Fluid Dynamics for Engineers

Olivier Cleynen
A one-semester course for students in
the Chemical and Energy Engineering program
at the University Otto von Guericke of Magdeburg
2015-2020

last edited October 27, 2022

https://fluidmech.ninja/
 Contents
Contents 2
About this course (syllabus) 6
1 Basic flow quantities 9
1.1 Concept of a fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2 Fluid dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3 Important concepts in mechanics . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4 Properties of fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.5 Forces on fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.6 Basic flow quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.7 Four balance equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.8 Classification of fluid flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.9 Limits of fluid dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.10 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.11 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2 Analysis of existing flows with one dimension 33
2.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2 One-dimensional flow problems . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.3 Balance of mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.4 Balance of momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.5 Balance of energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6 The Bernoulli equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.7 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.8 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3 Analysis of existing flows with three dimensions 51
3.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2 The Reynolds transport theorem . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.3 Balance of mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.4 Balance of momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.5 Balance of angular momentum . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.6 Balance of energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.7 Limits of integral analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.8 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.9 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4 Effects of pressure 73
4.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.2 Pressure forces on walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.3 Pressure fields in fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.4 Special case: pressure in static fluids . . . . . . . . . . . . . . . . . . . . . . . 79
4.5 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.6 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5 Effects of shear 91
5.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

3
5.2 Shear forces on walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.3 Shear fields in fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.4 Resistance to shear: viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.5 Special case: shear in simple laminar flows . . . . . . . . . . . . . . . . . . . 101
5.6 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6 Prediction of fluid flows 111
6.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.2 Organizing calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.3 Equations for all flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
6.4 Equations for incompressible flow . . . . . . . . . . . . . . . . . . . . . . . . 123
6.5 CFD: the Navier-Stokes equations in practice . . . . . . . . . . . . . . . . . . 127
6.6 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
7 Pipe flows 135
7.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
7.2 Frictionless flow in pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
7.3 Parameters to quantify losses in pipes . . . . . . . . . . . . . . . . . . . . . . 137
7.4 Laminar flow in pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
7.5 Turbulent flow in pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
7.6 Engineer’s guide to pipe flows . . . . . . . . . . . . . . . . . . . . . . . . . . 146
7.7 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
7.8 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
8 Engineering models 161
8.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
8.2 Comparing influences: the weighted momentum balance . . . . . . . . . . . 161
8.3 Making models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
8.4 Comparing results: coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . 169
8.5 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
8.6 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
9 Dealing with turbulence 179
9.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
9.2 Recognizing turbulence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
9.3 The effects of turbulence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
9.4 Quantifying turbulence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
9.5 Computing turbulent flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
9.6 Commented bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
9.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
10 Flow near walls 199
10.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
10.2 The concept of boundary layer . . . . . . . . . . . . . . . . . . . . . . . . . . 199
10.3 Laminar boundary layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
10.4 Boundary layer transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
10.5 Turbulent boundary layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
10.6 Flow separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
10.7 Solved problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
10.8 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
11 Large- and small-scale flows 221
11.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
11.2 Flow at large scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
11.3 Plotting velocity with functions . . . . . . . . . . . . . . . . . . . . . . . . . 223
11.4 Flow at very small scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
11.5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Appendix 245
A1 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
A2 Vector operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
A3 Field operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
A4 Derivations of the Bernoulli equation . . . . . . . . . . . . . . . . . . . . . . 253
A5 Flow parameters as force ratios . . . . . . . . . . . . . . . . . . . . . . . . . . 256
A6 Details of the winter 2020-2021 final examination (updated February 2021) . 259
A7 Example of previous examinations . . . . . . . . . . . . . . . . . . . . . . . . 261
A8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
 About this course (syllabus)
Fluid dynamics for engineers by Olivier Cleynen, PhD
University of Magdeburg, 2015-2020
https://fluidmech.ninja/

Welcome! From 2015 to 2020, these notes were the core of the Fluid Dynamics course of
the Chemical and Energy Engineering program at the University of Magdeburg, Germany.

Objectives
Starting with little or no experience with fluid mechanics, after taking this course:

• you should have a good understanding of what can, and cannot, be calculated with
fluid mechanics in engineering: how we approach problems depending on how
much information is available.

• you should be able to solve several real-world engineering fluid mechanics prob-
lems with confidence: calculating forces within fluids and on objects, predicting
flow in pipes, near walls, at small and large scales.

My objective is to enable you to get there with the minimum amount of your time and
energy (but not minimum power!).
If all goes well, at the end of the semester, you should be well-prepared to begin a course
in Computational Fluid Dynamics, where the knowledge and skills you acquire here can
be used to solve applied problems in great detail.

About these notes and the author

My name is Olivier Cleynen. I had the pleasure and
priviledge of directing this course from 2015 to 2020
while I was a PhD student at the University Otto
von Guericke of Magdeburg. In 2021, my colleagues
at the fluid dynamics laboratory took over. I am
no longer involved in teaching the course, but am
delighted that these course materials live on.
If you are a student enrolled in this course, you should definitely stop reading now and
check with your course instructor so you understand how the course will go (where to
get help, what chapters are examinable, and so on). I am not at all up-to-date with that
information, and unfortunately cannot help at all!
I obtained my Master’s in 2006, then went on to found and work for a non-profit
organization, and then became a university teacher in France. I arrived in Magdeburg in
2015, and currently live here with my partner and her twelve year-old child. I completed
my PhD in 2022, and I now am a math and physics teacher in a local high school. I am
delighted that the present course notes at https://fluidmech.ninja/ keep being useful to
students around the world.

Contact
My email is olivier cleynen.fr. I am available to answer queries about re-using the
course materials. Unfortunately, I cannot provide assistance with the content itself (e.g.
how to solve the end-of-chapter problems) or answer any questions about the course.
If you would like such assistance, please contact my colleagues at the fluid dynamics
laboratory of the University of Magdeburg.

Copyright, remixing, and authors

This document is mainly authored by Olivier Cleynen. Substantial contributions have
been made by colleagues Germán Santa-Maria [35], Jochen König, and Arjun Neyyathala.
Numerous improvements have been contributed by students over the years.
Many figures from authors not asso ciated with this course are included; the author,
license, and a link to the source are indicated every time. A few figures still remain which
are extracted from cited, fully-copyrighted works, as indicated.
Most portrait illustrations are authored by Oksana Grivina; they are fully-copyrighted
and used under a commercial license in this project. Other portrait illustrations are
authored by Agustin Dede Prasetyo and Olivier Cleynen under a CC-by license.
The text of this document is licensed under a Creative Commons CC-by-nc (attribution,
non-commercial) license. This means you are free to make copies of it, but not for com-
mercial purposes. This license also does not permit uploading this document to content
platforms such as Facebook, Academia.edu or Scribd. The Latex sources of this document
can be accessed from the git repository accessed from the course homepage. [34] A
number of YouTube videos, shot during the 2020 German lockdown, are linked along the
text.
If you use this document in other works, please cite it as “Olivier Cleynen. Fluid dynamics
for engineers. Under CC-by-nc license. 2020. url: https://fluidmech.ninja”.
 Conclusion

Welcome to this course!

(I recorded this YouTube video in April 2020)

I hope you have a great semester! Fluid mechanics is one of the most exciting disciplines
out there. Now, let us begin!

Olivier Cleynen
April 2020 (updated October 2022)

8
 Fluid Dynamics
Chapter 1 – Basic flow quantities
last edited April 18, 2021
by Olivier Cleynen — https://fluidmech.ninja/

1.1 Concept of a fluid 9

1.2 Fluid dynamics 9
1.2.1 Solution of a flow 9
1.2.2 Modeling of fluids 10
1.2.3 Theory, numerics, and experiment 11
1.3 Important concepts in mechanics 12
1.3.1 Position, velocity, acceleration 12
1.3.2 Forces and moments 12
1.3.3 Energy 13
1.4 Properties of fluids 14
1.4.1 Density 14
1.4.2 Phase 14
1.4.3 Temperature 15
1.4.4 Perfect gas model 15
1.4.5 Speed of sound 16
1.4.6 Viscosity 17
1.5 Forces on fluids 18
1.5.1 Gravity 18
1.5.2 Pressure 18
1.5.3 Shear 18
1.6 Basic flow quantities 19
1.7 Four balance equations 20
1.8 Classification of fluid flows 21
1.9 Limits of fluid dynamics 23
1.10 Solved problems 24
1.11 Problems 27

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

1.1 Concept of a fluid

We call fluidw a type of matter which is continuously deformable, and which
spontaneously tends to adapt its shape to its container by occupying all of
the space made available to it.

1.2 Fluid dynamics

1.2.1 Solution of a flow

Fluid dynamics (or generally, fluid mechanics) is the study of the movement
of fluids. The most common type of problem in this discipline is the search
for a complete description of the fluid flow around or through a solid object.
This problem is solved when the entire set of velocities of fluid particles
has been described. This set of velocities, which is often a function of time,
9
 can be described either as a set of discrete values (“pixelized” data) or as a
mathematical function; it is called the flow solution.
If the solution is known, the shear and pressure efforts generated on the
surface of the object can be calculated. Other quantities, such as the force
and moments applying on the object, or the fluid’s energy gains or losses,
can also be calculated.

1.2.2 Modeling of fluids

Like all matter, fluids are made of discrete, solid molecules. However, in fluid
mechanics, we work at the macroscopic scale: at that scale, matter can be
treated like a continuum, in which all physical properties of interest can be

There are about 2 ⋅ 1022 molecules in the air within an “empty” 1-liter bottle
continuously differentiated.

at ambient temperature and pressure. Even when the air within the bottle is
completely still, these molecules are constantly colliding with each other and

approximately 1 000 km/h.

with the bottle walls; on average, their speed is equal to the speed of sound:
Video: why not to calculate the
movement of molecules
by Olivier Cleynen (CC-by) Despite the complexity of individual molecule movements, even the most
https://youtu.be/37K7GxYnebk
turbulent flows can be accurately described and solved by considering the
velocities of groups of several millions of molecules collectively, which we
name fluid particles.w By doing so, we never find out the velocity of individual
molecules: instead, those are averaged in space and time and result in much
simpler and smoother trajectories, which are those we can observe with
macroscopic instruments such as video cameras and pressure probes.
Our selection of an appropriate fluid particle size (in effect defining the lower
boundary of the macroscopic scale), is illustrated in fig. 1.1. We choose to
reduce our volume of study to the smallest possible size before the effect of
individual molecules becomes meaningful.
Adopting this point of view, which is named the continuum abstraction, is not
a trivial decision, because the physical laws which determine the behavior
of molecules are very different from those which determine the behavior of
XKCD #2283: how (not) to pic-
ture big numbers elements of fluid. For example, in fluid mechanics we never consider any
by Randall Munroe (CC-by-nc)
https://xkcd.com/2283 inter-element attraction or repulsion forces; while new forces “appear” due
to pressure or shear effects that do not exist at a molecular level.

Figure 1.1: Measurement of the average value of a property (here, velocity 𝑉 ; but it
could be pressure, or temperature) inside a given volume. As the volume shrinks
towards zero, the fluid can no longer be treated as a continuum; and property
measurements will oscillate wildly.
Figure CC-by-sa Olivier Cleynen

10
A direct benefit of the continuum abstraction is that the mathematical com-
plexity of our problems is greatly simplified. Finding the solution for the

equation (eq. 4/15 p. 80) instead of a system of 2 ⋅ 1022 equations with 2 ⋅ 1022
bottle of “still air” mentioned above, for example, requires only a single

unknowns (all leading up to 𝑉⃗average,𝑥,𝑦,𝑧,𝑡 = 0⃗!).

Another consequence is that we cannot treat a fluid as if it were a mere set of
marbles with no interaction which would hit objects as they move by. Instead
we must think of a fluid –even a low-density fluid such as atmospheric air–
as an infinitely-flexible medium able to fill in almost instantly all of the space
made available to it.

1.2.3 Theory, numerics, and experiment

Today, fluid dynamicists are typically specializing in any one of three sub-
disciplines:

Analytical fluid mechanics which is the main focus of these lectures and
which consists in predicting fluid flows mathematically. As we shall
SMBC #2010-08-29: cooperation
see, it is only able to provide (exact) solutions for very simple flows. between theoretical and experi-
In fluid mechanics, theory nevertheless allows us to understand the mental scientists
by Zach Weinersmith
mechanisms of complex fluid phenomena, describe scale effects, and https://www.smbc-
comics.com/comic/2010-08-29
predict forces associated with given fluid flows;

Numerical fluid mechanics also called Computational Fluid Dynamicsw

or cfd, which consists in solving problems using very large numbers
of discrete values. Initiated as a research topic in the 1970s, cfd is now
omnipresent in the industry; it allows for excellent visualization and
parametric studies of very complex flows. Nevertheless, computational
solutions obtained within practical time frames are inherently approxi-
mate: they need to be challenged using analysis, and calibrated using
experimental measurements;

Experimental fluid mechanics which consists in reproducing phenom-

ena of interest within laboratory conditions and observing them using
experimental techniques. A very mature branch (it first provided useful
results at the end of the 19th century), it is unfortunately associated
with high human, equipment and financial costs. Experimental mea-
surements are indispensable for the validation of computational simu-
lations; meanwhile, the design of meaningful experiments necessitates
a good understanding of scale effects.

Our study of analytical fluid mechanics should therefore be a useful tool to

approach the other two sub-disciplines of fluid mechanics.

11
 1.3 Important concepts in mechanics
Mechanics in general deals with the study of forces and motion of bodies. A
few concepts relevant for us are recalled here.

1.3.1 Position, velocity, acceleration

The description of the movement of bodies (without reference to the causes
and effects of that movement) is called kinematics.w
The position in space of an object can be fully expressed using three com-

⃗ If the object moves, then this vector varies with time.

written 𝑥).
ponents (one for each dimension) This is usually done with a vector (here

The velocity 𝑉⃗ of the object is the rate of change in time of its position:

d𝑥⃗
𝑉⃗ ≡
d𝑡
(1/1)

The length 𝑉 of the velocity vector 𝑉⃗ is measured in m s−1 :

𝑉 ≡ ||𝑉⃗ || (1/2)

In this document, 𝑉 always is a positive number. Its formal name is speed,

but in practice the term velocity is used to designate either the vector or its
length, according to context.

(each having positive or negative values in m s−1 ) must be expressed. In this

In order to express the velocity vector completely, three distinct values

document, this is done with different notations. In Cartesian coordinates, we

have:
⎛ 𝑉𝑥 ⎞ ⎛ 𝑢 ⎞
𝑉⃗ = ⎜ 𝑉𝑦 ⎟ = ⎜ 𝑣 ⎟
⎜ ⎟ ⎜ ⎟
⎝ 𝑉𝑧 ⎠ ⎝ 𝑤 ⎠
In cylindrical coordinates, we write:

⎛ 𝑢𝑟 ⎞
𝑉 = ⎜ 𝑢𝜃
⃗ ⎟
⎜ ⎟
⎝ 𝑢𝑧 ⎠

The acceleration 𝑎⃗ of the object is the rate of change in time of its velocity:

d𝑉⃗
𝑎⃗ ≡
d𝑡
(1/3)

Acceleration is especially important in mechanics because it can be deduced

from Newton’s second law (see eq. 1/25 p. 20 below) if the forces applying
on the object are known. Acceleration can then be integrated with respect
to time to obtain velocity, which can be integrated with respect to time to

Like velocity, acceleration has three components (each measured in m s−2 ).

obtain position.

It shows at which rate each component of velocity is changing. It may not

always point in the same direction as velocity (fig. 1.2).

12
Figure 1.2: A body (black dot) is following a trajectory, plotted in blue. The accel-
eration vectors are plotted in red for two cases: on top, when the magnitude of
its velocity remains constant, and on the bottom, when this magnitude changes
continuously.
Figure CC-0 Olivier Cleynen

1.3.2 Forces and moments

as a magnitude, and so it is expressed with a vector (typically noted 𝐹⃗).

A force expresses an effort exerted on a body. Force has a direction as well

The three main types of forces relevant to fluid mechanics are those due
to pressure, those due to shear, and those due to gravity (the force due to

When a force exerts at a position 𝑟⃗ away from a reference point, it exerts a

gravity is called weight).

torsion (or “twisting”) effort named moment.w Like a force, a moment has a

vector 𝑀⃗ is expressed as the cross product of the arm 𝑟⃗ and the force 𝐹⃗:
direction as well as a magnitude, and so is best expressed with a vector. This

𝑀⃗ ≡ 𝑟⃗ ∧ 𝐹⃗ (1/4)

See Appendix A2.2 p. 248 for a short briefing about the cross product of
vectors.

1.3.3 Energy
Energy, measured in joules (J), is in most general terms the ability of a body
to set other bodies in motion. It can be accumulated or spent by bodies in a
large number of different ways. The most relevant forms of energy in fluid
mechanics are:

Kinetic energyw noted 𝐸k , accumulated as motion:

1
𝐸k ≡ 𝑚𝑉 2
2
(1/5)

13
 Workw noted 𝑊 , which is energy spent on displacing an object over a
distance 𝑙 with a force 𝐹 :

𝑊 ≡ 𝐹⃗ ⋅ 𝑙⃗ (1/6)
where 𝑊 is the work (J);
𝐹⃗ is the force (vector with magnitude in N);
and 𝑙⃗ is the movement distance (vector with magnitude in m).

See Appendix A2.1 p. 247 for a short briefing about the dot product of
vectors.

Internal energyw noted 𝐼 stored as heat within the body itself. As long
as no phase changes occurs, the internal energy 𝐼 of fluids is roughly
proportional to their absolute temperature 𝑇 .

1.4 Properties of fluids

Beyond velocity, which is the primary unknown for us in fluid mechanics,
there are a few other important fluid properties.

1.4.1 Density
The density 𝜌 (Greek letter rho) is the amount of mass per unit volume:
𝑚
𝜌≡

(1/7)

where 𝜌 is the density (kg m−3 );

𝑚 is the considered mass (kg);
and  is the considered volume (m3 ).

conditions, air has a density of approximately 𝜌air = 1,2 kg m−3 ; that of water
Two orders of magnitude that are useful to remember: at ambient atmospheric

is almost a thousand times greater, at 𝜌water = 1 000 kg m−3 .

Never confuse pressure 𝑝, which is how hard a fluid

Advice from an expert

pushes on walls, with density 𝜌, which is how much

of the fluid there is per unit volume. Not many
fluid dynamicists will confess to it, but most have, as
beginner students, mixed up the two symbols once
in a moment of weakness. Make sure you write
those two letters very clearly, so you distinguish them easily even when
stressed or distracted.

1.4.2 Phase
Fluids can be broadly classified into phases,w which are loosely-defined sets
of physical behaviors. Typically one distinguishes liquids which are fluids
with large densities on which surface tension effects play an important role,
from gases or vapors which have low densities and no surface tension effects.
Phase changes are often brutal (but under specific conditions can be blurred
14 or smeared-out); they usually involve large energy transfers. The presence of
multiple phases in a flow is an added layer of complexity in the description
of fluid phenomena.

1.4.3 Temperature
Temperaturew is a scalar property measured in Kelvins (an absolute scale). It
represents a body’s potential for receiving or providing heat and is defined,

We convert from Kelvins to degrees Celsius by subtracting 273,15 units:

in thermodynamics, based on the transformation of heat and work.

𝑇 (°C) = 𝑇 (K) − 273,15 (1/8)

Although we can “feel” temperature in daily life, it must be noted that the
human body is a very poor thermometer in practice. This is because we
constantly produce heat, and we infer temperature by the power our body
loses or gains as heat. This power not only depends on our own body
temperature (hot water “feels” hotter when we are cold), but also on the
heat capacity of the fluid (cold water “feels” colder than air at the same
temperature) and on the amount of convection taking place (ambient air
“feels” colder on a windy day).

is extremely high (𝑐air ≈ 1 kJ kg−1 K, 𝑐water ≈ 4 kJ kg−1 K). Unless very high
In spite of these impressions, the fact is that the heat capacity of fluids

velocities are attained, the temperature changes associated with fluid flow
are much too small to be measurable in practice.

Advice from an expert

Never make guesses about the dynamics of a flow using
your own perception of temperature. The human body
constantly rejects or absorbs heat, and makes for a very
bad thermometer. Wind may just feel cold because the
air movement increases heat transfer off your skin, and
not because of changes in air pressure or in density. If
needed, use a real thermometer!

1.4.4 Perfect gas model

When a gas has relatively simple molecules, moderate temperature and low

the perfect gas model.w Their absolute temperature 𝑇 is then modeled as a

pressure, several of its properties can be related easily to one another with

function of their pressure 𝑝 with a single, approximately constant parameter

𝑅specific ≡ 𝑝/𝜌𝑇 :
𝑝
= 𝑅specific 𝑇
𝜌
(1/9)

where 𝑅specific depends on the state and nature of the gas (J K−1 kg−1 );
and 𝑝 is the pressure (Pa, see §1.5.2 further down).

Note that 𝑅specific here is a specific gas constant (whose value depends on the
gas); chemists often instead use a universal definition of 𝑅 in J mol−1 K−1 .

an equation of state. When 𝑅 remains approximately constant, the fluid is

This type of model (relating temperature to pressure and density) is called
15
 said to behave as a perfect gas. The properties of air can satisfactorily be
predicted using this model. Other models existw which predict the properties
of gases over larger property ranges, at the cost of increased mathematical
complexity.
Many fluids, especially liquids, do not follow this equation and their tem-
perature must be determined in another way, most often with the help of
laboratory measurements.

1.4.5 Speed of sound

An important property of fluids is the speed at which pressure changes can
travel within the fluid (these pressure changes may for example be caused

molecules within the fluid, and it is called the speed of sound,w noted 𝑐.
by the movement of an object). This speed is equal to the average speed of

In fluid dynamics, we often quantify how fast the fluid is flowing relative to

ratio of the local fluid speed 𝑉 to the local speed of sound 𝑐:

the speed of sound. For this, we define the Mach numberw noted [Ma] as the

𝑉
[Ma] ≡
𝑐
(1/10)

Since both 𝑉 and 𝑐 can be functions of space in a given flow, [Ma] may not be
uniform (e.g. the Mach number around an aircraft in flight is different at the
nose and above its wings). Nevertheless, a single value is typically chosen to
identify “the” representative Mach number of any given flow.

at [Ma] ≤ 0,3, their density 𝜌 stays constant. Density variations in practice can
It is observed that providing no heat or work transfer occurs, when fluids flow

be safely neglected below [Ma] = 0,6. When the density is uniform, the flow
is said to be incompressible. Above these Mach numbers, it is observed that
when subjected to pressure variations, fluids exert work upon themselves,
which translates into measurable density and temperature changes: these are
called compressibility effects, and we will not study them in this course.
In most flows, the density of liquids is almost invariant – so that water flows
are generally entirely incompressible.

that 𝑐 depends only on the absolute temperature:

When the fluid is air (and generally within a perfect gas), it can be shown

√
𝑐 = 𝛾 𝑅𝑇 (1/11)

where 𝑐 is the local speed of sound (m s−1 );

in all cases for a perfect gas,

𝛾 is a gas property, approx. constant (dimensionless);

and 𝑇 is the local temperature (K).

Advice from an expert

When you are showering and decrease the temperature

colder: this is because it travels at speed 𝑉 in the pipe to

at the knob, it takes some time for the water to feel

the showerhead. But when you decrease the flow rate at

the decrease in pressure travels at speed 𝑐 in the pipe

the knob, the response is instantaneous: this is because

(5 000 km h−1 ). The dynamics of fluids become funky

16
 when the speed of the fluid and the speed of sound are comparable: this
is why the Mach number is so important to us.

1.4.6 Viscosity
We have said above that a fluid element can deform continuously under
pressure and shear efforts: it will never “snap” or break apart. However this
deformation is not “for free”: it will require force and energy inputs which
are not reversible (they are not reversed if the motion is reversed). Resistance
to straining in a fluid is measured with a property named viscosity.w Video: how to make sense of vis-
cosity
In informal terms, viscosity is the “stickiness” of fluids: for example, honey by Olivier Cleynen (CC-by)
https://youtu.be/5YUFt-V_kdk
and sugar syrups are more viscous than water.
More formally, viscosity is quantified as follows. Imagine a small brick-

in fig. 1.3. The continuous straining of the brick requires a force 𝐹 per unit
shaped element of fluid, which is deformed (strained) horizontally, as shown

area 𝐴, called shear stress 𝜏 (see also §1.5.3 further down). We expect that

and with the speed Δ𝑣 at which the brick is strained. Conversely, we expect
the shear increases with both the “stickiness” of the fluid —the viscosity—

that the shear will decrease when the element height Δ𝑦 is increased.
We define the viscosity 𝜇 as the ratio between the required shear stress,
𝜏 = 𝐹 /𝐴, and the rate at which the brick is strained, Δ𝑣/Δ𝑦:
𝜏
𝜇≡
( Δ𝑦 )
Δ𝑣
(1/12)

where 𝜇 is the viscosity (N s m−2 or Pa s);

𝜏
Δ𝑢 is the velocity difference between top and bottom planes in fig. 1.3 (m s−1 );
is the horizontal shear in fig. 1.3 (Pa);

and Δ𝑦 is the height difference between top and bottom planes in fig. 1.3 (m).

distance), and so this turns out as Pascal − seconds in si units. We will come
The dimension of viscosity is in (force per area) divided by (velocity per

back to shear and viscosity in chapter 5 (Effects of shear).

difference Δ𝑣 between the top and bottom surface. The horizontal force 𝐹 required
Figure 1.3: A brick-shaped element of fluid is strained, by applying a velocity
on the top edge, divided by the area 𝐴, is the shear stress 𝜏 (see also §1.5.3 p. 18
further down). The higher the required shear stress for a given strain rate, the more
viscous the fluid is.
Figure CC-0 Pm.schroeder/Olivier Cleynen
17
 1.5 Forces on fluids
Fluids are subjected to, and subject their surroundings and themselves to
forces. Identifying and quantifying those forces allows us to determine how
they will flow. Three types of forces are relevant in fluid dynamics: gravity,
pressure and shear.

1.5.1 Gravity

Earth) is expressed with a vector 𝑔⃗ pointing downwards. Within the Earth’s

Gravity (here, the attraction effort exerted on fluids at a distance by the

atmosphere, the magnitude 𝑔 of this vector varies extremely slightly with

altitude, and it may be considered constant at 𝑔 = 9,81 m s−2 . The weight
force exerted on an object of mass 𝑚 is then simply quantified as

𝐹⃗weight = 𝑚𝑔⃗ (1/13)

1.5.2 Pressure
The concept of pressurew can be approached with the following conceptual

the pressure 𝑝 will be the ratio of the perpendicular force 𝐹⟂ to the surface
experiment: if a flat solid surface is placed in a fluid at zero relative velocity,

area 𝐴:
𝐹⟂
𝑝≡
𝐴
(1/14)
where 𝑝 is the pressure (N m−2 or Pascals, 1 Pa ≡ 1 N m−2 );
𝐹⟂ is the component of force perpendicular to the surface (N);
and 𝐴 is the surface area (m2 ).

Although in si units pressure is measured in Pascals (1 Pa ≡ 1 N m−2 ), in prac-

tice it is often measured in bars (1 bar ≡ 1 ⋅ 105 Pa). Ambient atmospheric pres-
sure at normal altitude varies with the weather and is approximately 1 bar.
The pressure distribution within a fluid is related to their velocity distribution

pressure field 𝑝(𝑥,𝑦,𝑧,𝑡) , a function of space an time.

according to relations that we will study later. We shall then be looking for a

The pressure 𝑝 exerting on an element of fluid can be thought of as the

time- and space-average of the perpendicular component of impact force
of its molecules on its neighbors. It is strictly a macroscopic property, that
is, it cannot be defined at a microscopic level (there is no such thing as the
“pressure of a molecule”). In subsonic flows, it is a scalar property, meaning
that for a given particle, it is the same in all directions. Pressure effects are
explored in further detail in chapter 4 (Effects of pressure).

1.5.3 Shear
In the same thought experiment as above, the shearw , noted 𝜏 (greek letter
“tau”), expresses the efforts of a force parallel to a surface of interest:
𝐹∥
𝜏≡
𝐴
(1/15)

where 𝜏 is the shear (N m−2 or Pa);

𝐹∥ is the component of force parallel to the surface (N);
18 and 𝐴 is the surface area (m2 ).
Contrary to pressure, shear is not a scalar: it can (and often does) take

two components and could be represented by a vector 𝜏⃗ = (𝜏𝑥 , 𝜏𝑦 ). We will

different values in different directions: on the flat plate above, it would have

explore shear in further detail in chapter 5 (Effects of shear).

1.6 Basic flow quantities

A few fluid-flow related quantities can be quantified easily and are worth
listing here.

Mass floww is noted 𝑚̇ and represents the amount of mass flowing through
a chosen surface per unit time. When the velocity across the surface Video: how to deal with the ⟂
symbol when calculating mass
is uniform, it can be quantified as: flow

𝑚̇ = 𝜌𝑉⟂ 𝐴
by Olivier Cleynen (CC-by)
https://youtu.be/H5l–WlCZQ8
(1/16)
where 𝑚̇ is the mass flow (kg s−1 );
𝜌 is the fluid density (kg m−3 );
𝐴 is the area of the considered surface (m2 );
and 𝑉⟂ is the component of velocity perpendicular to the surface (m s−1 ).

Instead of 𝑉⟂ 𝐴, we might pick 𝑉𝐴⟂ : the velocity and the area of a

surface perpendicular to that velocity:

𝑚̇ = 𝜌𝑉𝐴⟂ (1/17)
where 𝑉 is the flow speed (m s−1 );
and 𝐴⟂ is the area of a surface perpendicular to the flow velocity (m2 ).

Volume floww is noted ̇ and represents the volume of the fluid flowing
through a chosen surface per unit time. Much like mass flow, when the
velocity is uniform, it is quantified as:
𝑚̇
̇ = 𝑉⟂ 𝐴 = 𝑉𝐴⟂ =
𝜌
(1/18)

where ̇ is the volume flow (m3 s−1 ).

𝑃̇ pressure necessary to force a mass flow of fluid through a chosen surface

Power to cross a surface is a time rate of energy transfer. The power

at a pressure 𝑝 is:

𝑃̇ pressure = 𝐹⃗pressure ⋅ 𝑉⃗fluid

= ̇ 𝑝
(1/19)

𝑚̇
(1/20)
= 𝑝
𝜌
(1/21)

where 𝑃̇ pressure is the power required to cross the surface (W);

and 𝑝 is the mean pressure at the surface (Pa).

If a fluid passes across a volume, the net power 𝑃̇ pressure, net required to
both enter and leave the volume may be expressed as
𝑚̇
𝑃̇ pressure, net = Δ𝑝
𝜌
(1/22)

where Δ𝑝 is the pressure difference between outlet and inlet.

19
 Power to increase temperature is also a time rate of energy transfer. If

long as no phase change occurs, the associated power 𝑃̇ temperature is:

the temperature of a fluid changes when it flows through a volume, as

𝑃̇ temperature = 𝑚̇ 𝑐fluid Δ𝑇 (1/23)

where 𝑃̇ temperature is the power to increase temperature (W);
Δ𝑇 is the temperature change occurring in the fluid (K);
and 𝑐fluid is the specific heat capacity of the fluid (J K−1 kg−1 ).

The heat capacityw 𝑐fluid of fluids varies strongly according to the

in a steady flow, the heat capacity is termed 𝑐𝑝 . In fluids such as liquid

amount of work that they are performing. When no work is performed

water and air, this capacity is almost independent of temperature.

As we will see in chapter 2 (Analysis of existing flows with one dimension), fluid
flow involves many forms of energy changes. We will learn to combine and
compare them progressively.

1.7 Four balance equations

Most problems in fluid mechanics are solved by applying basic physical prin-
ciples. We write out those principles in the form of four balance equations.

1. Mass balance:
The total amount of matter at hand in a given phenomenon must
remain constant (since in fluid mechanics, we do not usually consider
nuclear reactions). This statement can be expressed as:

𝑚system = cst
d𝑚system
=0
d𝑡
(1/24)

2. Balance of linear momentum:

law,w which states that the net force 𝐹⃗net ≡ Σ𝐹⃗ applying to any given
The momentum balance equation is a formulation of Newton’s second

system is equal to its mass 𝑚 times its acceleration. In fluid mechanics,

time of the system’s linear momentum,w the quantity 𝑚𝑉⃗ :

the most useful formulation of this physical law uses the change in

d
𝐹⃗net = 𝑚𝑉⃗ )
d𝑡 (
(1/25)

3. Balance of angular momentum:

This is a different form of Newton’s second law, useful in situations

plied about an axis. It states that the net moment 𝑀⃗ net, X ≡ Σ𝑀⃗X applied
where rotation about an axis, or moments (“twisting” efforts) are ap-

momentumw about this same point, the quantity 𝑟⃗ ∧ 𝑚𝑉⃗ :

on a system about a point X is equal to the time change of its angular

d
𝑀⃗ net, X = 𝑟⃗ ∧ 𝑚𝑉⃗ )
d𝑡 (
(1/26)

4. Balance of energy:
20 This equation, also known as the “first principle of thermodynamics”,w
 states that the total amount of energy within an isolated system must
remain constant:
d𝐸isolated system
=0
d𝑡
(1/27)

In special cases, further equations are used to describe other phenomena af-
fecting the fluid flow (e.g. chemical reactions, or interaction between phases).
In most cases however, the four equations above are the only important equa-
tions written in fluid mechanics. We usually apply those balance statements
to our problem in either one of two ways:

• We may have information about a flow which already exists, and want
to calculate how fluid properties change as it flows through the area
of interest, and what the related forces are. In that case, we write
the equations in an integral form: we will do this in chapters 2 and 3
(Analysis of existing flows).

• We may instead wish to predict how the fluid is going to flow through
our zone of interest. In order to do this, we need to calculate flow
properties in an extensive manner, aiming to obtain vector fields for
the velocity and pressure everywhere, at all times. To this effect, we
write the equations in a differential form: we will do this in chapter 6
(Prediction of fluid flows).

1.8 Classification of fluid flows

As we will see progressively, it is extremely difficult to obtain general so-
lutions for fluid flow. Thus, whenever possible or reasonable, simplifying
hypothesis are made about the behavior of any particular flow, that allow
us to proceed with the analysis and obtain a reasonable, if inexact, specific
solution. It is therefore a habit of fluid dynamicists to classify fluid flows in
various categories, which are not necessarily incompatible. When approach-
ing a given problem, we typically look out for the following characteristics:

Time dependence Flows which do not vary with time are called steady.
Steadiness is dependent on the chosen point of view: for example,
the air flow around an aircraft in cruise flight is seen to be steady
from within the aircraft, but is obviously highly unsteady from the
point of view of the air particles directly in the path of the airliner.
Steadiness is also dependent on the selection of a suitable time frame:
time-variations in a flow may be negligible if the time window used in
the study is short enough.

tive 𝜕/𝜕𝑡 or the total time derivative D/D𝑡 —concepts we will study
Mathematically, flows are steady when either the partial time deriva-

in chapter 6 (Prediction of fluid flows) — of properties is zero: this consid-

erably eases the search for solutions. It is important not to confuse
steadiness (the fields of properties remain the same as time passes)
with uniformity (the properties are the same everywhere in space).
To find out whether or not a flow is steady, compare instantaneous
representations (photos, measurement readings, vector field represen-
tations, etc.) taken at different times. If they are all identical, the flow
is steady. There does not exist an analytical method, however, that
allows us to predict whether a flow will be steady. 21
 Compressibility When the density of the fluid is uniform, the flow is said to
be incompressible. The term is treacherous, because it refers to density,
not pressure (incompressible flows almost always feature non-uniform
pressure fields).

number (eq. 1/10 p. 16). Below 0,3 the flow is always incompressible.
To find out whether a gas flow is compressible, compute the Mach

Compressibility effects can be reasonably neglected below [Ma] = 0,6.

Unless very specific phenomena such as phase changes or extreme
speeds occur, the flow of liquids is always incompressible.

Advice from an expert

friend: it has nothing to do with pressure 𝑝, and all

The word “incompressible” is a really mean false

with density 𝜌. Most flows with moderate speeds

and powers are incompressible: the density does
not change.
Pressure will always change in space or even in
time if there is a fluid flow, so there is no word for “pressure
remains the same” in fluid mechanics.

Temperature distribution In a fluid flow, temperature changes can occur

due to three phenomena:

• Heat transfer from solid bodies;

• Changes in pressure and density due to work being performed on
the fluid (by moving solid bodies, or by the fluid itself);
• Heat created through internal friction within the fluid.

To be certain that a fluid flow will have uniform temperature (i.e.

whether it is isothermal), therefore, we must take three steps:

1. Quantify the heat transfer from external bodies. If that is zero,

the flow is at least adiabatic, meaning there is no heat transfer;
2. Find out whether the flow is incompressible. As long as it is, the
density changes are negligibly small and no temperature changes
will occur due to compression or expansion of the fluid;
3. Quantify the mechanical energy lost every second by the fluid
as it flows through the domain of interest. This power can be
either transmitted to a moving part (e.g. a turbine), or dissipated
internally through friction, as heat.
Because the heat capacity of fluids is generally very high, temper-
ature changes due to internal friction are usually negligibly small.
This is assessed with an example in exercise 1.7 p. 29.

Turbulence One last characteristic that we systematically attempt to iden-

tify in fluid flows is turbulencew (or its opposite, laminarity). While
laminar flows are generally very smooth and steady, turbulent flows
feature multiple, chaotic velocity field variations in time and space.
We shall first approach the concept of turbulence in chapter 7 (Pipe
flows), and study it more formally in chapter 9 (Dealing with turbulence). In
the meantime, we can predict whether a flow will become turbulent
22
 by using a non-dimensional parameter named the Reynolds number,
noted [Re]:
𝜌𝑉 𝐿
[Re] ≡
𝜇
(1/28)

using the fluid properties 𝜌 and 𝜇, a representative fluid speed 𝑉 , and a

To find out whether a flow will become turbulent, quantify [Re] by

length 𝐿 which is representative of the flow domain (for example, the

of 104 or more, the flow is very likely to become turbulent over the
length or width of an obstacle in the flow). If the result is on the order

length 𝐿. By contrast, with [Re] of the order of 102 or less, the flow is
very likely to remain laminar over this length.
This crude quantification, of course, deserves more explanation —
we will be coming back to the Reynolds number in chapters 7 and
following.

Advice from an expert

The Reynolds number is the best mea-
sure of how complex a flow is — how
intricate and chaotic the movement
of the fluid will be. When they want
to compare themselves to their peers
(and maybe try to impress them), fluid
dynamicists won’t ask about flow speed or mass flow, but instead
just ask: “what’s your Reynolds number?”

1.9 Limits of fluid dynamics

Fluid dynamics is a complex discipline.
It is easy to observe that flows as ordinary as sea waves crashing on a reef,
water flowing down a river stream, or air blown into one’s hands, display
tremendous geometrical complexity. Even after choosing to describe only the
movement of macroscopic fluid particles instead of individual molecules (and
thereby avoiding studying thousands of billions of individual movements),
we still need to describe a three-dimensional field of properties (pressure,
temperature, etc.), one of which, velocity, is itself three-dimensional.
Thus, even before we begin describing the exact problem and a procedure to
obtain its solution, we know that the mere description of a solution can have
tremendous complexity.
Additionally, we have to admit that in practice much progress can be made in
the field of fluid dynamics. For example, our weather forecasts have almost no
value beyond one week, and aircraft manufacturers with budgets measured
in billions of dollars still make extensive use of wind tunnel models – this
despite our staggering continuous rate of progress in computing technology,
and many decades of efforts dedicated to analytical fluid dynamics.
In our present study of fluid dynamics, therefore, we shall proceed modestly,
and will always take care to show the limits of our analysis.

23
 1.10 Solved problems

Force due to pressure on a plate

A large, 3×3 m square aquarium
window (fig. 1.4) is subjected to
pressure from fluids on each side.

exerts uniform pressure 𝑝air as:

On the right side, the atmosphere

𝑝air = 𝑝atmosphere (1/29)

On the other side, water from

pressure 𝑝water expressed in Pascals plate: on the left, exerted by water; on

the aquarium exerts non-uniform Figure 1.4: Pressure distribution on a
the right, exerted by air.
as:

𝑝water = 1,2 ⋅ 105 + 9,81 ⋅ 103 × 𝑥

Figure CC-0 Olivier Cleynen

(1/30)

What is the force resulting from fluid pressure on each side of the plate?

See this solution worked out step by step on YouTube

https://youtu.be/2PE74f6fIMM (CC-by Olivier Cleynen)

incorrect! The correct result is 1,2124 MN. The method, numbers, equations etc.
Note: Unfortunately Olivier made an error in this video: the end computation is

are all ok — only the final result is affected. Many thanks to the students who
double-checked and reported the problem!

24
A sailboat travels at velocity 𝑉⃗
Power and moment resulting from a force

with 𝑉 =1,5 m s−1 (see fig. 1.5). The

relative wind comes from the back
and from the left; it acts on

namic force 𝐹⃗. The magnitude is

the sail and results in an aerody-

𝐹 =13 kN; the force acts at an angle

𝜃=30°, 2 m ahead of the center of
gravity.
What is the power contributed by
the wind, and what is the moment
exerted by the aerodynamic force
about the center of gravity?

eling at velocity 𝑉⃗ . The wind exerts a

Figure 1.5: A boat seen from above, trav-

force 𝐹⃗ on the boat, a little forward of

its center of gravity.
Figure CC-0 Olivier Cleynen

See this solution worked out step by step on YouTube

https://youtu.be/yLqsGF-R9ps (CC-by Olivier Cleynen)

25
26
 Problem sheet 1: Basic flow quantities
last edited September 3, 2020
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

1.1 Reading quiz

Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.

1.2 Compressibility effects

An aircraft is flying in air with density 0,9 kg m−3 and temperature −5 °C. Above which
flight speed would you expect the air flow over the wings to become compressible?

1.3 Pressure-induced force

A flat, 2 m-by-2 m panel is used as the wall of a swimming pool (fig. 1.6). On the left side,
the pressure is uniform at 1 bar.

Figure 1.6: Pressure distribution on a flat panel that is part of the wall of a swimming pool
Figure CC-0 Olivier Cleynen

1.3.1. What is the pressure force (i.e. the force resulting from the pressure) exerted on
the left side of the plate?

increases with depth. The relation, expressed in pascals, is:

On the right side of the plate, the water exerts a pressure which is not uniform: it

𝑝water = 1,3 ⋅ 105 − 9,81 ⋅ 103 × 𝑧 (1/31) 27

 1.3.2. What is the pressure force exerted on the right side of the plate?
[Hint: we will explore the required expression in chapter 4 (Effects of pressure) as
eq. 4/3 p. 74]

1.4 Shear-induced force

A fluid flows over a 3 m by 3 m flat horizontal plate, in the 𝑥-direction as shown in fig. 1.7.
Because of this flow, the plate is subjected to uniform shear 𝜏𝑧𝑥 = 1,65 Pa.

Figure 1.7: Shear force exerting on a plate

Figure CC-0 Olivier Cleynen

1.4.1. What is the shear force applying on the plate?

function of 𝑥 expressed (in pascals) as 𝜏𝑧𝑥 = 1,65 − 0,01 × 𝑥 2 ?

1.4.2. What would be the shear force if the shear was not uniform, but instead was a

[Hint: we will explore the required expression in chapter 5 (Effects of shear) as eq. 5/3
p. 92]

1.5 Speed of sound White [22] P1.87

Isaac Newton measured the speed of sound by timing the interval between observing

The cannon is shot 8,4 km away from Newton. What is the air temperature if the
smoke produced by a cannon blast and the hearing of the detonation.

measured interval is 24,2 s? What is the temperature if the interval is 25,1 s?

1.6 Wind on a truck

A truck moves with constant speed 𝑉⃗ on a road, with 𝑉 = 50 km h−1 . It experiences strong

force 𝐹⃗ with 𝐹 = 5 kN at an angle 𝜃 = 20°, as shown in fig. 1.8.

relative wind coming from the back and from the right: this results in an aerodynamic

Figure 1.8: Top view of a truck traveling at velocity 𝑉⃗ and subject to a aerodynamic force 𝐹⃗
Figure CC-0 Olivier Cleynen

28
 1.6.1. What is the power given by the wind to the truck?

The force 𝐹⃗ is applying at a distance 0,8 m behind the center of gravity of the truck.

namic force 𝐹⃗ about the center of gravity?

1.6.2. What are the magnitude and the direction of the moment exerted by the aerody-

1.7 Go-faster exhaust pipe

cylindrical outlet pipe, whose outlet is slanted at an angle 𝜃 = 25° to improve the good
The engine exhaust gases of a student’s hot-rod car are flowing quasi-steadily in a

looks of the car and provide the opportunity for an exercise.

Figure 1.9: Exhaust gas pipe of a car. The outlet cross-section is at an angle 𝜃 relative to the axis
of the pipe.
Figure CC-0 Olivier Cleynen
Photo cropped, mirrored and edited from an original CC-by-sa by kazandrew2

The outlet velocity is measured at 15 m s−1 , and the exhaust gas density is 1,1 kg m−3 . The
slanted outlet section area 𝐴 is 420 cm2 .

1.7.1. What is the mass flow 𝑚̇ through the pipe?

1.7.2. What is the volume flow ̇ of exhaust gases?

pressure loss of 21 Pa – we will learn to quantify this in chapter 7 (Pipe flows). In these
Because of the shear within the exhaust gases, the flow through the pipe induces a

conditions, the specific heat capacity of the exhaust gases is 𝑐𝑝gases = 1 100 J kg−1 K−1 .

1.7.3. What is the power required to carry the exhaust gases through the pipe?
1.7.4. What is the gas temperature increase due to the shear in the flow? 29
 1.8 Acceleration of a particle

particle of width 0,1 mm. The particle is accelerating at a rate of 2,5 m s−2 .
Inside a complex, turbulent water flow, we are studying the trajectory of a cubic fluid

1.8.1. What is the net force applying to the particle?

1.8.2. In practice, which types of forces could cause it to accelerate?

1.9 Flow classifications

1.9.1. Can an incompressible flow also be unsteady?

1.9.2. Can a very viscous fluid flow in a turbulent manner?

1.9.3. [more difficult] Can a compressible flow also be isothermal?

1.9.4. Give an example of an isothermal flow, of an unsteady flow, of a compressible

flow, and of an incompressible flow.

