 0.012593 0.081847 0.079722
0.24000 0.34775 1.002 0.17397 0.015027 0.089910 0.087372
0.25000 0.36329 1.003 0.18896 0.017813 0.098460 0.095449
0.26000 0.37896 1.003 0.20458 0.020986 0.10752 0.10397
0.27000 0.39478 1.003 0.22085 0.024585 0.11710 0.11294
0.28000 0.41073 1.004 0.23777 0.028651 0.12724 0.12239
0.29000 0.42683 1.005 0.25535 0.033229 0.13796 0.13232
0.30000 0.44309 1.005 0.27358 0.038365 0.14927 0.14276
0.31000 0.45951 1.006 0.29247 0.044110 0.16121 0.15372
0.32000 0.47609 1.007 0.31203 0.050518 0.17381 0.16522
0.33000 0.49285 1.008 0.33226 0.057647 0.18709 0.17728
0.34000 0.50978 1.009 0.35316 0.065557 0.20109 0.18992
0.35000 0.52690 1.011 0.37474 0.074314 0.21584 0.20316
0.36000 0.54422 1.012 0.39701 0.083989 0.23137 0.21703
0.37000 0.56172 1.013 0.41997 0.094654 0.24773 0.23155
0.38000 0.57944 1.015 0.44363 0.10639 0.26495 0.24674
0.39000 0.59736 1.017 0.46798 0.11928 0.28307 0.26264
0.40000 0.61550 1.019 0.49305 0.13342 0.30214 0.27926
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 67
Table -5.1: Fliegner’s number and other parameters as function of Mach number (continue)
³ ´2 ¡ ṁ ¢2 ¡ ṁ ¢2 ¡ ṁ ¢2
P 0 A∗ RT0 1 1
M Fn ρ̂ AP P2 A Rρ0 P A Rρ0 2 T A
0.41000 0.63386 1.021 0.51882 0.14889 0.32220 0.29663
0.42000 0.65246 1.023 0.54531 0.16581 0.34330 0.31480
0.43000 0.67129 1.026 0.57253 0.18428 0.36550 0.33378
0.44000 0.69036 1.028 0.60047 0.20442 0.38884 0.35361
0.45000 0.70969 1.031 0.62915 0.22634 0.41338 0.37432
0.46000 0.72927 1.035 0.65857 0.25018 0.43919 0.39596
0.47000 0.74912 1.038 0.68875 0.27608 0.46633 0.41855
0.48000 0.76924 1.042 0.71967 0.30418 0.49485 0.44215
0.49000 0.78965 1.046 0.75136 0.33465 0.52485 0.46677
0.50000 0.81034 1.050 0.78382 0.36764 0.55637 0.49249
0.51000 0.83132 1.055 0.81706 0.40333 0.58952 0.51932
0.52000 0.85261 1.060 0.85107 0.44192 0.62436 0.54733
0.53000 0.87421 1.065 0.88588 0.48360 0.66098 0.57656
0.54000 0.89613 1.071 0.92149 0.52858 0.69948 0.60706
0.55000 0.91838 1.077 0.95791 0.57709 0.73995 0.63889
0.56000 0.94096 1.083 0.99514 0.62936 0.78250 0.67210
0.57000 0.96389 1.090 1.033 0.68565 0.82722 0.70675
0.58000 0.98717 1.097 1.072 0.74624 0.87424 0.74290
0.59000 1.011 1.105 1.112 0.81139 0.92366 0.78062
0.60000 1.035 1.113 1.152 0.88142 0.97562 0.81996
0.61000 1.059 1.122 1.194 0.95665 1.030 0.86101
0.62000 1.084 1.131 1.236 1.037 1.088 0.90382
0.63000 1.109 1.141 1.279 1.124 1.148 0.94848
0.64000 1.135 1.151 1.323 1.217 1.212 0.99507
0.65000 1.161 1.162 1.368 1.317 1.278 1.044
0.66000 1.187 1.173 1.414 1.423 1.349 1.094
0.67000 1.214 1.185 1.461 1.538 1.422 1.147
0.68000 1.241 1.198 1.508 1.660 1.500 1.202
0.69000 1.269 1.211 1.557 1.791 1.582 1.260
0.70000 1.297 1.225 1.607 1.931 1.667 1.320
0.71000 1.326 1.240 1.657 2.081 1.758 1.382
0.72000 1.355 1.255 1.708 2.241 1.853 1.448
0.73000 1.385 1.271 1.761 2.412 1.953 1.516
0.74000 1.415 1.288 1.814 2.595 2.058 1.587
0.75000 1.446 1.305 1.869 2.790 2.168 1.661
0.76000 1.477 1.324 1.924 2.998 2.284 1.738
0.77000 1.509 1.343 1.980 3.220 2.407 1.819
0.78000 1.541 1.362 2.038 3.457 2.536 1.903
0.79000 1.574 1.383 2.096 3.709 2.671 1.991
68 CHAPTER 5. ISENTROPIC FLOW
Table -5.1: Fliegner’s number and other parameters as function of Mach number (continue)
³ ´2 ¡ ṁ ¢2 ¡ ṁ ¢2 ¡ ṁ ¢2
P0 A∗ RT0 1 1
M Fn ρ̂ AP P2 A Rρ0 P A Rρ0 2 T A
0.80000 1.607 1.405 2.156 3.979 2.813 2.082
0.81000 1.642 1.427 2.216 4.266 2.963 2.177
0.82000 1.676 1.450 2.278 4.571 3.121 2.277
0.83000 1.712 1.474 2.340 4.897 3.287 2.381
0.84000 1.747 1.500 2.404 5.244 3.462 2.489
0.85000 1.784 1.526 2.469 5.613 3.646 2.602
0.86000 1.821 1.553 2.535 6.006 3.840 2.720
0.87000 1.859 1.581 2.602 6.424 4.043 2.842
0.88000 1.898 1.610 2.670 6.869 4.258 2.971
0.89000 1.937 1.640 2.740 7.342 4.484 3.104
0.90000 1.977 1.671 2.810 7.846 4.721 3.244
0.91000 2.018 1.703 2.882 8.381 4.972 3.389
0.92000 2.059 1.736 2.955 8.949 5.235 3.541
0.93000 2.101 1.771 3.029 9.554 5.513 3.699
0.94000 2.144 1.806 3.105 10.20 5.805 3.865
0.95000 2.188 1.843 3.181 10.88 6.112 4.037
0.96000 2.233 1.881 3.259 11.60 6.436 4.217
0.97000 2.278 1.920 3.338 12.37 6.777 4.404
0.98000 2.324 1.961 3.419 13.19 7.136 4.600
0.99000 2.371 2.003 3.500 14.06 7.515 4.804
1.000 2.419 2.046 3.583 14.98 7.913 5.016

Example 5.5:
A gas flows in the tube with mass flow rate of 0.1 [kg/sec] and tube cross section
is 0.001[m2 ]. The temperature at Chamber supplying the pressure to tube is 27◦ C.
At some point the static pressure was measured to be 1.5[Bar]. Calculate for that
point the Mach number, the velocity, and the stagnation pressure. Assume that the
process is isentropic, k = 1.3, R = 287[j/kgK].

S OLUTION
The first thing that need to be done is to find the mass flow per area and it is
ṁ
= 0.1/0.001 = 100.0[kg/sec/m2 ]
A
It can be noticed that the total temperature is 300K and the static pressure is
1.5[Bar]. The solution is based on section equations (5.60) through (5.65). It is
fortunate that Potto-GDC exist and it can be just plug into it and it provide that

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.17124 0.99562 0.98548 3.4757 0.98116 3.4102 1.5392
5.2. ISENTROPIC CONVERGING-DIVERGING FLOW IN CROSS SECTION 69

The velocity can be calculated as

√ √
U = M c = kRT M = 0.17 × 1.3 × 287 × 300× ∼ 56.87[m/sec]
The stagnation pressure is
P
P0 = = 1.5/0.98116 = 1.5288[Bar]
P/P0

Flow with pressure losses

The expression for the mass flow rate (5.46) is appropriate regardless the flow is
isentropic or adiabatic. That expression was derived based on the theoretical total
pressure and temperature (Mach number) which does not based on the considera-
tions whether the flow is isentropic or adiabatic. In the same manner the definition
of A∗ referred to the theoretical minimum area (”throat area”) if the flow continues
to flow in an isentropic manner. Clearly, in a case where the flow isn’t isentropic or
adiabatic the total pressure and the total temperature will change (due to friction,
and heat transfer). A constant flow rate requires that ṁA = ṁB . Denoting sub-
script A for one point and subscript B for another point mass equation (5.47) can
be equated as
µ ¶µ ¶− 2(k−1)
k−1
kP0 A∗ k−1 2
1+ M = constant (5.71)
RT0 2
From equation (5.71), it is clear that the function f (P0 , T0 , A∗ ) = constant. There
are two possible models that can be used to simplify the calculations. The first
model for neglected heat transfer (adiabatic) flow and in which the total temperature
remained constant (Fanno flow like). The second model which there is significant
heat transfer but insignificant pressure loss (Rayleigh flow like).
If the mass flow rate is constant at any point on the tube (no mass loss occur) then
s
µ ¶ k−1
k+1

∗ k 2
ṁ = A P0 (5.72)
RT0 k + 1
For adiabatic flow, comparison of mass flow rate at point A and point B leads to
P0 A∗ |A = P0 A∗ |B

P0 |A A∗ |
; = ∗A (5.73)
P0 |B A |B
A∗
And utilizing the equality of A∗ = A A leads to
¯
A ¯
P0 |A A∗ MA A|A
= A¯ (5.74)
P0 |B ¯ A|B
A∗ M B
70 CHAPTER 5. ISENTROPIC FLOW

For a flow with a constant stagnation pressure (frictionless flow) and non adiabatic
flow reads
" B¯ #2
¯
T0 |A A∗ MB A|B
= ¯
A ¯
(5.75)
T0 |B A∗ M
A|A
A

Example 5.6:
At point A of the tube the pressure is 3[Bar], Mach number is 2.5, and the duct
section area is 0.01[m2 ]. Downstream at exit of tube, point B, the cross section
area is 0.015[m2 ] and Mach number is 1.5. Assume no mass lost and adiabatic
steady state flow, calculated the total pressure lost.

S OLUTION
Both Mach numbers are known, thus the area ratios can be calculated. The to-
tal pressure can be calculated because the Mach number and static pressure are
known. With these information, and utilizing equation (5.74) the stagnation pres-
sure at point B can be obtained.

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
1.5000 0.68966 0.39498 1.1762 0.27240 0.32039 0.55401
2.5000 0.44444 0.13169 2.6367 0.05853 0.15432 0.62693

First, the stagnation at point A is obtained from Table (5.2) as

¯
¯
¯
¯
¯
P ¯¯ 3
P0 |A = µ ¶ ¯ = = 51.25781291[Bar]
P ¯ 0.058527663
¯
P0 ¯
| {z } ¯¯
M =2.5 A

by utilizing equation (5.74) provides

1.1761671 0.01
P0 |B = 51.25781291 × × ≈ 15.243[Bar]
2.6367187 0.015
Hence
P0 |A − P0 |B = 51.257 − 15.243 = 36.013[Bar]
Note that the large total pressure loss is much larger than the static pressure loss
(Pressure point B the pressure is 0.27240307 × 15.243 = 4.146[Bar]).

5.3 Isentropic Tables

5.3. ISENTROPIC TABLES 71
Table -5.2: Isentropic Table k = 1.4

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.000 1.00000 1.00000 5.8E+5 1.0000 5.8E + 5 2.4E+5
0.050 0.99950 0.99875 11.59 0.99825 11.57 4.838
0.100 0.99800 0.99502 5.822 0.99303 5.781 2.443
0.200 0.99206 0.98028 2.964 0.97250 2.882 1.268
0.300 0.98232 0.95638 2.035 0.93947 1.912 0.89699
0.400 0.96899 0.92427 1.590 0.89561 1.424 0.72632
0.500 0.95238 0.88517 1.340 0.84302 1.130 0.63535
0.600 0.93284 0.84045 1.188 0.78400 0.93155 0.58377
0.700 0.91075 0.79158 1.094 0.72093 0.78896 0.55425
0.800 0.88652 0.73999 1.038 0.65602 0.68110 0.53807
0.900 0.86059 0.68704 1.009 0.59126 0.59650 0.53039
1.00 0.83333 0.63394 1.000 0.52828 0.52828 0.52828
1.100 0.80515 0.58170 1.008 0.46835 0.47207 0.52989
1.200 0.77640 0.53114 1.030 0.41238 0.42493 0.53399
1.300 0.74738 0.48290 1.066 0.36091 0.38484 0.53974
1.400 0.71839 0.43742 1.115 0.31424 0.35036 0.54655
1.500 0.68966 0.39498 1.176 0.27240 0.32039 0.55401
1.600 0.66138 0.35573 1.250 0.23527 0.29414 0.56182
1.700 0.63371 0.31969 1.338 0.20259 0.27099 0.56976
1.800 0.60680 0.28682 1.439 0.17404 0.25044 0.57768
1.900 0.58072 0.25699 1.555 0.14924 0.23211 0.58549
2.000 0.55556 0.23005 1.688 0.12780 0.21567 0.59309
2.500 0.44444 0.13169 2.637 0.058528 0.15432 0.62693
3.000 0.35714 0.076226 4.235 0.027224 0.11528 0.65326
3.500 0.28986 0.045233 6.790 0.013111 0.089018 0.67320
4.000 0.23810 0.027662 10.72 0.00659 0.070595 0.68830
4.500 0.19802 0.017449 16.56 0.00346 0.057227 0.69983
5.000 0.16667 0.011340 25.00 0.00189 0.047251 0.70876
5.500 0.14184 0.00758 36.87 0.00107 0.039628 0.71578
6.000 0.12195 0.00519 53.18 0.000633 0.033682 0.72136
6.500 0.10582 0.00364 75.13 0.000385 0.028962 0.72586
7.000 0.092593 0.00261 1.0E+2 0.000242 0.025156 0.72953
7.500 0.081633 0.00190 1.4E+2 0.000155 0.022046 0.73257
8.000 0.072464 0.00141 1.9E+2 0.000102 0.019473 0.73510
8.500 0.064725 0.00107 2.5E+2 6.90E−5 0.017321 0.73723
9.000 0.058140 0.000815 3.3E+2 4.74E−5 0.015504 0.73903
9.500 0.052493 0.000631 4.2E+2 3.31E−5 0.013957 0.74058
10.00 0.047619 0.000495 5.4E+2 2.36E−5 0.012628 0.74192
72 CHAPTER 5. ISENTROPIC FLOW

5.3.1 Isentropic Isothermal Flow Nozzle

5.3.2 General Relationship
In this section, the other extreme case model where the heat transfer to the gas
is perfect, (e.g. Eckert number is very small) is presented. Again in reality the
heat transfer is somewhere in between the two extremes. So, knowing the two
limits provides a tool to examine where the reality should be expected. The perfect
gas model is again assumed (later more complex models can be assumed and
constructed in a future versions). In isothermal process the perfect gas model
reads

P = ρRT ; dP = dρRT (5.76)

Substituting equation (5.76) into the momentum equation6 yields

RT dP
U dU + =0 (5.77)
P
Integration of equation (5.77) yields the Bernoulli’s equation for ideal gas in isother-
mal process which reads

U2 2 − U1 2 P2
; + RT ln =0 (5.78)
2 P1

Thus, the velocity at point 2 becomes

r
P2
U2 = 2RT ln − U1 2 (5.79)
P1

The velocity at point 2 for stagnation point, U1 ≈ 0 reads

r
P2
U2 = 2RT ln (5.80)
P1

Or in explicit terms of the stagnation properties the velocity is

r
P
U = 2RT ln (5.81)
P0

Transform from equation (5.78) to a dimensionless form becomes

constant constant
µ(M2 2 − M1 2 )
¡
kR¡
T P2
; ¡
µln
= R¡
T (5.82)
2 P1
6 The one dimensional momentum equation for steady state is U dU/dx = −dP/dx+0(other effects)

which are neglected here.

5.3. ISENTROPIC TABLES 73

Simplifying equation (5.82) yields

k(M2 2 − M1 2 ) P2
; = ln (5.83)
2 P1
Or in terms of the pressure ratio equation (5.83) reads
à 2
! k2
P2 k(M1 2 −M2 2 ) eM1
=e 2 = 2 (5.84)
P1 eM2

As oppose to the adiabatic case (T0 = constant) in the isothermal flow the stag-
nation temperature ratio can be expressed

1¡ ¢ ¡ ¢
k−1 2 k−1 2
T0 1 T1¢̧ 1 + 2 M1 1+ 2 M1
= ¢¡ k−1 2
¢ =¡ k−1 2
¢ (5.85)
T0 2 ¢T2 1 + 2 M2 1+ 2 M2

Utilizing conservation of the mass AρM = constant to yield

A1 M2 P2
= (5.86)
A2 M1 P1
Combining equation (5.86) and equation (5.84) yields
à 2
! k2
A2 M1 eM2
= 2 (5.87)
A1 M2 eM1

The change in the stagnation pressure can be expressed as

à ! k−1
k " 2
# k2
k−1 2
P0 2 P2 1+ 2 M2 eM1
= k−1 2 = 2 (5.88)
P0 1 P1 1+ 2 M1 eM1

The critical point, at this stage, is unknown (at what Mach number the nozzle is
choked is unknown) so there are two possibilities: the choking point or M = 1 to
normalize the equation. Here the critical point defined as the point where M = 1
so results can be compared to the adiabatic case and denoted by star. Again it
has to emphasis that this critical point is not really related to physical critical point
but it is arbitrary definition. The true critical point is when flow is choked and the
relationship between two will be presented.
The critical pressure ratio can be obtained from (5.84) to read

P ρ (1−M 2 )k

∗
= ∗ =e 2
P ρ
(5.89)
74 CHAPTER 5. ISENTROPIC FLOW

Equation (5.87) is reduced to obtained the critical area ratio writes

A 1 (1−M 2 )k
∗
= e 2
A M (5.90)

Similarly the stagnation temperature reads

¡ ¢ k
2 k−1
T0 2 1 + k−1
2 M1
=
T0 ∗ k+1 (5.91)

Finally, the critical stagnation pressure reads

à ¡ ¢ ! k−1
k
2
P0 (1−M 2 )k 2 1 + k−1 M1
∗ =e
2 2
P0 k+1
(5.92)

The maximum value of stagnation pressure ratio is obtained when M = 0 at which

is
¯ µ ¶ k−1
k
P0 ¯¯ k 2
=e 2 (5.93)
P0 ∗ ¯M =0 k+1

For specific heat ratio of k = 1.4, this maximum value is about two. It can be noted
that the stagnation pressure is monotonically reduced during this process.
Of course in isothermal process T = T ∗ . All these equations are plotted in Figure
(5.6). From the Figure 5.3 it can be observed that minimum of the curve A/A∗ isn’t
on M = 1. The minimum of the curve is when area is minimum and at the point
where the flow is choked. It should be noted that the stagnation temperature is not
constant as in the adiabatic case and the critical point is the only one constant.
The mathematical procedure to find the minimum is simply taking the derivative
and equating to zero as following
¡ A
¢ k(M 2 −1) k(M 2 −1)
d A∗ kM 2 e 2 −e 2
= =0 (5.94)
dM M2
Equation (5.94) simplified to
1
kM 2 − 1 = 0 ; M = √ (5.95)
k
It can be noticed that a similar results are obtained for adiabatic
√ flow. The velocity
at the throat of isothermal model
√ is smaller by a factor of k. Thus, dividing the
critical adiabatic velocity by k results in
√
Uthroatmax = RT (5.96)
5.3. ISENTROPIC TABLES 75
Isothermal Nozzle
k=14
4
*
P/P
3.5 *
A/A
*
P0 / P0
3 *
T 0 / T0
*
2.5 T/T

1.5

0.5

0
0 0.5 1 1.5 2 2.5 3 3.5 4
M

Tue Apr 5 10:20:36 2005

Fig. -5.6: Various ratios as a function of Mach number for isothermal Nozzle

On the other hand, the pressure loss in adiabatic flow is milder as can be seen in
Figure (5.7(a)).
It should be emphasized that the stagnation pressure decrees. It is convenient to
find expression for the ratio of the initial stagnation pressure (the stagnation pres-
sure before entering the nozzle) to the pressure at the throat. Utilizing equation
(5.89) the following relationship can be obtained
Pthroat P ∗ Pthroat
= =
P0initial P0initial P ∗
„ “ ”2 «
1 1− √1 k

(1−02 )k
e k 2
=
e 2

1
e− 2 = 0.60653 (5.97)

Notice that the critical pressure is independent of the specific heat ratio, k, as
opposed to the adiabatic case. It also has to be emphasized that the stagnation
values of the isothermal model are not constant. Again, the heat transfer is ex-
76 CHAPTER 5. ISENTROPIC FLOW

Isothermal Nozzle Comparison between the two models

k=14 k=14
4 5

4.5 M isoT
3.5
M isentropic
4 Uisntropic/UisoT
3
3.5
2.5
3

2 2.5
*
A / A iso 2
1.5
*
A / A adiabatic
* 1.5
1 P / P iso
*
P / P adiabatic 1
0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2
M Distance (normalized distance two scales)

Tue Apr 5 10:39:06 2005 Thu Apr 7 14:53:49 2005

(a) Comparison between the isothermal (b) The comparison of the adiabatic
nozzle and adiabatic nozzle in various model and isothermal model
variables

Fig. -5.7: The comparison of nozzle flow

pressed as

Q = Cp (T02 − T02 ) (5.98)

5.3. ISENTROPIC TABLES 77
For comparison between
the adiabatic model and the Comparison between the two models
isothermal a simple profile of k=14

nozzle area as a function of 1

the distance is assumed. This
P / P0 isentropic
profile isn’t an ideal profile 0.8 T / T0 isentropic
but rather a simple sample P / P0 isothermal
just to examine the difference 0.6
T/T0 isothermal

between the two models so in

an actual situation it can be 0.4
bounded. To make sense and
eliminate unnecessary details 0.2
the distance from the entrance
to the throat is normalized (to 0
one (1)). In the same fashion 0 0.5 1 1.5
Distance (normalized distance two scales)
2

the distance from the throat to

the exit is normalized (to one Fri Apr 8 15:11:44 2005

(1)) (it doesn’t mean that these

distances are the same). In this Fig. -5.8: Comparison of the pressure and temperature
drop as a function of the normalized length
comparison the entrance area
(two scales)
ratio and the exit area ratio are
the same and equal to 20. The Mach number was computed for the two models
and plotted in Figure (5.7(b)). In this comparison it has to be remembered that
critical area for the two models are different by about 3% (for k = 1.4). As can be
observed from Figure (5.7(b)). The Mach number for the isentropic is larger for
the supersonic branch but the velocity is lower. The ratio of the velocities can be
expressed as
√
Us Ms kRTs
= √ (5.99)
UT MT kRTs

It can be noticed that temperature in the isothermal model is constant while tem-
perature in the adiabatic model can be expressed as a function of the stagnation
temperature. The initial stagnation temperatures are almost the same and can be
canceled out to obtain
Us Ms
∼ q (5.100)
UT MT 1 + k−1 2
2 Ms

By utilizing equation (5.100) the velocity ratio was obtained and is plotted in Figure
(5.7(b)).
Thus, using the isentropic model results in under prediction of the actual results for
the velocity in the supersonic branch. While, the isentropic for the subsonic branch
will be over prediction. The prediction of the Mach number are similarly shown in
Figure (5.7(b)).
78 CHAPTER 5. ISENTROPIC FLOW

Two other ratios need to be examined: temperature and pressure. The initial stag-
nation temperature is denoted as T0 int . The temperature ratio of T /T0 int can be
obtained via the isentropic model as

T 1
= k−1
(5.101)
T0 int 1+ 2 M
2

While the temperature ratio of the isothermal model is constant and equal to one
(1). The pressure ratio for the isentropic model is

P 1
=¡ ¢ k−1 (5.102)
P0 int 1+ k−1 2
2 M
k

and for the isothermal process the stagnation pressure varies and has to be taken
into account as the following:

isentropic
z}|{
∗
Pz P0 P0 z Pz
= (5.103)
P0 int P0 int P0 ∗ P0 z

where z is an arbitrary point on the nozzle. Using equations (5.88) and the isen-
tropic relationship, the sought ratio is provided.
Figure (5.8) shows that the range between the predicted temperatures of the two
models is very large, while the range between the predicted pressure by the two
models is relatively small. The meaning of this analysis is that transferred heat
affects the temperature to a larger degree but the effect on the pressure is much
less significant.
To demonstrate the relativity of the approach advocated in this book consider the
following example.

Example 5.7:
Consider a diverging–converging nozzle made out of wood (low conductive mate-
rial) with exit area equal entrance area. The throat area ratio to entrance area is
1:4 respectively. The stagnation pressure is 5[Bar] and the stagnation temperature
is 27◦ C. Assume that the back pressure is low enough to have supersonic flow
without shock and k = 1.4. Calculate the velocity at the exit using the adiabatic
model. If the nozzle was made from copper (a good heat conductor) a larger heat
transfer occurs, should the velocity increase or decrease? What is the maximum
possible increase?

S OLUTION
The first part of the question deals with the adiabatic model i.e. the conservation
of the stagnation properties. Thus, with known area ratio and known stagnation
Potto–GDC provides the following table:
5.4. THE IMPULSE FUNCTION 79

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
0.14655 0.99572 0.98934 4.0000 0.98511 3.9405
2.9402 0.36644 0.08129 4.0000 0.02979 0.11915
With the known Mach number and temperature at the exit, the velocity can be
calculated. The exit temperature is 0.36644 × 300 = 109.9K. The exit velocity, then,
is √ √
U = M kRT = 2.9402 1.4 × 287 × 109.9 ∼ 617.93[m/sec]
Even for the isothermal model, the initial stagnation temperature is given as 300K.
Using the area ratio in Figure (5.6) or using the Potto–GDC obtains the following
table
T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
1.9910 1.4940 0.51183 4.0000 0.12556 0.50225
The exit Mach number is known and the initial temperature to the throat tempera-
ture ratio can be calculated as the following:
T0ini 1 1
∗ = k−1 1
= = 0.777777778
T0 1+ 2 k 1 + k−1
k

Thus the stagnation temperature at the exit is

T0ini
= 1.4940/0.777777778 = 1.921
T0exit
The exit stagnation temperature is 1.92 × 300 = 576.2K. The exit velocity can be
determined by utilizing the following equation
√ √
Uexit = M kRT = 1.9910 1.4 × 287 × 300.0 = 691.253[m/sec]

As was discussed before, the velocity in the copper nozzle will be larger than the
velocity in the wood nozzle. However, the maximum velocity cannot exceed the
691.253[m/sec]

5.4 The Impulse Function

5.4.1 Impulse in Isentropic Adiabatic Nozzle
One of the functions that is used in calculating the forces is the Impulse function.
The Impulse function is denoted here as F , but in the literature some denote this
function as I. To explain the motivation for using this definition consider the calcu-
lation of the net forces that acting on section shown in Figure (5.9). To calculate
the net forces acting in the x–direction the momentum equation has to be applied

Fnet = ṁ(U2 − U1 ) + P2 A2 − P1 A1 (5.104)

80 CHAPTER 5. ISENTROPIC FLOW

The net force is denoted here as Fnet . The mass conservation also can be applied
to our control volume
ṁ = ρ1 A1 U1 = ρ2 A2 U2 (5.105)
Combining equation (5.104) with equation (5.105) and by utilizing the identity in
equation (5.42) results in
Fnet = kP2 A2 M2 2 − kP1 A1 M1 2 + P2 A2 − P1 A1 (5.106)
Rearranging equation (5.106) and dividing it by P0 A∗ results in
f (M2 ) f (M1 )
z }| { z f (M
}|
2) z }| { f (M1 )
Fnet P2 A2 ¡ 2
¢{ P1 A1 ¡z }|
2
¢{
= 1 + kM 2 − 1 + kM 1 (5.107)
P0 A∗ P0 A∗ P0 A∗

Examining equation (5.107) shows that

the right hand side is only a function of x-direction
Mach number and specific heat ratio, k.
Hence, if the right hand side is only a
function of the Mach number and k than
the left hand side must be function of
only the same parameters, M and k.
Defining a function that depends only
Fig. -5.9: Schematic to explain the signifi-
on the Mach number creates the con- cances of the Impulse function
venience for calculating the net forces
acting on any device. Thus, defining the Impulse function as
¡ ¢
F = P A 1 + kM2 2 (5.108)
In the Impulse function when F (M = 1) is denoted as F ∗
F ∗ = P ∗ A∗ (1 + k) (5.109)
The ratio of the Impulse function is defined as
see function (5.107)
¡ ¢ z }| {
F P1 A1 1 + kM1 2 1 P1 A1 ¡ 2
¢ 1
= ∗ ∗ = 1 + kM1 (5.110)
F∗ P A (1 + k) P∗ P0 A∗ (1 + k)
P0
|{z}
k
( k+1
2
) k−1
This ratio is different only in a coefficient from the ratio defined in equation (5.107)
which makes the ratio a function of k and the Mach number. Hence, the net force
is
µ ¶ k µ ¶
k + 1 k−1 F2 F1
Fnet = P0 A∗ (1 + k) − (5.111)
2 F∗ F∗
5.4. THE IMPULSE FUNCTION 81

To demonstrate the usefulness of the this function consider a simple situation of

the flow through a converging nozzle

Example 5.8:
1
Consider a flow of gas into a 2
converging nozzle with a mass
ṁ = 1[kg/sec]
flow rate of 1[kg/sec] and the
A1 = 0.009m2
entrance area is 0.009[m2 ] and A2 = 0.003m2
2 T0 = 400K
the exit area is 0.003[m ]. The P2 = 50[Bar]

stagnation temperature is 400K

and the pressure at point 2 was
measured as 5[Bar] Calculate
the net force acting on the noz- Fig. -5.10: Schematic of a flow of a compressible sub-
zle and pressure at point 1. stance (gas) thorough a converging nozzle
for example (5.8)
S OLUTION
The solution is obtained by getting the data for the Mach number. To obtained the
Mach number, the ratio of P1 A1 /A∗ P0 is needed to be calculated. To obtain this
ratio the denominator is needed to be obtained. Utilizing Fliegner’s equation (5.51),
provides the following
√ √
∗ ṁ RT 1.0 × 400 × 287
A P0 = = ∼ 70061.76[N ]
0.058 0.058
and
A2 P2 500000 × 0.003
∗
= ∼ 2.1
A P0 70061.76

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.27353 0.98526 0.96355 2.2121 0.94934 2.1000 0.96666

A
With the area ratio of A? = 2.2121 the area ratio of at point 1 can be calculated.
A1 A2 A1 0.009
= ? = 2.2121 × = 5.2227
A? A A2 0.003
And utilizing again Potto-GDC provides

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.11164 0.99751 0.99380 5.2227 0.99132 5.1774 2.1949

The pressure at point 1 is

P0 P1
P1 = P2 = 5.0times0.94934/0.99380 ∼ 4.776[Bar]
P2 P0
82 CHAPTER 5. ISENTROPIC FLOW

The net force is obtained by utilizing equation (5.111)

µ ¶ k µ ¶
P0 A∗ k + 1 k−1 F2 F1
Fnet = P2 A2 (1 + k) − ∗
P2 A2 2 F∗ F
1 3.5
= 500000 × × 2.4 × 1.2 × (2.1949 − 0.96666) ∼ 614[kN ]
2.1

5.4.2 The Impulse Function in Isothermal Nozzle

Previously Impulse function was developed in the isentropic adiabatic flow. The
same is done here for the isothermal nozzle flow model. As previously, the defi-
nition of the Impulse function is reused. The ratio of the impulse function for two
points on the nozzle is

F2 P2 A2 + ρ2 U2 2 A2
= (5.112)
F1 P1 A1 + ρ1 U1 2 A1

Utilizing the ideal gas model for density and some rearrangement results in
U2 2
F2 P2 A2 1 + RT
= U1 2
(5.113)
F1 P1 A1 1 +
RT

Since U 2 /RT = kM 2 and the ratio of equation (5.86) transformed equation into
(5.113)

F2 M1 1 + kM2 2
= (5.114)
F1 M2 1 + kM1 2

At the star condition (M = 1) (not the minimum point) results in

F2 1 1 + kM2 2
∗
=
F M2 1 + k
(5.115)

5.5 Isothermal Table

Table -5.3: Isothermal Table

T0 P0 A P A×P F
M T0 ? P0 ? A? P? A∗ ×P0 F∗
0.00 0.52828 1.064 5.0E + 5 2.014 1.0E+6 4.2E+5
5.5. ISOTHERMAL TABLE 83
Table -5.3: Isothermal Table (continue)

T0 P0 A P A×P F
M T0 ? P0 ? A? P? A∗ ×P0 F∗
0.05 0.52921 1.064 9.949 2.010 20.00 8.362
0.1 0.53199 1.064 5.001 2.000 10.00 4.225
0.2 0.54322 1.064 2.553 1.958 5.000 2.200
0.3 0.56232 1.063 1.763 1.891 3.333 1.564
0.4 0.58985 1.062 1.389 1.800 2.500 1.275
0.5 0.62665 1.059 1.183 1.690 2.000 1.125
0.6 0.67383 1.055 1.065 1.565 1.667 1.044
0.7 0.73278 1.047 0.99967 1.429 1.429 1.004
0.8 0.80528 1.036 0.97156 1.287 1.250 0.98750
0.9 0.89348 1.021 0.97274 1.142 1.111 0.98796
1.00 1.000 1.000 1.000 1.000 1.000 1.000
1.10 1.128 0.97376 1.053 0.86329 0.90909 1.020
1.20 1.281 0.94147 1.134 0.73492 0.83333 1.047
1.30 1.464 0.90302 1.247 0.61693 0.76923 1.079
1.40 1.681 0.85853 1.399 0.51069 0.71429 1.114
1.50 1.939 0.80844 1.599 0.41686 0.66667 1.153
1.60 2.245 0.75344 1.863 0.33554 0.62500 1.194
1.70 2.608 0.69449 2.209 0.26634 0.58824 1.237
1.80 3.035 0.63276 2.665 0.20846 0.55556 1.281
1.90 3.540 0.56954 3.271 0.16090 0.52632 1.328
2.00 4.134 0.50618 4.083 0.12246 0.50000 1.375
2.50 9.026 0.22881 15.78 0.025349 0.40000 1.625
3.000 19.41 0.071758 90.14 0.00370 0.33333 1.889
3.500 40.29 0.015317 7.5E + 2 0.000380 0.28571 2.161
4.000 80.21 0.00221 9.1E + 3 2.75E−5 0.25000 2.438
4.500 1.5E + 2 0.000215 1.6E + 5 1.41E−6 0.22222 2.718
5.000 2.8E + 2 1.41E−5 4.0E + 6 0.0 0.20000 3.000
5.500 4.9E + 2 0.0 1.4E + 8 0.0 0.18182 3.284
6.000 8.3E + 2 0.0 7.3E + 9 0.0 0.16667 3.569
6.500 1.4E + 3 0.0 5.3E+11 0.0 0.15385 3.856
7.000 2.2E + 3 0.0 5.6E+13 0.0 0.14286 4.143
7.500 3.4E + 3 0.0 8.3E+15 0.0 0.13333 4.431
8.000 5.2E + 3 0.0 1.8E+18 0.0 0.12500 4.719
8.500 7.7E + 3 0.0 5.4E+20 0.0 0.11765 5.007
9.000 1.1E + 4 0.0 2.3E+23 0.0 0.11111 5.296
9.500 1.6E + 4 0.0 1.4E+26 0.0 0.10526 5.586
10.00 2.2E + 4 0.0 1.2E+29 0.0 0.100000 5.875
84 CHAPTER 5. ISENTROPIC FLOW

5.6 The effects of Real Gases

To obtained expressions for non–ideal gas it is commonly done by reusing the ideal
gas model and introducing a new variable which is a function of the gas properties
like the critical pressure and critical temperature. Thus, a real gas equation can
be expressed in equation (4.19). Differentiating equation (4.19) and dividing by
equation (4.19) yields

dP dz dρ dT
= + + (5.116)
P z ρ T

Again, Gibb’s equation (5.27) is reused to related the entropy change to the
change in thermodynamics properties and applied on non-ideal gas. Since ds = 0
and utilizing the equation of the state dh = dP/ρ. The enthalpy is a function of the
temperature and pressure thus, h = h(T, P ) and full differential is
µ ¶ µ ¶
∂h ∂h
dh = dT + dP (5.117)
∂T P ∂P T
∂h
The definition of pressure specific heat is Cp ≡ ∂T and second derivative is
Maxwell relation hence,
µ ¶ µ ¶
∂h ∂s
=v−T (5.118)
∂P T ∂T P

First, the differential of enthalpy is calculated for real gas equation of state as
µ ¶µ ¶
T ∂z dP
dh = Cp dT − (5.119)
Z ∂T P ρ

Equations (5.27) and (4.19) are combined to form

· µ ¶µ ¶ ¸
ds Cp dT T ∂z dP
= −z 1+ (5.120)
R R T Z ∂T P P

The mechanical energy equation can be expressed as

Z µ 2¶ Z
U dP
d =− (5.121)
2 ρ

At the stagnation the definition requires that the velocity is zero. To carry the
integration of the right hand side the relationship between the pressure and the
density has to be defined. The following power relationship is assumed
µ ¶ n1
ρ P
= (5.122)
ρ0 P0
5.6. THE EFFECTS OF REAL GASES 85

Notice, that for perfect gas the n is substituted by k. With integration of equation
(5.121) when using relationship which is defined in equation (5.122) results
Z P1 Z P µ ¶ n1
U2 dP 1 P0
= = dP (5.123)
2 P0 ρ P0 ρ0 P

Substituting relation for stagnation density (4.19) results

Z P µ ¶ n1
U2 z0 RT0 P0
= dP (5.124)
2 P0 P0 P

For n > 1 the integration results in

v " #
u µ ¶( n−1
u n )
t 2n P
U = z0 RT0 1− (5.125)
n−1 P0

For n = 1 the integration becomes

s µ ¶
P0
U= 2z0 RT0 ln (5.126)
P

It must be noted that n is a function of the critical temperature and critical pressure.
The mass flow rate is regardless to equation of state as following

ṁ = ρ∗ A∗ U ∗ (5.127)

Where ρ∗ is the density at the throat (assuming the chocking condition) and A∗ is
the cross area of the throat. Thus, the mass flow rate in our properties
U∗
ρ∗ z }| {
z }| {v "
µ ¶ n1 u µ ¶ n−1 #
P0 P u ( )
tz RT 2n 1 − P
n
ṁ = A∗ 0 0 (5.128)
z0 RT0 P0 n−1 P0

For the case of n = 1

ρ∗ U ∗∗
z }| { zs }| {
µ ¶ 1 µ ¶
P0 P n
P0
ṁ = A∗ 2z0 RT0 ln (5.129)
z0 RT0 P0 P

The Mach number can be obtained by utilizing equation (4.34) to defined the Mach
number as
U
M=√ (5.130)
znRT
86 CHAPTER 5. ISENTROPIC FLOW

Integrating equation (5.120) when ds = 0 results

Z T2 Z P2 µ µ ¶µ ¶ ¶
Cp dT T ∂z dP
= z 1+ (5.131)
T1 R T P1 Z ∂T P P

To carryout the integration of equation (5.131) looks at Bernnolli’s equation which

is
Z Z
dU 2 dP
=− (5.132)
2 ρ

After integration of the velocity

Z P/P0 µ ¶
dU 2 ρ0 P
=− d (5.133)
2 1 ρ P0

It was shown in Chapter (4) that (4.33) is applicable for some ranges of relative
temperature and pressure (relative to critical temperature and pressure and not the
stagnation conditions).
v
u
u µ ¶" µ ¶ n−1 #
2n P n
U = tz0 RT0 1− (5.134)
n−1 P0

When n = 1 or when n → 1
s µ ¶
P0
U= 2z0 RT0 ln (5.135)
P

The mass flow rate for the real gas ṁ = ρ∗ U ∗ A∗

r µ ¶ n1 · ¸
A∗ P0 2n P∗ P∗
ṁ = √ 1− (5.136)
z0 RT0 n−1 P0 P0

And for n = 1
r s µ ¶
A∗ P0 2n P0
ṁ = √ 2z0 RT0 ln (5.137)
z0 RT0 n−1 P

Fliegner’s number in this case is

r µ ¶ n1 · ¸
ṁc0 2n P∗ P∗
Fn = ∗ 1− (5.138)
A P0 n−1 P0 P0
5.6. THE EFFECTS OF REAL GASES 87

Fliegner’s number for n = 1 is

µ ¶2 µ ¶
ṁc0 P∗ P∗
Fn = =2 − ln (5.139)
A∗ P0 P0 P0
The critical ratio of the pressure is
µ ¶ n−1
n
P∗ 2
= (5.140)
P0 n+1
When n = 1 or more generally when n → 1 this is a ratio approach
P∗ √
= e (5.141)
P0

To obtain the relationship between the temperature and pressure, equation (5.131)
can be integrated
µ ¶ CR [z+T ( ∂T
∂z
)P ]
T0 P0 p
= (5.142)
T P

The power of the pressure ratio is approaching k−1k when z approaches 1. Note
that

T0 ³ z ´ µ P ¶ 1−n
n
0 0
= (5.143)
T z P

The Mach number at every point at the nozzle can be expressed as

v "
uµ
u ¶ µ ¶ 1−n #
t 2 z0 T0 P −0 n
M= 1− (5.144)
n−1 z T P

For n = 1 the Mach number is

r
z0 T0 P0
M= 2 ln (5.145)
z T P
The pressure ratio at any point can be expressed as a function of the Mach number
as
· ¸ n−1 z+T ( ∂T
n − 1 2 ( n )[ )P ]
∂z
T0
= 1+ M (5.146)
T 2
for n = 1
T0
eM [z+T ( )P ]
2 ∂z
∂T
= (5.147)
T
88 CHAPTER 5. ISENTROPIC FLOW

The critical temperature is given by

µ ¶ n z+T ( ∂T
1 + n ( 1−n )[ )P ]
∂z
T∗
= (5.148)
T0 2
and for n = 1
q
T∗
e−[z+T ( )P ]
∂z
∂T
= (5.149)
T0

The mass flow rate as a function of the Mach number is

s
µ ¶ n+1
P0 n n − 1 2 n−1
ṁ = M 1+ M (5.150)
c0 2
For the case of n = 1 the mass flow rate is
q s
µ ¶ n+1
P0 A∗ n n − 1 2 n−1
ṁ =
c0
M2
e 1+
2
M (5.151)

Example 5.9:
A design is required that at a specific point the Mach number should be M = 2.61,
the pressure 2[Bar], and temperature 300K.
i. Calculate the area ratio between the point and the throat.
ii. Calculate the stagnation pressure and the stagnation temperature.
iii. Are the stagnation pressure and temperature at the entrance different from the
point? You can assume that k = 1.405.

S OLUTION

1. The solution is simplified by using Potto-GDC for M = 2.61 the results are

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
2.6100 0.42027 0.11761 2.9066 0.04943 0.14366

2. The stagnation pressure is obtained from

P0 2.61
P0 = P = ∼ 52.802[Bar]
P 0.04943
The stagnation temperature is
T0 300
T0 = T = ∼ 713.82K
T 0.42027
3. Of course, the stagnation pressure is constant for isentropic flow.
 CHAPTER 6
Normal Shock

In this chapter the relationships between the two sides of normal shock are pre-
sented. In this discussion, the flow is assumed to be in a steady state, and the
thickness of the shock is assumed to be very small. A discussion on the shock
thickness will be presented in a forthcoming section1 .
A shock can occur in at
least two different mechanisms.
The first is when a large differ-
direction x
flow y
ence (above a small minimum
value) between the two sides Px Py
of a membrane, and when the Tx Ty
membrane bursts (see the dis-
cussion about the shock tube). c.v.
Of course, the shock travels
from the high pressure to the Fig. -6.1: A shock wave inside a tube, but it can also be
low pressure side. The sec- viewed as a one–dimensional shock wave.
ond is when many sound waves
“run into” each other and accumulate (some refer to it as “coalescing”) into a large
difference, which is the shock wave. In fact, the sound wave can be viewed as
an extremely weak shock. In the speed of sound analysis, it was assumed the
medium is continuous, without any abrupt changes. This assumption is no longer
valid in the case of a shock. Here, the relationship for a perfect gas is constructed.
In Figure (6.1) a control volume for this analysis is shown, and the gas
flows from left to right. The conditions, to the left and to the right of the shock, are

1 Currently under construction.

89
90 CHAPTER 6. NORMAL SHOCK

assumed to be uniform2 . The conditions to the right of the shock wave are uniform,
but different from the left side. The transition in the shock is abrupt and in a very
narrow width.
The chemical reactions (even condensation) are neglected, and the shock
occurs at a very narrow section. Clearly, the isentropic transition assumption is not
appropriate in this case because the shock wave is a discontinued area. Therefore,
the increase of the entropy is fundamental to the phenomenon and the understand-
ing of it.
It is further assumed that there is no friction or heat loss at the shock
(because the heat transfer is negligible due to the fact that it occurs on a relatively
small surface). It is customary in this field to denote x as the upstream condition
and y as the downstream condition.
The mass flow rate is constant from the two sides of the shock and there-
fore the mass balance is reduced to
ρx Ux = ρy Uy (6.1)

In a shock wave, the momentum is the quantity that remains constant be-
cause there are no external forces. Thus, it can be written that
¡ ¢
Px − Py = ρx Uy 2 − ρy Ux 2 (6.2)
The process is adiabatic, or nearly adiabatic, and therefore the energy equation
can be written as
Ux 2 Uy 2
Cp Tx + = Cp Ty + (6.3)
2 2
The equation of state for perfect gas reads
P = ρRT (6.4)

If the conditions upstream are known, then there are four unknown con-
ditions downstream. A system of four unknowns and four equations is solvable.
Nevertheless, one can note that there are two solutions because of the quadratic
of equation (6.3). These two possible solutions refer to the direction of the flow.
Physics dictates that there is only one possible solution. One cannot deduce the
direction of the flow from the pressure on both sides of the shock wave. The only
tool that brings us to the direction of the flow is the second law of thermodynamics.
This law dictates the direction of the flow, and as it will be shown, the gas flows
from a supersonic flow to a subsonic flow. Mathematically, the second law is ex-
pressed by the entropy. For the adiabatic process, the entropy must increase. In
mathematical terms, it can be written as follows:
sy − sx > 0 (6.5)
2 Clearly the change in the shock is so significant compared to the changes in medium before and

after the shock that the changes in the mediums (flow) can be considered uniform.
 91

Note that the greater–equal signs were not used. The reason is that the process
is irreversible, and therefore no equality can exist. Mathematically, the parameters
are P, T, U, and ρ, which are needed to be solved. For ideal gas, equation (6.5) is

Ty Py
ln − (k − 1) >0 (6.6)
Tx Px

It can also be noticed that entropy, s, can be expressed as a function of

the other parameters. Now one can view these equations as two different subsets
of equations. The first set is the energy, continuity, and state equations, and the
second set is the momentum, continuity, and state equations. The solution of ev-
ery set of these equations produces one additional degree of freedom, which will
produce a range of possible solutions. Thus, one can have a whole range of solu-
tions. In the first case, the energy equation is used, producing various resistance
to the flow. This case is called Fanno flow, and Chapter (10) deals extensively with
this topic. The mathematical explanation is given Chapter (10) in greater detail. In-
stead of solving all the equations that were presented, one can solve only four (4)
equations (including the second law), which will require additional parameters. If
the energy, continuity, and state equations are solved for the arbitrary value of the
Ty , a parabola in the T − −s diagram will be obtained. On the other hand, when
the momentum equation is solved instead of the energy equation, the degree of
freedom is now energy, i.e., the energy amount “added” to the shock. This situa-
tion is similar to a frictionless flow with the addition of heat, and this flow is known
as Rayleigh flow. This flow is dealt with in greater detail in Chapter (11).
Since the shock has
no heat transfer (a special M < 1 M = p1 subsonic
case of Rayleigh flow) and k flow supersonic
there isn’t essentially any Ty ; ; P y ; sy flow
momentum transfer (a spe- T
M =1
cial case of Fanno flow), shock jump

the intersection of these M =1

two curves is what really Fanno
Rayleigh
line
happened in the shock. In line
Figure (6.2), the intersec-
tion is shown and two solu-
M >1
Tx ; Px ; sx
tions are obtained. Clearly,
the increase of the entropy
determines the direction of s
the flow. The entropy in-
creases from point x to Fig. -6.2: The intersection of Fanno flow and Rayleigh flow
point y. It is also worth not- produces two solutions for the shock wave.
ing that the temperature at M = 1 on Rayleigh flow is larger than that on the Fanno
line.
92 CHAPTER 6. NORMAL SHOCK

6.1 Solution of the Governing Equations

6.1.1 Informal Model
Accepting the fact that the shock is adiabatic or nearly adiabatic requires that total
energy is conserved, T0 x = T0 y . The relationship between the temperature and
the stagnation temperature provides the relationship of the temperature for both
sides of the shock.
Ty 2
k−1
Ty T0 y 1+ 2 Mx
= Tx
= k−1 2 (6.7)
Tx T0 x 1+ 2 My

All the other relationships are essentially derived from this equation. The
only issue left to derive is the relationship between Mx and My . Note that the Mach
number is a function of temperature, and thus for known Mx all the other quantities
can be determined, at least, numerically. The analytical solution is discussed in
the next section.

6.1.2 Formal Model

Equations (6.1), (6.2), and (6.3) can be converted into a dimensionless form. The
reason that dimensionless forms are heavily used in this book is because by doing
so it simplifies and clarifies the solution. It can also be noted that in many cases
the dimensionless equations set is more easily solved.
From the continuity equation (6.1) substituting for density, ρ, the equation
of state yields
Px Py
Ux = Uy (6.8)
RTx RTy
Squaring equation (6.8) results in
Px 2 2 Py 2
Ux = Uy 2 (6.9)
R2 Tx 2 R2 Ty 2
Multiplying the two sides by the ratio of the specific heat, k, provides a way to
obtain the speed of sound definition/equation for perfect gas, c2 = kRT to be used
for the Mach number definition, as follows:
Px 2 Py 2
Ux 2 = Uy 2 (6.10)
Tx kRTx Ty kRTy
| {z } | {z }
cx 2 cy 2

Note that the speed of sound on the different sides of the shock is different. Utiliz-
ing the definition of Mach number results in
Px 2 Py 2
Mx 2 = My 2 (6.11)
Tx Ty
6.1. SOLUTION OF THE GOVERNING EQUATIONS 93

Rearranging equation (6.11) results in

µ ¶2 µ ¶2
Ty Py My
= (6.12)
Tx Px Mx

Energy equation (6.3) can be converted to a dimensionless form which can be

expressed as
µ ¶ µ ¶
k−1 2 k−1 2
Ty 1 + My = Tx 1 + Mx (6.13)
2 2

It can also be observed that equation (6.13) means that the stagnation temper-
ature is the same, T0 y = T0 x . Under the perfect gas model, ρU 2 is identical to
kP M 2 because
M2
z }| {
ρ
z}|{
2 P  2 
 U  kRT = kP M 2
ρU = (6.14)
RT  kRT
| {z }

c2

Using the identity (6.14) transforms the momentum equation (6.2) into

Px + kPx Mx 2 = Py + kPy My 2 (6.15)

Rearranging equation (6.15) yields

Py 1 + kMx 2
= (6.16)
Px 1 + kMy 2

The pressure ratio in equation (6.16) can be interpreted as the loss of the static
pressure. The loss of the total pressure ratio can be expressed by utilizing the
relationship between the pressure and total pressure (see equation (5.11)) as
¡ k−1 2
¢ k−1
k
P0 y Py 1 + 2 My
= ¡ ¢ k−1
k (6.17)
P0 x Px 1 + k−1 2
2 Mx

The relationship between Mx and My is needed to be solved from the above set
of equations. This relationship can be obtained from the combination of mass,
momentum, and energy equations. From equation (6.13) (energy) and equation
(6.12) (mass) the temperature ratio can be eliminated.
µ ¶2 k−1 2
Py M y 1+ 2 Mx
= k−1 2 (6.18)
Px M x 1+ 2 My
94 CHAPTER 6. NORMAL SHOCK

Combining the results of (6.18) with equation (6.16) results in

µ ¶2 µ ¶2 2
1 + kMx 2 Mx 1 + k−1
2 Mx
= (6.19)
1 + kMy 2 My 1 + k−1
2 My
2

Equation (6.19) is a symmetrical equation in the sense that if My is substituted

with Mx and Mx substituted with My the equation remains the same. Thus, one
solution is
My = Mx (6.20)
It can be observed that equation (6.19) is biquadratic. According to the Gauss
Biquadratic Reciprocity Theorem this kind of equation has a real solution in a cer-
tain range3 which will be discussed later. The solution can be obtained by rewriting
equation (6.19) as a polynomial¡(fourth order).
¢ It is also possible to cross–multiply
equation (6.19) and divide it by Mx 2 − My 2 results in
k−1¡ 2 ¢
1+ My + My 2 − kMy 2 My 2 = 0 (6.21)
2
Equation (6.21) becomes

2
Mx 2 + 2
k−1
My = 2k 2 (6.22)
k−1 Mx −1

The first solution (6.20) is the trivial solution in which the two sides are identical
and no shock wave occurs. Clearly, in this case, the pressure and the temperature
from both sides of the nonexistent shock are the same, i.e. Tx = Ty , Px = Py . The
second solution is where the shock wave occurs.
The pressure ratio between the two sides can now be as a function of only
a single Mach number, for example, Mx . Utilizing equation (6.16) and equation
(6.22) provides the pressure ratio as only a function of the upstream Mach number
as
Py 2k k−1
= Mx 2 − or
Px k+1 k+1

Py 2k ¡ 2 ¢
=1+ Mx − 1 (6.23)
Px k+1

The density and upstream Mach number relationship can be obtained in

the same fashion to became

ρy Ux (k + 1)Mx 2
= = (6.24)
ρx Uy 2 + (k − 1)Mx 2
3 Ireland, K. and Rosen, M. ”Cubic and Biquadratic Reciprocity.” Ch. 9 in A Classical Introduction to

Modern Number Theory, 2nd ed. New York: Springer-Verlag, pp. 108-137, 1990.
6.1. SOLUTION OF THE GOVERNING EQUATIONS 95

The fact that the pressure ratio is a function of the upstream Mach number, Mx ,
provides additional way of obtaining an additional useful relationship. And the tem-
perature ratio, as a function of pressure ratio, is transformed into
µ ¶ Ã k+1 Py !
Ty Py k−1 + Px
= k+1 Py
(6.25)
Tx Px 1 + k−1 Px

In the same way, the relationship between the density ratio and pressure ratio is
³ ´³ ´
k+1 Py
ρx 1 + k−1 Px
= ³ ´ ³ ´ (6.26)
ρy k+1
+ y
P
k−1 Px

which is associated with the shock wave.

