Full download textbook at textbookfull.

com

A New Hypothesis on the Anisotropic Reynolds

Stress Tensor for Turbulent Flows: Volume II:
Practical Implementation and Applications of
an Anisotropic Hybrid k-omega Shear-Stress
Transport/Stochastic Turbulence Model László
https://textbookfull.com/product/a-new-hypothesis-
Könözsy
on-the-anisotropic-reynolds-stress-tensor-for-
turbulent-flows-volume-ii-practical-
implementation-and-applications-of-an-anisotropic-
hybrid-k-omega-shear-stress-transport-stochastic/

Download more textbook from https://textbookfull.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

A New Hypothesis on the Anisotropic Reynolds Stress

Tensor for Turbulent Flows Volume I Theoretical
Background and Development of an Anisotropic Hybrid k
omega Shear Stress Transport Stochastic Turbulence
Model László Könözsy
https://textbookfull.com/product/a-new-hypothesis-on-the-
anisotropic-reynolds-stress-tensor-for-turbulent-flows-volume-i-
theoretical-background-and-development-of-an-anisotropic-hybrid-
k-omega-shear-stress-transport-stochastic-turbul/

Shear Love 1st Edition M K Moore

https://textbookfull.com/product/shear-love-1st-edition-m-k-
moore/

Coulson and Richardson’s Chemical Engineering, Fourth

Edition: Volume 3A: Chemical and Biochemical Reactors
and Reaction Engineering R. Ravi

https://textbookfull.com/product/coulson-and-richardsons-
chemical-engineering-fourth-edition-volume-3a-chemical-and-
biochemical-reactors-and-reaction-engineering-r-ravi/

Engineering of Thermoplastic Elastomer with Graphene

and Other Anisotropic Nanofillers Abhijit Bandyopadhyay

https://textbookfull.com/product/engineering-of-thermoplastic-
elastomer-with-graphene-and-other-anisotropic-nanofillers-
abhijit-bandyopadhyay/
India and the European Union in a Turbulent World
Rajendra K. Jain

https://textbookfull.com/product/india-and-the-european-union-in-
a-turbulent-world-rajendra-k-jain/

Computational Intelligence Theories Applications and

Future Directions Volume II ICCI 2017 Nishchal K. Verma

https://textbookfull.com/product/computational-intelligence-
theories-applications-and-future-directions-volume-ii-
icci-2017-nishchal-k-verma/

Lifelogging for Organizational Stress Measurement

Theory and Applications Thomas Fischer

https://textbookfull.com/product/lifelogging-for-organizational-
stress-measurement-theory-and-applications-thomas-fischer/

Yoga for Stress Magbook

https://textbookfull.com/product/yoga-for-stress-magbook/

Residual Stress, Thermomechanics & Infrared Imaging,

Hybrid Techniques and Inverse Problems, Volume 9:
Proceedings of the 2015 Annual Conference on
Experimental and Applied Mechanics 1st Edition Sven
Bossuyt
https://textbookfull.com/product/residual-stress-thermomechanics-
infrared-imaging-hybrid-techniques-and-inverse-problems-
volume-9-proceedings-of-the-2015-annual-conference-on-
Fluid Mechanics and Its Applications

László Könözsy

A New Hypothesis
on the Anisotropic
Reynolds Stress Tensor
for Turbulent Flows
Volume II: Practical Implementation
and Applications of an Anisotropic
Hybrid k-omega Shear-Stress Transport/
Stochastic Turbulence Model
Fluid Mechanics and Its Applications

Volume 125

Series Editor
André Thess, German Aerospace Center, Institute of Engineering
Thermodynamics, Stuttgart, Germany

Founding Editor
René Moreau, Ecole Nationale Supérieure d’Hydraulique,
St Martin D’heres Cedex, France
The purpose of this series is to focus on subjects in which fluid mechanics plays a
fundamental role. As well as the more traditional applications of aeronautics,
hydraulics, heat and mass transfer etc., books will be published dealing with topics,
which are currently in a state of rapid development, such as turbulence, suspensions
and multiphase fluids, super and hypersonic flows and numerical modelling
techniques. It is a widely held view that it is the interdisciplinary subjects that will
receive intense scientific attention, bringing them to the forefront of technological
advancement. Fluids have the ability to transport matter and its properties as well as
transmit force, therefore fluid mechanics is a subject that is particulary open to cross
fertilisation with other sciences and disciplines of engineering. The subject of fluid
mechanics will be highly relevant in such domains as chemical, metallurgical,
biological and ecological engineering. This series is particularly open to such new
multidisciplinary domains. The median level of presentation is the first year
graduate student. Some texts are monographs defining the current state of a field;
others are accessible to final year undergraduates; but essentially the emphasis is on
readability and clarity.

Springer and Professor Thess welcome book ideas from authors. Potential
authors who wish to submit a book proposal should contact Dr. Mayra Castro,
Senior Editor, Springer Heidelberg, e-mail: [email protected]
Indexed by SCOPUS, EBSCO Discovery Service, OCLC, ProQuest Summon,
Google Scholar and SpringerLink

More information about this series at http://www.springer.com/series/5980

László Könözsy

A New Hypothesis
on the Anisotropic Reynolds
Stress Tensor for Turbulent
Flows
Volume II: Practical Implementation
and Applications of an Anisotropic Hybrid
k-omega Shear-Stress Transport/Stochastic
Turbulence Model

123
László Könözsy
Centre for Computational
Engineering Sciences
Cranfield University
Cranfield, Bedfordshire, UK

Extra Material to this book can be downloaded on https://www.springer.com/in/book/9783030606022

ISSN 0926-5112 ISSN 2215-0056 (electronic)

Fluid Mechanics and Its Applications
ISBN 978-3-030-60602-2 ISBN 978-3-030-60603-9 (eBook)
https://doi.org/10.1007/978-3-030-60603-9
© Springer Nature Switzerland AG 2021
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
I dedicate this book to the memory of my
father—who was a mechanical engineer and
a mathematician—who passed away during
that period of time when I was working on
this manuscript. I will never forget our
fruitful discussions on scientific questions and
on the meaning of life.
Preface

The subject of this self-contained book is interdisciplinary which encompasses

mathematics, physics, computer programming, analytical solutions and numerical
modelling, industrial Computational Fluid Dynamics (CFD), academic benchmark
problems and engineering applications in conjunction with the research field of
anisotropic turbulence. In other words, this book focuses on theoretical approaches,
computational examples and numerical simulations, including computer program-
ming techniques, to demonstrate the strength of a new hypothesis and anisotropic
turbulence modelling approach for academic benchmark problems and industrially
relevant engineering applications. The reader can learn quickly how to use a new
turbulence model in engineering practice to obtain accurate and reliable numerical
results for elementary and advanced turbulent flow problems where the physics of
anisotropic turbulence is indispensable. This book contains MATLAB codes and C
programming language-based User-Defined Function (UDF) codes which can be
compiled in the ANSYS-FLUENT environment. The computer codes help to
understand and use efficiently a new concept which can also be implemented in any
other software packages. The simulation results are compared to classical analytical
solutions and experimental data taken from the literature. A particular attention is
paid to how to obtain accurate results within a reasonable computational time for
wide range of benchmark problems. The provided examples and programming
techniques help to graduate and postgraduate students, engineers and researchers to
further develop their technical skills and knowledge that have posed a challenge to
many people in the past and present.
The content of Chap. 1 returns to the original roots of turbulence modelling,
which, thus, intentionally goes back in time. For graduate and postgraduate stu-
dents, the understanding of classical analytical approaches to turbulence is indis-
pensable, because the current state-of-the-art turbulence models would not exist
without them. Osborne Reynolds (1883) [17] observed the transition from laminar
to turbulent flow in a pipe. Thus, turbulent flows in long horizontal circular smooth
pipes are one of the most well-known and experimentally studied problems in fluid
mechanics. Therefore, in addition to external flows, axisymmetric turbulent shear
flows in horizontal circular pipes and three-dimensional turbulent flows in a

vii
viii Preface

cylindrical pipe are investigated in Chaps. 4 and 5, respectively. To compare

simulation results of the anisotropic hybrid k-x SST/STM closure model [10] in
Chaps. 4 and 5 with classical analytical solutions for internal flows, the first chapter
focuses on the analytical solutions of the simplified Reynolds momentum equation
for wall-bounded turbulent flows. The introductory chapter is devoted to demystify
the strength and weaknesses of classical analytical solutions including even ele-
mentary intermediate mathematical derivation steps in the investigations which
cannot be found in most textbooks. It means that the full step-by-step mathematical
derivations of classical analytical solutions for the simplified Reynolds equation are
discussed in depth in which cases the hydrodynamic and hydraulic aspects of the
fluid flow are considered together. The derived analytical solutions are relying on
the momentum transfer theory of Prandtl (1925) [15] by employing a first-, second-
and third-order turbulent length-scale function proposed by Prandtl (1933) [16], von
Kármán (1930) [9] and Czibere (2001) [5], respectively. Using the Newton–
Raphson iterative method, the numerical solution of the implicit formula for the
resistance coefficient for smooth circular pipes is explained in detail. For educa-
tional purposes, the analytically derived fully developed dimensionless turbulent
mean velocity profiles have been implemented in a MATLAB code (see Appendix
B). The simplified analytical solutions derived in the first chapter can be used for
code verification and turbulence model validation purposes in an initial stage of a
model development and computer code implementation (see Chaps. 4 and 5). It is
important to note that these classical analytical approaches can predict the physi-
cally correct turbulent shear-stress distribution; however, these zeroth-order (alge-
braic) turbulence models are not capable of predicting the normal Reynolds
stresses. Therefore, the reader must keep in mind that real turbulent flows are
always three dimensional, unsteady and anisotropic.
In Chap. 2, a brief summary of the improved mathematical formulation of the
anisotropic hybrid k-x Shear-Stress Transport (SST)/ Stochastic Turbulence Model
(STM) related to a new hypothesis on the Reynolds stress tensor (2.41) proposed in
the first volume of this book [10] is presented. The aim of the second chapter is to
bring closer a novel anisotropic turbulence modelling approach to the wider
audience. The mathematical formulation of the anisotropic hybrid k-x SST/STM
closure model [10]—in conjunction with the anisotropic Reynolds stress tensor
(2.41)—is relying on the unification of the generalised Boussinesq hypothesis
(deformation theory) [1, 8] and the three-dimensional similarity theory of turbulent
velocity fluctuations [6, 7]. In other words, the anisotropic hybrid k-x SST/STM
closure model assumes that the anisotropic Reynolds stress tensor sRA is related to
a) the product of twice the dynamic eddy viscosity coefficient and the mean
rate-of-strain (deformation) tensor 2lt S; b) the anisotropic distribution of the
dominant turbulent Reynolds shear stress lH HG and c) the turbulent kinetic energy
" #
k term
ð2=3ÞqkI . To support the reader to understand quickly the computer

code implementation of the new hybrid model in Chap. 3, all differences and
Preface ix