30
Answers
1.2 If you adopt [Ma] = 0,6 as an upper limit, you will obtain 𝑉max = 709 km h−1 (eqs. 1/10
& 1/11 p. 16). Note that propellers, fan blades etc. will meet compressiblity effects
far sooner.

1.3 1) 𝐹left = 400 kN (eq. 1/14 p. 18); 2) 𝐹right = 480 kN (eq. 4/3 p. 74).

1.4 1) 𝐹1 = 14,85 N (eq. 1/15 p. 18); 2) 𝐹2 = 14,58 N (eq. 5/3 p. 92).

1.5 26,7 °C & 5,6 °C.

1.6 1) 𝑊̇ = 𝐹⃗aero ⋅ 𝑉⃗truck = 65,3 kW. See Appendix A2.1 p. 247 for a short briefing about

⎛ 0 ⎞
the dot product of vectors;

2) 𝑀 = ||⃗𝑟 ∧ 𝐹aero || = 1 368 N m, 𝑀 = ⎜ 0 ⎟ (points vertically upwards). See

⃗ ⃗
⎜ ⎟
⎝ −1 368 ⎠
Appendix A2.2 p. 248 for a short briefing about the cross product of vectors.

1) 𝑚̇ = 0,2929 kg s−1 (eq. 1/16 p. 19);

2) ̇ = 266,2 L s−1 (eq. 1/18 p. 19);
1.7

3) 𝑊̇ = 5,59 W (eq. 1/20 p. 19);

4) Δ𝑇 = +0,0174 K (eq. 1/23 p. 20), an illustration of remarks made in §1.8 p. 22
regarding temperature distribution.

1.8 1) 𝐹net = 2,5 ⋅ 10−9 N (eq 1/25 p. 20), such are the orders of magnitude involved in
cfd calculations!
2) Only three kinds: forces due to pressure, shear, and gravity.

1.9 1) yes, 2) yes if [Re] is high enough, 3) yes (in very specific cases such as high
pressure changes combined with high heat transfer or high irreversibility, therefore
generally no), 4) open the cap of a water bottle and turn it upside down: you have
an isothermal, unsteady, incompressible flow. An example of compressible flow
could be the expansion in a jet engine nozzle.

31
32
 Fluid Dynamics
Chapter 2 – Analysis of existing flows
with one dimension
last edited May 22, 2020
by Olivier Cleynen — https://fluidmech.ninja/

2.1 Motivation 33
2.2 One-dimensional flow problems 33
2.3 Balance of mass 35
2.3.1 Mass balance equation 35
2.3.2 Problems with the mass balance equation 36
2.4 Balance of momentum 37
2.5 Balance of energy 39
2.6 The Bernoulli equation 41
2.6.1 Theory 41
2.6.2 Reality 42
2.7 Solved problems 44
2.8 Problems 45

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

2.1 Motivation
In this chapter, we learn to analyze fluid flows for which a lot of information
is already available. We want, when confronted to a simple flow (for example,
flow entering and leaving a machine), to be able to answer three questions:

• What is the mass flow in each inlet and outlet?

• What is the force required to move the flow?

• What energy transfers are required for this movement?

2.2 One-dimensional flow problems

The method we develop here is called integral analysis, because it involves
calculating the overall (integral) effect of the fluid flowing through a consid-
ered volume. In this chapter, we consider one-dimensional flows (at least in a
loose definition); we will consider more advanced cases in chapter 3 (Analysis
of existing flows with three dimensions).
For now, we are interested in flows where four conditions are met:
Abstruse Goose #237: how many
dimensions are enough for you?
1. There is a clearly identified inlet and outlet; by an anonymous artist (CC-by-nc)
https://abstrusegoose.com/237

2. At inlet and outlet, the fluid properties are uniform, so that they can be
evaluated in bulk (e.g. the inlet has only one velocity, one temperature
etc.);

3. There are no significant changes in flow direction;

33
 4. A lot of information is available about the fluid properties at inlet.

Providing that those conditions are met, we can answer the question: what is
the net effect of the fluid flow through the considered volume?
In order to write useful equations, we need to begin with rigorous definitions,
with the help of figure 2.1:

• We call control volume a certain volume we are interested in. Fluid

flows through the control volume.
In this chapter, the control volume does not change with time. The
fluid flow does not change in time, either (i.e. the flow is steady). The
fluid enters and leaves the control volume at a clearly-identifiable inlet
and outlet.
• At a certain instant, the mass of fluid that is inside the control volume
is called the system. The system is traveling. At a later point in time, it
has moved and deformed.

The three equations that we write in this chapter state that basic physical
laws apply to the system. They are balance equations (see §1.7 p. 20). Each
time, we will express what is happening to the system, as a function of
the fluid properties at the inlet and outlet of the control volume. This
will allow us to answer three questions:

• What is the mass flow entering and leaving the control volume?
• What is the force required to move the flow through the control vol-
ume?
• What energy transfer is required to move the flow through the control
volume?

Figure 2.1: A control volume within a flow. The system is the amount of mass
included within the control volume at a given time. Because mass enters and leaves
the control volume, the system is being moved and deformed (bottom).
Figure CC-0 Olivier Cleynen
34
2.3 Balance of mass

2.3.1 Mass balance equation

How much mass flow is coming in and out of the control volume? We answer
this question by writing a mass balance equation. It compares the rate of
change of the system’s mass (which by definition is zero, see eq. 1/24 p. 20),
to the flow of mass through the borders of the control volume:
d𝑚sys d
=0= 𝑚CV + 𝑚̇ net
d𝑡 d𝑡
(2/1)

=0= +
the rate of change of the rate of change the net mass flow
the fluid’s mass of mass inside at the borders
as it transits the considered volume of the considered volume

Since we are here interested only in steady flows, d𝑚CV / d𝑡 = 0. Furthermore,

mass flow 𝑚̇ net . Equation 2/1 becomes:

we have clearly-identified inlets and outlets allowing us to re-express the net

0 = Σ𝑚̇ incoming + Σ𝑚̇ outgoing (2/2)

= +
the sum of the sum of
time rate of creation
incoming mass flows outgoing mass flows
or destruction of mass
(negative terms) (positive terms)

For steady flow through a fixed considered volume.

The sign convention is counter-intuitive: mass flows are negative inwards

and positive outwards. For example, in a case where there were two inlets
and two outlets, we could write:

0 = 𝑚̇ in 1 + 𝑚̇ in 2 + 𝑚̇ out 1 + 𝑚̇ out 2
̇ in 1 − |𝑚|
0 = − |𝑚| ̇ in 2 + |𝑚|
̇ out 1 + |𝑚|
̇ out 2

We can substitute 𝑚̇ = 𝜌𝑉⟂ 𝐴 (eq. 1/16 p. 19) into the equations above, obtain-
ing:

0 = Σ [𝜌𝑉⟂ 𝐴]incoming + Σ [𝜌𝑉⟂ 𝐴]outgoing (2/3)

For steady flow through a fixed considered volume.

Looking again at an example case where there were two inlets and two
outlets, this equation 2/3 would become:

0 = 𝜌in 1 𝑉⟂ in 1 𝐴in 1 + 𝜌in 2 𝑉⟂ in 2 𝐴in 2 + 𝜌out 1 𝑉⟂ out 1 𝐴out 1 + 𝜌out 2 𝑉⟂ out 2 𝐴out 2
0 = (𝜌𝑉⟂ 𝐴)in 1 + (𝜌𝑉⟂ 𝐴)in 2 + (𝜌𝑉⟂ 𝐴)out 1 + (𝜌𝑉⟂ 𝐴)out 2
0 = − (𝜌|𝑉⟂ |𝐴)in 1 − (𝜌|𝑉⟂ |𝐴)in 2 + (𝜌|𝑉⟂ |𝐴)out 1 + (𝜌|𝑉⟂ |𝐴)out 2
For steady flow through a fixed considered volume with two inlets and two outlets.

In a simple case where there is only one inlet and one outlet, this last equation
can be rewritten as

(𝜌|𝑉⟂ |𝐴)1 = (𝜌|𝑉⟂ |𝐴)2 (2/4) 35

 For steady flow through a fixed considered volume with one inlet and one outlet.

2.3.2 Problems with the mass balance equation

mis-use this equation is to draw the conclusion that “if 𝐴 decreases, then 𝑉
The equation 2/4 above is interesting, but also treacherous. The best way to

must increase”. This is only true some of the time, and here are two reasons
why:

1. The density 𝜌 may change between inlet and outlet. In low-speed flows
Video: Dangers associated with

without heat transfer, 𝜌 does not vary significantly (see §1.8 p. 1.8).
the mass balance equation
by Olivier Cleynen (CC-by)
https://youtu.be/M1Dw6qU-FOU

or when compressed air is expanded), 𝜌 may vary together with 𝐴

But in compressible flows (for example when combustion is involved,

and 𝑉 . Typically, in supersonic flows (where [Ma]>1, and the fluid

moves faster than the speed of sound), increases in 𝐴 lead to increases
in 𝑉 , because of a decrease in 𝜌.

flow, it may well be that reducing 𝐴2 leads to an increase in 𝑉2 , but

2. There is no causal relationship in equation 2/4. In an incompressible

other words, reducing 𝐴2 may both increase 𝑉2 and decrease 𝑚. ̇ In-

nothing guarantees that the product of the two remains constant. In

creases in velocity are not “for free”: they require force be applied and
energy be spent. The mass balance equation cannot account for those
phenomena.

Advice from an expert

Remember the title of the chapter: the tools here are
for analyzing existing flows: those for which we can,
if needed, gather more information by making mea-
surements. If you find yourself predicting velocity in a
machine you design with just a mass balance equation,
then you might quickly find yourself making unrealistic
assumptions. Immediately check what force and power
are required to generate this velocity. For this, you need a momentum
balance equation, and an energy balance equation.

36
2.4 Balance of momentum
What force is applied to the fluid for it to travel through the control volume?
We answer this question by writing a momentum balance equation. It
compares the rate of change of the system’s momentum (which by definition
is the net force applying to it, see eq. 1/25 p. 20), to the flow of momentum
through the borders of the control volume:

d(𝑚𝑉⃗sys ) d
= 𝐹⃗net = 𝑚𝑉⃗ ) + ⃗
(𝑚̇ 𝑉 )net
d𝑡 d𝑡 ( CV
(2/5)

= 𝐹⃗net = +
the rate of change of the rate of change the net flow of momentum
the fluid’s momentum of momentum through the boundaries
as it transits within the considered volume of the considered volume

Since we are here interested only in steady flows, and we have clearly-
identified inlets and outlets, this becomes:

𝐹⃗net = Σ (𝑚̇ 𝑉⃗ ) + Σ (𝑚̇ 𝑉⃗ )

incoming outgoing
(2/6)

= +
the vector sum the sum of incoming the sum of outgoing
of forces momentum flows momentum flows
on the fluid (with negative 𝑚̇ terms) (with positive 𝑚̇ terms)

The same convention as above is applied for the sign of the mass flow 𝑚. ̇
For example, in a case where there were one inlet and one outlet, we would
write:

𝐹⃗net = ⃗
(𝑚̇ 𝑉 )in + (𝑚̇ 𝑉⃗ )

𝐹⃗net = − (|𝑚| ̇ 𝑉⃗ )
̇ 𝑉⃗ ) + (|𝑚|
out

in out

For steady flow through a fixed considered volume with one inlet and one outlet.

As before, we can substitute 𝑚̇ = 𝜌𝑉⟂ 𝐴 (eq. 1/16 p. 19) into the equations
above, obtaining:

𝐹⃗net on fluid = Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ] + Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ] (2/7)

incoming outgoing

where 𝑉⟂ is negative inwards, positive outwards.

For steady flow through a fixed considered volume,

In the example case where there is one inlet and one outlet, we would write:

𝐹⃗net = ⃗
(𝜌𝑉⟂ 𝐴𝑉 )in + (𝜌𝑉⟂ 𝐴𝑉⃗ )

𝐹⃗net = − (𝜌|𝑉⟂ |𝐴𝑉⃗ ) + (𝜌|𝑉⟂ |𝐴𝑉⃗ )

out

in out

volume is traversed by a steady flow with mass flow 𝑚,̇ with one inlet (point 1)
To make clear a few things, let us focus on the simple case where a considered

⃗
and one outlet (point 2). The net force 𝐹net applying on the fluid is

𝐹⃗net = |𝑚|
̇ (𝑉⃗2 − 𝑉⃗1 ) (2/8)
37
 Three remarks can be made about this equation. First, we need to be aware

express 𝐹⃗net , we need to calculate its three components:

that this is not one, but three equations, one for each dimension. In order to

⎧
⎪ 𝐹 = |𝑚|̇ (𝑉2𝑥 − 𝑉1𝑥 )
⎪ net𝑥
⎨ 𝐹net𝑦 = |𝑚|̇ (𝑉2𝑦 − 𝑉1𝑦 )
⎪
⎪
Video: The net force thing in the

𝐹
⎩ net𝑧 = | ̇
𝑚| (𝑉2𝑧 − 𝑉1𝑧 )
momentum balance equation (2/9)
by Olivier Cleynen (CC-by)
https://youtu.be/AOcGNeY9ad0

Second, there are two reasons why we could calculate a non-zero net force
on the fluid in equation 2/8, as illustrated in fig. 2.2.

1. Even if 𝑉⃗2 is aligned and in the same direction as 𝑉⃗1 , they can be of
different magnitude. A force is required to accelerate or decelerate
the fluid (more precisely, an acceleration or deceleration of the fluid is
equivalent to a force);

2. Even if 𝑉⃗2 has the same magnitude as 𝑉⃗1 , they can have different
directions. A force is required to change the direction in which a flow
is flowing (or more precisely, a change of direction is equivalent to a
force).

We will explore these phenomena in greater detail in chapter 3 (Analysis of ex-

isting flows with three dimensions). In this current chapter, we are mostly interested
in one-dimensional flows, and it will suffice for us to solve equation 2/8 in
one suitable direction only, for example,

̇ (𝑉2𝑥 − 𝑉1𝑥 )
𝐹net𝑥 = |𝑚| (2/10)

The final remark is that the equation does not describe a cause-effect re-
lationship. The net force does not cause the change in velocity any more

and simultaneous. Similarly, we have no way to know what 𝐹⃗net is made of.
than the change in velocity causes the net force: they are both equivalent

The exact mechanism which adds up to a net force (pressure change, shear
applied through a static wall, the movement of a turbine, etc.) is “hidden” in
the control volume, and unknown to us. In order to find out what happens
in the control volume, we need a different type of analysis, which we will
approach in chapter 6 (Prediction of fluid flows).

Figure 2.2: Two reasons can explain why a net force 𝐹⃗net appears in eq. 2/8. On
the left, 𝑉⃗2 and 𝑉⃗1 are aligned, but have different lengths. On the right 𝑉⃗2 and 𝑉⃗1
have the same length, but different directions. We will look at the second case in
chapter 3 (Analysis of existing flows with three dimensions).
Figure CC-0 Olivier Cleynen
38
2.5 Balance of energy
What power is applied to the fluid for it to travel through the control volume?
We answer this question by writing an energy balance equation. It compares
the rate of change of the system’s energy, to the flow of energy through the

For this we prefer to express the energy 𝐸 as the specific energy 𝑒 (in J kg−1 )
borders of the control volume.

multiplied by the mass 𝑚 (kg). The time rate change of 𝑚𝑒 is measured in

watts (1 W ≡ 1 J s−1 ). The energy balance equation is then:
d𝑚𝑒sys d
= Σ (𝑄̇ + 𝑊̇ ) = (𝑚𝑒)CV + ̇ net
(𝑚𝑒)
d𝑡 d𝑡
(2/11)

= = +
the rate of change of the sum of the rate of change the net flow of energy
the fluid’s energy powers as of energy within through the boundaries
as it transits heat and work the considered volume of the considered volume

where 𝑄̇ is the power transferred as heat (W)

𝑊̇ is the power transferred as work (W)

Let us examine the terms of this equation.

The sum of powers as heat and work can be broken down as three compo-
nents:

• the net power transferred as heat 𝑄̇ net (positive inwards);

• the net power transferred as work with moving solid surfaces 𝑊̇ surfaces, net
(for example, a moving piston, turbine blade, or rotating shaft, positive
inwards);

• and the net power transferred to and from the fluid by the fluid itself,
in order to enter and leave the considered volume. This power is called

have 𝑊̇ pressure = −𝑚(𝑝/𝜌).

̇
power to cross a surface (see §1.6 p. 1.6); for each inlet or outlet we

We can thus write:

Σ (𝑄̇ + 𝑊̇ ) = 𝑄̇ net + 𝑊̇ shaft, net + 𝑊̇ pressure

𝑝
(2/12)

= 𝑄̇ net + 𝑊̇ shaft, net − 𝑚̇

( 𝜌 )net
(2/13)

Turning now to the specific energy 𝑒, we break it down into three components
(see also §1.3.3 p. 13):

• the specific internal energy 𝑖, which represents the energy per unit

ics, this is often noted 𝑢, but in fluid mechanics we reserve this symbol
mass contained as stored heat within the fluid itself. In thermodynam-

to note the 𝑥-component of velocity. In a perfect gas, 𝑖 is simply pro-

portional to absolute temperature (𝑖 = 𝑐𝑣 𝑇 ), but for other fluids such as
water, it cannot be easily measured, and precomputed tables relating 𝑖
to other properties must be used;

• the specific kinetic energy 𝑒𝑘 ,

1
𝑒𝑘 ≡ 𝑉 2
2
(2/14) 39
 • the specific potential energy 𝑒𝑝 , related to gravity 𝑔 (m s−2 ) and altitude
𝑧 (m) as:

𝑒𝑝 ≡ 𝑔𝑧 (2/15)

We thus write out specific energy 𝑒 (in J kg−1 ) as:

𝑒 ≡ 𝑖 + 𝑒𝑘 + 𝑒𝑝 (2/16)

Now, we focus on steady flows (for which energy in the control volume does
not change with time), and we can come back to eq. 2/11 to rewrite it as:

𝑄̇ net + 𝑊̇ shaft, net + 𝑊̇ presure = (𝑚̇ 𝑒)net

𝑝
𝑄̇ net + 𝑊̇ shaft, net = (𝑚̇ 𝑒)net + 𝑚̇
( 𝜌 )net
𝑝 1
𝑄̇ net + 𝑊̇ shaft, net = 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]net
(2/17)

Rewriting this into one general, usable form, we obtain:

𝑝 1
𝑄̇ net + 𝑊̇ shaft, net = Σ 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]in
𝑝 1
+Σ 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]out
(2/18)

where 𝑚̇ = 𝜌𝑉⟂ 𝐴 is negative inwards, positive outwards.

For steady flow through a fixed considered volume,

This equation 2/18 is known in thermodynamics as the steady flow energy

concept of enthalpy ℎ ≡ 𝑖 + 𝑝/𝜌, which we do not use here).

equation (in thermodynamics, it is usually expressed with the help of the

As usual, let us focus on a case where there is only one inlet and one outlet.
We obtain:
𝑝 1 𝑝 1
𝑄̇ net + 𝑊̇ shaft, net = − |𝑚|̇ 𝑖 + + 𝑉 2 + 𝑔𝑧 + |𝑚|̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]in [ ( 𝜌 2 )]out
𝑝 1
𝑄̇ net + 𝑊̇ shaft, net = |𝑚|
̇ Δ𝑖 + Δ + Δ ( 𝑉 2 ) + Δ(𝑔𝑧)
[ 𝜌 2 ]
(2/19)

This equation is very useful in principle, but not so much in practice, for two
reasons:

1. It contains a lot of terms. There are five fluid properties at inlet and
outlet which affect energy, and it is difficult to predict which one will

water pump with known powers 𝑄̇ in and 𝑊̇ shaft, in . An efficient pump

be affected by a heat or work transfer. For example, consider a simple

will generate large increases in 𝑝 (or 𝑉 and 𝑧), while an inefficient

pump will generate large increases in 𝑖 and 1/𝜌. The energy balance
equation, in this form, tells us nothing about how energy input to the
control volume is redistributed.

40
 Advice from an expert
Again, remember the title of the chapter. To cal-
culate the value of any one property in equa-
tion 2/18, you need to input the value of the eleven
other ones. It is tempting to take shortcuts while
doing so (“oh, the pressure is probably the same”),
with disastrous consequences. There is no solu-
tion to this. If you are attempting to predict fluid
flow, and are missing information, better stop without a result than
take hazardous attempts at using equation 2/18.

of ordinary fluids is very large, and so 𝑖 is usually hundreds of times

2. The terms have disproportionate values in practice. The heat capacity

for example, an increase of temperature of 0,1 °C (with the term Δ𝑖)

larger than the four terms in the brackets of equation 2/18. In water

requires the same energy as increasing its velocity from 30 km h−1 to

110 km h−1 (with the term Δ𝑒𝑘 ). This is not an issue in thermodynamics,
where heat, work and temperature are the most important parameters.
But in fluid mechanics, velocity is of great interest, and the energy
balance is not always useful to predict its changes.

Advice from an expert

In fluid dynamics, fluid movement usually in-
volves relatively small amounts of energy. You
can convince yourself of this by dropping milk

potential energy as Δ(𝑔𝑧) are converted into

into a bowl of water: minuscule amounts of

ties, before slowly dissipating into internal energy as Δ𝑖.

an incredibly complex distribution of veloci-

changes by measuring temperature differences (and thus Δ𝑖), in

While in principle, we could calculate pressure drops or velocity

practice this only works when very high powers are involved, such
as in a compressor or in a rocket engine nozzle. For ordinary flow
(say, air flow around a car, or water flow in a pipe), the temperature
changes are much too small to be measured. See exercise 2.3 p. 46
for an example of this.

2.6 The Bernoulli equation

2.6.1 Theory
The Bernoulli equation is the energy equation applied to specific cases.
To derive the Bernoulli equation, we will start from equation 2/18 and add
five constraints:

d(𝑚𝑒)CV
1. Steady flow.
d𝑡
(We had already implemented this restriction, when we set from
eq. 2/11 to zero in order to obtain eq. 2/18.)
41
 Thus, 𝜌 stays constant;
2. Incompressible flow.

Thus, both 𝑄̇ net and 𝑊̇ shaft, net are zero;

3. No heat or work transfer.

Thus, the fluid internal energy 𝑖 cannot increase;

4. No friction.

5. One-dimensional flow.
Thus, our considered volume has only one inlet (labeled 1) and one
outlet (labeled 2): all fluid particles move together with the same transit
time, and the overall trajectory is already known.

With these five restrictions, equation 2/18 simply becomes:

𝑝 1
0+0= 𝑚̇ 𝑖cst. + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌cst. 2 )]1
𝑝 1
+ 𝑚̇ 𝑖cst. + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌cst. 2 )]2

̇ and canceling 𝑖cst. , as follows,

Dividing by |𝑚|

𝑝 1 𝑝 1
0 = − 𝑖cst. + + 𝑉 2 + 𝑔𝑧 + 𝑖cst. + + 𝑉 2 + 𝑔𝑧
( 𝜌cst. 2 )1 ( 𝜌cst. 2 )2
𝑝 1 𝑝 1
0=− + 𝑉 2 + 𝑔𝑧 + + 𝑉 2 + 𝑔𝑧
( 𝜌cst. 2 )1 ( 𝜌cst. 2 )2

and multiplying by the (constant and uniform) density 𝜌, we obtain the

Bernoulli equation, with all terms having dimensions of pressure:
1 2 1 2
(𝑝 + 𝜌𝑉 + 𝜌𝑔𝑧 =
)1 (𝑝 + 𝜌𝑉 + 𝜌𝑔𝑧 )
2 2 2
(2/20)

This equation describes the properties of a fluid particle in a steady, incom-

pressible, friction-less flow with no energy transfer.

2.6.2 Reality
Let us insist on the incredibly frustrating restrictions brought by the five
conditions above:

1. Steady flow.
Video: The Bernoulli equation This constrains us to continuous flows with no transition effects, which
will kill you
by Olivier Cleynen (CC-by)
is a reasonable limit;
https://youtu.be/4RBxVepjcQc
2. Incompressible flow.

diffusers, nozzles, nor in flows where 𝑀 > 0,6.

We cannot use this equation to describe flow in compressors, turbines,

3. No heat or work transfer.

We cannot use this equation in a machine (e.g. in pumps, turbines,
combustion chambers, coolers).

4. No friction.
This is a tragic restriction! We cannot use this equation to describe a
42 turbulent or viscous flow, e.g. near a wall or in a wake.
 5. One-dimensional flow.
This equation is only valid if we know precisely the trajectory of the
fluid whose properties are being calculated.

Of course, we can overcome those shortcomings by adding one extra (nega-

tive) term called “Δ𝑝loss ” to eq. 2/20, which lumps together all of the effects
unaccounted for. In this way, we obtain the Bernoulli equation with losses:
1 2 1 2
(𝑝 + 2 𝜌𝑉 + 𝜌𝑔𝑧 )1 = (𝑝 + 2 𝜌𝑉 + 𝜌𝑔𝑧 )2 + Δ𝑝loss (2/21)

There are indeed cases where the pressure losses due to the imperfection
of the flow are well-understood, and can be easily quantified. This is true
of flow in pipes, for example (we study those in chapter 7). In those cases,
eq. 2/21 is extremely useful.
Nevertheless, this approach is also easily misused. In a fluid flow where
several of the restrictions above do not hold —and many such flows can be
found in everyday life as well as engineering applications— equation 2/21
will betray its users. Convince yourself that any wrong equation can be

2 + 3 = −18 + Δ𝑝loss .
made correct by adding an unknown “bucket” term at the end: for example

Advice from an expert

In case you are not sure whether the Bernoulli equa-
tion applies, start from an energy balance equation.
Crossing out the terms that do not apply will force
you to question their importance (e.g. is heat transfer
really negligible? etc.). If you do not come to a conclu-
sive end, do not remove terms that are inconvenient.
The unfortunate reality is that in fluid mechanics, the
energy balance equation contains many terms, with
disproportionate values, and using it alone is not enough to solve most
practical problems.

Advice from an expert

Among the five restrictions listed, the last is the most
severe, and the most often forgotten: the Bernoulli
equation does not allow us to predict the trajectory of
fluid particles. Just like all of the other equations
in this chapter, it requires a control volume with a
known inlet and a known outlet. If you find yourself
drawing out flow streamlines and interpreting the
result with the Bernoulli equation, you are running
astray. The tools you need to do this correctly are waiting for us in
chapter 6 (Prediction of fluid flows).

43
 2.7 Solved problems

Flow in a nozzle

from 2 m2 to 1 m2 . The flow is so smooth that energy dissipation due to

Water is flowing through a nozzle, where the diameter decreases gently

The water enters the nozzle with a uniform velocity of 3 m s−1 .

wall friction is negligible.

What is the mass flow? What is the outlet velocity? And what is the
pressure change across the pipe?

See this solution worked out step by step on YouTube

https://youtu.be/ZqvZTQu8SgA (CC-by Olivier Cleynen)

Flow through a valve

mass flow entering the pipe is 2 kg s−1 , and it enters the pipe with a
Water is flowing through a straight pipe with constant diameter. The

uniform velocity of 2 m s−1 .

pressure drop: Δ𝑝valve = −3,5 kPa.

In the middle of the pipe length, a valve is installed, which causes the

What is the outlet velocity? What is the net force on the fluid as it
transits? What is the power dissipated as friction?

See this solution worked out step by step on YouTube

https://youtu.be/OEfMpXtkCQM (CC-by Olivier Cleynen)

44
 Problem sheet 2: Analysis of existing flows
with one dimension
last edited April 15, 2021
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

Balance of mass in a fixed control volume with steady flow:

0 = Σ [𝜌𝑉⟂ 𝐴]incoming + Σ [𝜌𝑉⟂ 𝐴]outgoing (2/3)

where 𝑉⟂ is negative inwards, positive outwards.

Balance of momentum in a fixed control volume with steady flow:

𝐹⃗net on fluid = Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ] + Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ] (2/7)

incoming outgoing

where 𝑉⟂ is negative inwards, positive outwards.

Balance of energy in a fixed control volume with steady flow:

𝑝 1
𝑄̇ net + 𝑊̇ shaft, net = Σ 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]in
𝑝 1
+Σ 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]out
(2/18)

where 𝑚̇ is negative inwards, positive outwards.

2.1 Reading quiz

Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.

2.2 Pipe expansion without losses

is 8 cm, the water arrives with a uniform velocity of 1,5 m s−1 . The diameter increases
Water flows from left to right in a pipe, as shown in fig. 2.3. On the left, the diameter

gently until it reaches 16 cm; the expansion is smooth, so that losses (specifically, energy
losses due to wall friction and flow separation) are negligible.

2.2.1. What are the mass and volume flows at inlet and outlet?

2.2.2. What is the average velocity of the water at the right end of the expansion?
45
 Figure 2.3: A simple pipe expansion, with water flowing from left to right.
CC-0 Olivier Cleynen

2.2.3. What is the pressure change in the water across the expansion?

The volume flow of water in the pipe is now doubled.

2.2.4. What is the new pressure change?

The water is drained from the pipe, and instead, air with density 1,225 kg m−3 is flowed
in the pipe, incoming with a uniform velocity of 1,5 m s−1 .

2.2.5. What is the new pressure change?

2.3 Pipe flow with losses

of the pipe length (fig. 2.4). Water arrives the pipe with a uniform velocity of 1,5 m s−1
Water flows in a long pipe which has constant diameter; a valve is installed in the middle

and the pipe diameter is 250 mm.

using the dimensionless loss coefficient 𝐾valve (we later will later encounter it as eq. 7/6
The pipe itself and the valve, together, induce a pressure loss which can be quantified

p. 137). With this tool, the pressure loss Δ𝑝valve is related to the mean incoming speed
𝑉incoming as:
|Δ𝑝valve |
𝐾valve ≡ = 2,6
1
𝜌𝑉incoming
2
(2/22)
2

2.3.1. What is the outlet velocity of the water?

(note: this is a classical “trick” question! :-)

2.3.2. What is the drop in water pressure across the valve?

2.3.3. What is the power dissipated into internal energy?

2.3.4. If the heat losses of the pipe and valve are negligible, what is the temperature
increase of the water?

Figure 2.4: A simple, straight pipe, featuring a partially-open valve in the center
CC-0 Olivier Cleynen
46
2.4 Combustor from a jet engine
A jet engine is equipped with several combustors (sometimes also called combustion
chambers). We are interested in fluid flow through one such combustor, shown in fig. 2.5.
Air from the compressor enters the combustor, is mixed with fuel, and combustion occurs,
which greatly increases the temperature and specific volume of the mix, before it is run
through the turbine.
The conditions at inlet are as follows:

• Air mass flow: 0,5 kg s−1 ;

• Air properties: 25 bar, 1 050 °C, 12 m s−1

• Fuel mass flow: 5 g s−1 .

At the outlet, the hot gases have pressure 24,5 bar and temperature 1 550 °C, and exit
with a speed of 50 m s−1 .

properties: 𝑐v = 718 J kg−1 K−1 , 𝑅air = 287 J kg−1 K−1 .

We assume that the incoming air and outgoing gas have the same thermodynamic

2.4.1. What are the mass flows at inlet and outlet?

2.4.2. What are the volume flows at inlet and at outlet?

2.4.3. What is the power provided to the flow as heat?

2.4.4. What is the net force exerted on the gas as it travels through the combustor?

Figure 2.5: A combustor in a sectioned jet engine (here, a Turboméca Adour). Air enters from the
left, out of the compressor (whose blades are painted blue). It leaves the combustor on the right
side, into the turbine. In the combustor, high-temperature, steady combustion takes place.
Photo CC-by-sa Olivier Cleynen
47
 2.5 Water jet on a truck
A water nozzle shoots water towards the back of a small stationary van. It has a 3 cm2
cross-sectional area, and the water speed at the nozzle outlet is 𝑉jet = 20 m s−1 . As the
horizontal water jet hits the back of the van, it is split in two symmetrical vertical

(𝑉2 = 𝑉3 = 20 m s−1 ).
flows (fig. 2.6). The two opposite vertical jets have same mass flow and same velocity

2.5.1. What is the net force exerted on the water by the truck?

2.5.2. What is the net force exerted on the truck by the water?

𝑉truck = 15 m s−1 .
Now, the truck moves longitudinally in the same direction as the water jet, with a speed

(This is a crude conceptual setup, which allows us to approach conceptually the case
where water acts on the blades of a turbine.)

2.5.3. What is the new force exerted by the water on the truck?

2.5.4. What is the mechanical power transmitted to the truck?

2.5.5. How would the power be modified if the volume flow was kept constant, but
the diameter of the nozzle was reduced? (briefly justify your answer, e.g. in 30
words or less)

Figure 2.6: A water jet flowing out of a nozzle (left), and impacting the vertical back surface of a
small electric truck, on the right.
Figure CC-0 Olivier Cleynen
48
2.6 High-speed gas flow
Scientists build a very high-speed wind tunnel. For this, they build a large compressed air
tank. Air escapes from the tank into a pipe which decreasing cross-section, as shown in
fig. 2.7. The pipe diameter reaches a minimum (at the tunnel throat), and then it expands
again, before discharging into the atmosphere.

Figure 2.7: A converging-diverging nozzle. Air flows from the left tank to the right outlet, with a
contraction in the middle.
Figure CC-0 Olivier Cleynen

For simplicity, we assume that heat losses through the tunnel walls are negligible, and

In the tank (point 1), the air is stationary, with pressure 7,8 bar and temperature 246,6 °C.
that the fluid has uniformly-distributed velocity in cross-sections of the pipe.

At the throat (point 2), the pressure and temperature have dropped to 4,2 bar and 160 °C.
The velocity has reached 417,2 m s−1 . The throat cross-section is 0,01 m2 .

2.6.1. What is the mass flow through the tunnel?

2.6.2. What is the kinetic energy per unit mass of the air at the throat?

the air has seen its pressure and temperature drop to 1,38 bar and 43 °C.
Downstream of the throat, the pressure keeps dropping. By the time it reaches a point 3,

(if you need to convince yourself that 𝐴3 > 𝐴1 , you may also calculate the
2.6.3. What is the fluid velocity at point 3?

cross-section area)
2.6.4. What is the net force exerted on the fluid between the points 2 and 3?
2.6.5. What is the kinetic energy per unit mass of the air at point 3?

Once it has passed point 3, the air undergoes complex loss-inducing evolutions (including

discharges into the atmosphere (point 4) with pressure 1 bar and temperature 165 °C.
going through a shock wave, where its properties change very suddenly), before it

2.6.6. What is the fluid velocity at outlet?

2.6.7. What is the outlet cross-section area?
2.6.8. What is the net force exerted on the fluid between section 3 and the outlet?
2.6.9. What is the kinetic energy per unit mass of the air at the outlet? 49
 Answers
2.2 p. 45

2.2.1 At both inlet and outlet, 𝑚̇ = 7,53 kg s−1 and ̇ = 7,53 L s−1
2.2.2 𝑉2 = 0,375 m s−1
2.2.3 Δ𝑝1→2 = +1 054 Pa
2.2.4 Δ𝑝3→4 = +4 218 Pa
2.2.5 Δ𝑝5→6 = +1,29 Pa

2.3 p. 46

2.3.1 𝑉2 = 𝑉1 = 1,5 m s−1 by application of the mass balance equation; although

2.3.2 With eq. 2/22, Δ𝑝valve = −2 925 Pa

a mis-application of the energy equation would suggest otherwise

̇ = +215,37 W
2.3.3 𝑚Δ𝑖
2.3.4 With eq. 2/18, Δ𝑇 = +0,7 mK (very small!)

2.4 p. 47

2.4.1 |𝑚̇ 1 | = 0,005 + 0,5 kg s−1 and |𝑚̇ 2 | = 0,505 kg s−1

2.4.2 ̇ 1 = 0,0759 m3 s−1 & ̇ 2 = 0,1078 m3 s−1 (there is no volume balance equa-

2.4.3 𝑄̇ = +261 kW (using 𝑉2 = 50 m s−1 )

tion!)

2.4.4 𝐹net = +19,25 N (in flow-wise direction)

2.5 p. 48

2.5.1 𝐹net on water = −120 N

2.5.2 𝐹⃗water/truck = −𝐹⃗net on water
2.5.3 𝐹net on water = −7,5 N
2.5.4 𝑊̇ truck = 112,5 W

2.6 p. 49

2.6.1 With eq. 2/18, 𝑉2 = 417,2 m s−1 , and so 𝑚̇ 2 = 𝑚̇ = 14,1 kg s−1

2.6.2 𝑒𝑘2 = 87,03 kJ kg−1
2.6.3 With eq. 2/18, 𝑉3 = 638,71 m s−1 (you may then calculate 𝜌3 and obtain
𝐴3 > 𝐴2 even though 𝑉3 > 𝑉2 , a classical feature of supersonic flows)
2.6.4 𝐹net2→3 = +3,137 kN
2.6.5 𝑒𝑘3 = 204,61 kJ kg−1
2.6.6 𝑉4 = 405 m s−1
2.6.7 𝜌4 = 0,7952 kg s−1 and so 𝐴4 = 0,0438 m2
2.6.8 𝐹net3→4 = −3,309 kN (so, against the flow direction)
2.6.9 𝑒𝑘4 = 82,01 kJ kg−1
50
 Fluid Dynamics
Chapter 3 – Analysis of existing flows
with three dimensions
last edited May 22, 2020
by Olivier Cleynen — https://fluidmech.ninja/

3.1 Motivation 51
3.2 The Reynolds transport theorem 51
3.2.1 Control volume 51
3.2.2 Rate of change of an additive property 52
3.3 Balance of mass 54
3.4 Balance of momentum 56
3.5 Balance of angular momentum 57
3.6 Balance of energy 59
3.7 Limits of integral analysis 59
3.8 Solved problems 60
3.9 Problems 63

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

3.1 Motivation
In this chapter, we use the same tools that we developed in chapter 2, but we
improve them so we can apply them to more complex cases. Specifically, we
would like to answer the following questions:

1. What are the mass flows and forces involved when a flow has non-
uniform velocity?

2. What are the forces and moments involved when a flow changes direc-
tion?

3.2 The Reynolds transport theorem

3.2.1 Control volume

Let us begin, this time, by building a control volume in any arbitrary flow: we

we will write equations that work inside any generic velocity field 𝑉⃗ =
are no longer limited to one-inlet, one-oulet steady-flow situations. Instead,

(𝑢, 𝑣, 𝑤) which is a function of space and time: 𝑉⃗ = 𝑓 (𝑥, 𝑦, 𝑧, 𝑡).

Within this flow, we draw an arbitrary volume named control volume (CV)
which is free to move and change shape (fig. 3.1). We are going to measure
the properties of the fluid at the borders of this volume, which we call the
control surface (CS), in order to compute the net effect of the flow through
the volume.
At a given time, the control volume contains a certain amount of mass which
we call the system (sys). Thus the system is a fixed amount of mass transiting
51
 Figure 3.1: A control volume within an arbitrary flow (compare with figure 2.1 p. 34).
The system is the amount of mass included within the control volume at a given
time. At a later time (bottom), it may have left the control volume, and its shape and
properties may have changed. The control volume may also change shape with time,
although this is not represented here.
Figure CC-0 Olivier Cleynen

through the control volume at the exact time when we write the equation,
and its properties (volume, pressure, velocity etc.) may change in the process.
All along the chapter, we are focusing on the question: based on measured
fluid properties at some chosen area in space and time (the properties at the
control surface), how can we quantify what is happening to the system (the
mass inside the control volume)?

3.2.2 Rate of change of an additive property

methodology even more robust. We start with a “dummy” fluid property 𝐵,

In order to proceed with our calculations, we need to make our accounting

Let us therefore consider an arbitrary additive property 𝐵 of the fluid. By the

which we will later replace with physical variables of interest.

term additive property, we mean that the total amount of property is divided
if the fluid is divided. For instance, this works for properties such as mass,

The specific (i.e. per unit mass) value of 𝐵 is designated 𝑏 ≡ 𝐵/𝑚.

volume, energy, entropy, but not pressure or temperature.

We now want to compute the variation of a system’s property 𝐵 based on

measurements made at the borders of the control volume. We will achieve
this with an equation containing three terms:

• The time variation of the quantity 𝐵 within the system is measured

with the term d𝐵sys / d𝑡.
We will use this term to represent, for example, the rate of change of
the fluid’s energy or momentum as it travels through a jet engine.
52
 • Within the control volume, the enclosed quantity 𝐵CV can vary by

with compressed air): we measure this with the term d𝐵CV / d𝑡.
accumulation (for example, mass may be increasing in an air tank fed

control volume, carrying with it some amount of 𝐵 every second: we

• Finally, a mass flux may be flowing through the boundaries of the

write that net flow out of the system as 𝐵̇ net ≡ 𝐵̇ out − 𝐵̇ in .

We can now link these three terms with the simple equation:
d𝐵sys d𝐵CV
= + 𝐵̇ net
d𝑡 d𝑡
(3/1)

the net flow of 𝐵

= +
the rate of change the rate of change
of the fluid’s 𝐵 of 𝐵 inside at the borders
as it transits considered volume of the considered volume

Since 𝐵 may not be uniformly distributed within the control volume, we

like to express the term 𝐵CV as the integral of the volume density 𝐵/ with
respect to volume :
𝐵
𝐵CV = ∭ d = ∭ 𝜌𝑏 d

(3/2)
d𝐵CV d
= 𝜌𝑏 d
d𝑡 ∭CV
CV CV

d𝑡
(3/3)

and  is volume (m3 ).

where CV is the control volume,

The second term of eq. 3/1, 𝐵̇ net , can be evaluated by quantifying, for each
area element d𝐴 of the control volume’s surface, the surface flow rate 𝜌𝑏𝑉⟂
of property 𝐵 that flows through it, as shown in fig. 3.2. The integral over
the entire control volume surface CS of this term is:

𝐵̇ net = ∬ 𝜌𝑏𝑉⟂ d𝐴 = ∬ 𝜌𝑏 (𝑉⃗rel ⋅ 𝑛)

⃗ d𝐴 (3/4)
CS CS

𝑛⃗ is a unit vector on each surface element d𝐴 pointing outwards,

where CS is the control surface (enclosing the control volume CV),

𝑉⃗rel is the local velocity of fluid relative to the control surface,

and 𝑉⟂ ≡ 𝑉⃗rel ⋅ 𝑛⃗ is the local cross-surface speed (positive outwards, neg. inwards)

control surface with area d𝐴. The 𝑛⃗ vector defines the orientation of d𝐴 surface,
Figure 3.2: Part of the system may be flowing through an arbitrary piece of the

and by convention is always pointed outwards. The amount of 𝐵 flowing through

this small area per unit time is d𝐴 is 𝜌𝑉⟂ d𝐴𝑏
Figure CC-0 Olivier Cleynen
53
 By inserting equations 3/3 and 3/4 into equation 3/1, we obtain:

d𝐵sys d
= 𝜌𝑏 d + ∬ 𝜌𝑏 (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 d𝑡 ∭CV
(3/5)
CS

general, abstract accounting tool, but as we soon replace 𝐵 by meaningful

Equation 3/5 is named the Reynolds transport theoremw ; it stands now as a

variables, it will prove extremely useful, allowing us to quantify the net effect
of the flow of a system through a volume for which border properties are
known.
In the following sections we are going to use this equation to write out four
key balance equations (see §1.7 p. 20):

• balance of mass;

• balance of linear momentum;

• balance of angular momentum;

• balance of energy.

Advice from an expert

Take some time to observe the (sometimes curious!)
terms of equation 3/5, and learn the associated vocab-
ulary. Just like an accountant in a company would like
to have a very clear method for counting how money
is spent and earned, fluid dynamicists need very good
tools to describe what’s coming in and out of their
flows. When you’ll be hurling three-dimensional vector operations at
swirling flows down this chapter, you’ll be glad you learned about control
surfaces and sign conventions earlier on.

3.3 Balance of mass

What is the balance of mass for fluid flowing through any arbitrary volume?
We answer this question by writing out a mass balance equation in the

We now state that the placeholder variable 𝐵 is mass 𝑚. It follows that

template provided by the Reynolds transport theorem (eq. 3/5).

d𝐵/ d𝑡 becomes d𝑚sys / d𝑡, which by definition is zero (see eq. 1/24 p. 20). Also,
𝑏 ≡ 𝐵/𝑚 = 𝑚/𝑚 = 1 and now the Reynolds transport theorem becomes:

d𝑚sys d
=0= 𝜌 d + ∬ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 d𝑡 ∭CV CS
(3/6)

=0= +
the rate of change the rate of change the net mass flow
of the fluid’s mass of mass inside at the borders
as it transits considered volume of the considered volume

54
This equation 3/6 is often called continuity equation. It allows us to compare
the incoming and outgoing mass flows through the borders of the control
volume.

which the volume flow 𝜌(𝑉⃗rel ⋅ 𝑛)

⃗ is uniform: this is illustrated in figure 3.3.
Sometimes, the control volume has well-defined inlets and outlets through
Video: with sufficient skills (and
lots of practice!), it is possible
In that case equation 3/6 reduces to forms that we have already identified in for a musician to produce an un-
interrupted stream of air into
the previous chapter (see §2.3 p. 35):
d
an instrument while still con-

0= 𝜌 d + ∑ {𝜌𝑉⟂ 𝐴} + ∑ {𝜌𝑉⟂ 𝐴}
tinuing to breathe, a technique

d𝑡 ∭CV
called circular breathing.w Can
(3/7) you identify the different terms

d
of eq. 3/8 as they apply to the

= 𝜌 d + ∑ {𝜌|𝑉⟂ |𝐴} − ∑ {𝜌|𝑉⟂ |𝐴}

out in

d𝑡 ∭CV
saxophonist’s mouth?
by David Hernando Vitores (CC-by-sa)
https://frama.link/vyH-cxCL

d
0= ̇ − ∑ {|𝑚|}
𝑚CV + ∑ {|𝑚|} ̇
out in

d𝑡
(3/8)
out in

inlet and outlet. Here eq. 3/6 translates as 0 = d𝑡d ∭CV 𝜌 d + 𝜌3 |𝑉⟂3 |𝐴3 + 𝜌2 |𝑉⟂2 |𝐴2 −
Figure 3.3: A control volume for which the system’s properties are uniform at each

𝜌1 |𝑉⟂1 |𝐴1 .

As before, in equation 3/7, the term 𝜌𝑉⟂ 𝐴 at each inlet or outlet corresponds
Figure CC-0 Olivier Cleynen

to the local mass flow ±𝑚̇ (negative inwards, positive outwards) through the
boundary.