The Maximum Conditions

Shock Wave relationship
The maximum speed of sound My and P0y/P0x as a function of Mx
is when the highest tempera- 1
ture is achieved. The maxi- 0.9
mum temperature that can be 0.8 My
achieved is the stagnation tem- P0y/P0x
0.7
perature
0.6
r
2k
My

0.5
Umax = RT0 (6.27)
k−1 0.4

0.3
The stagnation speed of sound
0.2
is
p 0.1
c0 = kRT0 (6.28) 0
1 2 3 4 5 6 7 8 9 10
Fri Jun 18 15:47:34 2004 Mx
Based on this definition a new
Mach number can be defined Fig. -6.3: The exit Mach number and the stagnation
pressure ratio as a function of upstream
U Mach number.
M0 = (6.29)
c0

The Star Conditions

The speed of sound at the critical condition can also be a good reference velocity.
The speed of sound at that velocity is
√
c∗ = kRT ∗ (6.30)
96 CHAPTER 6. NORMAL SHOCK

In the same manner, an additional Mach number can be defined as

U
M∗ = (6.31)
c∗

6.1.3 Prandtl’s Condition

It can be easily observed that the temperature from both sides of the shock wave
is discontinuous. Therefore, the speed of sound is different in these adjoining
mediums. It is therefore convenient to define the star Mach number that will be
independent of the specific Mach number (independent of the temperature).

U c U c
M∗ = ∗
= ∗ = ∗M (6.32)
c c c c

The jump condition across the shock must satisfy the constant energy.

c2 U2 c∗ 2 c∗ 2 k + 1 ∗2
+ = + = c (6.33)
k−1 2 k−1 2 2(k − 1)

Dividing the mass equation by the momentum equation and combining it with the
perfect gas model yields

c1 2 c2 2
+ U1 = + U2 (6.34)
kU1 kU2

Combining equation (6.33) and (6.34) results in

· ¸ · ¸
1 k + 1 ∗2 k − 1 1 k + 1 ∗2 k − 1
c − U1 + U1 = c − U2 + U2 (6.35)
kU1 2 2 kU2 2 2

After rearranging and diving equation (6.35) the following can be obtained:

U1 U2 = c∗ 2 (6.36)

or in a dimensionless form

M ∗ 1 M ∗ 2 = c∗ 2 (6.37)
6.2. OPERATING EQUATIONS AND ANALYSIS 97

6.2 Operating Equations and Analysis

In Figure (6.3), the Mach number after the shock, My , and the ratio of the total
pressure, P0y /P0x , are plotted as a function of the entrance Mach number. The
working equations were presented earlier. Note that the My has a minimum value
which depends on the specific heat ratio. It can be noticed that the density ratio
(velocity ratio) also has a finite value regardless of the upstream Mach number.
The typical situations in which these equations can be used also include the mov-
ing shocks. The equations should be used with the Mach number (upstream or
downstream) for a given pressure ratio or density ratio (velocity ratio). This kind of
equations requires examining Table (6.1) for k = 1.4 or utilizing Potto-GDC for for
value of the specific heat ratio. Finding the Mach number for a pressure ratio of
8.30879 and k = 1.32 and is only a few mouse clicks away from the following table.

Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
2.7245 0.47642 2.1110 3.9360 8.3088 0.38109

To illustrate the use of the Shock Wave relationship

Py/Py, ρy/ρx and Ty/Tx as a function of Mx
above equations, an exam-
120.0
ple is provided.
110.0
100.0
Example 6.1: 90.0
Py/Px
Ty/Tx
Air flows with a Mach num- 80.0 ρy/ρx
ber of Mx = 3, at a pres- 70.0
sure of 0.5 [bar] and a 60.0
temperature of 0◦ C goes 50.0
through a normal shock. 40.0
Calculate the temperature, 30.0
pressure, total pressure, 20.0
and velocity downstream of 10.0
the shock. Assume that 0.0
k = 1.4. 1 2 3 4 5
Mx
6 7 8 9 10
Fri Jun 18 15:48:25 2004

S OLUTION Fig. -6.4: The ratios of the static properties of the two sides
Analysis: of the shock.
First, the known informa-
tion are Mx = 3, Px = 1.5[bar] and Tx = 273K. Using these data, the total pres-
sure can be obtained (through an isentropic relationship in Table (5.2), i.e., P0x is
known). Also with the temperature, Tx , the velocity can readily be calculated. The
relationship that was calculated will be utilized to obtain the ratios for the down-
stream of the normal shock. PP0x
x
= 0.0272237 =⇒ P0x = 1.5/0.0272237 = 55.1[bar]
√ √
cx = kRTx = 1.4 × 287 × 273 = 331.2m/sec
98 CHAPTER 6. NORMAL SHOCK
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
3.0000 0.47519 2.6790 3.8571 10.3333 0.32834

Ux = Mx × cx = 3 × 331.2 = 993.6[m/sec]
Now the velocity downstream is determined by the inverse ratio of ρy /ρx = Ux /Uy =
3.85714.

Uy = 993.6/3.85714 = 257.6[m/sec]
µ ¶
P0y
P0y = × P0x = 0.32834 × 55.1[bar] = 18.09[bar]
P0x

6.2.1 The Limitations of the Shock Wave

When the upstream Mach number becomes very large, the downstream Mach
number (see equation (6.22)) is limited by
: ∼0
»
»»»
2
2
1 +»
(k−1)Mx
2 k−1
My = ∼0 = (6.38)
2k ½
>
1½
2k
−
k−1 ½
Mx 2

This result is shown in Figure (6.3). The limits of the pressure ratio can be obtained
by looking at equation (6.16) and by utilizing the limit that was obtained in equation
(6.38).

6.2.2 Small Perturbation Solution

The small perturbation solution refers to an analytical solution where only a small
change (or several small changes) occurs. In this case, it refers to a case where
only a “small shock” occurs, which is up to Mx = 1.3. This approach had a major
significance and usefulness at a time when personal computers were not available.
Now, during the writing of this version of the book, this technique is used mostly in
obtaining analytical expressions for simplified models. This technique also has an
academic value and therefore will be described in the next version (0.5.x series).
The strength of the shock wave is defined as
Py − Px Py
P̂ = = −1 (6.39)
Px Px
By using equation (6.23) transforms equation (6.39) into
2k ¡ 2 ¢
P̂ = Mx − 1 (6.40)
k+1
6.2. OPERATING EQUATIONS AND ANALYSIS 99

or by utilizing equation (6.24) the following is obtained:

³ ´
2k ρy
k−1 ρx − 1
P̂ = ³ ´ (6.41)
2 ρy
k−1 − ρx − 1

6.2.3 Shock Thickness

The issue of shock thickness (which will be presented in a later version) is pre-
sented here for completeness. This issue has a very limited practical application
for most students; however, to convince the students that indeed the assumption
of very thin shock is validated by analytical and experimental studies, the issue
should be presented.
The shock thickness can be defined in several ways. The most common
definition is by passing a tangent to the velocity at the center and finding out where
the theoretical upstream and downstream conditions are meet.

6.2.4 Shock or Wave Drag

It is communally believed that regardless to the cause of the shock, the shock
creates a drag (due to increase of entropy). In this section, the first touch of this
phenomenon will be presented. The fact that it is assumed that the flow is friction-
less does not change whether or not shock drag occur. This explanation is broken
into two sections: one for stationary shock wave, two for moving shock shock wave.
A better explanation should appear in the oblique shock chapter.
Consider a normal shock as shown in figure (6.5). Gas flows in a supersonic

stream lines

U1 U2
ρ2
ρ1
A1 A2
P1
P2

Fig. -6.5: The diagram that reexplains the shock drag effect.

velocity around a two–dimensional body and creates a shock. This shock is an

oblique shock, however in this discussion, if the control volume is chosen close
enough to the body is can be considered as almost a normal shock (in the oblique
shock chapter a section on this issue will be presented that explains the fact that
shock is oblique, to be irrelevant).
The control volume that is used here is along two stream lines. The other two
boundaries are arbitrary but close enough to the body. Along the stream lines there
100 CHAPTER 6. NORMAL SHOCK

is no mass exchange and therefore there is no momentum exchange. Moreover,

it is assumed that the gas is frictionless, therefore no friction occurs along any
stream line. The only change is two arbitrary surfaces since the pressure, velocity,
and density are changing. The velocity is reduced Ux > Uy . However, the density is
increasing, and in addition, the pressure is increasing. So what is the momentum
net change in this situation? To answer this question, the momentum equation
F
must be written and it will be similar to equation (5.104). However, since F y∗ = FFx∗
there is no net force acting on the body. For example, consider upstream of Mx =
3. and for which

Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
3.0000 0.47519 2.6790 3.8571 10.3333 0.32834

and the corespondent Isentropic information for the Mach numbers is

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
3.0000 0.35714 0.07623 4.2346 0.02722 0.11528 0.65326
0.47519 0.95679 0.89545 1.3904 0.85676 1.1912 0.65326

Now, after it was established, it is not a surprising result. After all, the shock
analysis started with the assumption that no momentum is change. As conclusion
there is no shock drag at stationary shock. This is not true for moving shock as it
will be discussed in section (6.3.1).

6.3 The Moving Shocks

In some situations, the shock wave is not stationary. This kind of situation arises in
many industrial applications. For example, when a valve is suddenly 4 closed and
a shock propagates upstream. On the other extreme, when a valve is suddenly
opened or a membrane is ruptured, a shock occurs and propagates downstream
(the opposite direction of the previous case). In some industrial applications, a
liquid (metal) is pushed in two rapid stages to a cavity through a pipe system. This
liquid (metal) is pushing gas (mostly) air, which creates two shock stages. As a
general rule, the shock can move downstream or upstream. The last situation is the
most general case, which this section will be dealing with. There are more genera
cases where the moving shock is created which include a change in the physical
properties, but this book will not deal with them at this stage. The reluctance to deal
with the most general case is due to fact it is highly specialized and complicated
even beyond early graduate students level. In these changes (of opening a valve
and closing a valve on the other side) create situations in which different shocks are
moving in the tube. The general case is where two shocks collide into one shock
4 It will be explained using dimensional analysis what is suddenly open
6.3. THE MOVING SHOCKS 101

and moves upstream or downstream is the general case. A specific example is

common in die–casting: after the first shock moves a second shock is created in
which its velocity is dictated by the upstream and downstream velocities.
In cases where the shock ve-
locity can be approximated as a con-
stant (in the majority of cases) or as flow M
direction 
y M x
x
near constant, the previous analysis,
Px Ux
y
Py U s
0

equations, and the tools developed in Uy

0

this chapter can be employed. The Ty Tx

problem can be reduced to the pre-
viously studied shock, i.e., to the sta- c.v.
tionary case when the coordinates are Stationary Coordinates
attached to the shock front. In such a
case, the steady state is obtained in
the moving control value.
For this analysis, the coordi- !
U U !
0

Us Ux
0
s y
nates move with the shock. Here, the
prime ’ denote the values of the static U = 0  x Px
coordinates. Note that this notation P <P
x y Tx
is contrary to the conventional nota-
tion found in the literature. The rea-
son for the deviation is that this choice c.v.
reduces the programing work (espe- Moving Coordinates
cially for object–oriented programing
like C++). An observer moving with Fig. -6.6: Comparison between stationary
shock and moving shock in ducts
the shock will notice that the pressure
in the shock is
0 0
Px = Px Py = Py (6.42)

The temperature measured by the observer is

0 0
Tx = Tx Ty = Ty (6.43)

Assuming that the shock is moving to the right, (refer to Figure (6.6)) the velocity
measured by the observer is
0
Ux = Us − Ux (6.44)

Where Us is the shock velocity which is moving to the right. The “downstream”
velocity is
0
Uy = Us − Uy (6.45)

The speed of sound on both sides of the shock depends only on the tempera-
ture and it is assumed to be constant. The upstream prime Mach number can be
102 CHAPTER 6. NORMAL SHOCK

defined as
0 Us − Ux Us
Mx = = − Mx = Msx − Mx (6.46)
cx cx
It can be noted that the additional definition was introduced for the shock upstream
Mach number, Msx = Ucxs . The downstream prime Mach number can be expressed
as
0 Us − Uy Us
My = = − My = Msy − My (6.47)
cy cy
Similar to the previous case, an additional definition was introduced for the shock
downstream Mach number, Msy . The relationship between the two new shock
Mach numbers is
Us cy Us
=
cx cx cy
r
Ty
Msx = Msy (6.48)
Tx
The “upstream” stagnation temperature of the fluid is
µ ¶
k−1
T0x = Tx 1 + Mx 2
2
(6.49)
and the “upstream” prime stagnation pressure is
µ ¶ k−1
k
k−1 2
P0x = Px 1 + Mx (6.50)
2
The same can be said for the “downstream” side of the shock. The difference
between the stagnation temperature is in the moving coordinates
T0y − T0x = 0 (6.51)
It should be noted that the stagnation temperature (in the stationary coordinates)
rises as opposed to the stationary normal shock. The rise in the total temperature
is due to the fact that a new material has entered the c.v. at a very high velocity,
and is “converted” or added into the total temperature,
µ ´ ¶ µ ´ ¶
k−1³ 0 2 k−1³ 0 2
T0y − T0x =Ty 1 + Msy − My − Tx 1 + Msx − Mx
2 2
0
T0y
z µ }| ¶{
k−1 02 k−1
0 = Ty 1 + My +Ty Msy (Msy − 2My )
2 2
0
T0x
z µ }| ¶{
k−1 02 k−1
− Tx 1 + Mx −Tx Msx (Msx − 2Mx ) (6.52)
2 2
6.3. THE MOVING SHOCKS 103

and according to equation (6.51) leads to

µ ¶
0 0 Tx k − 1 Ty k − 1
T0y − T0x = Us (Msx − 2Mx ) − (Msy − 2My ) (6.53)
cx 2 cy 2

Again, this difference in the moving shock is expected because moving material
velocity (kinetic energy) is converted into internal energy. This difference can also
be viewed as a result of the unsteady state of the shock.

6.3.1 Shock or Wave Drag Result from a Moving Shock

stationary lines at the

speed of the object

moving U2 6= 0
U1 = 0 object ρ2
ρ1
A1 A2
P1
P2

stream lines

Fig. -6.7: The diagram that reexplains the shock drag effect of a moving shock.

In section (6.2.4) it was shown that there is no shock drag in stationary shock.
However, the shock or wave drag is very significant so much so that at one point
it was considered the sound barrier . Consider the figure (6.7) where the stream
lines are moving with the object speed. The other boundaries are stationary but the
velocity at right boundary is not zero. The same arguments, as discussed before in
the stationary case, are applied. What is different in the present case (as oppose to
the stationary shock), one side has increase the momentum of the control volume.
This increase momentum in the control volume causes the shock drag. In way, it
can be view as continuous acceleration of the gas around the body from zero. Note
this drag is only applicable to a moving shock (unsteady shock).
The moving shock is either results from a body that moves in gas or from a
sudden imposed boundary like close or open valve5 In the first case, the forces/energy
flows from body to gas and there for there is a need for large force to accelerate
the gas over extremely short distance (shock thickness). In the second case, the
gas contains the energy (as high pressure, for example in the open valve case)
and the energy potential is lost in the shock process (like shock drag).
5 According to my son, the difference between these two cases is the direction of the information.

Both case there essentially bodies, however, in one the information flows from inside the field to the
boundary while the other case it is the opposite.
104 CHAPTER 6. NORMAL SHOCK

For some strange reasons, this topic has several misconceptions that even
appear in many popular and good textbooks6 . Consider the following example
taken from such a book.

Fig. -6.8: The diagram for the common explanation for shock or wave drag effect a shock.
Please notice the strange notations (e.g. V and not U) and they result from a
verbatim copy.

Example 6.2:
A book explains the shock drag is based on the following rational: The body is
moving in a stationary frictionless fluid under one–dimensional flow. The left plane
is moving with body at the same speed. The second plane is located “downstream
from the body where the gas has expanded isotropically (after the shock wave) to
the upstream static pressure”. the bottom and upper stream line close the control
volume. Since the pressure is the same on the both planes there is no unbalanced
pressure forces. However, there is a change in the momentum in the flow direction
because U1 > U2 . The force is acting on the body. There several mistakes in this
explanation including the drawing. Explain what is wrong in this description (do not
describe the error results from oblique shock).

S OLUTION
Neglecting the mistake around the contact of the stream lines with the oblique
shock(see for retouch in the oblique chapter), the control volume suggested is
stretched with time. However, the common explanation fall to notice that when
the isentropic explanation occurs the width of the area change. Thus, the simple
explanation in a change only in momentum (velocity) is not appropriate. Moreover,
in an expanding control volume this simple explanation is not appropriate. Notice
that the relative velocity at the front of the control volume U1 is actually zero. Hence,
the claim of U1 > U2 is actually the opposite, U1 < U2 .

6 Similar situation exist in the surface tension area.

6.3. THE MOVING SHOCKS 105

6.3.2 Shock Result from a Sudden and Complete Stop

The general discussion can be simplified in the extreme case when the shock is
moving from a still medium. This situation arises in many cases in the industry, for
example, in a sudden and complete closing of a valve. The sudden closing of the
valve must result in a zero velocity of the gas. This shock is viewed by some as
a reflective shock. The information propagates upstream in which the gas velocity
is converted into temperature. In many such cases the steady state is established
quite rapidly. In such a case, the shock velocity “downstream” is Us . Equations
(6.42) to (6.53) can be transformed into simpler equations when Mx is zero and Us
is a positive value.

Mx My < Mx Us + Ux
!
x Us
0

y Ty
Px U s Py x Px
Ux Px < P y
0

Tx Tx
Uy
0

=0

c.v. c.v.
Stationary Coordinates Moving Coordinates

Fig. -6.9: Comparison between a stationary shock and a moving shock in a stationary
medium in ducts.

The “upstream” Mach number reads

Us + Ux
Mx = = Msx + Mx (6.54)
cx
The “downstream” Mach number reads
|Us |
My = = Msy (6.55)
cy

Again, the shock is moving to the left. In the moving coordinates, the observer
(with the shock) sees the flow moving from the left to the right. The flow is moving
to the right. The upstream is on the left of the shock. The stagnation temperature
increases by
µ ¶
Tx k − 1 Ty k − 1
T0y − T0x = Us (Msx + 2Mx ) − (Msy ) (6.56)
cx 2 cy 2

The prominent question in this situation is what will be the shock wave velocity for
0
a given fluid velocity, Ux , and for a given specific heat ratio. The “upstream” or
the “downstream” Mach number is not known even if the pressure and the tem-
perature downstream are given. The difficulty lies in the jump from the stationary
106 CHAPTER 6. NORMAL SHOCK

Mx My < Mx Us + Ux
!
x Us
0

y Ty
Px U s Py x Px
Ux Px < P y
0

Tx Tx
Uy
0

=0

c.v. c.v.
Stationary Coordinates Moving Coordinates

Fig. -6.10: Comparison between a stationary shock and a moving shock in a stationary
medium in ducts.

coordinates to the moving coordinates. It turns out that it is very useful to use the
dimensionless parameter Msx , or Msy instead of the velocity because it combines
the temperature and the velocity into one parameter.

The relationship between the Mach number on the two sides of the shock are tied
through equations (6.54) and (6.55) by

0
´2³
2
Mx + Msx + k−1
2
(My ) = ¡ 0 ¢2 (6.57)
2k
k−1 Mx + Msx −1

And substituting equation (6.57) into (6.48) results in

f (Msx )
zs}| { v ¡
u 0 ¢2 2
Tx u Mx + Msx + k−1
Mx = t (6.58)
Ty 2k
¡ 0 ¢2
k−1 Mx + Msx −1
6.3. THE MOVING SHOCKS 107

The temperature ratio in equation Shock in A Suddenly Close Valve

(6.58) and the rest of the right–hand k=14
side show clearly that Msx has four 3
Msx
possible solutions (fourth–order poly- Msy
nomial Msx has four solutions). Only
one real solution is possible. The so- 2

lution to equation (6.58) can be ob-

tained by several numerical methods.
Note, an analytical solution can be ob- 1

tained for equation (6.58) but it seems

utilizing numerical methods is much
more simple. The typical method is 0
0.1 1
the “smart” guessing of Msx . For very Mx

small values of the upstream Mach

0 Thu Aug 3 18:54:21 2006
number, Mx ∼ ² equation (6.58) pro-
vides that Msx ∼ 1 + 12 ² and Msy = Fig. -6.11: The moving shock Mach numbers as
1 − 21 ² (the coefficient is only approx- a result of a sudden and complete
imated as 0.5) as shown in Figure stop.
(6.11). From the same figure it can also be observed that a high velocity can result
in a much larger velocity for the reflective shock. For example, a Mach number
close to one (1), which can easily be obtained in a Fanno flow, the result is about
double the sonic velocity of the reflective shock. Sometimes this phenomenon can
have a tremendous significance in industrial applications.
Note that to achieve supersonic velocity (in stationary coordinates) a diverging–
converging nozzle is required. Here no such device is needed! Luckily and hope-
fully, engineers who are dealing with a supersonic flow when installing the nozzle
and pipe systems for gaseous mediums understand the importance of the reflec-
tive shock wave.
Two numerical methods and the algorithm employed to solve this problem for given,
0
Mx , is provided herein:

(a) Guess Mx > 1,

(b) Using shock table or use Potto–GDC to calculate temperature ratio and My ,
0
q
Tx
(c) Calculate the Mx = Mx − Ty My

0 0
(d) Compare to the calculated Mx to the given Mx . and adjust the new guess
Mx > 1 accordingly.

¡p¿ The second method is “successive substitutions,” which has better conver-
gence to the solution initially in most ranges but less effective for higher accuracies.
0
(a) Guess Mx = 1 + Mx ,
108 CHAPTER 6. NORMAL SHOCK

(b) using the shock table or use Potto–GDC to calculate the temperature ratio
and My ,
0
q
(c) calculate the Mx = Mx − TTxy My

(d) Compare the new Mx approach the old Mx , if not satisfactory use the new
0 0
Mx to calculate Mx = 1 + Mx then return to part (b).

6.3.3 Moving Shock into Stationary Medium (Suddenly Open

Valve)
General Velocities Issues

When a valve or membrane is suddenly opened, a shock is created and propagates

downstream. With the exception of close proximity to the valve, the shock moves
in a constant velocity (6.12(a)). Using a coordinates system which moves with the
shock results in a stationary shock and the flow is moving to the left see Figure
(6.12(b)). The “upstream” will be on the right (see Figure (6.12(b))).

Uy = Us U y
0
Ux = Us
Uy Ux
0

=0
0

Upstream
x Px  x Px
Us Tx Tx

c.v. c.v.

(a) Stationary coordinates (b) Moving coordinates

Fig. -6.12: A shock moves into a still medium as a result of a sudden and complete opening
of a valve

Similar definitions of the right side and the left side of the shock Mach numbers
can be utilized. It has to be noted that the “upstream” and “downstream” are the
reverse from the previous case. The “upstream” Mach number is

Us
Mx = = Msx (6.59)
cx
The “downstream” Mach number is
0
Us − Uy 0
My = = Msy − My (6.60)
cy

Note that in this case the stagnation temperature in stationary coordinates changes
(as in the previous case) whereas the thermal energy (due to pressure difference)
6.3. THE MOVING SHOCKS 109

is converted into velocity. The stagnation temperature (of moving coordinates) is

µ ¶ µ ¶
k−1 2 k−1 2
T0 y − T0 x = Ty 1 + (Msy − My ) − Tx 1 + (Mx ) = 0 (6.61)
2 2
A similar rearrangement to the previous case results in
µ ¶
0 0 k−1¡ ¢2
T0 y − T0 x = Ty 1 + −2Msy My + My 2 (6.62)
2

Shock in A Suddenly Open Valve Shock in A Suddenly Open Valve

k = 1 4, My’ = 0.3 k = 1 4, My’ = 1.3
1.75 4
Mx Mx
My 3.5 My
1.5 Ty/Tx Ty/Tx
3

2.5
1.25
2

1.5
1
1

0.5
0.75
0 10 0 5 10 15 20
Number of Iteration Number of Iteration

Wed Aug 23 17:20:59 2006 Wed Aug 23 17:46:15 2006

0 0
(a) My = 0.3 (b) My = 1.3

Fig. -6.13: The number of iterations to achieve convergence.

The same question that was prominent in the previous case appears now, what will
be the shock velocity for a given upstream Mach number? Again, the relationship
between the two sides is
v
u
u (Msx )2 + k−1
2
Msy = My + t 2k
0

2 (6.63)
k−1 (Msx ) − 1

Since Msx can be represented by Msy theoretically equation (6.63) can be solved.
It is common practice to solve this equation by numerical methods. One such
methods is “successive substitutions.” This method is applied by the following al-
gorithm:
(a) Assume that Mx = 1.0.
(b) Calculate the Mach number My by utilizing the tables or Potto–GDC.
(c) Utilizing r
Ty ³ 0
´
Mx = My + My
Tx
calculate the new “improved” Mx .
110 CHAPTER 6. NORMAL SHOCK

(d) Check the new and improved Mx against the old one. If it is satisfactory, stop
or return to stage (b).
0
To illustrate the convergence of the procedure, consider the case of My = 0.3 and
0
My = 0.3. The results show that the convergence occurs very rapidly (see Figure
0
(6.13)). The larger the value of My , the larger number of the iterations required to
achieve the same accuracy. Yet, for most practical purposes, sufficient results can
be achieved after 3-4 iterations.

Piston Velocity When a piston is moving, it creates a shock that moves at a

speed greater than that of the piston itself. The unknown data are the piston veloc-
ity, the temperature, and, other conditions ahead of the shock. Therefore, no Mach
number is given but pieces of information on both sides of the shock. In this case,
the calculations for Us can be obtained from equation (6.24) that relate the shock
velocities and Shock Mach number as
Ux Msx (k + 1)Msx 2
= 0 = (6.64)
Uy U
Msx − cyx 2 + (k − 1)Msx 2

Equation (6.64) is a quadratic equation for Msx . There are three solutions of
which the first one is Msx = 0 and this is immediately disregarded. The other two
solutions are
q£ ¤2
0 0
(k + 1)Uy ± Uy (1 + k) + 16cx 2
Msx = (6.65)
4 cx
The negative sign provides a negative value which is disregarded, and the only
solution left is
q£ ¤2
0 0
(k + 1)Uy + Uy (1 + k) + 16cx 2
Msx = (6.66)
4 cx
or in a dimensionless form
q£ ¤2
0 0
(k + 1)Myx + Myx (1 + k) + 16
Msx =
4 (6.67)
0 0
Where the “strange” Mach number is Msx = Uy /cx . The limit of the equation
when cx → ∞ leads to
0
(k + 1)Myx
Msx = (6.68)
4
As one additional “strange” it can be seen that the shock is close to the piston
when the gas ahead of the piston is very hot. This phenomenon occurs in many
6.3. THE MOVING SHOCKS 111

industrial applications, such as the internal combustion engines and die casting.
Some use equation (6.68) to explain the next Shock-Choke phenomenon.
In one of the best book in fluid mechanics provides a problem that is the
similar to the piston pushing but with a twist. In this section analysis will carried for
the error in neglecting the moving shock. This problem is discussed here because
at first glance looks a simple problem, however, the physics of the problem is a bit
complicated and deserve a discussion7 .

Fig. -6.14: Schematic of showing the piston pushing air.

A piston with a known dimensions (shown in Figure 6.3.3 is pushed by a

constant force. The gas (air) with an initial temperature is pushed through a con-
verging nozzle (shown in the original schematic). The point where the moving
shock reaches to the exit there are two situations:choked and unchoked flow. If the
flow is choked, then the Mach number at the exit is one. If the flow is unchoked,
then the exit Mach number is unknown but the pressure ratio is know. Assuming
the flow is choked
√ (see √ later for the calculation) the exit Mach number is 1 and
therefor, Ue = kRT = 1.4 × 287 × 0.833 × 293.15 ∼ 313[m/sec] The velocity at
the cylinder is assumed to be isentropic and hence area ratio is A/A∗ = 1600 the
condition at the cylinder can be obtained from Potto-GDC as

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
3.614E−4 1.0 1.0 1.6E+3 1.0 1.6E+3 6.7E+2

√
The piston velocity is then Upiston = 0.000361425 × 1.4 × 287 × 297.15 ∼
0.124[m/sec].
Before the semi state state is achieved, the piston is accelerated to the con-
stant velocity (or at least most constant velocity). A this stage, a shock wave is
moving away from piston toward the nozzle. If this shock reaches to exit before the
semi state is achieved, the only way to solve this problem is by a numerical method
(either characteristic methods or other numerical method) and it is out of the scope
of this chapter. The transition of the moving shock through the converging nozzle
is neglected in this discussion. However, if a quasi steady state is obtained, this
7 A student from France forward this problem to this author after argument with his instructor. The

instructor used the book’s manual solution and refused to accept the student improved solution which
he learned from this book/author. Therefore, this problem will be referred as the French problem.
112 CHAPTER 6. NORMAL SHOCK

discussion deals with that case. Before the shock is reaching to exit no flow occur
at the exit (as opposite to the solution which neglects the moving shock).
The first case (choked, which is the more common, for example, syringe
when pushing air has similar situations), is determined from the fact that pressure
at the cylinder can be calculated. If the pressure ratio is equal or higher than the
critical ratio then the flow is choked. For the unchoked case, the exit Mach number
is unknown. However, the pressure ratio between the cylinder and the outside
world is known. The temperature in the cylinder has to be calculated using moving
shock relationship.
In the present case, the critical force should be calculated first. The specific
heat ratio is k = 1.4 and therefore critical pressure ratio is 0.528282. The critical
force is
Pcritical
Fcritical = Pcritical Apiston = Pa Apiston (6.69)
Pa
In this case
π × 0.122
Fcritical = 101325(1/0.528282 − 1) × ∼ 1022.74[N ]
4
Since the force is 1100 [N], it is above the critical force the flow is chocked.
The pressure ratio between the cylinder and the choking point is the critical pres-
sure ratio. It should be noted that further increase of the force will not change the
pressure ratio but the pressure at the choking point (see the Figure below).
1100
101325 +
Pcylinder π×0.122
= 4
= 1.96
Pa 101325
The moving shock conditions are determined from the velocity of the piston.
As first approximation the piston Mach number is obtained from the area ratio in
0
isentropic flow (3.614E −4 ). Using this Mach number is My Potto-GDC provides
0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
1.0002 0.99978 0.0 0.000361 1.0 1.001 1.0

The improved the piston pressure ratio (“piston” pressure to the nozzle pres-
sure) is changed by only 0.1%. Improved accuracy can be obtained in the second
iteration by taking this shock pressure ratio into consideration. However, here, for
most engineering propose this improvement is insignificant. This information pro-
vides the ability to calculate the moving shock velocity.
√
Vshock = c Ms = c Mx = 1.0002 1.4 × 287 × 293.15 ∼ 343.3[m/sec]
The time for the moving shock to reach depends on the length of the cylinder as
Lcylinder
t= (6.70)
Vshock
6.3. THE MOVING SHOCKS 113

For example, in case the length is three times the diameter will result then the time
is
3 × 0.12
t= ∼ 0.001[sec]
343.3

In most case this time is insignificant, how- pressure

after
the steady state

ever, there are process and conditions that this Nozzle

shock reaches
Pressure
the nozzle

shock affects the calculations. In Figure 6.17 "initial"

pressure

shows the pressure at the nozzle and the piston unsteady

state
velocity. It can be observed that piston velocity
increase to constant velocity very fast. Initially Piston
Velocity
gradual

the transition continue until a quasi steady state pressure

increase

{
is obtained. This quasi steady state continues t 0 Time[Msec]
until the shock reaches to the nuzzle and the
pressure at the nozzle jump in a small amount
(see Figure shock:fig:cylinderNozzleResults. Fig. -6.15: Time the pressure at
the nozzle for the French
problem.
Shock–Choke Phenomenon

Assuming that the gas velocity is supersonic (in stationary coordinates) before the
shock moves, what is the maximum velocity that can be reached before this model
fails? In other words, is there a point where the moving shock is fast enough to
reduce the “upstream” relative Mach number below the speed of sound? This is
the point where regardless of the pressure difference is, the shock Mach number
cannot be increased.
This shock–choking phe- Shock in A Suddenly Open Valve
nomenon is somewhat simi- 2.5
Maximum M ’ possible
y

lar to the choking phenomenon 2.25 y(max)

M

that was discussed earlier in a 2

nozzle flow and in other pipe
Maximum My’

1.75
flow models (later chapters). 1.5

The difference is that the ac- 1.25

tual velocity has no limit. It 1

must be noted that in the previ- 0.75

ous case of suddenly and com- 0.5

The spesific heat ratio, k
pletely closing of valve results
in no limit (at least from the Thu Aug 24 17:46:07 2006

model point of view). To explain

this phenomenon, look at the Fig. -6.16: The maximum of “downstream” Mach num-
normal shock. Consider when ber as a function of the specific heat, k.
the “upstream” Mach approaches infinity, Mx = Msx → ∞, and the downstream
Mach number, according to equation (6.38), is approaching to (k − 1)/2k. One can
view this as the source of the shock–choking phenomenon. These limits determine
114 CHAPTER 6. NORMAL SHOCK

the maximum velocity after the shock since Umax = cy My . From the upstream side,
the Mach number is
∞
r µ ¶
µ
¡
Ty k − 1
Mx = Msx = ¡ (6.71)
¡ Tx 2k

Thus, the Mach number is approaching infinity because of the temperature ratio
but the velocity is finite.
To understand this limit, consider that the maximum Mach number is obtained
P
when the pressure ratio is approaching infinity Pxy → ∞. By applying equation
(6.23) to this situation the following is obtained:
s µ ¶
k+1 Px
Msx = −1 +1 (6.72)
2k Py

and the mass conservation leads to

Uy ρy = Us ρx
³ 0
´
Us − Uy ρy = Us ρx
r µ ¶
0 Ty ρx
My = 1− Msx (6.73)
Tx ρy

Substituting equations (6.26) and (6.25) into equation (6.73) results in

v v ³ ´³ ´
µ ¶u u Py
2k u 1 + k+1
0 1 Py ut k+1 u k−1 Px
My = 1− Py
×t ³ ´ ³ ´ (6.74)
k Px k−1 P
Px + k+1
k+1
+ y
k−1 Px

When the pressure ratio is approaching infinity (extremely strong pressure ratio),
the results is
s
0 2
My = (6.75)
k(k − 1)

What happens when a gas with a Mach number larger than the maximum
Mach number possible is flowing in the tube? Obviously, the semi steady state
described by the moving shock cannot be sustained. A similar phenomenon to
the choking in the nozzle and later in an internal pipe flow is obtained. The Mach
number is reduced to the maximum value very rapidly. The reduction occurs by an
increase of temperature after the shock or a stationary shock occurs as it will be
shown in chapters on internal flow.
6.3. THE MOVING SHOCKS 115
0 Ty
k Mx My My Tx
1.30 1073.25 0.33968 2.2645 169842.29
1.40 985.85 0.37797 1.8898 188982.96
1.50 922.23 0.40825 1.6330 204124.86
1.60 873.09 0.43301 1.4434 216507.05
1.70 833.61 0.45374 1.2964 226871.99
1.80 801.02 0.47141 1.1785 235702.93
1.90 773.54 0.48667 1.0815 243332.79
2.00 750.00 0.50000 1.00000 250000.64
2.10 729.56 0.51177 0.93048 255883.78
2.20 711.62 0.52223 0.87039 261117.09
2.30 695.74 0.53161 0.81786 265805.36
2.40 681.56 0.54006 0.77151 270031.44
2.50 668.81 0.54772 0.73029 273861.85

Table of maximum values of the shock-choking phenomenon.

The mass flow rate when the pressure ratio is approaching infinity, ∞, is
ρy
cy
z }| { z }| {
0 p
ṁ 0 0 Py
= Uy ρy = My cy ρy = My kRTy
A RTy
0√
My kPy
= p (6.76)
RTy

Equation (6.76) and equation (6.25) can be transferred for large pressure ratios
into
ṁ p Px k − 1
∼ Ty (6.77)
A Tx k + 1

p Since the right hand side of equation (6.77) is constant, with the exception of
Ty the mass flow rate is approaching infinity when the pressure ratio is approach-
ing infinity. Thus, the shock–choke phenomenon means that the Mach number is
only limited in stationary coordinates but the actual flow rate isn’t.

Moving Shock in Two and Three Dimensions

A moving shock into a still gas can occur in a cylindrical or a spherical coordi-
nates8 . For example, explosion can be estimated as a shock moving in a three
8 Dr. Attiyerah asked me to provide example for this issue. Explosion is not my area of research but it

turned to be similar to the author’s work on evacuation and filling of semi rigid chambers. It also similar
to shock tube and will be expanded later.
116 CHAPTER 6. NORMAL SHOCK

dimensional direction in uniform way. A long line of explosive can create a cylin-
drical moving shock. These shocks are similar to one dimensional shock in which
a moving gas is entering a still gas. In one dimensional shock the velocity of the
shock is constant. In two and three dimensions the pressure and shock velocity
and velocity behind the shock are function of time. These difference decrease the
accuracy of the calculation because the unsteady part is not accounted for. How-
ever, the gain is the simplicity of the calculations. The relationships that have been
developed so far for the normal shock are can be used for this case because the
shock is perpendicular to the flow. However, it has to be remembered that for very
large pressure difference the unsteadiness has to be accounted. The deviation
increases as the pressure difference decrease and the geometry became larger.
Thus, these result provides the limit for the unsteady state. This principle can be
demonstrated by looking in the following simple example.

Example 6.3:
After sometime after an explosion a spherical “bubble” is created with pressure
of 20[Bar]. Assume that the atmospheric pressure is 1[Bar] and temperature of
27◦ C Estimate the higher limit of the velocity of the shock, the velocity of the gas
inside the “bubble” and the temperature inside the bubble. Assume that k = 1.4
and R = 287[j/kg/K and no chemical reactions occur.

S OLUTION
The Mach number can be estimated from the pressure ratio
Pinside
= 20
Poutside
. One can obtain using Potto–gdc the following
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
4.1576 0.43095 4.2975 4.6538 20.0000 0.12155

or by using the shock dynamics section the following

0 0 Ty Py P0y
Mx My Mx My Tx Px P0 x
4.1576 0.43095 0 1.575 4.298 20 0.12155

The shock velocity estimate is then

Mx
z}|{ √
Us = Ms cy = 4.1576 × 1.4 × 287 × 300 ∼ 1443.47[m/sec]

The temperature inside the “bubble” is then

Ty
Ty = Tx = 4.298 × 300 ∼ 1289.4K
Tx
6.3. THE MOVING SHOCKS 117

The velocity of the gas inside the “bubble” is then

0 0 √
Uy = My cy = 1.575 × 1.4 × 287 × 1289.4 ∼ 1133.65[m/sec]

These velocities estimates are only the upper limits. The actual velocity will be
lower due to the unsteadiness of the situation.
End Solution
This problem is unsteady state but can be con- Toutside
sidered as a semi steady state. This kind of Poutside
analysis creates a larger error but gives the
trends and limits. The common problem is that r(t) Uy = Ux
for a given pressure ratio and initial radius (vol-
T(t)
ume) the shock velocity and inside gas velocity P(t)
inside are needed. As first approximation it can
be assumed material inside the “bubble’ is uni-
form and undergoes isentropic process. This is
similar to shock tube.
Fig. -6.17: Time the pressure at
the nozzle for the French
6.3.4 Partially Open Valve problem.
The previous case is a special case of the mov-
ing shock. The general case is when one gas
flows into another gas with a given velocity. The
only limitation is that the “downstream’ gas velocity is higher than the “upstream”
gas velocity as shown in Figure (6.20).

Uy = Us U y
′
0
Ux = Us − Ux
Uy
0
′
Ux Upstream
x Px ′ ′  x Px
Us Tx Uy > Ux
Tx

c.v. c.v.

(a) Stationary coordinates (b) Moving coordinates

Fig. -6.18: A shock moves into a moving medium as a result of a sudden and complete open
valve.

The relationship between the different Mach numbers on the “upstream” side is
0
Mx = Msx − Mx (6.78)

The relationship between the different Mach on the “downstream” side is

0
My = Msy − My (6.79)
118 CHAPTER 6. NORMAL SHOCK

An additional parameter has be supplied to solve the problem. A common problem

is to find the moving shock velocity when the velocity “downstream” or the pressure
is suddenly increased. It has to be mentioned that the temperature “downstream”
is unknown (the flow of the gas with the higher velocity). The procedure for the
calculations can be done by the following algorithm:
0
(a) Assume that Mx = Mx + 1.

(b) Calculate the Mach number My by utilizing the tables or Potto–GDC.

0
(c) Calculate the “downstream” shock Mach number Msy = My + My

(d) Utilizing r
Ty 0
Mx = (Msy ) − Mx
Tx
calculate the new “improved” Mx

(e) Check the new and improved Mx against the old one. If it is satisfactory, stop
or return to stage (b).

Earlier, it was shown that the shock chok- Shock in A Suddenly Open Valve
ing phenomenon occurs when the flow is 1
k=14

running into a still medium. This phe- 0.9

nomenon also occurs in the case where 0.8 x

M ’ = 0.9

a faster flow is running into a slower fluid. x

M ’ = 0.2
M ’ = 0.0
0.7 x

The mathematics is cumbersome but re-

My

sults show that the shock choking phe-

0.6

nomenon is still there (the Mach number is 0.5

limited, not the actual flow). Figure (6.19) 0.4

exhibits some “downstream” Mach num- 0

0.3
0.4 0.8 1.2 1.6 2 2.4 2.8

bers for various static Mach numbers, My , y

M’

and for various static “upstream” Mach Thu Oct 19 10:34:19 2006
0
numbers, Mx . The figure demonstrates Fig. -6.19: The results of the partial open-
that the maximum can also occurs in the ing of the valve.
vicinity of the previous value (see following question/example).

6.3.5 Partially Closed Valve

The totally closed valve is a special case of a partially closed valve in which there
is a sudden change and the resistance increases in the pipe. The information
propagates upstream in the same way as before. Similar equations can be written:

0
Ux = Us + Ux (6.80)
6.3. THE MOVING SHOCKS 119

′
′
′ ′ Ux = Us + Ux Uy = Us + Uy
Ux Uy
Upstream
x ρy P y
P
Tx Us
x
Ty

c.v. c.v.

(a) Stationary coordinates (b) Moving coordinates

Fig. -6.20: A shock as a result of a sudden and partially a valve closing or a narrowing the
passage to the flow

0
Uy = Us + Uy (6.81)

0
Mx = Ms + Mx (6.82)

0
My = Ms + My (6.83)

For given static Mach numbers the procedure for the calculation is as follows:
0
(a) Assume that Mx = Mx + 1.
(b) . Calculate the Mach number My by utilizing the tables or Potto–GDC
0
(c) Calculate the “downstream” shock Mach number Msy = My − My
(d) Utilizing r
Ty 0
Mx = (Msy ) + Mx
Tx
calculate the new “improved” Mx
(e) Check the new and improved Mx against the old one. If it is satisfactory, stop
or return to stage (b).

6.3.6 Worked–out Examples for Shock Dynamics

Example 6.4:
A shock is moving at a speed of 450 [m/sec] in a stagnated gas at pressure of
1[Bar] and temperature of 27◦ C. Compute the pressure and the temperature be-
hind the shock. Assume the specific heat ratio is k=1.3.
120 CHAPTER 6. NORMAL SHOCK

S OLUTION
It can be observed that the gas behind the shock is moving while the gas ahead of
the shock is still. Thus, it is the case of a shock moving into still medium (suddenly
opened valve case). First, the Mach velocity ahead of the shock has to calculated.
0 U 450
My = √ =√ ∼ 1.296
kRT 1.3 × 287 × 300

By utilizing Potto–GDC or Table (6.4) one can obtain the following table:

0 0 Ty Py P0y
Mx My Mx My Tx Px P0 x
2.4179 0.50193 0.0 1.296 1.809 6.479 0.49695

Using the above table, the temperature behind the shock is

0 Ty
Ty = Ty = Tx = 1.809 × 300 ∼ 542.7K
Tx

In same manner, it can be done for the pressure ratio as following

0 Py
Py = Py = Px = 6.479 × 1.0 ∼ 6.479[Bar]
Px

The velocity behind the shock wave is obtained (for confirmation)

0 0 √ hmi
Uy = My cy = 1.296 × 1.3 × 287 × 542.7 ∼ 450
sec

Example 6.5:
Gas flows in a tube with a velocity of 450[m/sec]. The static pressure at the tube
is 2Bar and the (static) temperature of 300K. The gas is brought into a complete
stop by a sudden closing a valve. Calculate the velocity and the pressure behind
the reflecting shock. The specific heat ratio can be assumed to be k = 1.4.

S OLUTION
0
The first thing that needs to be done is to find the prime Mach number Mx =
1.2961. Then, the prime properties can be found. At this stage the reflecting shock
velocity is unknown.
Simply using the Potto–GDC provides for the temperature and velocity the following
table:

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
2.0445 0.56995 1.2961 0.0 1.724 4.710 0.70009
6.3. THE MOVING SHOCKS 121
0
If you insist on doing the steps yourself, find the upstream prime Mach, Mx to be
1.2961. Then using Table (6.2) you can find the proper Mx . If this detail is not
sufficient then simply utilize the iterations procedure described earlier and obtain
the following:

Ty 0
i Mx My Tx My
0 2.2961 0.53487 1.9432 0.0
1 2.042 0.57040 1.722 0.0
2 2.045 0.56994 1.724 0.0
3 2.044 0.56995 1.724 0.0
4 2.044 0.56995 1.724 0.0

The table was obtained by utilizing Potto–GDC with the iteration request.

Example 6.6:
What should be the prime Mach number (or the combination of the velocity with
the temperature, for those who like an additional step) in order to double the tem-
perature when the valve is suddenly and totally closed?

S OLUTION
The ratio can be obtained from Table (6.3). It can also be obtained from the sta-
tionary normal shock wave table. Potto-GDC provides for this temperature ratio the
following table:

Ty ρy Py P0 y
Mx My Tx ρx Px P0x
2.3574 0.52778 2.0000 3.1583 6.3166 0.55832

using the required Mx = 2.3574 in the moving shock table provides

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
2.3574 0.52778 0.78928 0.0 2.000 6.317 0.55830

Example 6.7:
A gas is flowing in a pipe with a Mach number of 0.4. Calculate the speed of the
shock when a valve is closed in such a way that the Mach number is reduced by
half. Hint, this is the case of a partially closed valve case in which the ratio of the
prime Mach number is half (the new parameter that is added in the general case).

S OLUTION
Refer to section (6.3.5) for the calculation procedure. Potto-GDC provides the
solution of the above data
122 CHAPTER 6. NORMAL SHOCK
0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
1.1220 0.89509 0.40000 0.20000 1.0789 1.3020 0.99813

If the information about the iterations is needed please refer to the following table.
Ty Py 0
i Mx My Tx Px My
0 1.4000 0.73971 1.2547 2.1200 0.20000
1 1.0045 0.99548 1.0030 1.0106 0.20000
2 1.1967 0.84424 1.1259 1.5041 0.20000
3 1.0836 0.92479 1.0545 1.2032 0.20000
4 1.1443 0.87903 1.0930 1.3609 0.20000
5 1.1099 0.90416 1.0712 1.2705 0.20000
6 1.1288 0.89009 1.0832 1.3199 0.20000
7 1.1182 0.89789 1.0765 1.2922 0.20000
8 1.1241 0.89354 1.0802 1.3075 0.20000
9 1.1208 0.89595 1.0782 1.2989 0.20000
10 1.1226 0.89461 1.0793 1.3037 0.20000
11 1.1216 0.89536 1.0787 1.3011 0.20000
12 1.1222 0.89494 1.0790 1.3025 0.20000
13 1.1219 0.89517 1.0788 1.3017 0.20000
14 1.1221 0.89504 1.0789 1.3022 0.20000
15 1.1220 0.89512 1.0789 1.3019 0.20000
16 1.1220 0.89508 1.0789 1.3020 0.20000
17 1.1220 0.89510 1.0789 1.3020 0.20000
18 1.1220 0.89509 1.0789 1.3020 0.20000
19 1.1220 0.89509 1.0789 1.3020 0.20000
20 1.1220 0.89509 1.0789 1.3020 0.20000
21 1.1220 0.89509 1.0789 1.3020 0.20000
22 1.1220 0.89509 1.0789 1.3020 0.20000

Example 6.8:
A piston is pushing air that flows
in a tube with a Mach number of
M = 0.4 and 300◦ C. The piston is ′
′ Mx = 0.4
accelerated very rapidly and the air M y = 0.8
adjoined the piston obtains Mach
number M = 0.8. Calculate the ve-
locity of the shock created by the
piston in the air. Calculate the time
Fig. -6.21: Schematic of a piston pushing air in a
it takes for the shock to reach the tube.
end of the tube of 1.0m length. Assume that there is no friction and the Fanno flow
model is not applicable.
6.3. THE MOVING SHOCKS 123

S OLUTION
Using the procedure described in this section, the solution is

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
1.2380 0.81942 0.50000 0.80000 1.1519 1.6215 0.98860

The complete iteration is provided below.

Ty Py 0
i Mx My Tx Px My
0 1.5000 0.70109 1.3202 2.4583 0.80000
1 1.2248 0.82716 1.1435 1.5834 0.80000
2 1.2400 0.81829 1.1531 1.6273 0.80000
3 1.2378 0.81958 1.1517 1.6207 0.80000
4 1.2381 0.81940 1.1519 1.6217 0.80000
5 1.2380 0.81943 1.1519 1.6215 0.80000
6 1.2380 0.81942 1.1519 1.6216 0.80000

The time it takes for the shock to reach the end of the cylinder is

length 1
t= =√ = 0.0034[sec]
Us 1.4 × 287 × 300(1.2380 − 0.4)
|{z}
0
cx (Mx −Mx )

Example 6.9:
From the previous example (6.13) calculate the velocity difference between initial
piston velocity and final piston velocity.

S OLUTION
The stationary difference between the two sides of the shock is:
0 0 0 0
∆U =Uy − Ux = cy Uy − cx Ux

 q
Ty

Tx
q  z√ }| { 
 
= 1.4 × 287 × 300 0.8 × 1.1519 −0.5
 

∼ 124.4[m/sec]
124 CHAPTER 6. NORMAL SHOCK

Example 6.10:
An engine is designed so that
two pistons are moving toward
each other (see Figure (6.22)).
1 [Bar]
The air between the pistons is 300 K

at 1[Bar] and 300K. The dis- 40 m/sec 70 m/sec

shock
tance between the two pistons waves
is 1[m]. Calculate the time it will
take for the two shocks to col-
lide.
Fig. -6.22: Figure for Example (6.10)
S OLUTION
This situation is an open valve case where the prime information is given. The
solution is given by equation (6.66), and, it is the explicit analytical solution. For
this case the following table can easily be obtain from Potto–GDC for the left piston
0 0 Ty Py P0y 0
Mx My Mx My Tx Px P0 x Uy cx
1.0715 0.93471 0.0 0.95890 1.047 1.173 0.99959 40.0 347.

while the velocity of the right piston is

0 0 Ty Py P0 y 0
Mx My Mx My Tx Px P0 x Uy cx
1.1283 0.89048 0.0 0.93451 1.083 1.318 0.99785 70.0 347.