similarities between the anisotropic hybrid k-x SST/STM approach [10] and the
k-x SST model of Menter [12, 13] are highlighted in Chap. 2. Those derivations
and equations are summarised and discussed briefly which are relevant to com-
putational purposes.
The content of Chap. 3 focuses on the computer code implementation of the
anisotropic hybrid k-x Shear-Stress Transport/Stochastic Turbulence Model
(SST/STM) including computer programming aspects. Since graduate and post-
graduate students do not necessarily have extensive experience in anisotropic tur-
bulence modelling, including how to implement a new anisotropic engineering
turbulence model in an in-house, open-source or commercial software package, the
third chapter has a unique feature from this point of view. In the first part of Chap. 3,
the numerical computation of the elements of the symmetrical anisotropic similarity
tensor H (2.4) has been discussed through a modified version of the Stochastic
Turbulence Model (STM) of Czibere [6, 7]. The importance of the anisotropic
similarity tensor H (2.4) and its modified deviatoric part H H (2.30) is to provide
physically correct model constants to describe the mechanically similar local velocity
fluctuations (2.1) related to the new anisotropic Reynolds stress tensor (2.41)
(see Chap. 2). As a practical approach, the computation of the scalar elements of the
symmetrical anisotropic similarity tensor H (2.4) is explained by the implementation
of an example MATLAB code. The difference between the implemented STM and
the original STM of Czibere [6, 7] is that the Bradshaw constant a1 [2, 3, 12] has been
considered in the modified STM for the convergence criterion instead of the von
Kármán constant j. In the second part of Chap. 3, the implementation of the aniso-
tropic hybrid k-x SST/STM closure model [10] has been described through a
C programming language-based User-Defined Functions (UDF) code which can be
compiled in the ANSYS-FLUENT environment. This software package is widely
used for teaching engineering turbulence modelling, because already existing clas-
sical and advanced turbulence modelling approaches are available in this software
package. Furthermore, it is possible to implement new anisotropic turbulence models
through UDF C codes. However, a step-by-step guidance, tutorial and UDF C
computer code examples are still missing from the literature to support the reader to
understand quickly how to implement a new anisotropic turbulence model through
UDF C functions in the ANSYS-FLUENT environment. Therefore, the second part
of Chap. 3 is intended to explain the UDF C code of the new anisotropic hybrid k-x
SST/STM closure model [10] including the discussion on computer programming
techniques to teach how to implement a new anisotropic turbulence modelling
approach in the ANSYS-FLUENT environment. The additional source terms of the
anisotropic hybrid k-x SST/STM closure model [10] are added to k-x SST model of
Menter [12, 13, 14] in conjunction with the scalar momentum equations and the
additional production terms of the turbulent kinetic energy k and specific dissipation
rate x transport equations, respectively. The reader can gain experience how to use
existing turbulence models and learn how to implement a new anisotropic turbulence
model which can be implemented in any other software packages. Since there is no
practical guide currently available in the literature about how to implement
x Preface

technically an anisotropic turbulence model through UDF C functions in the

ANSYS-FLUENT environment, the objective of Chap. 3 is to fill the knowledge gap
in this subject area.
In Chap. 4, two-dimensional computational examples are presented using the
anisotropic hybrid k-x Shear-Stress Transport/Stochastic Turbulence Model
(SST/STM) [10]. It is important to highlight that the physics of turbulence is always
a three-dimensional phenomenon and strictly speaking two-dimensional turbulence
does not exist in nature at all. However, when a Reynolds-Averaged Navier–Stokes
(RANS) engineering turbulence model is under development, its validation for
two-dimensional well-established classical benchmark problems is still recom-
mended and is in the centre of interest. Therefore, numerical simulations have been
performed for turbulent flows (a) over a flat plate with zero pressure gradient
(Klebanoff problem), (b) over a NACA 0012 airfoil (NASA test problem), (c) in
axisymmetric straight smooth circular pipes at eight different Reynolds numbers,
(d) in a rotationally symmetric coaxial curved duct and (e) over a plane
backward-facing step, respectively. The description of these two-dimensional
numerical simulations can be considered as tutorials for solving simple anisotropic
turbulent flow benchmark problems with the use of the ANSYS-FLUENT R19.1
software package. In Chap. 4, all simulation results have been compared to
experimental data taken from the literature. For two-dimensional flows, the
dimensionless scalar weight parameter lH ð0\jlH j  1Þ and the anisotropic scale
factors kH H
11 and k22 in the modified deviatoric similarity tensor H
H
(3.99) of the
anisotropic hybrid k-x SST/STM closure model [10] are calibrated and validated
related to measurements. The simulation results indicate that the physical behaviour
of the turbulent mean flow is captured correctly with the isotropic k-x SST model
of Menter [12, 13, 14] and the anisotropic hybrid k-x SST/STM closure model [10]
as well. However, further improvements can be achieved with the anisotropic
approach in terms of predicting the physically correct Reynolds stress distributions
in the framework of RANS turbulence modelling. The objective of the fourth
chapter is to show the capabilities of the anisotropic approach. Therefore, 156
simulation results are presented and analysed in Chap. 4.
In the short closing Chap. 5, three-dimensional simulation results have been
presented with the anisotropic hybrid k-x Shear-Stress Transport/Stochastic
Turbulence Model (SST/STM) [10] to make an attempt and a first step towards
more complex real-world engineering applications. Therefore, the objective of the
fifth chapter is to shed light on possible research areas where further improvements
can be made as a future work on the new anisotropic turbulence modelling
approach investigated in this book. Numerical simulations have been performed for
three-dimensional turbulent flows (a) in a horizontal smooth cylindrical pipe,
(b) over a NACA 0013 wing and (c) over the Jetstream 31 aircraft, respectively. In
Chap. 5, all simulation results have been compared to the k-x SST model [12, 13,
14] and experimental data taken from the literature. The dimensionless scalar
weight parameter lH ð0\jlH j  1Þ and the anisotropic scale factors kH H H
11 , k22 and k33
H
in the modified deviatoric similarity tensor H (3.99) have been calibrated and
Preface xi

validated for three-dimensional turbulent flows related to the experimental work of

Laufer [11] in a straight smooth cylindrical pipe. In that case, the anisotropic hybrid
k-x SST/STM approach [10] captured the physically correct anisotropic Reynolds
stress distributions for y þ ’ 65. The simulation results suggest that the numerical
values of these scale factors have to be higher in the near-wall region ðy þ ’ 65Þ
than in the inertial sublayer to obtain the accurate order of magnitude of the ani-
sotropic Reynolds stresses. For the NACA 0013 wing and the Jetstream 31 aircraft
external flows, the lift coefficient cl (4.60) values were under-predicted with the k-x
SST [12, 13, 14] and the anisotropic hybrid k-x SST/STM [10] models compared
to the experimental data in [4, 18]. Therefore, further investigations on the math-
ematical properties of the modified deviatoric similarity tensor H H (3.97) related to
the anisotropic scale factors kH H H
11 , k22 and k33 are recommended. The possible future
work on the engineering research area of anisotropic turbulence modelling is
addressed in the end of Chap. 5.

Cranfield, UK Dr. László Könözsy

July 2020

References

1. Boussinesq J (1877) Théorie de l'Ecoulement tourbillant. Mem Présentés par Divers Savants
Acad Sci Inst Fr 23:46–50
2. Bradshaw P (1967) The Turbulence Structure of Equilibrium Boundary Layers. J Fluid Mech
29(Part 4):625–645
3. Bradshaw P, Ferriss DH, Atwell NP (1967) Calculation of Boundary-Layer Development
Using the Turbulent Energy Equation. J Fluid Mech 28(Part 3):593–616
4. Casadei L, Könözsy L, Lawson NJ (2019) Unsteady Detached-Eddy Simulation (DES) of the
Jetstream 31 Aircraft in One Engine Inoperative (OEI) Condition with Propeller Modelling.
Aerosp Sci Technol 91:287–300
5. Czibere T (2001 a ) Researches on Turbulence 1991-2001 (in Hungarian), Lecture Notes for
PhD Students, Manuscript, University of Miskolc, Department of Fluid and Heat Engineering,
Miskolc-Egyetemváros, H-3515, Hungary
6. Czibere T (2001b) Three Dimensional Stochastic Model of Turbulence. J Comput Appl Mech
2(1):7–20
7. Czibere T (2006) Calculating Turbulent Flows Based on a Stochastic Model. J Comput Appl
Mech 7(2):155–188
8. Davidson PA (2004) Turbulence. An Introduction for Scientists and Engineers. Oxford
University Press Inc.: New York, USA, ISBN 978-0-19-852949-1
9. von Kármán T (1931) Mechanical Similitude and Turbulence. National Advisory Committee
for Aeronautics (N A C A) Technical Memorandum No 611, Washington, USA; English
Translation by Vanier, J pp 1–21
10. Könözsy L (2019) A New Hypothesis on the Anisotropic Reynolds Stress Tensor for
Turbulent Flows, Volume I: Theoretical Background and Development of an Anisotropic
Hybrid k-omega Shear-Stress Transport/Stochastic Turbulence Model, Fluid Mechanics and
Its Application, vol 120, 1st edn. Springer International Publishing, Springer Nature
Switzerland AG, ISBN 978-3-030-13542-3
xii Preface

11. Laufer J (1954) The Structure of Turbulence in Fully Developed Pipe Flow. NACA Technical
Report 1174
12. Menter FR (1992) Improved Two-Equation k-x Turbulence Models for Aerodynamic Flows.
NASA Technical Memorandum 103975 pp 1–31
13. Menter FR (1994) Improved Two-Equation Eddy-Viscosity Turbulence Models for
Engineering Applications. AIAA J 32(8):1598–1605
14. Menter FR, Kuntz M, Langtry R (2003) 4th Internal Symposium, Turbulence, Heat and Mass
Transfer, 2003, Antalya, Turkey, vol 4, Begell House, Inc., New York, Wallingford, chap Ten
Years of Industrial Experience with the SST Turbulence Model, pp 625–632. ISBN
1567001963
15. Prandtl L (1925) Über die ausgebildete Turbulenz. ZAMM 5:136–139
16. Prandtl L (1933) Neuere Ergebnisse der Turbulenzforschung. Zeitschrift des Vereines deut-
scher Ingenieure 7(5):105–114, English translation: NACA-TM-720, 1933
17. Reynolds O (1883) An Experimental Investigation of the Circumstances Which Determine
Whether the Motion of Water Shall Be Direct or Sinuous, and of the Law of Resistance in
Parallel Channels. Phil Trans R Soc Lond 174:935–982
18. Teschner TR (2012) A Comparative Study between Winglet and Raked Wingtip Wing
Configurations. BSc Thesis, Hamburg University of Applied Sciences, Department of
Automotive and Aeronautical Engineering, Hamburg, Germany pp 1–62
Acknowledgements