Advice from an expert

If you are thinking that equation 3/6 is just a very fancy
way to write the “good” mass balance equation we al-
ready wrote in chapter 2 (eq. 2/3 p. 35), you are not
completely wrong. There are two useful improvements,

First, having ∬ 𝜌𝑉⟂ d𝐴 instead of just 𝜌𝑉⟂ 𝐴 allows us

however:

to handle cases where the incoming/outgoing velocities

are not uniform. Outside of fluid mechanics textbooks, very few flows

Second, there is an unsteady term d𝑚CV / d𝑡. It is not used to solve prob-
have a nice smooth uniform outlet!

lems in this course, but one day when you are confronted to a case where
your inlet and outlet mass flows are not equal, it will save your day!
55
 3.4 Balance of momentum
What force is applied to the fluid for it to travel through the control volume?
We answer this question by writing out a mass balance equation in the

We now state that the placeholder variable 𝐵 is momentum 𝑚𝑉⃗ . It follows

template provided by the Reynolds transport theorem (eq. 3/5).

that d𝐵/ d𝑡 becomes d𝑚𝑉⃗sys )/ d𝑡, which by definition is the net force 𝐹⃗net
applying to the system (see eq. 1/25 p. 20). Also, 𝑏 ≡ 𝐵/𝑚 = 𝑚𝑉⃗ /𝑚 = 𝑉⃗ and
Abstruse Goose #338: Newton’s
laws of motion almost didn’t
happen
by an anonymous artist (CC-by-nc) now the Reynolds transport theorem becomes:
https://abstrusegoose.com/338

d(𝑚𝑉⃗sys ) d
= 𝐹⃗net = 𝜌 𝑉⃗ d + ∬ 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 d𝑡 ∭CV CS
(3/9)

= +
the vector sum the rate of change the net flow of momentum
of forces of momentum through the boundaries
on the fluid within the considered volume of the considered volume

which the flow 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)

⃗ is uniform: this is illustrated again in figure 3.4.
Sometimes, the control volume has well-defined inlets and outlets through

In that case equation 3/9 reduces to forms that we have already identified in

{ } { }
the previous chapter (see §2.4 p. 37):
d
𝐹⃗net = 𝜌 𝑉⃗ d + ∑ (𝜌|𝑉⟂ |𝐴)𝑉⃗ − ∑ (𝜌|𝑉⟂ |𝐴)𝑉⃗
d𝑡 ∭CV
Video: Making sense of the 3D
linear momentum balance equa-

{ } { }
(3/10)

d
tion

= ⃗
𝑚𝑉 ) + ∑ |𝑚| ⃗
̇ 𝑉 − ∑ |𝑚| ⃗
̇ 𝑉
by Olivier Cleynen (CC-by) out in

d𝑡 (
https://youtu.be/iDCpqoJJSI4

(3/11)
CV out in

are uniform at each inlet and outlet, eq. 3/9 translates as 𝐹⃗net = d𝑡d ∭CV 𝜌 𝑉⃗ d +
Figure 3.4: The same control volume as in fig. 3.3. Here, since the system’s properties

𝜌3 |𝑉⟂3 |𝐴3 𝑉⃗3 + 𝜌2 |𝑉⟂2 |𝐴2 𝑉⃗2 − 𝜌1 |𝑉⟂1 |𝐴1 𝑉⃗1 .

Figure CC-0 Olivier Cleynen

56
To make clear a few things, let us focus on the simple case where a con-

equation 3/9, the net force 𝐹⃗net applying on the fluid is:
sidered volume has only one inlet (point 1) and one outlet (point 2). From

d
𝐹⃗net = 𝜌 𝑉⃗ d + ∬ 𝜌2 |𝑉⟂2 |𝑉⃗2 d𝐴2 − ∬ 𝜌1 |𝑉⟂1 |𝑉⃗1 d𝐴1
d𝑡 ∭CV
Video: as a person walks, the de-
flection of the air passing around
their body can be used to sustain
(3/12) the flight of a paper airplane (a

In this equation 3/12, what could cause 𝐹⃗net to be non-zero?

walkalong gliderw ). Can you fig-
ure out the momentum flow en-
tering and leaving a control vol-

• The first term, which we could informally write as d(𝑚𝑉⃗ )CV / d𝑡, could
ume surrounding the glider, and
the resulting net force?
by Y:sciencetoymaker (styl)
https://youtu.be/S6JKwzK37_8

volume changes. This may occur if the distribution of velocities 𝑉⃗

be non-zero. This happens when the momentum inside the control

within the control volume is changing, such as when the fluid in a tank
sloshes back and forth against the walls.

̇ 2 𝑉⃗2 − |𝑚|
|𝑚| ̇ 1 𝑉⃗1 , could also be non-zero. This happens when the flux
• The sum of the last two terms, which we could informally write as

of momentum entering the control volume is different from the one

leaving it:

– It may be because the mass flow 𝑚̇ is different at inlet and outlet,

even if the two velocity distributions are the same;
– It may be because the velocities have different length, and the
flow is speeding up or slowing down;
– It may be because the velocities are aligned differently, and the
flow is changing directions;
– It may be because the velocities are non-uniform and distributed
differently, even if their average (and thus the mass flows) are the
same.

As you can see, a lot of different things may be happening at once! We will
study (separately) the most relevant of those effects in the problem section of
this chapter.

3.5 Balance of angular momentum

What moment (“twisting effort”) is applied to the fluid for it to travel through
the control volume? We answer this question by writing an angular momen-
tum balance (see eq. 1/26 p. 20) in the template provided by the Reynolds

We position ourselves at a point 𝑋 , about which we measure all moments.

transport theorem (eq. 3/5).

All positions are measured with a position vector 𝑟⃗X𝑚 . We now state that
Video: rocket landing gone

the placeholder variable 𝐵 is angular momentum 𝑟⃗X𝑚 ∧ 𝑚𝑉⃗ . It follows that

wrong. Can you compute the
moment exerted by the top

d𝐵/ d𝑡 becomes d⃗𝑟X𝑚 ∧ 𝑚𝑉⃗sys / d𝑡, which by definition is the net moment 𝑀⃗ net
thruster around the base of the
rocket as it (unsuccessfully) at-

applying to the system (see again eq. 1/26 p. 20). Also, 𝑏 ≡ 𝐵/𝑚 = 𝑟⃗X𝑚 ∧ 𝑉⃗ and
tempts to compensate for the
collapsed landing leg?
by Y:SciNews (styl)
https://youtu.be/4cvGGxTsQx0

now the Reynolds transport theorem becomes:

57
 d(⃗𝑟X𝑚 ∧ 𝑚𝑉⃗ )sys d
= 𝑀⃗ net,X = 𝑟⃗X𝑚 ∧ 𝜌 𝑉⃗ d + ∬ 𝑟⃗X𝑚 ∧ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ 𝑉⃗ d𝐴
d𝑡 d𝑡 ∭CV CS
(3/13)

= +
the vector sum the rate of change of the net flow of angular momentum
of moments the angular momentum through the boundaries
on the fluid within the considered volume of the considered volume

in which 𝑟⃗X𝑚 is a vector giving the position of any mass 𝑚 relative to point 𝑋 .

which the flow 𝑟⃗X𝑚 ∧ 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)

⃗ is uniform: this is illustrated in figure 3.5.
Sometimes, the control volume has well-defined inlets and outlets through

{ } { }
In that case equation 3/13 reduces to a more readable form:

⃗ d ⃗ ⃗ ⃗
𝑀net,X = 𝑟⃗X𝑚 ∧ 𝜌 𝑉 d + ∑ 𝑟⃗X𝑚 ∧ |𝑚| ̇ 𝑉 − ∑ 𝑟⃗X𝑚 ∧ |𝑚| ̇ 𝑉
d𝑡 ∭
Video: Making sense of the an-
gular momentum balance equa- CV out in
tion
by Olivier Cleynen (CC-by) (3/14)
https://youtu.be/VR8LGr6PuRY

Figure 3.5: A control volume for which the properties of the system are uniform

𝑀⃗ net,X = d𝑡d ∭CV 𝑟⃗X𝑚 ∧ 𝜌 𝑉⃗ d + 𝑟⃗2 ∧ |𝑚̇ 2 |𝑉⃗2 − 𝑟⃗1 ∧ |𝑚̇ 1 |𝑉⃗1 .
at each inlet or outlet. Here the moment about point X in the bottom right is

Figure CC-0 Olivier Cleynen

The same remarks we made for linear momentum earlier apply here: there
are many possible phenomena which may equate to a moment on the fluid.
We will explore this equation rather shyly in the problem section of the
chapter.

Advice from an expert

The angular momentum balance equation is useful in
cases where we attempt to balance a machine using
fluid flows. Moments can be added and subtracted
as vectors, just like forces. It’s an extra layer of ab-
straction to learn (just about every fluid dynamicist
has suffered at first with the clockwise-is-positive
convention). You’ll be glad you learned it when you
attempt to prevent the helicopter from spinning un-
controllably — or when you ask your cfd software to calculate a moment,
and it spits out a vector!
58
3.6 Balance of energy
What power is applied to the fluid for it to travel through the control volume?
We can answer this question by writing an energy balance in the template
provided by the Reynolds transport theorem (eq. 3/5).
The derivation of the equation we look for is the same as the derivation of
equation 2/18 p. 40 in the last chapter. We may therefore directly jump to the
result:

d𝐸sys d
= 𝑄̇ net + 𝑊̇ shaft, net = 𝜌 𝑒𝑓 d + ∬ 𝜌 𝑒𝑓 (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 d𝑡 ∭CV CS
(3/15)

with the term 𝑒𝑓 carrying all of the energy terms relevant for us in fluid

𝑝 1 2
mechanics:

𝑒𝑓 ≡ 𝑖 + + 𝑉 + 𝑔𝑧
𝜌 2
(3/16)

It can be seen here that this equation only differs from equation 2/18 in the
last chapter in the two following ways:

• There is an unsteady term (first term on the right-hand side), account-

ing for the accumulation or depletion of energy within the control
volume;

• The energy flows through the inlet and outlet are expressed as integrals,
allowing us to account for non-uniform distributions.

Save for those two differences, the equation is not any different — and, it
must be admitted, not much more useful in practice. In this course, we will
not be using this equation to solve problems.

3.7 Limits of integral analysis

Integral analysis is an incredibly useful tool in fluid dynamics: in any given
problem, it allows us to rapidly describe and calculate the main fluid phenom-
ena at hand. The net force exerted on the fluid as it is deflected downwards
by a helicopter, for example, can be calculated using just a loosely-drawn
control volume and a single vector equation.
As we progress through the end-of-chapter problems, however, the limits of
this method slowly become apparent. There are two of them:

• First, we are confined to calculating the net effect of fluid flow. The net
force, for example, encompasses the integral effect of all forces —due to
pressure, shear, and gravity— applied on the fluid as it transits through
the control volume. Integral analysis gives us absolutely no way of
distinguishing between those sub-components. In order to do that (for
example, to calculate which part of a pump’s mechanical power is lost
to internal viscous effects), we would need to look within the control
volume. 59
 direction. The value d𝐵sys / d𝑡 of any finite integral cannot be used to
• Second, all four of our equations in this chapter only work in one

find which function 𝜌𝑏𝑉⟂ d𝐴 was integrated over the control surface to

which will result in a net force of +12 N. Knowing the net value of an
obtain it. For example, there are an infinite number of velocity profiles

integral, we cannot deduce the conditions which lead to it.

In practice, this is a major limitation on the use of integral analysis,
because it confines us to working with large swaths of experimental
data gathered at the borders of our control volumes. From the wake
below the helicopter, we deduce the net force; but the net force tells us
nothing about the shape of the wake.

Clearly, in order to overcome these limitations, we are going to need to

open up the control volume, and look at the details of the flow within —
perhaps by dividing it into a myriad of sub-control volumes. This is what
we set ourselves to in chapter 6 (Prediction of fluid flows), with a thundering and
formidable methodology we shall call derivative analysis.

3.8 Solved problems

Flow through pipe bend

A chemical is flowing through a pipe with 90° bend, with a mass flow
of 200 kg s−1 . Its incoming velocity (uniform) is 2 m s−1 , and its outgoing
velocity is 3 m s−1 .

What is the net force exerting on the fluid?

See this solution worked out step by step on YouTube

https://youtu.be/npEbzOx3qHY (CC-by Olivier Cleynen)

60
 Outlet with a non-uniform velocity

is uniform, with 𝑉1 = 20 m s−1 . At the outlet, the velocity is the same on

Water is flowing through a straight rectangular pipe. At inlet, its velocity

average; however, it is not uniformly distributed. We have 𝑉2 = 19 + 𝑦

(m s−1 ), with 𝑦 the vertical coordinate (in meters). The width of the pipe
in the 𝑧-direction is Δ𝑧 = 1 m.

What is the net force exerting on the fluid?

See this solution worked out step by step on YouTube

https://youtu.be/hPRvLVolHtY (CC-by Olivier Cleynen)

Force and power on a moving turbine blade

In a jet engine, a row of blades is moving with great speed (𝑉blade =
442,4 m s−1 ). Each blade receives 1,7 kg s−1 of high-speed, high-temperature,
high-pressure gas. Through both the movement and the shape of the
blade, this gas is deviated and their properties change. For simplicity, we
assume the properties at inlet and outlet are uniform.
The incoming and outgoing velocities are described below two times:
once (left) from the point of view of the stationary engine, and once
(right) from the point of view of the moving blades.

What is the net force exerting on each blade? What is the power trans-
mitted to the blade?
61
 See this solution worked out step by step on YouTube
https://youtu.be/BgUjpaBYeDc (CC-by Olivier Cleynen)

Mass flows in a boundary layer

A flat plate is positioned parallel to a steady, uniform air flow (𝑢1 = 𝑈 =
25 m s−1 ). At the trailing edge of the plate, the velocity 𝑢2 is measured. It is
a function of height 𝑦, with thickness 𝛿 = 2 cm, following the distribution:
𝑦 6
𝑢2 = 𝑈 ( )
1

𝛿
(3/17)

The plate has length 𝐿1 = 50 cm (in 𝑥-direction, along the flow) and
width 𝐿2 = 80 cm (in 𝑧-direction, across the flow). The density of air is
1,225 kg m−3 .

The layer has thickness 𝛿 at outlet. What is the inlet height of a control
volume that has the mass flow of this layer?

See this solution worked out step by step on YouTube

https://youtu.be/2ROV07UcjiI (CC-by Olivier Cleynen)

Shear force in a boundary layer

In the flat plate problem from above, what is the net force exerted on the
air by the plate?

See this solution worked out step by step on YouTube

https://youtu.be/Lu3GRcv3BSA (CC-by Olivier Cleynen)

62
 Problem sheet 3: Analysis of existing flows
with three dimensions
last edited May 15, 2020
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

Mass balance through an arbitrary volume:

d
0= 𝜌 d + ∬ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 ∭CV
(3/6)
CS

Momentum balance through an arbitrary volume:

d
𝐹⃗net = 𝜌 𝑉⃗ d + ∬ 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 ∭CV
(3/9)
CS

Angular momentum balance through an arbitrary volume:

d
𝑀⃗ net,X = 𝑟⃗X𝑚 ∧ 𝜌 𝑉⃗ d + ∬ 𝑟⃗X𝑚 ∧ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ 𝑉⃗ d𝐴
d𝑡 ∭CV
(3/13)
CS

3.1 Reading quiz

Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.
(the quiz for chapter 3 will open from May 10 to May 20)

3.2 Pipe bend

A pipe with diameter 30 mm has a bend with angle 𝜃 = 130°, as shown in fig. 3.6. Wa-
CC-0 Olivier Cleynen

ter enters and leaves the pipe with the same speed 𝑉1 = 𝑉2 = 1,5 m s−1 . The velocity
distribution at both inlet and outlet is uniform.

3.2.1. What is the mass flow traveling through the pipe?

3.2.2. What is the force exerted by the pipe bend on the water?

3.2.3. Represent the force vector qualitatively (i.e. without numerical data).

3.2.4. What would be the new force if all of the speeds were doubled?

63
 Figure 3.6: A pipe bend, through which water is flowing. We assume that the velocity distribution
at inlet and outlet is identical.
Figure CC-0 Olivier Cleynen

3.3 Exhaust gas deflector

A deflector is used behind a stationary aircraft during ground testing of a jet engine
(fig. 3.7).

Figure 3.7: A mobile exhaust gas deflector, used to deflect hot jet engine exhaust gases upwards
during ground tests.
Figure CC-0 Olivier Cleynen

speed is 𝑉jet = 600 km h−1 , temperature 400 °C and the pressure is atmospheric. As the
The deflector is fed with a horizontal air jet with a quasi-uniform velocity profile; the

with a 40° angle relative to the horizontal.

exhaust gases travel through the pipe, their heat losses are negligible. Gases are rejected

The inlet diameter is 1 m and the horizontal outlet surface is 6 m2 .

3.3.1. What is the force exerted on the ground by the deflection of the exhaust gases?
3.3.2. Describe qualitatively (i.e. without numerical data) a modification to the deflec-
tor that would reduce the horizontal component of force.
3.3.3. What would the force be if the deflector traveled rearwards (positive 𝑥-direction)
with a velocity of 10 m s−1 ?

3.4 Pelton water turbine White [22] P3.56

A water turbine is modeled as the following system: a water jet exiting a stationary

can be neglected, and the water jet is deflected entirely with a 180° angle.
nozzle hits a blade which is mounted on a rotor (fig. 3.8). In the ideal case, viscous effects

The nozzle has a cross-section diameter of 5 cm and produces a water jet with a speed
𝑉jet = 15 m s−1 . The rotor diameter is 2 m and the blade height is negligibly small.

stationary (𝑉blade = 0 m s−1 ).

We first study the case in which the rotor is prevented from rotating, so that the blade is
64
Figure 3.8: Schematic drawing of a water turbine blade. This type of turbine is called Pelton
turbine.
Figure CC-0 Olivier Cleynen

3.4.1. What is the force exerted by the water on the blade?

3.4.2. What is the moment exerted by the blade around the rotor axis?
3.4.3. What is the power transmitted to the rotor?

We now let the rotor rotate freely. Friction losses are negligible, and it accelerates until it
reaches maximum velocity.

3.4.4. What is the rotor rotation speed?

3.4.5. What is the power transmitted to the rotor?

The rotor is now coupled to an electrical generator.

for 𝑉blade = 13 𝑉water .

3.4.6. Show that the maximum power that can be transmitted to the generator occurs

3.4.7. What is the maximum power that can be transmitted to the generator?
3.4.8. How would the above result change if viscous effects were taken into account?
(briefly justify your answer, e.g. in 30 words or less)

3.5 Snow plow derived from Gerhart & Gross [7] Ex5.9
A road-based snow plow (fig. 3.9) is clearing up the snow on a flat surface. We wish to

The snow plow is advancing at 25 km h−1 ; its blade has a frontal-view width of 4 m.
quantify the power required for its operation.

The snow on the ground is 30 cm deep and has density 300 kg m−3 .
The snow is pushed along the blade and is rejected horizontally with a 30° angle to the
left of the plow. Its density has then risen to 450 kg m−3 . The cross-section area 𝐴outlet of
the outflowing snow in the 𝑥-𝑦 plane is 1,1 m2 .

3.5.1. What is the force exerted on the blade by the deflection of the snow?
(Indicate its magnitude and coordinates)
3.5.2. What is the power required for the operation of the snow plow?
3.5.3. If the plow velocity was increased by 10 %, what would be the increase in power?
65
 Figure 3.9: Outline schematic of a blade snow plow.
Figure CC-0 Olivier Cleynen

3.6 Inlet of a pipe

Water is circulated inside a cylindrical pipe with diameter 1 m (fig. 3.10).
Based on White [22]

At the entrance of the pipe, the speed is uniform: 𝑢1 = 𝑈av. = 5 m s−1 .

Shear applies on the fluid from the walls, where the velocity is zero. This strains the fluid
particles, and changes the velocity distribution. At the outlet of the pipe, the velocity
profile is no longer uniform. It can be modeled as a function of the radius with the

𝑟 7
relationship:
𝑢2(𝑟) = 𝑈center (1 − )
1

𝑅
(3/18)

What is the center velocity 𝑈center at the outlet?

66
Figure 3.10: Velocity profiles at the inlet and outlet of a circular pipe.
Figure CC-0 Olivier Cleynen

3.7 Drag on a cylindrical profile

In order to measure the drag on a cylindrical profile, a cylindrical tube is positioned
perpendicular to the air flow in a wind tunnel (fig. 3.11), and the longitudinal component
of velocity is measured across the tunnel section.

Figure 3.11: A cylinder profile set up in a wind tunnel, with the air flowing from left to right.
Figure CC-0 Olivier Cleynen

Upstream of the cylinder, the air flow velocity is uniform (𝑢1 = 𝑈 = 30 m s−1 ).
Downstream of the cylinder, the speed is measured across a 2 m height interval. Hori-
zontal speed measurements are gathered and modeled with the following relationship:

𝑢2(𝑦) = 29 + 𝑦 2 (3/19)

The width of the cylinder (perpendicular to the flow) is 2 m. The Mach number is very
low, and the air density remains constant at 𝜌 = 1,23 kg m−3 ; pressure is uniform all along
the measurement field.

3.7.1. What is the drag force applying on the cylinder?

and the function 𝑢2(𝑦) above only modeled time-averaged values of the horizontal
3.7.2. How would this value change if the flow in the cylinder wake was turbulent,

velocity? (briefly justify your answer, e.g. in 30 words or less)

67
 3.8 Drag on a flat plate

In order to achieve this, measurements of the horizontal velocity 𝑢 are made around the
We wish to measure the drag applying on a thin plate positioned parallel to an air stream.

plate (fig. 3.12).

Figure 3.12: Side view of a plate positioned parallel to the flow.

At the leading edge of the plate, the horizontal velocity of the air is uniform: 𝑢1 = 𝑈 =
Figure CC-0 Olivier Cleynen

10 m s−1 .

by the effect of shear. This layer, called boundary layer, has a thickness of 𝛿 = 1 cm. The
At the trailing edge of the plate, we observe that a thin layer of air has been slowed down

horizontal velocity profile can be modeled with the relation:

𝑦 7
𝑢2(𝑦) =𝑈( )
1

𝛿
(3/20)

The width of the plate (perpendicular to the flow) is 30 cm and it has negligible thickness.
The flow is incompressible (𝜌 = 1,23 kg m−3 ) and the pressure is uniform.

3.8.1. What is the drag force applying on the plate?

3.8.2. What is the power required to compensate the drag?

3.8.3. Under which form is the kinetic energy lost by the flow carried away? Can this
new form of energy be measured? (briefly justify your answer, e.g. in 30 words
or less)

68
3.9 Drag measurements in a wind tunnel
A group of students proceeds with speed measurements in a wind tunnel. The objective
is to measure the drag applying on a wing profile positioned across the tunnel test section
(fig. 3.13).

Figure 3.13: Wing profile positioned across a wind tunnel. The horizontal velocity distributions
upstream and downstream of the profile are also shown.
Figure CC-0 Olivier Cleynen

Upstream of the profile, the air flow velocity is uniform (𝑢1 = 𝑈 = 50 m s−1 ).
Downstream of the profile, horizontal velocity measurements are made every 5 cm across
the flow; the following results are obtained:

horizontal speed 𝑢2 (m s−1 )

0 50
vertical position (cm)

5 50
10 49
15 48
20 45
25 41
30 39
35 40
40 43
45 47
50 48
55 50
60 50

The width of the profile (perpendicular to the flow) is 50 cm. The airflow is incompressible
(𝜌 = 1,23 kg m−3 ) and the pressure is uniform across the measurement surface.

3.9.1. What is the drag applying on the profile?

3.9.2. How would the above calculation change if vertical speed measurements were
also taken into account?
69
 3.10 Moment on gas deflector non-examniable

deflector viewed from the side. The midpoint of the inlet is 2 m above and 5 m behind
We revisit the exhaust gas deflector of exercise 3.3 p. 64. Figure 3.14 below shows the

the wheel labeled “A”, while the center axis of the outlet passes 1,72 m away from it, as
represented in fig 3.14

Figure 3.14: Side view of the mobile exhaust gas deflector which was shown in fig. 3.7 p. 64
Figure CC-0 Olivier Cleynen

What is the moment generated by the gas flow about the axis of the wheel labeled “A”?

3.11 Helicopter tail moment non-examinable

In a helicopter, the role of the tail is to counter exactly the moment exerted by the main
rotor about the main rotor axis. This is usually done using a tail rotor which is rotating
around a horizontal axis.
A helicopter, shown in figure 3.15, is designed to use a tail without a rotor, so as to reduce
risks of accidents when landing and taking off. The tail is a long hollow cylindrical tube
with two inlets and one outlet. To simplify calculations, we consider that pressure is
atmospheric at every inlet and outlet.

• inlet A has a cross-section area of 0,2 m2 . It contributes hot exhaust gases of density
0,8 kg m−3 and velocity 12 m s−1 , aligned with the (𝑥) axis of the tail;

• inlet B contributes 25 kg s−1 of atmospheric air incoming at an angle 𝛼 = 130°

relative to the axis of the tail, with a velocity of 3 m s−1 .

with a fixed velocity of 45 m s−1 . The angle 𝜃 at which gases are rejected is controlled by
The mix of exhaust gases and atmospheric air is rejected at the tip of the tail (outlet C)

the flight computer.

3.11.1. What is the rejection angle 𝜃 required so that the tail generates a moment
of +6 kN m around the main rotor (𝑦) axis?
70
Figure 3.15: Top-view of a helicopter using a rotor-less tail.
Figure derived from a figure CC-by-sa by Commons User:FOX 52

3.11.2. Propose and quantify a modification to the tail geometry or operating conditions

𝑥-axis), while still generating the same moment.

that would allow the tail to produce no thrust (that is to say, zero force in the

Remark: this system is commercialized by MD Helicopters as the notar. The use of

exhaust gases was abandoned, however, a clever use of air circulation around the tail pipe
axis contributes to the generated moment; this effect is explored in chapter 11 (Large- and
small-scale flows) (§11.3.4 p. 228).

71
 Answers
1) 𝑚̇ = 1,0603 kg s−1
2) 𝐹⃗net = (−0,5681; −1,2184) N
3.2

4) The force is quadrupled.

1) 𝐹⃗net = (−9,532; +1,479) kN : ||𝐹⃗net || = 9,646 kN (force on ground is opposite: 𝐹⃗fluid on pipe =
−𝐹⃗pipe on fluid );
3.3

3) ||𝐹⃗net 2 || = 8,525 kN.

1) 𝐹net = (−883,6; 0) N;
2) 𝑀net 𝑋 = |𝐹net |𝑅 = 883,6 N m;
3.4

3) 𝑊̇ rotor = 0 W;
4) 𝜔 = 143,2 rpm (𝐹net = 0 N);
5) 𝑊̇ rotor = 0 W again;
6) 𝑊̇ rotor, max = 1,963 kW @ 𝑉blade, optimal = 13 𝑉water jet .

1) 𝑚̇ = 2 500 kg s−1 ; 𝐹net 𝑥 = +10,07 kN, 𝐹net 𝑧 = −12,63 kN (force on blade is opposite);
2) 𝑊̇ = 𝐹⃗net ⋅ 𝑉⃗plow = 𝐹net𝑥 |𝑉1 | = 69,94 kW
3.5

3) 𝑊̇ 2 = 1,13 𝑊̇ (+33 %)

3.6 𝑉center = 1,2245𝑈

3.7 𝐹net𝑥 = 𝜌𝐿 ∫𝑆2 (𝑢22 − 𝑈 𝑢2 ) d𝑦 = −95,78 N.

𝐹net𝑥 = 𝜌𝐿 ∫0 (𝑢(𝑦) − 𝑈 𝑢(𝑦) ) d𝑦 = −7,175 ⋅ 10−2 N : 𝑊̇ drag = 𝑈 |𝐹net𝑥 | = 0,718 W.

𝛿 2
3.8

3.9 𝐹net𝑥 ≈ 𝜌𝐿Σ𝑦 [(𝑢22 − 𝑈 𝑢2 ) δ𝑦 ] = −64,8 N.

3.10 Re-use 𝑚̇ = 67,76 kg s−1 , 𝑉1 = 166,7 m s−1 , 𝑉2 = 33,94 m s−1 from ex. 3.3. With 𝑅2⟂𝑉2 =
1,717 m, plug in numbers in eq. 3/14 p. 58: 𝑀net = +18,64 kN m in 𝑧-direction.

3.11 1) Work eq. 3/13 down to scalar equation (in 𝑦-direction), solve for 𝜃: 𝜃 = 123,1°.

be solved at the same time. 𝑟C can be shortened, the flow in C can be split into
2) There are multiple solutions which allow both moment and force equations to

forward and rearward components, or tilted downwards etc. Reductions in 𝑚̇ B or

𝑉C are also possible, but quantifying them requires solving both equations at once.

72
 Fluid Dynamics
Chapter 4 – Effects of pressure
last edited September 3, 2020
by Olivier Cleynen — https://fluidmech.ninja/

4.1 Motivation 73
4.2 Pressure forces on walls 73
4.2.1 Magnitude of the pressure force 73
4.2.2 Position of the pressure force 74
4.3 Pressure fields in fluids 75
4.3.1 The direction of pressure 75
4.3.2 Pressure on an infinitesimal volume 76
4.4 Special case: pressure in static fluids 79
4.4.1 Forces in static fluids 79
4.4.2 Pressure and depth 80
4.4.3 Buoyancy 82
4.5 Solved problems 83
4.6 Problems 85

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

4.1 Motivation
In fluid mechanics, only three types of forces apply to fluid particles: forces
due to gravity, pressure, and shear. This chapter focuses on pressure (we will
address shear in chapter 5), and should allow us to answer two questions:

• How is the effect of pressure described and quantified?

• What are the pressure forces generated on walls by static fluids?

4.2 Pressure forces on walls

4.2.1 Magnitude of the pressure force

When the pressure 𝑝 exerted is uniform and the wall is flat, the resulting
What is the force with which a fluid pushes against a wall?

force 𝐹 is easily calculated:

𝐹pressure = 𝑝uniform 𝑆flat wall (4/1) Video: when pressure-induced

forces in static fluids matter: 24

When the fluid pressure 𝑝 is not uniform (for example, as depicted on the
hours of heavy tonnage transit
through the Miraflores locks in
Panama. Can you quantify the
force applying on a single lock
right side of the wall in figure 4.1), the situation is more complex: the force door?

of area d𝑆, and the corresponding forces are summed up as:

must be obtained by integration. The surface is split in infinitesimal portions by Y:Pancho507 (styl)
https://youtu.be/LNKtS91jaxw

𝐹pressure = ∫ d𝐹pressure = ∫ 𝑝 d𝑆 (4/2)

𝑆 𝑆

for a flat surface.

73
 Figure 4.1: Pressure distribution on a flat plate. We already studied this situation in
chapter 1, problem 1.3 p. 27.
Figure CC-0 Olivier Cleynen

where the 𝑆-integral denotes an integration over the entire surface.

What is required to calculate the scalar 𝐹 in eq. 4/2 is an expression of 𝑝 as

a function of 𝑆. In a static fluid, this expression will often be easy to find,
as we see later on. Typically, in two dimensions 𝑥 and 𝑦 we re-write d𝑆 as
d𝑆 = d𝑥 d𝑦 and we may then proceed with the calculation starting from

𝐹pressure = ∬ 𝑝(𝑥,𝑦) d𝑥 d𝑦 (4/3)

The above equations work only for a flat surface. When we consider a two-
or three-dimensional object immersed in a fluid with non-uniform pressure,
the integration must be carried out with vectors.

𝐹⃗pressure = ∫ d𝐹⃗ = ∫ 𝑝 𝑛⃗ d𝑆 (4/4)

𝑆 𝑆

where the 𝑆-integral denotes an integration over the entire surface;

and 𝑛⃗ is a unit vector describing, on each infinitesimal surface element d𝑆, the
direction normal to the surface.

4.2.2 Position of the pressure force

We are often interested not only in the magnitude of the pressure force, but
also its position. This position can be evaluated by calculating the magnitude

This moment 𝑀⃗ X , using notation shown in fig. 4.2, is expressed as:

of the moment generated by the pressure forces about any chosen point X.

𝑀⃗ X = ∫ d𝑀⃗ X = ∫ 𝑟⃗X𝐹 ∧ d𝐹⃗ = ∫ 𝑟⃗X𝐹 ∧ 𝑝 𝑛⃗ d𝑆 (4/5)

𝑆 𝑆 𝑆

where 𝑟⃗X𝐹 is a vector expressing the position of each infinitesimal surface relative to
point X.

Much like eq. 4/4 above, this eq. 4/5 is easily implemented in a software
algorithm but not very approachable on paper. In this course however,
we want to study the simple case where the surface is flat, and where the
reference point X is in the same plane as the surface. Equation 4/5 is then a

74
on an arbitrary surface (left: perspective view; right: side view). The vector 𝑛⃗ is a
Figure 4.2: Moment generated about an arbitrary point X by the pressure exerted

convention unit vector everywhere perpendicular to the infinitesimal surface d𝑆

considered.
Figure CC-0 Olivier Cleynen

great deal simpler, and we can calculate the magnitude 𝑀X as:

𝑀X = ∫ d𝑀X = ∫ 𝑟X𝐹 d𝐹 = ∫ 𝑟X𝐹 𝑝 d𝑆 (4/6)

𝑆 𝑆 𝑆

for a flat surface, with X in the plane of the surface.

Once both 𝐹pressure and 𝑀X pressure have been quantified, the distance 𝑅X𝐹 be-
tween point X and the application point of the net pressure force is easily
computed:
𝑀X pressure
𝑅X𝐹 =
𝐹pressure
(4/7)

4.3 Pressure fields in fluids

We approached the concept of pressure in chapter 1 (Basic flow quantities) with

(eq. 1/14), for example a flat plate of area 𝐴:

the notion that it represented force perpendicular to a given flat surface

𝐹⟂
𝑝≡
𝐴
(4/8)

To appreciate the concept of pressure in fluid mechanics, we need to go

beyond this equation.

4.3.1 The direction of pressure

An important concept is that in continuum mechanics, the flat surface is
imaginary. More precisely, a fluid is able to exert pressure not only on solid 75
 surfaces, but also upon and within itself. In this context, we need to rework
eq. 4/8 so that now pressure is defined as perpendicular force per area on an
infinitesimally small surface of fluid:
𝐹⟂
𝑝 ≡ lim
𝐴→0 𝐴
(4/9)

Equation 4/9 may appear unsettling at first sight, because as 𝐴 tends to zero,
𝐹⟂ also tends to zero; nevertheless, in any continuous medium, the ratio of
these two terms tends to a single non-zero value: the local pressure.
This brings us to the second particularity of pressure in fluids: the pressure
on either side of the infinitesimal flat surface is the same regardless of its
Video: Two weird things about orientation. In other words, pressure has no direction: there is only one (scalar)
pressure in fluid dynamics
by Olivier Cleynen (CC-by)
https://youtu.be/0d8gfsKllmU
value for pressure at any one point in space.
Thus, in a fluid, pressure applies not merely on the solid surfaces of its

a scalar property field 𝑝(𝑥,𝑦,𝑧,𝑡) .

container, but also everywhere within itself. We need to think of pressure as

4.3.2 Pressure on an infinitesimal volume

While pressure has no direction, it may not have the same value everywhere
in a fluid, and so the gradient (the rate of change with respect to distance)
of pressure may not be null. For example, in a static water pool, pressure is
uniform in the two horizontal directions, but it increases along with depth.
Instead of a flat plate, let us now consider an infinitesimally small cube within
the fluid (fig. 4.3). Because the cube is placed in a scalar field, the pressure
exerting on each of its six faces may be different. The net effect of pressure

What are those three components? In the 𝑥-direction, the pressure on faces 1
will therefore have three components: one for each pair of opposing faces.

and 4 act upon a surface of area d𝑦 d𝑧:

𝐹net, pressure,𝑥 = d𝑦 d𝑧 [𝑝1 − 𝑝4 ]

Figure 4.3: The pressure on each face of an infinitesimal volume may have a different
These changes are labeled d𝑝|𝑖 in each of the 𝑖 = 𝑥, 𝑦, 𝑧 directions.
value. The net effect of pressure will depend on how the pressure varies in space.

figure CC-0 Olivier Cleynen

76
We express 𝑝4 − 𝑝1 as the derivative of pressure in the 𝑥-direction (𝜕𝑝/𝜕𝑥),
multiplied by the distance d𝑥 which separates points 1 and 4, obtaining:

𝜕𝑝
𝐹net, pressure,𝑥 = d𝑦 d𝑧 − d𝑥
[ 𝜕𝑥 ]
−𝜕𝑝
= d
𝜕𝑥
(4/10)

where d ≡ d𝑥 d𝑦 d𝑧 is the volume of the infinitesimal cube.

Now generalizing eq. 4/10 for the other two directions, we can write:
−𝜕𝑝
𝐹net, pressure,𝑥 = d
𝜕𝑥
−𝜕𝑝
𝐹net, pressure,𝑦 = d
𝜕𝑦
−𝜕𝑝
𝐹net, pressure,𝑧 = d
𝜕𝑧

This is tedious to write, but we recognize a pattern. And indeed, we introduce

the concept of gradient, a mathematical operator, defined as so (see also
Appendix A3 p. 250):
𝜕 𝜕 𝜕
⃗ ≡ 𝑖⃗ + 𝑗⃗ + 𝑘⃗
∇
𝜕𝑥 𝜕𝑦 𝜕𝑧
(4/11) Video: How many dimensions
does it take to describe the effect
of pressure?
by Olivier Cleynen (CC-by)
With this cool new tool, we elegantly re-write the group of equations above: https://youtu.be/wjRWNBOX3OM

𝐹⃗net, pressure = − d ∇
⃗𝑝 (4/12)

Finally, we obtain:

1 ⃗
⃗𝑝
𝐹net, pressure = −∇
d
(4/13)

This last equation reads “the pressure force per unit volume is the opposite
of the pressure gradient”. It shows us that in any fluid and any situation, the
force due to pressure points the opposite way of the pressure gradient. Thus,
if a particle of any kind is “dropped” into a fluid flow, we can quantify in
which direction, and with which magnitude, pressure (a scalar field) is going
to “push” it. This is given by equation 4/13, which quantifies this effect as a
vector field (see figures 4.4 & 4.5).

Advice from an expert

Ponder for a moment what the dimensions of
the physical properties in this equation 4/13
are. If you were to record on a usb stick the

you would have to store one value (in Pa) for

values for pressure in a given fluid domain,

each point in space (𝑥, 𝑦, 𝑧), once for each

time point (𝑡).
If you were to calculate the effect of pressure, you would calculate minus

store three values (the three components, in 𝑥, 𝑦, and 𝑧, each in Pa m−1 )

the gradient of pressure. To record this information, you would need to

for each point in space, once for each time point. This field of vectors
77
 would indicate the amount of force per unit volume with which pressure
is pushing the fluid.
Having a good grasp of the tools we are using here is important, because
things will soon get more complicated when shear is added to the equa-
tion (in chapter 5), ultimately leading to the all-powerful Navier-Stokes
equation in chapter 6.

Figure 4.4: Water flow in a two-dimensional water tank, visualized with a computa-

enters with a velocity of 10 m s−1 in a 10 m-high tunnel, and flows around a “bump” at
tional fluid dynamics (cfd) software package. The flow is from left to right: water

the bottom. On the top, the magnitude of velocity is represented (background color),
with white lines indicating flow direction. On the bottom, pressure is displayed. The
minimum pressure in the flow is zero Pascal.
values for pressure have been arbitrarily adjusted for visual purposes so that the

Figure CC-by by Arjun Neyyathala

78
Figure 4.5: The negative of the pressure gradient field in the two-dimensional flow
from figure 4.4. All the vectors are represented with the same length, but their
magnitude is coded as color. The arrows indicate the local force per unit volume
with which pressure is acting on the fluid.
Figure CC-by by Arjun Neyyathala

4.4 Special case: pressure in static fluids

4.4.1 Forces in static fluids

Fluid statics is the study of fluids at rest, i.e. those whose velocity field 𝑉⃗ is

{
everywhere null and constant:

𝑉⃗ = 0⃗
𝜕 𝑉⃗
= 0⃗
(4/14)
𝜕𝑡

We choose to study this type of problem now, because it makes for a concep-
tually and mathematically simple case with which we can practice calculating
fluid-induced forces.
What are the forces applying on an arbitrary particle in a static fluid?

• The force due to pressure is related to the pressure gradient: we just

quantified this with equation 4/13 above.

• The force due to shear is zero. We will indeed see in chapter 5 (Effects of
shear) that shear efforts can be expressed as a function of viscosity and
velocity. All ordinary fluids are unable to exert shear when they are
static.

• The force due to gravity is easy to quantify: it is the mass 𝑚 of the

⃗
fluid particle multiplied by the gravity vector 𝑔.

In a moving fluid, the sum of these forces would add up to the mass of the
particle times its acceleration. But in a static fluid, the velocity is zero and
never changes. We can thus write:

𝐹⃗net, pressure + 𝐹⃗shear + 𝐹⃗gravity = 0⃗

− d ∇ ⃗ 𝑝 + 0⃗ + 𝑚𝑔⃗ = 0⃗
−∇ ⃗ 𝑝 + 0⃗ + 𝜌 𝑔⃗ = 0⃗ 79
 We can finally rewrite this as:

⃗ 𝑝 = 𝜌 𝑔⃗
∇ (4/15)
in a static fluid.

This is a very useful equation, which states that in a static fluid, the only
parameter affecting pressure is gravity. More precisely, the fluid density
times the gravity vector is equal to the change in space of the pressure.
We will see in chapter 6 (Prediction of fluid flows) that equation 4/15 is the specific
case for a much larger general and powerful equation, the Navier-Stokes
equation. But more on that later!

Advice from an expert

What this equation 4/15 is really saying is that in a
static fluid, “pressure changes only with altitude”. This
sounds trivial, but consider the consequences. For
example, when you swim under an anchored boat, you
can’t “feel” the presence of the boat: only your own
depth matters. Or this: the pressure at the inlet of the
turbines of a gigawatt-class hydraulic dam power station depends on
the depth of the reservoir, but not at all on its overall size. So much
consequence for such a small equation!

4.4.2 Pressure and depth

Very often in studies of static fluids, the 𝑧-axis is oriented vertically, positive
It is now easy to quantify pressure everywhere inside a static fluid.

downwards. With this convention, there is no need for a vector equation to

quantify pressure, and equation 4/15 becomes:
d𝑝
= 𝜌𝑔
d𝑧
(4/16)

in a static fluid, when 𝑧 is oriented positive downwards.

The first consequence we draw from equation 4/16 is that in a static fluid
(e.g. in a glass of water, in a swimming pool, in a calm atmosphere), pressure
depends solely on height. Within a static fluid, at a certain altitude, we will
measure the same pressure regardless of the surroundings (fig. 4.6).

of liquid water is approximately constant: 𝜌water = 1 000 kg m−3 . In a water

How is pressure distributed within static liquid water bodies? The density

reservoir, equation 4/16 becomes:

d𝑝
= 𝜌water 𝑔
( d𝑧 )water
d𝑝
= 1 000 × 9,81 = 9,81 ⋅ 103 Pa m−1 = 9,81 ⋅ 10−2 bar m−1
( d𝑧 )water
(4/17)

80
the environment. Here, as long as the fluid remains static, 𝑝A = 𝑝B = 𝑝C = 𝑝D .
Figure 4.6: Pressure at a given depth (or height) in a static fluid does not depend on

Figure CC-0 Olivier Cleynen

Therefore, in static water, pressure increases by approximately 0,1 bar/m

as depth increases. For example, at a depth of 3 m, the pressure will be
approximately 1,3 bar (which is the atmospheric pressure plus Δ𝑧 × d𝑝/ d𝑧).
In the atmosphere, the situation is more complex, because the density 𝜌air of

once again orienting 𝑧 vertically downwards, we can express the pressure

atmospheric air is not uniform. If we model atmospheric air as a perfect gas,

gradient as:

d𝑝 1𝑔
= 𝜌air 𝑔 = 𝑝
( d𝑧 )atm. 𝑇𝑅
(4/18)

This time, the variation of pressure with respect to distance depends on

for ambient temperature and pressure (1 bar, 15 °C) yields:

pressure itself (and it is proportional to it). A quick numerical investigation

d𝑝 1 9, 81
= 1 ⋅ 105 × ×
( d𝑧 )atm. ambient 288,15 287
= 11,86 Pa m−1 = 1,186 ⋅ 10−4 bar m−1 (4/19)

This rate (approximately 0,1 mbar/m) is almost a thousand times smaller than
that of water (fig. 4.7).
Since the rate of pressure change depends on pressure, it also varies with
altitude, and the calculation of pressure differences in the atmosphere is a
little more complicated than for water.

temperature 𝑇 , the gravitational acceleration 𝑔 and the gas constant 𝑅 are

If we focus on a moderate height change, it may be reasonable to consider that

uniform. In this (admittedly restrictive) case, equation 4/18 can be integrated

as so:
d𝑝 𝑔
= 𝑝
d𝑧 𝑅𝑇cst.
XKCD #2153: effects of high alti-
tude

2
1 𝑔 2

∫ 𝑝 d𝑝 = 𝑅𝑇 ∫ d𝑧
by Randall Munroe (CC-by-nc)
https://xkcd.com/2153

𝑝2 𝑔
1 cst. 1

ln = Δ𝑧
𝑝1 𝑅𝑇cst.
𝑝2 𝑔Δ𝑧
= exp
𝑝1 [ 𝑅𝑇cst. ]
(4/20)

81
 Figure 4.7: Variation of pressure as a function of altitude for water and air at the
surface of a water reservoir. The gradient of pressure with respect to altitude is
almost a thousand times larger in water than in air.
Figure CC-0 Olivier Cleynen

This is only a simplified model: in reality, air temperature varies significantly

approximately −6 K km−1 ). Adapting equation 4/20 for a uniform temperature

within the atmosphere (at moderate altitudes the change with altitude is

gradient (instead of uniform temperature) is the subject of problem 4.8 p. 88.

Advice from an expert

Here, we see how the “𝑝=𝜌𝑔ℎ” equation that many of us

if 𝜌 is not uniform, everything falls apart. To be safe,

learned in our first class of fluid mechanics is dangerous:

tion 4/16. As you integrate 𝑝 with respect to 𝑧, you will

when you study a static fluid, always start from equa-

be forced to consider how 𝜌 falls into the picture. In wa-

ter, 𝜌 is uniform (just a number). But in the atmosphere,
air is “squished” by gravity, and much denser near the ground.