The time for the shocks to collide is

length 1[m]
t= = ∼ 0.0013[sec]
Usx 1 + Usx 2 (1.0715 + 1.1283)347.

6.4 Shock Tube

The shock tube is a study tool with very little practical purposes. It is used in many
cases to understand certain phenomena. Other situations can be examined and
extended from these phenomena. A cylinder with two chambers connected by a
diaphragm. On one side the pressure is high, while the pressure on the other side
is low. When the diaphragm is ruptured the gas from the high pressure section
flows into the low pressure section. When the pressure is high enough, a shock
is created that it travels to the low pressure chamber. This is the same case as in
the suddenly opened valve case described previously. At the back of the shock,
expansion waves occur with a reduction of pressure. The temperature is known to
reach several thousands degrees in a very brief period of time. The high pressure
chamber is referred to in the literature is the driver section and the low section is
referred to as the expansion section.
6.4. SHOCK TUBE 125

Initially, the gas from the driver 5expansion

front
4 3 2 1

section is coalescing from small

shock waves into a large shock Diaphragm

t
wave. In this analysis, it is
assumed that this time is es- some where reflective

sentially zero. Zone 1 is an reflective wave shock

wave

undisturbed gas and zone 2 is

e
ac
an area where the shock al-

rf
Su
t1

back

t
ready passed. The assumption

ac
e
wav

fr

nt
ck

on
sho

Co
t
is that the shock is very sharp
with zero width. On the other
side, the expansion waves are distance

moving into the high pressure

chamber i.e. the driver section. Fig. -6.23: The shock tube schematic with a pressure
The shock is moving at a super- ”diagram.”
sonic speed (it depends on the
definition, i.e., what reference
temperature is being used) and
the medium behind the shock is also moving but at a velocity, U2 , which can be
supersonic or subsonic in stationary coordinates. The velocities in the expansion
chamber vary between three zones. In zone 3 is the original material that was in
the high pressure chamber but is now the same pressure as zone 2. Zone 4 is
where the gradual transition occurs between original high pressure to low pres-
sure. The boundaries of zone 4 are defined by initial conditions. The expansion
front is moving at the local speed of sound in the high pressure section. The ex-
pansion back front is moving at the local speed of sound velocity but the actual gas
is moving in the opposite direction in U2 . In fact, material in the expansion chamber
and the front are moving to the left while the actual flow of the gas is moving to the
right (refer to Figure (6.23)). In zone 5, the velocity is zero and the pressure is in
its original value.
The properties in the different zones have different relationships. The relationship
between zone 1 and zone 2 is that of a moving shock into still medium (again, this
is a case of sudden opened valve). The material in zone 2 and 3 is moving at the
same velocity (speed) but the temperature and the entropy are different, while the
pressure in the two zones are the same. The pressure, the temperature and their
properties in zone 4 aren’t constant and continuous between the conditions in zone
3 to the conditions in zone 5. The expansion front wave velocity is larger than the
velocity at the back front expansion wave velocity. Zone 4 is expanding during the
initial stage (until the expansion reaches the wall).
The shock tube is a relatively small√length 1 − 2[m] and the typical velocity is in the
range of the speed of sound, c ∼ 340 thus the whole process takes only a few
milliseconds or less. Thus, these kinds of experiments require fast recording de-
vices (a relatively fast camera and fast data acquisition devices.). A typical design
problem of a shock tube is finding the pressure to achieve the desired temperature
126 CHAPTER 6. NORMAL SHOCK

or Mach number. The relationship between the different properties was discussed
earlier and because it is a common problem, a review of the material is provided
thus far.
The following equations were developed earlier and are repeated here for clarifica-
tion. The pressure ratio between the two sides of the shock is
µ ¶
P2 k−1 2k
= Ms1 2 − 1 (6.84)
P1 k+1 k−1

Rearranging equation (6.84) becomes

r
k − 1 k + 1 P2
Ms1 = + (6.85)
2k 2k P1

Or expressing the velocity as

r
k − 1 k + 1 P2
Us = Ms1 c1 = c1 + (6.86)
2k 2k P1

And the velocity ratio between the two sides of the shock is
k+1 P2
U1 ρ2 1 + k−1
= = k+1 P2P1 (6.87)
U2 ρ2 k−1 P 1

The fluid velocity in zone 2 is the same

µ ¶
U2
U2 0 = Us − U2 = Us 1 − (6.88)
Us

From the mass conservation, it follows that

U2 ρ1
= (6.89)
Us ρ2

v
r u k+1 P2
0 k − 1 k + 1 P2 u
t1 − k−1 + P1
U2 = c1 + k+1 P2
(6.90)
2k 2k P1 1 + k−1 P1

After rearranging equation (6.90) the result is

v
µ ¶u
u 2k
c1 P2
− 1 t P2k+1
0
U2 = k−1
(6.91)
k P1 P 1+k1
6.4. SHOCK TUBE 127

On the isentropic side, in zone 4, taking the derivative of the continuity equation,
d(ρU ) = 0, and dividing by the continuity equation the following is obtained:
dρ dU
=− (6.92)
ρ c
Since the process in zone 4 is isentropic, applying the isentropic relationship (T ∝
ρk−1 ) yields
r µ ¶ k−1
c T ρ 2
= = (6.93)
c5 T5 ρ5

From equation (6.92) it follows that

µ ¶ k−1
dρ ρ 2
dU = −c = c5 dρ (6.94)
ρ ρ5

Equation (6.94) can be integrated as follows:

Z U3 Z ρ3 µ ¶ k−1
ρ 2
dU = c5 dρ (6.95)
U5 =0 ρ5 ρ5

The results of the integration are

à µ ¶ k−1 !
2c5 ρ3 2
U3 = 1− (6.96)
k−1 ρ5

Or in terms of the pressure ratio as

à µ ¶ k−1 !
2c5 P3 2k
U3 = 1− (6.97)
k−1 P5
0
As it was mentioned earlier the velocity at points 2 and 3 are identical, hence
equation (6.97) and equation (6.91) can be combined to yield
à ! v
µ ¶ k−1 µ ¶u
u 2k
2c5 P3 2k c1 P2
1− = − 1 t P2k+1
k−1
(6.98)
k−1 P5 k P1 P 1+k 1

After some rearrangement, equation (6.98) is transformed into

 ³ ´ − k−1
2k

P5 P2  (k − −1 1) cc51

P5
P3
= 1 − r  (6.99)
P1 P1  √ ³ ´
P2
2k 2k + (k + 1) P1 − 1
128 CHAPTER 6. NORMAL SHOCK

Or in terms of the Mach number, Ms1

µ ¶" ¡ ¢ #− k−1
2k

P5 k1 − 1 2k
k−1 c1
k+1 c5 Ms1 2 − 1
= Ms1 2 − 1 1− (6.100)
P1 k+1+1 k1 − 1 Ms1

Using the Rankine–Hugoniot relationship and the perfect gas model, the following
is obtained:

k1 −1 P2
T2 1+ k1 +1 P1
= k1 −1 P1
(6.101)
T1 1+ k1 +1 P2

By utilizing the isentropic relationship for zone 3 to 5 results in

µ ¶ k5k−1 Ã P ! k5k−1
2 5
T3 P3 5
P1
= = P5
(6.102)
T5 P5 P1

Example 6.11:
A shock tube with an initial pressure ratio of P
P1 = 20 and an initial temperature of
5

300K. Find the shock velocity and temperature behind the shock if the pressure
ratio is P
P1 = 40?
5

S OLUTION

6.5 Shock with Real Gases

6.6 Shock in Wet Steam

6.7 Normal Shock in Ducts

The flow in ducts is related to boundary layer issues. For a high Reynolds number,
the assumption of an uniform flow in the duct is closer to reality. It is normal to
have a large Mach number with a large Re number. In that case, the assumptions
in construction of these models are acceptable and reasonable.
6.8. MORE EXAMPLES FOR MOVING SHOCKS 129

6.8 More Examples for Moving Shocks

Example 6.12:
This problem was taken from
the real industrial manufactur- distance
ing world. An engineer is re-
quired to design a cooling sys-
tem for a critical electronic de-
vice. The temperature should valve exit
not increase above a certain
value. In this system, air is Fig. -6.24: Figure for Example (6.13)
supposed to reach the pipe exit
as quickly as possible when the
valve is opened (see Figure (6.26)). opening valve problem The distance between
between the valve and the pipe exit is 3[m]. The conditions upstream of the valve
are 30[Bar] and 27◦ C . Assume that there isn’t any resistance whatsoever in the
pipe. The ambient temperature is 27◦ C and 1[Bar]. Assume that the time scale for
opening the valve is significantly smaller than the typical time of the pipe (totally un-
realistic even though the valve manufacture claims of 0.0002 [sec] to be opened).
After building the system, the engineer notices that the system does not cool the
device fast enough and proposes to increase the pressure and increase the diam-
eter of the pipe. Comment on this proposal. Where any of these advises make any
sense in the light of the above assumptions? What will be your recommendations
to the manufacturing company? Plot the exit temperature and the mass flow rate
as a function of the time.

S OLUTION
This problem is known as the suddenly open valve problem in which the shock
choking phenomenon occurs. The time it takes for the shock to travel from the
P
valve depends on the pressure ratio Pxy = 30
0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
5.0850 0.41404 0.0 1.668 5.967 30.00 0.057811

The direct calculation will be by using the “upstream” Mach number, Mx = Msx =
5.0850. Therefore, the time is
distance 3
t= √ = ∼ 0.0017[sec]
Msx kRTx 5.0850sqrt1.4 × 287 × 300
The mass flow rate after reaching the exit under these assumptions remains con-
stant until the uncooled material reaches the exit. The time it takes for the material
from the valve to reach the exit is
distance 3
t= 0p = ∼ 0.0021[sec]
My kRTy 1.668sqrt1.4 × 287 × 300 × 5.967
130 CHAPTER 6. NORMAL SHOCK

During that difference of time the ma-

terial is get heated instead of cooling
down because of the high temperature.
The suggestion of the engineer to in- Mass Flow
Rate
crease the pressure will decrease the
time but will increase the temperature
at the exit during this critical time pe- Velocity
riod. Thus, this suggestion contradicts
the purpose of the required manufactur-
ing needs. Time[Msec]

To increase the pipe diameter will not Fig. -6.25: The results for Example (6.13)
change the temperature and therefore
will not change the effects of heating. It
can only increase the rate after the initial heating spike

A possible solution is to have the valve very close to the pipe exit. Thus, the heating
time is reduced significantly. There is also the possibility of steps increase in which
every step heat released will not be enough to over heat the device. The last
possible requirement a programmable valve and very fast which its valve probably
exceed the moving shock the valve downstream. The plot of the mass flow rate
and the velocity are given in Figure (6.27).

Example 6.13:
This problem was taken from
the real industrial manufactur- distance
ing world. Engineer is given to
design a cooling system for crit-
ical electronic devise. The tem-
perature should not increase valve exit
above certain value. In this sys-
tem air is supposed to reach the Fig. -6.26: Figure for example (6.13)
pipe exit as quickly as possible
when the valve is opened (see
Figure (6.26)). The distance between the valve to the exit is 3[m]. The condi-
tions upstream the valve are 30[Bar] and 27◦ C . Assume that there isn’t resistance
whatsoever in the pipe. The ambient temperature is 27◦ C and 1[Bar]. Assume that
the time scale for opening the valve is significantly smaller than the typical time of
the pipe (totally unrealistic even though the valve manufacture claims that it takes
0.0002 [sec] to open completely the valve). After building the system the engi-
neer notices that the system does not cool the device fast enough and proposes
to increase the pressure and increase the diameter of the pipe. Comment on this
proposal. Did the suggestions make any sense in the light of the above assump-
tions? What will be your recommendations to the manufacturing company? Plot
the exit temperature and mass flow rate as function of the time.
6.8. MORE EXAMPLES FOR MOVING SHOCKS 131

S OLUTION
This problem is known as the suddenly open valve problem in which the shock
choking phenomenon occurs. The time it takes for the shock to travel from the
P
valve depends of the pressure ratio Pxy = 30.
0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
5.0850 0.41404 0.0 1.668 5.967 30.00 0.057811

The direct calculation will be by using the “upstream” Mach number, Mx = Msx =
5.0850. Therefore, the time is
distance 3
t= √ = ∼ 0.0017[sec]
Msx kRTx 5.0850sqrt1.4 × 287 × 300
The mass flow rate after reaching the exit under these assumptions remains con-
stant until the uncooled material reaches the exit. The time it takes the material
from the valve to reach the exit is
distance 3
t= 0p = ∼ 0.0021[sec]
My kRTy 1.668sqrt1.4 × 287 × 300 × 5.967
During that difference of time the ma-
terial is get heated instead of cold to
a high temperature. The suggestion of
the engineer to increase the pressure Mass Flow
will decrease the time but will increase Rate
the temperature at the exit during this
critical time period. Thus, this sugges-
tion contradicts the purpose of required Velocity
the manufacturing needs.
To increase the pipe diameter will not
change the temperature and therefore Time[Msec]

will not change the effects of heating. It

can only increase the rate after the ini- Fig. -6.27: The results for Example (6.13)
tial heating spike.
A possible solution is to have the valve very close to the pipe exit. Thus, the heating
time is reduced significantly. There is also the possibility of steps increase in which
every step heat released will not be enough to over heat the device. The last
possible requirement programmable value and very fast which its valve probably
exceed the moving shock the valve downstream. The plot of the mass flow rate
and the velocity are given in Figure (6.27).

Example 6.14:
Example (6.13) deals with a damaging of electronic product by the temperature
increase. Try to estimate the temperature increase of the product. Plot the pipe
exit temperature as a function of the time.
132 CHAPTER 6. NORMAL SHOCK

S OLUTION
To be developed

6.9 Tables of Normal Shocks, k = 1.4 Ideal Gas

Table -6.1: The shock wave table for k = 1.4

Ty ρy Py P0y
Mx My Tx ρx Px P0x
1.00 1.00000 1.00000 1.00000 1.00000 1.00000
1.05 0.95313 1.03284 1.08398 1.11958 0.99985
1.10 0.91177 1.06494 1.16908 1.24500 0.99893
1.15 0.87502 1.09658 1.25504 1.37625 0.99669
1.20 0.84217 1.12799 1.34161 1.51333 0.99280
1.25 0.81264 1.15938 1.42857 1.65625 0.98706
1.30 0.78596 1.19087 1.51570 1.80500 0.97937
1.35 0.76175 1.22261 1.60278 1.95958 0.96974
1.40 0.73971 1.25469 1.68966 2.12000 0.95819
1.45 0.71956 1.28720 1.77614 2.28625 0.94484
1.50 0.70109 1.32022 1.86207 2.45833 0.92979
1.55 0.68410 1.35379 1.94732 2.63625 0.91319
1.60 0.66844 1.38797 2.03175 2.82000 0.89520
1.65 0.65396 1.42280 2.11525 3.00958 0.87599
1.70 0.64054 1.45833 2.19772 3.20500 0.85572
1.75 0.62809 1.49458 2.27907 3.40625 0.83457
1.80 0.61650 1.53158 2.35922 3.61333 0.81268
1.85 0.60570 1.56935 2.43811 3.82625 0.79023
1.90 0.59562 1.60792 2.51568 4.04500 0.76736
1.95 0.58618 1.64729 2.59188 4.26958 0.74420
2.00 0.57735 1.68750 2.66667 4.50000 0.72087
2.05 0.56906 1.72855 2.74002 4.73625 0.69751
2.10 0.56128 1.77045 2.81190 4.97833 0.67420
2.15 0.55395 1.81322 2.88231 5.22625 0.65105
2.20 0.54706 1.85686 2.95122 5.48000 0.62814
2.25 0.54055 1.90138 3.01863 5.73958 0.60553
2.30 0.53441 1.94680 3.08455 6.00500 0.58329
2.35 0.52861 1.99311 3.14897 6.27625 0.56148
2.40 0.52312 2.04033 3.21190 6.55333 0.54014
2.45 0.51792 2.08846 3.27335 6.83625 0.51931
2.50 0.51299 2.13750 3.33333 7.12500 0.49901
2.75 0.49181 2.39657 3.61194 8.65625 0.40623
3.00 0.47519 2.67901 3.85714 10.33333 0.32834
3.25 0.46192 2.98511 4.07229 12.15625 0.26451
6.9. TABLES OF NORMAL SHOCKS, K = 1.4 IDEAL GAS 133
Table -6.1: The shock wave table for k = 1.4 (continue)

Ty ρy Py P0y
Mx My Tx ρx Px P0x
3.50 0.45115 3.31505 4.26087 14.12500 0.21295
3.75 0.44231 3.66894 4.42623 16.23958 0.17166
4.00 0.43496 4.04688 4.57143 18.50000 0.13876
4.25 0.42878 4.44891 4.69919 20.90625 0.11256
4.50 0.42355 4.87509 4.81188 23.45833 0.09170
4.75 0.41908 5.32544 4.91156 26.15625 0.07505
5.00 0.41523 5.80000 5.00000 29.00000 0.06172
5.25 0.41189 6.29878 5.07869 31.98958 0.05100
5.50 0.40897 6.82180 5.14894 35.12500 0.04236
5.75 0.40642 7.36906 5.21182 38.40625 0.03536
6.00 0.40416 7.94059 5.26829 41.83333 0.02965
6.25 0.40216 8.53637 5.31915 45.40625 0.02498
6.50 0.40038 9.15643 5.36508 49.12500 0.02115
6.75 0.39879 9.80077 5.40667 52.98958 0.01798
7.00 0.39736 10.46939 5.44444 57.00000 0.01535
7.25 0.39607 11.16229 5.47883 61.15625 0.01316
7.50 0.39491 11.87948 5.51020 65.45833 0.01133
7.75 0.39385 12.62095 5.53890 69.90625 0.00979
8.00 0.39289 13.38672 5.56522 74.50000 0.00849
8.25 0.39201 14.17678 5.58939 79.23958 0.00739
8.50 0.39121 14.99113 5.61165 84.12500 0.00645
8.75 0.39048 15.82978 5.63218 89.15625 0.00565
9.00 0.38980 16.69273 5.65116 94.33333 0.00496
9.25 0.38918 17.57997 5.66874 99.65625 0.00437
9.50 0.38860 18.49152 5.68504 105.12500 0.00387
9.75 0.38807 19.42736 5.70019 110.73958 0.00343
10.00 0.38758 20.38750 5.71429 116.50000 0.00304

Table -6.2: Table for a Reflective Shock from a suddenly closed end (k=1.4)

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
1.006 0.99403 0.01 0.0 1.004 1.014 1.00000
1.012 0.98812 0.02 0.0 1.008 1.028 1.00000
1.018 0.98227 0.03 0.0 1.012 1.043 0.99999
1.024 0.97647 0.04 0.0 1.016 1.057 0.99998
1.030 0.97074 0.05 0.0 1.020 1.072 0.99997
1.037 0.96506 0.06 0.0 1.024 1.087 0.99994
1.043 0.95944 0.07 0.0 1.028 1.102 0.99991
134 CHAPTER 6. NORMAL SHOCK
Table -6.2: Table for Reflective Shock from suddenly closed valve (end) (k=1.4)(continue)

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
1.049 0.95387 0.08 0.0 1.032 1.118 0.99986
1.055 0.94836 0.09 0.0 1.036 1.133 0.99980
1.062 0.94291 0.10 0.0 1.040 1.149 0.99973
1.127 0.89128 0.20 0.0 1.082 1.316 0.99790
1.196 0.84463 0.30 0.0 1.126 1.502 0.99317
1.268 0.80251 0.40 0.0 1.171 1.710 0.98446
1.344 0.76452 0.50 0.0 1.219 1.941 0.97099
1.423 0.73029 0.60 0.0 1.269 2.195 0.95231
1.505 0.69946 0.70 0.0 1.323 2.475 0.92832
1.589 0.67171 0.80 0.0 1.381 2.780 0.89918
1.676 0.64673 0.90 0.0 1.442 3.112 0.86537
1.766 0.62425 1.00 0.0 1.506 3.473 0.82755
1.858 0.60401 1.10 0.0 1.576 3.862 0.78652
1.952 0.58578 1.20 0.0 1.649 4.280 0.74316
2.048 0.56935 1.30 0.0 1.727 4.728 0.69834
2.146 0.55453 1.40 0.0 1.810 5.206 0.65290
2.245 0.54114 1.50 0.0 1.897 5.715 0.60761
2.346 0.52904 1.60 0.0 1.990 6.256 0.56312
2.448 0.51808 1.70 0.0 2.087 6.827 0.51996
2.552 0.50814 1.80 0.0 2.189 7.431 0.47855
2.656 0.49912 1.90 0.0 2.297 8.066 0.43921
2.762 0.49092 2.00 0.0 2.410 8.734 0.40213
3.859 0.43894 3.00 0.0 3.831 17.21 0.15637
5.000 0.41523 4.00 0.0 5.800 29.00 0.061716
6.162 0.40284 5.00 0.0 8.325 44.14 0.026517
7.336 0.39566 6.00 0.0 11.41 62.62 0.012492
8.517 0.39116 7.00 0.0 15.05 84.47 0.00639
9.703 0.38817 8.00 0.0 19.25 1.1E+2 0.00350
10.89 0.38608 9.00 0.0 24.01 1.4E+2 0.00204
12.08 0.38457 10.0 0.0 29.33 1.7E+2 0.00125

Table -6.3: Table for shock propagating from suddenly opened valve (k=1.4)

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0x
1.006 0.99402 0.0 0.01 1.004 1.014 1.00000
1.012 0.98807 0.0 0.02 1.008 1.028 1.00000
1.018 0.98216 0.0 0.03 1.012 1.043 0.99999
1.024 0.97629 0.0 0.04 1.016 1.058 0.99998
6.9. TABLES OF NORMAL SHOCKS, K = 1.4 IDEAL GAS 135
Table -6.3: Table for shock propagating from suddenly opened valve (k=1.4)

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
1.031 0.97045 0.0 0.05 1.020 1.073 0.99996
1.037 0.96465 0.0 0.06 1.024 1.088 0.99994
1.044 0.95888 0.0 0.07 1.029 1.104 0.99990
1.050 0.95315 0.0 0.08 1.033 1.120 0.99985
1.057 0.94746 0.0 0.09 1.037 1.136 0.99979
1.063 0.94180 0.0 0.10 1.041 1.152 0.99971
1.133 0.88717 0.0 0.20 1.086 1.331 0.99763
1.210 0.83607 0.0 0.30 1.134 1.541 0.99181
1.295 0.78840 0.0 0.40 1.188 1.791 0.98019
1.390 0.74403 0.0 0.50 1.248 2.087 0.96069
1.495 0.70283 0.0 0.60 1.317 2.441 0.93133
1.613 0.66462 0.0 0.70 1.397 2.868 0.89039
1.745 0.62923 0.0 0.80 1.491 3.387 0.83661
1.896 0.59649 0.0 0.90 1.604 4.025 0.76940
2.068 0.56619 0.0 1.00 1.744 4.823 0.68907
2.269 0.53817 0.0 1.100 1.919 5.840 0.59699
2.508 0.51223 0.0 1.200 2.145 7.171 0.49586
2.799 0.48823 0.0 1.300 2.450 8.975 0.38974
3.167 0.46599 0.0 1.400 2.881 11.54 0.28412
3.658 0.44536 0.0 1.500 3.536 15.45 0.18575
4.368 0.42622 0.0 1.600 4.646 22.09 0.10216
5.551 0.40843 0.0 1.700 6.931 35.78 0.040812
8.293 0.39187 0.0 1.800 14.32 80.07 0.00721
8.821 0.39028 0.0 1.810 16.07 90.61 0.00544
9.457 0.38870 0.0 1.820 18.33 1.0E + 2 0.00395
10.24 0.38713 0.0 1.830 21.35 1.2E + 2 0.00272
11.25 0.38557 0.0 1.840 25.57 1.5E + 2 0.00175
12.62 0.38402 0.0 1.850 31.92 1.9E + 2 0.00101
14.62 0.38248 0.0 1.860 42.53 2.5E + 2 0.000497
17.99 0.38096 0.0 1.870 63.84 3.8E + 2 0.000181
25.62 0.37944 0.0 1.880 1.3E+2 7.7E + 2 3.18E−5
61.31 0.37822 0.0 1.888 7.3E+2 4.4E + 3 0.0
62.95 0.37821 0.0 1.888 7.7E+2 4.6E + 3 0.0
64.74 0.37820 0.0 1.888 8.2E+2 4.9E + 3 0.0
66.69 0.37818 0.0 1.888 8.7E+2 5.2E + 3 0.0
68.83 0.37817 0.0 1.888 9.2E+2 5.5E + 3 0.0
71.18 0.37816 0.0 1.889 9.9E+2 5.9E + 3 0.0
73.80 0.37814 0.0 1.889 1.1E+3 6.4E + 3 0.0
76.72 0.37813 0.0 1.889 1.1E+3 6.9E + 3 0.0
136 CHAPTER 6. NORMAL SHOCK
Table -6.3: Table for shock propagating from suddenly opened valve (k=1.4)

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
80.02 0.37812 0.0 1.889 1.2E+3 7.5E + 3 0.0
83.79 0.37810 0.0 1.889 1.4E+3 8.2E + 3 0.0

Table -6.4: Table for shock propagating from a suddenly opened valve (k=1.3)

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
1.0058 0.99427 0.0 0.010 1.003 1.013 1.00000
1.012 0.98857 0.0 0.020 1.006 1.026 1.00000
1.017 0.98290 0.0 0.030 1.009 1.040 0.99999
1.023 0.97726 0.0 0.040 1.012 1.054 0.99998
1.029 0.97166 0.0 0.050 1.015 1.067 0.99997
1.035 0.96610 0.0 0.060 1.018 1.081 0.99995
1.042 0.96056 0.0 0.070 1.021 1.096 0.99991
1.048 0.95506 0.0 0.080 1.024 1.110 0.99987
1.054 0.94959 0.0 0.090 1.028 1.125 0.99981
1.060 0.94415 0.0 0.100 1.031 1.140 0.99975
1.126 0.89159 0.0 0.200 1.063 1.302 0.99792
1.197 0.84227 0.0 0.300 1.098 1.489 0.99288
1.275 0.79611 0.0 0.400 1.136 1.706 0.98290
1.359 0.75301 0.0 0.500 1.177 1.959 0.96631
1.452 0.71284 0.0 0.600 1.223 2.252 0.94156
1.553 0.67546 0.0 0.700 1.274 2.595 0.90734
1.663 0.64073 0.0 0.800 1.333 2.997 0.86274
1.785 0.60847 0.0 0.900 1.400 3.471 0.80734
1.919 0.57853 0.0 1.00 1.478 4.034 0.74136
2.069 0.55074 0.0 1.100 1.570 4.707 0.66575
2.236 0.52495 0.0 1.200 1.681 5.522 0.58223
2.426 0.50100 0.0 1.300 1.815 6.523 0.49333
2.644 0.47875 0.0 1.400 1.980 7.772 0.40226
2.898 0.45807 0.0 1.500 2.191 9.367 0.31281
3.202 0.43882 0.0 1.600 2.467 11.46 0.22904
3.576 0.42089 0.0 1.700 2.842 14.32 0.15495
4.053 0.40418 0.0 1.800 3.381 18.44 0.093988
4.109 0.40257 0.0 1.810 3.448 18.95 0.088718
4.166 0.40097 0.0 1.820 3.519 19.49 0.083607
4.225 0.39938 0.0 1.830 3.592 20.05 0.078654
4.286 0.39780 0.0 1.840 3.669 20.64 0.073863
4.349 0.39624 0.0 1.850 3.749 21.25 0.069233
6.9. TABLES OF NORMAL SHOCKS, K = 1.4 IDEAL GAS 137
Table -6.4: Table for shock propagating from a suddenly opened valve (k=1.3)

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
4.415 0.39468 0.0 1.860 3.834 21.90 0.064766
4.482 0.39314 0.0 1.870 3.923 22.58 0.060462
4.553 0.39160 0.0 1.880 4.016 23.30 0.056322
4.611 0.39037 0.0 1.888 4.096 23.91 0.053088
4.612 0.39035 0.0 1.888 4.097 23.91 0.053053
4.613 0.39034 0.0 1.888 4.098 23.92 0.053018
4.613 0.39033 0.0 1.888 4.099 23.93 0.052984
4.614 0.39031 0.0 1.888 4.099 23.93 0.052949
4.615 0.39030 0.0 1.889 4.100 23.94 0.052914
4.615 0.39029 0.0 1.889 4.101 23.95 0.052879
4.616 0.39027 0.0 1.889 4.102 23.95 0.052844
4.616 0.39026 0.0 1.889 4.103 23.96 0.052809
4.617 0.39025 0.0 1.889 4.104 23.97 0.052775
138 CHAPTER 6. NORMAL SHOCK
 CHAPTER 7
Normal Shock in Variable Duct
Areas

In the previous two chapters, the flow in a variable area duct and a normal shock
(discontinuity) were discussed. A discussion of the occurrences of shock in flow in
a variable is presented. As it is was presented before, the shock can occur only in
steady state when there is a supersonic flow. but also in steady state cases when
there is no supersonic flow (in stationary coordinates). As it was shown in Chapter
6, the gas has to pass through a converging–diverging nozzle to obtain a super-
sonic flow.
In the previous chapter,
the flow in a convergent–
divergent nuzzle was pre-
sented when the pressure
ratio was above or below
the special range. This
Chapter will present the
flow in this special range of
c PB = P
pressure ratios. It is inter- P a
0

esting to note that a normal 0P Subsonic

M <1
shock must occur in these d
ic flo
subson shock
a
w afte
r

situations (pressure ratios).

In Figure (7.1) the
Supersonic
M >1 b

reduced pressure distribu- distance, x

tion in the converging–

diverging nozzle is shown Fig. -7.1: The flow in the nozzle with different back pres-
in its whole range of pres- sures.

139
140 CHAPTER 7. NORMAL SHOCK IN VARIABLE DUCT AREAS

sure ratios. When the pres-

sure ratio, P B is between point “a” and point “b” the flow is different from what was
discussed before. In this case, no continuous pressure possibly can exists. Only
in one point where P B = Pb continuous pressure exist. If the back pressure, P B is
smaller than Pb a discontinuous point (a shock) will occur. In conclusion, once the
flow becomes supersonic, only exact geometry can achieve continuous pressure
flow.
In the literature, some refer to a nozzle with an area ratio such point b as
above the back pressure and it is referred to as an under–expanded nozzle. In
the under–expanded case, the nozzle doesn’t provide the maximum thrust possi-
ble. On the other hand, when the nozzle exit area is too large a shock will occur
and other phenomenon such as plume will separate from the wall inside the noz-
zle. This nozzle is called an over–expanded nozzle. In comparison of nozzle
performance for rocket and aviation, the over–expanded nozzle is worse than the
under–expanded nozzle because the nozzle’s large exit area results in extra drag.
The location of the shock is determined by geometry to achieve the right
back pressure. Obviously if the back pressure, P B , is lower than the critical value
(the only value that can achieve continuous pressure) a shock occurs outside of the
nozzle. If the back pressure is within the range of Pa to Pb than the exact location
determines that after the shock the subsonic branch will match the back pressure.
The first example is
for academic reasons. It
has to be recognized that the
troat
Aexit = 9[ m2℄
shock wave isn’t easily visible
(see Mach’s photography tech- P0 = 4[Bar℄
niques). Therefore, this exam- T0 = 308K exit
point "e"

ple provides a demonstration of

the calculations required for the x y

location even if it isn’t realis- A = 3[ m ℄

 2
Asho k = 6[ m2℄
tic. Nevertheless, this example
will provide the fundamentals to Fig. -7.2: A nozzle with normal shock
explain the usage of the tools
(equations and tables) that were developed so far.

Example 7.1:
A large tank with compressed air is attached into a converging–diverging nozzle
at pressure 4[Bar] and temperature of 35◦ C. Nozzle throat area is 3[cm2 ] and the
exit area is 9[cm2 ] . The shock occurs in a location where the cross section area
is 6[cm2 ] . Calculate the back pressure and the temperature of the flow. (It should
be noted that the temperature of the surrounding is irrelevant in this case.) Also
determine the critical points for the back pressure (point “a” and point “b”).

S OLUTION
Since the key word “large tank” was used that means that the stagnation tempera-
 141

ture and pressure are known and equal to the conditions in the tank.
First, the exit Mach number has to be determined. This Mach number can be
calculated by utilizing the isentropic relationship from the large tank to the shock
(point “x”). Then the relationship developed for the shock can be utilized to calcu-
late the Mach number after the shock, (point “y”). From the Mach number after the
shock, My , the Mach number at the exit can be calculated by utilizing the isentropic
relationship.
It has to be realized that for a large tank, the inside conditions are essentially the
stagnation conditions (this statement is said without a proof, but can be shown
that the correction is negligible for a typical dimension ratio that is over 100. For
example, in the case of ratio of 100 the Mach number is 0.00587 and the error
is less than %0.1). Thus, the stagnation temperature and pressure are known
T0 = 308K and P0 = 4[Bar]. The star area (the throat area), A∗ , before the shock
is known and given as well.
Ax 6
∗
= =2
A 3
With this ratio (A/A∗ = 2) utilizing the Table (6.1) or equation (5.48) or the GDC–
Potto, the Mach number, Mx is about 2.197 as shown table below:

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
2.1972 0.50877 0.18463 2.0000 0.09393 0.18787

With this Mach number, Mx = 2.1972 the Mach number, My can be obtained. From
equation (6.22) or from Table (5.2) My ∼
= 0.54746. With these values, the subsonic
branch can be evaluated for the pressure and temperature ratios.

Ty ρy Py P0y
Mx My Tx ρx Px P0 x
2.1972 0.54743 1.8544 2.9474 5.4656 0.62941

From Table (5.2) or from equation (5.11) the following Table for the isentropic rela-
tionship is obtained

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
0.54743 0.94345 0.86457 1.2588 0.81568 1.0268

Again utilizing the isentropic relationship the exit conditions can be evaluated. With
known Mach number the new star area ratio, Ay /A∗ is known and the exit area can
be calculated as
Ae Ae Ay 9
= × ∗ = 1.2588 × = 1.8882
A∗ Ay A 6
Ae
with this area ratio, A∗ = 1.8882, one can obtain using the isentropic relationship
as
142 CHAPTER 7. NORMAL SHOCK IN VARIABLE DUCT AREAS

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
0.32651 0.97912 0.94862 1.8882 0.92882 1.7538

Since the stagnation pressure is constant as well the stagnation temperature, the
exit conditions can be calculated.
µ ¶µ ¶µ ¶µ ¶
Pexit P0 Py Px
Pexit = P0
P0 Py Px P0
µ ¶
1
=0.92882 × × 5.466 × 0.094 × 4
0.81568
∼
=2.34[Bar]

The exit temperature is

µ ¶µ ¶µ ¶µ ¶
Texit T0 Ty Tx
Texit = T0
T0 Ty Tx T0
µ ¶
1
=0.98133 × × 1.854 × 0.509 × 308
0.951
∼
=299.9K

For the “critical” points ”a” and ”b” are the points that the shock doesn’t occur and
yet the flow achieve Mach equal 1 at the throat. In that case we don’t have to go
through that shock transition. Yet we have to pay attention that there two possible
back pressures that can “achieve” it or target. The area ratio for both cases, is
A/A∗ = 3 In the subsonic branch (either using equation or the isentropic Table or
GDC-Potto as

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
0.19745 0.99226 0.98077 3.0000 0.97318 2.9195
2.6374 0.41820 0.11310 3.0000 0.04730 0.14190

µ ¶
Pexit
Pexit = P0 = 0.99226 × 4 ∼
=3.97[Bar]
P0

For the supersonic sonic branch

µ ¶
Pexit
Pexit = P0 = 0.41820 × 4 ∼
=1.6728[Bar]
P0
 143

It should be noted that the flow rate is constant and maximum for any point beyond
the point ”a” even if the shock is exist. The flow rate is expressed as following
 ∗ 
P
z }| {
P∗ 
 
ρ∗
 P0 
z }| { M =1 P
 0  c ³ ∗ ´
z }| { P r
∗ ∗ P ∗ z}|{ √ P0 P0 T∗
ṁ = ρ A U = A cM =   A kRT ∗ = ³ ´ A kR T0
RT ∗ R T∗
T T0
T 0
T∗  0

R T0 T0 

| {z }
T∗

The temperature and pressure at the throat are:

µ ∗¶
∗ T
T = T0 = 0.833 × 308 = 256.7K
T0

The temperature at the throat reads

µ ∗¶
∗ P
P = P0 = 0.5283 × 4 = 2.113[Bar]
P0

The speed of sound is

√
c= 1.4 × 287 × 256.7 = 321.12[m/sec]

And the mass flow rate reads

4105
ṁ = 3 × 10−4 × 321.12 = 0.13[kg/sec]
287 × 256.7
It is interesting to note that in this case the choking condition is obtained (M = 1)
when the back pressure just reduced to less than 5% than original pressure (the
pressure in the tank). While the pressure to achieve full supersonic flow through
the nozzle the pressure has to be below the 42% the original value. Thus, over
50% of the range of pressure a shock occores some where in the nozzle. In fact
in many industrial applications, these kind situations exist. In these applications a
small pressure difference can produce a shock wave and a chock flow.
For more practical example1 from industrial application point of view.

Example 7.2:
In the data from the above example (7.1) where would be shock’s location when
the back pressure is 2[Bar]?
1 The meaning of the word practical is that in reality the engineer does not given the opportunity to

determine the location of the shock but rather information such as pressures and temperature.
144 CHAPTER 7. NORMAL SHOCK IN VARIABLE DUCT AREAS

S OLUTION
The solution procedure is similar to what was shown in previous Example (7.1).
The solution process starts at the nozzle’s exit and progress to the entrance.
The conditions in the tank are again the stagnation conditions. Thus, the exit pres-
sure is between point “a” and point “b”. It follows that there must exist a shock in
the nozzle. Mathematically, there are two main possible ways to obtain the solu-
tion. In the first method, the previous example information used and expanded. In
fact, it requires some iterations by “smart” guessing the different shock locations.
The area (location) that the previous example did not “produce” the “right” solution
(the exit pressure was 2.113[Bar]. Here, the needed pressure is only 2[Bar] which
means that the next guess for the shock location should be with a larger area2 .
The second (recommended) method is noticing that the flow is adiabatic and the
mass flow rate is constant which means that the ratio of the P0 × A∗ = Py0 × A∗ |@y
(upstream conditions are known, see also equation (5.71)).

Pexit Aexit Pexit Aexit 2×9

∗ = ∗ = = 1.5[unitless!]
Px 0 × Ax Py 0 × Ay 4×3

With the knowledge of the ratio PP0 A

A
∗ which was calculated and determines the exit

Mach number. Utilizing the Table (5.2) or the GDC-Potto provides the following
table is obtained

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.38034 0.97188 0.93118 1.6575 0.90500 1.5000 0.75158

With these values the relationship between the stagnation pressures of the shock
are obtainable e.g. the exit Mach number, My , is known. The exit total pressure
can be obtained (if needed). More importantly the pressure ratio exit is known. The
ratio of the ratio of stagnation pressure obtained by
f or Mexit
z }| ¶{ µ
µ ¶
P0 y P0 y Pexit 1 2
= = × = 0.5525
P0x Pexit P0x 0.905 4

Looking up in the Table (5.2) or utilizing the GDC-Potto provides

Ty ρy Py P0y
Mx My Tx ρx Px P0 x
2.3709 0.52628 2.0128 3.1755 6.3914 0.55250

With the information of Mach number (either Mx or My ) the area where the shock
(location) occurs can be found. First, utilizing the isentropic Table (5.2).
2 Of course, the computer can be use to carry this calculations in a sophisticate way.
7.1. NOZZLE EFFICIENCY 145

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
2.3709 0.47076 0.15205 2.3396 0.07158 0.16747

Approaching the shock location from the upstream (entrance) yields

A ∗
A= A = 2.3396 × 3 ∼
= 7.0188[cm2 ]
A∗
Note, as “simple” check this value is larger than the value in the previous exam-
ple.

7.1 Nozzle efficiency

Obviously nozzles are not perfectly efficient and there are several ways to define
the nozzleefficiency. One of the effective way is to define the efficiency as the
ratio of the energy converted to kinetic energy and the total potential energy could
be converted to kinetic energy. The total energy that can be converted is during
isentropic process is

E = h0 − hexits (7.1)

where hexits is the enthalpy if the flow was isentropic. The actual energy that was
used is

E = h0 − hexit (7.2)

The efficiency can be defined as

2
h0 − hexit (Uactual )
η= = 2 (7.3)
h0 − hexits (Uideal )
The typical efficiency of nozzle is ranged between 0.9 to 0.99. In the literature
some define also velocity coefficient as the ratio of the actual velocity to the ideal
velocity, Vc
s
2
√ (Uactual )
Vc = η = 2 (7.4)
(Uideal )
There is another less used definition which referred as the coefficient of discharge
as the ratio of the actual mass rate to the ideal mass flow rate.
ṁactual
Cd = (7.5)
ṁideal

7.2 Diffuser Efficiency

146 CHAPTER 7. NORMAL SHOCK IN VARIABLE DUCT AREAS

The efficiency of the diffuser is defined P01

as the ratio of the enthalpy change that h P02

occurred between the entrance to exit P2
01
stagnation pressure to the kinetic en- 02

ergy. 2 P1

2(h3 − h1 ) h3 − h1
η= 2 = (7.6)
U1 h0 1 − h1 1

For perfect gas equation (7.6) can be s,entropy

converted to
Fig. -7.3: Description to clarify the definition of
2Cp (T3 − T1 ) diffuser efficiency
η= (7.7)
U1 2
And further expanding equation (7.7) results in
³ ´ ³ ´ õ ¶ k−1 !
kR
2 k−1 T1 TT31 − 1 2 T3
k−1 T1 − 1 2 T3 k
η= = = −1 (7.8)
c1 2 M1 2 M1 2 M1 2 (k − 1) T1

Example 7.3:
A wind tunnel combined from
a nozzle and a diffuser (actu- nozzle Diffuser
ally two nozzles connected by a 1 2
constant area see Figure (7.4)) A n 3
Ad
4

the required condition at point

3 are: M = 3.0 and pres- capacitor
sure of 0.7[Bar] and tempera-
ture of 250K. The cross sec-
tion in area between the nuzzle Compressor
and diffuser is 0.02[m2 ]. What is
cooler
area of nozzle’s throat and what
is area of the diffuser’s throat to
maintain chocked diffuser with
subsonic flow in the expansion heat
section. k = 1.4 can be as- out

sumed. Assume that a shock

occurs in the test section. Fig. -7.4: Schematic of a supersonic tunnel in a contin-
uous region (and also for example (7.3)
S OLUTION
The condition at M = 3 is summarized in following table

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
3.0000 0.35714 0.07623 4.2346 0.02722 0.11528 0.65326
7.2. DIFFUSER EFFICIENCY 147

The nozzle area can be calculated by

A?
A∗ n = A = 0.02/4.2346 = 0.0047[m2 ]
A
In this case, P0 A∗ is constant (constant mass flow). First the stagnation behind the
shock will be

Ty ρy Py P0 y
Mx My Tx ρx Px P0x
3.0000 0.47519 2.6790 3.8571 10.3333 0.32834

P0 n ∗ 1
A∗ d = A n∼ 0.0047 ∼ 0.0143[m3 ]
P0 d 0.32834

Example 7.4:
A shock is moving at 200 [m/sec] in pipe with gas with k = 1.3, pressure of 2[Bar]
and temperature of 350K. Calculate the conditions after the shock.

S OLUTION
This is a case of completely and suddenly open valve with the shock velocity, tem-
perature and pressure “upstream” known. In this case Potto–GDC provides the
following table

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
5.5346 0.37554 0.0 1.989 5.479 34.50 0.021717

The calculations were carried as following: First calculate the Mx as

p
M x = Us / k ∗ 287. ∗ Tx

Then calculate the My by using Potto-GDC or utilize the Tables. For example
Potto-GDC (this code was produce by the program)

Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
5.5346 0.37554 5.4789 6.2963 34.4968 0.02172

The calculation of the temperature and pressure ratio also can be obtain by the
same manner. The “downstream” shock number is
Us
Msy = r ³ ´ ∼ 2.09668
Ty
k ∗ 287. ∗ Tx ∗ Tx
148 CHAPTER 7. NORMAL SHOCK IN VARIABLE DUCT AREAS

Finally utilizing the equation to calculate the following

0
My = Msy − My = 2.09668 − 0.41087 ∼ 1.989

Example 7.5:
An inventor interested in a design of tube and piston so that the pressure is doubled
in the cylinder when the piston is moving suddenly. The propagating piston is
assumed to move into media with temperature of 300K and atmospheric pressure
of 1[Bar]. If the steady state is achieved, what will be the piston velocity?

S OLUTION
This is an open valve case in which the pressure ratio is given. For this pressure
ratio of Py /Px = 2 the following table can be obtained or by using Potto–GDC
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
1.3628 0.75593 1.2308 1.6250 2.0000 0.96697

The temperature ratio and the Mach numbers for the velocity of the air (and the
piston) can be calculated. The temperature at “downstream” (close to the piston)
is
Ty
Ty = Tx = 300 × 1.2308 = 369.24[◦ C]
Tx
The velocity of the piston is then
√
Uy = My ∗ cy = 0.75593 ∗ 1.4 ∗ 287 ∗ 369.24 ∼ 291.16[m/sec]

Example 7.6:
A flow of gas is brought into a sudden stop. The mass flow rate of the gas is
2 [kg/sec] and cross section A = 0.002[m3 ]. The imaginary gas conditions are
temperature is 350K and pressure is 2[Bar] and R = 143[j/kg K] and k = 1.091
(Butane?). Calculate the conditions behind the shock wave.

S OLUTION
This is the case of a closed valve in which mass flow rate with the area given.
Thus, the “upstream” Mach is given.
0 ṁ ṁRT 2 × 287 × 350
Ux = = = ∼ 502.25[m/sec]
ρA PA 200000 × 0.002
0
Thus the static Mach number, Mx is
0
0 Ux 502.25
Mx = =√ ∼ 2.15
cx 1.091 × 143 × 350
7.2. DIFFUSER EFFICIENCY 149

With this value for the Mach number Potto-GDC provides

0 0 Ty Py P0 y
Mx My Mx My Tx Px P0 x
2.9222 0.47996 2.1500 0.0 2.589 9.796 0.35101

This table was obtained by using the procedure described in this book. The itera-
tion of the procedure are

Ty Py 0
i Mx My Tx Px My
0 3.1500 0.46689 2.8598 11.4096 0.0
1 2.940 0.47886 2.609 9.914 0.0
2 2.923 0.47988 2.590 9.804 0.0
3 2.922 0.47995 2.589 9.796 0.0
4 2.922 0.47996 2.589 9.796 0.0
5 2.922 0.47996 2.589 9.796 0.0
150 CHAPTER 7. NORMAL SHOCK IN VARIABLE DUCT AREAS
 CHAPTER 8

Nozzle Flow With External Forces

This chapter is under heavy construction. Please ignore. If you want to

contribute and add any results of experiments, to this chapter, please do so.
You can help especially if you have photos showing these effects.

In the previous chapters a simple model describing the flow in nozzle was
explained. In cases where more refined calculations have to carried the gravity or
other forces have to be taken into account. Flow in a vertical or horizontal nozzle
are different because the gravity. The simplified models that suggests them–self
are: friction and adiabatic, isothermal, seem the most applicable. These models
can served as limiting cases for more realistic flow.

The effects of the gravity of the nozzle flow in two models isentropic and
isothermal is analyzed here. The isothermal nozzle model is suitable in cases
where the flow is relatively slow (small Eckert numbers) while as the isentropic
model is more suitable for large Eckert numbers.

The two models produces slightly different equations. The equations re-
sults in slightly different conditions for the chocking and different chocking speed.
Moreover, the working equations are also different and this author isn’t aware of
material in the literature which provides any working table for the gravity effect.

151
152 CHAPTER 8. NOZZLE FLOW WITH EXTERNAL FORCES

8.1 Isentropic Nozzle (Q = 0)

The energy equation for isentropic nozzle provides

external
work or
potential
difference,
i.e. z × g
z }| {
dh + U dU = f (x)dx (8.1)

Utilizing equation (5.27) when ds = 0 leads to

dP
+ U dU = f (x0 )dx0 (8.2)
ρ

For the isentropic process dP = const × kρk−1 dρ when the const = P/ρk
at any point of the flow. The equation (8.2) becomes
dP
z }| {
any point RT
z}|{ z}|{
P ρk 1 P dρ
k dρ + U dU = k U dU =f (x0 )dx0 (8.3)
ρk ρ ρ ρ ρ
kRT dρ c2
+ U dU = dρ + U dU =f (x0 )dx0
ρ ρ

The continuity equation as developed earlier (mass conservation equation

isn’t effected by the gravity)
dρ dA dU
− = + =0 (8.4)
ρ A U
Substituting dρ/ρ from equation 8.3, into equation (8.2) moving dρ to the right hand
side, and diving by dx0 yields
· ¸
dU 1 dU 1 dA
U 0 = c2 + + f (x0 ) (8.5)
dx U dx0 A dx0
Rearranging equation (8.5) yields
· ¸
dU 2 dU c2 dA f (x0 )
= M + + (8.6)
dx0 dx0 AU dx0 U
And further rearranging yields
¡ ¢ dU c2 dA f (x0 )
1 − M2 = + (8.7)
dx0 AU dx0 U
8.1. ISENTROPIC NOZZLE (Q = 0) 153

Equation (8.7) can be rearranged as

· 2 ¸
dU 1 c dA f (x0 )
= + (8.8)
dx0 (1 − M 2 ) AU dx0 U

Equation (8.8) dimensionless form by utilizing x = x0 /` and ` is the nozzle length

 
dM 1  1 dA `f (x) 
=  +  (8.9)
dx (1 − M 2 )  AM dx cM 
c |{z}
U

And the final form of equation (8.9) is

¡ ¢ · ¸
d M2 2 1 dA `f (x)
= +
dx (1 − M 2 ) A dx c2
(8.10)

The term `fc(x)

2 is considered to be very small (0.1 × 10/100000 < 0.1%)
for “standard” situations. The dimensionless number, `fc(x) 2 sometimes referred as
Ozer number determines whether gravity should be considered in the calculations.
Nevertheless, one should be aware of value of Ozer number for large magnetic
fields (astronomy) and low temperature, In such cases, the gravity effect can be
considerable.
As it was shown before the transition must occur when M = 1. Conse-
quently, two zones must be treated separately. First, here the Mach number is
discussed and not the pressure as in the previous chapter. For M < 1 (the sub-
2
sonic branch) the term (1−M 2 ) is positive and the treads determined by gravity and

the area function. · ¸

1 dA `f (x)
+ > 0 =⇒ d(M 2 ) > 0
A dx c2
or conversely, · ¸
1 dA `f (x)
+ < 0 =⇒ d(M 2 ) < 0
A dx c2
2
For the case of M > 1 (the supersonic branch) the term (1−M 2 ) is negative and

therefore · ¸
1 dA `f (x)
+ > 0 =⇒ d(M 2 ) < 0
A dx c2
For the border case M = 1, the denominator 1 − M 2 = 0, is zero either d(M 2 ) = ∞
or · ¸
1 dA `f (x)
+ = 0.
A dx c2
154 CHAPTER 8. NOZZLE FLOW WITH EXTERNAL FORCES

And the dM is indeterminate. As it was shown in chapter (5) the flow is chocked
(M = 1) only when
· ¸
dA `f (x)
+ = 0. (8.11)
dx c2

It should be noticed that when f (x) is zero, e.g. horizontal flow, the equa-
tion (8.11) reduced into dA dx = 0 that was developed previously.
The ability to manipulate the location provides a mean to in-
crease/decrease the flow rate. Yet this ability since Ozer number is relatively very
small.
This condition means that the critical point can occurs in several locations
dA
that satisfies equation (8.11). Further, the critical point, sonic point is Ax 6= 0 If
f (x) is a positive function, the critical point happen at converging part of the nozzle
(before the throat) and if f (x) is a negative function the critical point is diverging
part of the throat. For example consider the gravity, f (x) = −g a flow in a nozzle
vertically the critical point will be above the throat.

8.2 Isothermal Nozzle (T = constant)

 CHAPTER 9
Isothermal Flow

In this chapter a model dealing with gas that flows through a long tube is described.
This model has a applicability to situations which occur in a relatively long distance
and where heat transfer is relatively rapid so that the temperature can be treated,
for engineering purposes, as a constant. For example, this model is applicable
when a natural gas flows over several hundreds of meters. Such situations are
common in large cities in U.S.A. where natural gas is used for heating. It is more
predominant (more applicable) in situations where the gas is pumped over a length
of kilometers.
The high speed of the gas is ob- w
tained or explained by the combination 
of heat transfer and the friction to the  +

direction P
flow

flow. For a long pipe, the pressure dif- P +
P

ference reduces the density of the gas.
For instance, in a perfect gas, the den-
T
U
g (M) +

+
g ( +
) UT UT M M
w
sity is inverse of the pressure (it has c.v.
to be kept in mind that the gas under-
goes an isothermal process.). To main- Fig. -9.1: Control volume for isothermal flow
tain conservation of mass, the velocity
increases inversely to the pressure. At
critical point the velocity reaches the speed of sound at the exit and hence the flow
will be choked1 .