I would like to give a special thanks to Prof. Tibor Czibere (former Minister of
Education of Hungary, 1988–1989) who taught me analytical solution methods of
fluid mechanics and thermodynamics for wide range of engineering problems and
other complex physical applications. I would like to acknowledge the support of
Prof. Karl W. Jenkins (Head of Centre for Computational Engineering Sciences at
Cranfield University) who encouraged me to establish the M.Sc. in Aerospace
Computational Engineering (M.Sc. in ACE) course what I am currently leading as a
Course Director at Cranfield University in the United Kingdom. Therefore, I have
continuous opportunity to work together with excellent postgraduate students who
are interested in the development of the next generation of anisotropic turbulence
models. My special thanks go to one of my former supervisors, Prof. Árpád Á. Fáy,
who gave me invaluable advices during the preparation of this manuscript. I would
like to acknowledge the encouragement of Dr. Tamás István Józsa, who was one of
my former M.Sc. students and who is currently undertaking a Postdoctoral
Research Fellow position at Oxford University, because his critical way of thinking
on wall-bounded turbulent flows inspired me to further extend the content of Chaps.
1, 4 and 5, respectively. Furthermore, my special thanks go to one of my former
M.Sc. and Ph.D. students, Dr. Jamie Francis Townsend, who wrote his Ph.D. thesis
in the field of computational relativistic hydrodynamics, and in addition to his
ongoing very intense research work, Dr. Townsend was reading the manuscript
carefully and gave me invaluable advices during the preparation of this book.
The present author was a Visiting Research Fellow from October to December of
2003 at the University of Siegen, Institute of Fluid- and Thermodynamics, in
Germany working on the mathematical modelling of anisotropic wall-bounded tur-
bulent flows in curved computational domains. Therefore, I would like to acknowl-
edge the sponsorship of the DAAD-MÖB 193/2002-2003 German-Hungarian
exchange programme as well as the research work with Prof. Dr.-Ing. habil. Szilárd
Szabó (former Head of Department of Fluid and Heat Engineering) from the
University of Miskolc, Hungary and Prof. Dr.-Ing. habil. Wolfram Frank, Dr.-Ing.
Jörg Franke and Dr.-Ing. Gábor Janiga from the University of Siegen, Germany.

xiii
xiv Acknowledgements

I have to admit that the constructive comments and remarkable questions of

Prof. Tamás Lajos, Prof. Aurél Galántai and Dr. Gergely Kristóf during my Ph.D.
viva in 2004 in Hungary boosted my mind for many years and inspired me to work
further on the theoretical background and practical aspects of the research area of
anisotropic turbulence modelling. Certainly and intentionally, I would have liked to
prove or disprove a couple of their comments over time on the three-dimensional
anisotropic similarity theory of turbulent velocity fluctuations. Therefore, I would
like to give them a special thanks, because they brought my attention to very
important points in conjunction with the numerical solution of partial differential
equations which are indispensable in the field of turbulence research.
The present author was a Visiting Research Fellow from end of October to
middle of November, 2005 at the Otto von Guericke University Magdeburg
(OVGU), The Laboratory of Fluid Dynamics and Technical Flows, Germany where
he was working on the numerical investigation of the effect of different turbulent
inlet boundary conditions and modelling non-isothermal turbulent flows. Therefore,
I would like to acknowledge the sponsorship of the DAAD-MÖB 34/2005-2006
German-Hungarian exchange programme as well as the research work with
Prof. Dr.-Ing. habil. Szilárd Szabó from the University of Miskolc, Hungary and
Prof. Dr.-Ing. habil. Dominique Thévenin and Dr.-Ing. Gábor Janiga from the Otto
von Guericke University Magdeburg (OVGU), Germany. My special thanks go to
Prof. Dr.-Ing. habil. Szilárd Szabó who provided me extensive Laser Doppler
Velocimetry (LDV) technique-based experimental data for computer code verifi-
cation and turbulence model validation purposes in a rotationally symmetric coaxial
curved duct. Furthermore, the technical discussions with Dr.-Ing. Gábor Janiga on
the experimental aspects of turbulent flows in curved pipes and ducts are greatly
acknowledged.
I would like to acknowledge the fruitful and encouraging discussions with
Prof. Nicholas Lawson (Head of National Flying Laboratory Centre at Cranfield
University) who provided me extensive experimental and computational input data
for simulating the Jetstream 31 aircraft owned by Cranfield University in the United
Kingdom. I would like to thank Prof. Thomas B. Gatski, Prof. Dr. Wolfgang Rodi,
Prof. Michael A. Leschziner and Prof. Dr.-Ing. habil. Suad Jakirlić for their valu-
able advices I received at different scientific conferences or workshops in Europe on
advanced turbulence modelling approaches for complex engineering applications.
I would like to acknowledge the support of Dr. Tom-Robin Teschner who
provided me experimental data, mesh file and operating conditions to the numerical
example for the external turbulent flow over the NACA 0013 wing which are
presented in Chap. 5. I would like to mention that Dr. Teschner was one of my
former M.Sc. and Ph.D. students who was the winner of the Lord Kings Norton
Medal in 2019 which is the prize awarded only for the best performing doctor-
ate student at Cranfield University in the United Kingdom. The unique advices of
Dr. Teschner in terms of computer programming encouraged me to further
improve the content of Chap. 3. I would like to greatly acknowledge the support of
Mr. David Pérez Sánchez, who was working on his M.Sc. thesis project under my
supervision on the implementation, validation and verification of a non-linear eddy
Acknowledgements xv

viscosity model for aerospace applications. Mr. Pérez Sánchez provided me com-
putational meshes and input data for different two-dimensional benchmark prob-
lems. I would like to mention that Mr. Pérez Sánchez was the winner of the
Vice-Chancellor’s prize in 2020 which is awarded annually to the most outstanding
taught course student at Cranfield University and he received the Course Director’s
prize for the best overall achievement on the M.Sc. in Aerospace Computational
Engineering (M.Sc. in ACE) course. The technical discussions with Mr. Pérez
Sánchez inspired me to further improve the accuracy of several numerical simu-
lations presented in this book.
My special thanks go to former Cranfield postgraduate M.Sc. students,
Mr. Adrian Dunne, Mr. Ionut-Andrei Cimpoeru, Mr. Nicolas Ledos, Mr. Albert
Herrando Moraira, Mr. Tony Gasana, Mr. Leonidas Theofilopoulos and
Mr. Hrishabh Chaudhary, because I had that great opportunity to supervise out-
standing postgraduate students in their M.Sc. thesis projects on hybrid turbulence
modelling and numerical method development for aerospace applications. My
special thanks go to Mr. Olaf Artur Popczyk who was working on his M.Sc. thesis
project under my supervision and won the Course Director’s prize for the best
M.Sc. thesis in 2019 on the M.Sc. in Computational Fluid Dynamics (M.Sc. in
CFD) course at Cranfield University in the United Kingdom. Mr. Olaf Artur
Popczyk provided me excellent quality of PGF/TikZ templates in the LaTeX
document environment which made me able to further improve the quality of many
figures presented in this book.
I would like to acknowledge the valuable support of Mr. Jordan Recouly and
Mr. Émile Moireau from Supméca Institute of Mechanics of Paris, in France,
because they provided me high-quality computational meshes during their intern-
ship programme in 2019 at Cranfield University.
My special thanks go to my wife, Mrs. Martina Könözsy-Kovács, my son,
Mr. Alexander Teodor Könözsy, my mother, Mrs. Judit Veress Könözsyné, and my
aunt, Dr. Klára Veress for their continuous support and endless patience when I was
working on the manuscript of this book.
The work presented here was carried out in addition to my university teaching
commitments at the University of Miskolc, Hungary, the University of Siegen, the
Otto von Guericke University Magdeburg, Germany and Cranfield University in the
United Kingdom, respectively. Therefore, I would like to thank again all of my
present and former colleagues and students for their support and encouragement
which created a positive atmosphere at work and motivated me to finish this book.
Contents

1 Introduction to Classical Analytical Solutions for Wall-Bounded

Turbulence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Governing Equations of Incompressible Turbulence . . . . . . . . . . . 4
1.3 Turbulent Flow in Straight Smooth Circular Pipes . . . . . . . . . . . . 6
1.3.1 Mathematical Formulation of the Fully-Developed
Turbulent Flow Problem in Straight Circular Pipes . . . .. 6
1.3.2 The Simplified Reynolds Momentum Equation . . . . . . .. 8
1.3.3 The Constant Pressure Gradient Assumption . . . . . . . . .. 11
1.3.4 Integrated Mathematical Form of the Simplified
Reynolds Momentum Equation . . . . . . . . . . . . . . . . . . .. 13
1.3.5 The Viscous Sublayer and the Definition of the Friction
(Shear) Velocity at the Wall . . . . . . . . . . . . . . . . . . . . .. 16
1.3.6 Hydraulic Loss and the Dimensionless Resistance
Coefficient (Coefficient of Friction) . . . . . . . . . . . . . . . .. 18
1.3.7 The Reynolds Number and Its Related Physical
Quantities for Turbulent Flows in Pipes . . . . . . . . . . . .. 21
1.3.8 Hydrodynamic Entrance (Development) Length for
Fully-Developed Turbulent Flows in Horizontal Pipes . .. 24
1.3.9 Relationships Amongst the Pressure Drop, the Reynolds
Number and the Volume Flow Rate . . . . . . . . . . . . . . .. 25
1.3.10 Mean Velocity Distribution in the Viscous Sublayer . . .. 28
1.3.11 Fluid Flow Regions of the Dimensionless Mean
Velocity Distribution in the Turbulent Boundary
Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 31
1.3.12 Turbulent Velocity Profiles Using a First-Order (Linear)
Mixing-Length Function (Prandtl’s Solution) . . . . . . . . .. 41
1.3.13 The Resistance Law for Smooth Circular Pipes . . . . . . .. 54
1.3.14 Computation of the Resistance Coefficient Using
the Newton–Raphson Iterative Method . . . . . . . . . . . . .. 57

xvii
xviii Contents

1.3.15 An Algorithm for Computing Turbulent Velocity

Profiles in Straight Smooth Circular Pipes . . . . . . . . . ... 59
1.3.16 Turbulent Velocity Profiles Using a Second-Order
Mixing-Length Function (Von Kármán’s Solution) . . ... 62
1.3.17 Turbulent Velocity Profiles Using a Third-Order
Mixing-Length Function (Czibere’s Solution) . . . . . . ... 76
1.3.18 Comparison of Analytically Computed Turbulent
Mean Velocity Profiles with Experimental Data . . . . . ... 89
1.4 Analogy Between Pipe and Channel Flows . . . . . . . . . . . . . . ... 94
1.4.1 Mathematical Formulation of the Fully-Developed
Turbulent Flow Problem in Straight Plane Channels . . ... 94
1.4.2 The Simplified Reynolds Momentum Equation
for Two-Dimensional Flows in Channels . . . . . . . . . . ... 97
1.4.3 Integrated Mathematical Form of the Simplified
Reynolds Momentum Equation for Plane Flows . . . . . . . . 101
1.4.4 The Definition of the Friction (Shear) Velocity . . . . . . . . 103
1.4.5 The Definition of the Reynolds Number for Turbulent
Flows in Straight Plane Channels . . . . . . . . . . . . . . . . . . 105
1.4.6 The Resistance Law for Smooth Channels . . . . . . . . . . . . 106
1.4.7 Remarks on Channel Flow Computations . . . . . . . . . . . . 107
1.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
2 The Anisotropic Hybrid k-x SST/Stochastic Turbulence Model . . . . 115
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
2.1.1 Three-Dimensional Map Space and the Similarity
Tensor of the Mechanically Similar Local Velocity
Fluctuations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
2.1.2 A New Hypothesis on the Anisotropic Reynolds Stress
Tensor for Incompressible Turbulent Flows . . . . . . . . . . . 119
2.1.3 The Matrix Form of the New Hypothesis on the
Anisotropic Reynolds Stress Tensor . . . . . . . . . . . . . . . . 124
2.2 Mathematical Formulation and Derivations . . . . . . . . . . . . . . . . . 127
2.2.1 Tensor Divergence of the New Anisotropic Reynolds
Stress Tensor for Incompressible Turbulent Flows . . . . . . 128
2.2.2 The New Turbulent Kinetic Energy Production Term . . . . 129
2.2.3 Dissipation and Diffusion Terms of the Turbulent
Kinetic Energy Transport Equation . . . . . . . . . . . . . . . . . 131
2.2.4 The New Specific Dissipation Rate Production Term . . . . 132
2.2.5 Dissipation and Diffusion Terms of the Transport
Equation of the Specific Dissipation Rate . . . . . . . . . . . . 133
2.3 Governing Equations of the Anisotropic Hybrid k-x SST/STM
Closure Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Contents xix