In practice, the atmosphere also features significant lateral pressure gradients

(which are strongly related to the wind) and its internal fluid mechanics are
complex and fascinating. Equation 4/20 is a useful and convenient model, but
refinements must be made if precise results are to be obtained.

4.4.3 Buoyancy
Any solid body immersed within a fluid is subjected to pressure on its walls.
When the pressure is not uniform (for example because the fluid is subjected
to gravity, although this may not be the only cause), then the net force due
to fluid pressure on the body walls will be non-zero.
Video: playing around with an
air pump and a vacuum chamber
When the fluid is purely static, this net pressure force is called buoyancy.
by Y:Roobert33 (styl) Since in this case, the only cause for the pressure gradient is gravity, the
https://youtu.be/ViuQKqUQ1U8
net pressure force is oriented upwards. The buoyancy force is completely
independent from (and may or may not compensate) the object’s weight.
Since it comes from equation 4/15 that the variation of pressure within a
fluid is caused solely by the fluid’s weight, we can see that the force exerted
on an immersed body is equal to the weight of the fluid it replaces (that is to
82 say, the weight of the fluid that would occupy its own volume were it not
there). This relationship is sometimes named Archimedes’ principle. The
force which results from the static pressure gradient applies to all immersed
bodies: a submarine in an ocean, an object in a pressurized container, and
of course, the reader of this document as presently immersed in the earth’s
atmosphere.

Figure 4.8: Immersion in a static fluid results in forces that depend on the body’s vol-
ume. They can evidenced by the removal of the fluid (for example in a depressurized
semi-spherical vessel).
Figure CC-0 Olivier Cleynen

4.5 Solved problems

Pressure at the bottom of a lake

A lake has the dimensions shown in the figure above. What is the pressure
at the bottom of the lake?

See this solution worked out step by step on YouTube

https://youtu.be/QZSGZWCVlRc (CC-by Olivier Cleynen)
83
 Pressure force on a wall

The side wall of a water tank has the dimensions shown in the figure
above. What is the force exerted due to the pressure of the water on the
wall?

See this solution worked out step by step on YouTube

https://youtu.be/gmEtw5lvJsM (CC-by Olivier Cleynen)

Position of pressure force on a wall

In the problem above, at what height above the ground does the force
due to pressure apply?

See this solution worked out step by step on YouTube

https://youtu.be/Ck3tAheuCZI (CC-by Olivier Cleynen)

84
 Problem sheet 4: Effects of pressure
last edited April 25, 2021
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

4.1 Reading quiz

Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.

4.2 Pressure in a static fluid

A small water container whose geometry is described in fig. 4.9 is filled with water. What
is the pressure at the bottom of the container?

Figure 4.9: A small water container.

Figure CC-0 Olivier Cleynen

4.3 Pressure measurement with a U-tube

with water. What is the pressure 𝑝int. in the vessel?

A tube is connected to a pressurized air vessel as shown in fig. 4.10. The U-tube is filled

pressure 𝑝atm. .
Figure 4.10: Working principle of a simple liquid tube manometer. The outlet is at atmospheric
Figure CC-0 Olivier Cleynen

(𝜌mercury = 13 600 kg m−3 ) was used instead of water?

What would be the height difference shown for the same internal pressure if mercury
85
 4.4 Straight water tank door

window is 3 m high, 4 m wide, and is positioned 0,4 m above the bottom of the tank.
A water tank has a window on one of its straight walls, as shown in figure 4.11. The

Figure 4.11: An aquarium tank with a window installed on one of its walls
Figure CC-0 Olivier Cleynen

4.4.1. What is the magnitude of the force applying on the window due to the pressure
exerted by the water?
4.4.2. At which height does this force apply?

4.5 Access door on a water channel wall

An open water channel used in a laboratory is filled with stationary water (fig. 4.12). An
observation window is installed on one of the walls of the channel, to enable observation

The hinge stands 1,5 m below the water surface. The window has a length of 0,9 m and
and measurements. The window is hinged on its bottom face.

a width of 2 m. The walls of the channel are inclined with an angle 𝜃 = 60° relative to
horizontal.

Figure 4.12: A door installed on the wall of a water channel. The water in the canal is perfectly
still.
Figure CC-0 Olivier Cleynen

4.5.1. Represent graphically the pressure of the water and atmosphere on each side of
the window.
4.5.2. What is the magnitude of the moment exerted by the pressure of the water
about the axis of the window hinge?
4.5.3. If the same door was positioned at the same depth, but the angle 𝜃 was decreased,
would the moment be modified? (briefly justify your answer, e.g. in 30 words or
86 less)
4.6 Pressure force on a cylinder
An idealized flow over a cylinder is depicted in figure 4.13. This is a very primitive
flow solution, obtained using a model (the potential flow model, which we mention in
chapter 11) which cannot account for viscous effects or flow separation. Nevertheless, it
provides a good first “ideal flow” situation to compute surface pressure forces in fluid
flows.

Figure 4.13: Idealized flow around a cylinder, as predicted by potential flow theory. The stream-
along the 𝑧 direction.
lines are represented only in a plane crossing the center of the cylinder, but they are identical all

Figure CC-by-sa by Commons User:Kraaiennest & Olivier Cleynen

The cylinder has diameter 10 cm, and it spans 50 cm across the flow (in the 𝑧 direction).

𝑉 = 0 m s−1 and the air pressure is everywhere 𝑝∞ = 1 bar. We would like to calculate the
We start by considering the case where there is no flow: the velocity is everywhere

forces applying represented in figure 4.14, both caused by the air pressure.

of the cylinder, on an area spanning 20° around the leading edge.

Figure 4.14: Left: vertical force on the top half of the cylinder. Right: horizontal force on “nose”

CC-0 Olivier Cleynen

4.6.1. What is the magnitude of 𝐹⃗top ?

4.6.2. What is the magnitude of 𝐹⃗upstream ?

87
 (a couple of hints to help with the algebra: ∫ sin 𝑥 d𝑥 = − cos 𝑥 + 𝑘 and ∫ sin3 𝑥 d𝑥 =
1
3
cos3 𝑥 − cos 𝑥 + 𝑘)
We now consider the case there there is fluid flow: air with density 𝜌atm. = 1,225 kg m−3
is coming in at 𝑉∞ = 50 km h−1 . In this case, the pressure 𝑝𝑠 on the surface of the cylinder

coordinate 𝜃 as:
is no longer uniform (see also problem 11.3 p. 236). It is expressed as a function of the

1
𝑝𝑠 = 𝑝∞ + 𝜌 (𝑉∞2 − 4𝑉∞2 sin2 𝜃 )
2
(4/21)

4.6.3. What is the new magnitude of 𝐹⃗top ?

4.6.4. What is the new magnitude of 𝐹⃗upstream ?

(a couple of hints to help with the algebra: ∫ cos 𝑥 d𝑥 = sin 𝑥 + 𝑘 and ∫ cos 𝑥 sin2 𝑥 d𝑥 =
1
3
sin3 𝑥 + 𝑘)

4.7 Buoyancy of a barge

A barge of very simple geometry is moored in a water reservoir (fig. 4.15).

Figure 4.15: Basic layout of a barge floating in water.

Figure CC-0 Olivier Cleynen

4.7.1. Sketch the distribution of pressure on each of the immersed walls of the barge
(left and right sides, rear , bottom and slanted front).

4.7.2. What is the magnitude of the force resulting from pressure efforts on each of
these walls?

4.7.3. What is the weight of the barge?

4.8 Atmospheric pressure distribution non-examinable

The integration we carried out in with equation 4/20 p. 81 to model the pressure distribu-

and constant (𝑇 = 𝑇cst. ). In practice, this may not always be the case.
tion in the atmosphere was based on the hypothesis that the temperature was uniform

of −7 K km−1 ), how can the pressure distribution be expressed analytically?

4.8.1. If the atmospheric temperature decreases with altitude at a constant rate (e.g.

88
Khalifa tower (800 m above the ground). Inside the tower, the temperature is controlled
A successful fluid dynamics lecturer purchases an apartment at the top of the Burj

everywhere at 18,5 °C. Outside, the ground temperature is 30 °C and it decreases linearly
with altitude (gradient: −7 K km−1 ).

(1 bar) is identical inside and outside of the tower.

A door is opened at the bottom of the tower, so that at zero altitude the air pressure

For the purpose of the exercise, we pretend the tower is entirely hermetic (meaning air is
prevented from flowing in or out of its windows).

4.8.2. What is the pressure difference between each side of the windows in the apart-
ment at the top of the tower?

89
 Answers
4.2 𝑝A = 𝑝atm. + 0,039 bar ≈ 1,039 bar.

1) 𝑝inside = 𝑝atm. + 0,0157 bar ≈ 1,0157 bar;

2) Δ𝑧2 = 1,1765 cm. Is having both of those results right enough to call yourself a
4.3

U-tube star?

1) 𝐹net = 𝜌𝑔𝐿 (𝑍max 𝐿max − 12 𝐿2max ) = 176,6 kN;

2) 𝑀net, bottom hinge = 𝜌𝑔𝐿 ( 12 𝑍max 𝐿2max − 31 𝐿3max ) = 176,6 kN m so the force exerts at 𝑅 =
4.4

𝑀net /𝐹net = 1 m above the bottom hinge.

2) 𝑀net = 7,79 kN m;
3) observe the equation used to calculate 𝑀net to answer this question. If needed,
4.5

ask for help in class!

1) 𝐹top = ∫𝜃=0 𝑅𝐿𝑝𝑠 sin 𝜃 d𝜃 = −5 000 N (downwards);

𝜃=𝜋

2) 𝐹front = +868 N (in downstream direction);

4.6

3) 𝐹top = −4 990 N downwards (a lift force of 10 N);

4) 𝐹front = +869 N (1 N of drag!).

2) 𝐹rear = 0,2453 MN, 𝐹side = 2,1714 MN, 𝐹bottom = 13,734 MN, 𝐹front = 0,3468 MN;
3) 𝐹buoyancy = 13,979 MN (1 425 t).
4.7

1) 𝑝𝑝21 = (1 + 𝑘𝑧
𝑇1 )
𝑔
2 𝑘𝑅

2) Δ𝑝 = +247,4 Pa inwards.
4.8

90
 Fluid Dynamics
Chapter 5 – Effects of shear
last edited September 19, 2020
by Olivier Cleynen — https://fluidmech.ninja/

5.1 Motivation 91
5.2 Shear forces on walls 91
5.2.1 Magnitude of the shear force 91
5.2.2 Direction and position of the shear force 92
5.3 Shear fields in fluids 92
5.3.1 The direction of shear 93
5.3.2 Shear on an infinitesimal volume 93
5.4 Resistance to shear: viscosity 97
5.4.1 Viscosity 97
5.4.2 Kinematic viscosity 98
5.4.3 Turbulent viscosity 99
5.4.4 Non-Newtonian fluids 100
5.4.5 The no-slip condition 100
5.5 Special case: shear in simple laminar flows 101
5.6 Solved problems 103
5.7 Problems 105

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

5.1 Motivation
In fluid mechanics, only three types of forces apply to fluid particles: forces
due to gravity, pressure, and shear. This chapter focuses on shear, and should
allow us to answer two questions:

• How is the effect of shear described and quantified?

• What are the shear forces generated on walls by simple flows?

5.2 Shear forces on walls

5.2.1 Magnitude of the shear force

When the shear 𝜏 exerted is uniform and the wall is flat, the resulting force 𝐹
What is the force which which a fluid shears (i.e. “rubs”) against a wall?

in the direction 𝑖 is easily calculated:

𝐹shear, direction 𝑖 = 𝜏uniform, direction 𝑖 𝑆flat wall (5/1)

When the shear 𝜏 exerted by the fluid is not uniform (for example, because
more friction is occurring on some parts of the surface than on others), the
situation is more complex: the force must be obtained by integration. The

91
 surface is split in infinitesimal portions of area d𝑆, and the corresponding
forces are summed up as:

𝐹shear, direction 𝑖 = ∫ d𝐹shear, direction 𝑖 (5/2)

𝑆

𝐹shear, direction 𝑖 = ∬ 𝜏direction 𝑖 d𝑆 (5/3)

𝑆

where the 𝑆-integral denotes an integration over the entire surface.

for a flat surface,

What is required to calculate the scalar 𝐹 in eq. 5/3 is an expression of 𝜏 as a

function of 𝑆. In a simple laminar flow, this expression will often be relatively
easy to find, as we see later on.

5.2.2 Direction and position of the shear force

The above equations work only for a flat surface, and in a chosen direction 𝑖.
When we consider a two- or three-dimensional object immersed in a fluid
with non-uniform shear, the integration must be carried out with vectors.
We will not attempt this in this course, but the expression is worth writing
out in order to understand how computational fluid dynamics (cfd) software

In a general case, the shear on any infinitesimal surface d𝑆 needs to be

will proceed with the calculation.

expressed as a vector 𝜏⃗𝑛 , where 𝑛 is the direction perpendicular to the surface.

The net force due to shear on the surface is then:

𝐹⃗shear = ∫ 𝜏⃗𝑛 d𝑆 (5/4)

𝑆

Much like equation 4/4 in the previous chapter, eq. 5/4 is not too hard to
implement as a software algorithm to obtain numerically, for example, the
force resulting from shear due to fluid flow around a body such as the body
of a car. Its computation by hand, however, is far too tedious for us to even
attempt.
The position of the shear force is obtained with two moment vector equations,
in a manner similar to that described in §4.2.2 p. 74 with pressure. This is
outside of the scope of this course.

5.3 Shear fields in fluids

We approached the concept of shear in chapter 1 (Basic flow quantities) with the

example a flat plate of area 𝐴:

notion that it represented force parallel to a given flat surface (eq. 1/15), for

𝐹∥
𝜏≡
𝐴
(5/5)

Like we did with pressure, to appreciate the concept of shear in fluid mechan-
ics, we need to go beyond this equation.
92
5.3.1 The direction of shear
Already from the definition in eq. 5/5 we can appreciate that “parallel to a
flat plate” can mean a multitude of different directions, and so that we need
more than one dimension to represent shear. Furthermore, much in the same
way as we did for pressure, we do away with the flat plate and accept that
shear is a field, i.e. it is an effort applying not only upon solid objects but also Video: cloud movements in a
time-lapse video on an interest-
upon and within fluids themselves. We replace eq. 5/5 with a more general ing day are evidence of a highly-
definition:
𝐹⃗∥
strained atmosphere: pilots and
meteorologists refer to this as

𝜏⃗ ≡ lim
wind shear.

𝐴→0 𝐴
by Y:StormsFishingNMore (styl)
(5/6) https://youtu.be/LjWeYPEmCk8

Contrary to pressure, shear is not a scalar, i.e. it can (and often does) take

it as a vector 𝜏⃗ = (𝜏𝑥 , 𝜏𝑦 , 𝜏𝑦 ), and in a fluid, there is a shear vector field:

different values in different directions. At a given point in space we represent

⎛ 𝜏𝑥 ⎞
𝜏⃗(𝑥,𝑦,𝑧,𝑡) ≡ ⎜ 𝜏𝑦 ⎟
⎜ ⎟
⎝ 𝜏𝑧 ⎠(𝑥,𝑦,𝑧,𝑡)
(5/7)

5.3.2 Shear on an infinitesimal volume

Describing the changes in space of the shear vector field requires another
mathematical dimension (called order). Instead of a flat plate, let us consider
an infinitesimally small cube within the fluid, as shown in figure 5.1. Because
the cube is immersed inside a vector field, it may have a different the shear
vector exerting on each of its six faces.
In order to express the efforts on any given face, we express a component of We the Robots #20071126: using
shear with two subscripts: the right terminology is impor-
tant (or not?).
by Chris Harding
http://www.wetherobots.com/2007-11-26-
• The first subscript indicates the direction normal to the surface in large/

which we are interested;

• The second subscript indicates the direction in which the shear is

applying.

faces 1 to 3 represented). The shear tensor 𝜏⃗𝑖𝑗 has six members of three components
Figure 5.1: Shear efforts on a cubic fluid particle (with only the efforts on the visible

each.
Figure CC-0 Olivier Cleynen

93
 For example, 𝜏⃗𝑥𝑦 represents the shear in the 𝑦-direction on a surface perpen-
dicular to the 𝑥-direction. On this face, the shear vector would be:

𝜏⃗𝑥𝑗 = 𝜏⃗𝑥𝑥 + 𝜏⃗𝑥𝑦 + 𝜏⃗𝑥𝑧

= 𝜏𝑥𝑥 𝑖⃗ + 𝜏𝑥𝑦 𝑗⃗ + 𝜏𝑥𝑧 𝑘⃗
(5/8)
Video: the net effect of shear (5/9)
where the subscript 𝑥𝑗 indicates all of the directions (𝑗 = 𝑥, 𝑦, 𝑧) on a face perpendic-
by Olivier Cleynen (CC-by)

ular to the 𝑥-direction.

https://youtu.be/KBPmTCKLFHc

In eq. 5/8, a surprising term appears: 𝜏𝑥𝑥 . It is the shear effort perpendicular
to the surface of interest. How is this possible? The answer is that the faces
of the infinitesimal cube studied here are not solid. They are permeable, and
the local velocity may include have a component through the face of the cube
Video: what is 𝜏𝑥𝑥 , perpendicu- (in fact, this must happen for any flow to occur at all). Therefore, there is no
lar shear? reason for the shear effort, which is three-dimensional, to be aligned along
by Olivier Cleynen (CC-by)
https://youtu.be/_3EYEySeG4g each flat surface. As the fluid travels across any face, it can be sheared and
strained in any arbitrary direction, and therefore, shear can and most often
does have a component (𝜏𝑖𝑖 ) perpendicular to an arbitrary surface inside a
fluid.
Now, the net shear effect on the cube will have eighteen components: one
tree-dimensional vector for each of the six faces. Each of those components

en entity —a tensor— containing six vectors 𝜏⃗1 , 𝜏⃗2 , 𝜏⃗3 . . . 𝜏⃗6 . By convention,
may take a different value. The net shear could perhaps be represented as

however, shear is notated using only three vector components: one for each

field 𝜏⃗𝑖𝑗 :
pair of faces. Shear efforts on a volume are thus represented with a tensor

⎛ 𝜏⃗𝑥𝑗 ⎞ ⎛ 𝜏⃗𝑥𝑗 {1,4} ⎞

𝜏⃗𝑖𝑗 ≡ ⎜ 𝜏⃗𝑦𝑗 ⎟≡⎜ 𝜏⃗𝑦𝑗 {2,5} ⎟
⎜ ⎟ ⎜ ⎟
⎝ 𝜏⃗𝑧𝑗 ⎠ ⎝ 𝜏⃗𝑧𝑗 {3,6} ⎠
⎛ 𝜏𝑥𝑥 𝜏𝑥𝑦 𝜏𝑥𝑧 ⎞
𝜏⃗𝑖𝑗 ≡ ⎜ 𝜏𝑦𝑥 𝜏𝑦𝑦 𝜏𝑦𝑧 ⎟
⎜ ⎟
⎝ 𝜏𝑧𝑥 𝜏𝑧𝑦 𝜏𝑧𝑧 ⎠
(5/10)

In this last equation 5/10, each of the nine components of the tensor acts as the
container for two contributions: one for each of the two faces perpendicular
to the direction expressed in its first subscript.
So much for the shear effort on an element of fluid. What about the net force
due to shear on the fluid element? Not every element counts: part of the
shear will accelerate (change the velocity vector) the particle, while part of it

making a careful selection within the eighteen components of 𝜏⃗𝑖𝑗 . We may

will merely strain (deform) the particle. Quantifying this force thus requires

start with the 𝑥-direction, which consists of the sum of the component of
shear in the 𝑥-direction on each of the six cube faces:

𝐹⃗shear 𝑥 = 𝑆3 𝜏⃗𝑧𝑥 3 − 𝑆6 𝜏⃗𝑧𝑥 6

+𝑆2 𝜏⃗𝑦𝑥 2 − 𝑆5 𝜏⃗𝑦𝑥 5
+𝑆1 𝜏⃗𝑥𝑥 1 − 𝑆4 𝜏⃗𝑥𝑥 4 (5/11)

94
Given that 𝑆3 = 𝑆6 = d𝑥 d𝑦, that 𝑆2 = 𝑆5 = d𝑥 d𝑧 and that 𝑆1 = 𝑆4 = d𝑧 d𝑦, this
is re-written as:

𝐹⃗shear 𝑥 = d𝑥 d𝑦 (⃗
𝜏𝑧𝑥 3 − 𝜏⃗𝑧𝑥 6 )
𝜏𝑦𝑥 2 − 𝜏⃗𝑦𝑥 5 )
+ d𝑥 d𝑧 (⃗
𝜏𝑥𝑥 1 − 𝜏⃗𝑥𝑥 4 )
+ d𝑧 d𝑦 (⃗ (5/12)

In the same way we did with pressure in chapter 4 (§4.4.2 p. 80), we express
each pair of values as a derivative with respect to space multiplied by an
infinitesimal distance:
𝜕 𝜏⃗𝑧𝑥 𝜕 𝜏⃗𝑦𝑥 𝜕 𝜏⃗𝑥𝑥
𝐹⃗shear 𝑥 = d𝑥 d𝑦 d𝑧 + d𝑥 d𝑧 d𝑦 + d𝑧 d𝑦 d𝑥
( 𝜕𝑧 ) ( 𝜕𝑦 ) ( 𝜕𝑥 )
𝜕 𝜏⃗𝑧𝑥 𝜕 𝜏⃗𝑦𝑥 𝜕 𝜏⃗𝑥𝑥
= d + +
( 𝜕𝑧 𝜕𝑦 𝜕𝑥 )
(5/13)

We can see with this equation 5/13 that shear in the 𝑥-direction has three

derivative in space of a shear component which points in the 𝑥-direction.

contributors (one for each pair of cube faces). Each of the contributors is a

⃗⋅ :
written ∇
Now, we introduce the operator divergent (see also Appendix A3 p. 250),

𝜕 ⃗ 𝜕 ⃗ 𝜕 ⃗
Video: the divergent of shear

⃗⋅ ≡
∇ 𝑖⋅ + 𝑗 ⋅ + 𝑘⋅
by Olivier Cleynen (CC-by)

𝜕𝑥 𝜕𝑦 𝜕𝑧
https://youtu.be/PvcT_E55Lgc

(5/14)
𝜕𝐴𝑥 𝜕𝐴𝑦 𝜕𝐴𝑧
⃗ ⋅ 𝐴⃗ ≡
∇ + +
𝜕𝑥 𝜕𝑦 𝜕𝑧
(5/15)

⎛ 𝜕𝑥 + 𝜕𝑦 + 𝜕𝑧 ⎞ ⎛
𝜕𝐴 𝜕𝐴𝑦𝑥 𝜕𝐴
⃗ ⋅ 𝐴⃗𝑖𝑥
∇ ⎞
⎜ ⎟ ⎜ ⎟
𝑥𝑥 𝑧𝑥

⃗ ⋅ 𝐴⃗𝑖𝑗 ≡ ⎜ 𝜕𝐴𝑥𝑦 + 𝜕𝐴𝑦𝑦 + 𝜕𝐴𝑧𝑦 ⎟ = ⎜

∇ ⃗ ⋅ 𝐴⃗𝑖𝑦
∇ ⎟
⎜ 𝜕𝐴𝑥𝑧 + 𝜕𝐴𝑦𝑧 + 𝜕𝐴𝑧𝑧 ⎟ ⎜ ⎟
𝜕𝑥 𝜕𝑦 𝜕𝑧
⃗ ⋅ 𝐴⃗𝑖𝑧
∇
(5/16)
⎝ 𝜕𝑥 𝜕𝑦 𝜕𝑧 ⎠ ⎝ ⎠

force in the 𝑥-direction is equal to the particle volume times the divergent of
With this new tool, we can go back to equation 5/13 to see that the net shear

the shear in the 𝑥-direction:

𝐹⃗shear 𝑥 = d ∇
⃗ ⋅ 𝜏⃗𝑖𝑥 (5/17)

So much for the 𝑥-direction. The 𝑦- and 𝑧-direction are taken care of in the
same fashion, so that we can gather up our puzzle pieces and express the
force per unit volume due to shear as the divergent of the shear tensor:

⎛ 𝐹shear 𝑥 ⎞ ⃗ ⋅ 𝜏⃗𝑖𝑥 |
⎛ |∇ ⎞
𝐹shear = ⎜ 𝐹shear 𝑦
⃗ ⎟ = d ⎜ |∇
⃗ ⋅ 𝜏⃗𝑖𝑦 | ⎟ = d ∇
⃗ ⋅ 𝜏⃗𝑖𝑗
⎜ ⎟ ⎜ ⎟
⎝ 𝐹shear 𝑧 ⎠ ⃗ ⋅ 𝜏⃗𝑖𝑧 |
⎝ |∇ ⎠
(5/18)

1 ⃗
⃗ ⋅ 𝜏⃗𝑖𝑗
𝐹net, shear = ∇
d
(5/19)

An example of a divergent of shear vector field is shown in figures 5.2 and 5.3.

This equation 5/19 is more than we really need to go through the problems
in this chapter, but we will come back to it when we will want to calculate 95
 in figure 4.4 p. 78. Vectors with magnitude lower than 1 Pa m−1 are not represented.
Figure 5.2: The divergent of shear in the flow field of the computed flow described

The arrows indicate the local force per unit volume with which shear is acting on
the fluid.
Figure CC-by by Arjun Neyyathala

Figure 5.3: The magnitude of the divergent of shear in the flow field of the computed

figure 5.2 above). In the top image, the color scale is saturated at 1 ⋅ 105 Pa m−1 , while
flow described in figure 4.4 p. 78 (the magnitude of the vector field represented in

on the bottom image, it reaches a value 20 times higher (2 ⋅ 106 Pa m−1 ), showing the
very high local values attained very close to the wall.
Figures CC-by by Arjun Neyyathala

the dynamics of fluid particles in chapter 6 (Prediction of fluid flows), where the
divergent of shear be a building block of the glorious Navier-Stokes equations.
For now, it is enough to sum up our findings as follows:

• Shear at a point in space has three components — it is a vector field;

• The effect of shear on a volume of fluid has eighteen components – it

is a second-order tensor field;

96
 • The net force due to shear on a volume of fluid, expressed using the
divergent of the shear tensor, has three components — it is a vector
field.

Advice from an expert

Again, take a moment to consider the dimen-
sions involved here. If you were to store in-
formation about shear in a flow in a file on
a usb key, how many columns would you

has three components (each in Pa) at every

need? Shear is a vector field, which means it

to store for each combination of 𝑥,𝑦,𝑧 and 𝑡.

point in space and time: that’s three values

If you were to calculate the effect of shear, then you would calculate its
divergent, which is also a three-dimensional vector field. To record this

⃗ ⋅ 𝜏⃗𝑖𝑥 |, |∇
of |∇ ⃗ ⋅ 𝜏⃗𝑖𝑦 |, and |∇
⃗ ⋅ 𝜏⃗𝑖𝑧 |, all in Pa m−1 ), for each combination of 𝑥,𝑦,𝑧
information, you would, likewise, have to store three values (one for each

and 𝑡. Just like the negative of the gradient of pressure, the divergent
of shear represents a force per unit volume, showing in which direction
shear is “pushing” the fluid particles as they flow.

5.4 Resistance to shear: viscosity

5.4.1 Viscosity
In chapter 1, we saw already that viscosity 𝜇 is a fluid property that quantifies
its resistance to shear (see §1.4.6 p. 17). More precisely, we quantified 𝜇 as
the ratio of shear stress to strain rate with equation 1/12, reproduced here:
𝜏
𝜇≡
( Δ𝑦 )
Δ𝑣
(5/20)

Now, we generalize this equation: we shrink down the “brick” of fluid from

(fig. 5.4). The strain rate is now 𝜕𝑉𝑗 /𝜕𝑖, which is the rate of change in the
chapter 1 down to an infinitesimal volume of fluid inside an arbitrary flow

𝑖-direction of the velocity in the 𝑗-direction. Viscosity is the ratio between

shear in the shear in the 𝑗-direction and this strain rate:
||⃗
𝜏𝑖𝑗 ||
𝜇≡
( 𝜕𝑖 )
𝜕𝑉𝑗 (5/21)

If we turn this equation around, we find that we can express the local shear
by differentiating the local velocity with respect to distance:

𝜕𝑉𝑗
||⃗
𝜏𝑖𝑗 || = 𝜇
𝜕𝑖
(5/22)

in which the subscript 𝑖 is an arbitrary direction (𝑥, 𝑦 or 𝑧) and 𝑗 is the direction

following it in order (e.g. 𝑗 = 𝑧 when 𝑖 = 𝑦);
and where 𝜇 is the viscosity (or “dynamic viscosity”) (Pa s).

97
 Figure 5.4: Any velocity gradient 𝜕𝑉𝑦 /𝜕𝑥 = 𝜕𝑣/𝜕𝑥 in the flow results in a shear force
𝐹∥ in the direction 𝑦. The ratio between the shear and the velocity gradient is called
viscosity.
Figure CC-0 Olivier Cleynen

Viscosity 𝜇 is measured in Pa s, which is the same as N s m−2 or kg m−1 s−1 . It

has historically been measured in poise (1 poise ≡ 0,1 Pa s).
The values of viscosity vary very strongly from one fluid to another: for
example, honey is roughly ten thousand times more viscous than water,
which is roughly a hundred times more viscous than ambient air. The
viscosities of four relevant fluids are quantified in figure 5.6 (where they
Video: viscosity for the engineer can easily be quantified) and figure 5.5 (where the relative values are better
by Olivier Cleynen (CC-by)
https://youtu.be/TR0R0aB_MP8
observed).

1.5×10−2
Viscosity 𝜇 of liquids in Pa s

10−2

5×10−3
Crude Oil

0
Water
Air CO2

−20 0 20 40 60 80 100 120

Temperature 𝑇 in degree Celsius ( C)
◦

Figure 5.5: The viscosity of four fluids (crude oil, water, air, and C02) as a function
of temperature, plotted on a linear scale. This makes clearly visible the difference
in the order of magnitudes of the viscosities of the four fluids, but air and CO2 are
indistinguishably close to the zero axis. A more useful version of this figure is shown
as figure 5.6.
Figure CC-by by Arjun Neyyathala & Olivier Cleynen

98
 2×10−2 2.4×10−5

10−2 2.2×10−5
9×10−3
8×10−3
7×10−3 ⟵ Crude Oil
6×10−3 Air ⟶
5×10−3
2×10−5
4×10−3
3×10−3
Viscosity 𝜇 of liquids in Pa s

Viscosity 𝜇 of gases in Pa s
2×10−3 1.8×10−5
CO2 ⟶

10−3 1.6×10−5
9×10−4
8×10−4
7×10−4
6×10−4
5×10−4
1.4×10−5
4×10−4 ⟵ Water
3×10−4

2×10−4 1.2×10−5

10−4 10−5
−20 0 20 40 60 80 100 120
Temperature 𝑇 in degree Celsius (◦C)

Figure 5.6: The viscosity of four fluids (crude oil, water, air, and C02) as a function of
temperature. The scale for liquids is logarithmic and displayed on the left; the scale
for gases is linear and displayed on the right.
Figure CC-by by Arjun Neyyathala & Olivier Cleynen

5.4.2 Kinematic viscosity

is written 𝜈: the Greek letter nu, an unfortunate choice because it is easy to

Sometimes, the concept of kinematic viscosity is used. Kinematic viscosity

mis-read as the 𝑦-component of velocity, 𝑣 ≡ 𝑉𝑦 . Kinematic viscosity 𝜈 is

𝜇
𝜈≡
defined as
𝜌
(5/23)

where 𝜈 is measured in m2 s−1 .

Kinematic viscosity is formulated that way because it is the part of the

Reynolds number (see eq. 1/28 p. 23) that depends on the fluid properties. We
do not make use of it in this course.

5.4.3 Turbulent viscosity

viscosity 𝜇T (in Pa s just like viscosity). This is because the bulk effect of
In computational fluid dynamics (cfd) simulations, use is made of turbulent

turbulence is to increase dissipation, in a way that is similar to viscous

dissipation. Turbulent viscosity is therefore a property of the flow, with
values that vary very strongly with space and time within the flow (unlike
viscosity which is usually only a property of the fluid). We will encounter it
in chapter 9 (Dealing with turbulence).
99
 5.4.4 Non-Newtonian fluids
Viscosity is said to be a property of the fluid, and not of the flow, because for
most fluids, it has a value that varies only with temperature. Those fluids are
formally called Newtonian fluids. Most fluids of interest in engineering fluid
mechanics (air, water, exhaust gases, pure gases) are Newtonian fluids.
Some fluids, however, feature a strain rate that depends on the value of shear
stress: one could say that they have “variable viscosity”, depending on how
they are strained. Those fluids are called non-Newtonian fluids. The study of
the viscosity characteristics of such fluids is called rheologyw .
Some fluids, like oil-based paint, jelly-based fluids, or tomato juice, are shear-
thinning: they become progressively less viscous when the strain rate is
increased.
Some fluids, like blood, are shear-thickening: they become progressively
thicker when the strain rate is increased. A few representative viscosity
characteristics are displayed in figure 5.7.

Advice from an expert

“Non-Newtonian” does not mean “sticky”; it merely
means that the fluid is “variably sticky”. Some, like
plant resin, resist being strained too quickly, while
others, like paint, resist being strained too slowly. It is
the behavior which matters, and not the actual value
of their viscosities (which may be generally high or
low).
Mix some starch and water in a bowl, and pour some
paint in another: time spent playing with those two fluids in your kitchen
totally counts as serious fluid mechanics study time.

slope of the curve (𝜇) varies with 𝜕𝑉𝑗 /𝜕𝑖 (so, those which do not feature straight lines
Figure 5.7: Various possible viscosity characteristics of fluids. Those for which the
on this diagram) are called non-Newtonian.
100 Figure CC-0 Olivier Cleynen
5.4.5 The no-slip condition
We observe that whenever we measure the velocity of a fluid flow along
a solid wall, the speed tends to zero as we approach the wall surface. In
other words, the fluid adheres to the surface regardless of the overall faraway
flow velocity. This phenomenon, called the no-slip condition, is extremely
important in fluid dynamics. One consequence of this is that fluid flows near
walls are dominated by viscous effects (internal friction) due to the large
strain rates there.

Advice from an expert

The no-slip condition is actually pretty freaky: no mat-
ter how fast you go, and no matter how smooth the
surface, the fluid velocity on the surface of a solid ob-
ject will always be zero! Engineers like to have polished,
low-roughness surfaces on machines like aircraft, be-
cause this reduces the thickness of the shear layer, and
how turbulent it becomes, thus reducing shear — we
will study this in chapter 10 (Flow near walls). But there is no escaping the
fact that fluid particles next to the wall will always feature zero relative
velocity.

5.5 Special case: shear in simple laminar flows

In any ordinary fluid flow, the velocity field is complex, and it is difficult to
express shear and its net effect on particles. Since this requires expressing
three values at each point in (three-dimensional) space and time, this would
require complex mathematics or large amounts of discrete data.
In simple cases, however, it is possible to express and calculate shear relatively
easily. This is especially true in simple, steady, laminar (smooth) flows
—typically flows for which the Reynolds number (eq.1/28 p. 23) is low.
In those cases, we can guess a reasonably realistic velocity distribution, and
then derive an expression for the distribution of shear from it.
One such classical example is the Couette flow, where fluid is imprisoned
between a static flat surface and another flat surface moving parallel to it, as

as well as the spacing 𝐻 , are known.

illustrated in figure 5.8. In this case, the bottom wall and top wall velocities,

Figure 5.8: A simple flow. The bottom wall is stationary, while the top wall slides
from left to right. In between the walls, fluid is strained uniformly.
Figure CC-by-sa Commons User:Kulmalukko

101
 {
A reasonable guess for the velocity distribution in steady laminar regime is:

𝑉𝑥 = 𝑉bottom wall + 𝑘𝑦
𝑉𝑦 = 0

By applying boundary conditions (𝑉bottom wall = 0 and 𝑉𝑥 @ 𝑦=𝐻 = 𝑉top wall ) we

{
can re-write this as:
𝑉𝑥 = 0 + top𝐻wall 𝑦
𝑉

𝑉𝑦 = 0

fluid can be computed. The shear in the 𝑥-direction is proportional to the

And now that the velocity field is known, the shear everywhere in the

derivative in the 𝑦-direction of the velocity in the 𝑥-direction:

d 𝑉top wall
𝜏𝑦𝑥 = 𝜇 0+ 𝑦
d𝑦 ( 𝐻 )
𝜇 𝑉top wall
=
𝐻

(it is independent of 𝑦 and 𝑥). A few slightly more complex cases are waiting
Thus, we see here that the shear applied in the fluid is the same everywhere

for us in the problem sheet; but to handle more realistic shear distributions,
what is needed is a software able to compute the behavior of fluids. The basic
but formidable equations to be solved for this are the topic of the upcoming
chapter 6 (Prediction of fluid flows).

Advice from an expert

Note that this method only works because we
have guessed the velocity field before we calcu-
lated shear. We can only do this when flows are
trivially simple. What if we don’t know what the
velocity is to begin with?
Things will be much harder once we have to eval-

that 𝜏⃗ and 𝑉⃗ depend on one another. Not to say,

uate shear based on a calculated velocity field, so

but it looks like the math in chapter 6 will be off

its kadoova.

102
5.6 Solved problems

Shear force on a flat plate

A flat plate with dimensions 200 mm×400 mm is moved 1 mm above the

ground at 0,5 m s−1 , in a fluid with viscosity 1 ⋅ 10−5 Pa s. What is the
shear force applying on the plate?

See this solution worked out step by step on YouTube

https://youtu.be/aPIz4ASi6cQ (CC-by Olivier Cleynen)

Cylindrical viscometer

A moment of 0,9 N m is required to turn the inner cylinder of a viscometer

(a device designed to measure viscosity) is turned at 150 rpm.
The inner cylinder has diameter 20 cm, the height of the liquid is 80 cm,

drawing) is 2 mm.
and the spacing between the two cylinders (greatly exaggerated on the

What is the viscosity of the fluid?

103
 See this solution worked out step by step on YouTube
https://youtu.be/jc9qIv-jzC4 (CC-by Olivier Cleynen)

Note: Unfortunately Olivier made an error in this video: the height 𝐻1 is

(𝜇 = 2,27 ⋅ 10−2 Pa s). Many thanks to the students who double-checked and
forgotten in the last few lines. The final numerical result is nevertheless correct

reported the problem!

104
 Problem sheet 5: Effects of shear
last edited June 5, 2020
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

Shear force on a flat solid surface:

𝐹shear, direction 𝑖 = ∬ 𝜏direction 𝑖 d𝑆 (5/3)

𝑆

Shear in the direction 𝑗, on a plane perpendicular to direction 𝑖:

𝜕𝑉𝑗
||⃗
𝜏𝑖𝑗 || = 𝜇
𝜕𝑖
(5/22)

2×10−2 2.4×10−5

10−2 2.2×10−5
9×10−3
8×10−3
7×10−3 ⟵ Crude Oil
6×10−3 Air ⟶
5×10−3
2×10−5
4×10−3
3×10−3
Viscosity 𝜇 of liquids in Pa s

Viscosity 𝜇 of gases in Pa s
2×10−3 1.8×10−5
CO2 ⟶

10−3 1.6×10−5
9×10−4
8×10−4
7×10−4
6×10−4
5×10−4
1.4×10−5
4×10−4 ⟵ Water
3×10 −4

2×10−4 1.2×10−5

10−4 10−5
−20 0 20 40 60 80 100 120
Temperature 𝑇 in degree Celsius ( C)
◦

Figure 5.9: The viscosity of four fluids (crude oil, water, air, and C02) as a function of temperature.
The scale for liquids is logarithmic and displayed on the left; the scale for gases is linear and
displayed on the right. This is a reproduction of figure 5.6 p. 99.
Figure CC-by by Arjun Neyyathala & Olivier Cleynen

105
 5.1 Quiz
Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.

5.2 Flow in between two plates Munson & al. [29] Ex1.5

flow is laminar (smooth and fully steady), with a univorm velocity profile 𝑢 = 𝑓 (𝑦) which
A fluid is forced to flow between two stationary plates (fig. 5.10). We observe that the

is linked to the average fluid velocity 𝑉average by the relationship:

3 𝑦 2
𝑢 = 𝑉average 1 − ( )
2 [ 𝐻 ]
(5/24)

where 𝑦 is measured from the middle of the gap;

and 𝐻 is half of the gap length.

Figure 5.10: Velocity distribution for laminar flow in between two plates, also known as Couette
flow.
Figure CC-0 Olivier Cleynen

The fluid is water at 40 °C, the average velocity is 0,6 m s−1 and the two plates are 1,5 mm
apart.

5.2.1. What is the shear effort 𝜏𝑦𝑥 plate generated on the lower plate?

5.2.2. What is the shear effort 𝜏𝑦𝑥 in the middle plane of the flow?

106
5.3 Friction on a plate
A plate the size of an A4 sheet of paper (210 mm × 297 mm) is moved horizontally at
constant speed above a large flat surface (fig. 5.11). We assume that the velocity profile of
the fluid betweeen the plate and the flat surface is entirely uniform, smooth, and steady.

Figure 5.11: A plate moved horizontally across a flat surface.

Figure CC-0 Olivier Cleynen

5.3.1. Express the force 𝐹𝑦𝑧 due to shear on the plate as a function of its velocity 𝑈plate ,
the gap height 𝐻 , and the properties of the fluid.
5.3.2. The plate speed is 𝑈plate = 1 m s−1 and the gap height is 𝐻 = 5 mm. What is the
shear force 𝐹𝑦𝑧 when the fluid is air at 40 °C, and when the fluid is crude oil at
the same temperature?
5.3.3. If a very long and thin plate with the same surface area was used instead of the
A4-shaped plate, would the shear force be different? (briefly justify your answer,
e.g. in 30 words or less)

5.4 Viscometer Çengel & al. [25] 2-78

An instrument designed to measure the viscosity of fluids (named viscometer) is made of
two coaxial cylinders (fig. 5.12). The inner cylinder is immersed in a liquid, and it rotates

The two cylinders are 75 cm tall. The inner cylinder diameter is 15 cm and the spacing
within the stationary outer cylinder.

is 1 mm.
When the inner cylinder is rotated at 300 rpm, a friction-generated moment of 0,8 N m is
measured.

5.4.1. If the flow in between the cylinders corresponds to the simplest possible flow
case (steady, uniform, fully-laminar), what is the viscosity of the fluid?
5.4.2. Would a non-Newtonian fluid induce a higher moment? (briefly justify your
answer, e.g. in 30 words or less)

[Note: in practice, when the inner cylinder is turned at high speed, the flow displays
mesmerizing patterns called Taylor—Couette vortices, the description of which is much more
complex!]
107
 Figure 5.12: Sketch of a cylinder viscometer. The width of the gap has been greatly exaggerated
for clarity.
Figure CC-0 Olivier Cleynen

5.5 Boundary layer

A laminar fluid flow occurs along a wall (fig. 5.13). Close to the wall (𝑦 < 𝛿), we observe
White [22] P1.56

layer. The speed 𝑢(𝑦) can then be modeled with the relation:
that viscous effects dominate the mechanics of the flow. This zone is designated boundary

𝜋𝑦
𝑢 = 𝑈 sin ( )
2𝛿
(5/25)

in which 𝑈 is the flow speed far away from the wall.

Figure 5.13: Velocity profile across the boundary layer.

Figure CC-0 Olivier Cleynen

The fluid is CO2 at 20 °C; measurements yield 𝑈 = 10,8 m s−1 and 𝛿 = 3 cm.

5.5.1. What is the shear effort 𝜏𝑦𝑥 on the wall?

5.5.2. At which height 𝑦1 above the surface will the shear effort be half of this value?

5.5.3. What would be the wall shear if the CO2 was replaced with water at the same
temperature?
108
5.6 Clutch Çengel & al. [25] 2-74
Two aligned metal shafts are linked by a clutch, which is made of two disks very close

disk diameters are both 30 cm and the gap between them is 2 mm; they are submerged in
one to another, rotating in the same direction at similar (but not identical) speeds. The

crude oil with temperature 80 °C.

Figure 5.14: Sketch of the two disks constituting the clutch. The gap width has been exaggerated
for clarity.
Figure CC-0 Olivier Cleynen

The power shaft rotates at 1 450 rpm, while the powered shaft rotates at 1 398 rpm. We
consider the simplest possible flow case (steady, laminar) in between the two disks.

5.6.1. What is the moment imparted by one disk to the other?

5.6.2. How would the moment change if the radius of each disk was doubled?

5.6.3. What is the transmitted power and the clutch efficiency?

5.6.4. Briefly (e.g. in 30 words or less) propose one reason why in practice the flow in
between the two disks may be different from the simplest-case flow used in this
exercise.

109
 Answers
1) 𝜏𝑦𝑥 |𝑦=−𝐻 = 1,44 N m−2 ;
2) 𝜏𝑦𝑥 |𝑦=0 = 0 N m−2 .
5.2

5.3 1) 𝐹𝑦𝑧 = 𝐿1 𝐿2 𝜇 𝐻𝑈 = 2,38 ⋅ 10−4 N for air, and 6,55 ⋅ 10−2 N for oil.

5.4 𝜇 = 𝑀𝐿Δ𝑅 / (2𝜋𝜔𝑅13 𝐻 ) = 1,281 ⋅ 10−2 N s m−2 .

1) 𝜏𝑦𝑥,wall CO2 = 8,369 ⋅ 10−3 N m−2 ;

2) 𝑦1 = 32 𝛿 = 2 cm;
5.5

3) and 𝜏𝑦𝑥,wall water = 0,565 N m−2

1) 𝑀 = 𝜋2 𝜇𝜔 𝑅 4 = 7,145 ⋅ 10−3 N m;
3) 𝑊̇ 2 = 𝜔2 𝑀 = 1,05 W; 𝜂clutch = 𝑊
ℎ
̇ 1 = 96,4 % (adequate for this very low-power, low-
5.6
𝑊̇ 2

relative speed, laminar-flow case).