1 This
√
explanation is not correct as it will be shown later on. Close to the critical point (about, 1/ k,
the heat transfer, is relatively high and the isothermal flow model is not valid anymore. Therefore, the
study of the isothermal flow above this point is only an academic discussion but also provides the upper
limit for Fanno Flow.

155
156 CHAPTER 9. ISOTHERMAL FLOW

9.1 The Control Volume Analysis/Governing equations

Figure (9.1) describes the flow of gas from the left to the right. The heat transfer up
stream (or down stream) is assumed to be negligible. Hence, the energy equation
can be written as the following:

dQ U2
= cp dT + d = cp dT0 (9.1)
ṁ 2

The momentum equation is written as the following

−AdP − τw dAwetted area = ṁdU (9.2)

where A is the cross section area (it doesn’t have to be a perfect circle; a close
enough shape is sufficient.). The shear stress is the force per area that acts on
the fluid by the tube wall. The Awetted area is the area that shear stress acts on.
The second law of thermodynamics reads
s2 − s1 T2 k − 1 P2
= ln − ln (9.3)
Cp T1 k P1

The mass conservation is reduced to

ṁ = constant = ρU A (9.4)

Again it is assumed that the gas is a perfect gas and therefore, equation of
state is expressed as the following:

P = ρRT (9.5)

9.2 Dimensionless Representation

In this section the equations are transformed into the dimensionless form and pre-
sented as such. First it must be recalled that the temperature is constant and
therefore, equation of state reads
dP dρ
= (9.6)
P ρ
It is convenient to define a hydraulic diameter

4 × Cross Section Area

DH = (9.7)
wetted perimeter
9.2. DIMENSIONLESS REPRESENTATION 157

Now, the Fanning friction factor2 is introduced, this factor is a dimensionless

friction factor sometimes referred to as the friction coefficient as
τw
f= 1 2 (9.8)
2 ρU

Substituting equation (9.8) into momentum equation (9.2) yields

ṁ
µ ¶ A
z}|{
4dx 1 2
−dP − f ρU = ρU dU (9.9)
DH 2

Rearranging equation (9.9) and using the identify for perfect gas M 2 = ρU 2 /kP
yields:
µ ¶
dP 4f dx kP M 2 kP M 2 dU
− − = (9.10)
P DH 2 U
Now the pressure, P as a function of the Mach number has to substitute along
with velocity, U .

U 2 = kRT M 2 (9.11)

Differentiation of equation (9.11) yields

¡ ¢
d(U 2 ) = kR M 2 dT + T d(M 2 ) (9.12)

d(M 2 ) d(U 2 ) dT
= − (9.13)
M2 U2 T

Now it can be noticed that dT = 0 for isothermal process and therefore

d(M 2 ) d(U 2 ) 2U dU 2dU
2
= = = (9.14)
M U2 U2 U

The dimensionalization of the mass conservation equation yields

dρ dU dρ 2U dU dρ d(U 2 )
+ = + = + =0 (9.15)
ρ U ρ 2U 2 ρ 2 U2
Differentiation of the isotropic (stagnation) relationship of the pressure (5.11)
yields
à !
1 2
dP0 dP 2 kM dM 2
= + k−1
(9.16)
P0 P 1 + 2 M2 M2
2 It should be noted that Fanning factor based on hydraulic radius, instead of diameter friction equa-

tion, thus “Fanning f” values are only 1/4th of “Darcy f” values.

158 CHAPTER 9. ISOTHERMAL FLOW

Differentiation of equation (5.9) yields:

µ ¶
k−1 2 k−1
dT0 = dT 1 + M +T dM 2 (9.17)
2 2

Notice that dT0 6= 0 in an isothermal flow. There is no change in the actual

temperature of the flow but the stagnation temperature increases or decreases
depending on the Mach number (supersonic flow of subsonic flow). Substituting T
for equation (9.17) yields:

T0 k−1
2 dM
2
M2
dT0 = ¡ ¢ (9.18)
1 + k−1
2 M
2 M2

Rearranging equation (9.18) yields

dT0 (k − 1) M 2 dM 2
= ¡ ¢ (9.19)
T0 2 1 + k−1
2
M2

By utilizing the momentum equation it is possible to obtain a relation be-

tween the pressure and density. Recalling that an isothermal flow (T = 0) and
combining it with perfect gas model yields

dP dρ
= (9.20)
P ρ

From the continuity equation (see equation (9.14)) leads

dM 2 2dU
2
= (9.21)
M U

The four equations momentum, continuity (mass), energy, state are de-
scribed above. There are 4 unknowns (M, T, P, ρ)3 and with these four equations
the solution is attainable. One can notice that there are two possible solutions (be-
cause of the square power). These different solutions are supersonic and subsonic
solution.
The distance friction, 4fDL , is selected as the choice for the independent
variable. Thus, the equations need to be obtained as a function of 4fDL . The
density is eliminated from equation (9.15) when combined with equation (9.20) to
become
dP dU
=− (9.22)
P U
3 Assuming the upstream variables are known.
9.2. DIMENSIONLESS REPRESENTATION 159

After substituting the velocity (9.22) into equation (9.10), one can obtain
µ ¶
dP 4f dx kP M 2 dP
− − = kP M 2 (9.23)
P DH 2 P

Equation (9.23) can be rearranged into

dP dρ dU 1 dM 2 kM 2 dx
= =− =− 2
=− 2
4f (9.24)
P ρ U 2 M 2 (1 − kM ) D

Similarly or by other path the stagnation pressure can be expressed as a function

of 4fDL
¡ ¢
dP0 kM 2 1 − k+1
2 M
2
dx
= ¡ k−1
¢ 4f (9.25)
P0 2
2 (kM − 1) 1 + 2 M 2 D

dT0 k (1 − k) M 2 dx
= ¡ k−1
¢ 4f (9.26)
T0 2
2 (1 − kM ) 1 + 2 M 2 D

The variables in equation (9.24) can be separated to obtain integrable form as

follows
Z L Z 1/k
4f dx 1 − kM 2
= dM 2 (9.27)
0 D M 2 kM 2

It can be noticed that at the entrance (x = 0) for which M = Mx=0 (the initial
velocity in the tube isn’t zero). The term 4fDL is positive for any x, thus, the term on
the other side has to be positive as well. To obtain this restriction 1 = kM 2 . Thus,
the value M = √1k is the limiting case from a mathematical point of view. When
Mach number larger than M > √1k it makes the right hand side of the integrate
negative. The physical meaning of this value is similar to M = 1 choked flow which
was discussed in a variable area flow in Chapter (5).
Further it can be noticed from equation (9.26) that when M → √1k the value
of right hand side approaches infinity (∞). Since the stagnation temperature (T0 )
has a finite value which means that dT0 → ∞. Heat transfer has a limited value
therefore the model of the flow must be changed. A more appropriate model is an
adiabatic flow model yet it can serve as a bounding boundary (or limit).
Integration of equation (9.27) requires information about the relationship
between the length, x, and friction factor f . The friction is a function of the
Reynolds number along the tube. Knowing the Reynolds number variations is
important. The Reynolds number is defined as

DU ρ
Re = (9.28)
µ
160 CHAPTER 9. ISOTHERMAL FLOW

The quantity U ρ is constant along the tube (mass conservation) under constant
area. Thus, the only viscosity is varied along the tube. However under the assump-
tion of ideal gas, viscosity is only a function of the temperature. The temperature
in isothermal process (the definition) is constant and thus the viscosity is constant.
In real gas the pressure effect is very minimal as described in “Basic of fluid me-
chanics” by this author. Thus, the friction factor can be integrated to yield
¯
4f L ¯¯ 1 − kM 2
= + ln kM 2 (9.29)
D ¯max kM 2

The definition for perfect gas yields M 2 = U 2 /kRT and noticing that √
T = constant is used to describe the relation of the properties at M = 1/ k.
By denoting the superscript symbol ∗ for the choking condition, one can obtain that
M2 1/k
2
= ∗2 (9.30)
U U
Rearranging equation (9.30) is transformed into
U √
= kM (9.31)
U∗
Utilizing the continuity equation provides
ρ 1
ρU = ρ∗ U ∗ ; =⇒ ∗ = √ (9.32)
ρ kM
Reusing the perfect–gas relationship
P ρ 1
= ∗ =√
P∗ ρ kM
(9.33)
Now utilizing the relation for stagnated isotropic pressure one can obtain
" # k
P0 P 1 + k−1 2 M 2 k−1
= ∗ (9.34)
P0∗ P 1 + k−1
2k
P
Substituting for P∗ equation (9.33) and rearranging yields
µ ¶ k−1
k µ ¶ k
P0 1 2k k − 1 2 k−1 1
∗ =√ 1+ M
P0 k 3k − 1 2 M
(9.35)
And the stagnation temperature at the critical point can be expressed as
µ ¶
T0 T 1 + k−1
2 M
2
2k k−1
= ∗ = 1+ M2
T0∗ T 1 + k−1
2k
3k − 1 2
(9.36)

These equations (9.31)-(9.36) are presented on in Figure (9.2)

9.3. THE ENTRANCE LIMITATION OF SUPERSONIC BRANCH 161

Isothermal Flow
* * *
P/P , ρ/ρ and T0/T0 as a function of M

1e+02
4fL

D
P or 
 ρ
* ∗
P ρ
1e+01 *
T0/T0
*
P0/P0

0.1

0.01
0.1 1 10
Fri Feb 18 17:23:43 2005 Mach number

Fig. -9.2: Description of the pressure, temperature relationships as a function of the Mach
number for isothermal flow

9.3 The Entrance Limitation of Supersonic Branch

Situations where the conditions at the tube exit have not arrived at the critical con-
ditions are discussed here. It is very useful to obtain the relationship between the
entrance and the exit condition for this case. Denote 1 and 2 as the conditions at
the inlet and exit respectably. From equation (9.24)
¯ ¯ µ ¶2
4f L 4f L ¯¯ 4f L ¯¯ 1 − kM1 2 1 − kM2 2 M1
= − = − + ln (9.37)
D D ¯max1 D ¯max2 kM1 2 kM2 2 M2

For the case that M1 >> M2 and M1 → 1 equation (9.37) is reduced into the
following approximation
∼0
z }| {
4f L 1 − kM2 2
= 2 ln M1 − 1 − (9.38)
D kM2 2
162 CHAPTER 9. ISOTHERMAL FLOW

Solving for M1 results in

e( )
1 4f L
2 D +1
M1 ∼ (9.39)

This relationship shows the maximum limit that Mach number can approach when
the heat transfer is extraordinarily fast. In reality, even small 4fDL > 2 results in
a Mach number which is larger than 4.5. This velocity requires a large entrance
length to achieve good heat transfer. With this conflicting mechanism obviously the
flow is closer to the Fanno flow model. Yet this model provides the directions of the
heat transfer effects on the flow.

9.4 Comparison with Incompressible Flow

The Mach number of the flow in some instances is relatively small. In these cases,
one should expect that the isothermal flow should have similar characteristics as
incompressible flow. For incompressible flow, the pressure loss is expressed as
follows
4f L U 2
P1 − P2 = (9.40)
D 2
Now note that for incompressible flow U1 = U2 = U and 4fDL represent the ratio of
the traditional h12 . To obtain a similar expression for isothermal flow, a relationship
between M2 and M1 and pressures has to be derived. From equation (9.40) one
can obtained that
P1
M2 = M1 (9.41)
P2
Substituting this expression into (9.41) yields
à µ ¶2 ! µ ¶2
4f L 1 P2 P2
= 1− − ln (9.42)
D kM1 2 P1 P1

Because f is always positive there is only one solution to the above equation even
though M2.
Expanding the solution for small pressure ratio drop, P1 − P2 /P1 , by some
mathematics.
denote
P1 − P2
χ= (9.43)
P1
Now equation (9.42) can be transformed into
à µ ¶2 ! à !2
4f L 1 P2 − P1 + P1 1
= 2 1− − ln P2 (9.44)
D kM1 P1 P1
9.4. COMPARISON WITH INCOMPRESSIBLE FLOW 163

µ ¶2
4f L 1 ³ 2
´ 1
= 1 − (1 − χ) − ln (9.45)
D kM1 2 1−χ

µ ¶2
4f L 1 ¡ ¢ 1
= 2 2χ − χ2 − ln (9.46)
D kM1 1−χ
now we have to expand into a series around χ = 0 and remember that
x2 ¡ ¢
f (x) = f (0) + f 0 (0)x + f 00 (0) + 0 x3 (9.47)
2
and for example the first derivative of
µ ¶2 ¯¯
d 1 ¯
ln ¯ =
dχ 1−χ ¯
χ=0
£ ¤ ¯
2 −3 ¯
(1 − χ) × (−2)(1 − χ) (−1)¯ = 2 (9.48)
χ=0

similarly it can be shown that f 00 (χ = 0) = 1 equation (9.46) now can be

approximated as
4f L 1 2
¡ 2
¢ ¡ ¢
= 2 (2χ − χ ) − 2χ − χ + f χ3 (9.49)
D kM1
rearranging equation (9.49) yields
4f L χ £ ¤ ¡ ¢
= 2 (2 − χ) − kM1 2 (2 − χ) + f χ3 (9.50)
D kM1
and further rearrangement yields
4f L χ £ ¡ ¢ ¤ ¡ ¢
= 2 2(1 − kM1 2 ) − 1 + kM1 2 χ + f χ3 (9.51)
D kM1

in cases that χ is small

4f L χ £ ¡ ¢ ¤
≈ 2(1 − kM1 2 ) − 1 + kM1 2 χ (9.52)
D kM1 2

The pressure difference can be plotted as a function of the M1 for given

4f L
value of D . Equation (9.52) can be solved explicitly to produce a solution for
s
1 − kM1 2 1 − kM1 2 kM1 2 4f L
χ= − − (9.53)
1 + kM1 2 1 + kM1 2 1 + kM1 2 D
A few observations can be made about equation (9.53).
164 CHAPTER 9. ISOTHERMAL FLOW

9.5 Supersonic Branch

Apparently, this analysis/model is over simplified for the supersonic branch and
does not produce reasonable results since it neglects to take into account the heat
transfer effects. A dimensionless analysis4 demonstrates that all the common ma-
terials that the author is familiar which creates a large error in the fundamental
assumption of the model and the model breaks. Nevertheless, this model can pro-
vide a better understanding to the trends and deviations of the Fanno flow model.
In the supersonic flow, the hydraulic entry length is very large as will be
shown below. However, the feeding diverging nozzle somewhat reduces the re-
quired entry length (as opposed to converging feeding). The thermal entry length
is in the order of the hydrodynamic entry length (look at the Prandtl number, (0.7-
1.0), value for the common gases.). Most of the heat transfer is hampered in the
sublayer thus the core assumption of isothermal flow (not enough heat transfer so
the temperature isn’t constant) breaks down5 .
The flow speed at the entrance is very large, over hundred of meters per
second. For example, a gas flows in a tube with 4fDL = 10 the required entry Mach
number is over 200. Almost all the perfect gas model substances dealt with in
this book, the speed of sound is a function of temperature. For this illustration,
for most gas cases the speed of sound is about 300[m/sec]. For example, even
with low temperature like 200K the speed of sound of air is 283[m/sec]. So, even
for relatively small tubes with 4fDL = 10 the inlet speed is over 56 [km/sec]. This
requires that the entrance length to be larger than the actual length of the tub for
air. Remember from Fluid Dynamic book
UD
Lentrance = 0.06 (9.54)
ν
The typical values of the the kinetic viscosity, ν, are 0.0000185 kg/m-sec at 300K
and 0.0000130034 kg/m-sec at 200K. Combine this information with our case of
4f L
D = 10
Lentrance
= 250746268.7
D
On the other hand a typical value of friction coefficient f = 0.005 results in
Lmax 10
= = 500
D 4 × 0.005
The fact that the actual tube length is only less than 1% of the entry length means
that the assumption is that the isothermal flow also breaks (as in a large response
time).
Now, if Mach number is changing from 10 to 1 the kinetic energy change is
about TT00∗ = 18.37 which means that the maximum amount of energy is insufficient.
4 This dimensional analysis is a bit tricky, and is based on estimates. Currently and ashamedly the

author is looking for a more simplified explanation. The current explanation is correct but based on
hands waving and definitely does not satisfy the author.
5 see Kays and Crawford “Convective Heat Transfer” (equation 12-12).
9.7. ISOTHERMAL FLOW EXAMPLES 165

Now with limitation, this topic will be covered in the next version because it
provide some insight and boundary to the Fanno Flow model.

9.6 Figures and Tables

Table -9.1: The Isothermal Flow basic parameters

4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.03000 785.97 28.1718 17.6651 28.1718 0.87516
0.04000 439.33 21.1289 13.2553 21.1289 0.87528
0.05000 279.06 16.9031 10.6109 16.9031 0.87544
0.06000 192.12 14.0859 8.8493 14.0859 0.87563
0.07000 139.79 12.0736 7.5920 12.0736 0.87586
0.08000 105.89 10.5644 6.6500 10.5644 0.87612
0.09000 82.7040 9.3906 5.9181 9.3906 0.87642
0.10000 66.1599 8.4515 5.3334 8.4515 0.87675
0.20000 13.9747 4.2258 2.7230 4.2258 0.88200
0.25000 7.9925 3.3806 2.2126 3.3806 0.88594
0.30000 4.8650 2.8172 1.8791 2.8172 0.89075
0.35000 3.0677 2.4147 1.6470 2.4147 0.89644
0.40000 1.9682 2.1129 1.4784 2.1129 0.90300
0.45000 1.2668 1.8781 1.3524 1.8781 0.91044
0.50000 0.80732 1.6903 1.2565 1.6903 0.91875
0.55000 0.50207 1.5366 1.1827 1.5366 0.92794
0.60000 0.29895 1.4086 1.1259 1.4086 0.93800
0.65000 0.16552 1.3002 1.0823 1.3002 0.94894
0.70000 0.08085 1.2074 1.0495 1.2074 0.96075
0.75000 0.03095 1.1269 1.0255 1.1269 0.97344
0.80000 0.00626 1.056 1.009 1.056 0.98700
0.81000 0.00371 1.043 1.007 1.043 0.98982
0.81879 0.00205 1.032 1.005 1.032 0.99232
0.82758 0.000896 1.021 1.003 1.021 0.99485
0.83637 0.000220 1.011 1.001 1.011 0.99741
0.84515 0.0 1.000 1.000 1.000 1.000

9.7 Isothermal Flow Examples

There can be several kinds of questions aside from the proof questions6 Generally,
the “engineering” or practical questions can be divided into driving force (pressure
6 The proof questions are questions that ask for proof or for finding a mathematical identity (normally

good for mathematicians and study of perturbation methods). These questions or examples will appear
in the later versions.
166 CHAPTER 9. ISOTHERMAL FLOW

difference), resistance (diameter, friction factor, friction coefficient, etc.), and mass
flow rate questions. In this model no questions about shock (should) exist7 .
The driving force questions deal with what should be the pressure differ-
ence to obtain certain flow rate. Here is an example.

Example 9.1:
A tube of 0.25 [m] diameter and 5000 [m] in length is attached to a pump. What
should be the pump pressure so that a flow rate of 2 [kg/sec] will be achieved?
Assume that friction factor f = 0.005 and the exit pressure is 1[bar]. hThe specific
i
J
heat for the gas, k = 1.31, surroundings temperature 27◦ C, R = 290 Kkg . Hint:
calculate the maximum flow rate and then check if this request is reasonable.

S OLUTION
If the flow was incompressible then for known density, ρ, the velocity can be calcu-
2
lated by utilizing ∆P = 4fDL U2g . In incompressible flow, the density is a function of
√
the entrance Mach number. The exit Mach number is not necessarily 1/ k i.e. the
flow is not choked. First, check whether flow is choked (or even possible).
Calculating the resistance, 4fDL
4f L 4 × 0.0055000
= = 400
D 0.25
Utilizing Table (9.1) or the program provides

4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.04331 400.00 20.1743 12.5921 0.0 0.89446

The maximum flow rate (the limiting case) can be calculated by utilizing the
above table. The velocity£ of ¤the gas at the entrance U = cM = 0.04331 ×
√
1.31 × 290 × 300 ∼ m
= 14.62 sec . The density reads
· ¸
P 2, 017, 450 ∼ kg
ρ= = = 23.19
RT 290 × 300 m3
The maximum flow rate then reads
· ¸
π × (0.25)2 kg
ṁ = ρAU = 23.19 × × 14.62 ∼
= 16.9
4 sec
The maximum flow rate is larger then the requested mass rate hence the flow is
not choked. It is note worthy to mention that since the isothermal model breaks
around the choking point, the flow rate is really some what different. It is more
appropriate to assume an isothermal model hence our model is appropriate.
7 Those who are mathematically inclined can include these kinds of questions but there are no real

world applications to isothermal model with shock.

9.7. ISOTHERMAL FLOW EXAMPLES 167

To solve this problem the flow rate has to be calculated as

· ¸
kg
ṁ = ρAU = 2.0
sec

P1 kU P1 kU P1
ṁ = A =√ A√ = AkM1
RT k kRT kRT c
Now combining with equation (9.41) yields

M2 P2 Ak
ṁ =
c

ṁc 2 × 337.59
M2 = = 2 = 0.103
P2 Ak 100000 × π×(0.25)
4 × 1.31
From Table (9.1) or by utilizing the program

4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.10300 66.6779 8.4826 5.3249 0.0 0.89567

The entrance Mach number is obtained by

¯
4f L ¯¯
= 66.6779 + 400 ∼
= 466.68
D ¯1

4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.04014 466.68 21.7678 13.5844 0.0 0.89442

The pressure should be

P = 21.76780 × 8.4826 = 2.566[bar]

Note that tables in this example are for k = 1.31

Example 9.2:
A flow of gas was considered for a distance of 0.5 [km] (500 [m]). A flow rate of
0.2 [kg/sec] is required. Due to safety concerns, the maximum pressure allowed
for the gas is only 10[bar]. Assume that the flow is isothermal and k=1.4, calculate
the required diameter of tube. The friction coefficient for the tube can be assumed
as 0.02 (A relative smooth tube of cast iron.). Note that tubes are provided in
increments of 0.5 [in]8 . You can assume that the soundings temperature to be
27◦ C.
8 It is unfortunate, but it seems that this standard will be around in USA for some time.
168 CHAPTER 9. ISOTHERMAL FLOW

S OLUTION
At first, the minimum diameter will be obtained when the flow is choked. Thus,
the maximum M1 that can be obtained when the M2 is at its maximum and back
pressure is at the atmospheric pressure.
Mmax
z}|{
P2 1 1
M1 = M2 = √ = 0.0845
P1 k 10
Now, with the value of M1 either by utilizing Table (9.1) or using the provided pro-
gram yields

4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.08450 94.4310 10.0018 6.2991 0.0 0.87625
¯
4f L ¯
With D ¯ = 94.431 the value of minimum diameter.
max

4f L 4 × 0.02 × 500
D= ¯ ' ' 0.42359[m] = 16.68[in]
4f L ¯ 94.43
D ¯
max

However, the pipes are provided only in 0.5 increments and the next size is 17[in]
or 0.4318[m]. With this pipe size the calculations are to be repeated in reverse and
produces: (Clearly the maximum mass is determined with)
√
P √ P AM k
ṁ = ρAU = ρAM c = AM kRT = √
RT RT
The usage of the above equation clearly applied to the whole pipe. The only point
that must be emphasized is that all properties (like Mach number, pressure and
etc) have to be taken at the same point. The new 4fDL is

4f L 4 × 0.02 × 500
= ' 92.64
D 0.4318

4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.08527 92.6400 9.9110 6.2424 0.0 0.87627

To check whether the flow rate satisfies the requirement

2 √
106 × π×0.4318
4 × 0.0853 × 1.4
ṁ = √ ≈ 50.3[kg/sec]
287 × 300

Since 50.3 ≥ 0.2 the mass flow rate requirement is satisfied.

9.7. ISOTHERMAL FLOW EXAMPLES 169

It should be noted that P should be replaced by P0 in the calculations. The speed

of sound at the entrance is
√ √ hmi
c = kRT = 1.4 × 287 × 300 ∼ = 347.2
sec
and the density is · ¸
P 1, 000, 000 kg
ρ= = = 11.61
RT 287 × 300 m3
The velocity at the entrance should be
hmi
U = M ∗ c = 0.08528 × 347.2 ∼
= 29.6
sec
The diameter should be
s r
4ṁ 4 × 0.2 ∼
D= = = 0.027
πU ρ π × 29.6 × 11.61

Nevertheless, for the sake of the exercise the other parameters will be calculated.
This situation is reversed question. The flow rate is given with the diameter of the
pipe. It should be noted that the flow isn’t choked.

Example 9.3:
A gas flows of from a station (a) with pressure of 20[bar] through a pipe with 0.4[m]
diameter and 4000 [m] length to a different station (b). The pressure at the exit
(station (b)) is 2[bar]. The gas and the sounding temperature can be assumed
to be 300 K. Assume that the flow is isothermal, k=1.4, and the average friction
f=0.01. Calculate the Mach number at the entrance to pipe and the flow rate.

S OLUTION
First, the information whether the flow is choked needs to be found. Therefore, at
first it will be assumed that the whole length is the maximum length.

¯
4f L ¯¯ 4 × 0.01 × 4000
= = 400
D ¯max 0.4
¯
4f L ¯
with D ¯ = 400 the following can be written
max

4fL T0 ρ P P0
M D T0 ∗T ρ∗T P∗T P0 ∗T
0.0419 400.72021 0.87531 20.19235 20.19235 12.66915

P0
From the table M1 ≈ 0.0419 ,and P0 ∗T
≈ 12.67

28
P0 ∗T ∼
= ' 2.21[bar]
12.67
170 CHAPTER 9. ISOTHERMAL FLOW

The pressure at point (b) by utilizing the isentropic relationship (M = 1) pressure

ratio is 0.52828.

P0 ∗T
P2 = ³ ´ = 2.21 × 0.52828 = 1.17[bar]
P2
P0 ∗T

As the pressure at point (b) is smaller than the actual pressure P ∗ < P2 than the
actual pressure one must conclude that the flow is not choked. The solution is an
iterative process.

1. guess reasonable value of M1 and calculate 4fDL

¯ ¯
¯ ¯
2. Calculate the value of 4fDL ¯ by subtracting 4fDL ¯ − 4f L
D
2 1

3. Obtain M2 from the Table ? or by using the Potto–GDC.

Calculate the pressure, P2 bear in mind that this isn’t the real pressure but
4.
based on the assumption.
Compare the results of guessed pressure P2 with the actual pressure and
5.
choose new Mach number M1 accordingly.

Now the process has been done for you and is provided in figure ??? or in the
table obtained from the provided program.
¯
4fL ¯
¯
M1 M2 ¯ 4fL P2
D max 1 D P1
0.0419 0.59338 400.32131 400.00000 0.10000

The flow rate is

√ √
P k π × D2 2000000 1.4
ṁ = ρAM c = √ M= √ π × 0.22 × 0.0419
RT 4 300 × 287
' 42.46[kg/sec]

In this chapter, there are no examples on isothermal with supersonic flow.

9.8 Unchoked situations in Fanno Flow

Table -9.2: The flow parameters for unchoked flow

¯
4fL ¯
¯
M1 M2 ¯ 4fL P2
D max 1 D P1
0.7272 0.84095 0.05005 0.05000 0.10000
0.6934 0.83997 0.08978 0.08971 0.10000
9.8. UNCHOKED SITUATIONS IN FANNO FLOW 171
Table -9.2: The flow parameters for unchoked flow (continue)

4fL ¯
¯ ¯
M1 M2 ¯ 4fL P2
D max 1 D P2
0.6684 0.84018 0.12949 0.12942 0.10000
0.6483 0.83920 0.16922 0.16912 0.10000
0.5914 0.83889 0.32807 0.32795 0.10000
0.5807 0.83827 0.36780 0.36766 0.10000
0.5708 0.83740 0.40754 0.40737 0.10000

One of the interesting feature of the isothermal flow is that Reynolds number re-
mains constant during the flow for an ideal gas material (enthalpy is a function
of only the temperature). This fact simplifies the calculation of the friction factor.
This topic has more discussion on the web than on “scientific” literature. Here is a
theoretical example for such calculation that was discussed on the web.

Example 9.4:
Air flows in a tube with 0.1[m] diameter and 100[m] in length. The relative rough-
ness, ²/D = 0.001 and the entrance pressure is P1 = 20[Bar] and the exit pressure
is P1 = 1[Bar] . The surroundings temperature is 27◦ C. Estimate whether the flow
is laminar or turbulent, estimate the friction factor, the entrance and exit Mach num-
bers and the flow rate.

S OLUTION
The first complication is the know what is flow regimes. The process is to assume
that the flow is turbulent (long pipe). In this case, for large Reynolds number the
friction factor is about 0.005. Now the iterative procedure as following;
Calculate the 4fDL .
4f L 4 × 0.005 × 100
D = 0.1
= 20

For this value and the given pressure ratio the flow is choked. Thus,

4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.17185 20.0000 4.9179 3.1460 4.9179 0.88017

For this iteration the viscosity of the air is taken from the Basics of Fluid Mechanics
by this author and the Reynolds number can be calculated as
√ 200000
DU ρ 0.1 × 0.17185 × 1.4 × 287 × 300 ×
Re = = 287 × 300 ∼ 17159.15
µ 0.0008

For this Reynolds number the fiction factor can be estimated by using the full Cole-
172 CHAPTER 9. ISOTHERMAL FLOW

M1 isothermal flow

0.9

0.8 P2 / P1 = 0.8
0.7 P2 / P1 = 0.5
P2 / P1 = 0.2
M1 0.6 P2 / P1 = 0.10
0.5

0.4

0.3

0.2

0.1

0
0 10 20 30 40 50 60 70 80 90 100
4fL

Fri Feb 25 17:20:14 2005 D

Fig. -9.3: The Mach number at the entrance to a tube under isothermal flow model as a
function 4fDL

brook’s equation
µ ¶
1 ε/Dh 2.51
√ = −2 log10 + √ (9.55)
f 3.7 Re f
or the approximated Haaland’s equation
"µ ¶1.11 #
1 ε/D 6.9
√ = −1.8 log10 + (9.56)
f 3.7 Re

which provide f = 0.0053 and it is a reasonable answer in one iteration. Repeating

the iteration results in
4f L 4 × 0.0053 × 100
D = 0.1
= 21.2

with
4fL P P0 ρ T0
M D P∗ P0 ∗ ρ∗ T0 ∗
0.16689 21.4000 5.0640 3.2357 5.0640 0.87987
9.8. UNCHOKED SITUATIONS IN FANNO FLOW 173

And the “improved” Reynolds number is

√ 200000
0.1 × 0.16689 × 1.4 × 287 × 300 ×
Re = 287 × 300 ∼ 16669.6
0.0008
And the friction number is .0054 which very good estimate compare with the as-
sumption that this model was built on.
End Solution
174 CHAPTER 9. ISOTHERMAL FLOW
 CHAPTER 10

Fanno Flow

w 
An adiabatic flow with friction is

named after Ginno Fanno a Jewish  +

direction P
flow
engineer. This model is the second P +
P
+

UT UT M M
pipe flow model described here.
The main restriction for this model
T
U
g (M) +
g ( +
)
is that heat transfer is negligible and w
can be ignored 1 . This model is ap- c.v.
plicable to flow processes which are No heat transer

very fast compared to heat trans-

fer mechanisms with small Eckert Fig. -10.1: Control volume of the gas flow in a con-
stant cross section
number.
This model explains many industrial flow processes which includes emp-
tying of pressured container through a relatively short tube, exhaust system of an
internal combustion engine, compressed air systems, etc. As this model raised
from need to explain the steam flow in turbines.

10.1 Introduction
Consider a gas flowing through a conduit with a friction (see Figure (10.1)). It
is advantages to examine the simplest situation and yet without losing the core
properties of the process. Later, more general cases will be examined2 .

1 Even the friction does not convert into heat

2 Not ready yet, discussed on the ideal gas model and the entry length issues.

175
176 CHAPTER 10. FANNO FLOW

10.2 Fanno Model

The mass (continuity equation) balance can be written as

ṁ = ρAU = constant (10.1)

,→ ρ1 U1 = ρ2 U2

The energy conservation (under the assumption that this model is adia-
batic flow and the friction is not transformed into thermal energy) reads

T01 = T02 (10.2)

2 2
U1 U2
,→ T1 + = T2 +
2cp 2cp
(10.3)

Or in a derivative form
µ ¶
U2
Cp dT + d =0 (10.4)
2

Again for simplicity, the perfect gas model is assumed3 .

P = ρRT (10.5)
P1 P2
,→ =
ρ1 T1 ρ2 T2

It is assumed that the flow can be approximated as one–dimensional. The

force acting on the gas is the friction at the wall and the momentum conservation
reads

−AdP − τw dAw = ṁdU (10.6)

It is convenient to define a hydraulic diameter as

4 × Cross Section Area

DH = (10.7)
wetted perimeter

Or in other words

πDH 2
A= (10.8)
4
3 The equation of state is written again here so that all the relevant equations can be found when this

chapter is printed separately.

10.3. NON–DIMENSIONALIZATION OF THE EQUATIONS 177

It is convenient to substitute D for DH and yet it still will be referred to the same
name as the hydraulic diameter. The infinitesimal area that shear stress is acting
on is

dAw = πDdx (10.9)

Introducing the Fanning friction factor as a dimensionless friction factor which is

some times referred to as the friction coefficient and reads as the following:
τw
f= 1 2
(10.10)
2 ρU

By utilizing equation (10.2) and substituting equation (10.10) into momentum

equation (10.6) yields

A w τ
z }| { z µ }| ¶{ ṁ
A
πD 2
1 2 z}|{
− dP − πDdx f ρU = A ρU dU (10.11)
4 2

Dividing equation (10.11) by the cross section area, A and rearranging

yields
µ ¶
4f dx 1 2
−dP + ρU = ρU dU (10.12)
D 2

The second law is the last equation to be utilized to determine the flow direction.

s2 ≥ s1 (10.13)

10.3 Non–Dimensionalization of the Equations

Before solving the above equation a dimensionless process is applied. By utilizing
the definition of the sound speed to produce the following identities for perfect gas
µ ¶2
U U2
M2 = = (10.14)
c k |{z}
RT
P
ρ

Utilizing the definition of the perfect gas results in

ρU 2
M2 = (10.15)
kP
178 CHAPTER 10. FANNO FLOW

Using the identity in equation (10.14) and substituting it into equation (10.11) and
after some rearrangement yields
ρU 2
µ ¶ 2 z }| { dU
4f dx 1 ρU
−dP + kP M 2 = dU = kP M 2 (10.16)
DH 2 U U
By further rearranging equation (10.16) results in
µ ¶
dP 4f dx kM 2 dU
− − = kM 2 (10.17)
P D 2 U
It is convenient to relate expressions of (dP/P ) and dU/U in terms of the Mach
number and substituting it into equation (10.17). Derivative of mass conservation
((10.2)) results in
dU
U
z }| {
dρ 1 dU 2
+ =0 (10.18)
ρ 2 U2
The derivation of the equation of state (10.5) and dividing the results by equation
of state (10.5) results
dP dρ dT
= + (10.19)
P ρ dT
Derivation of the Mach identity equation (10.14) and dividing by equation (10.14)
yields

d(M 2 ) d(U 2 ) dT
2
= − (10.20)
M U2 T
Dividing the energy equation (10.4) by Cp and by utilizing the definition Mach
number yields
µ 2¶
dT 1 1 U2 U
+µ ¶ 2
d =
T kR TU 2
(k − 1)
| {z }
Cp
µ ¶
dT (k − 1) U 2 U2
,→ + 2
d =
T kRT
| {z } U 2
c2
dT k − 1 2 dU 2
,→ + M =0 (10.21)
T 2 U2
Equations (10.17), (10.18), (10.19), (10.20), and (10.21) need to be solved. These
equations are separable so one variable is a function of only single variable (the
10.3. NON–DIMENSIONALIZATION OF THE EQUATIONS 179

chosen as the independent variable). Explicit explanation is provided for only two
variables, the rest variables can be done in a similar fashion. The dimensionless
friction, 4fDL , is chosen as the independent variable since the change in the dimen-
sionless resistance, 4fDL , causes the change in the other variables.
Combining equations (10.19) and (10.21) when eliminating dT /T results
dP dρ (k − 1)M 2 dU 2
= − (10.22)
P ρ 2 U2

The term dρρ can be eliminated by utilizing equation (10.18) and substituting it into
equation (10.22) and rearrangement yields
dP 1 + (k − 1)M 2 dU 2
=− (10.23)
P 2 U2
The term dU 2 /U 2 can be eliminated by using (10.23)
¡ ¢
dP kM 2 1 + (k − 1)M 2 4f dx
=− (10.24)
P 2(1 − M 2 ) D
The second equation for Mach number, M variable is obtained by combining equa-
tion (10.20) and (10.21) by eliminating dT /T . Then dρ/ρ and U are eliminated by
utilizing equation (10.18) and equation (10.22). The only variable that is left is P
(or dP/P ) which can be eliminated by utilizing equation (10.24) and results in
¡ ¢
4f dx 1 − M 2 dM 2
= (10.25)
D kM 4 (1 + k−1 2
2 M )

Rearranging equation (10.25) results in

¡ ¢
dM 2 kM 2 1 + k−1
2 M
2
4f dx
= (10.26)
M2 1 − M2 D

After similar mathematical manipulation one can get the relationship for the
velocity to read
dU kM 2 4f dx
= (10.27)
U 2 (1 − M 2 ) D
and the relationship for the temperature is
dT 1 dc k(k − 1)M 4 4f dx
= =− (10.28)
T 2 c 2(1 − M 2 ) D
density is obtained by utilizing equations (10.27) and (10.18) to obtain
dρ kM 2 4f dx
=− (10.29)
ρ 2 (1 − M 2 ) D
180 CHAPTER 10. FANNO FLOW

The stagnation pressure is similarly obtained as

dP0 kM 2 4f dx
=− (10.30)
P0 2 D

The second law reads

dT dP
ds = Cp ln − R ln (10.31)
T P

The stagnation temperature expresses as T0 = T (1 + (1 − k)/2M 2 ). Taking

derivative of this expression when M remains constant yields dT0 = dT (1 + (1 −
k)/2M 2 ) and thus when these equations are divided they yield

dT /T = dT0 /T0 (10.32)

In similar fashion the relationship between the stagnation pressure and the pres-
sure can be substituted into the entropy equation and result in

dT0 dP0
ds = Cp ln − R ln (10.33)
T0 P0

The first law requires that the stagnation temperature remains constant, (dT0 = 0).
Therefore the entropy change is

ds (k − 1) dP0
=− (10.34)
Cp k P0

Using the equation for stagnation pressure the entropy equation yields

ds (k − 1)M 2 4f dx
= (10.35)
Cp 2 D

10.4 The Mechanics and Why the Flow is Choked?

The trends of the properties can be examined by looking in equations (10.24)
through (10.34). For example, from equation (10.24) it can be observed that the
critical point is when M = 1. When M < 1 the pressure decreases downstream
as can be seen from equation (10.24) because f dx and M are positive. For the
same reasons, in the supersonic branch, M > 1, the pressure increases down-
stream. This pressure increase is what makes compressible flow so different from
“conventional” flow. Thus the discussion will be divided into two cases: One, flow
above speed of sound. Two, flow with speed below the speed of sound.
10.5. THE WORKING EQUATIONS 181

Why the flow is choked?

Here, the explanation is based on the equations developed earlier and there is no
known explanation that is based on the physics. First, it has to be recognized that
the critical point is when M = 1. It will be shown that a change in location relative to
this point change the trend and it is singular point by itself. For example, dP (@M =
1) = ∞ and mathematically it is a singular point (see equation (10.24)). Observing
from equation (10.24) that increase or decrease from subsonic just below one M =
(1 − ²) to above just above one M = (1 + ²) requires a change in a sign pressure
direction. However, the pressure has to be a monotonic function which means that
flow cannot crosses over the point of M = 1. This constrain means that because
the flow cannot “crossover” M = 1 the gas has to reach to this speed, M = 1 at
the last point. This situation is called choked flow.

The Trends

The trends or whether the variables are increasing or decreasing can be observed
from looking at the equation developed. For example, the pressure can be ex-
amined by looking at equation (10.26). It demonstrates that the Mach number
increases downstream when the flow is subsonic. On the other hand, when the
flow is supersonic, the pressure decreases.
The summary of the properties changes on the sides of the branch

Subsonic Supersonic
Pressure, P decrease increase
Mach number, M increase decrease
Velocity, U increase decrease
Temperature, T decrease increase
Density, ρ decrease increase
Stagnation Temperature, T0 decrease increase

10.5 The Working Equations

Integration of equation (10.25) yields

Z Lmax k+1 2
4 1 1 − M2 k+1 2 M
f dx = 2
+ ln k−1
(10.36)
D L k M 2k 1 + 2 M2

A representative friction factor is defined as

Z Lmax
1
f¯ = f dx (10.37)
Lmax 0
182 CHAPTER 10. FANNO FLOW

In the isothermal flow model it was shown that friction factor is constant through
the process if the fluid is ideal gas. Here, the Reynolds number defined in equation
(9.28) is not constant because the temperature is not constant. The viscosity even
for ideal gas is complex function of the temperature (further reading in “Basic of
Fluid Mechanics” chapter one, Potto Project). However, the temperature variation
is very limit. Simple improvement can be done by assuming constant constant
viscosity (constant friction factor) and find the temperature on the two sides of the
tube to improve the friction factor for the next iteration. The maximum error can be
estimated by looking at the maximum change of the temperature. The temperature
can be reduced by less than %20 for most range of the spesific heats ratio. The
viscosity change for this change is for many gases about 10%. For these gases
the maximum increase of average Reynolds number is only 5%. What this change
in Reynolds number does to friction factor? That depend in the range of Reynolds
number. For Reynolds number larger than 10,000 the change in friction factor can
be considered negligible. For the other extreme, laminar flow it can estimated that
change of 5% in Reynolds number change about the same amount in friction factor.
With the exception the jump from a laminar flow to a turbulent flow, the change is
noticeable but very small. In the light of the about discussion the friction factor
is assumed to constant. By utilizing the mean average theorem equation (10.36)
yields

4f¯Lmax 1 1 − M2 k+1 k+1

2 M
2
= + ln (10.38)
D k M2 2k 1 + k−1
2 M
2

It is common to replace the f¯ with f which is adopted in this book.

Equations (10.24), (10.27), (10.28), (10.29), (10.29), and (10.30) can be
solved. For example, the pressure as written in equation (10.23) is represented
by 4fDL , and Mach number. Now equation (10.24) can eliminate term 4fDL and
describe the pressure on the Mach number. Dividing equation (10.24) in equation
(10.26) yields

dP
1 + (k − 1M 2
P
=− ¡ ¢ dM 2 (10.39)
dM 2
M2
2M 2 1 + k−1
2 M 2

The symbol “*” denotes the state when the flow is choked and Mach number is
equal to 1. Thus, M = 1 when P = P ∗ equation (10.39) can be integrated to
yield:

s
k+1
P 1 2
∗
= k−1
P M 1+ 2 M
2
(10.40)
10.5. THE WORKING EQUATIONS 183

In the same fashion the variables ratio can be obtained

k+1
T c2 2
= =
T∗ c∗ 2 1 + k−1
2 M
2
(10.41)

s
k−1 2
ρ 1 1+ 2 M
∗
= k+1
ρ M 2
(10.42)

µ ¶−1 s
k+1
U ρ 2
= =M k−1
U∗ ρ∗ 1+ 2 M
2
(10.43)

The stagnation pressure decreases and can be expressed by

k
(1+ 1−k
2 M )
2 k−1

z}|{
P0
P0 P
= P (10.44)
P0 ∗ P0 ∗
P∗
P∗
|{z}
k
( k+1
2
) k−1

Using the pressure ratio in equation (10.40) and substituting it into equation
(10.44) yields
à ! k−1
k s
k−1 2 k−1 2
P0 1+ 2 M 1 1+ 2 M
= (10.45)
P0 ∗ k+1
2
M k+1
2

And further rearranging equation (10.45) provides

à ! 2(k−1)
k+1
k−1 2
P0 1 1+ 2 M
=
P0 ∗ M k+1
2
(10.46)

The integration of equation (10.34) yields

v
uà ! k+1
s−s ∗ u k + 1
k
2 t
= ln M ¡ ¢ (10.47)
cp 2M 2 1 + k−1
2 M
2
184 CHAPTER 10. FANNO FLOW

* *
Fanno* Flow
P/P , ρ/ρ and T/T as a function of M

1e+02
4fL

D
P
*
P
1e+01 *
T/T
*
P0/P0
U/U*
1

0.1

0.01
0.1 1 10
Tue Sep 25 10:57:55 2007 Mach number

Fig. -10.2: Various parameters in Fanno flow as a function of Mach number

The results of these equations are plotted in Figure (10.2) The Fanno flow is in
many cases shockless and therefore a relationship between two points should be
derived. In most times, the “star” values are imaginary values that represent the
value at choking. The real ratio can be obtained by two star ratios as an example
T ¯
¯
T2 T ∗ M2
= T ¯ (10.48)
T1 ¯
T∗ M 1

A special interest is the equation for the dimensionless friction as following

Z L2 Z Lmax Z Lmax
4f L 4f L 4f L
dx = dx − dx (10.49)
L1 D L1 D L2 D

Hence,
µ ¶ µ ¶
4f Lmax 4f Lmax 4f L
= − (10.50)
D 2 D 1 D
10.6. EXAMPLES OF FANNO FLOW 185

10.6 Examples of Fanno Flow

Example 10.1:
Air flows from a reservoir and enters a uni- 2 =09 M :

form pipe with a diameter of 0.05 [m] and P0 =? = 0 05[ ℄ D : m

L = 10[m℄
length of 10 [m]. The air exits to the at- T0 =?ÆC
mosphere. The following conditions pre-
vail at the exit: P2 = 1[bar] temperature T2 = 27ÆC
T2 = 27◦ C M2 = 0.94 . Assume that the P2 = 1[bar℄
average friction factor to be f = 0.004 and
that the flow from the reservoir up to the Fig. -10.3: Schematic of Example (10.1)
pipe inlet is essentially isentropic. Esti-
mate the total temperature and total pressure in the reservoir under the Fanno
flow model.

S OLUTION
For isentropic, the flow to the pipe inlet, the temperature and the total pressure
at the pipe inlet are the same as those in the reservoir. Thus, finding the total
pressure and temperature at the pipe inlet is the solution. With the Mach number
and temperature known at the exit, the total temperature at the entrance can be
obtained by knowing the 4fDL . For given Mach number (M = 0.9) the following is
obtained.

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
0.90000 0.01451 1.1291 1.0089 1.0934 0.9146 1.0327

So, the total temperature at the exit is

¯
T ∗ ¯¯ 300
T ∗ |2 = ¯ T2 = = 290.5[K]
T 2 1.0327

4f L
To “move” to the other side of the tube the D is added as
¯ ¯ 4 × 0.004 × 10
4f L ¯ 4f L 4f L ¯
D ¯ = D + D ¯ = + 0.01451 ' 3.21
1 2 0.05
4f L
The rest of the parameters can be obtained with the new D either from Table
(10.1) by interpolations or by utilizing the attached program.

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
0.35886 3.2100 3.0140 1.7405 2.5764 0.38814 1.1699
4 This property is given only for academic purposes. There is no Mach meter.
186 CHAPTER 10. FANNO FLOW

Note that the subsonic branch is chosen. The stagnation ratios has to be added
for M = 0.35886

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.35886 0.97489 0.93840 1.7405 0.91484 1.5922 0.78305

The total pressure P01 can be found from the combination of the ratios as follows:
P1
z }| {
∗
P
z }| ¯{ ¯ ¯
P ∗ ¯¯ P ¯¯ P0 ¯¯
P01 = P2
P ¯2 P ∗ ¯1 P ¯1
1 1
=1 × × 3.014 × = 2.91[Bar]
1.12913 0.915

T1
z }| {
∗
T
z }| ¯{ ¯ ¯
T ∗ ¯¯ T ¯¯ T0 ¯¯
T01 = T2
T ¯2 T ∗ ¯1 T ¯1
1 1
=300 × × 1.17 × ' 348K = 75◦ C
1.0327 0.975

Another academic question/example:

Example 10.2:
A system is composed of a convergent-
divergent nozzle followed by a tube with = 29 65[ ℄ = 3 0 Mx =?
P0 :
= 0 025[ ℄
bar
= 1 0[ ℄ M1 :
D

L :
:

m
m

length of 2.5 [cm] in diameter and 1.0 [m] T = 400K 0

shock
long. The system is supplied by a vessel.
atmosphere
conditions
d-c nozzle
The vessel conditions are at 29.65 [Bar],
400 K. With these conditions a pipe inlet
Mach number is 3.0. A normal shock wave Fig. -10.4: The schematic of Example
occurs in the tube and the flow discharges (10.2)
to the atmosphere, determine:

(a) the mass flow rate through the system;

(b) the temperature at the pipe exit; and

(c) determine the Mach number when a normal shock wave occurs [Mx ].
Take k = 1.4, R = 287 [J/kgK] and f = 0.005.
10.6. EXAMPLES OF FANNO FLOW 187

S OLUTION

(a) Assuming that the pressure vessel is very much larger than the pipe, there-
fore the velocity in the vessel can be assumed to be small enough so it
can be neglected. Thus, the stagnation conditions can be approximated
for the condition in the tank. It is further assumed that the flow through
the nozzle can be approximated as isentropic. Hence, T01 = 400K and
P01 = 29.65[P ar]
The mass flow rate through the system is constant and for simplicity point 1
is chosen in which,

ṁ = ρAM c

The density and speed of sound are unknowns and need to be computed.
With the isentropic relationship the Mach number at point one (1) is known,
then the following can be found either from Table (10.1) or the Potto–GDC

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
3.0000 0.35714 0.07623 4.2346 0.02722 0.11528 0.65326

The temperature is
T1
T1 = T01 = 0.357 × 400 = 142.8K
T01
Using the temperature, the speed of sound can be calculated as
√ √
c1 = kRT = 1.4 × 287 × 142.8 ' 239.54[m/sec]

The pressure at point 1 can be calculated as

P1
P1 = P01 = 0.027 × 30 ' 0.81[Bar]
P01
The density as a function of other properties at point 1 is
¯ · ¸
P ¯¯ 8.1 × 104 kg
ρ1 = = ' 1.97
RT ¯1 287 × 142.8 m3

The mass flow rate can be evaluated from equation (10.2)

· ¸
π × 0.0252 kg
ṁ = 1.97 × × 3 × 239.54 = 0.69
4 sec
188 CHAPTER 10. FANNO FLOW

(b) First, check whether the flow is shockless by comparing the flow resistance
and the maximum possible resistance. From the Table (10.1) or by using the
Potto–GDC, to obtain the following

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
3.0000 0.52216 0.21822 4.2346 0.50918 1.9640 0.42857

and the conditions of the tube are

4f L 4 × 0.005 × 1.0
D = = 0.8
0.025

Since 0.8 > 0.52216 the flow is choked and with a shock wave.

The exit pressure determines the location of the shock, if a shock exists,
by comparing “possible” Pexit to PB . Two possibilities are needed to be
checked; one, the shock at the entrance of the tube, and two, shock at
the exit and comparing the pressure ratios. First, the possibility that the
shock wave occurs immediately at the entrance for which the ratio for Mx
are (shock wave Table (6.1))

Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
3.0000 0.47519 2.6790 3.8571 10.3333 0.32834

After the shock wave the flow is subsonic with “M1 ”= 0.47519. (Fanno flow
Table (10.1))

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
0.47519 1.2919 2.2549 1.3904 1.9640 0.50917 1.1481

The stagnation values for M = 0.47519 are

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.47519 0.95679 0.89545 1.3904 0.85676 1.1912 0.65326
10.6. EXAMPLES OF FANNO FLOW 189

The ratio of exit pressure to the chamber total pressure is

1 1
zµ }| ¶{ µ ¶ µ ¶µ ¶ zµ }| ¶{
P2 P2 P∗ P1 P0y P0 x
=
P0 P∗ P1 P0 y P0 x P0
1
= 1× × 0.8568 × 0.32834 × 1
2.2549
= 0.12476

The actual pressure ratio 1/29.65 = 0.0338 is smaller than the case in which
shock occurs at the entrance. Thus, the shock is somewhere downstream.
One possible way to find the exit temperature, T2 is by finding the location
of the shock. To find the location of the shock ratio of the pressure ratio,
P2
P1 is needed. With the location of shock, “claiming” upstream from the exit
through shock to the entrance. For example, calculate the parameters for
shock location with known 4fDL in the “y” side. Then either by utilizing shock
table or the program, to obtain the upstream Mach number.