2.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
3 Implementation of the Anisotropic Hybrid k-x SST/STM
Closure Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.2 Implementation of the Three-Dimensional Stochastic Turbulence
Model (STM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.2.1 An Algorithm for Computing the Elements
of the Anisotropic Similarity Tensor and Its Deviatoric
Part . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.2.2 A MATLAB Code Implementation of the Stochastic
Turbulence Model (STM) and Its Explanation . . . . . . . . . 165
3.2.3 An Example How to Compute the Elements of the
Anisotropic Similarity Tensor and Its Deviatoric Part . . . . 176
3.3 A Computer Code for the Anisotropic Hybrid k-x SST/STM
Closure Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
3.3.1 A C Code Based User-Defined Function (UDF)
Implementation of the Anisotropic k-x SST/STM
Closure Model in the ANSYS-FLUENT Environment . . . 179
3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
4 Two-Dimensional Simulations with an Anisotropic
Hybrid k-x SST/STM Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
4.2 Two-Dimensional Classical Benchmark Problems . . . . . . . . . . . . . 220
4.2.1 Turbulent Flow over a Flat Plate with Zero Pressure
Gradient (The Klebanoff Problem) . . . . . . . . . . . . . . . . . 222
4.2.2 Turbulent Flow over a NACA 0012 Airfoil . . . . . . . . . . . 252
4.2.3 Axisymmetric Turbulent Shear Flows in Straight
Circular Pipes at Low and High Reynolds Numbers . . . . . 279
4.2.4 Turbulent Shear Flow in a Rotationally Symmetric
Coaxial Curved Duct with Varying Cross Sections . . . . . 310
4.2.5 Turbulent Flow over a Plane Backward-Facing Step . . . . 330
4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
5 Three-Dimensional Simulations with an Anisotropic
Hybrid k-x SST/STM Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
5.2 Turbulent Flow in a Horizontal Cylindrical Pipe . . . . . . . . . . . . . 362
5.2.1 Mean Pressure and Velocity Distributions . . . . . . . . . . . . 364
5.2.2 Anisotropic Reynolds Stress Distributions . . . . . . . . . . . . 370
xx Contents

5.2.3 Turbulent Energy Production and Dissipation . . . . . . . . . 376

5.2.4 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
5.3 Turbulent Flow over a NACA 0013 Wing . . . . . . . . . . . . . . . . . . 381
5.3.1 The Experimental Background . . . . . . . . . . . . . . . . . . . . 382
5.3.2 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
5.3.3 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
5.4 Turbulent Flow over the Jetstream 31 Aircraft . . . . . . . . . . . . . . . 394
5.5 Summary and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402

Appendix A: Supplementary Mathematical Derivations . . . . . . . . . . . . . 405

Appendix B: Supplementary Computer Codes . . . . . . . . . . . . . . . . . . . . . 435
Appendix C: Digitalised Experimental Data . . . . . . . . . . . . . . . . . . . . . . . 463
Acronyms

2D Two Dimensional
3D Three Dimensional
AoA Angle of Attack
BC Boundary Condition
CFD Computational Fluid Dynamics
CPU Central Processing Unit
DES Detached Eddy Simulation
DNS Direct Numerical Simulation
FFT Fast Fourier Transform
GCI Grid Convergence Index
LDV Laser Doppler Velocimetry
LES Large Eddy Simulation
MAC Mean Aerodynamic Chord
MG Multigrid
MUSCL Monotonic Upstream-centered Scheme for Conservation Laws
NACA National Advisory Committee for Aeronautics
NASA National Aeronautics and Space Administration
NFLC National Flying Laboratory Centre, Cranfield University, UK
QUICK Quadratic Upstream Interpolation for Convective Kinematics
RANS Reynolds-Averaged Navier–Stokes
RSM Reynolds Stress Model
RST Reynolds Stress Transport
RSTM Reynolds Stress Transport Model
SAS Scale-Adaptive Simulation
SIMPLE Semi-Implicit Pressure Linked Equations
SIMPLEC Semi-Implicit Pressure Linked Equations-Consistent
SST Shear-Stress Transport

xxi
xxii Acronyms

STM Stochastic Turbulence Model

TAS True Air Speed
TVD Total Variation Diminishing
URANS Unsteady Reynolds-Averaged Navier–Stokes
Chapter 1
Introduction to Classical Analytical
Solutions for Wall-Bounded Turbulence

Scientists study the world as it is, engineers create the world

that never has been.
—Theodore von Kármán, 1962

Abstract The content of this introductory chapter returns to the original roots of tur-
bulence modelling and intentionally goes back in time as well. For graduate students,
the understanding of classical analytical approaches to turbulence is indispensable,
because the current state-of-the-art turbulence models would not exist without them.
Osborne Reynolds [94] observed the transition from laminar to turbulent flow in a
pipe. Thus, turbulent flows in long straight smooth circular pipes are one of the most
well-known and experimentally studied problems in fluid mechanics. Therefore, in
addition to external flows, axisymmetric turbulent shear flows in straight circular
pipes and three-dimensional turbulent flows in a cylindrical pipe are investigated
in Chaps. 4 and 5, respectively. To compare simulation results of the anisotropic
hybrid k-ω SST/STM closure model [61] in Chaps. 4 and 5 with classical analyti-
cal solutions for internal flows, this chapter focuses on the analytical solutions of
the simplified Reynolds momentum equation for wall-bounded turbulent flows. This
includes intermediate mathematical derivation steps in the investigations which can-
not be found in most textbooks. The derived analytical solutions are relying on the
momentum transfer theory of Prandtl [87] by employing first-, second- and third-
order turbulent length-scale functions proposed by Prandtl [88], von Kármán [55] and
Czibere [33], respectively. For educational purposes, the analytically derived fully-
developed dimensionless turbulent mean velocity profiles have been implemented
in a MATLAB code (see Appendix B). The simplified analytical solutions derived
in this chapter can be used for code verification and turbulence model validation
purposes in an initial stage of a turbulence model development and computer code
implementation (see Chaps. 4 and 5). However, the reader must keep in mind that
real turbulent flows are always three-dimensional, unsteady and anisotropic.

© Springer Nature Switzerland AG 2021 1

L. Könözsy, A New Hypothesis on the Anisotropic Reynolds Stress Tensor
for Turbulent Flows, Fluid Mechanics and Its Applications 125,
https://doi.org/10.1007/978-3-030-60603-9_1
2 1 Introduction to Classical Analytical Solutions for Wall-Bounded Turbulence

Keywords Incompressible turbulent flows · Reynolds momentum equation ·

Momentum transfer theory · Analytical solutions · Turbulent flows in long straight
circular pipes · Wall-bounded turbulence

1.1 Introduction

Osborne Reynolds [94] first observed experimentally the transition from laminar
to turbulent flow in a pipe at the Victoria University of Manchester in the United
Kingdom. The analytical and numerical solutions of the Reynolds momentum equa-
tion have been at the centre of research interest since Reynolds [95] introduced the
Reynolds-Averaged Navier–Stokes (RANS) equations including the tensor diver-
gence of the anisotropic Reynolds stress tensor on the right hand side of the Navier–
Stokes equations. One of the main questions in the research field of turbulence is
how to describe mathematically the physically correct fluctuating behaviour of the
elements of the symmetrical anisotropic Reynolds stress tensor? The final answer to
this question as a heart of turbulence modelling is still open for researchers despite
of the fact that simplified and sophisticated turbulence models were developed from
the beginning of the twentieth century.
To overcome the difficulties of the analytical solution to the Reynolds momen-
tum equation, three main theories were developed in the beginning of the twentieth
century [98]. Taylor [102] and Prandtl [86, 87] proposed the momentum transfer
(mixing-length) theory considering different theoretical aspects. Von Kármán [53–
55, 59, 60] derived the two-dimensional similarity theory of turbulent oscillatory
motions, and Taylor [103] developed the vorticity transport theory of turbulent flows.
These three theories laid the foundations for the simplified analytical solutions of
the Reynolds equation of motion for incompressible turbulent wall-bounded flows in
plane channels and long straight circular pipes. Furthermore, these theories provided
the theoretical background for many turbulence models that were born in the second
half of the twentieth century.
Osborne Reynolds [94] observed the transition from laminar to turbulent flow in
a pipe. Therefore, turbulent flows in a long straight circular pipe are one of the most
well-known and experimentally studied problems in fluid mechanics. In the first
half of the twentieth century, e.g., Stanton [99], Fritsch [42], Nikuradse [81–83],
Reichardt [92] and Laufer [64, 65] experimentally investigated turbulent shear flows
in straight pipes and/or channels and their dataset is still used in the engineering
practice.
It is important to mention that Miller [73] first questioned the reliability and
credibility of the experimental data of Nikuradse [82] for smooth pipes. The rea-
son was that when Nikuradse [82] produced the dimensionless velocity profile data
for smooth pipes, he added a numerical value of 7 to his y + values without any
explanation as highlighted by Beattie [13]. The criticism against Nikuradse [82] was
that he altered his experimental data to match them to theoretical results (see more
details on this subject in [13]). Despite of this criticism against the data measured
1.1 Introduction 3