110
 Fluid Dynamics
Chapter 6 – Prediction of fluid flows
last edited September 24, 2021
by Olivier Cleynen — https://fluidmech.ninja/

6.1 Motivation 111

6.2 Organizing calculations 112
6.2.1 Problem description 112
6.2.2 The total time derivative 112
6.3 Equations for all flows 116
6.3.1 Balance of mass 116
6.3.2 Balance of linear momentum 118
6.3.3 Balance of energy 120
6.3.4 Other terms and equations 121
6.3.5 Interlude: where does this leave us? 122
6.4 Equations for incompressible flow 123
6.4.1 Balance of mass 123
6.4.2 Balance of linear momentum 124
6.4.3 The Bernoulli equation (again) 127
6.5 CFD: the Navier-Stokes equations in practice 127
6.6 Solved problems 128
6.7 Problems 131

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

6.1 Motivation
In this chapter we assign ourselves the daunting task of predicting the
movement of fluids completely. We wish to express formally, and calculate,
the dynamics of fluids —the velocity field as a function of time— in any
arbitrary situation. For this, we develop a methodology named derivative
analysis.
Let us start with the unfortunate truth: not only are the methods developed
here are incredibly complex, but they are also very ineffective to solve fluid
flow problems with a pen and paper. Despite this, this chapter is extraordi-
narily important, for two reasons:

• derivative analysis allows us to formally describe and relate the key

parameters that regulate fluid flow, and so, it is the key to developing an
understanding of any fluid phenomenon, even when solutions cannot
be derived;

• it is the backbone for computational fluid dynamics (cfd) in which flow

solutions are obtained using numerical procedures, in every problem
of interest in research and industry today.

111
 6.2 Organizing calculations

6.2.1 Problem description

From now on, we wish to describe the velocity and pressure fields of a fluid
with the highest possible resolution. For this, we aim to predict and describe
the trajectory of fluid particles (recall §1.2.2 p. 10) as they travel.
Newton’s second law (recall eq. 1/25 p. 20) allows us to quantify how the
velocity vector of a particle varies with time. If we know all of the forces to

𝑉⃗particle = (𝑢, 𝑣, 𝑤) = 𝑓 (𝑥, 𝑦, 𝑧, 𝑡) as the particle moves through our area of

which one particle is subjected, we can obtain a streak of velocity vectors

interest. This is a description of the velocity of one particle; we obtain a

travel: 𝑉⃗particle = 𝑓 (𝑥0 , 𝑦0 , 𝑧0 , 𝑡0 , 𝑡). This is the process used in solid mechanics
function of time which depends on where and when the particle started its

when we wish to describe the movement of one object, for example, a satellite
in orbit.
This kind of description, however, is poorly suited to the description of fluid
flow, for three reasons:
• Firstly, in order to describe a given fluid flow (e.g. air flow around the

(𝑥0 , 𝑦0 , 𝑧0 , 𝑡0 ), and we would then obtain as many trajectories 𝑉⃗particle . It

side mirror of a car), we would need a large number of initial points

then becomes very difficult to study and describe a problem that is local
in space (e.g. the wake immediately behind the car mirror), because
this requires finding out where the particles of interest originated, and
accounting for the trajectories of each of them.
• Secondly, the concept of a “fluid particle” is not well-suited to the draw-
ing of trajectories. Indeed, not only can particles strain indefinitely,
but they can also diffuse into the surrounding particles, “blurring” and
blending themselves one into another.
• Finally, the velocity of a given particle is very strongly affected by the
properties (velocity, pressure) of the surrounding particles. We have to
resolve simultaneously the movement equations of all of the particles.

properties at a chosen fixed point of coordinates 𝑥point , 𝑦point ,

A space-based description of properties —one in which we describe

𝑧point , 𝑡point — is much more useful than a particle-based description

which depends on departure points 𝑥0 , 𝑦0 , 𝑧0 , 𝑡0 . It is easier to determine
the acceleration of a particle together with that of its current neighbors,
than together with that of its initial (former) neighbors.

that is expressed in terms of a fixed observation point 𝑉⃗point = (𝑢, 𝑣, 𝑤) =

What we are looking for, therefore, is a description of the velocity fields

𝑓 (𝑥point , 𝑦point , 𝑧point , 𝑡), through which particles of many different origins may
be passing. This is termed a Eulerian flow description, as opposed to the

in our flow study zone, we will obtain a velocity field 𝑉⃗point that is a function
particle-based Lagrangian description.w Grouping all of the point velocities

of time.

6.2.2 The total time derivative

speed 𝑢 = 𝑈canal (fig. 6.1). The temperature 𝑇water of the water is constant
Let us imagine a canal in which water is flowing at constant and uniform
112
stationary probe, reading 𝑇probe = 𝑇water on the instrument. Even though the
(in time), but not uniform (in space). We measure this temperature with a

temperature 𝑇water is constant, when reading the value measured at the probe,

{
temperature will be changing with time:
𝜕𝑇water |
𝜕𝑡 |particle
= 0
𝜕𝑇water |
𝜕𝑡 |probe
= − 𝜕𝑇𝜕𝑥
water
𝑢water

Figure 6.1: A one-dimensional water flow, for example in a canal. The water has
a non-uniform temperature, which, even if it is constant in time, translates in a
temperature rate change in time at the probe.
Figure CC-0 Olivier Cleynen

Advice from an expert

The total time derivative does not describe some kind
of funky physics phenomenon, like general relativity.
It is very straightforward.
It is important to realize that properties are not changed
by the reference frame. In the canal example above, the
temperature measured at the probe (𝑇probe ) is always
the temperature of the water (𝑇water ), regardless of the
flow conditions. By contrast, it is the change in time of the properties
which is different. This is because as time passes, different particles keep
hitting the probe, and they already have differing temperatures. Nothing
spooky here!

Advice from an expert

Do not let yourself be fooled by the idea that this is just
an abstract curiosity. Using the total time derivative
is what enables modern computational fluid dynamics
(cfd) to work at all. In those simulations, cells in a fixed,
stationary grid are attributed coordinates. The flow
through those cells is computed without ever following
113
 the particles (this would rapidly become messy, as particles mix and
tangle up). Cool kids do not ever “flow” the fluid flow. Instead, they
oscillate vectors, like beautiful three-dimensional wheat fields gently
oscillating with the wind.

Let us now study the case where the water temperature, in addition to being
non-uniform, is also decreasing everywhere because the canal is cooling

d𝑇 / d𝑡, which also affects the reading at the probe:

down. A particle’s temperature will then be changing at a non-zero rate

{
𝜕𝑇water |
𝜕𝑡 |particle
≠ 0
Video: figuring out the total

𝜕𝑇water |
𝜕𝑡 |probe
= 𝜕𝑇𝜕𝑡water ||particle − 𝜕𝑇𝜕𝑥 𝑢water
time derivative
by Olivier Cleynen (CC-by)
https://youtu.be/T6POOdLK4ok water

Re-arranging this last equation, we obtain the time change of the particle’s
temperature, expressed from the reference frame of the probe:
𝜕𝑇water | 𝜕𝑇water |
𝜕𝑡 |particle
= 𝜕𝑡 |probe
+ 𝜕𝑇water
𝜕𝑥
𝑢water (6/1)

We must keep in mind that all those derivatives can themselves be functions
of time and space; in equation 6/1, it is their value at the position of the probe
and at the time of measurement which is taken into account.

property 𝐴 of the fluid (including vector properties). The property 𝐴 of

This line of thought can be generalized for three dimensions and for any

𝐴 = 𝑓 (𝑥, 𝑦, 𝑧, 𝑡) within the fluid. The time rate change of 𝐴 expressed at

one individual particle can vary as it is moving, so that it has a distribution

derivative 1 of 𝐴 and written D𝐴/D𝑡:

a point fixed in space is named the total time derivative or simply total

D 𝜕 𝜕 𝜕 𝜕
Video: half-century-old, but

≡ +𝑢 +𝑣 +𝑤
timeless didactic exploration of

D𝑡 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
the concept of Lagrangian and
Eulerian derivatives, with ac- (6/2)

D𝐴 𝜕𝐴 𝜕𝐴 𝜕𝐴 𝜕𝐴
companying notes by Lumley[4]

= +𝑢 +𝑣 +𝑤
by the National Committee for Fluid

D𝑡 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
Mechanics Films (ncfmf, 1969[21]) (styl)
https://youtu.be/mdN8OOkx2ko
(6/3)
⃗
D𝐴 𝜕 𝐴 ⃗ 𝜕𝐴⃗ 𝜕𝐴 ⃗ 𝜕 𝐴⃗
= +𝑢 +𝑣 +𝑤
D𝑡 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧
(6/4)

⎛ 𝜕𝑡𝑥 + 𝑢 𝜕𝑥𝑥 + 𝑣 𝜕𝑦𝑥 + 𝑤 𝜕𝐴

𝜕𝐴 𝜕𝐴 𝜕𝐴 ⎞
⎜ 𝜕𝐴𝑦 ⎟
𝑥
𝜕𝑧
= ⎜ 𝜕𝑡 + 𝑢 𝜕𝑥 + 𝑣 𝜕𝑦 + 𝑤 𝜕𝑧 ⎟
𝜕𝐴𝑦 𝜕𝐴𝑦 𝜕𝐴𝑦

⎜ 𝜕𝐴𝑧 + 𝑢 𝜕𝐴𝑧 + 𝑣 𝜕𝐴𝑧 + 𝑤 𝜕𝐴𝑧 ⎟

(6/5)
⎝ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ⎠

In equations 6/3 and 6/4, it is possible simplify the notation of the last three

velocity vector 𝑉⃗ (by definition), and the components of the operator gradient
terms. For this, we have to recall two ingredients: the coordinates of the

⃗ (previously introduced in eq. 4/11 p. 77), writing them out as so:

∇

𝑉⃗ ≡ 𝑖⃗ 𝑢 + 𝑗⃗ 𝑣 + 𝑘⃗ 𝑤
𝜕 ⃗ 𝜕 𝜕
⃗ ≡ 𝑖⃗
∇ +𝑗 + 𝑘⃗
𝜕𝑥 𝜕𝑦 𝜕𝑧

1 Unfortunately this term has many denominations across the literature, including
advective, convective, hydrodynamic, Lagrangian, particle, substantial, substantive, or
114 Stokes derivative. In this document, the term total derivative is used.
We can now define the advective operator,w written (𝑉⃗ ⋅ ∇
⃗ ) (see also Ap-
pendix A3.3 p. 251):
𝜕 𝜕 𝜕
𝑉⃗ ⋅ ∇
⃗≡𝑢 +𝑣 +𝑤
𝜕𝑥 𝜕𝑦 𝜕𝑧
(6/6)

We can now rewrite eqs. 6/2 and 6/3 in a more concise way:
D 𝜕
≡ + (𝑉⃗ ⋅ ∇
⃗)
D𝑡 𝜕𝑡
(6/7)

D𝐴 𝜕𝐴
≡ + (𝑉⃗ ⋅ ∇
⃗ )𝐴
D𝑡 𝜕𝑡
(6/8)

D𝐴⃗ 𝜕 𝐴⃗
≡ + (𝑉⃗ ⋅ ∇
⃗ )𝐴⃗
D𝑡 𝜕𝑡
(6/9)

Advice from an expert

To make sure you master the concept of total deriva-
tive, try thinking of examples where only one member
of equation 6/8 is zero.

particle (D/D𝑡 = 0) may be unsteady in the reference

A flow that is steady from the point of view of the

frame of the laboratory (𝜕/𝜕𝑡 ≠ 0). This happens for

example where a fluid with non-uniform temperature flows at constant
velocity.
Conversely, a flow that is unsteady from the point of view of the particle

jet engine nozzle, the air particles accelerate sharply (D/D𝑡 ≠ 0), however
may be steady in the reference frame of the laboratory. For example, in a

the flow is steady from the point of view of the jet engine (𝜕/𝜕𝑡 = 0).

The total time derivative is the tool that we were looking for. From now on,
we can study fluid flows from a stationary reference frame, instead of in the

the same, but all the time derivatives d/ d𝑡 are replaced with total derivatives
reference frame of a moving particle. When we do so, all properties remain

D/ D𝑡. This allows us to compute the change in time of a property locally,

without the need to track the movement of particles along our field of study.
All computational fluid dynamics (cfd) simulations work in this manner.
The first and most important such property we are interested in is accel-

(d𝑉⃗particle / d𝑡), we will instead calculate focus on calculating the acceleration

eration. Instead of solving for the acceleration of each of many particles

field D𝑉 / D𝑡:

D𝑉⃗ 𝜕 𝑉⃗
= + (𝑉⃗ ⋅ ∇
⃗ )𝑉⃗
D𝑡 𝜕𝑡
(6/10)

An illustration of an acceleration field is shown further down in figure 6.5

p. 126.

115
 6.3 Equations for all flows

6.3.1 Balance of mass

How can we write a balance of mass equation for an entire complex, unsteady,

volume d, at a given instant in time (fig. 6.2).

three-dimensional fluid flow? Let’s begin by considering a fluid particle of

We can reproduce our analysis from chapter 3 (Analysis of existing flows with
three dimensions) by quantifying the mass flows passing through an infinitesimal
volume. In the present case, the control volume is stationary and the particle
(our system) is flowing through it. We start with eq. 3/6 p. 54:
d𝑚particle d
=0= 𝜌 d + ∬ 𝜌(𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 d𝑡 ∭CV
(6/11)
CS

The first of these two integrals can be rewritten using the Leibniz integral
rule:
d 𝜕𝜌
∭ 𝜌 d = ∭ d + ∬ 𝜌𝑉𝑆 d𝐴
d𝑡 CV 𝜕𝑡
𝜕𝜌
=∭ d
CV CS

CV 𝜕𝑡
XKCD #1524: changes in time
by Randall Munroe (CC-by-nc)
https://xkcd.com/1524 (6/12)

where 𝑉𝑆 is the speed of the control volume wall;

and where the term∬CS 𝜌𝑉𝑆 d𝐴 is simply zero because we chose a fixed control
volume, such as a fixed computation grid.1

Now we turn to the second term of equation 6/11, ∬CS 𝜌(𝑉⃗rel ⋅ 𝑛)

⃗ d𝐴, which
represents the mass flow 𝑚̇ net flowing through the control volume.

Figure 6.2: Conservation of mass within a fluid particle. In the 𝑥-direction, a mass
flow 𝑚̇ 1 = ∬ 𝜌1 𝑢1 d𝑧 d𝑦 is flowing in, and a mass flow 𝑚̇ 2 = ∬ 𝜌2 𝑢2 d𝑧 d𝑦 is flowing
out. These two flows may not be equal, since mass may also flow in the 𝑦- and
𝑧-directions.
Figure CC-0 Olivier Cleynen

1 Ina cfd calculation in which the grid is deforming, this term ∬CS 𝜌𝑉𝑆 d𝐴 will have to
116 be re-introduced in the continuity equation.
In the direction 𝑥, the mass flow 𝑚̇ net 𝑥 flowing through our control volume
can be expressed as:

𝑚̇ net 𝑥 = ∬ −𝜌1 |𝑢1 | d𝑧 d𝑦 + ∬ 𝜌2 |𝑢2 | d𝑧 d𝑦

𝜕
=∬ ∫ (𝜌𝑢) d𝑥 d𝑧 d𝑦
CS CS

𝜕𝑥
𝜕
=∭ (𝜌𝑢) d
CS

CV 𝜕𝑥
(6/13)

The same applies for directions 𝑦 and 𝑧, so that we can write:

⃗ ⃗ d𝐴 = 𝑚̇ net = 𝑚̇ net 𝑥 + 𝑚̇ net 𝑦 + 𝑚̇ net 𝑧

∬ 𝜌(𝑉rel ⋅ 𝑛)
𝜕 𝜕 𝜕
=∭ (𝜌𝑢) + (𝜌𝑣) + (𝜌𝑤) d
CS

CV [ 𝜕𝑥 𝜕𝑦 𝜕𝑧 ]
=∭ ⃗ ⋅ (𝜌 𝑉⃗ ) d
∇ (6/14)
CV

(note that here we have again used the operator divergent, which we first
used in chapter 5 p. 95 — see also Appendix A3 p. 250)
Now, with these two equations 6/12 and 6/14, we can come back to equa-
tion 6/11, which becomes:
d𝑚particle d
=0= 𝜌 d + ∬ 𝜌(𝑉⃗rel ⋅ 𝑛)⃗ d𝐴
d𝑡 d𝑡 ∭CV
𝜕𝜌
0= ∭ d + ∭ ∇ ⃗ ⋅ (𝜌 𝑉⃗ ) d
CS

CV 𝜕𝑡 CV
(6/15)

Since we are only concerned with a very small volume d, we drop the
integrals, obtaining:

𝜕𝜌
⃗ ⋅ (𝜌 𝑉⃗ ) = 0
+∇
𝜕𝑡
(6/16)

for all flows, with all fluids.

This equation 6/16 is named continuity equationw and is of crucial importance

in fluid mechanics. It sums up two terms:

• The first term on the left, 𝜕𝜌/𝜕𝑡, is the local time-change of density. For
example, when a gas contracts as it cools down, its density increases,
and the term becomes positive. In an incompressible flow, it is always
zero.

• The second term is the divergent of density times velocity, ∇ ⃗ ⋅ 𝜌 𝑉⃗ =

𝜕𝜌𝑢/𝜕𝑥 + 𝜕𝜌𝑣/𝜕𝑦 + 𝜕𝜌𝑤/𝜕𝑧. It sums up the changes in space of the
mass fluxes 𝜌𝑉𝑖 . For example, if a particle in a static fluid is heated
up suddenly, it will expand and the divergent of 𝜌 𝑉⃗ will have positive
value.

117
 6.3.2 Balance of linear momentum
What is the force field applying to the fluid everywhere in space and time,
and how does that affect its velocity field? To answer this question, we write
out a momentum balance equation.

particle of mass 𝑚particle , as shown in fig. 6.3. Fundamentally, the forces on a

We start by writing Newton’s second law (eq. 1/25 p. 20) as it applies to a fluid

fluid particle are of only three kinds, namely weight, pressure, and shear: 1

d𝑉⃗ ⃗
𝑚particle = 𝐹weight + 𝐹⃗net, pressure + 𝐹⃗net, shear
d𝑡
(6/17)

infinitesimal volume d, which is traversed by a fluid particle. We measure

We now write this equation from the point of view of a stationary cube of

the time-change of velocity from the reference frame of the cube, making
good use of the total time derivative tool we developed earlier p. 115:

D𝑉⃗ ⃗
𝑚particle = 𝐹weight + 𝐹⃗net, pressure + 𝐹⃗net, shear
D𝑡
𝑚particle D𝑉⃗ 1 ⃗ 1 ⃗ 1 ⃗
= 𝐹weight + 𝐹net, pressure + 𝐹net, shear
d D𝑡 d d d
D𝑉⃗ 1 ⃗ 1 ⃗ 1 ⃗
𝜌 = 𝐹weight + 𝐹net, pressure + 𝐹net, shear
D𝑡 d d d
(6/18)

Now, we rewrite the forces term on the right, and the hard work we did in
the previous chapters is paying off.
The force due to gravity is of course the weight. We have:
1 ⃗ 1
𝐹weight = 𝑚𝑔⃗ = 𝜌 𝑔⃗
d d
(6/19)

Figure 6.3: In our study of fluid mechanics, we consider only forces due to gravity,
shear, or pressure.
Figure CC-0 Olivier Cleynen

1 In some special applications, additional forces may also apply, see §6.3.4 p. 121 further
118 down.
The force due to pressure was dealt with in chapter 4 (Effects of pressure). Back
then, we expressed it with the help of the gradient of pressure in equation 4/13
p. 77, which we repeat here:
1 ⃗
⃗𝑝
𝐹net, pressure = −∇
d
(6/20)

And finally, we had dealt with the shear force in chapter 5 (Effects of shear).
With somewhat effort, we had expressed it as a function of the divergent of
shear in equation 5/19 p. 95, which we repeat here:
1 ⃗
⃗ ⋅ 𝜏⃗𝑖𝑗
𝐹net, shear = ∇
d
(6/21)

Now, we can put together all of our findings back into equation 6/18, we
obtain the Cauchy equation:w

D𝑉⃗
𝜌 ⃗𝑝 + ∇
= 𝜌 𝑔⃗ − ∇ ⃗ ⋅ 𝜏⃗𝑖𝑗
D𝑡
(6/22)

for all flows, with all fluids.

The Cauchy equation is a formulation of Newton’s second law applied to

from a fixed reference frame (the acceleration field D𝑉⃗ /D𝑡) as a the sum
a fluid particle. It expresses the time change of the velocity field measured

of the contributions of gravity, pressure and shear effects. This is quite a

breakthrough. Within the chaos of an arbitrary flow, in which fluid particles
are shoved, pressurized, squeezed, and distorted, we know precisely what we
need to look for in order to quantify the time-change of velocity: gravity, the
gradient of pressure, and the divergent of shear.

for us. In our search for the velocity field 𝑉⃗ , the changes in time and space
Nevertheless, while it is an excellent start, this equation isn’t detailed enough

of the shear tensor 𝜏⃗𝑖𝑗 and pressure 𝑝 are unknowns. Ideally, those two
terms should be expressed solely as a function of the flow’s other properties.
Obtaining such an expression is what Claude-Louis Navier and Gabriel Stokes
set themselves to in the 19th century: we follow their footsteps in the next
paragraphs.
The Navier-Stokes equation is the Cauchy equation (eq. 6/22) applied to
Newtonian fluids. In Newtonian fluids, which we encountered in chapter 5
(§5.4.4 p. 100), shear efforts are simply proportional to the rate of strain; thus,

We had seen with eq. 5/22 p. 97 that the norm ||⃗ 𝜏𝑖𝑗 || of shear component in
the shear component of eq. 6/22 can be re-expressed usefully.

direction 𝑗 along a surface perpendicular to 𝑖 depended on the viscosity and

the velocity:
𝜕𝑢𝑗
||⃗
𝜏𝑖𝑗 || = 𝜇
𝜕𝑖

This is a one-dimensional (scalar) equation. Unfortunately, it does not trans-

late easily into three dimensions. The required vector algebra far exceeds our
level for this course, and we are interested only in the result (the derivation
of this equation in Cartesian coordinates is covered in Anderson [9] and Ver-
steeg & Malalasekra [20], and the vector form can be found in Batchelor [2]).
We obtain the heavy-handed result, in the form of a (three-dimensional)

119
 1
vector field:

⃗ 2 𝑉⃗ + 𝜇 ∇
⃗ ⋅ 𝜏⃗𝑖𝑗 = 𝜇 ∇
∇ ⃗ ⋅ 𝑉⃗
⃗ ∇
3 ( ) (6/23)

⃗ 2 ) do not
The details of the notation (which includes the Laplacian operator ∇
interest us at the moment; we will explore them later on.
Adding this relationship between shear and the velocity field into the last
term of equation 6/22, we obtain the Navier-Stokes equation for compressible
flow:w

D𝑉⃗ 1
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗ + 𝜇 ∇
⃗ 𝑝 + 𝜇∇ ⃗ ⋅ 𝑉⃗
⃗ ∇
D𝑡 3 ( ) (6/24)

for all flows of a Newtonian fluid.

followed by the velocity field 𝑉⃗ , in all possible flows of a Newtonian fluid.

This three-dimensional vector equation sets the conditions that are to be

We will use a simplified version of this equation in section 6.4.2 below.

6.3.3 Balance of energy

This topic is well covered in Anderson [9] and Versteeg & Malalasekra [20]

How much energy is expended or received by the fluid particles as they travel
through a complex, arbitrary flow? We answer this question with an energy
balance equation. Once again, we start from the analysis of transfers on an

the following form, naming them 𝐴, 𝐵 and 𝐶 for clarity:

infinitesimal control volume. We are going to relate three energy terms in

= +
the rate of change the net flux the rate of work done
of energy inside of heat into on the element due to

𝐴 = 𝐵 + 𝐶
the fluid element the element body and surface forces
(6/25)

Let us first evaluate term 𝐶. The rate of work done on the particle is the dot
product of its velocity 𝑉⃗ and the net force 𝐹⃗net applying to it:

𝐹⃗weight 𝐹⃗net, pressure 𝐹⃗net, shear

𝐶 = 𝑉⃗ ⋅ + + d
( d d d )

We replace the content of the parentheses with the right part of the Navier-
Stokes equation above, obtaining the scalar term:
1
𝐶 = 𝑉⃗ ⋅ [𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗ + 𝜇 ∇
⃗ 𝑝 + 𝜇∇ ⃗ ⋅ 𝑉⃗ d
⃗ ∇
3 ( )] (6/26)

We now turn to term 𝐵, the net flux of heat into the element. We attribute this
flux to two contributions (i.e. 𝐵 = 𝑄̇ radiation + 𝑄̇ conduction ). The first contributor
is the heat transfer 𝑄̇ radiation from the emission or absorption of radiation,
which we shyly express as:

𝑄̇ radiation = 𝜌 𝑞̇ radiation d (6/27)

in which 𝑞̇ radiation is the local power per unit mass (in W kg−1 ) transfered to the
element, to be determined from the boundary conditions and flow temperature
120 distribution.
The second contributor the term 𝐵 is the named 𝑄̇ conduction , attributed to
thermal conduction through the faces of the element. In the 𝑥-direction,

heat 𝑄̇ conduction,𝑥 expressed as a function of the power per area 𝑞 (in W m−2 )
thermal conduction through the faces of the element causes a net flow of

through each of the two faces perpendicular to 𝑥:

𝜕𝑞𝑥
𝑄̇ conduction,𝑥 = 𝑞𝑥 − 𝑞𝑥 + d𝑥 d𝑦 d𝑧
[ ( 𝜕𝑥 )]
𝜕𝑞𝑥
=− d𝑥 d𝑦 d𝑧
𝜕𝑥
𝜕𝑞𝑥
(6/28)

=− d
𝜕𝑥
(6/29)

In turn, the fluxes 𝑞𝑖 can be expressed as a function of the local temperature

gradients according to the Fourier law,
𝜕𝑇
𝑞𝑖 = −𝜅
𝜕𝑖
(6/30)

where 𝜅 is the conductivity of the fluid (W m−1 K−1 ).

so that now we may write the heat transfer due to conduction as

𝑄̇ conduction = 𝑄̇ conduction,𝑥 + 𝑄̇ conduction,𝑦 + 𝑄̇ conduction,𝑧

𝜕𝑞𝑥 𝜕𝑞𝑦 𝜕𝑞𝑧
=− + + d
[ 𝜕𝑥 𝜕𝑦 𝜕𝑧 ]
(6/31)

𝜕 2𝑇 𝜕 2𝑇 𝜕 2𝑇
=𝜅 + + d
[ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
(6/32)

Finally, term 𝐴, the rate of change of energy inside the fluid element, can
be expressed as a function of the specific internal energy 𝑖 specific kinetic
energy 𝑒𝑘 and :
D 1
𝐴=𝜌 ( 𝑖 + 𝑉 2 ) d
D𝑡 2
(6/33)
Abstruse Goose #275: how scien-
tists see the world
by an anonymous artist (CC-by-nc)
We are therefore able quantify the change of energy of fluid particle with a https://abstrusegoose.com/275

scalar field equation as follows:

D 1 2 𝜕 2𝑇 𝜕 2𝑇 𝜕 2𝑇
𝜌 𝑖 + 𝑉 = 𝜌 𝑞̇ + 𝜅 + +
D𝑡 ( 2 ) [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
1
radiation

+𝑉⃗ ⋅ [𝜌 𝑔⃗ − ∇ ⃗ 𝑝 + 𝜇∇ ⃗ 2 𝑉⃗ + 𝜇 ∇ ⃗ ∇⃗ ⋅ 𝑉⃗
3 ( )]
(6/34)

6.3.4 Other terms and equations

The more is happening in a given flow, and the more equations and equation
terms are needed to describe it. Depending on the applications, new forces
may become dominant and might be added to gravity, pressure and shear in
our equations. For example:

• In a description of an ocean flow, a tidal force may be added;

121
 • In a description of a large atmospheric flow, a Coriolis force may be
added;

• In a description of sap flow in a tree or plant, forces related to surface

tension may be added.

Additionally, modeling some more advanced fluid phenomena requires alto-

gether new equations, for example:

• Equations to describe chemical reactions, such as combustion;

• Equations to distinguish and model the interaction of several fluids (e.g.

in flow featuring drops, bubbles, or non-uniform mixes).

6.3.5 Interlude: where does this leave us?

So far, we have written five equations:

• One equation for balance of mass (eq. 6/16 p. 117);

• Three equations for balance of momentum (the three components of

the vector equation 6/24 p. 120);

• One equation for balance of energy (eq. 6/34 p. 121).

In order to solve an arbitrary flow involving transfers of both momentum and

energy —for example, the flow of air in a room when an electric fan heater is
turned on— we need to solve all five equations simultaneously. It is hard to
hide that the mathematical complexity of the problem is simply intractable.
No method is available to find a solution of this problem “by hand”, and
the computational costs to find numerical solutions with computers are
enormous.
This is the reason why in the section below, we narrow down our scope down
to more restrictive conditions: incompressible flows in which the energy
does not significantly vary. We will write again a balance of mass and a
balance of momentum, and take more time to explore the shape and behavior
of the resulting four equations.

Advice from an expert

Fluid dynamicists are not easy to observe in the wild,
but if you get the chance, you will likely find them
busy trying to not solve equations! At the start of
every problem, the effective fluid dynamicist will try
to figure which terms s/he can neglect, and reduce
the number of equations to solve. This is because
in practice, both experiments and computations are
prohibitively expensive, and answers need to be ob-
tained in reasonable amounts of time. In fact, we will soon see how
we can “weigh” the relative importance of equation terms. In chapter 8
(Engineering models) and following, we’ll be removing terms from eq. 6/24
until only the significant ones remain.

122
6.4 Equations for incompressible flow

(§1.8 p. 21): those for which 𝜌 is uniform and constant.

For the rest of this document, we focus on fluid flows that are incompressible

6.4.1 Balance of mass

To write a balance of mass for incompressible flow, we begin where we left
off with the general mass balance (eq. 6/16 p. 117), which we re-write here:
𝜕𝜌
⃗ ⋅ (𝜌 𝑉⃗ ) = 0
+∇
𝜕𝑡

In this equation, we see that if 𝜌 is uniform and constant, the first term
will vanish. Once this happens, 𝜌 can be simply dropped from the second
term. This leaves us with the (much simpler) mass balance equation for
incompressible flow, also called incompressible continuity equation:

⃗ ⋅ 𝑉⃗ = 0
∇ (6/35)

for any incompressible flow.

This equation states that “the divergent of velocity is zero” and it is a scalar
equation. In three Cartesian coordinates, it can be re-expressed as:
𝜕𝑢 𝜕𝑣 𝜕𝑤
+ + =0
𝜕𝑥 𝜕𝑦 𝜕𝑧
(6/36)

particular information about the orientation of 𝑉⃗ or about its change in time.

In spite of its elegance, equation 6/35 is not very talkative: it gives us no

How should we calculate 𝑢, 𝑣 or 𝑤 if we have only one information about

the sum of their derivatives in space?
In practice, the equation is insufficient to solve the majority of problems in
fluid mechanics, and it is used as a kinematic constraint to solutions used
to evaluate their physicality or the quality of the approximations made to
obtain them.

Advice from an expert

As strange as it sounds, the continuity equation is a

you already have the three fields 𝑢, 𝑣 or 𝑤, you can plug

constant source of frustration for fluid dynamicists. If

But if you don’t (for example, if you are missing 𝑤), how
them into equation 6/36, and hope it adds up to zero.

do you get to the missing element? Only one partial

derivative is specified for each of the elements, leaving a

𝑤 to 𝑥, 𝑦 and 𝑡).
lot of unknown dependencies (e.g. the functions relating

A great deal of completely nonsensical fluid flows will abide by the

continuity equation. They cannot be realized, because they would require
non-physical changes in fluid momentum. How do we determine what is
possible? We need a momentum balance equation.
123
 6.4.2 Balance of linear momentum
Here, we start from the balance of momentum equation which we obtained
for general flow as the Cauchy equation, eq. 6/22 p. 119:

D𝑉⃗
𝜌 ⃗𝑝 + ∇
= 𝜌 𝑔⃗ − ∇ ⃗ ⋅ 𝜏⃗𝑖𝑗
Video: worst one ever? The D𝑡
incompressible Navier-Stokes

⃗ ⋅ 𝜏⃗𝑖𝑗 , which has three components: ∇

∇ ⃗ ⋅ 𝜏⃗𝑖𝑥 , ∇
⃗ ⋅ 𝜏⃗𝑖𝑦 and ∇
⃗ ⋅ 𝜏⃗𝑖𝑧 . Let us first focus
equation
by Olivier Cleynen (CC-by) Clearly, the “hard” term in this equation is the last one: the divergent of shear

on the 𝑥-direction, using the diagram in figure 6.4.

https://youtu.be/QhjSwcSCeHQ

In the 𝑥-direction, the net effect of shear on the six faces is a vector expressed
as the divergent ∇⃗ ⋅ 𝜏⃗𝑖𝑥 , which, in incompressible flow, can be expressed with
with the help of viscosity 𝜇 as:

𝜕 𝜏⃗𝑥𝑥 𝜕 𝜏⃗𝑦𝑥 𝜕 𝜏⃗𝑧𝑥

⃗ ⋅ 𝜏⃗𝑖𝑥 =
∇ + +
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝑢 ⃗
𝜕 (𝜇 𝜕𝑢 𝑖⃗ 𝜕 (𝜇 𝜕𝑦 𝑖 ) 𝜕 (𝜇 𝜕𝑢 𝑖⃗)
𝜕𝑥 )
= + + 𝜕𝑧
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕 ( 𝜕𝑢 𝜕 ( 𝜕𝑦 ) 𝜕𝑢
𝜕 ( 𝜕𝑢
𝜕𝑥 ) ⃗ )
=𝜇 𝑖+𝜇 𝑖 + 𝜇 𝜕𝑧 𝑖⃗
⃗
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕 2𝑢 𝜕 2𝑢 𝜕 2𝑢 ⃗
=𝜇 + + 𝑖
( (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 )
(6/37)

To tidy up this tedious equation structure, we introduce the Laplacian oper-

ator (see also Appendix A3 p. 250) to represent the spatial variation of the
spatial variation of an object:

∇⃗2 ≡ ∇⃗ ⋅∇⃗
⃗ 2𝐴 ≡ ∇
∇ ⃗ ⋅∇⃗𝐴
(6/38)

⃗ 2 𝐴𝑥
⎛ ∇ ⎞ ⃗ ⋅∇
⎛ ∇ ⃗ 𝐴𝑥 ⎞
(6/39)

⃗ 2 𝐴⃗ ≡ ⎜ ∇
∇ ⃗ 2𝐴 ⎟ = ⎜ ∇
⃗ ⋅∇
⃗ 𝐴𝑦 ⎟
⎜ 2 𝑦 ⎟ ⎜ ⎟
⃗ 𝐴𝑧
⎝ ∇ ⎠ ⃗ ⋅∇
⎝ ∇ ⃗ 𝐴𝑧 ⎠
(6/40)

Figure 6.4: Shear efforts on a cubic fluid particle (already represented in fig. 5.1 p. 93).
Figure CC-0 Olivier Cleynen

124
And now, we can re-write eq. 6/37 more elegantly and generalize to three
dimensions:

⃗ ⋅ 𝜏⃗𝑖𝑥 = 𝜇 ∇
∇ ⃗ 2 𝑢 𝑖⃗ = 𝜇 ∇
⃗ 2 𝑢⃗
∇ ⃗ 2 𝑣 𝑗⃗ = 𝜇 ∇
⃗ ⋅ 𝜏⃗𝑖𝑦 = 𝜇 ∇ ⃗ 2 𝑣⃗
∇ ⃗ 2 𝑤 𝑘⃗ = 𝜇 ∇
⃗ ⋅ 𝜏⃗𝑖𝑧 = 𝜇 ∇ ⃗ 2 𝑤⃗

The three last equations are grouped together simply as

∇ ⃗ 2 𝑉⃗
⃗ ⋅ 𝜏⃗𝑖𝑗 = 𝜇 ∇ (6/41)

With this new expression, we can come back to the Cauchy equation (eq. 6/22
p. 119), in which we can replace the shear term with eq. 6/41. This produces
the beautiful Navier-Stokes equation for incompressible flow:

D𝑉⃗
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
D𝑡
(6/42)

for all incompressible flows of a Newtonian fluid.

This simplified but still formidable equation describes the property fields of
all incompressible flows of Newtonian fluids. It expresses the acceleration
field (left-hand side) as the sum of three contributions (right-hand side): those

for in equation 6/42 are the velocity (vector) field 𝑉⃗ = (𝑢, 𝑣, 𝑤) = 𝑓1 (𝑥, 𝑦, 𝑧, 𝑡)
of gravity, gradient of pressure, and divergent of shear. The solutions we look

and the pressure field 𝑝 = 𝑓2 (𝑥, 𝑦, 𝑧, 𝑡), given a set of constraints to represent XKCD #435: scientific fields
sorted by purity
the problem at hand. by Randall Munroe (CC-by-nc)
https://xkcd.com/435
Though it is without doubt charming, equation 6/42 should be remembered
for what it is really: a three-dimensional system of coupled equations. In
Cartesian coordinates this complexity is more apparent:

𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢 𝜕 2𝑢
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑥 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑥 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑝 𝜕 2𝑣 𝜕 2𝑣 𝜕 2𝑣
(6/43)
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑦 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑦 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]

𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑝 𝜕 𝑤 𝜕 𝑤 𝜕 𝑤
(6/44)
2 2 2
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑧 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑧 [ (𝜕𝑥) 2 (𝜕𝑦) 2 (𝜕𝑧)2 ]
(6/45)

Advice from an expert

The incompressible Navier-Stokes equation is
the crux of modern, practical fluid dynamics.
Don’t be one of those students who just casu-
ally dismiss it, saying “computers will solve it”.
To understand why real-world fluid dynamicists
work so hard rigging and configuring their sim-
ulations and experiments, and then spend so
much time waiting for their computational simulations to complete, you
must have minimal experience playing with the mathematics. Take
125
 the vector equation 6/42, and practice expanding it into eqs. 6/43-6/45.
It’s a necessary step, like the first time jumping in the deep end of the
swimming pool!

Today indeed, 150 years after it was first written, no general expression
has been found for velocity or pressure fields that would solve this vector
equation in the general case. Nevertheless, in this course we will use it
directly:

• to understand and quantify the importance of key fluid flow parameters,

in chapter 8 (Engineering models);

• to find analytical solutions to flows in a few selected cases, in the other

remaining chapters.

After this course, the reader might also engage into computational fluid
dynamics (cfd) a discipline entirely architectured around this equation, and
to which it purposes to find solutions as fields of discrete values.

Figure 6.5: The velocity (top) and acceleration (bottom) fields in the flow field of
the computed flow described in figure 4.4 p. 78. The velocity field is the result of a
computational fluid dynamics simulation (steady, Reynolds-Averaged: a modified
version of eq. 6/42). The acceleration field is obtained based on that solution, with
equation 6/10, but could also have been obtained with the right side of equation 6/42.
Figure CC-by by Arjun Neyyathala

126
6.4.3 The Bernoulli equation (again)
We had made clear in chapter 2 (Analysis of existing flows with one dimension) that
the Bernoulli equation was very limited in scope, and that it was always safer
to approach a problem from an energy equation instead (§2.6 p. 41). As a
reminder of this fact, and as an illustration of the bridges that can be built
between integral and derivative analysis, it can be instructive to derive the
Bernoulli equation directly from the Navier-Stokes equation. This derivation
Video: Bernoulli wants to help
is not difficult to follow; it is covered in Appendix A4.3 p. 253. Navier and Stokes
by Olivier Cleynen (CC-by)
https://youtu.be/qa5wvGOcg0Q

6.5 CFD: the Navier-Stokes equations in practice

This topic is well covered in Versteeg & Malalasekera [20]

In our analysis of fluid flow from a derivative perspective, our five physical
principles from §1.7 have been condensed into three balance equations (often
loosely referred together to as the Navier-Stokes equations). Out of these, the
first two, for conservation of mass (6/35) and linear momentum (6/42) in
incompressible flows, are often enough to characterize most free flows, and Video: the very basics of CFD

velocity field 𝑉⃗ :
should in principle be enough to find the primary unknown, which is the by Olivier Cleynen (CC-by)
https://youtu.be/bcmbRzF67Ig

0=∇⃗ ⋅ 𝑉⃗
⃗
D𝑉
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
D𝑡

We know of many individual analytical solutions to this mathematical prob-

lem: they apply to simple cases, and we shall describe several such flows in
the upcoming chapters. However, we do not have one general solution: one
that would encompass all of them. For example, in solid mechanics we have

solution 𝑥 = 𝑥0 + 𝑢0 𝑡 and 𝑦 = 𝑦0 + 𝑣0 𝑡 + 12 𝑔𝑡 2 , regardless of the particularities

long understood that all pure free fall movements can be described with the

of each fall. In fluid mechanics, even though our analysis was carried out in
the same manner, we have yet to find such one general solution — or even to
prove that one exists at all.
It is therefore tempting to attack the above pair of equations from the nu-

small increments 𝛿𝑥, 𝛿𝑦, 𝛿𝑧 and 𝛿𝑡, we could re-express the 𝑥-component of
merical side, with a computer algorithm. If one discretizes space and time in

eq. 6/42 as:

𝛿𝑢|𝑦 |
⎡ 𝛿 𝛿𝑢| 𝑥| 𝛿 𝑧| ⎤
𝛿𝑢|𝑡 𝛿𝑢|𝑥 𝛿𝑢|𝑦 𝛿𝑢|𝑧 𝛿𝑝|𝑥 |
𝛿𝑥 |𝑥 𝛿𝑦 ||𝑦 𝛿 𝛿𝑢|
𝛿𝑧 ||𝑧
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑥 − +𝜇⎢ + + ⎥
[ 𝛿𝑡 𝛿𝑥 𝛿𝑦 𝛿𝑧 ] 𝛿𝑥 ⎢ 𝛿𝑥 𝛿𝑦 𝛿𝑧 ⎥
⎣ ⎦
(6/46)

pressure, this equation 6/46 allows us to isolate and solve for 𝛿𝑢|𝑡 , and
If we start with a known (perhaps guessed) initial field for velocity and

therefore predict what the 𝑢 velocity field would look like after a time
increment 𝛿𝑡. The same can be done in the 𝑦- and 𝑧-directions. Repeating the

obtaining at every new time step the value of 𝑢, 𝑣 and 𝑤 at each position
process, we then proceed to the next time step and so on, marching in time,

within our computation grid.

The discretization schemes and solver algorithms used in practice in con-
temporary software are of course more subtle than those described here; 127
 nevertheless, this single-page brief does lay out the fundamental working
principle of computational fluid dynamics (cfd) today.
Diving into the intricacies of cfd is beyond the scope of our study. Never-
theless, the remarks in this last section should hopefully hint at the fact that
an understanding of the mathematical nature of the differential conservation
equations is of great practical importance in fluid dynamics. It is for that
reason that the problems in this chapter are dedicated to playing with the
mathematics of our two main equations.

6.6 Solved problems

A flow has the velocity field 𝑉⃗ = (2 + 3𝑥)𝑖⃗ + (4 − 3𝑦)𝑗⃗ (in si units).

Acceleration field

What is the acceleration field?

See this solution worked out step by step on YouTube

https://youtu.be/ZTtzavDri0o (CC-by Olivier Cleynen)

Playing with the continuity equation

An incompressible flow has the velocity field defined as follows:

𝑢 = 2𝑥 2 − 𝑦 2 + 𝑧 2
𝑣 = 3𝑥𝑦 + 3𝑦𝑧 + 𝑧
𝑤=?

How must 𝑤 be to satisfy the mass balance equation?

See this solution worked out step by step on YouTube

https://youtu.be/tdPhtzjE5W8 (CC-by Olivier Cleynen)

Note: Unfortunately Olivier made an error in this video: at 4:00 the term 2𝑥 2 is
incorrectly derived into 2𝑥. The final result should have −7𝑥𝑧 instead of −5𝑥𝑧
as the first term. Many thanks to the students who double-checked and reported
the problem!

128
 Playing with the Navier-Stokes equation
An incompressible flow with no gravity has the velocity field defined as
follows:

𝑉⃗ = (𝑎𝑥 + 𝑏)𝑖⃗ + (−𝑎𝑦 + 𝑐𝑥)𝑗⃗

Does a function exist to describe the pressure field, and if so, what is it?

See this solution worked out step by step on YouTube

https://youtu.be/qOfT7kPRl3s (CC-by Olivier Cleynen)

129
130
 Problem sheet 6: Prediction of fluid flows
last edited June 12, 2020
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

Continuity equation for incompressible flow:

⃗ ⋅ 𝑉⃗ = 0
∇ (6/35)

Navier-Stokes equation for incompressible flow:

D𝑉⃗
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
D𝑡
(6/42)

6.1 Quiz
Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.

6.2 Revision questions

For the continuity equation (eq. 6/35), and then for the incompressible Navier-Stokes
equation (eq. 6/42),

6.2.1. Write out the equation in its fully-developed form in three Cartesian coordinates;
6.2.2. State in which flow conditions the equation applies.

Also, in order to revise the notion of total (or substantial) derivative:

non-zero, even though the flow is entirely steady (𝜕/𝜕𝑡 = 0).

6.2.3. Describe a situation in which the total time derivative D/D𝑡 of a property is

6.2.4. Describe a situation in which the the flow is unsteady, although some property
of the fluid, when measured from the point of view of a fluid particle, is not
changing with time.

6.3 Acceleration field

A flow is described with the velocity field 𝑉⃗ = (0,5 + 0,8𝑥)𝑖⃗ + (1,5 − 0,8𝑦)𝑗⃗ (in si units, in
Çengel & al. [25] E4-3

What is the acceleration of a particle positioned at (2; 2; 2) at 𝑡 = 3 s ?

the laboratory frame of reference).

131
 6.4 Volumetric dilatation rate der. Munson & al. [29] 6.4
A flow is described by the following field (in si units):

𝑢 = 𝑥3 + 𝑦2 + 𝑧
𝑣 = 𝑥𝑦 + 𝑦𝑧 + 𝑧 3
𝑤 = −4𝑥 2 𝑧 − 𝑧 2 + 4

What is the volumetric dilatation rate field (the divergent of the velocity field)? What is
the value of this rate at {2;2;2}?

6.5 Incompressibility
Does the vector field 𝑉⃗ = (1,6 + 1,8𝑥)𝑖⃗ + (1,5 − 1,8𝑦)𝑗⃗ satisfy the continuity equation for
Çengel & al. [25] 9-28

two-dimensional incompressible flow?