The procedure for the calculations:

Calculate the entrance Mach number assuming the shock occurs at the exit:
0
1) a) set M2 = 1 assume the flow in the entire tube is supersonic:
0
b) calculated M1
Note this Mach number is the high Value.

Calculate the entrance Mach assuming shock at the entrance.

a) set M2 = 1
2) b) add 4fDL and calculated M1 ’ for subsonic branch
c) calculated Mx for M1 ’
Note this Mach number is the low Value.

According your root finding algorithm5 calculate or guess the shock location
and then compute as above the new M1 .
a) set M2 = 1
3) b) for the new 4f L and compute the new M ’ for the subsonic branch
D y
c) calculated Mx ’ for the My ’
4f L
d) Add the leftover of D and calculated the M1

4) guess new location for the shock according to your finding root procedure
and according to the result, repeat previous stage until the solution is ob-
tained.
190 CHAPTER 10. FANNO FLOW

4fL ¯
¯ ¯
4fL ¯
M1 M2 D up D down Mx My
3.0000 1.0000 0.22019 0.57981 1.9899 0.57910

(c) The ¯way of the numerical procedure for solving this problem is by finding
4f L ¯
D ¯ that will produce M1 = 3. In the process Mx and My must be calcu-
up
lated (see the chapter on the program with its algorithms.).

10.7 Supersonic Branch

In Chapter (9) it was shown that the isothermal model cannot describe adequately
the situation because the thermal entry length is relatively large compared to the
pipe length and the heat transfer is not sufficient to maintain constant temperature.
In the Fanno model there is no heat transfer, and, furthermore, because the very
limited amount of heat transformed it is closer to an adiabatic flow. The only limi-
tation of the model is its uniform velocity (assuming parabolic flow for laminar and
different profile for turbulent flow.). The information from the wall to the tube cen-
ter6 is slower in reality. However, experiments from many starting with 1938 work
by Frossel7 has shown that the error is not significant. Nevertheless, the com-
parison with reality shows that heat transfer cause changes to the flow and they
need/should to be expected. These changes include the choking point at lower
Mach number.

10.8 Maximum Length for the Supersonic Flow

It has to be noted and recognized that as opposed to subsonic branch the su-
personic branch has a limited length. It also must be recognized that there is a
maximum length for which only supersonic flow can exist8 . These results were
obtained from the mathematical derivations but were verified by numerous exper-
iments9 . The maximum length of the supersonic can be evaluated when M = ∞

6 The word information referred to is the shear stress transformed from the wall to the center of the

tube.
7 See on the web http://naca.larc.nasa.gov/digidoc/report/tm/44/NACA-TM-844.PDF
8 Many in the industry have difficulties in understanding this concept. The author seeks for a nice

explanation of this concept for non–fluid mechanics engineers. This solicitation is about how to explain
this issue to non-engineers or engineer without a proper background.
9 If you have experiments demonstrating this point, please provide to the undersign so they can be

added to this book. Many of the pictures in the literature carry copyright statements.
10.9. WORKING CONDITIONS 191

as follows:
k+1 2
4f Lmax 1 − M2 k+1 2 M
= + ln ¡ ¢=
D kM 2 2k 2 1 + k−1
2 M
2

4f L −∞ k + 1 (k + 1)∞
D (M → ∞) ∼ + ln
k×∞ 2k (k − 1)∞
−1 k + 1 (k + 1)
= + ln
k 2k (k − 1)
4f L
= D (M → ∞, k = 1.4) = 0.8215

The maximum length of the supersonic flow is limited by the above number. From
the above analysis, it can be observed that no matter how high the entrance Mach
number will be the tube length is limited and depends only on specific heat ratio, k
as shown in Figure (10.5).

The maximum length in supersonic flow

In Fanno Flow
1.5
1.4
1.3
1.2
4fLmax

1.1
maximum length, 
D

1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1.2 1.25 1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65
Thu Mar 3 16:24:00 2005 spesific heat, k

Fig. -10.5: The maximum length as a function of specific heat, k

10.9 Working Conditions

It has to be recognized that there are two regimes that can occur in Fanno flow
model one of subsonic flow and the other supersonic flow. Even the flow in the tube
starts as a supersonic in parts of the tube can be transformed into the subsonic
192 CHAPTER 10. FANNO FLOW

branch. A shock wave can occur and some portions of the tube will be in a subsonic
flow pattern.
The discussion has to differentiate between two ways of feeding the tube: converg-
ing nozzle or a converging-diverging nozzle. Three parameters, the dimensionless
friction, 4fDL , the entrance Mach number, M1 , and the pressure ratio, P2 /P1 are
controlling the flow. Only a combination of these two parameters is truly indepen-
dent. However, all the three parameters can be varied and they are discussed
separately here.

10.9.1 Variations of The Tube Length ( 4fDL ) Effects

fL 1 CCC <
s BBB 4fL2 CCC

0 1
4 0

s D A
B
B
1
B
 
D A
T0
Larger ) 4DfL
T

s
4f L
Fig. -10.6: The effects of increase of D
on the Fanno line

In the analysis of this effect, it should be assumed that back pressure is constant
and/or low as possible as needed to maintain a choked flow. First, the treatment of
the two branches are separated.

Fanno Flow Subsonic branch

For converging nozzle feeding, increasing the tube length results in increasing
the exit Mach number (normally denoted herein as M2 ). Once the Mach num-
ber reaches maximum (M = 1), no further increase of the exit Mach number can
be achieved. In this process, the mass flow rate decreases. It is worth noting that
entrance Mach number is reduced (as some might explain it to reduce the flow
rate). The entrance temperature increases as can be seen from Figure (10.7). The
10.9. WORKING CONDITIONS 193
constant pressure
lines

T0

T 1’’
2’’
1’
1 2’

Fanno lines

Fig. -10.7: The development properties in of converging nozzle

velocity therefore must decrease because the loss of the enthalpy (stagnation tem-
P
perature) is “used.” The density decrease because ρ = RT and when pressure
is remains almost constant the density decreases. Thus, the mass flow rate must
decrease. These results are applicable to the converging nozzle.
In the case of the converging–diverging feeding nozzle, increase of the dimension-
less friction, 4fDL , results in a similar flow pattern as in the converging nozzle. Once
the flow becomes choked a different flow pattern emerges.

Fanno Flow Supersonic Branch

There are several transitional points that change the pattern of the flow. Point a
is the choking point (for the supersonic branch) in which the exit Mach number
reaches to one. Point b is the maximum possible flow for supersonic flow and is
not dependent on the nozzle. The next point, referred here as the critical point
c, is the point in which no supersonic flow is possible in the tube i.e. the shock
reaches to the nozzle. There is another point d, in which no supersonic flow is
possible in the entire nozzle–tube system. Between these transitional points the
effect parameters such as mass flow rate, entrance and exit Mach number are
discussed.
At the starting point the flow is choked in the nozzle, to achieve supersonic flow.
The following ranges that has to be discussed includes (see Figure (10.8)):
194 CHAPTER 10. FANNO FLOW

³ ´
4f L 4f L
0 < D < D 0→a
³ ´ ³ ´choking
4f L 4f L 4f L
D < D < D a→b
³ ´ choking ³ ´shockless
4f L 4f L 4f L
D < D < D b→c
³ ´shockless chokeless
4f L 4f L
D < D < ∞ c→∞
chokeless

M 1

M M 2 M =1
all supersonic
flow b c
m m_ = onst
_
mixed supersonic
with subsonic
flow with a shock the nozzle
between is still
choked
M
fL
1
4
D
Fig. -10.8: The Mach numbers at entrance and exit of tube and mass flow rate for Fanno
Flow as a function of the 4fDL .

The 0-a range, the mass flow rate is constant because the flow is choked at the
nozzle. The entrance Mach number, M1 is constant because it is a function of the
nozzle design only. The exit Mach number, M2 decreases (remember this flow is
on the supersonic branch) and starts ( 4fDL = 0) as M2 = M1 . At the end of the
range a, M2 = 1. In the range of a − b the flow is all supersonic.
In the next range a − −b The flow is double choked and make the adjustment for
the flow rate at different choking points by changing the shock location. The mass
flow rate continues to be constant. The entrance Mach continues to be constant
and exit Mach number is constant. ³ ´
The total maximum available for supersonic flow b − −b0 , 4fDL , is only a
max
theoretical length in which the supersonic flow can occur if nozzle is provided with
a larger Mach number (a change to the nozzle area ratio which also reduces the
mass flow rate). In the range b − c, it is a more practical point.
In semi supersonic flow b − c (in which no supersonic is available in the tube but
only in the nozzle) the flow is still double choked and the mass flow rate is constant.
Notice that exit Mach number, M2 is still one. However, the entrance Mach number,
M1 , reduces with the increase of 4fDL .
It is worth noticing that in the a − c the mass flow rate nozzle entrance velocity and
10.9. WORKING CONDITIONS 195

the exit velocity remains constant!10

In the last range c − ∞ the end is really the pressure limit or the break of the model
and the isothermal model is more appropriate to describe the flow. In this range,
the flow rate decreases since (ṁ ∝ M1 )11 .
To summarize the above discussion, Figures (10.8) exhibits the development of
M1 , M2 mass flow rate as a function of 4fDL . Somewhat different then the subsonic
branch the mass flow rate is constant even if the flow in the tube is completely
subsonic. This situation is because of the “double” choked condition in the nozzle.
The exit Mach M2 is a continuous monotonic function that decreases with 4fDL .
The entrance Mach M1 is a non continuous function with a jump at the point when
shock occurs at the entrance “moves” into the nozzle.

Fanno Flow
M1 as a function of M2
1

0.9 4fL = 0.1


D
= 1.0
0.8
= 10.0
= 100.0
0.7
Entrace Mach number

0.6

0.5

0.4

0.3

0.2

0.1

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

Exit Mach number

Tue Oct 19 09:56:15 2004

4f L
Fig. -10.9: M1 as a function M2 for various D

10 On a personal note, this situation is rather strange to explain. On one hand, the resistance in-
creases and on the other hand, the exit Mach number remains constant and equal to one. Does
anyone have an explanation for this strange behavior suitable for non–engineers or engineers without
background in fluid mechanics?
11 Note that ρ increases with decreases of M but this effect is less significant.
1 1
196 CHAPTER 10. FANNO FLOW

Figure (10.9) exhibits the M1 as a function of M2 . The Figure

¯ was calculated by uti-
4f L ¯
lizing the data from Figure (10.2) by obtaining the D ¯ for M2 and subtracting
max
4f L
the given D and finding the corresponding M1 .

Fanno Flow
M1 as a function of M2 for the subsonic brench
5
4fL = 0.1

4.5
D
4 = 0.2
= 0.4
3.5 = 0.1 shock
= 0.4
3
M1

2.5

1.5

0.5

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
M2

Tue Jan 4 11:26:19 2005

4f L
Fig. -10.10: M1 as a function M2 for different D
for supersonic entrance velocity.

The Figure (10.10) exhibits the entrance Mach number as a function of the M2 .
Obviously there can be two extreme possibilities for the subsonic exit branch. Sub-
sonic velocity occurs for supersonic entrance velocity, one, when the shock wave
occurs at the tube exit and two, at the tube entrance. In Figure (10.10) only for
4f L 4f L 4f L
D = 0.1 and D = 0.4 two extremes are shown. For D = 0.2 shown with only
shock at the exit only. Obviously, and as can be observed, the larger 4fDL creates
larger differences between exit Mach number for the different shock locations. The
larger 4fDL larger M1 must occurs even for shock at the entrance.
For a given 4fDL , below the maximum critical length, the supersonic entrance flow
has three different regimes which depends on the back pressure. One, shockless
flow, tow, shock at the entrance, and three, shock at the exit. Below, the maximum
10.9. WORKING CONDITIONS 197

critical length is mathematically

4f L 1 1+k k+1
>− + ln
D k 2k k−1

For cases of 4fDL above the maximum critical length no supersonic flow can be
over the whole tube and at some point a shock will occur and the flow becomes
subsonic flow12 .

P2
10.9.2 The Pressure Ratio, P1
, effects
In this section the studied parameter is the variation of the back pressure and
thus, the pressure ratio P P1 variations. For very low pressure ratio the flow can be
2

assumed as incompressible with exit Mach number smaller than < 0.3. As the
pressure ratio increases (smaller back pressure, P2 ), the exit and entrance Mach
numbers increase. According to Fanno model the value of 4fDL is constant (fric-
tion factor, f , is independent of the parameters such as, Mach number, Reynolds
number et cetera) thus the flow remains on the same Fanno line. For cases where
the supply come from a reservoir with a constant pressure, the entrance pressure
decreases as well because of the increase in the entrance Mach number (velocity).
Again a differentiation of the feeding is important to point out. If the feeding nozzle
is converging than the flow will be only subsonic. If the nozzle is “converging–
diverging” than in some part supersonic flow is possible. At first the converging
nozzle is presented and later the converging-diverging nozzle is explained.

Choking explanation for pressure variation/reduction

Decreasing the pressure ratio or in actuality the back pressure, results in increase
of the entrance and the exit velocity until a maximum is reached for the exit ve-
locity. The maximum velocity is when exit Mach number equals one. The Mach
number, as it was shown in Chapter (5), can increases only if the area increase. In
our model the tube area is postulated as a constant therefore the velocity cannot
increase any further. However, for the flow to be continuous the pressure must de-
crease and for that the velocity must increase. Something must break since there
are conflicting demands and it result in a “jump” in the flow. This jump is referred
to as a choked flow. Any additional reduction in the back pressure will not change
the situation in the tube. The only change will be at tube surroundings which are
irrelevant to this discussion.
If the feeding nozzle is a “converging–diverging” then it has to be differentiated
between two cases; One case is where the 4fD¯L is short or equal to the critical
¯
length. The critical length is the maximum 4fDL ¯ that associate with entrance
max
Mach number.
12 See more on the discussion about changing the length of the tube.
198 CHAPTER 10. FANNO FLOW
P1
P P2

a shock in
the nozzle
fully subsoinic
flow

P2
P1

critical Point a
criticalPoint b critical Point c

4fL
D
critical Point d

4f L 4f L
Fig. -10.11: The pressure distribution as a function of D
for a short D

4f L
Short D

Figure (10.12) shows different pressure profiles for different back pressures. Before
the flow reaches critical point a (in the Figure) the flow is subsonic. Up to this stage
the nozzle feeding the tube increases the mass flow rate (with decreasing back
pressure). Between point a and point b the shock is in the nozzle. In this range
and further reduction of the pressure the mass flow rate is constant no matter how
low the back pressure is reduced. Once the back pressure is less than point b the
supersonic reaches to the tube. Note however that exit Mach number, M2 < 1 and
is not 1. A back pressure that is at the critical point c results in a shock wave that
is at the exit. When the back pressure is below point c, the tube is “clean” of any
shock13 . The back pressure below point c has some adjustment as it occurs with
exceptions of point d.

13 It is common misconception that the back pressure has to be at point d.

10.9. WORKING CONDITIONS 199
P1
P P2

maximum riti al B
0
4 fL 1CA
D
a shock in
the nozzle
fully subsoinic
flow

B 4
0 1
fun tion of M1 ; and
fL

D
CA

P2
P1

critical Point a
criticalPoint b
1

{
M = 1
0B4fL 1CA
D
critical Point c

4fL
D
4f L 4f L
Fig. -10.12: The pressure distribution as a function of D
for a long D

4f L
Long D
¯
4f L 4f L ¯
In the case of D > D ¯ reduction of the back pressure results in the same
max
4f L
process as explained in the short up to point c. However,
D ¯ point c in this case is
4f L 4f L ¯
different from point c at the case of short tube D < D ¯ . In this point the exit
max
Mach number is equal to 1 and the flow is double shock. Further reduction of the
back pressure at this stage will not “move” the shock wave downstream the nozzle.
At point c or location of the shock wave, is a function entrance Mach number, M1
and the “extra” 4fDL . The is no analytical solution for the location of this point c. The
procedure is (will be) presented in later stage.

10.9.3 Entrance Mach number, M1 , effects

In this discussion, the effect of changing the throat area on the nozzle efficiency
is neglected. In reality these effects have significance and needs to be accounted
200 CHAPTER 10. FANNO FLOW
Mach number in Fanno Flow

4fL

D
2

1.8

1.6
shock at
1.4
75%
50%
Mach Number
1.2
5%
1

0.8

0.6

0.4

0.2

0
0 0.05 0.1 0.15 0.2 0.25
4fL

D
Tue Jan 4 12:11:20 2005
4f L
Fig. -10.13: The effects of pressure variations on Mach number profile as a function of D
when the total resistance 4fDL = 0.3 for Fanno Flow

for some instances. This dissection deals only with the flow when it reaches the
supersonic branch reached otherwise the flow is subsonic with regular effects. It is
P2
assumed that in this discussion that the pressure ratio P 1
is large enough to create
4f L
a choked flow and D is small enough to allow it to happen.
The entrance Mach number, M1 is a function of the ratio of the nozzle’s throat area
to the nozzle exit area and its efficiency. This effect is the third parameter discussed
here. Practically, the nozzle area ratio is changed by changing the throat area.
As was shown before, there are two different maximums for 4fDL ; first is the total
maximum 4fDL of the supersonic which depends only on the specific heat, k, and
second the maximum depends on the entrance Mach number, ¯ M1 . This analysis
4f L 4f L ¯
deals with the case where D is shorter than total D ¯ .
max ¯
¯
Obviously, in this situation, the critical point is where 4fDL is equal to 4fDL ¯ as a
max
result in the entrance Mach number.
The process of decreasing the converging–diverging nozzle’s throat increases the
10.9. WORKING CONDITIONS 201
P2/P1 Fanno Flow

4fL

D
4.8
4.4
4
3.6
3.2 5%
50 %
2.8 75 %
P2/P1

2.4
2
1.6
1.2
0.8
0.4
0
0 0.05 0.1 0.15 0.2 0.25
4fL

D
Fri Nov 12 04:07:34 2004
4f L 4f L
Fig. -10.14: Mach number as a function of D
when the total D
= 0.3

entrance14 Mach number. If the tube contains no supersonic flow then reducing
the nozzle throat area wouldn’t increase the entrance Mach number.
This part is for the case where some part of the tube is under supersonic regime
and there is shock as a transition to subsonic branch. Decreasing the nozzle throat
area moves the shock location downstream. The “payment” for increase in the
supersonic length is by reducing the mass flow. Further, decrease of the throat
area results in flushing the shock out of the tube. By doing so, the throat area de-
creases. The mass flow rate is proportionally linear to the throat area and therefore
the mass flow rate reduces. The process of decreasing the throat area also results
in increasing the pressure drop of the nozzle (larger resistance in the nozzle15 )16 .

14 The word “entrance” referred to the tube and not to the nozzle. The reference to the tube is because

it is the focus of the study.

15 Strange? Frictionless nozzle has a larger resistance when the throat area decreases
16 It is one of the strange phenomenon that in one way increasing the resistance (changing the throat

area) decreases the flow rate while in a different way (increasing the 4fDL ) does not affect the flow rate.
202 CHAPTER 10. FANNO FLOW

4 fL 


0
fL CA

B 4D
1

D max1





4fL 



D retreat





M = 1 or less Mx M y
M =1
shock

Fig. -10.15: Schematic of a “long” tube in supersonic branch

¯
4f L 4f L ¯
In the case of large tube D > D ¯ the exit Mach number increases with the
max
decrease of the throat area. Once the exit Mach number reaches one no further
increases is possible. However, the location of the shock wave approaches to the
theoretical location if entrance Mach, M1 = ∞.

0
fL CA

B 4D
1

M 1 =8

M 1 =5

M =1
1

fL
4


 4fL 



D retreat D max
 
 
 
 

4f L
Fig. -10.16: The extra tube length as a function of the shock location, D
supersonic branch

The maximum location of the shock The main point in this discussion how-
ever, is to find
³ the
´ furthest shock location downstream. Figure (10.16) shows the
possible ∆ 4fDL as function of retreat of the location of the shock wave from the
10.9. WORKING CONDITIONS 203

maximum location. When the entrance Mach number is infinity, M1 = ∞, if the

shock location is at the maximum length, then shock at Mx = 1 results in My = 1.
The proposed procedure is based on Figure (10.16).
i) Calculate the extra 4fDL and subtract the actual extra 4f L
D assuming shock at
the left side (at the max length).

ii) Calculate the extra 4fDL and subtract the actual extra 4f L
D assuming shock at
the right side (at the entrance).
iii) According to the positive or negative utilizes your root finding procedure.

M 1 max

4fL
4f L 
D



D max∞


4f L
Fig. -10.17: The maximum entrance Mach number, M1 to the tube as a function of D
supersonic branch

From numerical point of view, the Mach number equal infinity when left side as-
sumes result in infinity length of possible extra (the whole flow in the tube is sub-
sonic). To overcome this numerical problem it is suggested to start the calculation
from ² distance from the right hand side.
Let denote
µ ¶ ¯
4f L 4f¯L 4f L ¯¯
∆ = − (10.51)
D D actual D ¯sup
¯ ¯
¯ ¯
Note that 4fDL ¯ is smaller than 4fDL ¯ . The requirement that has to be
sup ¯ max∞
¯
satisfied is that denote 4fDL ¯ as difference between the maximum possible
retreat
204 CHAPTER 10. FANNO FLOW

of length in which the supersonic flow is achieved and the actual length in which
the flow is supersonic see Figure (10.15). The retreating length is expressed as
subsonic but
¯ ¯ ¯
4f L ¯¯ 4f L ¯¯ 4f L ¯¯
= − (10.52)
D ¯retreat D ¯max∞ D ¯sup

Figure (10.17) shows the entrance Mach number, M1 reduces after the maximum
length is exceeded.

Example 10.3:
4f L
Calculate the shock location for entrance Mach number M1 = 8 and for D = 0.9
assume that k = 1.4 (Mexit = 1).

S OLUTION ¯
4f L ¯
The solution is obtained by an iterative process. The maximum D ¯ for k = 1.4
max
is 0.821508116. Hence, 4fDL exceed the maximum length 4f L
D for this entrance
Mach³ number.
´ The maximum for M1 = 8 is 4fDL = 0.76820, thus the extra tube
4f L
is ∆ D = 0.9 − 0.76820 = 0.1318. The left side is when the shock occurs at
4f L
D = 0.76820 (flow is choked and no additional 4fDL ). Hence, the value of left side
is −0.1318. The right side is when the shock is at the entrance at which the extra
4f L
D is calculated for Mx and My is

Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
8.0000 0.39289 13.3867 5.5652 74.5000 0.00849

With (M1 )0

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
0.39289 2.4417 2.7461 1.6136 2.3591 0.42390 1.1641
³ ´
4f L
The extra ∆ D is 2.442 − 0.1318 = 2.3102 Now the solution is somewhere
between the negative of left side to the positive of the right side17 .
In a summary of the actions is done by the following algorithm:
(a) check if the 4fDL exceeds the maximum 4f L
D max for the supersonic flow.
Accordingly continue.
¯
¯
(b) Guess 4fDL up = 4fDL − 4fDL ¯
max
17 What if the right side is also negative? The flow is chocked and shock must occur in the nozzle

before entering the tube. Or in a very long tube the whole flow will be subsonic.
10.9. WORKING CONDITIONS 205
4f L
(c) Calculate the Mach number corresponding to the current guess of D up ,

(d) Calculate the associate Mach number, Mx with the Mach number, My cal-
culated previously,
4f L
(e) Calculate D for supersonic branch for the Mx
4f L
(f) Calculate the “new and improved” D up

4f L 4f L 4f L
(g) Compute the “new D down = D − D up
¯
4f L ¯
(h) Check the new and improved D ¯ against the old one. If it is satisfac-
down
tory stop or return to stage (b).
Shock location are:

4fL ¯
¯ ¯
4fL ¯
M1 M2 D up D down Mx My
8.0000 1.0000 0.57068 0.32932 1.6706 0.64830

The iteration summary is also shown below

¯
4fL ¯
¯
4fL ¯ 4fL
i D up D down Mx My D
0 0.67426 0.22574 1.3838 0.74664 0.90000
1 0.62170 0.27830 1.5286 0.69119 0.90000
2 0.59506 0.30494 1.6021 0.66779 0.90000
3 0.58217 0.31783 1.6382 0.65728 0.90000
4 0.57605 0.32395 1.6554 0.65246 0.90000
5 0.57318 0.32682 1.6635 0.65023 0.90000
6 0.57184 0.32816 1.6673 0.64920 0.90000
7 0.57122 0.32878 1.6691 0.64872 0.90000
8 0.57093 0.32907 1.6699 0.64850 0.90000
9 0.57079 0.32921 1.6703 0.64839 0.90000
10 0.57073 0.32927 1.6705 0.64834 0.90000
11 0.57070 0.32930 1.6706 0.64832 0.90000
12 0.57069 0.32931 1.6706 0.64831 0.90000
13 0.57068 0.32932 1.6706 0.64831 0.90000
14 0.57068 0.32932 1.6706 0.64830 0.90000
15 0.57068 0.32932 1.6706 0.64830 0.90000
16 0.57068 0.32932 1.6706 0.64830 0.90000
17 0.57068 0.32932 1.6706 0.64830 0.90000

This procedure rapidly converted to the solution.

206 CHAPTER 10. FANNO FLOW

10.10 The Practical Questions and Examples of Sub-

sonic branch
The Fanno is applicable also when the flow isn’t choke18 . In this case, several
questions appear for the subsonic branch. This is the area shown in Figure (10.8)
in beginning for between points 0 and a. This kind of questions made of pair
given information to find the conditions of the flow, as oppose to only one piece of
information given in choked flow. There many combinations that can appear in this
situation but there are several more physical and practical that will be discussed
here.

4f L
10.10.1 Subsonic Fanno Flow for Given D
and Pressure Ratio
This pair of parameters is the
most natural to examine because, M
P
4f L
D
1
1
M
P ∆
4f L
2
2

∗
D P =P
in most cases, this information is
M =1
the only information that ³is pro-´ hypothetical section

vided. For a given pipe 4fDL ,

neither the entrance Mach num- Fig. -10.18: Unchoked flow calculations showing
ber nor the exit Mach number are the hypothetical “full” tub when choked
given (sometimes the entrance
Mach number is give see the next section). There is no exact analytical solution.
There are two possible approaches to solve this problem: one, by building a rep-
resentative function and find a root (or roots) of this representative function. Two,
the problem can be solved by an iterative procedure. The first approach require
using root finding method and either method of spline method or the half method
found to be good. However, this author experience show that these methods in
this case were found to be relatively slow. The Newton–Rapson method is much
faster but not were found to be unstable (at lease in the way that was implemented
by this author). The iterative method used to solve constructed on the properties
of several physical quantities must be in a certain range.. The first fact is that the
pressure ratio P2 /P1 is always between 0 and 1 (see Figure (10.18)). In the figure,
a theoretical extra tube is added in such a length that cause the flow to choke (if it
really was there). This length is always positive (at minimum is zero).
The procedure for the calculations is as the following:
¯ 0
0
4f L 4f L ¯
1) Calculate the entrance Mach number, M1 assuming the D = D ¯
max
(chocked flow);
0
2) Calculate the minimum pressure ratio (P2 /P1 )min for M1 (look at table (10.1))
18 This questions were raised from many who didn’t found any book that discuss these practical as-

pects and send questions to this author.

 10.10. PRACTICAL EXAMPLES FOR SUBSONIC FLOW 207

3) Check if the flow is choked:

There are two possibilities to check it.

4f L 4f L
a) Check if the given D is smaller than D obtained from the given P1 /P2 , or
b) check if the (P2 /P1 )min is larger than (P2 /P1 ),

continue if the criteria is satisfied. Or if not satisfied abort this procedure

and continue to calculation for choked flow.

4) Calculate the M2 based on the (P ∗ /P2 ) = (P1 /P2 ),

5) calculate ∆ 4fDL based on M2 ,

³³ ´ ³ ´ ´
4f L 4f L
6) calculate the new (P2 /P1 ), based on the new f D , D ,
³ ´ 1 2
(remember that ∆ 4fDL = 4fDL ),
2

7) calculate the corresponding M1 and M2 ,

8) calculate the new and “improve” the ∆ 4fDL by

³ ´
µ ¶ µ ¶ P2
4f L 4f L P1
given
∆ = ∆ ∗ ³ ´ (10.53)
D new D old P2
P1
old

Note, when the pressure ratios are matching also the ∆ 4fDL will also match.

9) Calculate the “improved/new” M2 based on the improve ∆ 4fDL

³ ´ ³ ´
4f L 4f L 4f L 4f L
10) calculate the improved D as D = D +∆ D
given new

4f L
11) calculate the improved M1 based on the improved D .

12) Compare the abs ((P2 /P1 )new − (P2 /P1 )old ) and if not satisfied
returned to stage (6) until the solution is obtained.

To demonstrate how this procedure is working consider a typical example of 4fDL =

1.7 and P2 /P1 = 0.5. Using the above algorithm the results are exhibited in the
following figure. Figure (10.19) demonstrates that the conversion occur at about
7-8 iterations. With better first guess this conversion procedure will converts much
faster (under construction).
208 CHAPTER 10. FANNO FLOW

3.0

Conversion occurs around 7-9 times

M1
2.5 M2
4f L
D
P2/P1
2.0 ∆ 4fDL

1.5

1.0

0.5

0 10 20 30 40 50 60 70 80 90 100 110 120 130

Number of Iterations, i

October 8, 2007

Fig. -10.19: The results of the algorithm showing the conversion rate for unchoked Fanno
flow model with a given 4fDL and pressure ratio.

10.10.2 Subsonic Fanno Flow for a Given M1 and Pressure Ra-

tio

This situation pose a simple mathematical problem while the physical situation
occurs in cases where a specific flow rate is required with a given pressure ratio
(range) (this problem was considered by some to be somewhat complicated). The
specific flow rate can be converted to entrance Mach number and this simplifies
the problem. Thus, the problem is reduced to find for given entrance Mach, M1 ,
and given pressure ratio calculate the flow parameters, like the exit Mach number,
M2 . The procedure is based on the fact that the entrance star pressure ratio can
be calculated using M1 . Thus, using the pressure ratio to calculate the star exit
pressure ratio provide the exit Mach number, M2 . An example of such issue is the
following example that combines also the “Naughty professor” problems.

Example 10.4:
Calculate the exit Mach number for P2 /P 1 = 0.4 and entrance Mach number M1 =
0.25.

S OLUTION
The star pressure can be obtained from a table or Potto-GDC as

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
0.25000 8.4834 4.3546 2.4027 3.6742 0.27217 1.1852
10.10. PRACTICAL EXAMPLES FOR SUBSONIC FLOW 209

And the star pressure ratio can be calculated at the exit as following

P2 P2 P1
= = 0.4 × 4.3546 = 1.74184
P∗ P1 P ∗
And the corresponding exit Mach number for this pressure ratio reads

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
0.60694 0.46408 1.7418 1.1801 1.5585 0.64165 1.1177

A bit show off the Potto–GDC can carry these calculations in one click as

4fL P2
M1 M2 D P1
0.25000 0.60693 8.0193 0.40000

While the above example show the most simple form of this question, in reality this
question is more complicated. One common problem is situation that the diameter
is not given but the flow rate and length and pressure (stagnation or static) with
some combination of the temperature. The following example deal with one of
such example.

Example 10.5:
A tank filled with air at stagnation pressure, 2[Bar] should be connected to a pipe
withh a friction
i factor, f = 0.005, and and length of 5[m]. The flow rate is (should be)
kg
0.1 sec and the static temperature at the entrance of the pipe was measured to be
27◦ C. The pressure ratio P2 /P1 should not fall below 0.9 (P2 /P1 > 0.9). Calculate
the exit Mach number, M2 , flow rate, and minimum pipe diameter. You can assume
that k = 1.4.

S OLUTION
The direct mathematical solution isn’t possible and some kind of iteration proce-
dure or root finding for a representative function. For the first part the “naughty
professor” procedure cannot be used because ṁ/A is not provided and the other
hand 4fDL is not provided (missing Diameter). One possible solution is to guess
the entrance Mach and check whether and the mass flow rate with the “naughty
professor” procedure are satisfied. For Fanno flow at for several Mach numbers
the following is obtained

4fL P2
M1 M2 D P1 Diameter
0.10000 0.11109 13.3648 0.90000 0.00748
0.15000 0.16658 5.8260 0.90000 0.01716
0.20000 0.22202 3.1887 0.90000 0.03136
210 CHAPTER 10. FANNO FLOW

0.4

guessed M1
calculated M1
0.3

Entrace Mach Nubmer

0.2

0.1

Solution
0.1 0.15 0.2 0.25 0.3
Conversion of the guesing the Mach Number

October 18, 2007

Fig. -10.20: Diagram for finding solution when the pressure ratio and entrance properties (T
and P0 are given

From the last table the diameter can be calculated for example for M1 = 0.2 as
4f L
D= 4f L
= 4 × 0.005 × 5/3.1887 = 0.03136[m]
D

The same was done for all the other Mach number. Now the area can be calcu-
lated and therefor the ṁ/A can be calculated. With this information the “naughty
professor” is given and the entrance Mach number can be calculated. For example
for M1 = 0.2 one can obtain the following:

ṁ/A = 0.1/(π × 0.031362 /4) ∼ 129.4666798

The same order as the above table it shown in “naughty professor” (isentropic
table).

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
1.5781 0.66752 0.36404 1.2329 0.24300 0.29960 0.56009
0.36221 0.97443 0.93730 1.7268 0.91334 1.5772 0.77785
0.10979 0.99760 0.99400 5.3092 0.99161 5.2647 2.2306

The first result are not reasonable and this process can continue until the satisfac-
tory solution is achieved. Here an graphical approximation is shown.
From this exhibit it can be estimated that M1 = 0.18. For this Mach number the
following can be obtained

4fL P2
M1 M2 D P1
0.18000 0.19985 3.9839 0.90000

Thus, the diameter can be obtained as D ∼ 0.0251[m]

The flow rate is ṁ/A ∼ 202.1[kg/sec × m2 ]
10.11. THE APPROXIMATION OF THE FANNO FLOW BY ISOTHERMAL FLOW211

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.17109 0.99418 0.98551 3.4422 0.97978 3.3726 1.4628

The exact solution is between 0.17 to 0.18 if better accuracy is needed.

10.11 The Approximation of the Fanno Flow by Isother-

mal Flow
The isothermal flow model has equations that theoreticians find easier to use and
to compare to the Fanno flow model.
One must notice that the maximum temperature at the entrance is T0 1 . When the
Mach number decreases the temperature approaches the stagnation temperature
(T → T0 ). Hence, if one allows certain deviation of temperature, say about 1% that
flow can be assumed to be isothermal. This tolerance requires that (T0 − T )/T0 =
0.99 which requires that enough for M1 < 0.15 even for large k = 1.67. This
requirement provides that somewhere (depend) in the vicinity of 4fDL = 25 the flow
can be assumed isothermal. Hence the mass flow rate is a function of 4fDL because
M1 changes. Looking at the table or Figure (10.2) or the results from Potto–GDC
attached to this book shows that reduction of the mass flow is very rapid. As it
can be seen for the Figure (10.21) the dominating parameter is 4fDL . The results
are very similar for isothermal flow. The only difference is in small dimensionless
friction, 4fDL .

10.12 More Examples of Fanno Flow

Example 10.6:
To demonstrate the utility in Figure (10.21) consider the following example. Find
the mass flow rate for f = 0.05, L = 4[m], D = 0.02[m] and pressure ratio P2 /P1 =
0.1, 0.3, 0.5, 0.8. The stagnation conditions at the entrance are 300K and 3[bar] air.

S OLUTION
4f L
First calculate the dimensionless resistance, D .

4f L 4 × 0.05 × 4
= = 40
D 0.02
From Figure (10.21) for P2 /P1 = 0.1 M1 ≈ 0.13 etc.
or accurately by utilizing the program as in the following table.
4fL
¯
4fL ¯
¯
4fL ¯ P2
M1 M2 D D 1 D 2 P1
0.12728 1.0000 40.0000 40.0000 0.0 0.11637
0.12420 0.40790 40.0000 42.1697 2.1697 0.30000
0.11392 0.22697 40.0000 50.7569 10.7569 0.50000
0.07975 0.09965 40.0000 107.42 67.4206 0.80000
212 CHAPTER 10. FANNO FLOW
M1 Fanno flow
with comperison to Isothermal Flow

0.4

P2 / P1 = 0.1 iso
P2 / P1 = 0.8 iso
0.3 P2 / P1 = 0.1
P2 / P1 = 0.2

M1
P2 / P1 = 0.5
0.2 P2 / P1 = 0.8

0.1

0
0 10 20 30 40 50 60 70 80 90 100
4fL

Wed Mar 9 11:38:27 2005 D

Fig. -10.21: The entrance Mach number as a function of dimensionless resistance and com-
parison with Isothermal Flow

Only for the pressure ratio of 0.1 the flow is choked.

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
0.12728 0.99677 0.99195 4.5910 0.98874 4.5393
0.12420 0.99692 0.99233 4.7027 0.98928 4.6523
0.11392 0.99741 0.99354 5.1196 0.99097 5.0733
0.07975 0.99873 0.99683 7.2842 0.99556 7.2519

Therefore, T ≈ T0 and is the same for the pressure. Hence, the mass rate is
a function of the Mach number. The Mach number is indeed a function of the
pressure ratio but mass flow rate is a function of pressure ratio only through Mach
number.
The mass flow rate is
r r µ ¶
k π × 0.022 1.4 kg
ṁ = P AM = 300000 × × 0.127 × ≈ 0.48
RT 4 287300 sec
10.13. THE TABLE FOR FANNO FLOW 213

and for the rest

µ ¶ µ ¶
P2 0.1242 kg
ṁ = 0.3 ∼ 0.48 × = 0.468
P1 0.1273 sec
µ ¶ µ ¶
P2 0.1139 kg
ṁ = 0.5 ∼ 0.48 × = 0.43
P1 0.1273 sec
µ ¶ µ ¶
P2 0.07975 kg
ṁ = 0.8 ∼ 0.48 × = 0.30
P1 0.1273 sec

10.13 The Table for Fanno Flow

Table -10.1: Fanno Flow Standard basic Table

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
0.03 787.08 36.5116 19.3005 30.4318 0.03286 1.1998
0.04 440.35 27.3817 14.4815 22.8254 0.04381 1.1996
0.05 280.02 21.9034 11.5914 18.2620 0.05476 1.1994
0.06 193.03 18.2508 9.6659 15.2200 0.06570 1.1991
0.07 140.66 15.6416 8.2915 13.0474 0.07664 1.1988
0.08 106.72 13.6843 7.2616 11.4182 0.08758 1.1985
0.09 83.4961 12.1618 6.4613 10.1512 0.09851 1.1981
0.10 66.9216 10.9435 5.8218 9.1378 0.10944 1.1976
0.20 14.5333 5.4554 2.9635 4.5826 0.21822 1.1905
0.25 8.4834 4.3546 2.4027 3.6742 0.27217 1.1852
0.30 5.2993 3.6191 2.0351 3.0702 0.32572 1.1788
0.35 3.4525 3.0922 1.7780 2.6400 0.37879 1.1713
0.40 2.3085 2.6958 1.5901 2.3184 0.43133 1.1628
0.45 1.5664 2.3865 1.4487 2.0693 0.48326 1.1533
0.50 1.0691 2.1381 1.3398 1.8708 0.53452 1.1429
0.55 0.72805 1.9341 1.2549 1.7092 0.58506 1.1315
0.60 0.49082 1.7634 1.1882 1.5753 0.63481 1.1194
0.65 0.32459 1.6183 1.1356 1.4626 0.68374 1.1065
0.70 0.20814 1.4935 1.0944 1.3665 0.73179 1.0929
0.75 0.12728 1.3848 1.0624 1.2838 0.77894 1.0787
0.80 0.07229 1.2893 1.0382 1.2119 0.82514 1.0638
0.85 0.03633 1.2047 1.0207 1.1489 0.87037 1.0485
0.90 0.01451 1.1291 1.0089 1.0934 0.91460 1.0327
0.95 0.00328 1.061 1.002 1.044 0.95781 1.017
1.00 0.0 1.00000 1.000 1.000 1.00 1.000
2.00 0.30500 0.40825 1.688 0.61237 1.633 0.66667
3.00 0.52216 0.21822 4.235 0.50918 1.964 0.42857
214 CHAPTER 10. FANNO FLOW
Table -10.1: Fanno Flow Standard basic Table (continue)

4fL P P0 ρ U T
M D P∗ P0 ∗ ρ∗ U∗ T∗
4.00 0.63306 0.13363 10.72 0.46771 2.138 0.28571
5.00 0.69380 0.089443 25.00 0.44721 2.236 0.20000
6.00 0.72988 0.063758 53.18 0.43568 2.295 0.14634
7.00 0.75280 0.047619 1.0E+2 0.42857 2.333 0.11111
8.00 0.76819 0.036860 1.9E+2 0.42390 2.359 0.086957
9.00 0.77899 0.029348 3.3E+2 0.42066 2.377 0.069767
10.00 0.78683 0.023905 5.4E+2 0.41833 2.390 0.057143
20.00 0.81265 0.00609 1.5E+4 0.41079 2.434 0.014815
25.00 0.81582 0.00390 4.6E+4 0.40988 2.440 0.00952
30.00 0.81755 0.00271 1.1E+5 0.40938 2.443 0.00663
35.00 0.81860 0.00200 2.5E+5 0.40908 2.445 0.00488
40.00 0.81928 0.00153 4.8E+5 0.40889 2.446 0.00374
45.00 0.81975 0.00121 8.6E+5 0.40875 2.446 0.00296
50.00 0.82008 0.000979 1.5E+6 0.40866 2.447 0.00240
55.00 0.82033 0.000809 2.3E+6 0.40859 2.447 0.00198
60.00 0.82052 0.000680 3.6E+6 0.40853 2.448 0.00166
65.00 0.82066 0.000579 5.4E+6 0.40849 2.448 0.00142
70.00 0.82078 0.000500 7.8E+6 0.40846 2.448 0.00122

10.14 Appendix
The friction factor in equation (10.25) was assumed constant. In Chapter 9 it was
shown that the Reynolds number remains constant for ideal gas fluid. However, in
Fanno flow the temperature does not remain constant hence as it was discussed
before the Reynolds number is increasing. Thus, the friction decreases with the
exception of the switch in the flow pattern (laminar to turbulent flow). For relatively
large relative roughness larger ²/D > 0.004 of 0.4% the friction factor is constant.
For smother pipe ²/D < 0.001 and Reynolds number between 10,000 to a million
the friction factor vary between 0.007 to 0.003 with is about factor of two. Thus,
the error of 4fDL is limited by a factor of two (2). For this range, the friction factor
can be estimated as a linear function of the log10 (Re). The error in this assumption
is probably small of the assumption that involve in fanno flow model construction.
Hence,

f = A log10 (Re) + B (10.54)

Where the constant A and B are function of the relative roughness. For most
practical purposes the slop coefficient A can be further assumed constant. The
slop coefficient A = −0.998125 Thus, to carry this calculation relationship between
the viscosity and the temperature. If the viscosity expanded as Taylor or Maclaren
10.14. APPENDIX 215

Almost Constant Constant

Zone Zone

Linear Representation
Zone

Small Error
Due to Linear
Assumption

Fig. -10.22: “Moody” diagram on the name Moody who netscape H. Rouse work to claim
as his own. In this section the turbulent area is divided into 3 zones, constant,
semi–constant, and linear After S Beck and R. Collins.

series then
µ A1 T
= A0 + + ··· (10.55)
µ1 T0
Where µ1 is the viscosity at the entrance temperature T1 .
Thus, Reynolds number is
DρU
Re = (10.56)
A0 + AT10T + · · ·

Substituting equation (10.56) into equation (10.54) yield

à !
DρU
f = A log10 +B (10.57)
A0 + AT1 1T2 + · · ·

Left hand side of equation (10.25) is a function of the Mach number since it con-
tains the temperature. If the temperature functionality will not vary similarly to the
case of constant friction factor then the temperature can be expressed using equa-
tion (10.41).
   
constant
z }| {
4 

 DρU 



A log10  2  + B  (10.58)
D  1 + k−1
2 M1  
A0 + A1 2 + ···
1 + k−1
2 M2
216 CHAPTER 10. FANNO FLOW

Equation (10.58) is only estimate of the functionally however, this estimate is al-
most as good as the assumptions of Fanno flow. Equation fanno:eq:fld2 can be
improved by using equation (10.58)
   
constant
z }| {
4 Lmax 


 DρU 

 1 1 − M2
 k+1 k+1
2 M
2
A log10   + B  ∼ + ln
D   1 + k−12 M
2   k M2 2k 1 + k−1
2 M
2
A0 + A1 k−1
1+ 2
(10.59)

In the most complicate case where the flow pattern is change from laminar flow
to turbulent flow the whole Fanno flow model is questionable and will produce poor
results.
In summary, in the literature there are three approaches to this issue of non con-
stant friction factor. The friction potential is recommended by a researcher in Ger-
many and it is complicated. The second method substituting this physical approach
with numerical iteration. In the numerical iteration method, the expression of the
various relationships are inserted into governing differential equations. The nu-
merical methods does not allow flexibility and is very complicated. The methods
described here can be expended (if really really needed) and it will be done in very
few iteration as it was shown in the Isothermal Chapter.
 CHAPTER 11
Rayleigh Flow

Rayleigh flow is model describing a frictionless flow with heat transfer through a
pipe of constant cross sectional area. In practice Rayleigh flow isn’t a really good
model for the real situation. Yet, Rayleigh flow is practical and useful concept
in a obtaining trends and limits such as the density and pressure change due to
external cooling or heating. As opposed to the two previous models, the heat
transfer can be in two directions not like the friction (there is no negative friction).
This fact creates a situation different as compare to the previous two models. This
model can be applied to cases where the heat transfer is significant and the friction
can be ignored.

11.1 Introduction
The third simple model for one– 2
1 flow
dimensional flow with constant heat transfer P1
direction
P2
for frictionless flow. This flow is referred in T1 T2
the literature as Rayleigh Flow (see historical
notes). This flow is another extreme case in Q
which the friction effect is neglected because heat transfer (in and out)
their relative effect is much smaller than the
heat transfer effect. While the isothermal flow Fig. -11.1: The control volume of
model has heat transfer and friction, the main Rayleigh Flow
assumption was that relative length is enables significant heat transfer to occur
between the surroundings and tube. In contrast, the heat transfer in Rayleigh flow
occurs between unknown temperature and the tube and the heat flux is maintained
constant. As before, a simple model is built around the assumption of constant
properties (poorer prediction to case where chemical reaction take a place).

217
218 CHAPTER 11. RAYLEIGH FLOW

This model is used to roughly predict the conditions which occur mostly in
situations involving chemical reaction. In analysis of the flow, one has to be aware
that properties do change significantly for a large range of temperatures. Yet, for
smaller range of temperatures and lengths the calculations are more accurate.
Nevertheless, the main characteristic of the flow such as a choking condition etc.
is encapsulated in this model.
The basic physics of the flow revolves around the fact that the gas is
highly compressible. The density changes through the heat transfer (tempera-
ture change). Contrary to Fanno flow in which the resistance always oppose the
flow direction, Rayleigh flow, also, the cooling can be applied. The flow velocity
acceleration change the direction when the cooling is applied.

11.2 Governing Equation

The energy balance on the control volume reads

Q = Cp (T0 2 − T0 1 ) (11.1)

the momentum balance reads

A(P1 − P2 ) = ṁ(V2 − V1 ) (11.2)

The mass conservation reads

ρ1 U1 A = ρ2 U2 A = ṁ (11.3)

Equation of state

P1 P2
= (11.4)
ρ1 T1 ρ2 T2

There are four equations with four unknowns, if the upstream conditions are known
(or downstream conditions are known). Thus, a solution can be obtained. One can
notice that equations (11.2), (11.3) and (11.4) are similar to the equations that were
solved for the shock wave.

P2 1 + kM1 2
= (11.5)
P1 1 + kM2 2

The equation of state (11.4) can further assist in obtaining the temperature ratio
as

T2 P2 ρ1
= (11.6)
T1 P1 ρ2
11.2. GOVERNING EQUATION 219

The density ratio can be expressed in terms of mass conservation as

U2 √
r kRT2 r
ρ1 U2 kRT2 M2 T2
= = q = (11.7)
ρ2 U1 U1 M1 T1
r kRT1
kRT1

or
r
ρ1 U2 M2 T2
= =
ρ2 U1 M1 T1
(11.8)
or Substituting equations (11.5) and (11.8) into equation (11.6) yields
r
T2 1 + kM1 2 M2 T2
= (11.9)
T1 1 + kM2 2 M1 T1
Transferring the temperature ratio to the left hand side and squaring the results
gives
· ¸2 µ ¶2
T2 1 + kM1 2 M2
= (11.10)
T1 1 + kM2 2 M1

M = p1k
P = P
M <1
T
M =1
ne
li
re
su
es
Pr
nt

M >1
ta
ns
co

Fig. -11.2: The temperature entropy diagram for Rayleigh line

The Rayleigh line exhibits two possible maximums one for dT /ds = 0 and for
ds/dT = 0. The second maximum can be expressed as dT /ds = ∞. The second
220 CHAPTER 11. RAYLEIGH FLOW

law is used to find the expression for the derivative.

s1 − s2 T2 k − 1 P2
= ln − ln (11.11)
Cp T1 k P1

· ¸ · ¸
s1 − s2 1 + kM1 2 ) M2 k−1 1 + kM 212
= 2 ln ( + ln
Cp (1 + kM2 2 ) M1 k 1 + kM1 2
(11.12)
Let the initial condition M1 , and s1 be constant and the variable parameters are
M2 , and s2 . A derivative of equation (11.12) results in
1 ds 2(1 − M 2 )
= (11.13)
Cp dM M (1 + kM 2 )
Taking the derivative of equation (11.13) and letting the variable parameters be
T2 , and M2 results in
dT 1 − kM 2
= constant × 3 (11.14)
dM (1 + kM 2 )
Combining equations (11.13) and (11.14) by eliminating dM results in
dT M (1 − kM 2 )
= constant × (11.15)
ds (1 − M 2 )(1 + kM 2 )2
On T-s diagram a family of curves can be drawn for a given constant. Yet for every
curve, several observations can√ be generalized. The derivative is equal to zero
when 1 − kM 2 = 0 or M = 1/ k or when M → 0. The derivative is equal to infinity,
dT /ds = ∞ when M = 1. From thermodynamics, increase of heating results in
increase of entropy. And cooling results in reduction of entropy. Hence, when
cooling is applied to a tube the velocity decreases
√ and when heating is applied the
velocity increases. At peculiar point of M = 1/ k when additional heat is applied
the temperature decreases. The derivative is negative, dT /ds < 0, yet note this
point is not the choking point. The choking occurs only when M = 1 because it
violates the second law. The transition to supersonic flow occurs when the area
changes, somewhat similarly to Fanno flow. Yet, choking can be explained by the
fact that increase of energy must be accompanied by increase of entropy. But
the entropy of supersonic flow is lower (see Figure (11.2)) and therefore it is not
possible (the maximum entropy at M = 1.).
It is convenient to refer to the value of M = 1. These values are referred to
as the “star”1 values. The equation (11.5) can be written between choking point
and any point on the curve.