by Nikuradse [82], most of the classical theoretical achievements are still relying on
the experimental work of Nikuradse [82] for smooth pipes.
Another important point is that the classical semi-empirical theories [53–55, 59,
60, 87, 102, 103] can only be used for two-dimensional flows in which case a sim-
ple geometry is considered. However, two-dimensional turbulence does not exist
in reality, because turbulence is always a three-dimensional fluid flow phenomenon
[33]. For turbulent shear flows, the simplified analytical solutions for the Reynolds
momentum equation may predict the physically correct velocity and shear stress dis-
tributions, however, these analytical approaches fail to predict the physically correct
turbulent fluctuations as pointed out by Shih-I Pai [98]. Furthermore, the isotropic tur-
bulence assumption of Taylor [104–108, 110]—which is a significant simplification
for real turbulent flows—can only be valid when the effect of turbulent shear stresses
is neglected. Strictly speaking, none of these classical theoretical approaches to tur-
bulence are satisfactory nowadays. Therefore, the development of three-dimensional
anisotropic turbulence modelling approaches could be a way to take into account
the real Reynolds stress distribution to capture the behaviour of internal and external
turbulent flows in a physically correct way (see, e.g., Chap. 2).
For graduate and postgraduate students, the understanding of the classical ana-
lytical solutions for the simplified Reynolds momentum equation is indispensable,
because these analytical solutions for wall-bounded turbulent flows can be used for
code verification and turbulence model validation purposes. Relying on the momen-
tum transfer theory [86, 87], analytical solutions for the simplified Reynolds momen-
tum equation, e.g., were derived by von Kármán [55], Prandtl [88] and Czibere [33],
respectively. Taylor [109] derived analytical solutions for wall-bounded turbulent
flows in pipes and between parallel planes based on the vorticity transport theory
[103]. The objective of this chapter is to provide a consistent mathematical deriva-
tion of the momentum transfer theory [86, 87] based classical analytical solutions for
the simplified Reynolds momentum equation for turbulent flows in straight circular
pipes. In other words, this chapter aims to demystify the strengths and weaknesses
of classical analytical approaches for internal turbulent flows. The hydrodynamic
and hydraulic aspects of the fluid flow are considered together in this work, because
these areas of fluid mechanics are not necessarily well-integrated in the literature
[16]. The simulation results performed with the k-ω SST model of Menter [70, 71]
and the anisotropic hybrid k-ω SST/STM closure model [61] are compared to ana-
lytical solutions and experimental data of Nikuradse [82] and Laufer [65] in Chaps. 4
and 5, respectively.
The theoretical achievements in the research field of classical analytical and semi-
empirical theories [53–55, 59, 60, 87, 102, 103] are discussed, e.g., by Pai [98],
Prandtl and Tietjens [89, 90], Schlichting [97], Monin and Yaglom [77, 78], Gruber
and Blahó [44], Cheremisinoff [26], Czibere [32, 33], Rohsenow et al. [96], Bobok
[16, 17], Johnson et al. [49], Hughes and Brighton [47], Lajos [101], Davidson [37],
Potter and Wiggert [85], and Rennels and Hudson [93], White [117] and Çengel and
Cimbala [25].
4 1 Introduction to Classical Analytical Solutions for Wall-Bounded Turbulence

1.2 Governing Equations of Incompressible Turbulence

The general set of Reynolds-averaged governing equations for incompressible tur-

bulent flows has been derived and discussed in [61]. The system of governing equa-
tions consists of the mass conservation (continuity) equation (1.1), the Reynolds-
Averaged Navier–Stokes (RANS) momentum equation (1.2)—which is also called
as the Reynolds momentum equation (1.2)—, and the turbulent kinetic energy trans-
port equation (1.5). For incompressible turbulent flows, the mass conservation (con-
tinuity) equation is
∇ · u = 0, (1.1)

where ∇ is the Hamilton (nabla/del) vector-type differential operator [1, 52], and u is
the mean velocity field. For incompressible flows, the continuity equation (1.1) of the
mean velocity field u has to satisfy the divergence-free (incompressibility) constraint
which means that the divergence of the mean velocity field u has to vanish. The
unsteady RANS (URANS/Reynolds) momentum equation with invariant (Gibbs)
notation can be written as
∂u
ρ + ρu · (∇ ⊗ u) = ρg − ∇ p + μ∇ 2 u + ∇ · τ R A , (1.2)
∂t
where the symbol ‘⊗’ denotes the dyad (tensor) product and the twice underline ‘=’
refers to a second-rank tensor, ρ is the density of the fluid, g is the gravitational body
force vector, p is the mean pressure field, μ is the dynamic viscosity of the fluid,
∇ 2 is the scalar-type second-order Laplace differential operator [1, 52], and τ R A is
the symmetrical anisotropic Reynolds stress tensor of incompressible turbulent flows
which can be given by
⎡ ⎤
u 1 u 1 u 1 u 2 u 1 u 3
⎢ ⎥
τ R A = −ρu ⊗ u = −ρ ⎢       ⎥
⎣ u2u1 u2u2 u2u3 ⎦ . (1.3)

u 3 u 1 u 3 u 2 u 3 u 3

The upper bar notation in Eq. (1.3) denotes Reynolds- or time-averaged physical
quantities [95, 98]. On the left hand side of the Reynolds momentum equation
(1.2), the first term represents the unsteadiness of the fluid flow field and the sec-
ond term is the non-linear advective/convective term which represents the spatial
change of the momentum. The velocity gradient tensor ∇ ⊗ u in the non-linear
advective/convective term can be defined by the dyad (tensor) product of the nabla
∇ vector-type differential operator and the velocity vector u which can be written in
a matrix form by
1.2 Governing Equations of Incompressible Turbulence 5
⎡ ∂ ⎤ ⎡ ∂u ∂u 2 ∂u 3 ⎤
1
⎢ ∂ x1 ⎥ ⎢ ∂ x1 ∂ x1 ∂ x1 ⎥
⎢ ⎥ ⎢ ⎥
⎢ ∂ ⎥  ⎢ ∂u ∂u 2 ∂u 3 ⎥
⎢ ⎥ ⎢ 1 ⎥
∇ ⊗u =⎢ ⎥ · u1 u2 u3 = ⎢ ⎥, (1.4)
⎢ ∂ x2 ⎥ ⎢ ∂ x2 ∂ x2 ∂ x2 ⎥
⎢ ⎥ ⎢ ⎥
⎣ ∂ ⎦ ⎣ ∂u ∂u 2 ∂u 3 ⎦
1
∂ x3 ∂ x3 ∂ x3 ∂ x3

where each component is expressed by using the Cartesian index notation. On the
right hand side of the Reynolds momentum equation (1.2), the first term represents
the effect of the gravity field, the second term is the pressure gradient term, the third
term is the viscous diffusion term, and the fourth term is the tensor divergence of
the Reynolds stress tensor (1.3) which is the source of the effects of the Reynolds
stresses. The turbulent kinetic energy transport equation of incompressible turbulent
flows can be given by

∂k
ρ + ρ (u · ∇) k = τ R A · · (∇ ⊗ u) − ρν(∇ ⊗ u ) · · (∇ ⊗ u )T +
∂t

1
+ ∇ · μ∇k − ρu · (u · u ) − u p  , (1.5)
2

where the symbol ‘··’ denotes the double dot scalar (inner) product of two second-
rank tensors, the superscript ‘T ’ stands for the transpose of a tensor, k is the turbulent
kinetic energy, ∇ ⊗ u is the velocity gradient tensor, ν is the kinematic viscosity of
the fluid, u is the Galilean invariant velocity fluctuation vector, and p  represents the
fluctuating pressure field. The turbulent kinetic energy k in the transport equation
(1.5) is the kinetic energy of turbulent fluctuations per unit mass which is the half of
the sum of the diagonal elements of the anisotropic Reynolds stress tensor (1.3) as

1  
3
u · u 1   1
k= = u 1 u 1 + u 2 u 2 + u 3 u 3 = u i · u i ≡ u i u i , (1.6)
2 2 2 i=1 2

where the equivalent symbol ‘≡’ stands for the use of Einstein’s summation con-
vention [40] where one can omit the summation symbol using the Cartesian index
notation [1, 52]. For incompressible turbulent flows, there are more unknowns than
partial differential transport equations in the Reynolds-averaged governing Eqs. (1.1),
(1.2) and (1.5) which is known as the closure problem of turbulence. In order to solve
the general set of Reynolds-averaged governing Eqs. (1.1), (1.2) and (1.5), a hypoth-
esis has to be imposed on the physical description of the anisotropic Reynolds stress
tensor (1.3) (see Chap. 2). To derive simplified analytical solutions for wall-bounded
turbulent shear flows in straight circular pipes, the scalar set of Reynolds-averaged
governing Eqs. (1.1) and (1.2) has been considered in the cylindrical (curvilinear)
coordinate system in Sect. 1.3.
6 1 Introduction to Classical Analytical Solutions for Wall-Bounded Turbulence

1.3 Turbulent Flow in Straight Smooth Circular Pipes

For axisymmetric turbulent shear flows in straight smooth circular pipes, analytical
solutions of the simplified Reynolds momentum equation (1.31) are presented in this
section using the momentum transfer theory of Prandtl [86, 87, 98]. The prepara-
tory mathematical derivations have been carried out in the cylindrical coordinate
system in Sects. 1.3.1–1.3.10. The analytical solutions for the simplified Reynolds
momentum equation (1.31) are relying on first-, second- and third-order turbulent
mixing-length functions proposed by Prandtl [88], von Kármán [55] and Czibere [33],
respectively. The dimensionless turbulent velocity profiles are derived step by step in
Sects. 1.3.12–1.3.17. The analytically derived dimensionless turbulent velocity pro-
files have been implemented in a MATLAB code (see Appendix B) and compared to
the experimental data of Nikuradse [82] and Laufer [65] at eight different Reynolds
numbers in Sect. 1.3.18. The analogy between turbulent flows in straight smooth
circular pipes and plane channels are discussed briefly in Sect. 1.4. The analytical
solutions of the simplified Reynolds equation (1.31) derived in this section can be
used for code verification and turbulence model validation purposes.

1.3.1 Mathematical Formulation of the Fully-Developed

Turbulent Flow Problem in Straight Circular Pipes

For the sake of clarity, the position vector x can be defined with the coordinate
notation r , ϕ, z in the cylindrical coordinate system as

x = x1 · e1 + x2 · e2 + x3 · e3 = r · er + ϕ · eϕ + z · ez , (1.7)

where r = x1 , ϕ = x2 and z = x3 are spatial coordinates, respectively. According

to the Reynolds decomposition [95, 98] of instantaneous physical quantities, the
instantaneous (turbulent) velocity field can be written as

uT = u + u = (u T )r · er + (u T )ϕ · eϕ + (u T )z · ez , (1.8)

where u is the mean velocity field, u is the fluctuating velocity field, and (u T )r , (u T )ϕ ,
(u T )z are the scalar components of the instantaneous (turbulent) velocity field (1.8)
in the cylindrical (curvilinear) coordinate system. The mean velocity field u can be
expressed by
u = u r · er + u ϕ · eϕ + u z · ez , (1.9)

where u r , u ϕ , u z are the scalar components of the mean velocity field u (1.9) and the
fluctuating velocity field u can be defined by

u = u r · er + u ϕ · eϕ + u z · ez , (1.10)
1.3 Turbulent Flow in Straight Smooth Circular Pipes 7