6.6 Missing components Munson & al. [29] E6.2 + Çengel & al. [25] 9-4
Two flows are described by the following fields:

𝑢1 = 𝑥 2 + 𝑦 2 + 𝑧 2
𝑣1 = 𝑥𝑦 + 𝑦𝑧 + 𝑧
𝑤1 = ?

𝑢2 = 𝑎𝑥 2 + 𝑏𝑦 2 + 𝑐𝑧 2
𝑣2 = ?
𝑤2 = 𝑎𝑥𝑧 + 𝑏𝑦𝑧 2

What must 𝑤1 and 𝑣2 be so that these flows be incompressible?

6.7 Another acceleration field

Given the velocity field 𝑉⃗ = (3𝑡)𝑖⃗ + (𝑥𝑧)𝑗⃗ + (𝑡𝑦 2 )𝑘⃗ (si units), what is the acceleration field,
White [22] E4.1

and what is the value measured at {2;4;6} and 𝑡 = 5 s?

6.8 Vortex Çengel & al. [25] 9.27

A vortex is modeled with the following two-dimensional flow:
𝑦
𝑢=𝐶
𝑥2
+ 𝑦2
𝑥
𝑣 = −𝐶 2
𝑥 + 𝑦2

Verify that this field satisfies the continuity equation for incompressible flow.
132
6.9 Pressure fields Çengel & al. [25] E9-13, White [22] 4.32 & 4.34
We consider the four (separate and independent) incompressible flows below:

𝑉⃗1 = (𝑎𝑥 + 𝑏)𝑖⃗ + (−𝑎𝑦 + 𝑐𝑥)𝑗⃗

𝑉⃗2 = (2𝑦)𝑖⃗ + (8𝑥)𝑗⃗
𝑉⃗3 = (𝑎𝑥 + 𝑏𝑡)𝑖⃗ + (𝑐𝑥 2 + 𝑒𝑦)𝑗⃗
𝑥 𝑦
𝑉⃗4 = 𝑈0 (1 + ) 𝑖⃗ − 𝑈0 𝑗⃗
𝐿 𝐿

The influence of gravity is neglected on the first three fields.

Does a function exist to describe the pressure field of each of these flows, and if so, what
is it?

133
 Answers
6.2 1) Continuity: eq. 6/36 p. 123. Navier-Stokes: see eqs. 6/43, 6/44 and 6/45 p. 125;
2) Read §6.3.1 p. 116 for continuity, and §6/22 p. 119 for Navier-Stokes;
3) and 4) see §6.2.2 p. 112.
D𝑉⃗
D𝑡
= (0,4 + 0,64𝑥)𝑖⃗ + (−1,2 + 0,64𝑦)𝑗⃗. At the probe it takes the value 1,68𝑖⃗ + 0,08𝑗⃗
(length 1,682 m s−2 ).
6.3

⃗ ⋅ 𝑉⃗ = −𝑥 2 + 𝑥 − 𝑧; thus at the probe it takes the value ∇

∇ ⃗ ⃗
6.4 ( ⋅ 𝑉) = −4 s−1 .

Apply equation 6/36 p. 123 to 𝑉⃗ : the answer is yes.

probe

6.5

1) Applying equation 6/36: 𝑤1 = −3𝑥𝑧 − 12 𝑧 2 + 𝑓(𝑥,𝑦,𝑡) ;

2) idem, 𝑣2 = −3𝑎𝑥𝑦 − 𝑏𝑧𝑦 2 + 𝑓(𝑥,𝑧,𝑡) .
6.6

D𝑉⃗ ⃗ At the probe it takes the value 3𝑖⃗+250𝑗⃗+496𝑘.

= (3)𝑖⃗+(3𝑧 +𝑦 2 𝑥)𝑡 𝑗⃗+(𝑦 2 +2𝑥𝑦𝑧𝑡)𝑘. ⃗
D𝑡
6.7

6.8 Apply equation 6/36 to 𝑉⃗ to verify incompressibility.

Note: the constant (initial) value 𝑝𝑖𝑛𝑖 is sometimes implicitly written in the un-
known functions 𝑓 .
6.9

1) 𝑝 = −𝜌 [𝑎𝑏𝑥 + 21 𝑎2 𝑥 2 + 𝑏𝑐𝑦 + 12 𝑎2 𝑦 2 ] + 𝑝𝑖𝑛𝑖 + 𝑓(𝑡) ;

2) 𝑝 = −𝜌 (8𝑥 2 + 8𝑦 2 ) + 𝑝𝑖𝑛𝑖 + 𝑓(𝑡) ; 3) 𝜕𝑥𝜕 ( 𝜕𝑦
𝜕𝑝
) ≠ 𝜕𝑦 ( 𝜕𝑥 ), thus we cannot
𝜕 𝜕𝑝

4) 𝑝 = −𝜌 [ 𝑈𝐿0 (𝑥 + + − 𝑔𝑥 𝑥 − 𝑔𝑦 𝑦 ] + 𝑝𝑖𝑛𝑖 + 𝑓(𝑡) .

describe the pressure with a mathematical function;
2𝐿 )
2
𝑥2 𝑦2
2𝐿

134
 Fluid Dynamics
Chapter 7 – Pipe flows
last edited September 3, 2020
by Olivier Cleynen — https://fluidmech.ninja/

7.1 Motivation 135

7.2 Frictionless flow in pipes 135
7.3 Parameters to quantify losses in pipes 137
7.4 Laminar flow in pipes 137
7.4.1 Laminar flow between plates 137
7.4.2 Laminar flow in pipes 139
7.5 Turbulent flow in pipes 142
7.5.1 When is a pipe flow turbulent? 142
7.5.2 Characteristics of turbulent flow 143
7.5.3 Velocity profile in turbulent pipe flow 144
7.5.4 Pressure losses in turbulent pipe flow 144
7.6 Engineer’s guide to pipe flows 146
7.6.1 Summary so far 146
7.6.2 Choosing laminar or turbulent flow 146
7.6.3 Pressure losses in laminar flow 147
7.6.4 Pressure losses in turbulent flow 147
7.6.5 Calculating pumping and turbining power 148
7.7 Solved problems 148
7.8 Problems 151

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

7.1 Motivation
In this chapter we focus on fluid flow in pipes. This topic allows us to
explore several important phenomena with only very modest mathematical
complexity. In particular, we are trying to answer two questions:

1. What does it take to describe fluid flow in ducts?

2. How can we quantify pressure changes in pipes and the power neces-
sary to overcome them?

7.2 Frictionless flow in pipes

We begin with the simplest possible ducted flow case: a purely hypothetical
fully-inviscid, incompressible, steady fluid flow in a one-dimensional pipe.
Since there are no shear forces, the velocity profile across the duct remains

If the cross-sectional area 𝐴 is changed, then the principle of mass conserva-

uniform (flat) all along the flow, as shown in figure 7.1.

𝑢 = 𝑉:
tion (eqs. 1/24, 3/6) is enough to allow us to compute the change in velocity

𝜌𝑉1 𝐴1 = 𝜌𝑉2 𝐴2 (7/1)

in steady pipe flow.
135
 Figure 7.1: Inviscid fluid flow in a one-dimensional duct. In this purely hypothetical
case, the velocity distribution is uniform across a cross-section of the duct. The

𝑝0 = 𝑝 + 12 𝜌𝑉av.
2 remains constant.
average velocity and pressure change with cross-section area, but the total pressure

Figure CC-0 Olivier Cleynen

This information, in turn allows us to compute the pressure change between

two sections of different areas by using the principle of energy conservation.
We notice that the flow is so simple that the five conditions associated with
Abstruse Goose #517: Daniel
Bernoulli was part of a big the use of the Bernoulli equation (see §2.6 p. 41) are fulfilled: the flow is
familyw steady, incompressible, one-dimensional, has known trajectory, and does not
by an anonymous artist (CC-by-nc)
https://abstrusegoose.com/517 feature friction or energy transfer. A simple application of eq. 2/20 p. 42
between any two points 1 and 2 gives us:
1
𝑝2 − 𝑝1 = − 𝜌 [𝑉22 − 𝑉12 ] − 𝜌𝑔(𝑧2 − 𝑧1 )
2
(7/2)

in steady, incompressible, inviscid pipe flow without heat or work transfer.

Thus, in this kind of simple flow, pressure increases everywhere the velocity
decreases, and vice-versa.

the total or dynamic pressure 𝑝total ≡ 𝑝0 ≡ 𝑝 + 12 𝜌𝑉 2 remains constant:

Another way of writing this equation is by stating that at constant altitude,

𝑝0 = cst. (7/3)
at constant altitude, in laminar inviscid straight pipe flow.

Inviscid flows are nice, but real flows are more interesting. Real flows feature
losses due to viscosity, and with viscous effects, one key assumption of
the Bernoulli equation breaks down. An additional term will appear in the

What does this extra term Δ𝑝loss depend on, and how can we quantify it?
Bernoulli equation, as we have seen in chapter 2 with equation 2/21 p. 43.

This is what the rest of the chapter is about.

136
7.3 Parameters to quantify losses in pipes
Hydraulics is the oldest branch of fluid dynamics, and much of the notation
used to describe pressure losses predates modern applications. The most
widely-used parameters for quantifying losses due to friction in a duct are
the following:

of friction in pipes. Engineers and physicists usually quantify Δ𝑝loss

The pressure loss is the most intuitive way of quantifying the net effect

as a negative number (i.e. Δ𝑝loss ≡ 𝑝2 − 𝑝1 with 𝑝2 < 𝑝1 ). However, in

hydraulics, the historical precedent is to quantify pressure loss with a

to pressure loss as |Δ𝑝loss |.

positive number. To make clear this convention, we will always refer

The elevation loss which we note |Δ𝑙| (in the literature, often noted Δℎ), is
defined as
|Δ𝑝loss |
|Δ𝑙| ≡
𝜌𝑔
(7/4)

by the fluid flow in the duct, and is measured in meters. The reference
It represents the hydrostatic height loss (with a positive number) caused

density 𝜌 in this definition is taken as the density of the fluid, the

density of water, or the density of mercury, depending on cases. We do
not use this definition in this course.

The Darcy friction factor noted 𝑓 is defined as

|Δ𝑝loss |
𝑓 ≡ 𝐿1
𝜌𝑉av.
2
(7/5)
𝐷2

where 𝑉av. is the average flow velocity in the pipe.

In pipe flows of relevance to the engineer, 𝑓 has values between 5 ⋅ 10−5

and 5 ⋅ 10−2 . Those values can be calculated or read from experimental
data, as explained further down.

The loss coefficient noted 𝐾𝐿 is defined as

|Δ𝑝loss |
𝐾𝐿 ≡ 1
𝜌𝑉av.
2
(7/6)
2

Components found in pipe networks, such as bends, filter screens,

proportional to the square of the average velocity. Typically, 𝐾𝐿 values

valves, or junction screens, all result in losses that remain roughly

range between 0,3 (smooth bend) and 2 (partially-closed valve).

7.4 Laminar flow in pipes

7.4.1 Laminar flow between plates

Before we study fluid flow in a circular pipe, let us begin with a simpler case:
flow between two parallel plates. This is a good place to start, because we
can work with Cartesian coordinates, and focus on two dimensions only.
137
 Let us first try a qualitative description of the flow, as displayed in figure 7.2.
Because of the no-slip condition at the walls, the velocity distribution within
any cross-flow section cannot be uniform. Shear occurs, which translates into
a pressure decrease along the flow. The faster the flow, and the higher the
gradient of velocity. Thus, shear within the flow, and the resulting pressure
loss, both increase when the cross-sectional area is decreased.

Figure 7.2: Viscous laminar fluid flow in a one-dimensional pipe. This time, the
no-slip condition at the wall creates a viscosity gradient across the duct cross-section.
This in turn translates into pressure loss. Sudden duct geometry changes such as
represented here would also disturb the flow further, but the effect was neglected
here.
Figure CC-0 Olivier Cleynen

How can we now describe quantitatively the velocity profile and the pressure
loss? We need to clearly sketch the flow we are interested in, which we do in
figure 5.10.
We also need a powerful, extensive mathematical tool to describe the flow:
we turn to the Navier-Stokes equation which we derived in the previous
chapter as eq. 6/42 p. 125:

D𝑉⃗
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
D𝑡
(7/7)

Figure 7.3: Two-dimensional laminar flow between two plates, also called Poiseuille
flow. We already studied this flow case in fig. 5.10 p. 106; this time, we wish to derive
an expression for the velocity distribution.
Figure CC-0 Olivier Cleynen
138
Since we are applying this tool to the simple case of fully-developed, two-
dimensional incompressible fluid flow between two parallel plates (fig. 7.3),
we need only two Cartesian coordinates, so that the vector equation trans-
lates to:
𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢
𝜌 +𝑢 +𝑣 = 𝜌𝑔𝑥 − +𝜇 +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 ] 𝜕𝑥 [ (𝜕𝑥)2 (𝜕𝑦)2 ]
(7/8)

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑝 𝜕 2𝑣 𝜕 2𝑣
𝜌 +𝑢 +𝑣 = 𝜌𝑔𝑦 − +𝜇 +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 ] 𝜕𝑦 [ (𝜕𝑥)2 (𝜕𝑦)2 ]
(7/9)

In this particular flow, we have restricted ourselves to a fully-steady (d/ d𝑡 =

0), horizontal (𝑔 = 𝑔𝑦 ), one-directional flow (𝑣 = 0). When the flow is fully de-
veloped, 𝜕𝑢/𝜕𝑥 = 0 and 𝜕 2 𝑢/(𝜕𝑥)2 = 0, and the system above shrinks down to:

𝜕𝑝 𝜕 2𝑢
0=− +𝜇
𝜕𝑥 [ (𝜕𝑦)2 ]
(7/10)
𝜕𝑝
0 = 𝜌𝑔 −
𝜕𝑦
(7/11)

We only have to integrate equation 7/10 twice with respect to 𝑦 to come to

the velocity profile across two plates separated by a height 2𝐻 :

1 𝜕𝑝
𝑢= (𝑦 2 − 𝐻 2 )
2𝜇 ( 𝜕𝑥 )
(7/12)

Now, the longitudinal pressure gradient 𝜕𝑝/𝜕𝑥 can be evaluated by working

out the volume flow rate ̇ for any given width 𝑍 with one further integration
of equation 7/12:

̇ 2 𝐻 2𝐻 3 𝜕𝑝
= ∫ 𝑢𝑍 d𝑦 = −
𝑍 𝑍 0 3𝜇 ( 𝜕𝑥 )
𝜕𝑝 3 𝜇 ̇
=− 
𝜕𝑥 2 𝑍𝐻3
(7/13)

In this section, the overall process is more important than the result: by
starting with the Navier-Stokes equations, and adding known constraints
that describe the flow of interest, we can predict analytically all of the
characteristics of a laminar flow.

7.4.2 Laminar flow in pipes

We now turn to studying flow in cylindrical pipes, which are widely used;
first considering laminar flow, and then expanding to turbulent flow.
The process is identical to above, only applied to cylindrical instead of
Cartesian coordinates. We focus on the fully-developed laminar flow of a
fluid in a cylindrical pipe without gravity (fig. 7.4).
For this flow, we wish to work out the velocity profile and calculate the
pressure loss related to the flow.

139
 Figure 7.4: A cylindrical coordinate system to study laminar flow in a cylindrical
duct.
Figure CC-0 Olivier Cleynen

We once again start from the Navier-Stokes vector equation, choosing this
time to develop it using cylindrical coordinates:

𝜕𝑣𝑟 𝜕𝑣𝑟 𝑣𝜃 𝜕𝑣𝑟 𝑣𝜃2 𝜕𝑣𝑟

𝜌 + 𝑣𝑟 + − + 𝑣𝑧
[ 𝜕𝑡 𝜕𝑟 𝑟 𝜕𝜃 𝑟 𝜕𝑧 ]
𝜕𝑝 1 𝜕 𝜕𝑣𝑟 𝑣𝑟 1 𝜕 2 𝑣𝑟 2 𝜕𝑣𝜃 𝜕 2 𝑣𝑟
= 𝜌𝑔𝑟 − +𝜇 𝑟 − 2 + 2 − +
𝜕𝑟 [ 𝑟 𝜕𝑟 ( 𝜕𝑟 ) 𝑟 𝑟 (𝜕𝜃)2 𝑟 2 𝜕𝜃 (𝜕𝑧)2 ]

𝜕𝑣𝜃 𝜕𝑣𝜃 𝑣𝜃 𝜕𝑣𝜃 𝑣𝑟 𝑣𝜃 𝜕𝑣𝜃

(7/14)
𝜌 + 𝑣𝑟 + + + 𝑣𝑧
[ 𝜕𝑡 𝜕𝑟 𝑟 𝜕𝜃 𝑟 𝜕𝑧 ]
1 𝜕𝑝 1 𝜕 𝜕𝑣𝜃 𝑣𝜃 1 𝜕 2 𝑣𝜃 2 𝜕𝑣𝑟 𝜕 2 𝑣𝜃
= 𝜌𝑔𝜃 − +𝜇 𝑟 − 2 + 2 + +
𝑟 𝜕𝜃 [ 𝑟 𝜕𝑟 ( 𝜕𝑟 ) 𝑟 𝑟 (𝜕𝜃)2 𝑟 2 𝜕𝜃 (𝜕𝑧)2 ]

𝜕𝑣𝑧 𝜕𝑣𝑧 𝑣𝜃 𝜕𝑣𝑟 𝜕𝑣𝑧

(7/15)
𝜌 + 𝑣𝑟 + + 𝑣𝑧
[ 𝜕𝑡 𝜕𝑟 𝑟 𝜕𝜃 𝜕𝑧 ]
𝜕𝑝 1 𝜕 𝜕𝑣𝑧 1 𝜕 2 𝑣𝑧 𝜕 2 𝑣𝑧
= 𝜌𝑔𝑧 − +𝜇 𝑟 + 2 +
𝜕𝑧 [ 𝑟 𝜕𝑟 ( 𝜕𝑟 ) 𝑟 (𝜕𝜃)2 (𝜕𝑧)2 ]
(7/16)

This mathematical arsenal does not frighten us, for the simplicity of the

we have 𝑔 = 0. Second, we have 𝑣𝑟 = 0 and 𝑣𝜃 = 0 everywhere. Thus, by

flow we are studying allows us to bring in numerous simplifications. First,

continuity, 𝜕𝑣𝑧 /𝜕𝑧 = 0.

Furthermore, since our flow is symmetrical, 𝑣𝑧 is independent from 𝜃. With
these two conditions, the above system shrinks down to:

0=0
XKCD #1230: polar coordinates
by Randall Munroe (CC-by-nc)

0=0
https://xkcd.com/1230
(7/17)

𝜕𝑝 1 𝜕 𝜕𝑣𝑧
(7/18)
0= − +𝜇 𝑟
𝜕𝑧 [ 𝑟 𝜕𝑟 ( 𝜕𝑟 )]
(7/19)

Now, with equation 7/19, we work towards obtaining an expression for 𝑣𝑧 by

integrating twice our expression for 𝜕𝑣𝑧 /𝜕𝑟:

𝜕 𝜕𝑣𝑧 𝑟 𝜕𝑝
𝑟 =
𝜕𝑟 ( 𝜕𝑟 ) 𝜇 𝜕𝑧
𝜕𝑣𝑧 𝑟 2 𝜕𝑝
𝑟 = + 𝑘1
( 𝜕𝑟 ) 2 𝜇 ( 𝜕𝑧 )
𝑟 2 𝜕𝑝
𝑣𝑧 = + 𝑘1 ln 𝑟 + 𝑘2
4 𝜇 ( 𝜕𝑧 )
(7/20)
140
integration constants 𝑘1 and 𝑘2 .
We have to use boundary conditions so as to unburden ourselves from

By setting 𝑣𝑧@𝑟=0 as finite, we deduce that 𝑘1 = 0 (because ln(0) → −∞).

By setting 𝑣𝑧@𝑟=𝑅 = 0 (no-slip condition), we obtain 𝑘2 = − 4𝑅 𝜇 𝜕𝑝
2
𝜕𝑧
.

sion for the velocity profile across a pipe of radius 𝑅 when the flow is laminar:
This simplifies eq. (7/20) and brings us to our objective, an extensive expres-

1 𝜕𝑝
𝑣𝑧 = 𝑢(𝑟) = − (𝑅 2 − 𝑟 2 )
4𝜇 ( 𝜕𝑧 )
(7/21)

This equation is parabolic (fig. 7.5). It tells us that in a pipe of given length 𝐿
and radius 𝑅, a given velocity profile will be achieved which is a function
only of the ratio Δ𝑝/𝜇.

Figure 7.5: The velocity profile across a cylindrical pipe featuring laminar viscous
flow.
Figure CC-0 Olivier Cleynen

We can also express the pressure gradient in the pipe as a function of the
volume flow rate. This is done through integration of velocity with respect
to density and cross-section area. We obtain:

𝜋𝜌𝐷 4 𝜕𝑝
𝑚̇ = −
128 𝜇 ( 𝜕𝑧 )
128 𝜇𝐿𝑚̇
Δ𝑝loss =−
𝜋 𝜌𝐷 4
(7/22)

This equation is interesting in several respects. For a given pipe length 𝐿 and
pressure drop Δ𝑝loss , the volume flow ̇ increases with the power 4 of the
diameter 𝐷. In other words, the volume flow is multiplied by 16 every time
the diameter is doubled.
We also notice that the pipe wall roughness does not appear in equation 7/22.
In a laminar flow, increasing the pipe roughness has no effect on the velocity
distribution in the pipe.
Out of curiosity, we may translate the result in equation 7/22 into a friction
factor equation, with the help of definition 7/5, obtaining
32𝑉av. 𝜇𝐿 𝜇 64
𝑓laminar cylinder flow = = 64 =
𝐿1
𝜌𝑉av.
2 𝐷2 𝜌𝑉av. 𝐷
(7/23)
𝐷2
[Re]𝐷
̇
in which we inserted the average velocity 𝑉av. = 𝜋𝑅
 = − Δ𝑝 𝐷 . 2
2 32𝜇𝐿

141
 7.5 Turbulent flow in pipes

7.5.1 When is a pipe flow turbulent?

It has long been observed that pipe flow can have different regimes. In
some conditions, the flow is unable to remain laminar (one-directional, fully-
steady); it becomes turbulent. Although the flow is steady when it is averaged
over short time period (e.g. a few seconds), it is subject to constant, small-
scale, chaotic and spontaneous velocity field changes in all directions.
Video: very basic visualization
of laminar and turbulent flow In 1883, Osborne Reynolds published the results of a meticulous investigation
regimes in a transparent pipe
by Engineering Fundamentals (styl) into the conditions in which the flow is able, or not, to remain laminar
https://youtu.be/56AyTIhNQBo
(figs. 7.6 and 7.7). He showed that they could be predicted using a single
non-dimensional parameter, later named Reynolds number, which, as we
have seen already with eq. 1/28 p. 23, is expressed as:
𝜌𝑉 𝐿
[Re] ≡
𝜇
(7/24)

In the case of pipe flow, the representative length 𝐿 is conventionally set to

the pipe diameter 𝐷 and the velocity to the cross-section average velocity:
𝜌 𝑉av. 𝐷
[Re]𝐷 ≡
𝜇
(7/25)

where 𝑉av. is the average velocity in the pipe (m s−1 ),

and 𝐷 is the pipe diameter (m).

The occurrence of turbulence is very well documented. The following values

are widely accepted:

• Pipe flow is laminar for [Re]𝐷 . 2 300 ;

Figure 7.6: Illustration published by Reynolds in 1883 showing the installation he set
up to investigate the onset of turbulence. Water flows from a transparent rectangular
tank down into a transparent drain pipe, to the right of the picture. Colored die
is injected at the center of the pipe inlet, allowing for the visualization of the flow
regime.
Image by Osborne Reynolds (1883, public domain)

142
Figure 7.7: Illustration published by Reynolds in 1883 showing two different flow
regimes observed in the installation from fig. 7.6.
Image by Osborne Reynolds (1883, public domain)

• Pipe flow is turbulent for [Re]𝐷 & 4 000.

The significance of the Reynolds number extends far beyond pipe flow; we
shall explore this in chapter 8 (Engineering models).

7.5.2 Characteristics of turbulent flow

This topic is well covered in Tennekes & Lumley [5]

Turbulence is a complex topic which is still not fully described analytically

today. Although it may display steadiness when time-averaged, a turbulent
flow is highly three-dimensional, unsteady, and chaotic in the sense that
the description of its velocity field is carried out with statistical, instead of
Video: turbulence for those who
analytic, methods. don’t have time to read chapter 9
by Y:Veritasium (styl)
In the scope of our study of fluid dynamics, the most important characteristics https://youtu.be/5zI9sG3pjVU

associated with turbulence are the following:

• A strong increase in mass and energy transfer within the flow. Slow and
rapid fluid particles have much more interaction (especially momentum
transfer) than within laminar flow;

• A strong increase in losses due to friction (typically by a factor 2). The

increase in momentum exchange within the flow creates strong dissipa-
tion through viscous effects, and thus transfer (as heat) of macroscopic
forms of energy (kinetic and pressure energy) into microscopic forms
(internal energy, translating as temperature);
143
 • Internal flow movements appear to be chaotic (though not merely
random, as would be white noise), and we do not have mathematical
tools to describe them analytically.

Consequently, solving a turbulent flow requires taking account of flow in all

three dimensions even for one-directional flow! We will come back to this
topic in chapter 9 (Dealing with turbulence).

7.5.3 Velocity profile in turbulent pipe flow

pipes, we split each velocity component 𝑣𝑖 in two parts, a time-averaged

In order to deal with the vastly-increased complexity of turbulent flow in

component 𝑣𝑖 and an instantaneous fluctuation 𝑣𝑖′ :

𝑣𝑟 = 𝑣𝑟 + 𝑣𝑟′
𝑣𝜃 = 𝑣𝜃 + 𝑣𝜃′
𝑣𝑧 = 𝑣𝑧 + 𝑣𝑧′
Video: highly-turbulent flow ex-
iting the flood discharge ducts of

In our case, 𝑣𝑟 and 𝑣𝜃 are both zero, but the fluctuations 𝑣𝑟′ and 𝑣𝜃′ are not, and
the Tarbela dam in northeastern
Pakistan

will cause 𝑣𝑧 to differ from the laminar flow case. The extent of turbulence is
by Y:Beauty Of Pakistan (styl)
https://youtu.be/13tBWzKajqw

often measured with the concept of turbulence intensity 𝐼 :

[ 𝑣𝑖 ]
1
′2 2
𝐼 ≡
𝑣𝑖
(7/26)

Regrettably, we have not found a general analytical solution to turbulent pipe

flow — note that if we did, it would likely exhibit complexity in proportion
to that of such flows. A widely-accepted average velocity profile constructed
from experimental observations is:

𝑟 7
𝑢 (𝑟) = 𝑣𝑧 = 𝑣𝑧 max (1 −
1

𝑅)
(7/27)

While it closely and neatly matches experimental observations, this model

is nowhere as potent as an analytical one and must be seen only as an
approximation. For example, it does not allow us to predict internal energy

wall shear stress (because it yields (𝜕𝑢/𝜕𝑟)𝑟=𝑅 = ∞, which is not physical).

dissipation rates (because it describes only time-averaged velocity), or even

The following points summarize the most important characteristics of turbu-

lent velocity profiles:

• They continuously fluctuate in time and we have no means to predict

them extensively;

• They are much “flatter” than laminar profiles (fig. 7.8);

• They depend on the wall roughness;

• They result in shear and dissipation rates that are markedly higher
than laminar profiles.

144
Figure 7.8: Velocity profiles for laminar (A), and turbulent (B and C) flows in a
cylindrical pipe. B represents the time-averaged velocity distribution, while C shows
several arbitrary instantaneous distributions (blurred) as well as their average in
time. Turbulent flow in a pipe also features velocities in the radial and angular
directions, which are not shown here.
Figure CC-0 Olivier Cleynen

7.5.4 Pressure losses in turbulent pipe flow

Losses caused by turbulent flow depend on the wall roughness 𝜖 and on the
diameter-based Reynolds number [Re]𝐷 .

friction factor 𝑓 anymore. Several empirical models can be built to obtain 𝑓 ,

For lack of an analytical solution, we are not able to predict the value of the

the most important of which is known as the Colebrook equation expressed

as:

1 1 𝜖 2,51
√ = −2 log + √
𝑓 ( 3,7 𝐷 [Re]𝐷 𝑓 )
(7/28)

The structure of this equation makes it inconvenient to solve for 𝑓 . To

circumvent this difficulty, equation 7/28 can be solved graphically on the

values for 𝑓 (and thus predict the pressure losses Δ𝑝loss ) if we know the
Moody diagram, fig. 7.9. This classic document allows us to obtain numerical

diameter-based Reynolds number [Re]𝐷 and the relative roughness 𝜖/𝐷.

145
 Figure 7.9: A Moody diagram, which presents values for 𝑓 measured experimentally,
as a function of the diameter-based Reynolds number [Re]𝐷 , for different values of
the relative roughness 𝜖/𝐷. This figure is reproduced with a larger scale as figure 7.11
p. 152.
Diagram CC-by-sa S Beck and R Collins, University of Sheffield

7.6 Engineer’s guide to pipe flows

7.6.1 Summary so far

The world of pipe flows is a good playground for us to experiment with
practical fluid mechanics. The main lessons learned in this chapter are as
follows:
• From an engineering point of view, pipe flows are one-dimensional.
We push a volume flow of fluid in on one side and expect to receive it

doing so is quantified with a single value, Δ𝑝loss .

on the other, perhaps with different properties. The integral “cost” of

• When the flow is laminar, we can solve for the flow in the pipe. For
this, we write the Navier-Stokes equations and apply known boundary

us a wealth of information about the flow, including the Δ𝑝loss .

conditions. We obtain the velocity everywhere in the pipe, which gives

• When the flow is turbulent, this analysis method does not work any-
more. Even though the conditions are in principle simple, the flow is
mesmerizingly complex. We resort to building models based on time-
averaged measurement data. Those models work for a wide number of

dimensionalized. We obtain the desired Δ𝑝loss with a single diagram

pipes and conditions because they relate parameters which are non-

reading and a simple algebraic manipulation.

These broad trends apply to many other sub-areas of fluid mechanics.

7.6.2 Choosing laminar or turbulent flow

When designing a pipe system, an engineer may have the opportunity to
146 create laminar or turbulent flow. Inserting an expression for mass flow
𝑚̇ = 𝜌𝑉av. (𝜋/4)𝐷 2 into the definition of the Reynolds number, we obtain, for
pipe flow:
𝑚̇ 4
[Re] =
𝐷 𝜋𝜇
(7/29)

This equation is telling: for a given fluid (𝜇) and a given mass flow (𝑚), ̇ the

increase the diameter 𝐷. Doing so also increases installation costs; therefore,

only way to ensure that the flow is laminar (with low Reynolds number) is to

there is a balance to strike between initial installation costs (increasing with

diameter) and operating costs due to pressure losses (which decrease as the
diameter increases). In practice, except for cases where very small mass flows
and velocities are involved (e.g. medical fluid flows), most piping installations
feature turbulent flow.

7.6.3 Pressure losses in laminar flow

The pressure losses in laminar flow are summed up with equation 7/22 p. 141,
which we repeat here with the most important terms positioned first:
𝑚̇ 128𝜇
Δ𝑝loss = −𝐿
𝐷 4 𝜋𝜌
(7/30)

for laminar pipe flow.

increase linearly with mass flow 𝑚, ̇ and with the power −4 (!) of the diame-
Again, it is visible here that in laminar pipe flow, losses per unit pipe length

ter 𝐷.

7.6.4 Pressure losses in turbulent flow

The picture for losses in turbulent flow is harder to draw. Re-arranging the
definition 7/5 p. 137 to include the mass flow, we obtain:

𝑚̇ 2 8
Δ𝑝loss = −𝐿 𝑓 2
𝐷 5 𝜋 𝜌
(7/31)

for turbulent pipe flow.

The factor 𝑓 in this equation generally varies according to the Reynolds

At very turbulent regimes (in the upper right area of the diagram), 𝑓 becomes
number and to the roughness of the pipe, as described in the Moody diagram.

independent of [Re]𝐷 and proportional to the relative roughness, so that we

can write:
𝑚̇ 2 𝜖 8
Δ𝑝loss = −𝐿 ( 𝑐1 + 𝑐2 ) 2
𝐷 5 𝐷 𝜋 𝜌
(7/32)

where 𝑐1 and 𝑐2 are approximately constant.

for turbulent pipe flow in very turbulent regimes,

pipe flow increase approximately with the square of mass flow 𝑚̇ and the
Based on this equation, we can see that pressure losses in highly-turbulent

power −5 of the diameter 𝐷. In between this regime and the laminar regime,
a variety of intermediary states are quantified using the Moody diagram.

147
 7.6.5 Calculating pumping and turbining power
The tools above are all we need to calculate, given the geometry of an
installation, how much pumping or turbining power is involved in moving a
given mass flow of liquid though a pipe system. The steps are as follows:

1. Calculate the hydrostatic pressure drop across the device.

This is done by imagining that there is no flow, and calculating the
static pressure which would then exert on each side of the device, using
equation 4/15 p. 80.

2. Calculate the pressure losses due to friction.

diagram to read the corresponding friction factor 𝑓 , and calculating the

This is done by calculating the Reynolds number, using the Moody

corresponding Δ𝑝loss using the definition 7/5. Care must be taken with
the sign of Δ𝑝loss , which is always negative by definition, but very often
expressed as a positive number in the literature.

calculated using their 𝐾𝐿 values and the definition 7/6.

Pressure losses induced by bends, junctions and obstacles are likewise

The complete pressure difference across the device is Δ𝑝device = −Δ𝑝losses +

3. Summing up the pressure differences.

Δ𝑝hydrostatic . Pressure losses due to friction are always negative, and

The power is recovered using equation 1/22 p. 19, 𝑃̇ device = Δ𝑝device 𝑚/𝜌.
̇
hydrostatic pressure differences may have either sign.

If this power is negative, the liquid is losing energy, and the device is
acting as a turbine. If the power is positive, the liquid is gaining energy,
and the device is acting as a pump.

7.7 Solved problems

Hydrostatic pressure on a turbine

A turbine is installed as shown above. What is the hydrostatic pressure

difference available to the turbine?

148
 See this solution worked out step by step on YouTube
https://youtu.be/3lyKby0fS-8 (CC-by Olivier Cleynen)

Pressure loss in a pipe

In the piping installation from the previous example, water at 20 °C is
flowing with a volume flow of 800 L s−1 .
The pipe has roughness 𝜖 = 0,25 mm and a diameter 𝐷 = 1,1 m. The bends
each induce a loss coefficient 𝐾𝐿 = 0,75.
What is the pressure drop due to friction losses in the pipe?

See this solution worked out step by step on YouTube

https://youtu.be/Tp6a_50uqUc (CC-by Olivier Cleynen)

Turbining power
In the piping installation from the previous examples, what is the power
made available to the turbine?

See this solution worked out step by step on YouTube

https://youtu.be/BOew8INdQ54 (CC-by Olivier Cleynen)

Pressure distribution in a pipe

In the piping installation from the previous examples, what is the pres-
sure distribution along the pipe?

See this solution worked out step by step on YouTube

https://youtu.be/q1xOmWaYZLg (CC-by Olivier Cleynen)

149
150
 Problem sheet 7: Pipe flows
last edited June 30, 2021
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

In cylindrical pipe flow, we assume the flow is always laminar for [Re]𝐷 . 2 300, and
always turbulent for [Re]𝐷 & 4 000. The Darcy friction factor 𝑓 is defined as:
|Δ𝑝loss |
𝑓 ≡ 𝐿1
𝜌𝑉av.
2
(7/5)
𝐷2

The loss coefficient 𝐾𝐿 is defined as:

|Δ𝑝loss |
𝐾𝐿 ≡ 1
𝜌𝑉av.
2
(7/6)
2

Viscosities of various fluids are given in fig. 7.10. Pressure losses in cylindrical pipes
can be calculated with the help of the Moody diagram presented in fig. 7.11 p. 152.

2×10−2 2.4×10−5

10−2 2.2×10−5
9×10−3
8×10−3
7×10−3 ⟵ Crude Oil
6×10−3 Air ⟶
5×10−3
2×10−5
4×10−3
3×10−3
Viscosity 𝜇 of liquids in Pa s

Viscosity 𝜇 of gases in Pa s

2×10−3 1.8×10−5
CO2 ⟶

10−3 1.6×10−5
9×10−4
8×10−4
7×10−4
6×10−4
5×10−4
1.4×10−5
4×10−4 ⟵ Water
3×10−4

2×10−4 1.2×10−5

10−4 10−5
−20 0 20 40 60 80 100 120
Temperature 𝑇 in degree Celsius ( C)
◦

Figure 7.10: The viscosity of four fluids (crude oil, water, air, and C02) as a function of temperature.
The scale for liquids is logarithmic and displayed on the left; the scale for gases is linear and 151
displayed on the right.
Figure reproduced from figure 5.6 p. 99; CC-by by Arjun Neyyathala & Olivier Cleynen
 Figure 7.11: A Moody diagram, which presents values for 𝑓 measured experimentally, as a
function of the diameter-based Reynolds number [Re]𝐷 , for different relative roughness values.
Diagram CC-by-sa S Beck and R Collins, University of Sheffield
152
7.1 Reading quiz
Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.

7.2 Revision questions non-examinable

The Moody diagram (fig. 7.11 p. 152) is simple to use, yet it takes practice to understand
it fully. . . here are three questions to guide your exploration. They can perhaps be
answered as you work through the other examples.
7.2.1. Why is there no zero on the diagram?
7.2.2. Why are the curves sloped downwards — should friction losses not instead

̇ but
increase with increasing Reynolds number?
7.2.3. Why can the pressure losses Δ𝑝losses be calculated given the volume flow ,
not the other way around?

7.3 Air flow in a small pipe Munson & al. [29] E8.5

through a 10 cm-long cylindrical pipe with a 4 mm diameter, roughness 0,0025 mm, at an

A machine designed to assemble micro-components uses an air jet. This air is driven

average speed of 50 m s−1 .

The inlet air pressure and temperature are 1,2 bar and 60 °C; the viscosity of air is
quantified in fig. 7.10 p. 151.
7.3.1. What is the pressure loss caused by the flow through the pipe?
7.3.2. What is the maximum average flow speed for which the flow would remain
laminar?
7.3.3. What would this speed be if the pipe diameter was 4 cm instead of 4 mm?

7.4 Water piping

total length of the pipe is 10 km, its diameter is 0,5 m, and its roughness is 𝜖 = 0,5 mm. It
A long pipe is installed to carry water from one large reservoir to another (fig. 7.16). The

The pump must be powerful enough to push 1 m3 s−1 of water at 20 °C.

must climb over a hill, so that the altitude changes along with distance.

Figure 7.10 p. 151 quantifies the viscosity of various fluids, and fig. 7.11 p. 152 quantifies
losses in cylindrical pipes.

7.4.1. Will the flow in the water pipe be turbulent?

7.4.2. What is the pressure loss caused by the flow through the pipe?
7.4.3. What is the pumping power required to meet the design requirements?
7.4.4. What would be the power required for the same volume flow if the pipe diameter
was doubled?
153
 Figure 7.12: Layout of the water pipe. For clarity, the vertical scale is greatly exaggerated. The
diameter of the pipe is also exaggerated.
Figure CC-0 Olivier Cleynen

7.5 Design of a water piping system

You start your career as a junior engineer in a company that designs piping and pumping

Your first assignment is to choose the dimensions of a system which should carry 3 m3 h−1
systems.

of water (10−3 Pa s) across a horizontal distance of 1 km. Fresh from reading through
chapter 7, you design the system to feature laminar flow only.
You begin with a revision of the relevant theory, starting, of course, from the Navier-
Stokes equations and a simple diagram (fig. 7.13), obtaining the velocity distribution for
laminar flow in a circular pipe:

1 𝜕𝑝
𝑣𝑧 = 𝑢(𝑟) = − (𝑅 2 − 𝑟 2 )
4𝜇 ( 𝜕𝑧 )
(7/21)

Figure 7.13: A cylindrical coordinate system to study laminar flow in a horizontal cylindrical
duct.
Figure CC-0 Olivier Cleynen

7.5.1. What is the minimum pipe diameter for which the flow will remain laminar?

pipe with laminar flow is expressed as a function of the volume flow ̇ and the
7.5.2. Starting from equation 7/21, show that the pressure loss per unit length in the

diameter 𝐷 as:
128 𝜇𝐿𝑚̇
Δ𝑝loss = −
𝜋 𝜌𝐷 4
(7/22)

7.5.3. With the diameter chosen above, what is the pressure loss in the pipe?
7.5.4. What is the pumping power required?

With those results in hand, you turn to your colleagues — but with a smile, they suggest
154 you try a design with turbulent flow instead.
 7.5.5. What would be the pressure loss if an 8 cm-diameter plastic pipe was used?

7.5.6. What would then be the pumping power required?

7.5.7. What is one advantage of using a pipe with smaller diameter? (briefly justify
your answer, e.g. in 30 words or less)

7.6 Major oil pipeline

Strong from your experience working through problem 7.5, you join the team in charge

Your design must safely carry 700 thousand barrels of oil (110 000 m3 ) per day along a
of designing one very large oil pipeline system (fig. 7.14).

length of 1 200 km. The crude oil has density 900 kg m−3 and its viscosity is quantified in
fig. 7.10 p. 151. The average temperature of the oil during the transit is 60 °C.
The landscape is flat for most of the journey, with a 200 km-wide mountain range in the
middle that reaches 1 400 m altitude.
Your team selects a cylindrical, smooth steel duct with 1,22 m diameter, average roughness
𝜖 = 0,15 mm. Because the pipeline passes through ecologically fragile areas, as well as a
seismically-active region, you decide to never exceed 200 psi (13,8 bar) of gauge pressure

you decide to never reach below 0,8 bar of absolute pressure in the pipeline.
in the pipeline. To prevent oil cavitation (a change of state with destructive consequences),

7.6.1. How much time does the average oil particle need to travel across the line?

7.6.2. How much pumping power is required in total?

7.6.3. How far apart should the pumping stations be laid out in the flat sections of the
pipeline?

7.6.4. How far apart should the pumping stations be laid out in the ascending section
of the pipeline?

7.6.5. Propose a pumping station arrangement, and calculate the power required for
each pump.

Figure 7.14: The Trans-Alaska Pipeline System, which inspired this problem. It was built in the
1970s, at tremendous financial, political and social cost.
Photo CC-by-sa by Luca Galuzzi – www.galuzzi.it

155
 Before you start building the pipeline, the operator would like to know how the system
would perform at half-capacity (i.e. with half the volume flow).

7.6.6. If none of the other input data changes, what is the new pumping power?

7.6.7. Propose one reason why in practice, the pumping power may be higher than
you just calculated (briefly justify your answer, e.g. in 30 words or less).

7.7 Pump with pipe expansion

A pump is used to carry a volume flow of 200 L s−1 from one large water reservoir to
another (fig. 7.15). The altitude of the water surface in both reservoirs is the same.

Figure 7.15: Layout of the water pipe. For clarity, the diameter of the pipe and the vertical scale
are exaggerated.
Figure CC-0 Olivier Cleynen

The pipe connecting the reservoirs is made of concrete (𝜖 = 0,25 mm); it has a diameter
of 50 cm on the first half, and 100 cm on the second half. In the middle, the conical
expansion element induces a loss coefficient of 0,8. At the outlet (at point D), the pressure

The inlet is 14 m below the surface. The total pipe length is 400 m; the altitude change
is approximately equal to the corresponding hydrostatic pressure in the outlet tank.

between inlet and outlet is 12 m.

7.7.1. Represent qualitatively (that is to say, showing the main trends, but without
displaying accurate values) the water pressure as a function of pipe distance,
when the pump is turned off.

7.7.2. On the same graph, represent qualitatively the water pressure when the pump
is switched on.

7.7.3. What is the water pressure at points A, B, C and D?

7.8 Piping and power of a water turbine from 2018-07 final examination
A water turbine is installed between two reservoirs in order to extract power from the

The pipe is made of reinforced concrete, with a total length 𝐿 = 0,8 km, a diameter 𝐷 =
flow of water. The water is guided to the turbine through a pipe, as shown in figure 7.16.

𝑑 = 1,2 m, and an interior surface roughness 𝜖 = 6 mm. The pipe has altitude variations
along its length, as indicated in figure 7.16. The turbine is designed to handle 5 000 L s−1
156 of water at 20 °C.
Figure 7.16: Layout of the water pipe. For clarity, the vertical scale and the diameter of the pipe
are greatly exaggerated.

7.8.1. On a diagram, represent qualitatively (i.e. without numerical data) the pressure
distribution along the length of the pipe.

7.8.2. What is the pressure drop due to friction losses generated by the water flow in
the pipe?

7.8.3. What is the power developed by the turbine?

7.8.4. What would be the new power developed by the turbine if the volume flow was
divided by two?

7.9 Politically incorrect fluid mechanics non-examinable

In spite of the advice of their instructor, a group of students attempts to apply fluid
mechanics to incommendable activities. Their objective is to construct a drinking straw
piping system that can mix a drink of vodka and tonic water in the correct proportions
(fig. 7.17). They use a “Strawz” kit of connected drinking straws, two bottles, and a glass
full of ice and liquid water to cool the mix.
For simplicity, the following information is assumed about the setup:

Figure 7.17: Conceptual sketch of a student experiment.

Figure CC-0 Olivier Cleynen

157
 Viscosity in centipoise
0 1,005
Percentage of alcohol by weight

10 1,538
20 2,183
30 2,71
40 2,91
50 2,87
60 2,67
70 2,37
80 2,008
90 1,61
100 1,2

Table 7.1: Viscosity of a mix of ethanol and water at 20 °C.

data from Bingham, Trans. Chem. Soc, 1902, 81, 179.

• Vodka is modeled as 40 % pure alcohol (ethanol) with 60 % water by volume;

• Ethanol density is 0,8 kg m−3 ;

• Tonic water is modeled as pure water;

• The viscosity of water and alcohol mixes is described in table 7.1 (use the nearest
relevant value);

• The pipe bends induce a loss coefficient factor 𝐾𝐿bend = 0,5 each;

• The pipe T-junction induces a loss coefficient factor 𝐾𝐿 = 0,3 in the line direction
and 1 in the branching flow;

• The pipe has inner diameter 𝐷 = 3 mm and roughness 𝜂 = 0,0025 mm.

The students wish to obtain the correct mix: one quarter vodka, three quarters tonic
water. For given levels of liquid in the bottles, is there a straw pipe network configuration
that will yield the correct mix, and if so, what is it?