P∗ 1 + kM1 2
=
P1 1+k
(11.16)
1 The star is an asterisk.
11.3. RAYLEIGH FLOW TABLES 221

The temperature ratio is

µ ¶2
T∗ 1 1 + kM1 2
= 2
T1 M 1+k
(11.17)

∗ √
√U kRT ∗
ρ1 U∗ kRT ∗
= = √ (11.18)
ρ∗ U1 √ U1 kRT1
kRT1

or
r
ρ1 U∗ 1 T∗
∗
= =
ρ U1 M1 T1
(11.19)

¡ 2
¢
T0 1 T1 1 + k−1
2 M1
= ¡ ¢ (11.20)
T0 ∗ T ∗ 1+k
2

or explicitly
µ ¶
T0 1 2(1 + k)M1 2 k−1
= 1+ M1 2
T0 ∗ (1 + kM 2 )2 2
(11.21)

The stagnation pressure ratio reads

¡ 2
¢
P0 1 P1 1 + k−1
2 M1
= ¡ ¢ (11.22)
P0 ∗ P ∗ 1+k
2

or explicitly

µ ¶Ã ! k−1
k

P0 1 1+k 1 + kM1 2
=
P0 ∗ 1 + kM1 2 (1+k)
2
(11.23)

11.3 Rayleigh Flow Tables

The “star” values are tabulated in Table (11.1). Several observations can be made
in regards to the stagnation temperature.
222 CHAPTER 11. RAYLEIGH FLOW
Table -11.1: Rayleigh Flow k=1.4

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.03 0.00517 0.00431 2.397 1.267 0.00216
0.04 0.00917 0.00765 2.395 1.266 0.00383
0.05 0.014300 0.011922 2.392 1.266 0.00598
0.06 0.020529 0.017119 2.388 1.265 0.00860
0.07 0.027841 0.023223 2.384 1.264 0.011680
0.08 0.036212 0.030215 2.379 1.262 0.015224
0.09 0.045616 0.038075 2.373 1.261 0.019222
0.10 0.056020 0.046777 2.367 1.259 0.023669
0.20 0.20661 0.17355 2.273 1.235 0.090909
0.25 0.30440 0.25684 2.207 1.218 0.13793
0.30 0.40887 0.34686 2.131 1.199 0.19183
0.35 0.51413 0.43894 2.049 1.178 0.25096
0.40 0.61515 0.52903 1.961 1.157 0.31373
0.45 0.70804 0.61393 1.870 1.135 0.37865
0.50 0.79012 0.69136 1.778 1.114 0.44444
0.55 0.85987 0.75991 1.686 1.094 0.51001
0.60 0.91670 0.81892 1.596 1.075 0.57447
0.65 0.96081 0.86833 1.508 1.058 0.63713
0.70 0.99290 0.90850 1.423 1.043 0.69751
0.75 1.014 0.94009 1.343 1.030 0.75524
0.80 1.025 0.96395 1.266 1.019 0.81013
0.85 1.029 0.98097 1.193 1.011 0.86204
0.90 1.025 0.99207 1.125 1.005 0.91097
0.95 1.015 0.99814 1.060 1.001 0.95693
1.0 1.00 1.00 1.00 1.00 1.000
1.1 0.96031 0.99392 0.89087 1.005 1.078
1.2 0.91185 0.97872 0.79576 1.019 1.146
1.3 0.85917 0.95798 0.71301 1.044 1.205
1.4 0.80539 0.93425 0.64103 1.078 1.256
1.5 0.75250 0.90928 0.57831 1.122 1.301
1.6 0.70174 0.88419 0.52356 1.176 1.340
1.7 0.65377 0.85971 0.47562 1.240 1.375
1.8 0.60894 0.83628 0.43353 1.316 1.405
1.9 0.56734 0.81414 0.39643 1.403 1.431
2.0 0.52893 0.79339 0.36364 1.503 1.455
2.1 0.49356 0.77406 0.33454 1.616 1.475
2.2 0.46106 0.75613 0.30864 1.743 1.494
2.3 0.43122 0.73954 0.28551 1.886 1.510
2.4 0.40384 0.72421 0.26478 2.045 1.525
11.4. EXAMPLES FOR RAYLEIGH FLOW 223
Table -11.1: Rayleigh Flow k=1.4 (continue)

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
2.5 0.37870 0.71006 0.24615 2.222 1.538
2.6 0.35561 0.69700 0.22936 2.418 1.550
2.7 0.33439 0.68494 0.21417 2.634 1.561
2.8 0.31486 0.67380 0.20040 2.873 1.571
2.9 0.29687 0.66350 0.18788 3.136 1.580
3.0 0.28028 0.65398 0.17647 3.424 1.588
3.5 0.21419 0.61580 0.13223 5.328 1.620
4.0 0.16831 0.58909 0.10256 8.227 1.641
4.5 0.13540 0.56982 0.081772 12.50 1.656
5.0 0.11111 0.55556 0.066667 18.63 1.667
5.5 0.092719 0.54473 0.055363 27.21 1.675
6.0 0.078487 0.53633 0.046693 38.95 1.681
6.5 0.067263 0.52970 0.039900 54.68 1.686
7.0 0.058264 0.52438 0.034483 75.41 1.690
7.5 0.050943 0.52004 0.030094 1.0E+2 1.693
8.0 0.044910 0.51647 0.026490 1.4E+2 1.695
8.5 0.039883 0.51349 0.023495 1.8E+2 1.698
9.0 0.035650 0.51098 0.020979 2.3E+2 1.699
9.5 0.032053 0.50885 0.018846 3.0E+2 1.701
10.0 0.028972 0.50702 0.017021 3.8E+2 1.702
20.0 0.00732 0.49415 0.00428 1.1E+4 1.711
25.0 0.00469 0.49259 0.00274 3.2E+4 1.712
30.0 0.00326 0.49174 0.00190 8.0E+4 1.713
35.0 0.00240 0.49122 0.00140 1.7E+5 1.713
40.0 0.00184 0.49089 0.00107 3.4E+5 1.714
45.0 0.00145 0.49066 0.000846 6.0E+5 1.714
50.0 0.00117 0.49050 0.000686 1.0E+6 1.714
55.0 0.000971 0.49037 0.000567 1.6E+6 1.714
60.0 0.000816 0.49028 0.000476 2.5E+6 1.714
65.0 0.000695 0.49021 0.000406 3.8E+6 1.714
70.0 0.000600 0.49015 0.000350 5.5E+6 1.714

The data is presented in Figure (11.3).

11.4 Examples For Rayleigh Flow

Illustrative example
The typical questions that are raised in Rayleigh Flow are related to the max-
imum heat that can be transferred to gas (reaction heat) and to the flow rate.
224 CHAPTER 11. RAYLEIGH FLOW
Rayleigh Flow
k=14
4
*
T/T
*
T0 / T0
3 *
P/P
*
P0 / P0
∗
ρ/ρ
2
@( M = ∞ ) 1.7143

1
@( M = ∞ ) 0.4898

0
0.1 1 10
M

Fri May 6 11:11:44 2005

Fig. -11.3: The basic functions of Rayleigh Flow (k=1.4)

Example 11.1:
Air enters a pipe with pressure of 3[bar] and temperature of 27◦ C at Mach number
of M = 0.25. Due to internal combustion heat was released and the exit tempera-
ture was found to be 127◦ C. Calculate the exit Mach number, the exit pressure, the
total exit pressure, and heat released and transferred to the air. At what hamount
i of
kJ
energy the exit temperature will start to decrease? Assume CP = 1.004 kg◦ C

S OLUTION
The entrance Mach number and the exit temperature are given and from Table
(11.1) or from the program the initial ratio can be calculated. From the initial values
the ratio at the exit can be computed as the following.

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.25000 0.30440 0.25684 2.2069 1.2177 0.13793
11.4. EXAMPLES FOR RAYLEIGH FLOW 225

and
T2 T1 T2 400
= ∗ = 0.304 × = 0.4053
T∗ T T1 300

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.29831 0.40530 0.34376 2.1341 1.1992 0.18991

The exit Mach number is known, the exit pressure can be calculated as
P ∗ P2 1
P2 = P1 =3× × 2.1341 = 2.901[Bar]
P1 P ∗ 2.2069
For the entrance the stagnation values are

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.25000 0.98765 0.96942 2.4027 0.95745 2.3005 1.0424

The total exit pressure, P02 can be calculated as the following:

isentropic
z}|{
P 01 P0 ∗ P02 1 1
P02 = P1 ∗ =3× × × 1.1992 = 3.08572[Bar]
P1 P01 P0 0.95745 1.2177
The heat released (heat transferred) can be calculated from obtaining the stag-
nation temperature from both sides. The stagnation temperature at the entrance,
T01
isentropic
z}|{
T 01
T01 = T1 = 300/0.98765 = 303.75[K]
T1
The isentropic conditions at the exit are

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.29831 0.98251 0.95686 2.0454 0.94012 1.9229 0.90103

The exit stagnation temperature is

isentropic
z}|{
T 02
T02 = T2 = 400/0.98765 = 407.12[K]
T2
The heat released becomes
· ¸
Q kJ
= Cp (T02 − T01 ) 1 × 1.004 × (407.12 − 303.75) = 103.78
ṁ seckg ◦ C
The
√ maximum temperature occurs at the point where the Mach number reaches
1/ k and at this point the Rayleigh relationship are:
226 CHAPTER 11. RAYLEIGH FLOW

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.84515 1.0286 0.97959 1.2000 1.0116 0.85714

The maximum heat before the temperature can be calculated as following:

T ∗ Tmax 300
Tmax = T1 × 1.0286 = 1013.7[K]
T1 T ∗ 0.3044
The isentropic relationship at the maximum are

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.84515 0.87500 0.71618 1.0221 0.62666 0.64051 0.53376

The stagnation temperature for this point is

T0max 1013.7
T0max = Tmax ∗ = = 1158.51[K]
Tmax 0.875
The maximum heat can be calculated as
· ¸
Q kJ
= Cp (T0max − T01 ) = 1 × 1.004 × (1158.51 − 303.75) = 858.18
ṁ kgsecK
Note that this point isn’t the choking point.

Example 11.2:
Heat is added to the air until the flow is choked in amount of 600 [kJ/kg]. The
exit temperature is 1000 [K]. Calculate the entrance temperature and the entrance
Mach number.

S OLUTION
The solution involves finding the stagnation temperature at the exit and subtracting
the heat (heat equation) to obtain the entrance stagnation temperature. From the
Table (11.1) or from the Potto-GDC the following ratios can be obtained.

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
1.0000 0.83333 0.63394 1.0000 0.52828 0.52828 0.52828

The stagnation temperature

T02 1000
T02 = T2 = = 1200.0[K]
T2 0.83333
The entrance temperature is
T01 Q/ṁ 600 ∼
=1− = 1200 − = 0.5016
T02 T02 CP 1200 × 1.004
11.4. EXAMPLES FOR RAYLEIGH FLOW 227
T
It must be noted that T02 = T0 ∗ . Therefore with T001∗ = 0.5016 either by using Table
(11.1) or by Potto-GDC the following is obtained

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.34398 0.50160 0.42789 2.0589 1.1805 0.24362

Thus, entrance Mach number is 0.38454 and the entrance temperature can be
calculated as following

T1
T1 = T ∗ = 1000 × 0.58463 = 584.6[K]
T∗
The difference between the supersonic branch to subsonic branch

Example 11.3:
Air with Mach 3 enters a frictionless duct with heating. What is the maximum heat
that can be added so that there is no subsonic flow? If a shock occurs immediately
at the entrance, what is the maximum heat that can be added?

S OLUTION
To achieve maximum heat transfer the exit Mach number has to be one, M2 = 1.
µ ¶
Q T0
= Cp (T02 − T01 ) = Cp T0 ∗ 1 − 1∗
ṁ T0

The table for M = 3 as follows

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
3.0000 0.28028 0.65398 0.17647 3.4245 1.5882

The higher the entrance stagnation temperature the larger the heat amount that
can be absorbed by the flow. In subsonic branch the Mach number after the shock
is

Ty ρy Py P0y
Mx My Tx ρx Px P0 x
3.0000 0.47519 2.6790 3.8571 10.3333 0.32834

With Mach number of M = 0.47519 the maximum heat transfer requires information
for Rayleigh flow as the following

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.33138 0.47519 0.40469 2.0802 1.1857 0.22844
228 CHAPTER 11. RAYLEIGH FLOW

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.47519 0.75086 0.65398 1.8235 1.1244 0.41176

It also must be noticed that stagnation temperature remains constant across shock
wave. ¯ ³ ´
Q¯ T
ṁ ¯ 1 − T001∗ 1 − 0.65398
¯ subsonic = ³ ´ subsonic = =1
Q¯ T01 1 − 0.65398
ṁ ¯ 1 − T0 ∗
supersonic supersonic

It is not surprising for the shock wave to be found in the Rayleigh flow.

Fuel
injection

M1 = 0.3

P1 = 15[Bar]
T1 = 350[K]

Fig. -11.4: Schematic of the combustion chamber.

Example 11.4:
One of the reason that Rayleigh flow model was invented is to be analyzed the
flow in a combustion chamber. Consider a flow of air in conduct with a fuel injected
into the flow as shown in Figure 11.4. Calculate what the maximum fuel–air ratio.
Calculate the exit condition for half the fuel–air ratio. Assume that the mixture
properties are of air. Assume that the combustion heat is 25,000[KJ/kg fuel] for
the average temperature range for this mixture. Neglect the fuel mass addition and
assume that all the fuel is burned (neglect the complications of the increase of the
entropy if accrue).

S OLUTION
Under these assumptions the maximum fuel air ratio is obtained when the flow is
choked. The entranced condition can be obtained using Potto-GDC as following

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.30000 0.40887 0.34686 2.1314 1.1985 0.19183

The choking condition are obtained using also by Potto-GDC as

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
11.4. EXAMPLES FOR RAYLEIGH FLOW 229

And the isentropic relationships for Mach 0.3 are

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
0.30000 0.98232 0.95638 2.0351 0.93947 1.9119 0.89699

The maximum fuel-air can be obtained by finding the heat per unit mass.
µ ¶
Q̇ Q T01
= = Cp (T02 − T01 ) = Cp T1 1 − ∗
ṁ m T

Q̇
= 1.04 × 350/0.98232 × (1 − 0.34686) ∼ 242.022[kJ/kg]
ṁ
The fuel–air mass ratio has to be
mf uel needed heat 242.022
= = ∼ 0.0097[kg f uel/kg air]
mair combustion heat 25, 000

If only half of the fuel is supplied then the exit temperature is

Q 0.5 × 242.022
T02 = + T01 = + 350/0.98232 ∼ 472.656[K]
mCp 1.04

The exit Mach number can be determined from the exit stagnation temperature as
following:
T2 T01 T02
= ∗
T∗ T0 T01
The last temperature ratio can be calculated from the value of the temperatures
T2 472.656
= 0.34686 × ∼ 0.47685
T∗ 350/0.98232

The Mach number can be obtained from a Rayleigh table or using Potto-GDC

T T0 P P0 ρ∗
M T∗ T0 ∗ P∗ P0 ∗ ρ
0.33217 0.47685 0.40614 2.0789 1.1854 0.22938

It should be noted that this example is only to demonstrate how to carry the calcu-
lations.
End Solution
230 CHAPTER 11. RAYLEIGH FLOW
 CHAPTER 12
Evacuating and Filling a Semi Rigid
Chambers

Fanno model

In some ways the next two Chapters for relatively short tube

contain materials is new to the tradi- V = f (t)

tional compressible flow text books1 . It Isothermal model

was the undersigned experience, that for relatively long tube

in traditional classes for with compress- Volume forced models

External forces that control the tank volume
ible flow (sometimes referred to as gas
dynamics) don’t provide a demonstra- Fanno model
for relativly short tube
tion to applicability of the class mate-
= ( )
rial aside to aeronautical spectrum even V f P; et

such as turbomachinery. In this Chap- Isothermal model

for relativly long tube
ter a discussion on application of com- Volume is a function of pressure or rigid
pressible flow to other fields like manu- (the volume can be also a function of inertia and etc)

facturing is presented2 . Semi rigid tank

There is a significant impor- Fig. -12.1: The two different classifications of

tance to the “pure” models such Isother- models that explain the filling or
mal flow and Fanno flow which have im- evacuating of a single chamber
mediate applicability. However, in many
instances, the situations, in life, are far
1 After completion of these Chapters, the undersigned discover two text books which to include some

material related to this topic. These books are OCR, J. A., Fundamentals of Gas Dynamics, Interna-
tional Textbook Co., Scranton, Pennsylvania, 1964. and “Compressible Fluid Flow,” 2nd Edition, by M.
A. Saad, Prentice Hall, 1985. However, these books contained only limit discussions on the evacuation
of chamber with attached nozzle.
2 Even if the instructor feels that their students are convinced about the importance of the compress-

ible, this example can further strength and enhance this conviction.

231
232 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS

more complicate. Combination of gas compressibility in the chamber and flow out
or through a tube post a special interest and these next two Chapters are deal-
ing with these topics. In the first Chapter models, were the chamber volume is
controlled or a function of the pressure, are discussed. In the second Chapter,
models, were the chamber’s volume is a function of external forces, are presented
(see Figure (12.1)).

12.1 Governing Equations and Assumptions

The process of filing or evacuating a semi flexible (semi rigid) chamber through a
tube is very common in engineering. For example, most car today equipped with
an airbag. For instance, the models in this Chapter are suitable for study of the
filling the airbag or filling bicycle with air. The analysis is extended to include a
semi rigid tank. The term semi rigid tank referred to a tank that the volume is either
completely rigid or is a function of the chamber’s pressure.
As it was shown in this book the most appropriate model for the flow in
the tube for a relatively fast situation is Fanno Flow. The Isothermal model is more
appropriate for cases where the tube is relatively long in–which a significant heat
transfer occurs keeping the temperature almost constant. As it was shown in Chap-
ter (10) the resistance, 4fDL , should be larger than 400. Yet Isothermal flow model
is used as the limiting case.

fanno model fanno model

reduced for relatively short tube for relatively short tube
connection 1 2 1 2
V = f (P; et ) V = f (P; et )

Isothermal model Isothermal model

for a relatively long tube for a relatively long tube

The connection is through a narrow passage A schematic of a direct connection

Fig. -12.2: A schematic of two possible connections of the tube to a single chamber

The Rayleigh flow model requires

that a constant heat transfer supplied either = ( )
m
V f P; t

out

by chemical reactions or otherwise. This

author isn’t familiar with situations in which
Control volume for the evacuating case
Rayleigh flow model is applicable. And
therefore, at this stage, no discussion is of-
fered here. = ( ) m V f P; t
in

Fanno flow model is the most ap-

propriate in the case where the filling and
evacuating is relatively fast. In case the Control volume for the filling case
4f L
filling is relatively slow (long D than the
Fig. -12.3: A schematic of the control vol-
Isothermal flow is appropriate model. Yet
umes used in this model
as it was stated before, here Isothermal flow
and Fanno flow are used as limiting or bounding cases for the real flow. Addition-
12.1. GOVERNING EQUATIONS AND ASSUMPTIONS 233

ally, the process in the chamber can be limited or bounded between two limits of
Isentropic process or Isothermal process.
In this analysis, in order to obtain the essence of the process, some simpli-
fied assumptions are made. The assumptions can be relaxed or removed and the
model will be more general. Of course, the payment is by far more complex model
that sometime clutter the physics. First, a model based on Fanno flow model is
constructed. Second, model is studied in which the flow in the tube is isothermal.
The flow in the tube in many cases is somewhere between the Fanno flow model
to Isothermal flow model. This reality is an additional reason for the construction
of two models in which they can be compared.
Effects such as chemical reactions (or condensation/evaporation) are ne-
glected. There are two suggested itself possibilities to the connection between the
tube to the tank (see the Figure 12.2): one) direct two) through a reduction. The
direct connection is when the tube is connect straight to tank like in a case where
pipe is welded into the tank. The reduction is typical when a ball is filled trough
an one–way valve (filling a baseball ball, also in manufacturing processes). The
second possibility leads itself to an additional parameter that is independent of the
resistance. The first kind connection tied the resistance, 4fDL , with the tube area.
The simplest model for gas inside the chamber as a first approximation is
the isotropic model. It is assumed that kinetic change in the chamber is negligible.
Therefore, the pressure in the chamber is equal to the stagnation pressure, P ≈ P0
(see Figure (12.4)). Thus, the stagnation pressure at the tube’s entrance is the
same as the pressure in the chamber.

The mass in the chamber and

mass flow out are expressed in
terms of the chamber variables P  P0
(see Figure 12.3. The mass in P1  P
the tank for perfect gas reads
U 0 1 2
dm
− ṁout = 0 (12.1)
dt Fig. -12.4: The pressure assumptions in the chamber
and tube entrance
And for perfect gas the mass at
any given time is

P (t)V (t)
m= (12.2)
RT (t)

The mass flow out is a function of the resistance in tube, 4fDL and the pressure
difference between the two sides of the tube ṁout ( 4fDL , P1 /P2 ). The initial condi-
tions in the chamber are T (0), P (0) and etc. If the mass occupied in the tube is
234 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS

neglected (only for filling process) the most general equation ideal gas (12.1) reads
ṁout
z }| {
m U
zµ }| ¶{ z }| {
d PV 4f L P2
± ρ1 A c1 M1 ( , )=0 (12.3)
dt RT D P1
When the plus sign is for filling process and the negative sign is for evacuating
process.

12.2 General Model and Non-dimensioned

It is convenient to non-dimensioned the properties in chamber by dividing them by
their initial conditions. The dimensionless properties of chamber as
T (t = t̄)
T̄ = (12.4a)
T (t = 0)
V (t = t̄)
V̄ = (12.4b)
V (t = 0)
P (t = t̄)
P̄ = (12.4c)
P (t = 0)
t
t̄ = (12.4d)
tc
where tc is the characteristic time of the system defined as followed
V (0)
tc = p (12.5)
AMmax kRT (0))
The physical meaning of characteristic time, tc is the time that will take to evacuate
the chamber if the gas in the chamber was in its initial state, the flow rate was at
its maximum (choking flow), and the gas was incompressible in the chamber.
Utilizing these definitions (12.4) and substituting into equation (12.3) yields
ρ c(t)
z }| { z
µ ¶ q }| {
P (0)V (0) d P̄ V̄ P̄1 P (0)
± A kRT̄1 T (0)Mmax M̄ (t̄) = 0 (12.6)
tc RT (0) dt̄ T̄ RT̄1 T (0)
where the following definition for the reduced Mach number is added as
M1 (t)
M̄ = (12.7)
Mmax
After some rearranging equation (12.6) obtains the form
µ ¶ p
d P̄ V̄ tc AMmax kRT (0) P¯1 M̄1
± p M̄ = 0 (12.8)
dt̄ T̄ V (0) T¯1
12.2. GENERAL MODEL AND NON-DIMENSIONED 235

and utilizing the definition of characteristic time, equation (12.5), and substituting
into equation (12.8) yields
µ ¶
d P̄ V̄ P¯1 M̄
± p =0 (12.9)
dt̄ T̄ T¯1

Note that equation (12.9) can be modified by introducing additional parameter

which referred to as external time, tmax 3 . For cases, where the process time is
important parameter equation (12.9) transformed to

µ ¶
d P̄ V̄ tmax P¯1 M̄
± p =0 (12.10)
dt̃ T̄ tc T¯1

when P̄ , V̄ , T̄ , and M̄ are all are function of t̃ in this case. And where t̃ = t/tmax .
It is more convenient to deal with the stagnation pressure then the actual pressure
at the entrance to the tube. Utilizing the equations developed in Chapter 5 be-
tween the stagnation condition, denoted without subscript, and condition in a tube
¯
denoted with subscript 1. The ratio of √P1¯ is substituted by
T1

· ¸ −(k+1)
P¯1 P̄ k − 1 2 2(k−1)
p =√ 1+ M (12.11)
T¯1 T̄ 2

It is convenient to denote
· ¸ −(k+1)
k−1 2 2(k−1)
f [M ] = 1 + M (12.12)
2

Note that f [M ] is a function of the time. Utilizing the definitions (12.11) and sub-
stituting equation (12.12) into equation (12.9) to be transformed into
µ ¶
d P̄ V̄ P̄ M̄ (t̄)f [M ]
± √ =0 (12.13)
dt̄ T̄ T̄

Equation (12.13) is a first order nonlinear differential equation that can be solved
for different initial conditions. At this stage, the author isn’t aware that there is a
general solution for this equation4 . Nevertheless, many numerical methods are
available to solve this equation.

3 This notation is used in many industrial processes where time of process referred to sometime as

the maximum time.

4 To those mathematically included, find the general solution for this equation.
236 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS

12.2.1 Isentropic Process

The relationship between the pressure and the temperature in the chamber can be
approximated as isotropic and therefore
· ¸ k−1
T (t) P (t) k k−1
T̄ = = = P̄ k (12.14)
T (0) P (0)

The ratios can be expressed in term of the reduced pressure as followed:

P̄ P̄ 1
= k−1 = P̄ k (12.15)
T̄ P̄ k
and
P̄ k+1
√ = P̄ 2k (12.16)
T̄

So equation (12.13) is simplified into three different forms:

d³ 1
´ k+1
V̄ P̄ k ± P̄ 2k M̄ (t̄)f [M ] = 0 (12.17a)
dt̄

1 1−k dP̄ 1 dV̄ k+1

P̄ k V̄ + P̄ k ± P̄ 2k M̄ (t̄)f [M ] = 0 (12.17b)
k dt̄ dt̄

dP̄ dV̄ 3k−1

V̄ + k P̄ ± k P̄ 2k M̄ (t̄)f [M ] = 0 (12.17c)
dt̄ dt̄
Equation (12.17) is a general equation for evacuating or filling for isentropic pro-
cess in the chamber. It should be point out that, in this stage, the model in the tube
could be either Fanno flow or Isothermal flow. The situations where the chamber
undergoes isentropic process but the flow in the tube is Isothermal are limited.
Nevertheless, the application of this model provide some kind of a limit where to
expect when some heat transfer occurs. Note the temperature in the tube entrance
can be above or below the surrounding temperature. Simplified calculations of the
entrance Mach number are described in the advance topics section.

12.2.2 Isothermal Process in The Chamber

12.2.3 A Note on the Entrance Mach number
The value of Mach number, M1 is a function of the resistance, 4fDL and the ratio
of pressure in the tank to the back pressure, PB /P1 . The exit pressure, P2 is
12.3. RIGID TANK WITH NOZZLE 237

different from PB in some situations. As it was shown before, once the flow became
choked the Mach number, M1 is only a function of the resistance, 4fDL . These
statements are correct for both Fanno flow and the Isothermal flow models. The
method outlined in Chapters ?? and 10 is appropriate for solving for entrance Mach
number, M1 .
Two equations must be solved for the Mach numbers at the duct entrance and exit
when the flow is in a chokeless condition. These equations are combinations of
the momentum and energy equations in terms of the Mach numbers. The charac-
teristic equations for Fanno flow (10.50), are
· ¯ ¸ · ¯ ¸
4f L 4f L ¯¯ 4f L ¯¯
= − (12.18)
D D ¯max 1 D ¯max 2
and
v
· ¸ 1−k u" # k−1
k+1

M1 u
k
k−1 2
P2 k−1 2 t 1+ 2 M2
= 1+ M2 k−1 2 (12.19)
P0 (t) 2 M2 1+ 2 M1

where 4fDL is defined by equation (10.49).

The solution of equations (12.18) and (12.19) for given 4fDL and P
P0 (t) yields the
exit

entrance and exit Mach numbers. See advance topic about approximate solution
for large resistance, 4fDL or small entrance Mach number, M1 .

12.3 Rigid Tank with Nozzle

he most simplest possible combination is discussed here before going trough the
more complex cases A chamber is filled or evacuated by a nozzle. The gas in the
chamber assumed to go an isentropic processes and flow is bounded in nozzle
between isentropic flow and isothermal flow5 . Here, it also will be assumed that
the flow in the nozzle is either adiabatic or isothermal.

12.3.1 Adiabatic Isentropic Nozzle Attached

The mass flow out is given by either by Fliegner’s equation (5.46) or simply use
cM ρA∗ and equation (12.17) becomes
1 1−k dP̄ k+1
P̄ k ± P̄ 2k (t̄)f [M ] = 0 (12.20)
k dt̄
It was utilized that V̄ = 1 and M̄ definition is simplified as M̄ = 1. It can be noticed
that the characteristic time defined in equation (12.5) reduced into:
V (0)
tc = p (12.21)
A kRT (0))
5 This work is suggested by Donald Katze the point out that this issue appeared in Shapiro’s Book

Vol 1, Chapter 4, p. 111 as a question 4.31.

238 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS

Also it can be noticed that equation (12.12) simplified into

· ¸ −(k+1) µ ¶ −(k+1)
k − 1 2 2(k−1) k + 1 2(k−1)
f [M ] = 1 + 1 = (12.22)
2 2

Equation (12.20) can be simplified as

1 ³ 1−k ´
P 2k dP ± f [m]dt̄ = 0 (12.23)
k
Equation (12.23) can be integrated as
Z P̄ Z t
1−k
P 2k dP ± dt = 0 (12.24)
1 0

The integration limits are obtained by simply using the definitions of reduced pres-
sure, at P (t̄ = 0) = 1 and P (t̄ = t̄) = P̄ . After the integration, equation (12.24) and
rearrangement becomes
· µ ¶ ¸ 1−k
2k
k−1
P̄ = 1 ± f [M ]t̄ (12.25)
2

Example 12.1:
A chamber is connected to a main line with pressure line with a diaphragm and
nozzle. The initial pressure at the chamber is 1.5[Bar] and the volume is 1.0[m3 ].
Calculate time it requires that the pressure to reach 5[Bar] for two different noz-
zles throat area of 0.001, and 0.1 [m2 ] when diaphragm is erupted. Assumed the
stagnation temperature at the main line is the ambient of 27[◦ C].

S OLUTION
The characteristic time is
V V 1.0
tmax = = ∗ = √ = 0.028[sec] (12.26)
A∗ c A c 0.1 1.4 × 287 × 300
And for smaller area
1.0
tmax = √ = 2.8[sec]
0.001 1.4 × 287 × 300

P (t) 4.5
P̄ = = = 3.0
P (0) 1.5
The time is
h 1−k i µ k + 1 ¶−()
t = tmax P̄ k − 1 (12.27)
2
12.4. RAPID EVACUATING OF A RIGID TANK 239

Substituting values into equation (12.27) results

h 1−1.4
i µ 2.4 ¶ −2.4
0.8
t = 0.028 3 2.8 −1 = 0.013[sec] (12.28)
2

Filling/Evacuating The Chamber Under Upchucked Condition

The flow in the nozzle can became upchucked and it can be analytically solved.
Owczarek [1964] found an analytical solution which described here.

12.3.2 Isothermal Nozzle Attached

In this case the process in nozzle is assumed to isothermal but the process in the
chamber is isentropic. The temperature in the nozzle is changing because the
temperature in the chamber is changing. Yet, the differential temperature change
in the chamber is slower than the temperature change in nozzle. For rigid volume,
V̄ = 1 and for isothermal nozzle T̄ = 1 Thus, equation (12.13) is reduced into

dP̄
= ±f [M ]P̄ = 0 (12.29)
dt̄
Separating the variables and rearranging equation (12.29) converted into
Z P̄ Z t̄
dP̄
± f [M ] dt̄ = 0 (12.30)
1 P̄ 0

Here, f [M ] is expressed by equation (12.22). After the integration, equation

(12.30) transformed into
µ ¶ −(k+1)
k+1 2(k−1)
ln P̄ = t̄
2
" #
−(k+1)
( k+1
2 )
2(k−1) t̄

P̄ = e (12.31)

12.4 Rapid evacuating of a rigid tank

12.4.1 With Fanno Flow
The relative Volume, V̄ (t) = 1, is constant and equal one for a completely rigid
tank. In such case, the general equation (12.17) “shrinks” and doesn’t contain the
relative volume term.
240 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS

A reasonable model for the tank is isentropic (can be replaced polytropic relation-
ship) and Fanno flow are assumed for the flow in the tube. Thus, the specific
governing equation is

dP̄ 3k−1
− k M̄ f [M ]P̄ 2k = 0 (12.32)
dt̄
For a choked flow the entrance Mach number to the tube is at its maximum, Mmax
and therefore M̄ = 1. The solution of equation (12.32) is obtained by noticing that
M̄ is not a function of time and by variables separation results in
Z t̄ Z P̄ Z P̄
dP̄ 1 1−3k
dt̄ = 3k−1 = P̄ 2k dP̄ (12.33)
0 1 k M̄ f [M ]P̄ 2k k M̄ f [M ] 1

direct integration of equation (12.33) results in

2 h 1−k i
t̄ = P̄ 2k − 1 (12.34)
(k − 1)M̄ f [M ]

It has to be realized that this is “reversed” function i.e. t̄ is a function of P and can
be reversed for case. But for the chocked case it appears as
· ¸ 2k
(k − 1)M̄ f [M ] 1−k
P̄ = 1 + t̄ (12.35)
2

The function is drawn as shown here in Figure (12.5). The Figure (12.5) shows
that when the modified reduced pressure equal to one the reduced time is zero.
The reduced time increases with decrease of the pressure in the tank.
At certain point the flow becomes chokeless flow (unless the back pressure is
complete vacuum). The transition point is denoted here as chT . Thus, equation
(12.34) has to include the entrance Mach under the integration sign as
Z P̄
1 1−3k
t̄ − t̄chT = P̄ 2k dP̄ (12.36)
PchT k M̄ f [M ]

For practical purposes if the flow is choked for more than 30% of the charecteristic
time the choking equation can be used for the whole range, unless extra long
time or extra low pressure is calculated/needed. Further, when the flow became
chokeless the entrance Mach number does not change much from the choking
condition.
Again, for the special cases where the choked equation is not applicable the inte-
gration has to be separated into zones: choked and chokeless flow regions. And
in the choke region the calculations can use the choking formula and numerical
calculations for the rest.
12.4. RAPID EVACUATING OF A RIGID TANK 241
1.0

P(t) V(t) = P (t)

0.8

V(t) = P (0)

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1.0

t¯

Fig. -12.5: The reduced time as a function of the modified reduced pressure

Example 12.2:
A chamber with volume of 0.1[m3 ] is filled with air at pressure of 10[Bar]. The
chamber is connected with a rubber tube with f = 0.025, d = 0.01[m] and length of
L = 5.0[m]

S OLUTION
4f L 4f L
The first parameter that calculated is D D =5

12.4.2 Filling Process

The governing equation is

dP̄ 3k−1
− k M̄ f [M ]P̄ 2k = 0 (12.37)
dt̄
For a choked flow the entrance Mach number to the tube is at its maximum, Mmax
and therefore M̄ = 1. The solution of equation (12.37) is obtained by noticing that
M̄ is not a function of time and by variable separation results in
Z t̄ Z P̄ Z P̄
dP̄ 1 1−3k
dt̄ = 3k−1 = P̄ 2k dP̄ (12.38)
0 1 k M̄ f [M ]P̄ 2k k M̄ f [M ] 1

direct integration of equation (12.38) results in

2 h 1−k i
t̄ = P̄ 2k − 1 (12.39)
(k − 1)M̄ f [M ]
242 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS

It has to be realized that this is a reversed function. Nevertheless, with today

computer this should not be a problem and easily can be drawn as shown here in
Figure (12.5). The Figure shows that when the modified reduced pressure equal

1
V(t) = P(t)
V(t) = V(0)

0.8

t

0.6

0.4

0.2

0 0.2 0.4 0.6 0.8 1

k
( f M ℄ P or
1) [ max k
( + 1)(3k 1) [f M ℄ P
max
2 k
2 2

Fig. -12.6: The reduced time as a function of the modified reduced pressure

to one the reduced time is zero. The reduced time increases with decrease of the
pressure in the tank.
At some point the flow becomes chokeless flow (unless the back pressure is a
complete vacuum). The transition point is denoted here as chT . Thus, equation
(12.39) has to include the entrance Mach under the integration sign as
Z P̄
1 1−3k
t̄ − t̄chT = P̄ 2k dP̄ (12.40)
PchT k M̄ f [M ]

12.4.3 The Isothermal Process

For Isothermal process, the relative temperature, T̄ = 1. The combination of the
isentropic tank and Isothermal flow in√the tube is different from Fanno flow in that
the chocking condition occurs at 1/ k. This model is reasonably appropriated
when the chamber is insulated and not flat while the tube is relatively long and the
process is relatively long.
It has to be remembered that the chamber can undergo isothermal process. For
12.4. RAPID EVACUATING OF A RIGID TANK 243

the double isothermal (chamber and tube) the equation (12.6) reduced into
ρ
¡ ¢ z }| { z c(0)}| {
P (0)V (0) d P̄ V̄ P̄1 P (0) p
± A kRT (0)Mmax M̄ (t̄) = 0 (12.41)
tc RT (0) dt̄ R T (0)

12.4.4 Simple Semi Rigid Chamber

A simple relation of semi rigid chamber when the volume of the chamber is linearly
related to the pressure as

V (t) = aP (t) (12.42)

where a is a constant that represent the physics. This situation occurs at least
in small ranges for airbag balloon etc. The physical explanation when it occurs
beyond the scope of this book. Nevertheless, a general solution is easily can be
obtained similarly to rigid tank. Substituting equation (12.42) into yields
d ³ 1+k ´ k+1
P̄ k − P̄ 2k M̄ f [M ] = 0 (12.43)
dt̄

Carrying differentiation result in

1 + k 1 dP̄ k+1
P̄ k − P̄ 2k M̄ f [M ] = 0 (12.44)
k dt̄

Similarly as before, the variables are separated as

Z t̄ Z P̄ k−1
k P̄ 2k dP̄
dt = (12.45)
0 1+k 1 M̄ f [M ]

The equation (12.45) integrated to obtain the form

2k 2 h 3k−1
i
t̄ = 1 − P̄ 2k (12.46)
M̄ f [M ](3k − 1)(1 + k)

The physical meaning that the pressure remains larger thorough evacuating pro-
cess, as results in faster reduction of the gas from the chamber.

12.4.5 The “Simple” General Case

The relationship between the pressure and the volume from the physical point of
view must be monotonous. Further, the relation must be also positive, increase of
the pressure results in increase of the volume (as results of Hook’s law. After all, in
244 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS

the known situations to this author pressure increase results in volume decrease
(at least for ideal gas.).
In this analysis and previous analysis the initial effect of the chamber container
inertia is neglected. The analysis is based only on the mass conservation and if
unsteady effects are required more terms (physical quantities) have taken into ac-
count. Further, it is assumed the ideal gas applied to the gas and this assumption
isn’t relaxed here.
Any continuous positive monotonic function can be expressed into a polynomial
function. However, as first approximation and simplified approach can be done by
a single term with a different power as
V (t) = aP n (12.47)
When n can be any positive value including zero, 0. The physical meaning of n = 0
is that the tank is rigid. In reality the value of n lays between zero to one. When
n is approaching to zero the chamber is approaches to a rigid tank and vis versa
when the n → 1 the chamber is flexible like a balloon.
There isn’t a real critical value to n. Yet, it is convenient for engineers to further
study the point where the relationship between the reduced time and the reduced
pressure are linear6 Value of n above it will Convex and below it concave.

d ³ 1+nk−k ´ k+1
P̄ k − P̄ 2k M̄ f [M ] = 0 (12.48)
dt̄

Notice that when n = 1 equation (12.49) reduced to equation (12.43).

After carrying–out differentiation results
1 + nk − k 1+nk−2k dP̄ k+1
P̄ k − P̄ 2k M̄ f [M ] = 0 (12.49)
k dt̄

Again, similarly as before, variables are separated and integrated as follows

Z t̄ Z 1+2nk−5k
1 + nk − k P̄ P̄ 2k dP̄
dt = (12.50)
0 k 1 M̄ f [M ]
Carrying–out the integration for the initial part if exit results in
2k 2 h 3k−2nk−1
i
t̄ = 1 − P̄ 2k (12.51)
M̄ f [M ](3k − 2nk − 1)(1 + k)
The linear condition are obtain when
3k − 2
3k − 2nk − 1 = 1 −→ n = (12.52)
2k
That is just bellow 1 (n = 0.785714286) for k = 1.4.
6 Some suggested this border point as infinite evocation to infinite time for evacuation etc. This

undersigned is not aware situation where this indeed play important role. Therefore, it is waited to find
such conditions before calling it as critical condition.
12.5. ADVANCE TOPICS 245

12.5 Advance Topics

The term 4fDL is very large for small values of the entrance Mach number which
requires keeping many digits in the calculation. For small values of the Mach num-
bers, equation (12.18) can be approximated as

4f L 1 Mexit 2 − Min 2
= (12.53)
D k Mexit 2 Min 2

and equation (12.19) as

Pexit Min
= . (12.54)
P0 (t) Mexit

The solution of two equations (12.53) and (12.54) yields

v
u h i2
u
u1 − P exit
P0 (t)
Min = t 4f L
. (12.55)
k D

This solution should used only for Min < 0.00286; otherwise equations (12.18)
and (12.19) must be solved numerically.
The solution of equation (12.18) and (12.19) is described in “Pressure die casting:
a model of vacuum pumping” Bar-Meir, G; Eckert, E R G; Goldstein, R. J. Journal
of Manufacturing Science and Engineering (USA). Vol. 118, no. 2, pp. 259-265.
May 1996.
246 CHAPTER 12. EVACUATING SEMIRIGID CHAMBERS
 CHAPTER 13
Evacuating/Filing Chambers under
External Volume Control

This chapter is the second on the section dealing with filling and evacuating cham-
bers. Here the model deals with the case where the volume is controlled by exter-
nal forces. This kind of model is applicable to many manufacturing processes such
as die casting, extraction etc. In general the process of the displacing the gas (in
many cases air) with a liquid is a very common process. For example, in die cast-
ing process liquid metal is injected to a cavity and after the cooling/solidification
period a part is obtained in near the final shape. One can also view the exhaust
systems of internal combustion engine in the same manner. In these processes,
sometime is vital to obtain a proper evacuation of the gas (air) from the cavity.

13.1 General Model

In this analysis, in order to obtain the essence of the process, some simplified
assumptions are made. It simplest model of such process is when a piston is
displacing the gas though a long tube. It assumed that no chemical reaction (or
condensation/evaporation) occur in the piston or the tube 1 . It is further assumed
that the process is relatively fast. The last assumption is a appropriate assumption
in process such as die casting.
Two extreme possibilities again suggest themselves: rapid and slow pro-
cesses. The two different connections, direct and through reduced area are com-
bined in this analysis.
1 such reaction are possible and expected to be part of process but the complicates the analysis and

do not contribute to understand to the compressibility effects.

247
248 CHAPTER 13. Evacuating under External Volume Control

13.1.1 Rapid Process

Clearly under the assumption of rapid process the heat transfer can be neglected
and Fanno flow can be assumed for the tube. The first approximation isotropic
process describe the process inside the cylinder (see Figure (13.1)).

isontropic process 1 2
V = f (t)
Fanno model

Fig. -13.1: The control volume of the “Cylinder”.

Before introducing the steps of the analysis, it is noteworthy to think about

the process in qualitative terms. The replacing incompressible liquid enter in the
same amount as replaced incompressible liquid. But in a compressible substance
the situation can be totally different, it is possible to obtain a situation where that
most of the liquid entered the chamber and yet most of the replaced gas can be
still be in the chamber. Obtaining conditions where the volume of displacing liquid
is equal to the displaced liquid are called the critical conditions. These critical con-
ditions are very significant that they provide guidelines for the design of processes.
Obviously, the best ventilation is achieved with a large tube or area. In
manufacture processes to minimize cost and the secondary machining such as
trimming and other issues the exit area or tube has to be narrow as possible. In the
exhaust system cost of large exhaust valve increase with the size and in addition
reduces the strength with the size of valve2 . For these reasons the optimum size
is desired. The conflicting requirements suggest an optimum area, which is also
indicated by experimental studies and utilized by practiced engineers.
The purpose of this analysis to yields a formula for critical/optimum vent
area in a simple form is one of the objectives of this section. The second objective
is to provide a tool to “combine” the actual tube with the resistance in the tube,
thus, eliminating the need for calculations of the gas flow in the tube to minimize
the numerical calculations.
A linear function is the simplest model that decibels changes the volume.
In reality, in some situations like die casting this description is appropriate. Nev-
ertheless, this model can be extended numerical in cases where more complex
function is applied.
· ¸
t
V (t) = V (0) 1 − (13.1)
tmax
2 After certain sizes, the possibility of crack increases.
13.1. GENERAL MODEL 249

Equation (13.1) can be non–dimensionlassed as

V̄ (t̄) = 1 − t̄ (13.2)

The governing equation (12.10) that was developed in the previous Chap-
ter (12) obtained the form as
½ ¾
£ ¤ k1 1 V̄ dP̄ dV̄ tmax M̄ f (M ) £ ¤ k+1
P̄ + + P̄ 2k = 0 (13.3)
k P̄ dt dt̄ tc

where t̄ = t/tmax . Notice that in this case that there are two different char-
acteristic times: the “characteristic” time, tc and the “maximum” time, tmax . The
first characteristic time, tc is associated with the ratio of the volume and the tube
characteristics (see equation (12.5)). The second characteristic time, tmax is as-
sociated with the imposed time on the system (in this case the elapsed time of the
piston stroke).
Equation (13.3) is an nonlinear first order differential equation and can be
rearranged as follows
dP̄ dt̄
³ ´ = ; P̄ (0) = 1. (13.4)
k 1− tmax k−1
1 − t̄
tc M̄ f [M ]P̄ P̄
2k

Equation (13.4) is can be solved only when the flow is chocked In which case f [m]
isn’t function of the time.
The solution of equation (13.4)) can be obtained by transforming and by
k−1 2k
introducing a new variable ξ = P̄ 2k and therefore P̄ = [ξ] k−1 . The reduced Pres-
sure derivative, dP̄ = 2k [ξ]( k−1 ) dξ Utilizing this definition and there implication
2k
−1
k−1
reduce equation (13.4)

2 [ξ]( k−1 )
2k
−1
dξ dt̄
2k = (13.5)
(k − 1) (1 − Bξ) [ξ] k−1 1 − t̄
tmax
where B = tc M̄ f [M ] And equation (13.5) can be further simplified as

2dξ dt̄
= (13.6)
(k − 1) (1 − Bξ) ξ 1 − t̄
Equation (13.6) can be integrated to obtain
¯ ¯
2 ¯ 1 − Bξ ¯
ln ¯ ¯ = − ln t̄ (13.7)
(k − 1)B ¯ ξ ¯
or in a different form
¯ ¯ 2
¯ 1 − Bξ ¯ (1−k)B
¯ ¯ = t̄ (13.8)
¯ ξ ¯
250 CHAPTER 13. Evacuating under External Volume Control

Now substituting to the “preferred” variable

" # 2 ¯1
tmax k−1 t
(1−k) max M̄ f [M ]
¯
1− tc M̄ f [M ]P̄
2k tc ¯
¯ = t̄ (13.9)
k−1 ¯
P̄ 2k ¯
P̄

The analytical solution is applicable only in the case which the flow is choked
thorough all the process. The solution is applicable to indirect connection. This
happen when vacuum is applied outside the tube (a technique used in die casting
and injection molding to improve quality by reducing porosity.). In case when the
flow chokeless a numerical integration needed to be performed. In the literature,
to create a direct function equation (13.4) is transformed into
³ k−1
´
tmax
dP̄ k 1 − tc M̄ f [M ] P̄ 2k

= (13.10)
dt̄ 1 − t̄
with the initial condition of

P (0) = 1 (13.11)

The analytical solution also can be approximated by a simpler equation as

tmax
P̄ = [1 − t] tc
(13.12)

The results for numerical evaluation in the case when cylinder is initially at an
atmospheric pressure and outside tube is also at atmospheric pressure are pre-
sented in Figure (13.2). In this case only some part of the flow is choked (the later
part). The results of a choked case are presented in Figure (13.3) in which outside
tube condition is in vacuum. These Figures (13.2) and 13.3 demonstrate the im-
portance of the ratio of tmax
tc . When tc
tmax
> 1 the pressure increases significantly
and verse versa. Thus, the question remains how the time ratio can be transfered
to parameters that can the engineer can design in the system.
Denoting the area that creates the ratio tmax
tc = 1 as the critical area, Ac
provides the needed tool. Thus the exit area, A can be expressed as

A
A= Ac (13.13)
Ac
¯
tmax ¯
The actual times ratio tc ¯ can be expressed as
@A

1
¯ ¯ z }|¯ {
tmax ¯¯ tmax ¯¯ tmax ¯¯
¯ = (13.14)
tc @A tc @A tc ¯@Ac
¯
13.1. GENERAL MODEL 251

According to equation (12.5) tc is inversely proportional to area, tc ∝ 1/A. Thus,

equation (13.14) the tmax is canceled and reduced into
¯
tmax ¯¯ A
¯ = (13.15)
tc @A Ac

Parameters influencing the process are the area ratio, AAc , and the friction
parameter, 4fDL . From other detailed calculations the author thesis (later to be
published on this www.potto.org). it was found that the influence of the parameter
4f L
D on the pressure development in the cylinder is quite small. The influence
is small on the residual air mass in the cylinder but larger on the Mach number,
Mexit . The effects of the area ratio, AAc , are studied here since it is the dominant
parameter.
It is important to point out the significance of the tmax
tc . This parameter rep-
resents the ratio between the filling time and the evacuating time, the time which
would be required to evacuate the cylinder for constant mass flow rate at the max-
imum Mach number when the gas temperature and pressure remain in their initial
values. This parameter also represents the dimensionless area, AAc , according to
the following equation
Figure (13.4) describes the pressure as a function of the dimensionless
time for various values of AAc . The line that represents AAc = 1 is almost straight.
For large values of AAc the pressure increases the volume flow rate of the air until
a quasi steady state is reached. This quasi steady state is achieved when the vol-
umetric air flow rate out is equal to the volume pushed by the piston. The pressure
and the mass flow rate are maintained constant after this state is reached. The
pressure in this quasi steady state is a function of AAc . For small values of AAc there
is no steady state stage. When AAc is greater than one the pressure is concave up-
ward and when AAc is less than one the pressure is concave downward as shown
in Figures (13.4), which was obtained by an integration of equation (13.9).

13.1.2 Examples
Example 13.1:
Calculate the minimum required vent area for die casting process when the die vol-
ume is 0.001[m3 ] and 4fDL = 20. The required solidification time, tmax = 0.03[sec].

S OLUTION

13.1.3 Direct Connection

In the above analysis is applicable to indirect connection. It should be noted that
critical area, Ac , is not function of the time. The direct connection posts more math-
ematical difficulty because the critical area is not constant and time dependent.
252 CHAPTER 13. Evacuating under External Volume Control

To continue

13.2 Summary
The analysis indicates there is a critical vent area below which the ventilation is
poor and above which the resistance to air flow is minimal. This critical area de-
pends on the geometry and the filling time. The critical area also provides a mean
to “combine” the actual vent area with the vent resistance for numerical simulations
of the cavity filling, taking into account the compressibility of the gas flow.
13.2. SUMMARY 253

5.0
Dimensionless ∗ `
4.6 Area, A/Ac
∇
∗ `
4.2 0.0 ∗
0.2 ` ∇
∗ `
3.8 0.5 ∇
∇
1.0 ∗ `
3.4 1.2 a ∇
`
h 2.0 o
∗
`
∇
P (t )
P (0)
3.0 5.0 g ∗
`
∇
∗ ∇
4fL ` ∇
2.6 = 100.0 ∗
D ` ∇
∗ ` ∇
∗ ` ∇
2.2 ∗ ` ∇
∇
∗ ` ∇
∗ ` ∇
∗ ` ∇
1.8 ∇
∗ ` ` ∇ ∇
∗
∗ ` ` ∇ ∇ ∇
a a a a a a
1.4 `∗ ∇
`∗ ∇
∗ ∇
` ∇ ∇
a a a a a
a a a a a
∗ `
` ∇ a a a a a a a
∗
` `∗ ∇
∇ a a a a a a
`∗a ∇
∗a ∇
`o a a o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
a o
`∗a ∇
g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g
a o o
1.0 `o
∇ `∗a ∇
∗a ∇
o o o o

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Figure a

5.0
Dimensionless ∗ `
4.6 Area, A/Ac
∇
∗ `
4.2 0.0 ∗
0.2 ` ∇
∗ `
3.8 0.5 ∇
` ∇
1.0 ∗
3.4 1.2 a
` ∇
h 2.0 o
∗
` ∇
P (t )
P (0)
3.0 5.0 g ∗
∇
∗ `
∇
4fL = 5.0 `
2.6 ∗ ∇
D ` ∇
∗
` ∇
∗ ` ∇
2.2 ∗ ` ∇
∗ ` ∇
` ∇
∗ ∇
1.8 ∗ ` ∇
∗ ` ∇ ∇
∗ ` ` ∇
∗ ∇ ∇
∗ ` ` ∇ ∇
1.4 ` ∇
`∗ ∇ ∇ a a a a a a a
`∗ ∇ a a a a a a
`∗ ∇ a a a a a a
∗ ∗ ∇
` a a a a a a a
` ∇
`∗a ∇
∇ a a o a a a
∗
`
∇ a
g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g
a
`∗a oa o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
1.0 `o
∇ `∗a ∇
∗a ∇
o o o o

0.0 0.1 0.2 0.3 0.4 0.5 0.6

h 0.7
Dimensionless Time, t,
0.8 0.9 1.0

or, Cylinder Volume Fraction

Figure b

Fig. -13.2: The pressure ratio as a function of the dimensionless time for chokeless condition
254 CHAPTER 13. Evacuating under External Volume Control

2.2

∗
.
2 .. ∗
.. .
. ∗
..
..
.. ∗
1.8 .. .
.. ∗
.
.. ∗
.. ∗
..
1.6 .. .
.. . ∗ hhh
A
= 0.0
. ∗ Ac
.. . ∗ 0.1 . . . . . .. . . . .
.. . ∗
..
. ∗
1.4 .. . ∗ 0.5 ∗
. ∗
... ∗
... ∗ 1.0
..
. ∗
.. ∗
1.2
..
..
∗ ∗∗ 1.5 ×

∗∗
.. .
. ... ∗ ∗ 4.0 a
... ∗ ∗
P(t) .
1 ∗a×. . ∗
× ×
P(0) × ×
a × ×
× ×
× ×
0.8 a × ×
× ×
a × ×
× ×
a × ×
×
× ×
0.6 a ×
×
a ×
×
×
a ×
×
a ×
×
×
0.4 a ×
a ×
×
a ×
a
a
a
0.2 a a
a a
a a
a a a
a a a
a a a a a
0 a a a a a a

0 0.1 0.2 0.3 0.4

h 0.5 0.6 0.7 0.8
DIMENSIONLESS TIME, t, or, CYLINDER VOLUME FRACTION
0.9 1

Fig. -13.3: The pressure ratio as a function of the dimensionless time for choked condition

5.0
Dimensionless ∗ `
4.6 Area, A/Ac
∇
∗ `
4.2 0.0 ∗
0.2 ` ∇
∗ `
3.8 0.5 ∇
` ∇
1.0 ∗
3.4 1.2 a
` ∇
h 2.0 o
∗
` ∇
P (t )
P (0) 3.0 5.0 g ∗
∇
∗
`
∇
4fL = 5.0 `
2.6 ∗ ∇
D ` ∇
∗
` ∇
∗ ` ∇
2.2 ∗ ` ∇
∗ ` ∇
` ∇
∗ ∇
1.8 ∗ ` ∇
∇
∗ ` `
∗ ∇ ∇
∗ ` ` ∇ ∇ ∇
∗
∗ ` ` ∇ a a a a a a a
1.4
∗ ∗ ` ∇ ∇ ∇
` a a a a a a
a a a a a a
` ∇ ∇ a a a a a
`∗ ∇
`∗ ∇ a a a a
∗
`
∇ `∗a ∇
∇ a a a a o
g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g g
`∗a oa o o oa o o o o o o o o o o o o o o o o o o o o o o o o o o o
1.0 `o
∇ `∗a ∇
∗a ∇
o o o o o

0.0 0.1 0.2 0.3 0.4 0.5 0.6

h 0.7
Dimensionless Time, t,
0.8 0.9 1.0

or, Cylinder Volume Fraction

Fig. -13.4: The pressure ratio as a function of the dimensionless time

 CHAPTER 14

Oblique Shock

14.1 Preface to Oblique Shock

In Chapter (6), discussion on a normal
shock was presented. A normal shock
is a special type of shock wave. The U1 U2
other type of shock wave is the oblique
shock. In the literature oblique shock,
normal shock, and Prandtl–Meyer func- 

tion are presented as three separate Æ = 0

and different issues. However, one can

view all these cases as three different Fig. -14.1: A view of a straight normal shock
regions of a flow over a plate with a de- as a limited case for oblique shock
flection section. Clearly, variation of the
deflection angle from a zero (δ = 0) to a
positive value results in oblique shock. Further changing the deflection angle to a
negative value results in expansion waves. The common representation is done
by not showing the boundaries of these models. However, this section attempts to
show the boundaries and the limits or connections of these models1 .