Fig. 1.1 Sketch of a fully-developed axisymmetric turbulent shear flow in a straight smooth circular
pipe where τw denotes the wall shear stress

where u r , u ϕ , u z are the scalar components of the fluctuating velocity field u (1.10) in
the cylindrical (curvilinear) coordinate system. The sketch of an axisymmetric turbu-
lent shear flow in a straight smooth circular pipe—including the graphical represen-
tation of the Reynolds velocity triangles defined by Eqs. (1.8), (1.9) and (1.10)—is
shown in Fig. 1.1.
To derive simplified analytical solutions of the Reynolds momentum equation
(1.2) for fully-developed turbulent shear flows in straight smooth circular pipes, the
scalar set of the governing equations (1.1) and (1.2) has to be simplified through
plausible physical and geometrical assumptions. For turbulent shear flows in straight
circular pipes, it can be assumed as a simplification of the physical problem that the
radial and azimuthal velocity components u r and u ϕ are equal to zero, respectively.
The reason is that the fully-developed turbulent velocity profile u z = u z (r ) in the
axial direction of the circular pipe z can be considered as a single variable function
of the radial spatial coordinate r . Therefore, it is reasonable to ignore the azimuthal
coordinate direction ϕ as a further simplification. The fluctuating velocity compo-
nents u r , u ϕ , u z and the fluctuating pressure field p  can also be considered as a single
variable function of the radial coordinate r . The fully-developed turbulent shear flow
in straight circular pipes can be investigated as an axisymmetric flow on the coor-
dinate plane (r, z), and its physical and geometrical assumptions can be formulated
as
u r = 0, u ϕ = 0, u z = u z (r ) , p = p (r, z) , (1.11)

u r = u r (r ) , u ϕ = u ϕ (r ) , u z = u z (r ) , p  = p  (r ) . (1.12)

Using these physical constraints (1.11) and (1.12), the elements of the viscous stress
tensor τ and the Reynolds stress tensor τ R A (1.3) can be given by

du
τr z = τzr = τzr (r ) = μ , and τrRz A = τzrR A (r ) =  (r ) = −ρu r u z , (1.13)
dr
8 1 Introduction to Classical Analytical Solutions for Wall-Bounded Turbulence

where μ is the dynamic viscosity of the fluid, and τrRz A (r ) =  (r ) represents the
principal (dominant) turbulent shear stress along the radial direction r . The physi-
cal assumptions—defined by Eqs. (1.12) and (1.13)—imply that the fully-developed
axisymmetric turbulent shear flow in straight circular pipes is considered as a steady-
state (stationary) turbulent flow, because the time coordinate t is excluded from
Eqs. (1.12) and (1.13), thus

∂u ∂u r ∂u ϕ ∂u z
= · er + · eϕ + · ez = 0. (1.14)
∂t ∂t ∂t ∂t
It means that the time derivative of the turbulent mean velocity field u (1.9) vanishes.
The investigated circular pipe is a horizontal pipe, therefore, the impact of the gravity
force g on the fluid flow can be neglected, thus

g = gr · er + gϕ · eϕ + gz · ez = 0. (1.15)

1.3.2 The Simplified Reynolds Momentum Equation

Using the physical constraints based on Eqs. (1.11)–(1.15) for fully-developed tur-
bulent shear flows in straight circular pipes, the scalar set of the general Reynolds-
averaged governing equations (1.1) and (1.2) can be formulated in the cylindrical
(curvilinear) coordinate system. Note that the cylindrical form of each individual term
in the Reynolds-averaged governing equations (1.1) and (1.2) has been derived in
Appendix A. The mass conservation (continuity) equation of incompressible turbu-
lent flows (1.1)—using all notations introduced in Eqs. (1.11)–(1.15)—can be written
as
1 ∂ (r u r ) 1 ∂u ϕ ∂u z (r )
+ + = 0, (1.16)
r ∂r
      r ∂ϕ ∂z
=0 =0

which means that the continuity equation (1.1) is satisfied automatically by apply-
ing all physical constraints defined by Eqs. (1.11)–(1.15). The Reynolds momentum
equation (1.2) of the velocity component u r can be given by
 
∂u r ∂u r u ϕ ∂u r u 2ϕ ∂u r ∂p
ρ + ρ ur + − + u z (r ) = ρgr − +
∂t ∂r r ∂ϕ r ∂z  ∂r
 
=0
   =0
=0
=0

 
1 ∂ ∂u r 1 ∂ 2 ur ∂ 2 ur ur 2 ∂u ϕ
+μ r + 2 + − 2 − 2
r ∂r ∂r r ∂ϕ 2 ∂z 2 r r ∂ϕ
  
=0
Another random document with
no related content on Scribd:
migrations; in most cases they are undertaken simply to gratify a
passing desire.
Quite otherwise is it with those mammals which, every year about
the same time, leave their habitat for some other region often far
distant, from which at a definite time they will return to their former
abode. These migrate; for they do not seize a chance opportunity,
but obey, consciously or unconsciously, a compelling necessity.
The fundamental cause of all true migration among mammals, is
some very distinct and decided seasonal change. In countries of
everlasting spring true migrations do not take place, for want is never
imminent. Summer must contrast with winter, whether the latter bring
frost and snow, or heat and drought; scarcity must alternate with
superfluity before the sluggish mammal makes up its mind to
migrate.
To a slight extent migrations take place among all mountain animals.
The chamois, the steinbock, the Alpine hare, the marmot all migrate
when the snow begins to melt, or a little later; they clamber over
hillsides and glaciers to the heights above, where the pasturage,
now laid bare, promises rich and abundant nourishment, and they
return to the lower slopes of the mountain before winter sets in. The
bear, by nature omnivorous, by habit a thief, undertakes a similar
migration at the same season, and completes it before winter sets in,
—at least so it is in the mountains of Siberia; the various wild cats
and dogs which live among the mountains do the same. Such
changes of residence occur also on the mountains of southern
countries, even of those lying within the tropics. In India and Africa
certain species of monkey ascend and descend the mountains at
regular intervals; elephants seek the high grounds on the approach
of summer, the low grounds in winter; on the Andes in South
America the guanacos flee before the snow into the valleys, and
before the summer-heat to the shoulders of the mountains. All these
migrations are confined by the mountains within comparatively
narrow limits. They only involve a change of altitude of from three to
nine thousand feet, or a journey which may be accomplished in a
few hours, or, at most, in a few days. They have, however, the
regularity characteristic of true migrations, especially in the precise
periodicity of their occurrence, and not less in the constant choice of
the same routes.
Highlands and plain, sea and air, offer a much wider field than the
mountains, and therefore the migrations of the animals inhabiting, or
temporarily traversing these can be more easily observed, and they
are more appropriately termed migratory animals than the dwellers
among the mountains. In the tundras of Russia and Siberia, the
reindeer, which, in Scandinavia, never leaves the mountains,
migrates to a great distance every autumn and returns the following
spring to his former summer haunts.[57] About the same time it
leaves Greenland, and, crossing the sea on a bridge of ice, reaches
the continent of America, where it spends the whole winter, only
returning to the hills of its native peninsula the following April. In both
cases, dread of the approaching winter does not seem to be the sole
cause of migration; there is at the same time a further incentive
supplied by a plague much feared in the far north. For the short
summer on these expanses calls to life an insect-world poor in
species, but endlessly rich in individuals, particularly an
indescribable number of mosquitoes and bot-flies, which make life a
burden to the reindeer, as well as to man. To escape these the
reindeer forsakes the marshy tundra, over which dense clouds of
mosquitoes hover during summer, and hies to where the scourge is
less severely felt—to the Alpine heights, which, in the summer
season afford their most fragrant pasturage. From inherited habit,
the reindeer migrate not only at the same time, but along the same
paths, thus forming tracks which may be distinctly traced, traversing
the tundra for many miles, and crossing streams and rivers at
definite places. At the beginning of the journey, the cows with their
calves arrange themselves in herds of from ten to a hundred, and
precede the young stags and hinds, which are followed again by the
old stags. One troop follows directly behind another, and the
observer can count thousands as they pass. All hurry incessantly on,
turning aside neither for the mountains nor the broad streams which
cross their path, and resting only when they have reached their
winter-quarters. Packs of wolves, bears, and gluttons follow close on
their heels and often pursue them no small part of the way. In spring,
on the return journey, the animals keep to the same order, but the
herds are much smaller, and they travel in a much more leisurely
fashion, and keep less strictly to the paths by which they went.
Journeys still longer than those of the reindeer are taken by the
American bison, the “buffalo” of the prairies.[58] What distance
individual animals travel cannot be stated with certainty, but herds in
course of migrating have been met from Canada to Mexico, from the
Missouri to the Rocky Mountains, and it may be assumed that a
single herd traverses a considerable part of the country lying
between these limits. The bisons have been seen in summer
scattered over the boundless prairie, and in winter in the same
places, but assembled in many thousands; their migrations have
been observed, for they have been followed for hundreds of miles
along the tracks—the so-called “buffalo-paths”, trodden out straight
across plains and over mountains. We learn from eye-witnesses that
a stream a mile wide is to them no barrier, scarcely even a
hindrance, for they throw themselves into it like an irresistible
avalanche, so that the water is covered with the dark, moving throng;
that the animals associate and separate again, the herds increase
and diminish; that old, surly, tyrannous, malevolent bulls avoid the
other bisons, having perhaps been expelled from the herd, and
compelled, probably only after protracted struggles, to live in hermit-
fashion until the following summer; and that, during heavy snowfalls,
the herds take shelter in the forests or on the slopes of the
mountains. From July onwards they begin their migrations from the
north towards the south. Small companies, which, till then, have
been leading a comfortable summer life, combine with others and set
out on the journey with them; other troops join the band, which grows
as it presses on, until there is, at length, formed one of those
extraordinary herds which, united till the next spring, moves and acts
as if animated by one soul. When the winter is safely past, the army
gradually breaks up, probably in exactly reverse order, into herds,
and these divide more and more until at length only small companies
are left. This breaking-up takes place during the course of the return
journey. Both in going and returning, one herd follows another at
some distance, but more or less along the same paths. Specially
favourable places, such as low grounds covered with rich grass,
cause a temporary damming up of the living stream. In such places
incalculable herds assemble together, spend days in the same spot,
and break up only again when all the grass has been eaten, and
hunger urges them to continue their journey. As they march the
wolves and bears follow their track, while eagles and vultures, birds
of ill omen, circle over their heads.
 Fig. 36.—A Herd of American Bison or Buffalo.

Scarcity of water, as well as of food, is often a cause of regular

migrations. When winter approaches in the south-east of Siberia,
more particularly in the high Gobi steppe, all the non-hibernating
mammals are compelled, by the peculiar circumstances of these
highlands, to seek refuge in lower-lying regions. The winter in these
high grounds of Central Asia is not more severe than in districts lying
further to the north or north-east, but it is usually almost snowless,
and such pools of water as have been formed by the extremely slight
fall of rain or snow, are covered with a thick sheet of ice. As soon as
this sheet becomes so strong that the animals inhabiting the Gobi
are unable to break it, they are obliged to change their quarters, and
they travel not only to southern but to northern lands, whose only
advantage is that they are covered with snow, for this affords ready
refreshment to the parched tongues of the wanderers, and offers
less resistance to their weak feet than the hard, unbreakable, and
less easily melted ice. This is the explanation of the fact that the
antelope, of which great numbers are found in the Gobi, forsakes a
land which, save for the lack of snow and therefore of available
water, is exactly the same as that which it chooses for its winter
quarters. Not hunger, but thirst, drives it from its home. At the
beginning of winter, the antelopes, at all times gregarious, assemble
in herds of many thousands, spreading over all the low grounds
around their native plateau; they often travel at the rate of fifty or
sixty miles in a single night, and extend their wanderings many
hundreds of miles beyond the boundaries of their proper habitat. The
observer who follows them can detect their tracks everywhere, and
in such numbers that it seems as though vast herds of sheep, far
exceeding in number any ordinary flock, had just passed by.
 Fig. 37.—Wild Horses crossing a River during a Storm.