158
Answers
1) Calculating inlet density with the perfect gas model, [Re]𝐷 = 14 263 (turbulent),
a Moody diagram read gives 𝑓 ≈ 0,029, so Δ𝑝friction = −1 292,6 Pa = −0,0129 bar.
7.3

7.4 |Δ𝑝alt. | = 𝜌𝑔(26 − 8 + 5 − 7) = 1,57 bar and |Δ𝑝friction | = 51,87 bar : 𝑊̇ pump = 5,345 MW.

7.5 1) 𝐷min = 0,46 m

2) Follow the process used p. 139 to go from eq. 7/10 to eq. 7/13: the only difference

3) |Δ𝑝|loss = 0,75 Pa
is the use of cylindrical (instead of rectangular) coordinates.

4) 𝑊̇ = 0,63 mW
5) |Δ𝑝|loss 2 = 4,82 kPa (8 000 times more)
6) 𝑊̇ 2 = 4,2 W

2) 𝑊̇ total = 10,02 MW
7.6 1) Approximately 12 days 18 hours

3) Δ𝐿flat terrain = 213 km

4) Δ𝐿ascending terrain = 11,96 km
7) Hint: in normal operation, what happens with the 10 MW of power — in which
form is this energy converted?

3) Δ𝑝𝑓 A→B = −3 735 Pa, Δ𝑝B→C = +71 Pa (the sum of losses due to friction and gains
due to decrease in kinetic energy), Δ𝑝𝑓 C→D = −117 Pa. Add hydrostatic pressure
7.7

changes, working backwards from D to A, to obtain pressure in all four points.

2) Δ𝑝friction losses = −2,0196 bar

3) 𝑊̇ turbine = −1,0503 MW
7.8

4) The power will decrease by 14 %

7.9 The author cannot remember which exercise you are referring to.

159
160
 Fluid Dynamics
Chapter 8 – Engineering models
last edited April 3, 2021
by Olivier Cleynen — https://fluidmech.ninja/

8.1 Motivation 161

8.2 Comparing influences: the weighted momentum balance161
8.2.1 Principle 161
8.2.2 The non-dimensional Navier-Stokes equation 162
8.2.3 The flow parameters of Navier-Stokes 164
8.2.4 Flow parameters obtained as force ratios 166
8.2.5 The Reynolds number in practice 166
8.3 Making models 167
8.4 Comparing results: coefficients 169
8.4.1 Principle 169
8.4.2 Force coefficients 169
8.4.3 Power coefficient, and other coefficients 170
8.4.4 Non-dimensionalizing all the problems 171
8.5 Solved problems 172
8.6 Problems 173

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

8.1 Motivation
In this chapter, we develop tools to analyze scale effects in fluid mechanics.
This study should allow us to answer two questions:

• How can we adequately reproduce a flow in an enlarged or reduced

version?
• How do forces and powers change when the flow is scaled?

8.2 Comparing influences:

the weighted momentum balance

8.2.1 Principle
In many practical engineering situations which involve fluid flow, investi-
gating the flow in the real application is impossible or impractical. In those
cases it is common practice to build scaled-up or scaled-down versions of
the flow in a laboratory. This brings up the question: how should the model
flow properties be adapted to represent the original flow? For example, if Video: when their movement is
filmed and then viewed sped-up,
the model is half as small as the original, should the velocity be halved? Or fog banks can appear to flow like
perhaps doubled? water. What time-lapse rate is
required to achieve physical sim-
ilarity?
The answer to this problem is as follows: two flows are dynamically similar by Simon Christen (stvl)
(i.e. representative of one another) when their flow parameters are the same. https://vimeo.com/69445362

In order to understand what this means, we will need to look back at the
Navier-Stokes equation (the momentum balance equation we derived in
chapter 6), re-writing it in a non-dimensional form. Onwards! 161
 Advice from an expert
Making models is useful when you have a large
machine, say, a jet airliner or a submarine,
and you want to study different configurations
without building a new full-scale machine each
time. But there are other applications: imagine
trying to study the flow of molten metal in a
furnace, the flow of blood in a beating heart,
or the flow of air around a mosquito’s wings.
Making models and understanding scale effects will allow you to investi-
gate fluid flow in all kind of inaccessible locations. That makes the math
worth the effort!

8.2.2 The non-dimensional Navier-Stokes equation

Here, we want to obtain an expression for the Navier-Stokes equation for
incompressible flow which allows for an easy comparison of its constituents.
We start with the original equation, which we derived in chapter 6 as equa-
tion 6/42 p. 125:

𝜕 𝑉⃗
𝜌 + 𝜌(𝑉⃗ ⋅ ∇
⃗ )𝑉⃗ = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
𝜕𝑡
(8/1)

this equation. The principle is to express each vector 𝐴⃗ as the multiple of its
What we would now like to do is separate the geometry from the scalars in

length 𝐴 and a non-dimensional vector 𝐴⃗∗ , which has the same direction as
𝐴⃗ but only unit length.

terms, starting with non-dimensional time 𝑡 ∗ , defined as time 𝑡 multiplied by

In order to achieve this, we introduce a series of non-dimensional physical

the frequency 𝑓 (in Hertz) at which the flow repeats itself:

𝑡∗ ≡ 𝑓 𝑡 (8/2)

A flow with a very high frequency is highly unsteady, and the changes in
time of the velocity field will be relatively important. On the other hand, the

we observe the flow, non-dimensional time 𝑡 ∗ progresses from 0 to 1, after

lower the frequency, and the closer the flow is to being steady. In all cases, as

We then introduce non-dimensional velocity 𝑉⃗ ∗ , a unit vector field equal to

which the solution is repeated.

the velocity vector field divided by its own (scalar field) length 𝑉 :

𝑉⃗
𝑉⃗ ∗ ≡
𝑉
(8/3)

Pressure 𝑝 is non-dimensionalized differently, since in fluid mechanics, it is

⃗ 𝑝 can be replaced by ∇
the velocity field. For example, in eq. 8/1 ∇ ⃗ (𝑝 − 𝑝∞ )
the pressure changes across a field, not their absolute value, that influence

(in which 𝑝∞ can be any constant faraway pressure). Now, the pressure field
𝑝 −𝑝∞ is non-dimensionalized by dividing it by a reference pressure difference

162
 𝑝0 − 𝑝∞ , obtaining:
𝑝 − 𝑝∞
𝑝∗ ≡
𝑝0 − 𝑝∞
(8/4)

If 𝑝0 is taken to be the maximum pressure in the studied field, then 𝑝 ∗ is a

Non-dimensional gravity 𝑔⃗∗ is simply a unit vector equal to the gravity

scalar field whose values can only vary between 0 and 1.

vector 𝑔⃗ divided by its own length 𝑔:

𝑔⃗
𝑔⃗∗ ≡
𝑔
(8/5)

⃗ ∗,
And finally, we define a new operator, the non-dimensional del ∇

⃗∗ ≡ 𝐿 ∇
∇ ⃗ (8/6)

which ensures that vector fields obtained with a non-dimensional gradient,

by a reference length 𝐿.
and the scalar fields obtained with a non-dimensional divergent, are “scaled”

These new terms allow us to replace the constituents of equation 8/1 each by
a non-dimensional “unit” term multiplied by a scalar term representing its
length or value:
𝑡∗
𝑡=
𝑓
𝑉⃗ = 𝑉 𝑉⃗ ∗
𝑝 − 𝑝∞ = 𝑝 ∗ (𝑝0 − 𝑝∞ )
𝑔⃗ = 𝑔 𝑔⃗∗
1 ∗
⃗= ∇
∇ ⃗
𝐿

Now, inserting these in equation 8/1, and re-arranging, we obtain:

𝜕 ∗ 1 ⃗∗ 1 ∗ ∗ ⃗ ∗2
∇
𝜌 1 ∗ (𝑉 𝑉 ) + 𝜌 (𝑉 𝑉 ⋅ ∇ ) (𝑉 𝑉 ) = 𝜌𝑔 𝑔⃗ − ∇ [𝑝 (𝑝0 − 𝑝∞ ) + 𝑝∞ ] + 𝜇 2 (𝑉 𝑉⃗ ∗ )
⃗ ∗ ⃗ ⃗ ∗ ∗ ⃗
𝜕𝑓 𝑡 𝐿 𝐿 𝐿
𝜕 𝑉⃗ ∗ 1 1 ∗ ∗ 1 ∗2 ⃗ ∗
𝜌𝑉 𝑓 + 𝜌𝑉 𝑉 (𝑉⃗ ∗ ⋅ ∇ ⃗ ∗ 𝑉⃗ ∗ = 𝜌𝑔 𝑔⃗∗ − ∇
)
⃗ [𝑝 (𝑝0 − 𝑝∞ )] + 𝜇𝑉 ∇ ⃗ 𝑉
𝜕𝑡 ∗ 𝐿 𝐿 𝐿2
𝜕 𝑉⃗ ∗ 𝜌𝑉 2 𝑝 − 𝑝∞ 𝜇𝑉
[𝜌𝑉 𝑓 ] + (𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ = [𝜌𝑔] 𝑔⃗∗ − 0
) [ ]
⃗ ∗𝑝∗ +
∇ ⃗ ∗2 𝑉⃗ ∗
∇
𝜕𝑡 ∗ [ 𝐿 ] 𝐿 [ 𝐿 2 ]
(8/7)

In equation 8/7, the terms in brackets each appear in front of non-dimen-

i.e. kg m−2 s−2 . Multiplying each by 𝜌𝑉𝐿 2 (of dimension m2 s2 kg−1 ), we obtain
sional (unit) vectors. These bracketed terms all have the same dimension,

a purely non-dimensional equation:

Video: Figuring out the non-di-

𝑓 𝐿 𝜕 𝑉⃗ ∗ 𝑔𝐿 𝑝0 − 𝑝∞ 𝜇
mensional Navier-Stokes equa-

+ [1] (𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ =
) 𝑔⃗∗ − ⃗ ∗𝑝∗ +
∇ ⃗ ∗2 𝑉⃗ ∗
∇
tion

[ 𝑉 ] 𝜕𝑡 ∗ [𝑉 ]
2 [ 𝜌𝑉 ]2 [ 𝜌𝑉 𝐿 ]
by Olivier Cleynen (CC-by)
https://youtu.be/KD3MIqUgN1A

(8/8)

163
 Equation 8/8 does not bring any information on top of the original incom-
pressible Navier-Stokes equation (eq. 8/1). Instead, it merely separates it into
two distinct kinds of components. The first, in square brackets, are a scalar
fields (purely numbers), which indicate the magnitude of the acceleration
field. The others are unit vector fields (fields of oscillating vectors, all of
length 1, and noted with stars), which represent the geometry (direction) of
the acceleration field. In this form, we can more easily observe and quantify
the weight of the different terms relative to one another. At this point, it is
time to introduce the notion of flow parameters.

8.2.3 The flow parameters of Navier-Stokes

From here on, in equation 8/8, we convene to call the terms written in
brackets flow parameters, and label them with the following definitions:
𝑓 𝐿
[St] ≡
𝑉
𝑝0 − 𝑝∞
(8/9)

[Eu] ≡
𝜌 𝑉2
(8/10)
𝑉
[Fr] ≡ √
𝑔𝐿
(8/11)

𝜌𝑉 𝐿
[Re] ≡
𝜇
(8/12)

This allows us to re-write eq. 8/8 as:

𝜕 𝑉⃗ ∗ ⃗∗ ⋅ ∇ 1 1
+ [1] 𝑉 ⃗ ∗ 𝑉⃗ ∗ = ⃗
𝑔 − ⃗ ∗𝑝∗ +
∇ ⃗ ∗2 𝑉⃗ ∗
∇
[ ] [ Re ]
∗
𝜕𝑡 ∗ 2
[St] [Eu]
Fr

And finally, we arrive at the simple, elegant and formidable non-dimensional

incompressible Navier-Stokes equation expressed with flow parameters:

𝜕 𝑉⃗ ∗ 1 1 ∗2 ⃗ ∗
+ [1] 𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ = ⃗ ∗𝑝∗ +
⃗∗ − [Eu] ∇
2 𝑔
⃗ 𝑉
∇
𝜕𝑡 ∗
[St]
[Fr] [Re]
(8/13)

Equation 8/13 is an incredibly useful tool in the study of fluid mechanics, for
two reasons:

1. It allows us to quantify the relative weight of the different terms in

XKCD #762: comparing things
a given flow.
by Randall Munroe (CC-by-nc) In this way, it serves as a compass, helping us determine which terms
https://xkcd.com/762
can safely be neglected in our attempts to find a particular flow solution,
merely by quantifying the four parameters [St], [Eu], [Fr], and [Re].

2. It allows us to obtain dynamic similarity between two flows at

different scales.
In order that a fluid flow be representative of another flow (for example

tunnel), it must generate the same vector field 𝑉⃗ ∗ . In order to do this,

in order to simulate airflow around an aircraft with a model in a wind

we must generate an incoming flow with the same four parameters [St],
[Eu], [Fr], and [Re].
164
Let us therefore take the time to explore the signification of these four
parameters.

The Strouhal number [St] ≡ 𝑓 𝐿/𝑉 (eq. 8/9) quantifies the influence of un-

the importance of the representative frequency 𝑓 at which the flow

steadiness effects over the acceleration field. It does this by evaluating

pattern repeats itself. Very high frequencies denote highly unsteady

flows. When the frequency is very low, [St] is very small; the flow is
then quasi-steady and it can be solved at a given moment in time as if
it was entirely steady.

The Euler number [Eu] ≡ 𝑝0 − 𝑝∞ /𝜌 𝑉 2 (eq. 8/10) quantifies the influence

comparing the largest relative pressure 𝑝0 −𝑝∞ to the flow of momentum

of the pressure gradient over the acceleration field. It does this by

velocity field 𝑉⃗ are likely to be caused by pressure gradients rather

in the flow field. The greater [Eu] is, and the more the changes in the

√
than viscosity, convection or unsteadiness.

The Froude number [Fr] ≡ 𝑉 / 𝑔 𝐿 (eq. 8/11) quantifies the relative impor-
tance of gravity effects. In practice, gravity effects only play an impor-
tant role in free surface flows, such as waves on the surface of a water
reservoir. In most other cases, gravity contributes only to a hydrostatic
effect, which has little influence over the velocity field.

The Reynolds number [Re] ≡ 𝜌 𝑉 𝐿/𝜇 (eq. 8/12, also eqs. 1/28 p. 23 & 7/24

It does this by comparing the magnitude of inertial effects (𝜌𝑉 𝐿) to

p. 142) quantifies the influence of viscosity over the acceleration field.

viscous effects (𝜇). When [Re] is very large, viscosity plays a negligible
role and the velocity field is mostly dictated by the inertia of the fluid.
We return to the significance of the Reynolds number in §8.2.5 below.

The Mach number [Ma] ≡ 𝑉 /𝑐 (eq. 1/10 p. 16) compares the flow speed
𝑉 with that of the molecules within the fluid particles (the speed of
sound 𝑐). [Ma] does not appear in equation 8/13, because we already
decided to restrict ourselves to non-compressible flows. If we hadn’t,
we would be re-expressing the pressure term in that equation as a
function of [Ma], quantifying the effect of changes in density.

These five flow parameters should be thought of scalar fields within the
studied flow domain: there is one distinct Reynolds number, one Mach
number etc. for each point in space and time. Nevertheless, when describing
fluid flows, the custom is to choose for each parameter a single representative
value for the whole flow. For example, when describing pipe flow, it is
customary to quantify a representative Reynolds number [Re]𝐷 based on
the average flow velocity 𝑉av. and the pipe diameter 𝐷 (as we have seen

over an aircraft wing is often based on the free-stream velocity 𝑉∞ and the
with eq. 7/25 p. 142), while the representative Reynolds number for flow

wing chord length. Similarly, the flight Mach number [Ma]∞ displayed on

air speed 𝑉∞ and the free-stream speed of sound 𝑐∞ , rather than particular
an aircraft cockpit instrument is computed using the relative free-stream

values measured closer to the aircraft.

165
 8.2.4 Flow parameters obtained as force ratios
Instead of the mathematical approach covered above, the concept of flow
parameter can be approached by comparing forces in fluid flows. This method
is described for reference in Appendix A5 p. 256.

8.2.5 The Reynolds number in practice

Among the five non-dimensional parameters described above, the Reynolds
number [Re] is by far the most relevant in the study of most fluid flows, and
it deserves a few additional remarks. As we have seen, the Reynolds number
is a measure of how little effect the viscosity has on the time-change of the
velocity vector field:

• With low [Re], the viscosity 𝜇 plays an overwhelmingly large role, and
Video: half-century-old, but
timeless didactic exploration of
how the dynamics of fluids
change with the Reynolds num-
ber, with accompanying notes the velocity of fluid particles is largely determined by that of their own
by G. I. Taylor[3]
neighbors;

• With high [Re], the momentum 𝜌𝑉 of the fluid particles plays a more
by the National Committee for Fluid
Mechanics Films (ncfmf, 1967[21]) (styl)

important role than the viscosity 𝜇, and the inertia of fluid particles
https://youtu.be/51-6QCJTAjU

affects their trajectory much more than the velocity of their neighbors.

In turn, this gives the Reynolds number a new role in characterization of

over the length 𝐿. Indeed, from a kinematic point of view, viscous effects are
flows: it can be thought of as the likeliness of the flow being turbulent

highly stabilizing: they tend to harmonize the velocity field and smooth out
disturbances. On the contrary, when these effects are overwhelmed by inertial
effects, velocity non-uniformities have much less opportunity to dissipate,
and at the slightest disturbance the flow will become and remain turbulent
(fig. 8.1). This is the reason why the quantification of a representative
Reynolds number is often the first step engineers and scientists take when
studying a fluid flow.

166
Figure 8.1: A viscous opaque fluid is dropped into a clearer receiving static fluid with
identical viscosity. The image shows four different experiments photographed after
yielding Reynolds numbers of 0,05, 10, 200 and 3 000 respectively.
the same amount of time has elapsed. The viscosity is decreased from left to right,
As described in eq. A/36 p. 256, a low Reynolds number indicates that viscous effects
dominate the acceleration field. As the Reynolds number increases, the nature of the
velocity field changes until it becomes clearly turbulent.
By the National Committee for Fluid Mechanics Films (ncfmf, 1967[21]), with accompanying notes by Taylor[3]
Image ©1961-1969 Educational Development Center, Inc., reproduced under Fair Use doctrine
A screen capture from film Low Reynolds Number Flow at http://web.mit.edu/hml/ncfmf.html

8.3 Making models

We now know that when we create a scaled-down or scaled-up version of one
flow (as shown for example in fig. 8.2), we attempt to keep all flow parameters
identical. In practice, this is extremely difficult to do, as we will see while
going through this chapter’s problem sheet, especially if ordinary fluids (air
or water) are to be used. The practice in science and engineering is usually
to focus on one or two key parameters, while ignoring others. If the budget
allows for it, several models may be built, each focusing on one parameter
(e.g. one model for compressibility effects, one model for viscous effects).

the flow around the real-size aircraft (here, a 48 m-wide Lockheed C-141 Starlifter),
Figure 8.2: In order for the flow around a wind tunnel model to be representative of
dynamic similarity must be obtained. The value of all flow parameters must be kept
identical. This is not always feasible in practice.
Wind tunnel photo by NASA (public domain)
Full-size aircraft photo CC-by by Peter Long

167
 In the design and construction of models, practical constraints must be
balanced against the need to reproduce the dynamics of fluids accurately.
They include:

• Cost of production. The volume of a model typically increases with its

length cubed, i.e. doubling its length multiplies its volume by a factor 8;

• Precision of manufacturing. Usually, the smaller the model, and the

smaller the geometrical accuracy that can be reached;

• Ease of instrumenting. Carrying out measurements over extremely

large or extremely small models may be challenging;

• Ease of optical access. Optical flow measurement devices are usually

preferred because they do not obstruct the flow, but when the flow
is internal to a model, they require transparent, flat walls to work
correctly;

• Performance of large-scale laboratory equipment. Wind tunnels and

water channels are expensive installations. Often, the characteristics
of models is adapted to the equipment’s availability and performance,
rather than the other way around.

Finally, it must be noted that in this chapter, we focus on reproducing flow

dynamics only. Other kinds of forces applying on models, and other physical
phenomena, are not taken into account. For example, an airplane model in
a tunnel may not be able to fly, and a boat model may not be able to float,
because they are made of different materials than the original objects, and
because weight and lift forces do not scale together (this is also the reason
why large birds such as condors or swans do not look like, and cannot fly as
slowly as mosquitoes and bugs! — these ideas are beautifully and smartly
explored by Hendrik Tennekes [8, 23]). Similarly, heat transfer or chemical
reaction rates, may be completely off in the model, unless the physical laws
that govern them are also taken into account.

Advice from an expert

When your small wind tunnel already cost half a million
euros to build, is run by fans with hundreds of kilowatts
of power, and your new laser-Doppler velocity mea-
surement system cost the lab 200 k€, it is clear that the
available laboratory equipment will dictate the model
size and the experimental flow conditions, not the other
way around.
A well-designed experiment will always take advantage of the best per-
formance of the available equipment. This is why fluid dynamicists like
to measure the performance of their wind tunnel not with size or speed,
but with a tunnel Reynolds number — the maximum Reynolds number
you can subject a model to in the test section.

168
8.4 Comparing results: coefficients

8.4.1 Principle
The general rule that we follow when we compare measurements from
experiments at different scale is that

in two physically similar experiments, the force and power coefficients

are the same.

In other words, for scientists and engineers, the hard work is making a model
where the flow coefficients are the same (and thus the original flow and the
model are physically similar). Once this is done, the comparison of results is
easy: the force and power coefficients are the same on the model and in the
original flow.

8.4.2 Force coefficients

In order to compare forces, we want to have a sense of how flow-induced
forces scale when fluid flows are scaled. A look back on chapter 2 (Analysis
of existing flows with one dimension), and in particular eq. 2/7 p. 37, shows us that
for a steady flow through a given control volume, the net force induced on
the fluid is expressed by: Video: Comparing fluid-induced

𝐹⃗net on fluid = Σnet [𝜌𝑉⟂ 𝐴𝑉⃗ ]

forces: using force coefficients
by Olivier Cleynen (CC-by)
(8/14) https://youtu.be/uGr_dloHWnI

This equation tells us that the norm of the net force vector, 𝐹net , is directly
related to a term with dimensions of 𝜌|𝑉⟂ |𝐴𝑉 . In our selection of a scale by
which to measure 𝐹net , it is therefore sensible to include a term proportional to
the the density 𝜌, a term proportional to the area 𝐴, and a term proportional to
the square of velocity 𝑉 . This “scale of fluid-induced force” is conventionally
measured using the force coefficient 𝐶𝐹 :

𝐹
𝐶𝐹 ≡ 1
𝜌𝑆𝑉 2
(8/15)
2

where 𝐹
𝜌 is a reference fluid density (kg m−3 );
is the considered fluid-induced force (N);

𝑆 is a reference surface area (m2 );

and 𝑉 is a reference velocity (m s−1 ).

In effect, the force coefficient relates the magnitude of the force exerted by

(𝜌𝑆𝑉 2 = 𝑚𝑉̇ ).
the fluid on an object (𝐹 ) to the rate of flow of momentum towards the object

It is worth making a few remarks about this equation. First, it is important

to realize that eq. 8/15 is a definition: while the choice of terms is guided

reasons to expect 𝐶𝐹 (which has no dimension, and thus no unit) to reach

by physical principles, it is not a physical law in itself, and there are no

any particular value in any given case. The choice of terms is also worth
commenting:

169
 • 𝐹 can be any fluid-induced force; generally we are interested in quan-
tifying either the drag 𝐹D (the force component parallel to the free-
stream velocity) or the lift 𝐹L (force component perpendicular to the
free-stream velocity);

• the reference area 𝑆 is chosen arbitrarily. It matters only that this

area grow and shrink in proportion to the studied flow case. In the
automobile industry, it is customary to choose the vehicle frontal area

area is customarily used. The square of a convenient reference length 𝐿

as a reference, while in the aeronautical industry the top-view wing

can also be chosen instead;

• the choice of reference velocity 𝑉 and density 𝜌 is also arbitrary, since

both these properties may vary in time and space within the studied
fluid flow case. It matters merely that the chosen reference values are

tions 𝑉∞ and 𝜌∞ are used;

representative of the case; typically the free-stream (faraway) condi-

• the term 1/2 in the denominator is a purely arbitrary and conventional

value.

Force coefficients are meaningful criteria to compare and relate what is going
on in the wind tunnel and on the full-size object: in each case, we scale the
measured force according to the relevant local flow conditions.

Just like when selecting the length 𝐿 in the Reynolds

Advice from an expert

indicate which reference 𝑆 or 𝑉 should be taken when

number, there is no absolute physical law that would

calculating a force coefficient. The only important

thing is that the definition should remain the same
across all scaled flows.
However, a large body of professional customs and
cultures usually exists in each sub-area of fluid dynamics. Make sure you
consult the available literature and field conventions, so your results fit
among them. Calculating the drag coefficient of a car using the car side-
view area is not wrong, but it would go a long way to annoy colleagues
and competitors!

For example, a flow case around a car A may be studied using a model B.

then the flow dynamics will be identical. The drag force 𝐹D B measured on
If dynamic similarity is maintained (and that is by no means an easy task!),

the model can then be compared to the force 𝐹D A on the real car, using
coefficients: since 𝐶𝐹D B = 𝐶𝐹D A we have:
1 1 𝐹D B 1
Video: practical application of

surements made on a 60 % size

𝐹D A = 𝐶𝐹D A 𝜌𝐴 𝑆A 𝑉A2 = 𝐶𝐹D B 𝜌𝐴 𝑆A 𝑉A2 = 1 𝜌𝐴 𝑆A 𝑉A2
scale effects: translating mea-

2 2 𝜌 𝑆 𝑉 2
2
formula one car wind tunnel
2 𝐵 B B
𝜌𝐴 𝑆A 𝑉A2
model into their “real size” race

𝐹D A = 𝐹
car values, as in problem 8.5

( 𝜌𝐵 𝑆B 𝑉B2 ) D B
by the Sauber F1 team (styl)
https://youtu.be/FQxmOQmnaGw

8.4.3 Power coefficient, and other coefficients

170
During our investigation of the flow using a small-scale model, we may be
interested in measuring not only forces, but also other quantities such as
power — this would be an important parameter, for example, when studying
a pump, an aircraft engine, or a wind turbine.
We again go back to chapter 2 (Analysis of existing flows with one dimension), and Video: Understanding scale ef-
fects will save you a lot of money
in particular eq. 2/18 p. 40. This helps us recall that power gained or lost by and embarrassment: the caution-
ary tale of Howard Hughes
a fluid flowing steadily through a control volume could be expressed as:

𝑝 1
by Olivier Cleynen (CC-by)

𝑄̇ net + 𝑊̇ shaft, net = Σnet 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧

https://youtu.be/fF07XkAUELE

[ ( 𝜌 2 )]
(8/16)

of the scalar Σnet 𝜌|𝑉⟂ |𝐴𝑉 2 . A meaningful ”scale” for the power of a machine
Thus, the power gained or lost by the fluid is directly related to the magnitude

every second, a quantity that grows proportionally to 𝜌𝐴𝑉 3 .

can therefore be the amount of energy in the fluid that is made available to it

power coefficient 𝐶P :
This “scale of fluid flow-related power” is conventionally measured using the

𝑊̇
𝐶P ≡ 1
𝜌𝑆𝑉 3
(8/17)
2

where 𝑊̇ is the power added to or subtracted from the fluid (W).

8.4.4 Non-dimensionalizing all the problems

In this chapter, we have focused on scaling forces and powers from one flow
to the other. There are of course other parameters of interest, and for each, it

loss coefficient 𝐾𝐿 ≡ |Δ𝑝|/ 12 𝜌𝑉av.

is common practice to define a coefficient. We have already used the pressure
2

soon use the shear coefficient 𝑐𝑓 in the forthcoming chapter (eq. 10/6 p. 203).
in chapter 7 (eq. 7/6 p. 137), and we shall
XKCD #687: abusing dimen-
sional analysis
by Randall Munroe (CC-by-nc)
There are many other examples; in fact, in fluid mechanics, the non-dimen- https://xkcd.com/687

sionalization of problems is a very important work methodology. This helps

us extrapolate from one or two experiments up to entire families of flows.

Advice from an expert

Now you should begin to see that there is more to di-
mensional analysis than just making models. Being
good at comparing the influences of various factors
on a fluid flow phenomenon means that you can gen-
eralize results. Pressure loss in a pipe, for example,

eter 𝑓 , irrespective of say the flow speed or fluid

is quantified with a single, suitably-designed param-

density. Take a look back on the Moody diagram

page 152: it has non-dimensional terms on all three sides! This is the
reason why we only need one diagram, instead of a catalog of diagrams
(e.g. one for each flow speed and diameter).
And so, you should not be surprised when, after you show them the
results of your first experimental fluid flow measurement campaign,
your colleagues in the fluid dynamics laboratory ask you: “can you
non-dimensionalize that diagram?”

171
 8.5 Solved problems

Scale model of a car

If we want to reproduce turbulent effects in the airflow around a car

using a 1/20 model, how fast should the flow around the model be?

See this solution worked out step by step on YouTube

https://youtu.be/4p99bCkZdho (CC-by Olivier Cleynen)

Wind tunnel speed

We reproduce the airflow around a full-size car moving in air at 25 °C.
We use a 1/20 model in a wind tunnel. This tunnel is limited to 20 m s−1 ,
but the air is cooled (𝑇 = 0 °C).
What is the real-car speed that is being simulated in the wind tunnel?

See this solution worked out step by step on YouTube

https://youtu.be/gHf8ouKfAAU (CC-by Olivier Cleynen)

Wind tunnel measurement

In the setup from the previous example, how should the force measure-
ments on the car be translated to values corresponding to the real car?

See this solution worked out step by step on YouTube

https://youtu.be/Fn-zxYT9-KY (CC-by Olivier Cleynen)

172
 Problem sheet 8: Engineering models
last edited September 19, 2020
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

The non-dimensional incompressible Navier-Stokes equation:

𝜕 𝑉⃗ ∗ 1 1 ∗2 ⃗ ∗
+ [1] 𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ = ⃗ ∗𝑝∗ +
⃗∗ − [Eu] ∇
2 𝑔
⃗ 𝑉
∇
𝜕𝑡 ∗
[St] (8/13)
[Fr] [Re]

in which [St] ≡ 𝑓𝑉𝐿 , [Eu] ≡ 𝑝𝜌0 −𝑝

𝑉2
∞
, [Fr] ≡ √𝑉 , and [Re] ≡ 𝜌 𝑉𝜇 𝐿 .
𝑔𝐿

The force coefficient 𝐶𝐹 and power coefficient 𝐶P are defined as:

𝐹 𝑊̇
𝐶𝐹 ≡ 𝐶P ≡
1
𝜌𝑆𝑉 2 1
𝜌𝑆𝑉 3
(8/17)
2 2

The speed of sound 𝑐 in air is modeled as:

√
𝑐 = 𝛾 𝑅𝑇 (1/11)

Figure 8.3 quantifies the viscosity of various fluids as a function of temperature.

173
 2×10−2 2.4×10−5

10−2 2.2×10−5
9×10−3
8×10−3
7×10−3 ⟵ Crude Oil
6×10−3 Air ⟶
5×10−3
2×10−5
4×10−3
3×10−3
Viscosity 𝜇 of liquids in Pa s

Viscosity 𝜇 of gases in Pa s
2×10−3 1.8×10−5
CO2 ⟶

10−3 1.6×10−5
9×10−4
8×10−4
7×10−4
6×10−4
5×10−4
1.4×10−5
4×10−4 ⟵ Water
3×10−4

2×10−4 1.2×10−5

10−4 10−5
−20 0 20 40 60 80 100 120
Temperature 𝑇 in degree Celsius (◦C)

Figure 8.3: The viscosity of four fluids (crude oil, water, air, and CO2 ) as a function of temperature.
The scale for liquids is logarithmic and displayed on the left; the scale for gases is linear and
displayed on the right.
Figure reproduced from figure 5.6 p. 99; CC-by by Arjun Neyyathala & Olivier Cleynen

8.1 Reading quiz

Once you are done with reading the content of this chapter, you can go take
the associated quiz at https://elearning.ovgu.de/course/view.php?id=7199
In the winter semester, quizzes are not graded.

8.2 Scaling a golf ball CC-0 Olivier Cleynen

8.2.1. Write the definition of the Reynolds number and indicatie the si units for each
term.

8.2.2. What is the consequence on the velocity field of having a low Reynolds number?
(Briefly justify your answer, e.g. in 30 words or less)

8.2.3. Give one example of a high-Reynolds number flow, and one of a low-Reynolds
174 number flow.
The standard golf ball has a diameter of 42,67 mm and a mass of 45,93 g (figure 8.4). A
typical maximum velocity for such balls is 200 km h−1 .

Figure 8.4: A golf ball

Photo CC-by by Bradley P. Johnson

Using a 3D printer, s/he prints an enlarged model with a diameter of 50 cm.

A student wishes to investigate the flow over a golf ball using a model in a wind tunnel.

8.2.4. If the atmospheric conditions are identical, what flow velocity needs to be
generated during the experiment in order to reproduce the flow patterns around
the real ball?
8.2.5. Would the Mach number for the real ball then be reproduced?
8.2.6. If the temperature on the golf course is −5 °C, and the temperature in the wind
tunnel is 31 °C (with identical pressure, 1 bar), what should the new wind tunnel
velocity be?
8.2.7. If the enlarged model was made out of the same materials as the real ball, how
heavy would it be?

8.3 Fluid mechanics of a giant airlinernon-examinable, CC-0 Olivier Cleynen

An Airbus A380 airliner has an 80 m-wingspan and is designed to cruise at [Ma] = 0,85
where the air temperature and pressure are −40 °C and 0,25 bar.

whose test section has a diameter of 1 m (in which, obviously, the model has to fit!).
A group of students wishes to study the flow field around the aircraft, with a wind tunnel

is always 1,5 ⋅ 10−5 Pa s

Together, they brainstorm various possibilities. They assume that the viscosity of the air

8.3.1. If the temperature inside the tunnel is maintained at 𝑇tunnel = 20 °C, propose a
combination of wind tunnel velocity and pressure which would enable the team
to adequately model the effects of viscosity.
8.3.2. Which flow conditions would be required in a water tunnel with the same
dimensions?

Realizing the conditions calculated in the questions above cannot be realized in practice,
the group of students tries instead to study the effect of compressibility (while accepting
that viscous effects may not be adequately modeled).

8.3.3. If the maximum velocity attainable in the wind tunnel is 80 m s−1 , which tunnel
air temperature is required for compressibility effects to be modeled?
175
 8.4 Scale effects on a dragonfly
A dragonfly (sketched in fig. 8.5) has a 10 cm wingspan, a mass of 80 mg, and cruises
at 4 m s−1 , beating its four wings 20 times per second.

Figure 8.5: Plan view of a dragonfly

Figure by Olivier Cleynen, A. Plank, Drury, Dru, Westwood, J. O. (CC-0)

have access to a wind tunnel with a test section of diameter 1 m.

You are tasked with the investigation of the flight performance of the dragonfly, and

8.4.1. Which model size and flow velocity would you use for this experiment?

8.4.2. How many wing beats per second would then be required on the model?

8.4.3. What would be the lift developed by the model during the experiment?

8.4.4. How much mechanical power would the model require, compared to the real
dragonfly?

8.5 Formula One testing

team races a car that is 5,1 m long, 1,8 m wide, has a 610 kW power plant, and a mass
You are leading the Aerodynamics team in a successful Formula One racing team. Your

of 750 kg (figure 8.6). The car can reach a top speed of 310 km h−1 ; at that speed, the drag
force is 7,1 kN.
In order to test different aerodynamic configurations for the car, your team invests in a

the choice between two model sizes: a 60 % model and a (smaller) 50 % model.
50-million-euro wind tunnel, in which you will run tests on a model. You currently have

8.5.1. What would be the frontal area of each model, in proportion to the frontal area
of the real car?

8.5.2. What would be the volume of each model, compared to the volume of the
real car?

8.5.3. How much less weight would the 50 % model have than the 60 % model?

176
Figure 8.6: A 2017 Formula One racing car, driven by Pascal Wehrlein
Photo CC-by by Jake Archibald

From here on, your team decides to use the 50 % car model.

8.5.4. If the ambient atmospheric conditions cannot be changed, which flow speed in
the wind tunnel is required, so that the air flow around the real car is reproduced
around the model?

You are preparing for a race where the air temperature will be 30 °C. The Formula One
regulations forbid you from running your wind tunnel faster than 50 m s−1 .

8.5.5. Your team considers modifying the wind tunnel air temperature, in order to

be controlled between −10 °C and 40 °C, but the pressure remains atmospheric
compensate for the limit in the air speed. If the temperature in the tunnel can

(1 bar), what is the maximum race-track speed that can be simulated in the wind
tunnel?

8.5.6. In that case, by which factor should the model drag force measurements be
multiplied in order to correspond to the real car?

8.5.7. The wind tunnel has a 2 m by 2 m square test cross-section. What is the power
required to bring stationary atmospheric air outside (𝑇atm. = 15 °C, 𝑉atm. = 0 m s−1 )
to the desired speed and temperature in the test section?

NB: this exercise is inspired by an informative and entertaining video by the Sauber F1 team about
their wind tunnel testing, which you are encouraged to watch at https://youtu.be/KC0E0wU6inU.

177
 Answers
8.2 1) See equation 8/12 and subsequent comments;
2) See §8.2.5, in practice this can be summarized in two or three sentences;
3) High-Re: air flow around an airliner in cruise; Low-Re: air flow around a dust

4) Match [Re]: 𝑉B = 4,74 m s−1

particle falling to the ground.

6) Increase the velocity by 25 %

5) No, but it is very low anyway (no compressibility effects to be reproduced)

7) 𝑚C = 73,9 kg

1) With a half-aircraft model of half-width 𝐿2 /2 = 80 cm, at identical speed and

ambient temperature, we would need an air pressure 𝑝2 ≈ 20 bar!
8.3

𝑉3 = 418 m s−1 !
2) In water, the density cannot be reasonably controlled, and we need a velocity

3) In air, at 𝑉4 = 80 m s−1 , Mach number can be reproduced at 𝑇4 = −251 °C (although

the Reynolds number is off). Wind tunnels used to investigate compressible flow
around aircraft have very powerful coolers.

1) With e.g. a model of span 60 cm, match the Reynolds number: 𝑉2 = 0,67 m s−1 ;
2) Match the Strouhal number: 𝑓2 = 0,56 Hz. Mach, Froude and Euler numbers will
8.4

have no effect here;

3) and 4) are left as a surprise!

3) The third model would have 42 % less mass than the second;
4) Maintaining [Re] requires 𝑉3 = 620 km h−1 (fast!);
8.5

5) Now the fastest track speed that can be studied is 31,7 m s−1 ;
6) Multiply force measurements by 1,39 to scale up to reality;
7) The required work as power is 331 kW and the power to change temperature is
−6,654 MW (better keep the cooled air in a closed circuit than feed the tunnel with
outside air!)

178
 Fluid Dynamics
Chapter 9 – Dealing with turbulence
last edited August 13, 2019
by Olivier Cleynen — https://fluidmech.ninja/

9.1 Motivation 179

9.2 Recognizing turbulence 180
9.2.1 A brief definition 180
9.2.2 Chaos, not randomness 180
9.2.3 Growth and decay 181
9.2.4 A cascade of vortices 182
9.2.5 Not turbulence 182
9.3 The effects of turbulence 183
9.3.1 Dissipation (losses) 183
9.3.2 Main flow patterns 183
9.3.3 Mixing 184
9.4 Quantifying turbulence 184
9.4.1 Average and fluctuation 184
9.4.2 Turbulence intensity 185
9.4.3 The size of eddies 185
9.4.4 Turbulent kinetic energy and dissipation rate 186
9.4.5 Turbulence anisotropy and inhomogeneity 188
9.5 Computing turbulent flow 189
9.5.1 Basic premise 189
9.5.2 Accounting for turbulence 189
9.6 Commented bibliography 190
9.7 Problems 193

9.1 Motivation
Most flows of interest to engineers and scientists are turbulent. Fluid flow
in industrial and domestic piping, in engines and in turbomachinery, is
turbulent. Flow close to solid surfaces and in the wake of objects is turbulent
at all but the slowest speeds. Blood flow in our largest veins and arteries,
and air flow in our nostrils and tracheae, are turbulent. River flows, ocean
currents, and all but the calmest winds are turbulent.
Turbulence may be ubiquitous, but it remains an incredibly complex phe-
nomenon, and describing it accurately requires either extraordinarily power-
ful numerical computations, or advanced mathematics. Neither of those is
available in this course.
We are going to treat turbulence not as a topic of research, but instead merely
as an occurrence that, as engineers, we need to account for. We will try to
answer the following two questions: In a flow, how to quantify and measure
the degree of turbulence? And how to predict its degree and its rate of decay?
In answering those, we will keep in mind that at our level, turbulence char-
acterization is not an exact science: we expect that our models may be off
by a factor of 10 or so. An imprecise or inexact understanding is better than
none at all, and, at the very least, we are developing some familiarity with
turbulence that will be useful in further studies, especially when exploring
computational fluid dynamics or experimental fluid dynamics.

179
 Figure 9.1: Plankton blooming in the Atlantic ocean reveals the complexity of the
covers approximately 500 km in this take.
flow passing over the coast of Argentina. The scale of the image is so that the height

Image by Jacques Descloitres, MODIS Rapid Response Team NASA GSFC (public domain)

9.2 Recognizing turbulence

This topic is well covered in Tennekes & Lumley [5]

9.2.1 A brief definition

We define turbulence as fluid motion in multiple scales which is chaotic and
dissipative.

9.2.2 Chaos, not randomness

This topic is well covered in Leschziner [32]

Turbulence is made possible by kinematic instability in the flow; this means

that small disturbances, such as non-uniformities in the velocity distribution,

ity 𝜇, and so the main parameter which determines how stable a given flow
are not significantly damped. The main dampening factor in fluids is viscos-

is the Reynolds number (eq. 1/28):

𝜌𝑉 𝐿
[Re] ≡
𝜇
(9/1)

Typically, when [Re] exceeds 1 000, the flow is very likely to be or become

is because the magnitude of the Laplacian of the velocity field is 1 000 times
turbulent (figure 9.2). We have seen in chapter 8 (Engineering models) that this

smaller than the magnitude of its advective. (Dampening factors other than
viscosity sometimes also exist, such as density gradients or interaction with
soft solid surfaces: in those cases, other non-dimensional parameters are
used).
This instability is what gives turbulence its chaotic characteristic. If a rider-
less bicycle is rolled forward and left to itself, it will continue rolling and
eventually fall to the side. Which side, left or right, depends on the initial
180
conditions: even a minute modification to the start position is likely to
influence the result. The fall is deterministic, and can be calculated very
precisely, but with a very strong dependence on the initial conditions.

Figure 9.2: A column of hot air from burning incense rising through cold air is an
unstable situation. After a certain length, the flow breaks down into chaotic patterns.
The occurrence is predicted across all fluids, plume diameters, and velocities: every
time, the Reynolds number is the determining parameter.
Photo CC-by by Rafa Espada

Turbulent motion in fluids has the same properties. Fluid motion follows
laws which are fully deterministic (the Navier-Stokes equation, eq. 6/42), but
the exact patterns in situations where the Reynolds number is high cannot
be predicted because, much like the for the bicycle above, they depend very
Video: simulation of two misci-
minutely on the initial configuration. ble fluids of different densities
layered one on top of the other
Thus, in two identical turbulent flow experiments, the details of the flow will (color representing density). The
“perfect” uniform initial situa-
be different. In this sense, turbulent flow is chaotic (depending extremely tion is unstable and leads to
sensitively on initial conditions) but not random: it remains predictable, chaotic (hard to predict) pat-
terns whose details will depend
governed by well-known deterministic laws in which chance does not play strongly on minute changes in
the initial conditions. (A 2d dns
a role. The effective engineer will determine 1) what general characteristics simulation performed with Mi-
of turbulence do remain identical in both flows, and 2) how they affect the croHH)
by Chiel van Heerwaarden (CC-by)
main, global flow characteristics. https://vimeo.com/84518319

9.2.3 Growth and decay

Turbulence occurs because of shear (sometimes also pressure) applied non-
uniformly in a fluid. Typical sources are a sudden turn at the downstream
side of an obstacle, and shear alongside solid walls.
Turbulence is only sustained if shear is continually applied to the flow (for
example by the walls of a long pipe). If no sustaining source is provided,
turbulence decays and eventually dies. This is a critical property: turbulence
is dissipative. Large-scale motion created by an initial obstacle continually 181
 breaks down into ever smaller-scale motion. Ultimately, the motion becomes
so small and slow that the energy is dissipated by viscosity into molecular-
scale motion.
In this sense, turbulence is not magic: it occurs in circumstances that are well
understood and documented. Turbulence takes energy out of the main flow
(this is usually measurable as a streamwise pressure drop) and ultimately
transforms it into heat.

9.2.4 A cascade of vortices

This topic is well covered in Davidson [31]

It is useful to think of turbulence in terms of vortices (also named eddies):

parcels of rotating fluid. In a first approach, we can consider a turbulent fluid
flow as made of two components: one main, mostly steady component (the
time-average), and an additional secondary component, made of a complex
chaotic tangle of vortices of multiple sizes.
Turbulence begins when large vortices are created, whose size is approxi-
mately that of the largest obstacle in the flow path. When those vortices are
stretched, they speed up (by virtue of conservation of angular momentum).
When they are compressed, they buckle and twist, deforming into complex
shapes and interacting with other vortices; this leads to their breakup into
smaller vortices.
A turbulent flow thus consists of a cascade of structures of decreasing size.
Energy is always passed down from larger into smaller structures. Smaller
vortices have lower diameter and feature lower velocities: they are then more
strongly affected by viscosity. Ultimately, the smallest vortices disappear,
their energy dissipated down into heat.