1 In this chapter, even the whole book, a very limited discussion about reflection shocks and collisions

of weak shock, Von Neumann paradox, triple shock intersection, etc are presented. The author believes
that these issues are not relevant to most engineering students and practices. Furthermore, these
issues should not be introduced in introductory textbook of compressible flow. Those who would like
to obtain more information, should refer to J.B. Keller, “Rays, waves and asymptotics,” Bull. Am. Math.
Soc. 84, 727 (1978), and E.G. Tabak and R.R. Rosales, “Focusing of weak shock waves and the Von
Neuman paradox of oblique shock reflection,” Phys. Fluids 6, 1874 (1994).

255
256 CHAPTER 14. OBLIQUE SHOCK

14.2 Introduction
14.2.1 Introduction to Oblique Shock
A normal shock occurs when there is a disturbance downstream which imposes
a boundary condition on the flow in which the fluid/gas can react only by a sharp
change in the flow direction. As it may be recalled, normal shock occurs when a
wall is straight/flat (δ = 0) as shown in Figure (14.1) which occurs when somewhere
downstream a disturbance2 appears. When the deflection angle is increased, the
gas flow must match the boundary conditions. This matching can occur only when
there is a discontinuity in the flow field. Thus, the direction of the flow is changed
by a shock wave with an angle to the flow. This shock is commonly referred to
as the oblique shock. Alternatively, as discussed in Chapter (1)3 the flow behaves
as it does in a hyperbolic field. In such a case, the flow field is governed by a
hyperbolic equation which deals with the case when information (like boundary
conditions) reaches from downstream only if they are within the range of influence.
For information such as the disturbance (boundary condition) to reach deep into
the flow from the side requires time. During this time, the flow moves downstream
and creates an angle.

14.2.2 Introduction to Prandtl–Meyer Function

Decreasing the deflection angle re- 0◦
sults in the same effects as before. ν∞(k)
Prandtl
Meyer
Oblique
θmax(k)
Function
The boundary conditions must match Shock

the geometry. Yet, for a negative de-

flection angle (in this section’s nota- No Shock

tion), the flow must be continuous. zone

The analysis shows that the flow ve- Fig. -14.2: The regions where oblique shock or
locity must increase to achieve this re- Prandtl–Meyer function exist. Notice
quirement. This velocity increase is that both have a maximum point and
referred to as the expansion wave. As a “no solution” zone, which is around
it will be shown in the next chapter, zero. However, Prandtl-Meyer func-
as opposed to oblique shock analy- tion approaches closer to a zero de-
sis, the increase in the upstream Mach flection angle.
number determines the downstream Mach number and the “negative” deflection
angle.
It has to be pointed out that both the oblique shock and the Prandtl–Meyer
function have a maximum point for M1 → ∞. However, the maximum point for the
Prandtl–Meyer function is much larger than the oblique shock by a factor of more
than 2. What accounts for the larger maximum point is the effective turning (less
entropy production) which will be explained in the next chapter (see Figure (14.2)).
2 Zero velocity, pressure boundary conditions, and different inclination angle, are examples of forces

that create shock. The zero velocity can be found in a jet flowing into a still medium of gas.
3 This section is under construction and does not appear in the book yet.
14.3. OBLIQUE SHOCK 257

14.2.3 Introduction to Zero Inclination

What happens when the inclination angle is zero? Which model is correct to use?
Can these two conflicting models, the oblique shock and the Prandtl–Meyer func-
tion, co-exist? Or perhaps a different model better describes the physics. In some
books and in the famous NACA report 1135 it was assumed that Mach wave and
oblique shock co–occur in the same zone. Previously (see Chapter 6), it was as-
sumed that normal shock occurs at the same time. In this chapter, the stability
issue will be examined in greater detail.

14.3 Oblique Shock

The shock occurs in real- θ−δ
ity in situations where the U1 n
shock has three–dimensional U1 t
θ U2
effects. The three–dimensional U1 U2 = U 1 t t
U2 n
ine

effects of the shock make it

on L

appear as a curved plane.

arsi

However, for a chosen arbi-

Comp


trary accuracy it requires a
Æ
specific small area, a one–
dimensional shock can be
considered. In such a case, Fig. -14.3: A typical oblique shock schematic
the change of the orientation
makes the shock considerations two–dimensional. Alternately, using an infinite
(or a two–dimensional) object produces a two–dimensional shock. The two–
dimensional effects occur when the flow is affected from the “side,” i.e., a change
in the flow direction4 .
To match the boundary conditions, the flow turns after the shock to be parallel
to the inclination angle. Figure (14.3) exhibits the schematic of the oblique shock.
The deflection angle, δ, is the direction of the flow after the shock (parallel to the
wall). The normal shock analysis dictates that after the shock, the flow is always
subsonic. The total flow after the oblique shock can also be supersonic, which
depends on the boundary layer.
Only the oblique shock’s normal component undergoes the “shock.” The tan-
gent component does not change because it does not “move” across the shock
line. Hence, the mass balance reads

ρ1 U1 n = ρ2 U2 n (14.1)

4 The author begs for forgiveness from those who view this description as offensive (There was an

unpleasant email to the author accusing him of revolt against the holy of the holies.). If you do not
like this description, please just ignore it. You can use the traditional explanation, you do not need the
author’s permission.
258 CHAPTER 14. OBLIQUE SHOCK

The momentum equation reads

P1 + ρ1 U1 n 2 = P2 + ρ2 U2 n 2 (14.2)

The momentum equation in the tangential direction yields

U1t = U2t (14.3)

The energy balance reads

U1 n 2 U2 2
Cp T1 + = Cp T2 + n (14.4)
2 2
Equations (14.1), (14.2), and (14.4) are the same as the equations for normal
shock with the exception that the total velocity is replaced by the perpendicular
components. Yet the new relationship between the upstream Mach number, the
deflection angle, δ, and the Mach angle, θ has to be solved. From the geometry it
can be observed that
U1 n
tan θ = (14.5)
U1 t

and
U2n
tan(θ − δ) = (14.6)
U2t

Unlike in the normal shock, here there are three possible pairs5 of solutions to
these equations. The first is referred to as the weak shock; the second is the strong
shock; and the third is an impossible solution (thermodynamically)6 . Experiments
and experience have shown that the common solution is the weak shock, in which
the shock turns to a lesser extent7 .
tan θ U1 n
= (14.7)
tan(θ − δ) U2 n

The above velocity–geometry equations can also be expressed in term of Mach

number, as

M1 n
sin θ = (14.8)
M1
5 Thisissue is due to R. Menikoff, who raised the solution completeness issue.
6 The solution requires solving the entropy conservation equation. The author is not aware of “simple”
proof and a call to find a simple proof is needed.
7 Actually this term is used from historical reasons. The lesser extent angle is the unstable angle and

the weak angle is the middle solution. But because the literature referred to only two roots, the term
lesser extent is used.
14.3. OBLIQUE SHOCK 259

M2n
sin(θ − δ) = (14.9)
M2

M1 t
cos θ = (14.10)
M1

M2 t
cos(θ − δ) = (14.11)
M2

The total energy across an oblique shock wave is constant, and it follows
that the total speed of sound is constant across the (oblique) shock. It should
be noted that although, U1 t = U2 t the Mach number is M1 t 6= M2 t because the
temperatures on both sides of the shock are different, T1 6= T2 .
As opposed to the normal shock, here angles (the second dimension) have to
be determined. The solution from this set of four equations, (14.8) through (14.11),
is a function of four unknowns of M1 , M2 , θ, and δ. Rearranging this set utilizing
geometrical identities such as sin α = 2 sin α cos α results in
· ¸
M1 2 sin2 θ − 1
tan δ = 2 cot θ (14.12)
M1 2 (k + cos 2θ) + 2

The relationship between the properties can be determined by substituting

M1 sin θ for of M1 into the normal shock relationship, which results in

P2 2kM1 2 sin2 θ − (k − 1)
=
P1 k+1
(14.13)

The density and normal velocity ratio can be determined by the following equation

ρ2 U1 n (k + 1)M1 2 sin2 θ
= =
ρ1 U2 n (k − 1)M1 2 sin2 θ + 2
(14.14)

The temperature ratio is expressed as

£ ¤
T2 2kM1 2 sin2 θ − (k − 1) (k − 1)M1 2 + 2
=
T1 (k + 1)2 M1
(14.15)
260 CHAPTER 14. OBLIQUE SHOCK

Prandtl’s relation for oblique shock is

k−1 2
Un1 Un2 = c2 − Ut (14.16)
k+1
The Rankine–Hugoniot relations are the same as the relationship for the normal
shock
P2 − P1 P2 − P1
=k (14.17)
ρ2 − ρ1 ρ2 − ρ1

14.4 Solution of Mach Angle

Oblique shock, if orientated to a coordinate perpendicular and parallel shock plane
is like a normal shock. Thus, the relationship between the properties can be de-
termined by using the normal components or by utilizing the normal shock table
developed earlier. One has to be careful to use the normal components of the
Mach numbers. The stagnation temperature contains the total velocity.
Again, the normal shock is a one–dimensional problem, thus, only one pa-
rameter is required (to solve the problem). Oblique shock is a two–dimensional
problem and two properties must be provided so a solution can be found. Prob-
ably, the most useful properties are upstream Mach number, M1 and the deflec-
tion angle, which create a somewhat complicated mathematical procedure, and
this will be discussed later. Other combinations of properties provide a relatively
simple mathematical treatment, and the solutions of selected pairs and selected
relationships will be presented.

14.4.1 Upstream Mach Number, M1 , and Deflection Angle, δ

Again, this set of parameters is, perhaps, the most common and natural to ex-
amine. Thompson (1950) has shown that the relationship of the shock angle is
obtained from the following cubic equation:

x3 + a1 x2 + a2 x + a3 = 0 (14.18)
where
x = sin2 θ (14.19)
and
M1 2 + 2
a1 = − − k sin2 δ (14.20)
M1 2
· ¸
2M1 2 + 1 (k + 1)2 k−1
a2 = − + + sin2 δ (14.21)
M1 4 4 M1 2
cos2 δ
a3 = − (14.22)
M1 4
14.4. SOLUTION OF MACH ANGLE 261

Equation (14.18) requires that x has to be a real and positive number to

obtain a real deflection angle8 . Clearly, sin θ must be positive, and the negative
sign refers to the mirror image of the solution. Thus, the negative root of sin θ must
be disregarded
The solution of a cubic equation such as (14.18) provides three roots9 . These
roots can be expressed as

1
x1 = − a1 + (S + T ) (14.23)
3

1 1 1 √
x2 = − a1 − (S + T ) + i 3(S − T ) (14.24)
3 2 2

and

1 1 1 √
x3 = − a1 − (S + T ) − i 3(S − T ) (14.25)
3 2 2

Where
q
3 √
S= R+ D, (14.26)

q
3 √
T = R− D (14.27)

and where the definition of the D is

D = Q3 + R2 (14.28)

and where the definitions of Q and R are

3a2 − a1 2
Q= (14.29)
9
and
9a1 a2 − 27a3 − 2a1 3
R= (14.30)
54
Only three roots can exist for the Mach angle, θ. From a mathematical point of
view, if D > 0, one root is real and two roots are complex. For the case D = 0, all
8 This point was pointed out by R. Menikoff. He also suggested that θ is bounded by sin−1 1/M
1
and 1.
9 The highest power of the equation (only with integer numbers) is the number of the roots. For

example, in a quadratic equation there are two roots.

262 CHAPTER 14. OBLIQUE SHOCK

the roots are real and at least two are identical. In the last case where D < 0, all
the roots are real and unequal.
The physical meaning of the above analysis demonstrates that in the range
where D > 0 no solution can exist because no imaginary solution can exist10 . D >
0 occurs when no shock angle can be found, so that the shock normal component
is reduced to subsonic and yet parallel to the inclination angle.
Furthermore, only in some cases when D = 0 does the solution have a
physical meaning. Hence, the solution in the case of D = 0 has to be examined in
the light of other issues to determine the validity of the solution.
When D < 0, the three unique roots are reduced to two roots at least for the
steady state because thermodynamics dictates11 that. Physically, it can be shown
that the first solution(14.23), referred sometimes as a thermodynamically unstable
root, which is also related to a decrease in entropy, is “unrealistic.” Therefore, the
first solution does not occur in reality, at least, in steady–state situations. This root
has only a mathematical meaning for steady–state analysis12 .
These two roots represent two different situations. First, for the second root,
the shock wave keeps the flow almost all the time as a supersonic flow and it is
referred to as the weak solution (there is a small section that the flow is subsonic).
Second, the third root always turns the flow into subsonic and it is referred to as
the strong solution. It should be noted that this case is where entropy increases in
the largest amount.
In summary, if a hand moves the shock angle starting from the deflection an-
gle and reaching the first angle that satisfies the boundary condition, this situation
is unstable and the shock angle will jump to the second angle (root). If an addi-
tional “push” is given, for example, by additional boundary conditions, the shock
angle will jump to the third root13 . These two angles of the strong and weak shock
are stable for a two–dimensional wedge (see the appendix of this chapter for a
limited discussion on the stability14 ).

10 A call for suggestions, to explain about complex numbers and imaginary numbers should be in-
cluded. Maybe insert an example where imaginary solution results in no physical solution.
11 This situation is somewhat similar to a cubical body rotation. The cubical body has three symmet-

rical axes which the body can rotate around. However, the body will freely rotate only around two axes
with small and large moments of inertia. The body rotation is unstable around the middle axes. The
reader can simply try it.
12 There is no experimental or analytical evidence, that the author has found, showing that it is to-

tally impossible. The “unstable” terms can be thermodynamically stable in unsteady case. Though,
those who are dealing with rapid transient situations should be aware that this angle of oblique shock
can exist. There is no theoretical evidence that showing that in strong unsteady state this angle is
unstable. The shock will initially for a very brief time transient in it and will jump from this angle to the
thermodynamically stable angles.
13 See the discussion on the stability. There are those who view this question not as a stability equa-

tion but rather as under what conditions a strong or a weak shock will prevail.
14 This material is extra and not recommended for standard undergraduate students.
14.4. SOLUTION OF MACH ANGLE 263

14.4.2 When No Oblique Shock Exist or When D > 0

Large deflection angle for given, M1

The first range is when the deflection angle

reaches above the maximum point. For a
given upstream Mach number, M1 , a change
in the inclination angle requires a larger en-
ergy to change the flow direction. Once, the
inclination angle reaches the “maximum po-
tential energy,” a change in the flow direction
is no longer possible. In the alternative view,
the fluid “sees” the disturbance (in this case,
the wedge) in front of it and hence the normal Fig. -14.4: Flow around spherically
blunted 30◦ cone-cylinder
shock occurs. Only when the fluid is away from
with Mach number 2.0. It
the object (smaller angle) liquid “sees” the ob- can be noticed that the
ject in a different inclination angle. This differ- normal shock, the strong
ent inclination angle is sometimes referred to shock, and the weak shock
as an imaginary angle. coexist.

The simple procedure For example, in Figure (14.4) and (14.5), the imaginary
angle is shown. The flow is far away from the object and does not “see’ the object.
For example, for, M1 −→ ∞ the maximum deflection angle is calculated when
D = Q3 + R2 = 0. This can be done by evaluating the terms a1 , a2 , and a3 for
M1 = ∞.
a1 = −1 − k sin2 δ
2
(k + 1) sin2 δ
a2 =
4
a3 = 0
With these values the coefficients R and Q are
³ 2 2
´
9(−)(1 + k sin2 δ) (k+1)4 sin δ − (2)(−)(1 + k sin2 δ)2
R=
54
and
(1 + k sin2 δ)2
Q=
9
Solving equation (14.28) after substituting these values of Q and R provides series
of roots from which only one root is possible. This root, in the case k = 1.4, is just
above δmax ∼ π4 (note that the maximum is also a function of the heat ratio, k).
While the above procedure provides the general solution for the three roots,
there is simplified transformation that provides solution for the strong and and weak
264 CHAPTER 14. OBLIQUE SHOCK

The fluid doesn’t ’’see’

the object

M∞

} The fluid ‘‘sees’’

the object with
"imaginary" inclination
angle

} Intermediate zone

}
The fluid "sees"
the object infront

Fig. -14.5: The view of a large inclination angle from different points in the fluid field.

solution. It must be noted that in doing this transformation the first solution is “lost”
supposedly because it is “negative.” In reality the first solution is not negative
but rather some value between zero and the weak angle. Several researchers15
suggested that instead Thompson’s equation should be expressed by equation
(14.18) by tan θ and is transformed into
µ ¶ µ ¶
k−1 ¡ ¢ k+1
1+ M1 2 tan δ tan3 θ − M1 2 − 1 tan2 θ + 1 + tan δ tan θ + 1 = 0
2 2
(14.31)

The solution to this equation (14.31) for the weak angle is

 ³ −1
´
M1 2 − 1 + 2f1 (M1 , δ) cos 4π+cos 3(f2 (M1 ,δ))
θweak = tan−1  ¡ 2
¢  (14.32)
3 1 + k−1 2 M1 tan δ

15 A whole discussion on the history of this can be found in “Open content approach to academic

writing” on http://www.potto.org/obliqueArticle.php
14.4. SOLUTION OF MACH ANGLE 265

³ −1 ´
M1 2 − 1 + 2f1 (M1 , δ) cos cos (f23(M1 ,δ))
θstrong = tan−1 ¡ 2
¢ (14.33)
3 1 + k−12 M1 tan δ

where these additional functions are

s µ ¶µ ¶
¡ 2 ¢2 k−1 2 k+1 2
f1 (M1 , δ) = M1 − 1 − 3 1 + M1 1+ M1 tan2 δ (14.34)
2 2

and
¡ ¢3 ¡ ¢¡ ¢
M1 2 − 1 −9 1+ k−1
2 M1
2
1 + k−1 2
2 M1 +
k+1
2 M1
4
tan2 δ
f2 (M1 , δ) =
f1 (M1 , δ)3
(14.35)

Figure (14.6) typical results for oblique shock for two deflection angle of 5
and 25 degree. Generally, the strong shock is reduced as increase of the Mach
number while the weak shock is increase. The impossible shock for unsteady state
is almost linear function of the upstream Mach number and almost not affected by
the deflection angle.

50

45 My s δ = 5◦ k = 1.2
My w δ = 5◦
40
My i δ = 5◦
35 My s δ = 25◦
My w δ = 25◦
30 My i δ = 25◦
My

25

20

15

10

10 20 30 40 50
Mx

November 28, 2007

Fig. -14.6: The three different Mach numbers after the oblique shock for two deflection an-
gles

The Procedure for Calculating The Maximum Deflection Point

266 CHAPTER 14. OBLIQUE SHOCK

The maximum is obtained when D = 0. When the right terms defined in

(14.20)-(14.21), (14.29), and (14.30) are substituted into this equation and utilizing
the trigonometrical sin2 δ + cos2 δ = 1 and other trigonometrical identities results in
Maximum Deflection Mach Number’s equation in which is
M1 2 (k + 1) (M1n 2 + 1) = 2(kM1n 4 + 2M1n 2 − 1) (14.36)
This equation and its twin equation can be obtained by an alternative procedure
proposed by someone16 who suggested another way to approach this issue. It
can be noticed that in equation (14.12), the deflection angle is a function of the
Mach angle and the upstream Mach number, M1 . Thus, one can conclude that the
maximum Mach angle is only a function of the upstream Much number, M1 . This
can be shown mathematically by the argument that differentiating equation (14.12)
and equating the results to zero creates relationship between the Mach number,
M1 and the maximum Mach angle, θ. Since in that equation there appears only the
heat ratio k, and Mach number, M1 , θmax is a function of only these parameters.
The differentiation of the equation (14.12) yields
³ ´ ³ ´
4 4 (k+1) 2 2 2 (k+1) 2
d tan δ kM 1 sin θ + 2 − 2 M1 M 1 sin θ − 1 + 2 M1
= h i (14.37)
dθ 4 4
kM sin θ − (k − 1) + (k+1)2 M1 2 2 2
M sin θ − 1
1 4 1

Because tan is a monotonous function, the maximum appears when θ has its
maximum. The numerator of equation (14.37) is zero at different values of the
denominator. Thus, it is sufficient to equate the numerator to zero to obtain the
maximum. The nominator produces a quadratic equation for sin2 θ and only the
positive value for sin2 θ is applied here. Thus, the sin2 θ is
r h
2 2
¡ k+1 ¢4 i
k+1
−1 + 4 M1 + (k + 1) 1 + k−1 2 M1 + 2 M1
sin2 θmax = (14.38)
kM1 2
Equation (14.38) should be referred to as the maximum’s equation. It should be
noted that both the Maximum Mach Deflection equation and the maximum’s equa-
tion lead to the same conclusion that the maximum M1n is only a function of up-
stream the Mach number and the heat ratio k. It can be noticed that the Maximum
Deflection Mach Number’s equation is also a quadratic equation for M1n 2 . Once
M1n is found, then the Mach angle can be easily calculated by equation (14.8).
To compare these two equations the simple case of Maximum for an infinite Mach
number is examined. It must be pointed out that similar procedures can also be
proposed (even though it does not appear in the literature). Instead, taking the
derivative with respect to θ, a derivative can be taken with respect to M1 . Thus,
d tan δ
=0 (14.39)
dM1
16 At first, it was seen as C. J.Chapman, English mathematician to be the creator but later an earlier

version by several months was proposed by Bernard Grossman. At this stage it is not clear who was
the first to propose it.
14.4. SOLUTION OF MACH ANGLE 267

and then solving equation (14.39) provides a solution for Mmax .

A simplified case of the Maximum Deflection Mach Number’s equation for
large Mach number becomes
r
k+1
M1n = M1 for M1 >> 1 (14.40)
2k
q
Hence, for large Mach numbers, the Mach angle is sin θ = k+1 2k (for k=1.4), which
makes θ = 1.18 or θ = 67.79◦ .
With the value of θ utilizing equation (14.12), the maximum deflection angle
can be computed. Note that this procedure does not require an approximation of
M1n to be made. The general solution of equation (14.36) is
rq
2
(k + 1) M1 4 + 8 (k 2 − 1) M1 2 + 16 (k + 1) + (k + 1) M1 2 − 4
M1n = √
2 k
(14.41)

Note that Maximum Deflection Mach Number’s equation can be extended to deal
with more complicated equations of state (aside from the perfect gas model).
This typical example is for those who like mathematics.

Example 14.1:
Derive the perturbation of Maximum Deflection Mach Number’s equation for the
case of a very small upstream Mach number number of the form M1 = 1 + ². Hint,
Start with equation (14.36) and neglect all the terms that are relatively small.

S OLUTION
The solution can be done by substituting (M1 = 1 + ²) into equation (14.36) and it
results in
sp
²(k) + ²2 + 2 ² − 3 + k²2 + 2 k² + k
M1n =
4k
(14.42)

where the epsilon function is

²(k) =(k 2 + 2k + 1) ²4 + (4 k 2 + 8 k + 4) ²3 +
2
(14 k 2 + 12 k − 2) ²2 + (20 k 2 + 8 k − 12) ² + 9 (k + 1) (14.43)

Now neglecting all the terms with ² results for the epsilon function in
2
²(k) ∼ 9 (k + 1) (14.44)
268 CHAPTER 14. OBLIQUE SHOCK

And the total operation results in

r
3 (k + 1) − 3 + k
M1n = =1 (14.45)
4k
Interesting to point out that as a consequence of this assumption the maximum
shock angle, θ is a normal shock. However, taking the second term results in
different value. Taking the second term in the explanation results in
vq
u
u 2
t 9 (k + 1) + (20 k 2 + 8 k − 12) ² − 3 + k + 2 (1 + k)²
M1n = (14.46)
4k

Note this equation (14.46) produce an un realistic value and additional terms are
required to obtained to produce a realistic value.

The case of D ≥ 0 or 0 ≥ δ

The second range in which D > 0 is when δ < 0. Thus, first the transition line
in which D = 0 has to be determined. This can be achieved by the standard
mathematical procedure of equating D = 0. The analysis shows regardless of the
value of the upstream Mach number D = 0 when δ = 0. This can be partially
demonstrated by evaluating the terms a1 , a2 , and a3 for the specific value of M1 as
following

M1 2 + 2
a1 =
M1 2
2M1 2 + 1
a2 = −
M1 4
1
a3 = − (14.47)
M1 4

With values presented in equations (14.47) for R and Q becoming

³ ´³ ´ ³ ´ ³ ´2
M1 2 +2 2M1 2 +1 −1 M1 2 +2
9 M1 2 M1 4
− 27 M1 4
−2 M1 2
R=
54
¡ ¢¡ ¢ ¡ ¢2
9 M1 2 + 2 2M1 2 + 1 + 27M1 2 − 2M1 2 M1 2 + 2
= (14.48)
54M1 6

and
³ ´ ³ ´3
2M1 2 +1 M1 2 +2
3 M1 4
− M1 2
Q= (14.49)
9
14.4. SOLUTION OF MACH ANGLE 269

Substituting the values of Q and R equations (14.48) (14.49) into equation (14.28)
provides the equation to be solved for δ.
 ³ ´ ³ 2 ´3 3
2M1 2 +1
 3 M1 4 − MM1 +2
1
2

  +
9
" ¡ ¢¡ ¢ ¡ ¢2 #2
9 M1 2 + 2 2M1 2 + 1 + 27M1 2 − 2M1 2 M1 2 + 2
= 0 (14.50)
54M1 6

The author is not aware of any analytical demonstration in the literature which
shows that the solution is identical to zero for δ = 017 . Nevertheless, this identity
can be demonstrated by checking several points for example, M1 = 1., 2.0, ∞.
Table (14.7) is provided for the following demonstration. Substitution of all the
above values into (14.28) results in D = 0.
Utilizing the symmetry and antisymmetry of the qualities of the cos and sin for
δ < 0 demonstrates that D > 0 regardless of Mach number. Hence, the physical
interpretation of this fact is that either no shock exists and the flow is without any
discontinuity or that a normal shock exists18 . Note that, in the previous case, with a
positive large deflection angle, there was a transition from one kind of discontinuity
to another.

In the range where δ ≤ 0, the

question is whether it is possi- XX
XXcoefficients
XXX
ble for an oblique shock to ex- a a2 a3
M1 XXX 1
ist? The answer according to
this analysis and stability anal- 1.0 -3 -1 - 32
ysis is no. And according to
this analysis, no Mach wave 9
2.0 3 0 16
can be generated from the wall
with zero deflection. In other 1
∞ -1 0 - 16
words, the wall does not emit
any signal to the flow (assum-
ing zero viscosity), which con- Fig. -14.7: The various coefficients of three different
tradicts the common approach. Mach numbers to demonstrate that D is
Nevertheless, in the literature, zero
there are several papers suggesting zero strength Mach wave; others suggest a
singular point19 . The question of singular point or zero Mach wave strength are
only of mathematical interest.
17 A mathematical challenge for those who like to work it out.
18 There are several papers that attempt to prove this point in the past. Once this analytical solu-
tion was published, this proof became trivial. But for non ideal gas (real gas) this solution is only an
indication.
19 See for example, paper by Rosles, Tabak, “Caustics of weak shock waves,” 206 Phys. Fluids 10 (1)

, January 1998.
270 CHAPTER 14. OBLIQUE SHOCK
Suppose that there is a Mach wave at
the wall at zero inclination (see Fig-
ure (14.8)). Obviously, another Mach
wave occurs after a small distance.
But because the velocity after a Mach
wave (even for an extremely weak shock µ1 µ2 µ3 µ∞
wave) is reduced, thus, the Mach an-
gle will be larger (µ2 > µ1 ). If the sit-
Fig. -14.8: The Mach waves that are sup-
uation keeps on occurring over a finite
posed to be generated at zero in-
distance, there will be a point where the clination.
Mach number will be 1 and a normal shock will occur, according the common ex-
planation. However, the reality is that no continuous Mach wave can occur because
of the viscosity (boundary layer).
In reality, there are imperfections in the wall and in the flow and there is the ques-
tion of boundary layer. It is well known, in the engineering world, that there is no
such thing as a perfect wall. The imperfections of the wall can be, for simplicity’s
sake, assumed to be as a sinusoidal shape. For such a wall the zero inclination
changes from small positive value to a negative value. If the Mach number is large
enough and the wall is rough enough, there will be points where a weak20 weak
will be created. On the other hand, the boundary layer covers or smooths out the
bumps. With these conflicting mechanisms, both will not allow a situation of zero
inclination with emission of Mach wave. At the very extreme case, only in several
points (depending on the bumps) at the leading edge can a very weak shock oc-
cur. Therefore, for the purpose of an introductory class, no Mach wave at zero
inclination should be assumed.
Furthermore, if it was assumed that no boundary layer exists and the wall is perfect,
any deviations from the zero inclination angle creates a jump from a positive angle
(Mach wave) to a negative angle (expansion wave). This theoretical jump occurs
because in a Mach wave the velocity decreases while in the expansion wave the
velocity increases. Furthermore, the increase and the decrease depend on the
upstream Mach number but in different directions. This jump has to be in reality
either smoothed out or has a physical meaning of jump (for example, detach normal
shock). The analysis started by looking at a normal shock which occurs when there
is a zero inclination. After analysis of the oblique shock, the same conclusion must
be reached, i.e. that the normal shock can occur at zero inclination. The analysis of
the oblique shock suggests that the inclination angle is not the source (boundary
condition) that creates the shock. There must be another boundary condition(s)
that causes the normal shock. In the light of this discussion, at least for a simple
engineering analysis, the zone in the proximity of zero inclination (small positive
and negative inclination angle) should be viewed as a zone without any change
unless the boundary conditions cause a normal shock.
Nevertheless, emission of Mach wave can occur in other situations. The approxi-

20 It is not a mistake, there are two “weaks.” These words mean two different things. The first “weak”

means more of compression “line” while the other means the weak shock.
14.4. SOLUTION OF MACH ANGLE 271

mation of weak weak wave with nonzero strength has engineering applicability in a
very limited cases, especially in acoustic engineering, but for most cases it should
be ignored.

Oblique Shock
k = 1 4 Mx=3
3 90 0.001

80
2.5
70
0.0005 Myw θs
2 60 Mys θw

50
1.5 0
40 0 10 20 30

1 30
-0.0005
20
0.5
10

0 0 -0.001
0.0 10.0 δ 20.0 30.0

Wed Jun 22 15:03:35 2005

Fig. -14.9: The calculation of D (possible error), shock angle, and exit Mach number for
M1 = 3

14.4.3 Upstream Mach Number, M1 , and Shock Angle, θ

The solution for upstream Mach number, M1 , and shock angle, θ, are far much
simpler and a unique solution exists. The deflection angle can be expressed as a
function of these variables as
· ¸
(k + 1)M1 2
cot δ = tan θ −1
2(M1 2 sin2 θ − 1)
(14.51)
272 CHAPTER 14. OBLIQUE SHOCK

or
2 cot θ(M1 2 sin2 θ − 1)
tan δ = (14.52)
2 + M1 2 (k + 1 − 2 sin2 θ)

The pressure ratio can be expressed as

P2 2kM1 2 sin2 θ − (k − 1)
=
P1 k+1
(14.53)

The density ratio can be expressed as

ρ2 U1 n (k + 1)M1 2 sin2 θ
= =
ρ1 U2 n (k − 1)M1 2 sin2 θ + 2
(14.54)

The temperature ratio expressed as

¡ ¢¡ ¢
T2 c2 2 2kM1 2 sin2 θ − (k − 1) (k − 1)M1 2 sin2 θ + 2
= 2 =
T1 c1 (k + 1)M1 2 sin2 θ
(14.55)

The Mach number after the shock is

(k − 1)M1 2 sin2 θ + 2
M2 2 sin(θ − δ) =
2kM1 2 sin2 θ − (k − 1)
(14.56)

or explicitly

(k + 1)2 M1 4 sin2 θ − 4(M1 2 sin2 θ − 1)(kM1 2 sin2 θ + 1)

M2 2 = ¡ ¢¡ ¢ (14.57)
2kM1 2 sin2 θ − (k − 1) (k − 1)M1 2 sin2 θ + 2

The ratio of the total pressure can be expressed as

· ¸ k−1
k · ¸ k−1
1
P02 (k + 1)M1 2 sin2 θ k+1
=
P01 (k − 1)M1 2 sin2 θ + 2 2kM1 2 sin2 θ − (k − 1)
(14.58)

Even though the solution for these variables, M1 and θ, is unique, the possible
range deflection angle, δ, is limited. Examining equation (14.51) shows that the
shock angle, θ , has to be in the range of sin−1 (1/M1 ) ≥ θ ≥ (π/2) (see Figure
14.10).
q The range of given θ, upstream Mach number M1 , is limited between ∞
and 1/ sin2 θ.
14.4. SOLUTION OF MACH ANGLE 273
subsonic
weak
solution

1.0 < M1 < ∞

Defection angle

supersonic
strong
solution −1 1 weak
θmin = sin soution
M1

no
possible solution solution
zone

θ, Shock angle
π π θ=0
θ= θmax ∼
2 2
Fig. -14.10: The possible range of solutions for different parameters for given upstream
Mach numbers

14.4.4 Given Two Angles, δ and θ

It is sometimes useful to obtain a relationship where the two angles are known.
The first upstream Mach number, M1 is

2(cot θ + tan δ)
M1 2 =
sin 2θ − (tan δ)(k + cos 2θ)
(14.59)

The reduced pressure difference is

2(P2 − P1 ) 2 sin θ sin δ

= (14.60)
ρU 2 cos(θ − δ)

The reduced density is

ρ2 − ρ1 sin δ
= (14.61)
ρ2 sin θ cos(θ − δ)

For a large upstream Mach number M1 and a small shock angle (yet not approach-
ing zero), θ, the deflection angle, δ must also be small as well. Equation (14.51)
274 CHAPTER 14. OBLIQUE SHOCK

can be simplified into

k+1
θ∼
= δ (14.62)
2

The results are consistent with the initial assumption which shows that it was an
appropriate assumption.

Fig. -14.11: Color-schlieren image of a two dimensional flow over a wedge. The total de-
flection angel (two sides) is 20◦ and upper and lower Mach angel are ∼ 28◦ and
∼ 30◦ , respectively. The image show the end–effects as it has thick (not sharp
transition) compare to shock over a cone. The image was taken by Gary Settles
at Gas Dynamics laboratory Penn State University.

14.4.5 Flow in a Semi–2D Shape

Example 14.2:
In Figure 14.11 exhibits wedge in a supersonic flow with unknown Mach number.
Examination of the Figure reveals that it is in angle of attack. 1) Calculate the Mach
number assuming that the lower and the upper Mach angles are identical and equal
to ∼ 30◦ each (no angle of attack). 2) Calculate the Mach number and angle of
attack assuming that the pressure after the shock for the two oblique shocks is
equal. 3) What kind are the shocks exhibits in the image? (strong, weak, unteady)
4) (Open question) Is there possibilty to estimate the air stagnation temperature
from the information provided in the image. You can assume that spesific heats, k
is a monotonic increasing function of the temperature.
14.4. SOLUTION OF MACH ANGLE 275

S OLUTION
Part (1)
The Mach angle and deflection angle can be obtained from the Figure 14.11. With
this data and either using equation (14.59) or potto-GDC results in
P0 y
M1 Mx My s My w θs θw δ P0 x
2.6810 2.3218 0 2.24 0 30 10 0.97172

The actual Mach number after the shock is then

M2 n 0.76617
M2 = = = 0.839
sin (θ − δ) sin(30 − 10)
The flow after the shock is subsonic flow.
Part (2)
For the lower part shock angle of ∼ 28◦ the results are
P0 y
M1 Mx My s My w θs θw δ P0 x
2.9168 2.5754 0 2.437 0 28 10 0.96549

From the last table, it is clear that Mach number is between the two values of
2.9168 and 2.6810 and the pressure ratio is between 0.96549 and 0.97172. One
of procedure to calculate the attack angle is such that pressure has to match by
“guessing” the Mach number between the extreme values.
Part (3)
The shock must be weak shock because the shock angle is less than 60◦ .
End Solution
The discussion so far was about the no shock

straight infinite long wedge21 which is ysis

flow direction
anal
iate
rmid range
inte
a “pure” 2–D configuration. Clearly, sis
{

naly
al a
norm range
{

for any finite length of the wedge, the

{

no shock
{

analysis needs to account for edge ef-

{

fects. The end of the wedge must

have a different configuration (see Fig-
ure (14.12)). Yet, the analysis for the
middle section produces a close re-
2-D oblique shock
sult to reality (because of symmetry). s
on both sides

ysi
anal
The section where the current anal- edge range

ysis is close to reality can be esti-

mated from a dimensional analysis for
Fig. -14.12: Schematic of finite wedge with
21 Even finite wedge with limiting wall can be considered as anzero
example forof
this discussion if the B.L.
angle attack.
is neglected.
276 CHAPTER 14. OBLIQUE SHOCK

the required accuracy or by a numer-

ical method. The dimensional analy-
sis shows that only the doted area to
be area where current solution can be
assumed as correct22 . In spite of the
small area were the current solution can be assumed, this solution is also act as
a “reality check” to any numerical analysis. The analysis also provides additional
value of the expected range. In Figure 14.11 shows that “shock angle” is not sharp.
The thickness (into page) of the wedge is only one half times the the wedge itself23 .
Even for this small ratio two dimensional it provide very good results.
Another geometry that can be considered as two–dimensional is the cone (some
referred to it as Taylor–Maccoll flow). Even though, the cone is a three–
dimensional problem, the symmetrical nature of the cone creates a semi–2D prob-
lem. In this case there are no edge effects and the geometry dictates slightly
different results. The mathematics is much more complicated but there are three
solutions. As before, the first solution is thermodynamical unstable. Experimental
and analytical work shows that the weak solution is the stable solution and a dis-
cussion is provided in the appendix of this chapter. As opposed to the weak shock,
the strong shock is unstable, at least, for steady state and no known experiments
showing that it exist can be found in the literature. All the literature, known to the
author, reports that only a weak shock is possible.

14.4.6 Small δ “Weak Oblique shock”

This interest in this topic is mostly from an academic point of view. It is recom-
mended that this issue be skipped and the time be devoted to other issues. The
author is not aware of any single case in which this topic is used in real–world cal-
culations. In fact, after the explicit analytical solution has been provided, studying
this topic seems to come at the expense of other more important topics. However,
the author admits that as long as there are instructors who examine their students
on this issue, it should be covered in this book.
For small deflection angles, δ, and small normal upstream Mach numbers, M1 ∼
1 + ²,

1
tan θ = p (14.63)
M1 2 − 1

... under construction.

22 At this stage, dimensional analysis is not completed. The author is not aware of any such analysis
in literature. The common approach is to carry out numerical analysis. In spite of recent trends, for
most engineering applications, a simple tool is sufficient for limit accuracy. Additionally, the numerical
works require many times a “reality check.”
23 This information is according to Gary Settles which he provided the estimate only.
14.4. SOLUTION OF MACH ANGLE 277

14.4.7 Close and Far Views of the Oblique Shock

In many cases, the close proximity view
provides a continuous turning of the de-
flection angle, δ. Yet, the far view shows
a sharp transition. The traditional ap-
proach to reconcile these two views is
by suggesting that the far view shock is
θ
a collection of many small weak shocks
(see Figure 14.13). At the local view
close to the wall, the oblique shock is a
δ
weak “weak oblique” shock. From the
far view, the oblique shock is an accu-
mulation of many small (or again weak)
“weak shocks.” However, these small Fig. -14.13: /; A local and a far view of the
oblique shock.
“shocks” are built or accumulate into a
large and abrupt change (shock). In this theory, the boundary layer (B.L.) does not
enter into the calculation. In reality, the boundary layer increases the zone where a
continuous flow exists. The boundary layer reduces the upstream flow velocity and
therefore the shock does not exist at close proximity to the wall. In larger distance
from the wall, the shock becomes possible.

14.4.8 Maximum Value of Oblique shock

The maximum values are summarized in the following Table .

Table -14.1: Table of maximum values of the oblique Shock k=1.4

Mx My δmax θmax
1.1000 0.97131 1.5152 76.2762
1.2000 0.95049 3.9442 71.9555
1.3000 0.93629 6.6621 69.3645
1.4000 0.92683 9.4272 67.7023
1.5000 0.92165 12.1127 66.5676
1.6000 0.91941 14.6515 65.7972
1.7000 0.91871 17.0119 65.3066
1.8000 0.91997 19.1833 64.9668
1.9000 0.92224 21.1675 64.7532
2.0000 0.92478 22.9735 64.6465
2.2000 0.93083 26.1028 64.6074
2.4000 0.93747 28.6814 64.6934
2.6000 0.94387 30.8137 64.8443
2.8000 0.94925 32.5875 65.0399
278 CHAPTER 14. OBLIQUE SHOCK
Table -14.1: Maximum values of oblique shock (continue) k=1.4

Mx My δmax θmax
3.0000 0.95435 34.0734 65.2309
3.2000 0.95897 35.3275 65.4144
3.4000 0.96335 36.3934 65.5787
3.6000 0.96630 37.3059 65.7593
3.8000 0.96942 38.0922 65.9087
4.0000 0.97214 38.7739 66.0464
5.0000 0.98183 41.1177 66.5671
6.0000 0.98714 42.4398 66.9020
7.0000 0.99047 43.2546 67.1196
8.0000 0.99337 43.7908 67.2503
9.0000 0.99440 44.1619 67.3673
10.0000 0.99559 44.4290 67.4419

It must be noted that the calculations are for the perfect gas model. In some cases,
this assumption might not be sufficient and different analysis is needed. Henderson
and Menikoff24 suggested a procedure to calculate the maximum deflection angle
for arbitrary equation of state25 .

14.5 Detached Shock

When the mathematical quantity D be-
comes positive, for large deflection an-

k
oc
sh
gle, there isn’t a physical solution to an

k
ea
w
M >1
oblique shock. Since the flow “sees” the Strong Shock Supersoini Area
zone C
obstacle, the only possible reaction is by Upstream U∞

a normal shock which occurs at some zone B Subsonic Area θ

distance from the body. This shock is Normal Shock zone A
M<
referred to as the detach shock. The de- 1

tached shock’s distance from the body Deta hment Distan e

For symmetri al bodys
is a complex analysis and should be
left to graduate class and researchers
in this area. Nevertheless, a graph and
a general explanation to engineers is
provided. Even though this topic has
few applications, some might be used in Fig. -14.14: The schematic for a round–tip
certain situations which the author isn’t bullet in a supersonic flow.
aware of.
24 Henderson and Menikoff ”Triple Shock Entropy Theorem” Journal of Fluid Mechanics 366 (1998)

pp. 179–210.
25 The effect of the equation of state on the maximum and other parameters at this state is unknown

at this moment and there are more works underway.

14.5. DETACHED SHOCK 279

Analysis of the detached shock can be carried out by looking at a body with a
round section moving in a supersonic flow (the absolute velocity isn’t important for
this discussion). Figure 14.14 exhibits a round–tip bullet with a detached shock.
The distance of the detachment is determined to a large degree by the upstream
Mach number. The zone A is zone where the flow must be subsonic because at
the body the velocity must be zero (the no–slip condition). In such a case, the gas
must go through a shock. While at zone C the flow must be supersonic. The weak
oblique shock is predicted to flow around the cone. The flow in zone A has to go
through some acceleration to became supersonic flow. The explanation to such a
phenomenon is above the level of this book (where is the “throat” area question26 .
Yet, it can be explained as the subsonic is “sucked” into gas in zone C. Regardless
of the explanation, these calculations can be summarized by the flowing equation

detachment distance
= constant × (θ − f (M∞ )) (14.64)
body thickness

where f (M∞ ) is a function of the upstream Mach number which tabulated in the
literature.
The constant and the function are different for different geometries. As a general
rule, the increase in the upstream Mach results in a decrease of the detachment
distance. Larger shock results in a smaller detachment distance, or, alternatively,
the flow becomes “blinder” to obstacles. Thus, this phenomenon has a larger im-
pact for a relatively smaller supersonic flow.

14.5.1 Issues Related to the Maximum Deflection Angle

The issue of maximum deflection
oblique sho ks
has a practical application aside δ1
θ1
from the obvious configuration used
as a typical simple example. In the
typical example, a wedge or a cone U B 2 C
θ2 Slip Plane
moves into a still medium or gas A
flows into it. If the deflection angle δ2
exceeds the maximum possible, a
detached shock occurs. However,
there are configurations in which a
detached shock occurs in design
Fig. -14.15: The schematic for a symmetrical suc-
and engineers need to take it into tion section with Mach reflection.
consideration. Such configurations
seem sometimes at first glance not related to the detached shock issue. Consider,
for example, a symmetrical suction section in which the deflection angle is just be-
tween the maximum deflection angle and above half of the maximum deflection
angle. In this situation, at least two oblique shocks occur and after their interaction
26 See example 14.6.
280 CHAPTER 14. OBLIQUE SHOCK

is shown in Figure (14.15). No detached shock issues are raised when only the
first oblique shock is considered. However, the second oblique shock complicates
the situation and the second oblique shock can cause a detached shock. This
situation is referred to in the scientific literature as the Mach reflection.
It can be observed that the maximum of the oblique shock for the perfect gas model
depends only on the upstream Mach number i.e., for every upstream Mach number
there is only one maximum deflection angle.

δmax = f (M1 ) (14.65)

Additionally, it can be observed for

a maximum oblique shock that a con-
stant deflection angle decrease of the
Mach number results in an increase of oblique sho ks
Mach angle (weak shock only) M1 > θ1 δ1
M2 =⇒ θ1 < θ2 . The Mach number
decreases after every shock. There-
U B C
fore, the maximum deflection angle de-
creases with a decrease the Mach num- A sub sonic
ber. Additionally, due to the symmetry flow
a slip plane angle can be guessed to
be parallel to original flow, hence δ1 =
δ2 . Thus, this situation causes the de-
tached shock to appear in the second Fig. -14.16: The “detached” shock in a com-
oblique shock. This detached shock plicated configuration sometimes
manifested itself in a form of curved referred to as Mach reflection.
shock (see Figure 14.16).
The analysis of this situation is logically very simple, yet the mathematics is
somewhat complicated. The maximum deflection angle in this case is, as before,
only a function of the upstream Mach number. The calculations for such a case
can be carried out by several approaches. It seems that the most straightforward
method is the following:
(a) Calculate M1 B ;
(b) Calculate the maximum deflection angle, θ2 , utilizing (14.36) equation
(c) Calculate the deflection angle, δ2 utilizing equation (14.12)
(d) Use the deflection angle, δ2 = δ1 and the Mach number M1 B to calculate
M1 B . Note that no maximum angle is achieved in this shock. Potto–GDC
can be used to calculate this ratio.
This procedure can be extended to calculate the maximum incoming Mach number,
M1 by checking the relationship between the intermediate Mach number to M1 .
In discussing these issues, one must be aware that there are zones of dual solu-
tions in which sharp shock line coexists with a curved line. In general, this zone
14.5. DETACHED SHOCK 281

increases as Mach number increases. For example, at Mach 5 this zone is 8.5◦ .
For engineering purposes when the Mach number reaches this value, it can be
ignored.

14.5.2 Oblique Shock Examples

Example 14.3:
Air flows at Mach number (M1 ) or Mx = 4 is approaching a wedge. What is the
maximum wedge angle at which the oblique shock can occur? If the wedge angle
is 20◦ , calculate the weak, the strong Mach numbers, and the respective shock
angles.

S OLUTION
The maximum wedge angle for (Mx = 4) D has to be equal to zero. The wedge
angle that satisfies this requirement is by equation (14.28) (a side to the case
proximity of δ = 0). The maximum values are:

Mx My δmax θmax
4.0000 0.97234 38.7738 66.0407

To obtain the results of the weak and the strong solutions either utilize the equation
(14.28) or the GDC which yields the following results

Mx My s My w θs θw δ
4.0000 0.48523 2.5686 1.4635 0.56660 0.34907

Fig. -14.17: Oblique shock occurs around a cone. This photo is courtesy of Dr. Grigory
Toker, a Research Professor at Cuernavaco University of Mexico. According to
his measurement, the cone half angle is 15◦ and the Mach number is 2.2.
282 CHAPTER 14. OBLIQUE SHOCK

Example 14.4:
A cone shown in Figure (14.17) is exposed to supersonic flow and create an
oblique shock. Is the shock shown in the photo weak or strong shock? Explain.
Using the geometry provided in the photo, predict at which Mach number was the
photo taken based on the assumption that the cone is a wedge.

S OLUTION
The measurement shows that cone angle is 14.43◦ and the shock angle is 30.099◦ .
With given two angles the solution can be obtained by utilizing equation (14.59) or
the Potto-GDC.

P0 y
M1 My s My w θs θw δ P0 x
3.2318 0.56543 2.4522 71.0143 30.0990 14.4300 0.88737

Because the flow is around the cone it must be a weak shock. Even if the cone
was a wedge, the shock would be weak because the maximum (transition to a
strong shock) occurs at about 60◦ . Note that the Mach number is larger than the
one predicted by the wedge.

Oblique Shock
k=14
3 90

80
2.5
70
θ
2 60 δ
My
50
1.5
40

1 30

20
0.5
10

0 0
2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Mx

Thu Jun 30 15:14:53 2005

Fig. -14.18: Maximum values of the properties in an oblique shock

14.5. DETACHED SHOCK 283

14.5.3 Application of Oblique Shock

One of the practical applications of the
oblique shock is the design of an in-  1

let suction for a supersonic flow. It is

normal sho k oblique sho ks 
suggested that a series of weak shocks 3


should replace one normal shock to 2

increase the efficiency (see Figure

(14.19))a . Clearly, with a proper de-
sign, the flow can be brought to a sub-
sonic flow just below M = 1. In such Fig. -14.19: Two variations of inlet suction for
a case, there is less entropy production supersonic flow.
(less pressure loss). To illustrate the design significance of the oblique shock, the
following example is provided.
a In fact, there is general proof that regardless to the equation of state (any kind of gas), the entropy

is to be minimized through a series of oblique shocks rather than through a single normal shock. For
details see Henderson and Menikoff “Triple Shock Entropy Theorem,” Journal of Fluid Mechanics 366,
(1998) pp. 179–210.

Example 14.5:
The Section described in Figure 14.20 air is flowing into a suction section at M =
2.0, P = 1.0[bar], and T = 17◦ C. Compare the different conditions in the two
different configurations. Assume that only a weak shock occurs.

1
neglect
the detached
distance

7◦
normal sho k oblique sho ks
2 Normal shock
3 4
1
2 7◦

Fig. -14.20: Schematic for Example (14.5).