Before the Chinese antelope begins its migration, restlessness

seizes the kulan or dziggetai, probably the ancestor of our horse,
and certainly the most beautiful and the proudest of all wild horses.
The foals of the summer are by autumn strong enough to be able to
endure a long journey with quick marches, and to bid defiance to all
the accidents and dangers of a wandering life. The young stallions
attain their full strength at the end of their fourth year, and towards
the end of September they leave the parent-herd and press forward.
Finally, the impulse to mate begins to animate the older stallions and
mares, and with it comes unrest and the desire to wander. Thus the
fleet, enterprising animals begin their annual migration long before
winter has set in, before even its approach has become at all
apparent; and on this account their migrations at first lack steadiness
and regularity, and have something of the character of journeys in
search of adventure. With the intention of shaking off the
burdensome yoke of the leader and absolute lord of the herd, and of
becoming independent and in their turn equally despotic, the young
stallions forsake the herd, and thenceforward traverse the sandy
steppes singly. All the younger mares who are mature, and even
many of the older ones, seem to be animated by the same feeling as
the young stallions, and they attempt to escape from the rule of their
tyrant and join his young rival, to fall immediately under his dominion.
But not without a struggle does the new candidate for leadership
gain his troop of mares; the old leader does not readily relinquish his
rights. For hours together the stallion stands on the top of a hill or on
the shoulder of a ridge, keenly scanning the country around. His eye
wanders over the desert, his dilated nostrils are turned towards the
wind, his ears are directed forwards on the alert. Eager for battle, he
rushes at full gallop towards every herd which approaches, every
adversary who shows himself; and a furious struggle takes place for
the possession of the mares, who always attach themselves to the
victor. Such combats and strife set the herd in motion, detach it from
the place where the summer has been spent, and lead on to
migrations which become gradually regular, prolonged, persistent,
and almost uninterrupted. In the course of these, if not before the
end of the combats just described, the kulan troops assemble in
ever-growing numbers, until at length herds of more than a thousand
head set out together for fields which give promise of pasturage.
They do not break up while in their winter quarters, and they are thus
compelled to be continually on the move in order to find sufficient
nourishment. The combined tread of the army, as they gallop on in
their usual furious fashion, rings dully out, and more than once, in
Russia, the sound has called the Cossacks of the military cordon to
arms. No wolf ventures to attack such a herd, for the courageous
wild horses know so well how to use their hoofs against him that he
soon gives up any attempt; it is only the sick and exhausted horses
which become his prey, as he follows the wandering herd. Even man
can do them no great damage, for their caution and shyness render
them difficult of approach. But winter, especially if much snow falls,
brings them much suffering. The pasture, at all times scanty, is
exhausted the more quickly the more numerous the herd which
feeds on it. Then the animals devour indiscriminately all the
vegetable substances they can find. For months together they have
to maintain life on leafless shoots. Their bodies cease to be fat and
plump, till at length they are like wandering skeletons. The mother,
herself starving, is no longer able to nourish her foal, for the milk-
yielding udder dries up in times of such need. Many a one whose
tender youth is unable to endure the hard fare dies of starvation.
Even the old horses suffer from the poverty and treachery of the
winter. Snow-storms blowing over their feeding-ground for days at a
time depress their usually cheerful courage, and increase the
boldness of the wolves, which, even if they do not fall upon the
already exhausted horses, persecute and annoy to the utmost those
who are not yet worn out. But as soon as circumstances begin to
improve, the wiry, weather-hardened, enduring creatures recover
their high spirits, and, when the snow begins to melt, they set out on
the return journey, reaching their summer home in about a month’s
time. There they break up into single herds, recuperate among the
luxuriantly sprouting, fragrant pasture, and, in a surprisingly short
time, become fat and plump again. Soon the want and misery of the
winter are forgotten.
Great as are the distances often traversed by all the mammals
already mentioned, they can scarcely be compared with those
covered by seals and whales. The water favours all the movements
of animals adapted to aquatic life, and offers everywhere the same
general conditions of life and the same amenities. Thus it renders
the migrations of its inhabitants easier, less toilsome and hazardous
than those of any other wanderers. Nevertheless it is somewhat
surprising to learn that many sea-mammals, and particularly the
whales, are among the most nomadic of all creatures; in fact that
many, if not most of them, pass their whole life in travelling. Strictly
speaking, no whale has a permanent place of sojourn for the whole
year, but passes singly, in pairs, with its young, or in more or less
numerous companies—the so-called schools—from one part of the
ocean to another, visiting certain favourite haunts in regular order,
and choosing different haunts in summer and in winter. The seas
inhabited by the same species of whale in winter and in summer
often lie farther apart than people seem to suppose, for some whales
travel, twice a year, more than a quarter of the earth’s circumference;
they are to be found in summer among the ice-floes of the Arctic
Ocean, and in winter on the other side of the Equator. The female
whales, who are in the highest degree sociable, and attached to their
young with the tenderest, most devoted love, assemble together in
surprising numbers, and under the guidance of a few males, traverse
the ocean by definite routes and at definite times, some keeping to
the open sea, others making their way along the coasts. Storms may
force them to change their route, or delay in the appearance of the
animals on which they feed, whose occurrence and disappearance is
obviously the chief cause of their migration, may to some extent
influence their course and the time of their visiting certain spots; but,
as a general rule, their migrations are so systematic that on northern
and southern coasts people look for the arrival of the whale on a
particular day, and place watches so that they may be able to begin
the long-desired chase without loss of time. Whales, which are
recognized by the dwellers on the coast by some mark, such as
mutilated fins, and which have been several times pursued in vain,
have been known to appear several years in succession at the same
time and at the same place; and the chase after these most valuable
and therefore increasingly persecuted animals takes place with the
same regularity as do hare-hunts on land, though at any other time
of the year it would be vain to look for them. “After Twelfth Day,” says
old Pontoppidan, “the Norwegians watch from all the hills for the
whale, whose arrival is announced by the herring.” First appears the
killer, then, three or four days, or at the most a fortnight later, the
rorqual, though, apparently, one comes from Davis Straits and the
other from Greenland. On the south coasts of the Faroë Islands, and
especially in the Qualbenfjord, from three to six bottle-nose whales
still appear every year about Michaelmas, as they did a hundred and
ninety years ago. In a Scottish bay there appeared twenty years in
succession a rorqual, which was generally known by the name of
“Hollie Pyke”, and was pursued every year and finally captured. On
the coast of Iceland single whales choose the same bays for a
temporary sojourn every year in the same months, and even weeks,
so that the inhabitants have got to know them individually, and have
given them special names. Certain well-known mother-whales visit
the same bays every year to bring forth their young, and they
themselves are spared, but they have to purchase their own lives,
dearly enough, at the cost of that of their young ones, which are
regularly taken captive. It is very unusual for the migrating whales to
keep neither to time nor to route; in general, their journeys are as
regular as if they were arranged according to the position of the
stars, and as if they took place along laid-out paths bounded on both
sides. No other mammal migrates more regularly; indeed, their
wanderings may be compared with the migrations of birds.
The seals, like the whales, migrate every year, on the whole with
great regularity, though not to such a distance. Those species which
inhabit inland seas cannot, of course, leave these, but they traverse
them every year in regular order, or at certain times ascend the
rivers flowing into them; all the ocean species, on the other hand, set
out every autumn and spring, by definite routes, to certain regions or
localities. All the seals in the far north, as well as those in the seas
about the South Pole, are forced to migrate by the extension of the
ice in winter, and may travel with it towards temperate zones,
returning towards the poles again as the ice melts. But they, like all
other members of their order, are impelled to travel for another not
less weighty reason; they require the mainland, or at least large,
spreading, fixed masses of ice, on which to bring forth and nurture
their young, until these are able to follow them into the water, there
to shift for themselves. Thus every year thousands and hundreds of
thousands of seals appear on certain islands and ice-banks,
covering some of these birth-places of their race in such crowds that
every available spot must be utilized in order to secure space for all
to bring forth their young. They pass weeks, even months, on land or
on the ice without hunting, descending into the sea, or taking food;
they suckle their young, then mate, and by degrees break up their
great assemblage, distributing themselves over the wide ocean to
resume their former manner of life, or setting out with their young,
who still require training, on more or less extensive foraging
expeditions.[59]

Fig. 38.—Flying Foxes.