S OLUTION
The first configuration is of a normal shock for which the results27 are

27 The results in this example are obtained using the graphical interface of POTTO–GDC thus, no

input explanation is given. In the past the input file was given but the graphical interface it is no longer
needed.
284 CHAPTER 14. OBLIQUE SHOCK
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
2.0000 0.57735 1.6875 2.6667 4.5000 0.72087

In the oblique shock, the first angle shown is

P0 y
Mx My s My w θs θw δ P0 x
2.0000 0.58974 1.7498 85.7021 36.2098 7.0000 0.99445

and the additional information by the minimal info in the Potto-GDC is

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
2.0000 1.7498 36.2098 7.0000 1.2485 1.1931 0.99445

In the new region, the new angle is 7◦ + 7◦ with new upstream Mach number of
Mx = 1.7498 resulting in

P0 y
Mx My s My w θs θw δ P0 x
1.7498 0.71761 1.2346 76.9831 51.5549 14.0000 0.96524

And the additional information is

Py Ty P0 y
Mx My w θw δ Px Tx P0x
1.7498 1.5088 41.8770 7.0000 1.2626 1.1853 0.99549

An oblique shock is not possible and normal shock occurs. In such a case, the
results are:
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
1.2346 0.82141 1.1497 1.4018 1.6116 0.98903

With two weak shock waves and a normal shock the total pressure loss is
P0 4 P0 P0 P0
= 4 3 2 = 0.98903 × 0.96524 × 0.99445 = 0.9496
P0 1 P0 3 P0 2 P0 1
The static pressure ratio for the second case is
P4 P4 P3 P2
= = 1.6116 × 1.2626 × 1.285 = 2.6147
P1 P3 P2 P1
The loss in this case is much less than in a direct normal shock. In fact, the loss in
the normal shock is above than 31% of the total pressure.
14.5. DETACHED SHOCK 285

Example 14.6: My w

A supersonic flow is approaching a very long two– A∗

10 ◦

dimensional bland wedge body and creates a detached M ys

shock at Mach 3.5 (see Figure 14.21). The half wedge

angle is 10◦ . What is the requited “throat” area ra-
tio to achieve acceleration from the subsonic region Fig. -14.21: Schematic for
to the supersonic region assuming the flow is one– Example (14.6).
dimensional?

S OLUTION
The detached shock is a normal shock and the results are
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
3.5000 0.45115 3.3151 4.2609 14.1250 0.21295

Now utilizing the isentropic relationship for k = 1.4 yields

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
0.45115 0.96089 0.90506 1.4458 0.86966 1.2574

Thus the area ratio has to be 1.4458. Note that the pressure after the weak shock
is irrelevant to the area ratio between the normal shock and the “throat” according
to the standard nozzle analysis.

Example 14.7:
D4 Slip Plane
The effects of a double
wedge are explained in the B P3 = P4

government web site as weak

3
oblique weak
shown in Figure (14.22). shock oblique
shock E
Adopt this description and or expension
wave
M 2
assume that the turn of
1

1
◦ 0
6 is made of two equal C
angles of 3◦ (see Figure A

14.22). Assume that there

are no boundary layers and Fig. -14.22: Schematic of two angles turn with two weak
all the shocks are weak and shocks.
straight. Perform the calcu-
lation for M1 = 3.0. Find
the required angle of shock
BE. Then, explain why this description has internal conflict.

S OLUTION
The shock BD is an oblique shock with a response to a total turn of 6◦ . The condi-
tions for this shock are:
286 CHAPTER 14. OBLIQUE SHOCK
P0 y
Mx My s My w θs θw δ P0 x
3.0000 0.48013 2.7008 87.8807 23.9356 6.0000 0.99105

The transition for shock AB is

P0y
Mx My s My w θs θw δ P0 x
3.0000 0.47641 2.8482 88.9476 21.5990 3.0000 0.99879

For the shock BC the results are

P0 y
Mx My s My w θs θw δ P0 x
2.8482 0.48610 2.7049 88.8912 22.7080 3.0000 0.99894

And the isentropic relationships for M = 2.7049, 2.7008 are

T ρ A P A×P
M T0 ρ0 A? P0 A∗ ×P0
2.7049 0.40596 0.10500 3.1978 0.04263 0.13632
2.7008 0.40669 0.10548 3.1854 0.04290 0.13665

The combined shocks AB and BC provide the base of calculating the total pressure
ratio at zone 3. The total pressure ratio at zone 2 is
P0 2 P0 P0
= 2 1 = 0.99894 × 0.99879 = 0.997731283
P0 0 P0 1 P0 0
On the other hand, the pressure at 4 has to be
P4 P4 P0 4
= = 0.04290 × 0.99105 = 0.042516045
P0 1 P0 4 P0 1
The static pressure at zone 4 and zone 3 have to match according to the govern-
ment suggestion hence, the angle for BE shock which cause this pressure ratio
needs to be found. To do that, check whether the pressure at 2 is above or below
or above the pressure (ratio) in zone 4.
P2 P0 P2
= 2 = 0.997731283 × 0.04263 = 0.042436789
P0 2 P0 0 P0 2

Since PP022 < PP041 a weak shock must occur to increase the static pressure (see
Figure 6.4). The increase has to be

P3 /P2 = 0.042516045/0.042436789 = 1.001867743

To achieve this kind of pressure ratio the perpendicular component has to be

14.5. DETACHED SHOCK 287
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
1.0008 0.99920 1.0005 1.0013 1.0019 1.00000

The shock angle, θ can be calculated from

θ = sin−1 1.0008/2.7049 = 21.715320879◦

The deflection angle for such shock angle with Mach number is

P0 y
Mx My s My w θs θw δ P0x
2.7049 0.49525 2.7037 0.0 21.72 0.026233 1.00000

From the last calculation it is clear that the government proposed schematic of the
double wedge is in conflict with the boundary condition. The flow in zone 3 will flow
into the wall in about 2.7◦ . In reality the flow of double wedge will produce a curved
shock surface with several zones. Only when the flow is far away from the double
wedge, the flow behaves as only one theoretical angle of 6◦ exist.

Example 14.8:
Calculate the flow deflection angle and other parameters downstream when the
Mach angle is 34◦ and P1 = 3[bar], T1 = 27◦ C, and U1 = 1000m/sec. Assume
k = 1.4 and R = 287J/KgK

S OLUTION
The Mach angle of 34◦ is below maximum deflection which means that it is a weak
shock. Yet, the Upstream Mach number, M1 , has to be determined

U1 1000
M1 = √ = = 2.88
kRT 1.4 × 287 × 300

Using this Mach number and the Mach deflection in either using the Table or the
figure or POTTO-GDC results in

P0y
Mx My s My w θs θw δ P0 x
2.8800 0.48269 2.1280 0.0 34.00 15.78 0.89127

The relationship for the temperature and pressure can be obtained by using equa-
tion (14.15) and (14.13) or simply converting the M1 to perpendicular component.

M1n = M1 ∗ sin θ = 2.88 sin(34.0) = 1.61

From the Table (6.1) or GDC the following can be obtained.

288 CHAPTER 14. OBLIQUE SHOCK
Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
1.6100 0.66545 1.3949 2.0485 2.8575 0.89145

The temperature ratio combined upstream temperature yield

T2 = 1.3949 × 300 ∼ 418.5K
and the same for the pressure
P2 = 2.8575 × 3 = 8.57[bar]
And the velocity
√ √
Un2 = My w kRT = 2.128 1.4 × 287 × 418.5 = 872.6[m/sec]

Example 14.9:
For Mach number 2.5 and wedge with a total angle of 22◦ , calculate the ratio of the
stagnation pressure.

Utilizing GDC for Mach number 2.5 and the angle of 11◦ results in
P0y
Mx My s My w θs θw δ P0 x
2.5000 0.53431 2.0443 85.0995 32.8124 11.0000 0.96873

Example 14.10:
What is the maximum pressure ratio that can be obtained on wedge when the gas
is flowing in 2.5 Mach without any close boundaries? Would it make any difference
if the wedge was flowing into the air? If so, what is the difference?

S OLUTION
It has to be recognized that without any other boundary condition, the shock is
weak shock. For a weak shock the maximum pressure ratio is obtained at the
deflection point because it is closest to a normal shock. To obtain the maximum
point for 2.5 Mach number, either use the Maximum Deflection Mach number’s
equation or the Potto–GDC
Py Ty P0 y
Mx My max θmax δ Px Tx P0x
2.5000 0.94021 64.7822 29.7974 4.3573 2.6854 0.60027

In these calculations, Maximum Deflection Mach’s equation was used to calculate

the normal component of the upstream, then the Mach angle was calculated us-
ing the geometrical relationship of θ = sin−1 M1n /M1 . With these two quantities,
utilizing equation (14.12) the deflection angle, δ, is obtained.
14.5. DETACHED SHOCK 289

Example 14.11:
Consider the schematic shown in the following figure.

3
stream line
2
1
θ
M1 = 4
δ

Assume that the upstream Mach number is 4 and the deflection angle is δ = 15◦ .
Compute the pressure ratio and the temperature ratio after the second shock
(sometimes referred to as the reflective shock while the first shock is called the
incidental shock).

S OLUTION
This kind of problem is essentially two wedges placed in a certain geometry. It is
clear that the flow must be parallel to the wall. For the first shock, the upstream
Mach number is known together with deflection angle. Utilizing the table or the
Potto–GDC, the following can be obtained:

P0 y
Mx My s My w θs θw δ P0 x
4.0000 0.46152 2.9290 85.5851 27.0629 15.0000 0.80382

And the additional information by using minimal information ratio button in Potto–
GDC is
Py Ty P0 y
Mx My w θw δ Px Tx P0x
4.0000 2.9290 27.0629 15.0000 1.7985 1.7344 0.80382

With a Mach number of M = 2.929, the second deflection angle is also 15◦ . With
these values the following can be obtained:

P0y
Mx My s My w θs θw δ P0 x
2.9290 0.51367 2.2028 84.2808 32.7822 15.0000 0.90041

and the additional information is

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
2.9290 2.2028 32.7822 15.0000 1.6695 1.5764 0.90041
290 CHAPTER 14. OBLIQUE SHOCK

With the combined tables the ratios can be easily calculated. Note that hand cal-
culations requires endless time looking up graphical representation of the solution.
Utilizing the POTTO–GDC which provides a solution in just a few clicks.
P1 P1 P2
= = 1.7985 × 1.6695 = 3.0026
P3 P2 P3
T1 T1 T2
= = 1.7344 × 1.5764 = 2.632
T3 T2 T3

Example 14.12:
A similar example as before but here Mach angle is 29◦ and Mach number is 2.85.
Again calculate the downstream ratios after the second shock and the deflection
angle.

S OLUTION
Here the Mach number and the Mach angle are given. With these pieces of infor-
mation by utilizing the Potto-GDC the following is obtained:
P0 y
Mx My s My w θs θw δ P0 x
2.8500 0.48469 2.3575 0.0 29.00 10.51 0.96263

and the additional information by utilizing the minimal info button in GDC provides

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
2.8500 2.3575 29.0000 10.5131 1.4089 1.3582 0.96263

With the deflection angle of δ = 10.51 the so called reflective shock gives the
following information
P0 y
Mx My s My w θs θw δ P0x
2.3575 0.54894 1.9419 84.9398 34.0590 10.5100 0.97569

and the additional information of

Py Ty P0y
Mx My w θw δ Px Tx P0 x
2.3575 1.9419 34.0590 10.5100 1.3984 1.3268 0.97569

P1 P1 P2
= = 1.4089 × 1.3984 ∼ 1.97
P3 P2 P3
T1 T1 T2
= = 1.3582 × 1.3268 ∼ 1.8021
T3 T2 T3
14.5. DETACHED SHOCK 291

Example 14.13:
Compare a direct normal shock to oblique shock with a normal shock. Where will
the total pressure loss (entropy) be larger? Assume that upstream Mach number
is 5 and the first oblique shock has Mach angle of 30◦ . What is the deflection angle
in this case?

S OLUTION
For the normal shock the results are

Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
5.0000 0.41523 5.8000 5.0000 29.0000 0.06172

While the results for the oblique shock are

P0y
Mx My s My w θs θw δ P0 x
5.0000 0.41523 3.0058 0.0 30.00 20.17 0.49901

And the additional information is

Py Ty P0 y
Mx My w θw δ Px Tx P0x
5.0000 3.0058 30.0000 20.1736 2.6375 2.5141 0.49901

The normal shock that follows this oblique is

Ty ρy Py P0 y
Mx My Tx ρx Px P0 x
3.0058 0.47485 2.6858 3.8625 10.3740 0.32671

The pressure ratios of the oblique shock with normal shock is the total shock in the
second case.
P1 P1 P2
= = 2.6375 × 10.374 ∼ 27.36
P3 P2 P3
T1 T1 T2
= = 2.5141 × 2.6858 ∼ 6.75
T3 T2 T3
Note the static pressure raised is less than the combination shocks as compared
to the normal shock but the total pressure has the opposite result.

Example 14.14:
A flow in a tunnel ends up with two deflection angles from both sides (see the
following Figure (14.14)).
292 CHAPTER 14. OBLIQUE SHOCK

C δ2

stream line θ2 D
1 4
slip plane
B
0 3 φ

2 θ1 F

stream line
A δ1

Illustration for example (14.14)

For upstream Mach number of 5 and deflection angle of 12◦ and 15◦ , calculate the
pressure at zones 3 and 4 based on the assumption that the slip plane is half of the
difference between the two deflection angles. Based on these calculations, explain
whether the slip angle is larger or smaller than the difference of the deflection
angle.

S OLUTION
The first two zones immediately after are computed using the same techniques
that were developed and discussed earlier.
For the first direction of 15◦ and Mach number =5.

P0 y
Mx My s My w θs θw δ P0 x
5.0000 0.43914 3.5040 86.0739 24.3217 15.0000 0.69317

And the additional conditions are

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
5.0000 3.5040 24.3217 15.0000 1.9791 1.9238 0.69317

For the second direction of 12◦ and Mach number =5.

P0 y
Mx My s My w θs θw δ P0 x
5.0000 0.43016 3.8006 86.9122 21.2845 12.0000 0.80600

And the additional conditions are

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
5.0000 3.8006 21.2845 12.0000 1.6963 1.6625 0.80600
14.5. DETACHED SHOCK 293

The conditions in zone 4 and zone 3 have two things that are equal. They are the
pressure and the velocity direction. It has to be noticed that the velocity magnitudes
in zone 3 and 4 do not have to be equal. This non–continuous velocity profile can
occur in our model because it is assumed that fluid is non–viscous.
If the two sides were equal because of symmetry the slip angle is also zero. It is
to say, for the analysis, that only one deflection angle exist. For the two different
deflection angles, the slip angle has two extreme cases. The first case is where
match lower deflection angle and second is to match the higher deflection angle.
In this case, it is assumed that the slip angle moves half of the angle to satisfy both
of the deflection angles (first approximation). Under this assumption the conditions
in zone 3 are solved by looking at the deflection angle of 12◦ + 1.5◦ = 13.5◦ which
results in

P0y
Mx My s My w θs θw δ P0 x
3.5040 0.47413 2.6986 85.6819 27.6668 13.5000 0.88496

with the additional information

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
3.5040 2.6986 27.6668 13.5000 1.6247 1.5656 0.88496

And in zone 4 the conditions are due to deflection angle of 13.5◦ and Mach 3.8006

P0 y
Mx My s My w θs θw δ P0 x
3.8006 0.46259 2.9035 85.9316 26.3226 13.5000 0.86179

with the additional information

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
3.8006 2.9035 26.3226 13.5000 1.6577 1.6038 0.86179

From these tables the pressure ratio at zone 3 and 4 can be calculated

P3 P3 P2 P0 P1 1 1
= = 1.6247 × 1.9791 ∼ 1.18192
P4 P2 P0 P1 P4 1.6963 1.6038

To reduce the pressure ratio the deflection angle has to be reduced (remember
that at weak weak shock almost no pressure change). Thus, the pressure at zone
3 has to be reduced. To reduce the pressure the angle of slip plane has to increase
from 1.5◦ to a larger number.
294 CHAPTER 14. OBLIQUE SHOCK

Example 14.15:
The previous example gave rise to another question on the order of the deflection
angles. Consider the same values as previous analysis, will the oblique shock with
first angle of 15◦ and then 12◦ or opposite order make a difference (M = 5)? If not
what order will make a bigger entropy production or pressure loss? (No general
proof is needed).

S OLUTION
Waiting for the solution

14.5.4 Optimization of Suction Section Design

Under heavy construction please ignore
The question raised is what is the optimum design for inlet suction unit?
There are several considerations that have to be taken into account besides su-
personic flow which includes for example the material strength consideration and
the operation factors.
The optimum deflection angle is a function of the Mach number range in
which the suction section is operated in. There are researchers which suggest
that the numerical work is the solution.

14.5.5 Retouch of Shock or Wave Drag

stationary control
volume

moving
object U2 6= 0
U1 = 0 ρ2
ρ1
A1 A2
P1
P2

stream lines

Fig. -14.23: The diagram that explains the shock drag effect of a moving shock considering
the oblique shock effects.

Since it was established that the common explanation is erroneous and the
steam lines are bending/changing direction when they touching the oblique shock
(compare with figure (6.7)). The correct explanation is that increase of the mo-
mentum into control volume is either requires increase of the force and/or results
in acceleration of gas. So, what is the effects of the oblique shock on the Shock
Drag? Figure (14.23) exhibits schematic of the oblique shock which show clearly
14.6. SUMMARY 295

that stream lines are bended. There two main points that should be discussed
in this context are the additional effects and infinite/final structure. The additional
effects are the mass start to have a vertical component. The vertical component
one hand increase the energy needed and thus increase need to move the body
(larger shock drag) (note the there is a zero momentum net change for symmetrical
bodies.). However, the oblique shock reduces the normal component that under-
goes the shock and hence the total shock drag is reduced. The oblique shock
creates a finite amount of drag (momentum and energy lost) while a normal shock
as indirectly implied in the common explanation creates de facto situation where
the shock grows to be infinite which of course impossible. It should be noted that,
oblique shock becomes less “oblique” and more parallel when other effects start to
kick in.

14.6 Summary

As with normal shock, the oblique shock with upstream Mach number, M1 is al-
ways greater than 1. However, in oblique, as oppose to the normal shock, the
downstream Mach number, M2 could be larger or smaller then 1. The perpendi-
cular component of the downstream Mach number, M1 n is always smaller than 1.
Given M1 and the deflection angle, δ there could be three solutions: the first one is
the “impossible” solution in the case where D is negative, two is weak shock, and
three is strong shock. When D is positive there is no physical solution and only
normal shock exist. When D is equal to zero, a special case is created because
the weak and strong solutions are equal (for large deflection angle). When D > 0,
for large deflection angle, there is a possibility of no two–dimensional solution re-
sulting in a detached shock case.
296 CHAPTER 14. OBLIQUE SHOCK

θ -δ -Mach number relationship

60
k = 1.4
1.05
55
1.10
1.15
1.20
1.3
50
1.4
1.5
1.6
45
1.7
1.8
1.9
40 2.0
2.2
2.4
δ, deflection angle
35 2.6
2.8
3.0
30 3.2
3.4
3.6
25
3.8
4.0
4.5
5
20
6
7
8
15
9
10
15
10 800

0 10 20 30 40 50 60 70 80 90
θ , shock angle

December 4, 2007

The relationship between the shock wave angle, θ and deflection angle, δ, and
Mach number for k=1.4. Graphs for other value can be found in separate
accompanying book for gas dynamics tables. Several additional graphs (different
k) are in a Gas Dynamics Tables by the same author.

14.7 Appendix: Oblique Shock Stability Analysis

The stability analysis is an analysis
which answers the question of what
happens if for some reason, the situa- Unstable Stable

tion moves away from the expected so-

lution. If the answer turns out to be
that the situation will return to its origi-
Fig. -14.24: Typical examples of unstable
and stable situations.
14.7. APPENDIX: OBLIQUE SHOCK STABILITY ANALYSIS 297

nal state then it is referred to as the sta-

ble situation. On the other hand, if the
answer is negative, then the situation is
referred to as unstable. An example to this situation, is a ball shown in the Figure
(14.24). Instinctively, the stable and unstable can be recognized. There is also the
situation where the ball is between the stable and unstable situations when the ball
is on a plane field which is referred to as the neutrally stable. In the same manner,
the analysis for the oblique shock wave is carried out. The only difference is that
here, there are more than one parameter that can be changed, for example, the
shock angle, deflection angle, and upstream Mach number. In this example only
the weak solution is explained. The similar analysis can be applied to strong shock.
Yet, in that analysis it has to be remembered that when the flow becomes subsonic
the equation changes from hyperbolic to an elliptic equation. This change compli-
cates the explanation and is omitted in this section. Of course, in the analysis the
strong shock results in an elliptic solution (or region) as opposed to a hyperbolic
in weak shock. As results, the discussion is more complicated but similar analysis
can be applied to the strong shock.
The change in the in- ∆θ +
clination angle results
in a different upstream ∆θ −

Mach number and a

∆δ −
different pressure. On
the other hand, to main-
∆δ +
tain the same direction
stream lines, the virtual
change in the deflec-
tion angle has to be in Fig. -14.25: The schematic of stability analysis for oblique shock.
the opposite direction
of the change of the shock angle. The change is determined from the solution
provided before or from the approximation (14.62).

k+1
∆θ = ∆δ (14.66)
2
Equation (14.66) can be applied for either positive, ∆θ+ or negative ∆θ− values.
The pressure difference at the wall becomes a negative increment which tends
to pull the shock angle to the opposite direction. The opposite happens when
the deflection increment becomes negative, the deflection angle becomes positive
which increases the pressure at the wall. Thus, the weak shock is stable.
Please note that this analysis doesn’t apply to the case of the close proximity of
the δ = 0. In fact, the shock wave is unstable according to this analysis to one
direction but stable to the other direction. Yet, it must be pointed out that it doesn’t
mean that the flow is unstable but rather that the model is incorrect. There isn’t any
known experimental evidence to show that flow is unstable for δ = 0.
298 CHAPTER 14. OBLIQUE SHOCK
 CHAPTER 15
Prandtl-Meyer Function

positive
angle

15.1 Introduction
As discussed in Chapter (14) when the de-

e
gl
flection turns to the opposite direction of m
mu
an
the flow, the flow accelerates to match the
xi
ma

boundary condition. The transition, as op-

posed to the oblique shock, is smooth, with- Fig. -15.1: The definition of the angle for
out any jump in properties. Here because of the Prandtl–Meyer function.
the tradition, the deflection angle is denoted
as a positive when it is away from the flow (see Figure (15.1)). In a somewhat
a similar concept to oblique shock there exists a “detachment” point above which
this model breaks and another model has to be implemented. Yet, when this model
breaks down, the flow becomes complicated, flow separation occurs, and no known
simple model can describe the situation. As opposed to the oblique shock, there
is no limitation for the Prandtl-Meyer function to approach zero. Yet, for very small
angles, because of imperfections of the wall and the boundary layer, it has to be
assumed to be insignificant.
Supersonic expansion and isentropic com-  U
pression (Prandtl-Meyer function), are an exten-
c
sion of the Mach line concept. The Mach line M
shows that a disturbance in a field of supersonic 1

flow moves in an angle of µ, which is defined as

(as shown in Figure (15.2)) 
p
µ ¶ M 2 1
−1 1
µ = sin (15.1) Fig. -15.2: The angles of the Mach
M
line triangle

299
300 CHAPTER 15. PRANDTL-MEYER FUNCTION

or
1
µ = tan−1 √ (15.2)
M1 − 1
A Mach line results because of a small disturbance in the wall contour. This Mach
line is assumed to be a result of the positive angle. The reason that a “negative”
angle is not applicable is that the coalescing of the small Mach wave which results
in a shock wave. However, no shock is created from many small positive angles.
The Mach line is the chief line in the analysis because of the wall contour
shape information propagates along this line. Once the contour is changed, the
flow direction will change to fit the wall. This direction change results in a change
of the flow properties, and it is assumed here to be isotropic for a positive angle.
This assumption, as it turns out, is close to reality. In this chapter, a discussion on
the relationship between the flow properties and the flow direction is presented.

15.2 Geometrical Explanation

The change in the flow direction is assume
to be result of the change in the tangential U y
x
dx = dU cos(90 − µ)
y

component. Hence, the total Mach num- Ux U + dU y y

ber increases. Therefore, the Mach angle U (90 )

dν
increase and result in a change in the di-  dy
 U + dU
rection of the flow. The velocity compo-
nent in the direction of the Mach line is as-

ne
li
(U + dU ) os(d) U

ch
sumed to be constant to satisfy the assump-

Ma
tion that the change is a result of the contour
Fig. -15.3: The schematic of the turning
only. Later, this assumption will be exam-
flow.
ined. The typical simplifications for geomet-
rical functions are used:
dν ∼ sin(dν); (15.3)
cos(dν) ∼ 1
These simplifications are the core reasons why the change occurs only in the
perpendicular direction (dν << 1). The change of the velocity in the flow direction,
dx is
dx = (U + dU ) cos ν − U = dU (15.4)
In the same manner, the velocity perpendicular to the flow, dy, is
dy = (U + dU ) sin(dν) = U dν (15.5)
The tan µ is the ratio of dy/dx (see Figure (15.3))
dx dU
tan µ = = (15.6)
dy U dν
15.2. GEOMETRICAL EXPLANATION 301

The ratio dU/U was shown to be

dU dM 2
= ¡ ¢ (15.7)
U 2M 2 1 + k−1
2 M
2

Combining equations (15.6) and (15.7) transforms it into

√
M 2 − 1dM 2
dν = − ¡ ¢ (15.8)
2M 1 + k−1
2
2 M
2

After integration of equation (15.8) becomes

r r
k+1 k−1 p
ν(M ) = − tan−1 (M 2 − 1) + tan−1 (M 2 − 1) + constant
k−1 k+1
(15.9)
The constant can be chosen in a such a way that ν = 0 at M = 1.

15.2.1 Alternative Approach to Governing Equations

In the previous sec-
back
tion, a simplified ver- Mach
line r Ur
sion was derived based Front
on geometrical argu- Uθ
Mach
line
ments. In this section, θ

a more rigorous ex-

planation is provided.
It must be recognized
that here the cylindrical
coordinates are advan-
tageous because the Fig. -15.4: The schematic of the coordinate based on the math-
flow turns around a sin- ematical description.
gle point.
For this coordinate system, the mass conservation can be written as
∂ (ρrUr ) ∂ (ρUθ )
+ =0 (15.10)
∂r ∂θ
The momentum equations are expressed as
∂Ur Uθ ∂Ur Uθ 2 1 ∂P c2 ∂ρ
Ur + − =− =− (15.11)
∂r r ∂θ r ρ ∂r ρ ∂r
and
∂Uθ Uθ ∂Uθ Uθ Ur 1 ∂P c2 ∂ρ
Ur + − =− =− (15.12)
∂r r ∂θ r rρ ∂θ rρ ∂θ
302 CHAPTER 15. PRANDTL-MEYER FUNCTION

If the assumption is that the flow isn’t a function of the radius, r, then all the
derivatives with respect to the radius will vanish. One has to remember that when
r enters to the function, like the first term in the mass equation, the derivative isn’t
zero. Hence, the mass equation is reduced to

∂ (ρUθ )
ρUr + =0 (15.13)
∂θ
Equation (15.13) can be rearranged as transformed into
µ ¶
1 ∂Uθ 1 ∂ρ
− Ur + = (15.14)
Uθ ∂θ ρ ∂θ

The momentum equations now obtain the form of

Uθ ∂Ur Uθ 2
− =0
rµ ∂θ r¶
∂Ur
Uθ − Uθ = 0 (15.15)
∂θ

Uθ ∂Uθ Uθ Ur c2 ∂ρ
− =−
r ∂θ r rρ ∂θ
µ ¶
∂Uθ c2 ∂ρ
Uθ − Ur = − (15.16)
∂θ ρ ∂θ
1 ∂ρ
Substituting the term ρ ∂θ from equation (15.14) into equation (15.16) results in
µ ¶ µ ¶
∂Uθ c2 ∂Uθ
Uθ − Ur = Ur + (15.17)
∂θ Uθ ∂θ
or
µ ¶ µ ¶
∂Uθ ∂Uθ
Uθ 2 Ur + = c2 Ur + (15.18)
∂θ ∂θ

And an additional rearrangement results in

µ ¶
¡ 2 2
¢ ∂Uθ
c − Uθ Ur + =0 (15.19)
∂θ

From equation (15.19) it follows that

Uθ = c (15.20)

It is remarkable that the tangential velocity at every turn is at the speed of sound!
It must be pointed out that the total velocity isn’t at the speed of sound, but only
15.2. GEOMETRICAL EXPLANATION 303

the tangential component. In fact, based on the definition of the Mach angle, the
component shown in Figure (15.3) under Uy is equal to the speed of sound, M = 1.
After some additional rearrangement, equation (15.15) becomes
µ ¶
Uθ ∂Ur
− Uθ = 0 (15.21)
r ∂θ
If r isn’t approaching infinity, ∞ and since Uθ 6= 0 leads to
∂Ur
= Uθ (15.22)
∂θ
In the literature, these results are associated with the characteristic line. This
analysis can be also applied to the same equation when they are normalized by
Mach number. However, the non–dimensionalization can be applied at this stage
as well.
The energy equation for any point on a stream line is
Uθ 2 + Ur 2
h(θ) + = h0 (15.23)
2
Enthalpy in perfect gas with a constant specific heat, k, is
c(θ)2
z }| {
k
z}|{
R 1 Cp c2
h(θ) = Cp T = Cp T = RT = (15.24)
R (k − 1) Cv k−1
and substituting this equality, equation (15.24), into equation (15.23) results in
c2 Uθ 2 + Ur 2
+ = h0 (15.25)
k−1 2
Utilizing equation (15.20) for the speed of sound and substituting equation (15.22)
which is the radial velocity transforms equation (15.25) into
¡ ∂U ¢2 ¡ ∂U ¢2
∂θ
r r
+ Ur 2
+ ∂θ = h0 (15.26)
k−1 2
After some rearrangement, equation (15.26) becomes
µ ¶2
k + 1 ∂Ur
+ Ur 2 = 2h0 (15.27)
k − 1 ∂θ
Note that Ur must be positive. The solution of the differential equation (15.27)
incorporating the constant becomes
à r !
p k−1
Ur = 2h0 sin θ (15.28)
k+1
304 CHAPTER 15. PRANDTL-MEYER FUNCTION

which satisfies equation (15.27) because sin2 θ+cos2 θ = 1. The arbitrary constant
in equation (15.28) is chosen such that Ur (θ = 0) = 0. The tangential velocity
obtains the form
r à r !
∂Ur k − 1p k−1
Uθ = c = = 2 h0 cos θ (15.29)
∂θ k+1 k+1

The Mach number in the turning area is

µ ¶2
Uθ 2 + Ur 2 Uθ 2 + Ur 2 Ur
M2 = = =1+ (15.30)
c2 Uθ 2 Uθ

Now utilizing the expression that was obtained for Ur and Uθ equations (15.29)
and (15.28) results for the Mach number is
à r !
2 k+1 2 k−1
M =1+ tan θ (15.31)
k−1 k+1

or the reverse function for θ is

r Ãr !
k+1 k−1¡ 2 ¢
θ= tan−1 M −1 (15.32)
k−1 k+1

What happens when the upstream Mach number is not 1? That is when
the initial condition for the turning angle doesn’t start with M = 1 but is already
at a different angle. The upstream Mach number is denoted in this segment as
Mstarting . For this upstream Mach number (see Figure (15.2))
q
tan ν = Mstarting 2 − 1 (15.33)

The deflection angle ν, has to match to the definition of the angle that is chosen
here (θ = 0 when M = 1), so

ν(M ) = θ(M ) − θ(Mstarting ) (15.34)

r Ãr !
k+1 k − 1p 2 p
ν(M ) = tan−1 M −1 − tan−1 M2 − 1 (15.35)
k−1 k+1

These relationships are plotted in Figure (15.6).

15.3. THE MAXIMUM TURNING ANGLE 305

15.2.2 Comparison And Limitations between the Two Ap-

proaches
The two models produce exactly the same results, but the assumptions for the con-
struction of these models are different. In the geometrical model, the assumption
is that the velocity change in the radial direction is zero. In the rigorous model,
it was assumed that radial velocity is only a function of θ. The statement for the
construction of the geometrical model can be improved by assuming that the frame
of reference is moving radially in a constant velocity.
Regardless of the assumptions that were used in the construction of these
models, the fact remains that there is a radial velocity at Ur (r = 0) = constant. At
this point (r = 0) these models fail to satisfy the boundary conditions and some-
thing else happens there. On top of the complication of the turning point, the
question of boundary layer arises. For example, how did the gas accelerate to
above the speed of sound when there is no nozzle (where is the nozzle?)? These
questions are of interest in engineering but are beyond the scope of this book (at
least at this stage). Normally, the author recommends that this function be used
everywhere beyond 2-4 the thickness of the boundary layer based on the upstream
length.
In fact, analysis of design commonly used in the industry and even questions
posted to students show that many assume that the turning point can be sharp. At
a small Mach number, (1 + ²) the radial velocity is small ². However, an increase
in the Mach number can result in a very significant radial velocity. The radial ve-
locity is “fed” through the reduction of the density. Aside from its close proximity to
turning point, mass balance is maintained by the reduction of the density. Thus,
some researchers recommend that, in many instances, the sharp point should be
replaced by a smoother transition.

15.3 The Maximum Turning Angle

The maximum turning angle is obtained when the starting Mach number is 1 and
the end Mach number approaches infinity. In this case, Prandtl–Meyer function
becomes
"r #
π k+1
ν∞ = −1
2 k−1
(15.36)

The maximum of the deflection point and the maximum turning point are only
a function of the specific heat ratios. However, the maximum turning angle is much
larger than the maximum deflection point because the process is isentropic.
What happens when the deflection angel exceeds the maximum angle? The
flow in this case behaves as if there is almost a maximum angle and in that region
beyond the flow will became vortex street see Figure (15.5)
306 CHAPTER 15. PRANDTL-MEYER FUNCTION

sl Maximum
ip
li turning
ne

Fig. -15.5: Expansion of Prandtl-Meyer function when it exceeds the maximum angle.

15.4 The Working Equations for the Prandtl-Meyer

Function
The change in the deflection angle is calculated by

ν2 − ν1 = ν(M2 ) − ν(M1 ) (15.37)

15.5 d’Alembert’s Paradox

In ideal inviscid incom-
pressible flows, the move- 3
ment of body does not
1 2
encounter any resistance.
4
This result is known as
d’Alembert’s Paradox, and
θ1 w θ2
this paradox is examined
θ1 θ2
here.
Supposed that a two– 4
dimensional diamond–shape 1 2
body is stationed in a su- 3
personic flow as shown in
Figure (15.7). Again, it is Fig. -15.7: A simplified diamond shape to illustrate the su-
assumed that the fluid is in- personic d’Alembert’s Paradox
viscid. The net force in flow
15.5. D’ALEMBERT’S PARADOX 307
Prandtl-Meyer Angle

100

80
k=1.4

60
θ

40

20

0
1 2 3 4 5 6 7 8 9 10
Mach Number

Fri Jul 8 15:39:06 2005

Fig. -15.6: The angle as a function of the Mach number

direction, the drag, is

³w ´
D=2 (P2 − P4 ) = w(P2 − P4 ) (15.38)
2

It can be observed that only the area that “seems” to be by the flow was
used in expressing equation (15.38). The relation between P2 and P4 is such that
the flow depends on the upstream Mach number, M1 , and the specific heat, k.
Regardless in the equation of the state of the gas, the pressure at zone 2, P2 , is
larger than the pressure at zone 4, P4 . Thus, there is always drag when the flow
is supersonic which depends on the upstream Mach number, M1 , specific heat, k,
and the “visible” area of the object. This drag is known in the literature as (shock)
wave drag.
308 CHAPTER 15. PRANDTL-MEYER FUNCTION

15.6 Flat Body with an Angle of Attack

Previously, the thickness of a body was
shown to have a drag. Now, a body with 1 2
3
zero thickness but with an angle of at-
tack will be examined. As opposed to ℓ

w
the thickness of the body, in addition 5
4 α plane
to the drag, the body also obtains lift. Slip

Again, the slip condition is such that the 7

pressure in region 5 and 7 are the same, 6

and additionally the direction of the ve-

locity must be the same. As before, the
Fig. -15.8: The definition of the angle for the
magnitude of the velocity will be differ-
Prandtl–Meyer function.
ent between the two regions.

15.7 Examples For Prandtl–Meyer Function

Example 15.1:
A wall is included with 20.0◦ an inclination. A flow of air with a temperature of 20◦ C
and a speed of U = 450m/sec flows (see Figure 15.9). Calculate the pressure
reduction ratio, and the Mach number after the bending point. If the air flows in an
imaginary two–dimensional tunnel with width of 0.1[m] what will the width of this
imaginary tunnel after the bend? Calculate the “fan” angle. Assume the specific
heat ratio is k = 1.4.

x1 = 0:1[m℄
U = 450m=se x2 =?
T = 20ÆC 1 2

 = 20ÆM =?

Fig. -15.9: The schematic of Example 15.1

S OLUTION
First, the initial Mach number has to be calculated (the initial speed of sound).
√ √
a= kRT = 1.4 ∗ 287 ∗ 293 = 343.1m/sec
15.7. EXAMPLES FOR PRANDTL–MEYER FUNCTION 309

The Mach number is then

450
M= = 1.31
343.1
this Mach number is associated with
P T ρ
M ν P0 T0 ρ0 µ
1.3100 6.4449 0.35603 0.74448 0.47822 52.6434

The “new” angle should be

ν2 = 6.4449 + 20 = 26.4449◦
and results in
P T ρ
M ν P0 T0 ρ0 µ
2.0024 26.4449 0.12734 0.55497 0.22944 63.4620

Note that P0 1 = P0 2
P2 P01 P2 0.12734
= = = 0.35766
P1 P1 P02 0.35603
The “new” width can be calculated from the mass conservation equation.
r
ρ1 M1 T1
ρ1 x1 M1 c1 = ρ2 x2 M2 c2 =⇒ x2 = x1
ρ2 M2 T2
r
0.47822 1.31 0.74448
x2 = 0.1 × × = 0.1579[m]
0.22944 2.0024 0.55497
Note that the compression “fan” stream lines are note and their function can be
obtain either by numerical method of going over small angle increments. The other
alternative is using the exact solution1 . The expansion “fan” angle changes in the
Mach angle between the two sides of the bend
fan angle = 63.4 + 20.0 − 52.6 = 30.8◦

Reverse the example, and this time the pressure on both sides are given and the
angle has to be obtained2 .

Example 15.2:
Gas with k = 1.67 flows over bend (see Figure 15.2). Compute the Mach number
after the bend, and the bend angle.
1 Itisn’t really different from this explanation but shown in a more mathematical form, due to Landau
and friends. It will be presented in the future version. It isn’t present now because of the low priority to
this issue.
2 This example is for academic understanding. There is very little with practicality in this kind of

problem.
310 CHAPTER 15. PRANDTL-MEYER FUNCTION

M = 1:4
P = 1:2[bar ℄ 1 2
P = 1[bar ℄
M =?

Fig. -15.10: The schematic for the reversed question of example (15.2)

S OLUTION
The Mach number is determined by satisfying the condition that the pressure down-
stream are and Mach given. The relative pressure downstream can be calculated
by the relationship

P2 P2 P1 1
= = × 0.31424 = 0.2619
P0 2 P1 P0 1 1.2

P T ρ
M ν P0 T0 ρ0 µ
1.4000 7.7720 0.28418 0.60365 0.47077 54.4623

With this pressure ratio P̄ = 0.2619 require either locking in the table or using the
enclosed program.

P T ρ
M ν P0 T0 ρ0 µ
1.4576 9.1719 0.26190 0.58419 0.44831 55.5479

For the rest of the calculation the initial condition is used. The Mach number after
the bend is M = 1.4576. It should be noted that specific heat isn’t k = 1.4 but
k = 1.67. The bend angle is

∆ν = 9.1719 − 7.7720 ∼ 1.4◦

∆µ = 55.5479 − 54.4623 = 1.0◦

15.8. COMBINATION OF THE OBLIQUE SHOCK AND ISENTROPIC EXPANSION311

15.8 Combination of the Oblique Shock and Isentropic

Expansion
Example 15.3:
Consider two–dimensional flat thin plate at an angle of attack of 4◦ and a Mach
number of 3.3. Assume that the specific heat ratio at stage is k = 1.3, calculate
the drag coefficient and lift coefficient.

S OLUTION
For M = 3.3, the following table can be obtained:
P T ρ
M ν P0 T0 ρ0 µ
3.3000 62.3113 0.01506 0.37972 0.03965 73.1416

With the angle of attack the region 3 will be at ν ∼ 62.31 + 4 for which the following
table can be obtained (Potto-GDC)

P T ρ
M ν P0 T0 ρ0 µ
3.4996 66.3100 0.01090 0.35248 0.03093 74.0528

On the other side, the oblique shock (assuming weak shock) results in
P0y
Mx My s My w θs θw δ P0 x
3.3000 0.43534 3.1115 88.9313 20.3467 4.0000 0.99676

and the additional information, by clicking on the minimal button, provides

Py Ty P0 y
Mx My w θw δ Px Tx P0 x
3.3000 3.1115 20.3467 4.0000 1.1157 1.1066 0.99676

The pressure ratio at point 3 is

P3 P3 P03 P01 1
= = 0.0109 × 1 × ∼ 0.7238
P1 P03 P01 P1 0.01506
The pressure ratio at point 4 is
P3
= 1.1157
P1
µ ¶
2 2 P4 P3 2
dL = (P4 −P 3 ) cos α = − cos α = (1.1157 − 0.7238) cos 4◦ ∼ .054
kP1 M1 2 kM1 2 P1 P1 1.33.32
µ ¶
2 P4 P3 2
dd = − sin α = (1.1157 − 0.7238) sin 4◦ ∼ .0039
kM1 2 P1 P1 1.33.32
This shows that on the expense of a small drag, a large lift can be obtained. Dis-
cussion on the optimum design is left for the next versions.
312 CHAPTER 15. PRANDTL-MEYER FUNCTION
Psurroundings

Aexit β
A∗

Mjet1
β
Slip lines
expenssion
lines

Fig. -15.11: Schematic of the nozzle and Prandtle–Meyer expansion.

Example 15.4:
To understand the flow after a nozzle consider a flow in a nozzle shown in Figure
15.4. The flow is choked and additionally the flow pressure reaches the nozzle
exit above the surrounding pressure. Assume that there is an isentropic expansion
(Prandtl–Meyer expansion) after the nozzle with slip lines in which there is a the-
oretical angle of expansion to match the surroundings pressure with the exit. The
ratio of exit area to throat area ratio is 1:3. The stagnation pressure is 1000 [kPa].
The surroundings pressure is 100[kPa]. Assume that the specific heat, k = 1.3.
Estimate the Mach number after the expansion.

S OLUTION
The Mach number a the nozzle exit can be calculated using Potto-GDC which
provides

T ρ A P A×P F
M T0 ρ0 A? P0 A∗ ×P0 F∗
1.7632 0.61661 0.29855 1.4000 0.18409 0.25773 0.57478

Thus the exit Mach number is 1.7632 and the pressure at the exit is
P − exit
Pexit = P0 = 1000 × 0.18409 = 184.09[kP a]
P −0
This pressure is higher than the surroundings pressure and additional expansion
must occur. This pressure ratio is associated with a expansion angle that Potto-
GDC provide as

P T ρ
M ν P0 T0 ρ0 µ
1.7632 19.6578 0.18409 0.61661 0.29855 60.4403

The need additional pressure ratio reduction is

Psurroundings Psurroundings Pexit 100
= = × 0.18409 = 0.1
P0 Pexit P0 184.09
15.8. COMBINATION OF THE OBLIQUE SHOCK AND ISENTROPIC EXPANSION313

Potto-GDC provides for this pressure ratio

P T ρ
M ν P0 T0 ρ0 µ
2.1572 30.6147 0.10000 0.51795 0.19307 65.1292

The change of the angle is

∆angle = 30.6147 − 19.6578 = 10.9569

Thus the angle, β is

β = 90 − 10.9569 ∼ 79.0
The pressure at this point is as the surroundings. However, the stagnation pressure
is the same as originally was enter the nozzle! This stagnation pressure has to
go through serious of oblique shocks and Prandtl-Meyer expansion to match the
surroundings stagnation pressure.
End Solution
314 CHAPTER 15. PRANDTL-MEYER FUNCTION
 APPENDIX A
Computer Program

A.1 About the Program

The program is written in a C++ language. This program was used to generate all
the data in this book. Some parts of the code are in FORTRAN (old code especially
for chapters 12 and 13 and not included here.1 . The program has the base class
of basic fluid mechanics and utilities functions to calculate certain properties given
data. The derived class are Fanno, isothermal, shock and others.
At this stage only the source code of the program is available no binary available.
This program is complied under gnu g++ in /Gnu/Linux system. As much support
as possible will be provided if it is in Linux systems. NO Support whatsoever will
be provided for any Microsoft system. In fact even PLEASE do not even try to use
this program under any Microsoft window system.

A.2 Usage
To use the program some information has to be provided. The necessary input
parameter(s), the kind of the information needed, where it has to be in a LATEX
format or not, and in many case where it is a range of parameter(s).

machV The Mach number and it is used in stagnation class

4f L
fldV The D and it is used in Fanno class isothermal class

p2p1V The pressure ratio of the two sides of the tubes

M1V Entrance Mach M1 to the tube Fanno and isothermal classes

1 when will be written in C++ will be add to this program.

315
316 APPENDIX A. COMPUTER PROGRAM

M1ShockV Entrance Mach M1 when expected shock to the tube Fanno and isothermal
classes
FLDShockV FLD with shock in the in Fanno class
4f L
M1fldV both M1 and D are given
P1 4f L
M1fldP2P1V three part info P2 , M1 and D are given
MxV Mx or My

infoStagnation print standard (stagnation) info

infoStandard standard info for (Fanno, shock etc)
infoTube print tube side info for (Fanno, etc) including
infoShock print shock sides info
infoTubeShock print tube info shock main info
infoTubeProfile the Mach number and pressure ratio profiles
infoTubeShockLimits print tube limits with shock

To get the shock results in LATEX of Mx The following lines have to be inserted in
the end of the main function.
int isTex = yes;
int isRange = no;
whatInfo = infoStandard ;
variableName = MxV;
Mx = 2.0 ;
s.makeTable(whatInfo, isRange, isTex, variableName, variableValue);

*******************************************
The following stuff is the same as above/below
if you use showResults with showHeads but the
information is setup for the latex text processing.
You can just can cut and paste it in your latex file.
You must use longtable style file and dcolumn
style files.

*******************************************
\setlongtables
\begin{longtable}
{|D..{1.4}|D..{1.4}|D..{1.4}|D..{1.4}|D..{1.4}|D..{1.4}|D..{1.4}|}
\caption{ ?? \label{?:tab:?}}\\
A.3. PROGRAM LISTINGS 317

\hline
\multicolumn{1}{|c|} {$\rule[-0.1in]{0.pt}{0.3 in}\mathbf{M} $} &
\multicolumn{1}{|c|} {$\mathbf{4fL \over D} $} &
\multicolumn{1}{|c|} {$\mathbf{P \over P^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{P_0 \over {P_0}^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{\rho \over \rho^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{U \over {U}^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{T \over T^{*}} $}

\\\hline

\endfirsthead
\caption{ ?? (continue)} \\\hline
\multicolumn{1}{|c|} {$\rule[-0.1in]{0.pt}{0.3 in}\mathbf{M} $} &
\multicolumn{1}{|c|} {$\mathbf{4fL \over D} $} &
\multicolumn{1}{|c|} {$\mathbf{P \over P^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{P_0 \over {P_0}^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{\rho \over \rho^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{U \over {U}^{*}} $} &
\multicolumn{1}{|c|} {$\mathbf{T \over T^{*}} $}

\\\hline
\endhead
2.176& 2.152& 0.3608& 1.000& 0.5854& 3.773& 0.6164 \\
\hline\end{longtable}

A.3 Program listings

Can be download from www.potto.org.
 CompressibleFlow
basic functions
virtual functions
318

Interpolation (root finding)

LaTeX functions
Representation functions

pipe flow stagnation discontinuity real fluids

common common common
functions functions functions

only
contain

P-M flow

specific
functions

Fanno Isothermal Rayleigh normal shock oblique shock

the actual the actual the actual specific specific
functions functions functions functions functions

Fig. -A.1: Schematic diagram that explains the structure of the program
APPENDIX A. COMPUTER PROGRAM
SUBJECTS INDEX 319

Subjects Index

A entrance Mach number calcula-

adiabatic nozzle, 54 tions, 189, 207
airbag, 232 entropy, 180
angle of attack, 308 shockless, 188
star condition, 182
B Fliegner, 3
Balloon Problem, 243 Fliegner experiment, 9
Bar-Meir’s solution to Oblique shock, friction factor, 13
8
Bernoulli’s equation, 36 G
Gibbs, function, 39
C gravity, 151
chamber controlled volume, 247
clasifications of chambers, 232 H
converging–diverging nozzle, 49 Hydraulic Jump, see discontinuity

D I
d’Alembert’s Paradox, 306 internal energy, 5
Darcy friction factor, 157 intersection of Fanno and Rayleigh
de Laval, Carl Gustaf Patrik, 9 lines, 7
deflection angle, 256 Isothermal Flow, 2, 3, see Shapiro
deflection angle range, 272 flow
deLavel’s nozzle, see de Laval, Garl isothermal flow, 155
Gustaf Patrik entrance issues, 161
detached shock, 278 entrance length limitation, 161
diffuser efficiency, 145 maximum , 4fDL 160
discontinuity, 1 table, 165

E L
Eckert number, 10 large deflection angle, 263
Emanuel’s partial solution to oblique line of characteristic, 303
shock, 8 long pipe flow, 155
External flow, 13
M
F Mach, 3
Fanning Friction factor, 157 maximum deflection angle, 265
fanno maximum turning angle, 305
second law, 177 Moody diagram, 13
Fanno flow, 12 moving shock, 9
fanno flow, 175, 4fDL 179 piston velocity, 110
choking, 180 solution for closed valve, 107
average friction factor, 182 stagnation temperature, 102
320 APPENDIX A. COMPUTER PROGRAM

N shock drag, see wave drag

NACA 1135, 8, 257 Shock in cylendrical coordinates, 115
negative deflection angle, 256 Shock in spherical coordinates, 115
normal components, 258 shock tube, 124
nozzle efficiency, 145 shock wave, 89
perturbation, 98
O solution, 94
Oblique shock star velocity, 96
stability, 296 table
oblique shock basic, 132
conditions for solution, 262 thickness, 99
normal shock, 255 trivial solution, 94
Prandtl–Meyer function, 255 small deflection angles, 276
oblique shock governing equations, sonic transition, 58
259 speed of sound, 4
Oblique shock stability, 8 ideal gas, 37
linear temperature, 39
P liquid, 43
Partially open value, 117 real gas, 39
perpendicular components, 258 solid, 44
piston velocity, 110 star, 53
Prandtl-Meyer flow, 299 steam table, 38
Prandtl-Meyer function two phase, 45
small angle, 299 speed of sound, what, 36
tangential velocity, 302 stagnation state, 49
strong solution, 262
R sub, 129
Rayleigh Flow, 12 supersonic tunnel, 146
negative friction, 217
Rayleigh flow, 217 T
rayleigh flow, 217 table
entrance Mach number, 227 shock choking, 115
second law, 220 shock wave
tables, 221 partial close valve, 122
two maximums, 219 Taylor–Maccoll flow, 276
Romer, see isothermal nozzle throat area, 58

S U
science disputes, 5 Upsteam Mach number, 271
semi rigid chamber, 232
semirigid tank V
limits, 233 von Neumann paradox, 255
Shapiro Flow, 3
Shapiro flow, 13 W
shock angle, 260 weak solution, 262
SUBJECTS INDEX 321

Y
Young’s Modulus, 44

Z
zero deflection angle, 270
322 APPENDIX A. COMPUTER PROGRAM

Authors Index

B R
Boyle, Robert, 4 Rankine, John Macquorn, 6
Rayleigh, 5
C Riemann, 5
Challis, 5 Rouse, 5
Converdill, 10
S
E Shapiro, 4
Eckert, E.R.G, 10 Stodola, 7
Stokes, 5
F
Fanno, Gino Girolamo, 7 T
Taylor, G. I., 7
G
Galileo Galilei, 4 V
Van Karman, 4
H
Henderson, 283 W
Hugoniot, Pierre Henri, 6 Wright brothers, 14

K
Kutta-Joukowski, 14

L
Landau, Lev, 7
Leonardo Da Vinci, 4

M
Mach, Ernest, 5
Menikoff, 283
Mersenne, Marin, 4
Meyer, Theodor, 7
Moody, 5

N
Newton, 4

O
Owczarek, 239

P
Poisson, 5, 6
Prandtl, Ludwig, 4, 14