As is well known, there are many mammals which have the habit of
hibernating, which pass the severe part of the year well protected in
deep and carefully closed burrows, and are thus spared the
necessity of leaving their haunts. Even among these, however, at
least among those living in the temperate zones, there are some
which migrate during their waking time, namely, the bats. Defective
as the wing of a bat must appear when compared with that of a bird,
it is nevertheless of such assistance in flight, that it makes journeys
possible which seem out of all proportion to the size of the animal.
Another fact makes travelling easier to the restless bat; it is not tied
down by its offspring to any particular spot, for the young one
attaches itself directly after birth to the breast of the mother, and is
borne by her through the air till it is capable of independent life. The
bat is thus one of the best-adapted of migratory mammals, and,
under some circumstances, it makes full use of its advantages. As a
general rule, the wanderings of the different species of bat are to be
regarded simply as excursions made with a view to taking advantage
of any district which is, for the time being, particularly rich in food; but
they do sometimes become really long journeys, which lead some
species to far distant lands, and they are then not without the
regularity characteristic of all true migrations. The largest bats, the
flying foxes, fly long distances every evening in search of the fruits
on which they chiefly subsist; they do not hesitate to cross an arm of
the sea fifty or sixty miles in breadth, and they must even have
traversed the distance between Southern Asia and the East of
Africa, as certain species occur in both these regions. The bats
proper accomplish at least as much. Following the reappearance of
the insects, which occurs at different times in regions of different
altitude, they ascend from the plains to the mountain heights, and
descend in autumn to the low grounds again; they pursue the
numerous flies which congregate about the wandering cattle-herds
of Central Africa, and they migrate also from the south towards the
north and return southwards again, or in reverse order. The boreal
bat appears at the beginning of the bright nights in the north of
Scandinavia and Russia, and leaves these districts, which may be
considered its head-quarters, towards the end of summer, to spend
the winter among the mountains of Central Germany and the Alps.
The pond-bat is regularly seen on the plains of North Germany
during summer, but only exceptionally at that period among the
mountains of Central Germany, in whose caverns it spends the
winter. That other species of bat occurring in Germany change their
place of abode in a similar manner can scarcely be doubted.
In the cases cited, which have been selected from a mass of
available material, I have given examples of those migrations of
mammals which we may call voluntary, because of their regularity;
but in so doing I have by no means completed my task. Hunger and
thirst, the poverty and temporary inhospitableness of a particular
region, sometimes press so severely on certain mammals that they
endeavour, as if despairing, to save themselves by flight. Abundant
nourishment and good weather favour the increase of all animals,
and affect that of a few plant-eating mammals to such an
extraordinary degree, that, even under propitious conditions, their
habitat must be extended. But if one or more rich years—in some
cases a few favourable months—be followed by a sudden reverse,
the famine soon passes all bounds, and robs the creatures not only
of the possibility of subsistence, but also of all hope, or at least of all
presence of mind.
It is under circumstances such as these that the field-voles of our
own country, and the Siberian voles, assemble in enormous
multitudes, leave their native haunts and migrate to other districts,
turning back for no obstacles, avoiding the water as little as the
forbidding mountains or the gloomy forest, fighting to the last against
hunger and misery, but perishing hopelessly from diseases and
epidemics which rage among them like plagues, reducing armies of
millions to a few hundreds. Thus, too, the squirrels of Siberia, which,
in ordinary years, undertake, at the most, only short excursions,
assemble in vast armies, hurry in troops or companies from tree to
tree, in compact masses from forest to forest, swim across rivers and
streams, throng into towns and villages, lose their lives by
thousands; but suffer no obstacle or hindrance—not even the most
obvious dangers—to delay them or divert them from their path. The
soles of their feet become worn and cracked, their nails ground
down, the hairs of their usually smooth fur rough and matted.
Through the forest lynxes and sables, in the open fields gluttons,
foxes and wolves, eagles, falcons, owls and ravens follow them
closely; pestilence claims more victims from their ranks than the
teeth and claws of beast of prey or the guns and cudgels of men, yet
they press on and on, apparently without hope of return. A Siberian
sportsman of my acquaintance gave me a verbal account of the
appearance of such an army of squirrels, in August 1869, in the town
of Tapilsk, among the Ural Mountains. It was only one wing of a
migrating army, of which the main body travelled through the forest
about five miles farther north. Sometimes in single file, sometimes in
companies of varying strength, but in unbroken succession, the
animals pressed on, crowding as densely through the town as
through the neighbouring forest; used the streets, as well as the
hedges, and the roofs of buildings as paths; filled every court-yard,
thronged through windows and doors into the houses, and created
quite an uproar among the inhabitants—much more among the
dogs, which killed thousands of them, evincing an unbridled
bloodthirstiness till then unsuspected. The squirrels, however, did not
seem to concern themselves in the least about the innumerable
victims falling in their midst; in fact, they took no notice of anything,
and allowed nothing to divert them from their route. The procession
lasted for three whole days, from early morning till late in the
evening, and only after nightfall each day was there a break in the
continuity of the stream. All travelled in exactly the same direction,
from south to north, and those that came last took the same paths as
their predecessors. The rushing Tchussoveia proved no obstacle, for
all that reached the bank of that rapid mountain-river plunged without
hesitation into its whirling and seething waters, and swam, deeply
sunk and with their tails laid across their backs, to the opposite bank.
My informant, who had been watching the procession with growing
attention and sympathy, rowed out into the midst of the throng. The
tired swimmers, to whom he stretched out an oar, climbed up by it
into the boat, where, apparently exhausted, they sat quietly and
confidingly, until it came alongside a larger vessel, when they
climbed into that, and remained on it for some time as indifferent as
before. As soon as the boat touched the bank they sprang ashore,
and proceeded on their journey as unconcernedly as if it had
suffered no interruption.
It must be similar circumstances which compel the lemmings to the
migrations which have been known for centuries. For many
successive years the heights in the tundras of Scandinavia, North
Russia, and the North of Siberia afford them comfortable quarters
and abundant nourishment; for the broad ridges of the fjelds and the
extensive plains between them, the highlands and the low grounds,
offer room and maintenance for millions of them. But not every year
do they enjoy the accustomed abundance for the whole summer. If a
winter in which much snow falls, and which is therefore favourable to
them, as they live safely below the snow, be followed by an early,
warm, and agreeable spring, their extraordinary fertility and power of
increase seem to have almost no bounds, and the tundra literally
teems with lemmings. A fine warm summer increases their numbers
past computing, but it also accelerates the life-course of all the
plants on which they feed, and before it is over these are partly
withered, partly devoured by the greedy teeth of the insatiable
rodents. Scarcity of food begins to be felt, and their comfortable life
comes to an end in panic. Their fearless, bold demeanour gives
place to a general uneasiness, and soon a mad anxiety for the future
takes possession of them. Then they assemble together and begin
to migrate. The same impulse animates many simultaneously, and
from them it spreads to others; the swarms become armies; they
arrange themselves in ranks, and a living stream flows like running
water from the heights to the low grounds. All hurry onwards in a
definite direction, but this often changes according to locality and
circumstances. Gradually long trains are formed in which lemming
follows lemming so closely that the head of one seems to rest on the
back of the one in front of it; and the continuous tread of the light,
little creatures hollows out paths deep enough to be visible from a
long distance in the mossy carpet of the tundra. The longer the
march lasts, the greater becomes the haste of the wandering
lemmings. Eagerly they fall upon the plants on and about their path
and devour whatever is edible; but their numbers impoverish even a
fresh district within a few hours, and though a few in front may pick
up a little food, nothing is left for those behind; the hunger increases
every minute, and the speed of the march quickens in proportion;
every obstacle seems surmountable, every danger trifling, and
thousands rush on to death. If men come in their way they run
between their legs; they face ravens and other powerful birds of prey
defiantly; they gnaw through hay-stacks, climb over mountains and
rocks, swim across rivers, and even across broad lakes, arms of the
sea, and fjords. A hostile company, like that behind the migrating
squirrels, follows in their wake: wolves and foxes, gluttons, martens
and weasels, the ravenous dogs of the Lapps and Samoyedes,
eagles, buzzards, and snowy owls, ravens and hooded crows fatten
on the innumerable victims which they seize without trouble from the
moving army; gulls and fishes feast on those which cross the water.
Diseases and epidemics, too, are not awanting, and probably
destroy more than all their enemies together. Thousands of
carcasses lie rotting on the wayside, thousands are carried away by
the waves; whether indeed any are left, and whether these return
later to their native Alpine heights, or whether all, without exception,
perish in the course of their journey, no one can say with certainty;
but so much I know, that I have traversed great tracts of the tundra of
Lapland where the paths and other traces of a great migrating army
were to be seen almost everywhere, while not a single lemming
could be discovered. Such tracts, I have been told, remain thus for
several successive years, and only after long periods become
gradually repeopled with the busy little rodents.[60]
What hunger causes in the North is brought about by the tortures of
thirst in the richer South. As the brackish pools which have afforded
water to the zebras, quaggas, antelopes, buffaloes, ostriches, and
other animals of the steppes, dry up more and more under the
burning heat of a South African winter, all the animals whose
necessities have hitherto been supplied by the steppes assemble
about the pools which still contain a little water, and these become
scenes of stirring, active life. But when these, too, evaporate, the
animals which have congregated around them are compelled to
migrate, and it may happen that despair takes possession of them,
as it did of the little rodents already described, and that, collecting in
herds like the wild horses and Chinese antelopes (dzieren) of the
steppes of Central Asia, or the bisons of the North American prairies,
they rush straight on for hundreds of miles, to escape the hardships
of winter.
In the South, too, the wild horses are the first to turn their backs on
the inhospitable country. Till the drought sets in, these beautifully-
marked, strong, swift, self-confident children of the Karroo, the
zebra, quagga, and dauw, wander careless and free through their
vast domain, each herd going its own way under the guidance of an
old, experienced, and battle-tried stallion. Then the cares of the
winter season begin to make themselves felt. One water-pool after
another disappears, and the herds which gather about those which
remain become more and more numerous. The general distress
makes even the combative stallions forget to quarrel and fight.
Instead of small companies, herds of more than a hundred head are
formed, and these move and act collectively, and finally forsake the
wintry region altogether before want has enfeebled their powers or
broken their stubborn wills. Travellers describe with enthusiasm the
spectacle presented by such a herd of wild horses on the march. Far
into the distance stretches the sandy plain, its shimmering red
ground-colour interrupted here and there by patches of sunburnt
grass, its scanty shade supplied by a few feathery-leaved mimosas,
and, as far away as the eye can reach, the horizon is bounded by
the sharp lines of mountains quivering in a bluish haze. In the midst
of this landscape appears a cloud of dust which, disturbed by no
breath of air, ascends to the blue heavens like a pillar of smoke.
Nearer and nearer the cloud approaches, until at length the eye can
distinguish living creatures moving within it. Soon the brightly-
coloured and strangely-marked animals present themselves clearly
to the spectator’s gaze; in densely thronged ranks, with heads and
tails raised, neck and neck with the quaintly-shaped gnus and
ostriches which have joined their company, they rush by on their way
to a new, and possibly far-distant feeding-ground, and ere the
onlooker has recovered himself, the wild army has passed by and is
lost from view in the immeasurable steppe.
 Fig. 39.—Springbok Antelopes.

The antelopes, which are also driven out by winter, do not always
follow the same paths, but usually travel in the same direction. None
is more numerous or more frequently seen than the springbok, one
of the most graceful and beautiful gazelles with which we are
acquainted. Its unusual beauty and agility strike everyone who sees
it in its wild state, now walking with elastic step, now standing still to
feed, now springing about in playful leaps, and thus disclosing its
greatest ornament, a mane-like snow-white tuft of hair, which at a
quieter pace is hidden in a longitudinal groove of the back. None of
the other antelopes, when forced to migrate, assemble in such
numerous herds as this one. Even the most vivid description cannot
convey to one who has not seen a herd of springboks on their
journey any adequate idea of the wonderful spectacle. After having
congregated for weeks, perhaps waiting for the first shower of rain,
the springboks at last resolve to migrate. Hundreds of the species
join other hundreds, thousands other thousands, and the more
threatening the scarcity, the more torturing the thirst, the longer the
distances which they cover; the flocks become herds, the herds
armies, and these resemble the swarms of locusts which darken the
sun. In the plains they cover square miles; in the passes between
the mountains they throng together in a compact mass which no
other creature can resist; over the low grounds they pour, like a
stream which has overflowed its banks and carries all before it.
Bewildering, intoxicating, and stupefying even the calmest of men,
the throng surges past for hours, perhaps days together.[61] Like the
greedy locusts, the famishing animals fall upon grass and leaves,
grain, and other fruits of the field; where they have passed, not a
blade is left. The man who comes in contact with them is at once
thrown to the ground, and so sorely wounded by the tread of their
hoofs, light indeed, but a thousand times repeated, that he may be
glad if he escapes with his life; a herd of sheep feeding in the way is
surrounded and carried off, never to be seen again; a lion, who
thought to gain an easy prey, finds himself forced to relinquish his
victim, and to travel with the stream. Unceasingly those behind press
forward, and those in front yield slowly to the pressure; those cooped
up in the middle strive continually to reach the wings, and their
efforts are strenuously resisted. Above the clouds of dust raised by
the rushing army the vultures circle; flanks and rear are attended by
a funeral procession of various beasts of prey; in the passes lurk
sportsmen, who send shot after shot into the throng. So the tortured
animals travel for many miles, till at length spring sets in and their
armies are broken up.
Shall I go on to consider other compulsory migrations, such as those
of the arctic foxes and polar bears when an ice-floe on which they
were hunting is loosened and floated off by the waves till, under
favourable circumstances, it touches some island? I think not, for
journeys such as these are not migrations, they are simply passive
driftings.