9.2.5 Not turbulence

Not all unsteady flows are turbulent. Well-known patterns such as a von
Kármán vortex street (figure 9.3) or a series of wave clouds, for example, are
not turbulent. Those oscillations occur at a single recognizable frequency
and scale, and once the phenomena has begun, their evolution is easily
predictable.
Most surface waves on a body of water (e.g. waves in open sea, figure 9.3)
are not turbulence: although they may be partly chaotic, they dissipate very
little energy (unless they crash on a shore) and propagate over very large
distances.
Molecular motion is, at the macroscopic scale, completely random, and will
appear in measurements as Gaussian white noise with no distinguishable
range of frequencies, and no dissipation phenomenon: it is not turbulence,
either.
Although more formal definitions of turbulence exist (see e.g. [13, 12, 31]),
we are satisfied with focusing on just three requisites: a flow is turbulent if
the motion contains many scales, is chaotic, and is dissipative.

182
Figure 9.3: Not turbulence: surface waves on the ocean (top) are complex, but
not fully-chaotic, and they feature very little dissipation. Well-known oscillatory
patterns such as the von Kármán vortex street feature one dominant frequency and
one dominant vortex size: they are not turbulent either.
Sea wave photo CC-by-sa by Tiago Fioreze (cropped)
Wake photo CC-by-sa by Jürgen Wagner

9.3 The effects of turbulence

This topic is well covered in Tennekes & Lumley [5]

9.3.1 Dissipation (losses)

Turbulence extracts energy from the main flow and dissipates it at a molecular
level. This energy is never recovered. Turbulence therefore causes additional
losses in the flow, that would not be present if it were to remain laminar.
Typically, friction losses are increased by a factor 2 to 10.
In bulk, turbulence has an effect similar to an increase in viscosity. A word
of warning must follow this sentence: the degree of turbulence in a given
flow is rarely uniform, and so this effect can rarely be quantified as a single
value for the entire flow; instead, it must be computed locally.
The increase in temperature caused by turbulent dissipation is not usually
measurable, due to the high heat capacity of fluids. It usually does not
significantly alter their density. Turbulence therefore acts as a mechanical
energy dissipator for the flow.
In some special cases, turbulence leads to a reduction in losses. We shall
study one such occurrence in chapter 10 (Flow near walls), where we sometimes
intentionally create turbulent boundary layers. This is not important at the
moment.

9.3.2 Main flow patterns

In any given flow, the laminar and turbulent regimes result in markedly
different flow patterns. Turbulent flows always lead to wider shear zones and
very unsteady patterns, which, when averaged over time, have more uniform
velocity profiles. 183
 When the flow is clearly turbulent, increases in the Reynolds number do

[Re]𝑑 = 105 and 107 have nearly identical spread, length, and time-averaged
not lead to main flow pattern changes anymore. For example, jet plumes at

velocity distribution. Some properties of turbulence still change with [Re], as

we will see below.

9.3.3 Mixing
Turbulence tremendously increases mixing. The large range of the scales of
motion (i.e. the many different sizes of vortices which occur simultaneously)
increases the contact surface between two mixing fluids, for example. This
makes turbulence a desired property in many chemical reactors, or in cases
where pollutants have to be dissipated (e.g. for exhaust gases).
The same features of turbulence greatly enhance heat transfer compared to
laminar flow. Most heat exchangers for which space is important, such as
radiators, feature turbulent flow.
The increased mixing also affects exchange of momentum, with increase in
the interaction between slower and faster fluid particles. This tends to widen
areas of interaction between fluids of different velocities, such as plumes,
exhaust stacks and shear layers.

9.4 Quantifying turbulence

9.4.1 Average and fluctuation

For the purpose of quantifying turbulence, we distinguish, in a given flow, be-
tween the average velocity and the “turbulent part” of velocity, as illustrated

one is the average flow (𝑢, 𝑣, 𝑤), and the other the instantaneous fluctuation
in figure 9.4. We thus decompose the velocity field into two components:

flow (𝑢 ′ , 𝑣 ′ , 𝑤 ′ ):

𝑢𝑖 ≡ 𝑢 𝑖 + 𝑢𝑖′
𝑢𝑖′ ≡ 0
(9/2)
(9/3)

here for temperature. The instantaneous temperature 𝑇 is decomposed as the sum of

Figure 9.4: An example of the separation between instantaneous and average values,

the time-averaged temperature 𝑇 (blue curve) and the fluctuation 𝑇 ′ , whose average
𝑇 ′ is zero (red curve).
Figure CC-0 Olivier Cleynen

184
9.4.2 Turbulence intensity

We define turbulence intensity 𝐼 as the average of the root-mean-square of

This topic is well covered in De Nevers [17]

the velocity fluctuations:

1 1
1

𝐼 ≡ [ [(𝑢 ′2 ) + (𝑣 ′2 ) + (𝑤 ′2 )]]
2

𝑉 3
(9/4)

Turbulence intensity is a dimensionless local property representing the

“strength” or “violence” of turbulence, usually expressed as a percentage.
At each point in space, the root-mean-square of each of the three compo-

Typically, a low-turbulence channel flow would feature 𝐼 = 3 %, while a

nents of the velocity fluctuation is compared to the magnitude of the velocity.

highly-turbulent channel flow could display for example 𝐼 = 80 %.

9.4.3 The size of eddies

This topic is well covered in Leschziner [32]

When observing turbulence from the engineer’s point of view, it is useful to

quantify two characteristic lengths: the size of the largest turbulent structures,
and the size of the smallest structures. In that view, the cascade of turbulent
eddies is contained between two limits: the turbulence-producing (large)
scale, and the viscous dissipation (small) scale.

nated as 𝐿max ≡ Λ. It is observed that Λ is typically 50 % of the largest scale

The largest structures have a length whose order of magnitude is desig-

of the flow (somewhat more in unconfined flows, and somewhat less in very
confined flows). For example, wind flow around a building will feature eddies

a characteristic velocity 𝑢Λ (typically, the maximum measurable velocity in

whose maximum size is half the size of the building. Those eddies also feature

the eddy) and a characteristic time scale 𝑡Λ (typically, a duration sufficient to

describe their movement fully), linked by a straightforward relation:
Λ
𝑣Λ =
𝑡Λ
(9/5)

It follows that one may quantify a large-scale eddy Reynolds number based
on those quantities:
𝜌𝑢Λ Λ
[Re]Λ ≡
𝜇
(9/6)

nated as 𝐿min ≡ 𝜂. Likewise, they feature a characteristic velocity 𝑢𝜂 and a

The smallest structures have a length whose order of magnitude is desig-

characteristic time scale 𝑡𝜂 .

In the 1940s, Andrey Kolmogorov and his team developed a model to relate
both the large and the small scales in the simplest occurrences of turbulence.

of the smallest vortices (based on their characteristic length 𝜂 and speed 𝑢𝜂 )

In particular, Kolmogorov postulated that the characteristic Reynolds number

is approximately equal to 1:
𝜌𝑢𝜂 𝜂
[Re]𝜂 ≡ ≈1
𝜇
(9/7)

185
 Figure 9.5: Surface-relative vorticity in the Atlantic ocean. Blue color indicates
clockwise rotation, and red color anticlockwise rotation. Rotating structures of
many different sizes can be observed. In homogeneous isentropic turbulence (and
in this flow case by approximation), the size of the largest and smallest vortices are
related to one another through the Reynolds number.
doi:10.1038/s41467-018-02983-w CC-by by Z. Su, J. Wang, P. Klein, A. F. Thompson & D. Menemenlis [33]

Based on this postulate, when the turbulence has been given time and space
enough to develop fully, is homogeneous, and isotropic (has identical proper-
ties in all three directions) —these are important restrictions—, Kolmogorov
and his peers showed using dimensional analysis that

𝐿min 𝜂
= = [Re]Λ
−3/4
𝐿max Λ
𝑢𝜂
(9/8)

= [Re]Λ
−1/4
𝑢Λ
𝑡𝜂
(9/9)

= [Re]Λ
−1/2
𝑡Λ
(9/10)

These three equations are a very important result. They show that as the
Reynolds number of a flow increases, the size and time scale of the smallest
structures in the flow decreases. The higher the Reynolds number, the more
complex and more minute the details of the flow become.

9.4.4 Turbulent kinetic energy and dissipation rate

We define turbulent kinetic energy 𝑘 as:

This topic is well covered in De Nevers [17]

1
𝑘≡ ((𝑢 ′2 ) + (𝑣 ′2 ) + (𝑤 ′2 ))
2
(9/11)

186
Turbulent kinetic energy, measured in J kg−1 , represents the amount of energy
per unit mass contained in the chaotic (turbulent) component of the fluid

To convince oneself that the parameters are related, one may insert 𝐼 (eq. 9/4)
flow velocity.

into eq. 9/11 to show that 𝑘 = 3/2 𝑉 𝐼 2 .

2

We define the turbulent dissipation rate 𝜖 as the rate at which turbulent

that turbulence is simply left to decay, then 𝜖 is the time rate change of 𝑘:
kinetic energy is dissipating to heat. When no turbulence is produced, so

𝜕𝑘
𝜖=−
𝜕𝑡
(9/12)

when no new turbulence is produced.

The dissipation rate is measured in W kg−1 and represents the local amount
of turbulent kinetic energy that is currently being converted to heat through
viscosity.
Through dimensional analysis, Kolmogorov and his peers showed that in ho-
mogeneous, fully-developed and isotropic turbulence, the size, characteristic
velocity, and characteristic time scale of the smallest eddies could be related
to the dissipation rate with the relationships:

𝜇3 1 4
1

𝜂=
( 𝜌3 𝜖 )
(9/13)

𝜇1 4
1

𝑢𝜂 =
(𝜌 𝜖 )
(9/14)

𝜇1 2
1

𝑡𝜂 =
(𝜌 𝜖 )
(9/15)

In other words, for completely-developed, homogeneous isotropic turbulence,

if the characteristic size and speed of the largest turbulent structures are
known, then the size of the smallest structures and the dissipation power can
be known.
In homogeneous, fully-developed, isotropic turbulence, it is interesting to
observe how the turbulent kinetic energy and dissipation rate are distributed

𝐸 as the kinetic energy of all the eddies in an eddy-size increment d𝑙 in a

across the scales of the vortices. One can define the kinetic energy density

volume of interest, divided by that increment:

𝜕𝑘
𝐸≡
𝜕𝑙
In this way, the kinetic energy 𝑘 is recovered as the integral of 𝐸 with respect
to d𝑙:
Λ
𝑘=∫ 𝐸 d𝑙 (9/16)
𝜂

In the same way, we may define the dissipation density 𝐷 as 𝜕𝜖/𝜕𝑙, so that
the dissipation rate would be recovered as:
Λ
𝜖=∫ 𝐷 d𝑙 (9/17)
𝜂 187
 figure 9.6, where the distribution of 𝑘 and 𝜖 across the scales of eddies is
These two integral equations are only useful to understand the meaning of

plotted.

Figure 9.6: Distribution of turbulent kinetic energy (left) and of turbulent dissipation
rate (right) in fully-developed homogeneous isotropic turbulence. The top diagrams
diagrams, the horizontal axis displays 1/𝑙, so that the small-scale eddies are on the
are in linear scale, while the bottom diagrams are in logarithmic scale. In those
right side, and large-scale eddies are on the left side.
turbulence; their features (in particular, the curves’ slopes and the ratios between Λ
Those energy and dissipation distributions are for the simplest occurrences of
and 𝜖) are used as reference cases in the study of more complex cases.
Figure CC-by-sa by Olivier Cleynen

In this figure 9.6, one may see that most of the energy is contained in the large
eddies, while most of the dissipation occurs in the small eddies.

9.4.5 Turbulence anisotropy and inhomogeneity

The relationships between largest and smallest scales derived by Kolmogorov
only hold true for very “straightforward” turbulence. In most flows of interest,
turbulence is being created and dissipated in more complex ways. In such
cases, which are outside the scope of this course, important properties to
quantify are:

Inhomogeneity (uneven distribution in space). The properties quantified

above are observed in different places, often revealing that production
and dissipation of turbulence occur in different areas;

Anisotropy (dependence on direction). The properties quantified above are

observed in different directions. Anisotropy is most often quantified by
evaluating correlations between velocity fluctuations, and the skewness
(asymmetry) of their distributions.
188
 Development. Sometimes, not enough time has passed for the full spectrum
of scales to appear, because the energy has not cascaded down to the
smallest scales yet.
In most of those cases, the basic models studied above serve as reference
cases, against which comparisons can be made.

9.5 Computing turbulent flow

9.5.1 Basic premise

for any given flow by solving for the change in time of the unknowns 𝑢, 𝑣
We have seen in §6.5 that in principle, the flow of fluids can be computed

and 𝑤 in the Navier-Stokes equation. Such a formulation is called a Direct

Numerical Simulation (dns); it allows solving for all flows and will very well
describe turbulent flows.
Unfortunately, the reality is that the computational cost of dns is enormous,
and precludes us from solving most flows of interest. Turbulence exacerbates
the problem. With eqs. 9/8 & 9/10 we can see that as the Reynolds number
increases, the spatial and temporal discretization of the computation must
increase, too. Every decrease in the size of the grid cell and in the length Video: dns simulation of air

algorithm. Halving each of 𝛿𝑥, 𝛿𝑦, 𝛿𝑧 and 𝛿𝑡 multiplies the total number
low speeds ([Re] = 4 ⋅ 105 , so 𝑉 ≈
flow over an airfoil at relatively
of the time step increases the total number of equations to be solved by the
50 km h−1 ). Because the com-
of equations by 16, so that soon enough the designer of the simulation will solved, 35 million cpu-hours
plete details of the flow are

were needed for this calculation

wish to know what maximum (coarsest) grid size is appropriate or tolerable. in 2015. On an ordinary desktop
Furthermore, in many practical cases, we may not even be interested in an computer, this would take 500
years to complete.
exhaustive description of the velocity field, and just wish to obtain a general, by Linné FLOW Centre (styl)
https://youtu.be/aR-hehP1pTk
coarse description of the fluid flow.

9.5.2 Accounting for turbulence

The decomposition of the flow into two components 𝑢 and 𝑢 ′ which we
This topic is well covered in Versteeg et al. [20], Leschziner [32], and Wilkes [19]

performed in eq. 9/2 earlier is useful when we wish to simulate the flow

say that 𝑢 is the component of velocity that is captured by the simulation,

numerically (performing computational fluid dynamics or cfd). Let us now

while 𝑢 ′ is the component which is too small, or occurring too quickly, for
the simulation to capture. Inserting the definition 9/2 into the 𝑥-component
of the Navier-Stokes equation for incompressible flow, we obtain:

𝜕(𝑢 + 𝑢 ′ ) 𝜕(𝑢 + 𝑢 ′ ) 𝜕(𝑢 + 𝑢 ′ ) 𝜕(𝑢 + 𝑢 ′ )

𝜌 + (𝑢 + 𝑢 ′ ) + (𝑣 + 𝑣 ′ ) + (𝑤 + 𝑤 ′ )
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ]
𝜕(𝑝 + 𝑝 )
′
𝜕 (𝑢 + 𝑢 ) 𝜕 (𝑢 + 𝑢 ) 𝜕 (𝑢 + 𝑢 ′ )
2 ′ 2 ′ 2
= 𝜌𝑔𝑥 − +𝜇 + +
𝜕𝑥 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]

Taking the average of this equation —thus expressing the dynamics of the
flow as we calculate them with a finite, coarse grid— yields, after some
intimidating but easily conquerable algebra:

𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 ′ 𝜕𝑢 ′ 𝜕𝑢 ′
𝜌 +𝑢 +𝑣 +𝑤 + 𝜌 𝑢′ + 𝑣′ + 𝑤′
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] [ 𝜕𝑥 𝜕𝑦 𝜕𝑧 ]
𝜕𝑝 𝜕 𝑢
2
𝜕 𝑢
2
𝜕 𝑢
2
= 𝜌𝑔𝑥 − +𝜇 + +
𝜕𝑥 [ (𝜕𝑥)2 (𝜕𝑦) 2 (𝜕𝑧)2 ]
(9/18)
189
 Equation 9/18 is the 𝑥-component of the Reynolds-averaged Navier-Stokes
equation (rans). It shows that when one observes the flow in terms of the
sum of an average and an instantaneous component, the dynamics cannot be
expressed solely according to the average component. Comparing eqs. 9/18

term, called the Reynolds stress, is often re-written as 𝜌 𝜕𝑢𝑖′ 𝑢𝑗′ /𝜕𝑗. In turbulent
and 6/43 we find that an additional term has appeared on the left side. This

𝑢 ′ and 𝑣 ′ ) are strongly correlated: they are each zero on average, but their
flow, it is not zero, because the instantaneous fluctuations of velocity (e.g.

multiples are not.

The difference between eqs. 9/18 and 6/43 can perhaps be expressed differ-
ently: the time-average of a turbulent flow cannot be calculated by solving for
the time-average velocities. Or, more bluntly: the average of the solution can-
not be obtained with only the average of the flow. This is a tremendous burden
in computational fluid dynamics, where limits on the available computational
power prevent us in practice from solving for these fluctuations.
In the overwhelming majority of computations, the Reynolds stress has to

is, a local value for 𝜌 𝜕𝑢𝑖′ 𝑢𝑗′ /𝜕𝑗 is estimated everywhere, depending on the
be approximated in bulk with schemes named turbulence models. That

average values (𝑢𝑖 ). The most well-known method for doing this is the
𝑘-epsilon turbulence model, which involves solving partly arbitrary transport
equations for both 𝑘 and 𝜖. The delights, shortcomings and mysteries of
that method and more are left for the reader to discover in a good hands-on
course on the youngest and most promising area of this discipline, cfd.

9.6 Commented bibliography

Most books on turbulence spill into mathematical intricacies which are irrel-
evant to the engineer; it is regrettable that some of the very best documenta-
tion on turbulence is in the user manuals of cfd software and experimental
measurement devices. It is also regrettable that few books provide actual
applied problems to be solved quantitatively. A commented (and necessarily
subjective) bibliography for learning about turbulence is proposed here:

• General understanding (encyclopedic knowledge)

The introductions of Mathieu & Scott [12], of Tennekes & Lumley [5],
and of Davidson [31] will provide excellent information.

• Reference works
Unfortunately, no book truly aimed at engineers is known to the au-
thor. The following books provide in-depth insight over the physics of
turbulence:

– Tennekes & Lumley [5]: despite its age, an outstanding book, in

particular for its first chapter.
– Leschziner [32]: despite its focus on cfd, the book serves as a
great step-by-step exploration of turbulence.
– Davidson [31] (reference book). Exquisitely referenced.
– Mathieu & Scott [12] (reference book). An appropriate amount of
comment is provided around mathematical expressions.
– Pope [13] (reference book). Terser than its counterparts above.
190
• Works containing useful problems
De Nevers [17] (a single chapter within) is a very applied, down-to-
earth and useful treatment; Tennekes & Lumley [5] also contains good
examples in the first chapter.

• Dealing with turbulence in cfd

The best coverage is probably in Leschziner [32]. Some useful (passing)
information can be found in Tu et al. [30] and Versteeg et al. [20].

• Books with parts useful for specific purposes include:

– Bernard & Wallace [14], for theory about experiments;

– Libby [10], for its introduction to applied statistics;
– Sagaut [18], for its coverage of the notion of spectra, and its
excellent illustrations;
– Cebeci [16], for its rigorous introduction into basic metrics, and
excellent fundamental data and diagrams.

191
192
 Problem sheet 9: Dealing with turbulence
last edited August 13, 2019
by Olivier Cleynen — https://fluidmech.ninja/

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

Turbulence intensity 𝐼 :

1 1
1

𝐼 ≡ [ [(𝑢 ′2 ) + (𝑣 ′2 ) + (𝑤 ′2 )]]
2

𝑉 3
(9/4)

Turbulent kinetic energy 𝑘:

1
𝑘≡ ((𝑢 ′2 ) + (𝑣 ′2 ) + (𝑤 ′2 ))
2
(9/11)

In homogeneous, isotropic, fully-developed turbulence, the following relationships

apply between the largest-scale and smallest-scale eddies:
𝐿min 𝜂
= = [Re]Λ
−3/4
𝐿max Λ
𝑢𝜂
(9/8)

= [Re]Λ
−1/4
𝑢Λ
𝑡𝜂
(9/9)

= [Re]Λ
−1/2
𝑡Λ
(9/10)

𝜇3 1 4
1

𝜂=
( 𝜌3 𝜖 )
(9/13)

𝜇1 4
1

𝑢𝜂 =
(𝜌 𝜖 )
(9/14)

𝜇1 2
1

𝑡𝜂 =
(𝜌 𝜖 )
(9/15)

193
 9.1 Hypothetical flow
We imagine a turbulent flow described at some point with the equations (in m s−1 )
From De Nevers [17] Ex 18.1

𝑢 = 10 + sin 𝑡
𝑣=0
(9/19)

𝑤=0
(9/20)
(9/21)

(No real turbulent flow can be described by equations this simple — but this is a nice first

What are the values of 𝑢, 𝑢 ′ , 𝑢 ′ , 𝐼𝑥 , 𝐼 , and 𝑘?

basis for practice)

Hint: ∫ sin2 𝑥 d𝑥 = 12 (𝑥 + 2 )
sin 2𝑥
+𝑏

9.2 Turbulent channel flow

A wind tunnel carries air through a channel which is 1 m wide and 0,24 m high. The
From De Nevers [17] Ex 18.2

average velocity is 0,82 m s−1 . The pressure drop caused by both friction on the walls and
turbulent dissipation is measured at −0,0286 Pa m−1 .

9.2.1. What is the non-turbulent kinetic energy per unit mass of the flow?
9.2.2. At what average rate does this kinetic energy degrade into heat?
9.2.3. If there was no heat transfer, what would be the rate of temperature increase of
the air?

Measurements are carried out to measure the turbulent intensity through the channel.
Those are displayed in figure 9.7.

Figure 9.7: Measurements of turbulent intensity in 𝑥 and 𝑦 directions in a rectangular channel

1 m wide and 0,24 m wide, in which the centerline velocity 𝑣 is 1 m s−1 . Here 𝑢 ′ and 𝑣 ′ are written
𝑣𝑥 and 𝑣𝑦 respectively.
Figure extracted from De Nevers [17], with source data from Reichardt 1938, Naturwissenschaften 26:407

194 9.2.4. What is the value of 𝑘 at a point 2 cm from the wall?

9.3 Cumulus cloud From Tennekes & Lumley [5] P1.1

size of a sphere of diameter 𝐷 = 50 m. To a good approximation, it features isotropic,

A cumulus cloud (one of those “fluffy” summer clouds, figure 9.8) has roughly the

reach a maximum velocity 𝑉 = 3 m s−1 .

homogeneous, fully-developed turbulence. The largest-scale air currents in the cloud

Figure 9.8: Summer clouds (Cumulus humilis) form when hot moist air convected from the ground
is cooled down when it rises.
Photo CC-by-sa by en:Wikipedia User:Dwindrim

9.3.1. What is approximately the size of the smallest eddies in the cloud?
9.3.2. What is approximately the dissipation power, per unit mass of air and for the
entire cloud?

𝐷2 = 100 m?
9.3.3. What will those three values become once the cloud has grown to a diameter of

9.4 Reactor tank

𝐿 = 2 m (figure 9.9). The tank is filled with a water-like liquid and vigorously stirred
A tank used to store chemical reactants has roughly the size of a cube of side length

maximum fluid velocity of 1,5 m s−1 .

with a large agitator propeller for a prolonged amount of time. The propeller induces a

9.4.1. What is approximately the size of the smallest eddies in the tank?
9.4.2. What is approximately the specific dissipation power?

A full-scale simulation (dns) of the flow was carried out, which required 500 hours of
computing time on a supercomputer. Now the same simulation is to be carried out again,
for the same flow, but with a fluid whose viscosity is half that of water.

9.4.3. What do you expect the new computation time to be?

195
 Figure 9.9: Schematic drawing of a cubic tank stirred with an agitator propeller
Figure CC-0 Olivier Cleynen

196
Answers
9.1 p. 194

9.1.1 𝑢 = 10 m s−1
9.1.2 𝑢 ′ = sin(𝑡)
9.1.3 𝑢 ′ = 0 m s−1 (as always)
9.1.4 𝐼𝑥 = 7,07 %
9.1.5 𝐼 = 4,08 %
9.1.6 𝑘 = 0,25 J kg−1

9.2 p. 194

9.2.1 𝑒̇ 𝑚main = 0,336 J kg−1

9.2.2 𝜖 = 0,0191 W kg−1
9.2.3 𝑇̇ = 0,02 mK s−1
9.2.4 𝑘 = 5,76 mJ kg−1

9.3 p. 195

9.3.1 Since [Re]Λ ≡ 6,1 ⋅ 103 , 𝜂 ≈ 0,2 mm

9.3.2 𝜖 ≈ 1,147 W kg−1 & 𝑊̇ 𝜖 ≈ 92 kW
9.3.3 𝜂2 = 0,6𝜂1 , 𝜖2 = 8𝜖1 , 𝑊̇ 𝜖2 = 64𝑊̇ 𝜖1

9.4 p. 195

9.4.1 𝜂 ≈ 0,023 mm (width of human hair)

9.4.2 𝜖 ≈ 3,4 W kg−1
9.4.3 Increase resolution in all three directions according to 𝜂, and the time
resolution according to 𝑡𝜂 : the computation time increases by a factor 6,7.

197
198
 Fluid Dynamics
Chapter 10 – Flow near walls
last edited September 19, 2020
by Olivier Cleynen — https://fluidmech.ninja/

10.1 Motivation 199

10.2 The concept of boundary layer 199
10.2.1 Rationale 199
10.2.2 Why do we study the boundary layer? 201
10.2.3 Characterization of the boundary layer 201
10.3 Laminar boundary layers 203
10.3.1 Governing equations 203
10.3.2 Blasius’ solution 204
10.4 Boundary layer transition 205
10.5 Turbulent boundary layers 206
10.6 Flow separation 207
10.7 Solved problems 210
10.8 Problems 213

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

10.1 Motivation
In this chapter, we focus on fluid flow close to solid walls. In these regions,
viscous effects dominate the dynamics of fluids. This study should allow us
to answer two questions:

• How can we quantify shear-induced friction on solid walls?

• How can we describe and predict flow separation?

10.2 The concept of boundary layer

10.2.1 Rationale
At the very beginning of the 20th century, Ludwig Prandtl observed that for
most ordinary fluid flows, viscous effects played almost no role outside of a
very small layer of fluid along solid surfaces. In this area, shear between the
zero-velocity solid wall and the outer flow dominates the flow structure. He
named this zone the boundary layer.w
We indeed observe that around any solid object within a fluid flow, there
exists a thin zone which is significantly slowed down because of the object’s
presence. This deceleration can be visualized by measuring the velocity
profile (fig. 10.1).

(termed 𝛿, as we will see in §10.2.3) is defined as the distance where the fluid
The boundary layer is a concept, a thin invisible layer whose upper limit

velocity in direction parallel to the wall is 99 % of the outer (undisturbed)

flow velocity.
199
 component of velocity (𝑉𝑥 = 𝑢) is represented.
Figure 10.1: A typical velocity profile in a boundary layer. Only the horizontal

Figure CC-by Olivier Cleynen

incoming flow velocity 𝑈∞ .

Figure 10.2: The thickness of the boundary layer depends strongly on the faraway

Figure CC-by-sa Olivier Cleynen & Commons User:F l a n k e r

Upon investigation, we observe that the boundary layer thickness depends

strongly on the main flow characteristics. In particular, it decreases when
speed increases or when viscosity is decreased (fig. 10.2).
As we travel downstream along a boundary layer, we observe experimentally
that the flow regime is always laminar at first. Then, at some distance
downstream which we name transition point, the boundary layer becomes
turbulent. The flow-wise position of the transition point depends on the flow
properties and is somewhat predictable. Further downstream, the boundary
layer becomes fully turbulent. It has larger thickness than in the laminar
regime, and grows at a faster rate. Like all turbulent flows, it then features
strong energy dissipation and its analytical description becomes much more
difficult.
The flow within the boundary layer, and the main external flow (outside of it)
affect one another, but may be very different in nature. A laminar boundary
layer may exist within a turbulent main flow; while turbulent boundary
layers are commonplace in laminar flows.

200
10.2.2 Why do we study the boundary layer?
Expending our energy on solving such a minuscule area of the flow may
seem counter-productive, yet three great stakes are at play here:

• First, a good description allows us to avoid having to solve the

Navier-Stokes equations in the whole flow. Video: the basics of the bound-
ary layer
Indeed, outside of the boundary layer and of the wake areas, viscous

⃗ 𝑝 (the Euler equation, which we will introduce as eq. 11/3

by Olivier Cleynen (CC-by)

𝜌 DD𝑡𝑉 = −∇
effects can be safely neglected. Fluid flow can then be described with
⃗
https://youtu.be/cgWIYSnvIEg

in the next chapter) with acceptably low inaccuracy. As we shall

soon see, this allows us to find many interesting solutions, all within
reach of human comprehension and easily obtained with computers.
Unfortunately, they cannot account for shear on walls.
Thus, solving the flow within the boundary layer, whether analytically
or experimentally, allows us to solve the rest of the flow in a simplified
manner (fig. 10.3).

Figure 10.3: Fluid flow around a wing profile. When analyzing the flow, whether
analytically or within a computational fluid dynamics (cfd) simulation, the flow
domain is frequently split into three distinct areas. In the boundary layer (B), fluid
flow is dominated by viscosity. Outside of the boundary layer (A), viscous effects
are very small, and the flow can be approximately solved using the Euler equation.
Lastly, in the turbulent wake (C), characterization of the flow is very difficult and
typically necessitates experimental investigations.
Figure CC-by-sa Olivier Cleynen

• Secondly, the boundary layer is the key to quantifying friction. A

good resolution of the boundary layer allows us to precisely quantify
the shear forces generated by a fluid on an object.

• Finally, a good understanding of the mechanisms at hand within the

boundary layer allows us to predict flow separation, which is the di-
vergence of streamlines relative to the object. Control of the boundary
layer is key to ensuring that a flow will follow a desired trajectory!

10.2.3 Characterization of the boundary layer

Three different parameters are typically used to quantify how thick a bound-
ary layer is at any given position.
201
 The first is the thickness 𝛿,

𝛿 ≡ 𝑦|𝑢=0,99𝑈 (10/1)

where the speed 𝑢 is 99 % of 𝑈 .

which is, as we have seen above, equal to the distance away from the wall

The second is the displacement thickness 𝛿 ∗ , which corresponds to the vertical

distance by which streamlines outside of the boundary layer have been
displaced. This vertical “shifting” of the flow occurs because the inner fluid is
slowed down near the wall, creating some blockage for the outer flow, which
then proceeds to avoid it partially by deviating outwards (fig. 10.4).
Integral analysis performed on a control volume enclosing the boundary
layer (as for example in exercise 3.8 from chapter 3) allows us to quantify the
displacement thickness as:
∞
𝑢
𝛿 ≡∫
∗
(1 − 𝑈 ) d𝑦 (10/2)
0

the ∞ limit), as long as the upper limit exceeds the boundary layer thickness.
In practice, this integral can be calculated on a finite interval (instead of using

The third and last parameter is the momentum thickness 𝛿 ∗∗ (sometimes

written 𝜃) which is equal to the thickness of a corresponding layer of fluid
at velocity 𝑈 which would possess the same amount of momentum as the
boundary layer. The momentum thickness can be thought of as the thickness
of the fluid that would need to be entirely stopped (for example by pumping
it outside of the main flow) in order to generate the same drag as the local
boundary layer.
A review of the experience we gathered in chapter 3 while solving prob-
lems 3.7 to 3.9 p. 67 allows us to to quantify the momentum thickness as:

𝛿
𝑢 𝑢
𝛿 ∗∗ ≡ ∫ 1 − d𝑦
𝑈 ( 𝑈)
(10/3)
0

with the same remark regarding the upper limit. The momentum thickness,
in particular when compared to the displacement thickness, is an important
parameter used in prediction models for boundary layer separation.

for a quantification of the shear term 𝜏wall . Since we are working with the
Once these three thicknesses have been quantified, we are generally looking

Figure 10.4: Thickness 𝛿 and displacement thickness 𝛿 ∗ of a boundary layer. It is

(drawn blue) penetrate it and the vertical velocity 𝑣, although very small compared
important to understand that the boundary layer is not a closed zone: streamlines
to 𝑢, is not zero.
Figure CC-by-sa Olivier Cleynen & Commons User:F l a n k e r

202
hypothesis that the fluid is Newtonian, we merely have to know 𝑢(𝑦) to
quantify shear, according equation 5/22 which we wrote way back p. 97:
𝜕𝑢
𝜏wall 𝑦𝑥 = 𝜇
𝜕𝑦
(10/4)

In a boundary layer, the shear 𝜏wall will decrease with longitudinal distance 𝑥,
because the velocity gradient above it also decreases. Consequently, 𝜏wall will
become a function of 𝑥, so that the entire shear force will be obtained by
integration (reusing eq. 5/3 p. 92):

𝐹shear𝑦𝑥 = ∫ 𝜏wall 𝑦𝑥 d𝑥 d𝑧 (10/5)

𝑆

As we have seen in the previous chapter, fluid dynamicists like to quantify

boundary layer is typically non-dimensionalized with the shear coefficient 𝑐𝑓 ,

phenomena with non-dimensional parameters. The wall shear exerted by the

𝜏wall
𝑐𝑓(𝑥) ≡ 1
𝜌𝑈 2
(10/6)
2

where 𝑈 is the outer-layer (free-stream) velocity.

distance 𝑥.
The shear coefficient, just like the shear, remains a function of the flow-wise

10.3 Laminar boundary layers

10.3.1 Governing equations

What is happening inside a laminar, steady boundary layer? We begin by
writing out the Navier-Stokes for incompressible isothermal flow in two
Cartesian coordinates (eqs. 6/42 p. 125):

𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢
𝜌 +𝑢 +𝑣 = 𝜌𝑔𝑥 − +𝜇 +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 ] 𝜕𝑥 [ (𝜕𝑥)2 (𝜕𝑦)2 ]
(10/7)

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑝 𝜕 2𝑣 𝜕 2𝑣
𝜌 +𝑢 +𝑣 = 𝜌𝑔𝑦 − +𝜇 +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 ] 𝜕𝑦 [ (𝜕𝑥)2 (𝜕𝑦)2 ]
(10/8)

Building from these two equations, we are going to add three simplifications,
which are hypotheses based on experimental observation of fluid flow in
boundary layers:

1. Gravity plays a negligible role;

tion, 𝑣) is very small (𝑣 ≪ 𝑢).

2. The component of velocity perpendicular to the wall (in our conven-

Thus, its stream-wise spatial variations can also be neglected: 𝜕𝑣/𝜕𝑥 ≈ 0

and 𝜕 2 𝑣/(𝜕𝑥)2 ≈ 0. The same goes for the derivatives in the 𝑦-direction:
𝜕𝑣
𝜕𝑦
≈ 0 and 𝜕 2 𝑣/(𝜕𝑦)2 ≈ 0.

3. The component of velocity parallel to the wall (in our convention, 𝑢)

varies much more strongly in the 𝑦-direction than in the 𝑥-direction:
𝜕 2 𝑢/(𝜕𝑥)2 ≪ 𝜕 2 𝑢/(𝜕𝑦)2 . 203
 With all of these simplifications, equation 10/8 shrinks down to
𝜕𝑝
≈0
𝜕𝑦
(10/9)

which tells us that pressure is a function of 𝑥 only (𝜕𝑝/𝜕𝑥 = d𝑝/ d𝑥).

applying it outside of the boundary layer where 𝑢 = 𝑈 :

We now turn to equation 10/7, first to obtain an expression for pressure by

d𝑝 d𝑈
= −𝜌𝑈
d𝑥 d𝑥
(10/10)

and secondly to obtain an expression for the velocity profile:

𝜕𝑢 𝜕𝑢 1 𝜕𝑝 𝜇 𝜕 2 𝑢
𝑢 +𝑣 =− +
𝜕𝑥 𝜕𝑦 𝜌 𝜕𝑥 𝜌 (𝜕𝑦)2
d𝑈 𝜇 𝜕 2 𝑢
=𝑈 +
d𝑥 𝜌 (𝜕𝑦)2
(10/11)

Thus, the velocity field 𝑉⃗ = (𝑢; 𝑣) = 𝑓 (𝑥, 𝑦) in a steady laminar boundary

layer is driven by the two following equations: a balance of momentum, and
a balance of mass:
𝜕𝑢 𝜕𝑢 d𝑈 𝜇 𝜕 2 𝑢
𝑢 +𝑣 =𝑈 +
𝜕𝑥 𝜕𝑦 d𝑥 𝜌 (𝜕𝑦)2
(10/12)
𝜕𝑢 𝜕𝑣
+ =0
𝜕𝑥 𝜕𝑦
(10/13)

layer, 𝑢(𝑥,𝑦) . Unfortunately, over a century after it has been written, we still
The main unknown in this system is the longitudinal speed profile across the

have not found an analytical solution to it.

10.3.2 Blasius’ solution

Heinrich Blasius undertook a PhD thesis under the guidance of Prandtl, in
which he focused on the characterization of laminar boundary layers. As
part of his work, he showed that the geometry of the velocity profile (i.e.

regardless of the flow velocity or the position, 𝑢 can be simply expressed as a

the velocity distribution) within such a layer is always the same, and that

function of non-dimensionalized distance away from the wall termed 𝜂:

√
𝜌𝑈
𝜂≡𝑦
𝜇𝑥
(10/14)

Blasius was able to show that 𝑢 is a function such that 𝑢/𝑈 = 𝑓(𝜂) ′

𝑓 + 2 𝑓 𝑓 = 0. Unfortunately, no known analytical solution to this equation

, with
′′′ 1 ′′

for 𝑓 ′ at selected positions 𝜂. Those are plotted in fig. 10.5.

is known. However, it has now long been possible to obtain numerical values

Based on this work, it can be shown that for a laminar boundary layer flowing
along a smooth wall, the four parameters of interest for the engineer are

204
Figure 10.5: The velocity profile obtained by Blasius (an exact solution to the Navier-
Stokes equations simplified with laminar boundary-layer hypothesis).
Figure CC-0 Olivier Cleynen

simply functions of the distance-based Reynolds number [Re]𝑥 :

𝛿 4,91
=√
𝑥
(10/15)
[Re]𝑥
𝛿∗ 1,72
=√
𝑥
(10/16)
[Re]𝑥
𝛿 ∗∗ 0,664
=√
𝑥
(10/17)
[Re]𝑥
0,664
𝑐𝑓(𝑥) = √ (10/18)
[Re]𝑥

10.4 Boundary layer transition

After it has traveled a certain length along the wall, the boundary layer
becomes very unstable and it transits rapidly from a laminar to a turbulent
regime (fig. 10.6). We have already described the characteristics of turbulence
in broadly in chapter 7 (Pipe flows) and more extensively in chapter 9 (Dealing
with turbulence); they apply to turbulence within the boundary layer. It is worth
reminding ourselves that the boundary layer may be turbulent in a globally
laminar flow (e.g. around an aircraft in flight, the boundary layer is turbulent,
but the main flow is laminar). Here, we refer to the regime of the boundary

We observe that the distance 𝑥transition at which the boundary layer changes
layer only, not the outer flow.

regime is reduced when the velocity is increased, or when the viscosity is

number [Re]𝑥 ≡ 𝜌𝑈 𝑥/𝜇. The most commonly accepted prediction for the
decreased. In practice this distance depends on the distance-based Reynolds

transition position is:

[Re]𝑥 transition ≈ 5 ⋅ 105 (10/19) 205

 Figure 10.6: Transition of a boundary layer from laminar to turbulent regime.
Figure CC-by Olivier Cleynen

Transition can be generated earlier if the surface roughness is increased, or

if obstacles (e.g. turbulators, vortex generators, trip wires) are positioned
within the boundary layer. Conversely, a very smooth surface and a very
steady, uniform incoming flow will result in delayed transition.

10.5 Turbulent boundary layers

The extensive description of turbulent flows remains an unsolved problem.
As we have seen in chapter 9 (Dealing with turbulence), by contrast with laminar
counterparts, turbulent flows result in
• increased mass, energy and momentum exchange;
Video: visualizing the turbu-
lent boundary layer on flat body
moved in a stationary tank • increased losses to friction;
by Lee, Kwon, Hutchins, and Monty [28]
(styl)
https://youtu.be/e1TbkLIDWys • apparently chaotic internal movements.
Instead of resolving the entire time-dependent flow in the boundary layer, we

speed, 𝑢. A widely-accepted velocity model is:

satisfy ourselves with describing the average component of the longitudinal

𝑢 𝑦 7
≈( )
1

𝑈 𝛿
(10/20)

for turbulent boundary layer flow over a smooth surface.

This profile has a much flatter geometry near the wall than its laminar
counterpart (fig. 10.7).
In the same way that we have worked with the laminar boundary layer
profiles, we can derive models for our characteristics of interest from this
velocity profile:
𝛿 0,16
≈
𝑥 [Re] 17
Video: a look inside a turbu-
lent boundary layer, with a com-
(10/21)
𝑥
pletely resolved (dns) fluid flow

𝛿 ∗
0,02
simulation

≈
𝑥 [Re] 17
by Linné Flow Centre
& Serc KTH [26, 27] (styl)
https://youtu.be/4KeaAhVoPIw (10/22)
𝑥
𝛿 ∗∗ 0,016
≈
𝑥 1 (10/23)
[Re]𝑥7
0,027
𝑐𝑓(𝑥) ≈ 1 (10/24)
[Re]𝑥7
206
Figure 10.7: Comparison of laminar and turbulent boundary layer profiles, both
boundary layers are much thicker than laminar ones, and that 𝑢 is only a time-
scaled to the same size. It is important to remember that in practice turbulent
average velocity.
Figure CC-0 Olivier Cleynen

10.6 Flow separation

Under certain conditions, fluid flow separates from the wall. The boundary
layer then disintegrates and we observe the appearance of a turbulent wake
near the wall. Separation is often an undesirable phenomenon in fluid
mechanics: it may be thought of as the point where we fail to impart a

When the main flow speed 𝑈 along the wall is varied, we observe that
desired trajectory to the fluid. Video: flow separation: what it
is, how to predict and avoid it
by Olivier Cleynen (CC-by)

speed gradient (d𝑈 / d𝑥 > 0), and the flatter the profile becomes. Conversely,
https://youtu.be/zB5g4aB1Sus
the geometry of the boundary layer changes. The greater the longitudinal

when the longitudinal speed gradient is negative, the boundary layer velocity
profile straightens up. When it becomes perfectly vertical at the wall, it
is such that streamlines separate from the wall: this is called separationw
(fig. 10.8).
The occurrence of separation can be predicted if we have a robust model
for the velocity profile inside the boundary layer. For this, we go back to
fundamentals, stating that at the separation point, the shear effort on
the surface must be zero:
𝜕𝑢
𝜏wall at separation = 0 = 𝜇
( 𝜕𝑦 )@𝑦=0
(10/25)

At the wall surface (𝑢 = 0 and 𝑣 = 0), equation 10/12 p. 204 becomes:

𝜕 2𝑢 d𝑝 d𝑈
𝜇 = = −𝜌𝑈
( (𝜕𝑦)2 )@𝑦=0 d𝑥 d𝑥
(10/26)

Thus, as we progressively increase the term d𝑝/ d𝑥, the term 𝜕 2 𝑢/(𝜕𝑦)2
reaches higher (positive) values on the wall surface. Nevertheless, we know
207
 and flowing towards a region of increasing pressure. For clarity, the 𝑦-scale is greatly
Figure 10.8: Separation of the boundary layer. The main flow is from left to right,
exaggerated.
Figure CC-by Olivier Cleynen

that it must take a negative value at the exterior boundary of the boundary

where 𝜕 2 𝑢/(𝜕𝑦)2 changes sign is called inflexion point.

layer. Therefore, it must change sign somewhere in the boundary. This point

at the wall (𝑦 = 0) the term 𝜕𝑢/𝜕𝑦 tends towards ever smaller values. Given
The existence of the inflection point within the boundary layer tells us that

enough distance 𝑥, it will reach zero value, and the boundary layer will
separate (fig. 10.9). Therefore, the longitudinal pressure gradient, which in
practice determines the longitudinal velocity gradient, is the key factor in
the analytical prediction of separation.
We shall remember two crucial points regarding the separation of boundary
layers:

1. Separation occurs in the presence of a positive pressure gradient,

which is sometimes named adverse pressure gradient.
Separation points along a wall (e.g. a car bodywork, an aircraft wing,

increases (positive d𝑝/ d𝑥 or negative d𝑈 / d𝑥). If pressure remains

rooftops, mountains) are always situated in regions where pressure

constant, or if it decreases, then the boundary layer cannot separate.

Figure 10.9: The inflection point within a boundary layer about to separate.
Figure CC-by Olivier Cleynen
208
 2. Laminar boundary layers are much more sensitive to separation
than turbulent boundary layers (fig. 10.10).
A widely-used technique to reduce or delay the occurrence of separa-
tion is to make boundary layers turbulent, using low-height artificial
obstacles positioned in the flow. By doing so, we increase shear-based
friction (which increases with turbulence) as a trade-off for better
resistance to stall.

Figure 10.10: The effect of decreasing Reynolds number on flow attachment over an
airfoil at constant angle of attack, with the transition point highlighted. Laminar
boundary layers are much more prone to separation than turbulent boundary layers.
Figure CC-by-sa Olivier Cleynen, based on Barlow & Pope 1999 [11]

Predicting in practice the position of a separation point is difficult, because

an intimate knowledge of the boundary layer profile and of the (external-
flow-generated) pressure field are required — and as the flow separates, these
are no longer independent. Resorting to experimental measurements, in this
case, is often a wise idea! Video: a football with a rough
edge on one side will see tur-
bulent boundary layer on that
side and laminar boundary layer
on the other. When the lami-
nar layer separates before the
turbulent one, the main flow
around the ball becomes asym-
metric and deviates the ball side-
ways, a phenomenon that can be
exploited to perform impressive
tricks.
by freekikerz (styl)
https://youtu.be/rzRuZNvkMbc

209
 10.7 Solved problems

Advertising board on a car

board above their car. They drive at 10 m s−1 ; the board is 3 m long and
A successful fluid dynamics professor advertises for their course using a

1,5 m high.

Drawing CC-by-sa by Imad Kharkouk

Will the boundary layer on the board become turbulent? How thick will
it become?

See this solution worked out step by step on YouTube

https://youtu.be/ML5dqjHeqPE (CC-by Olivier Cleynen)

Shear force on a board (laminar part)

In the example above, what is the shear force on the laminar layer part
of the board?

See this solution worked out step by step on YouTube

https://youtu.be/xxoqwzihRMc (CC-by Olivier Cleynen)

correct, but improperly calculated. The correct result is 0,107 N. Many thanks
Note: Unfortunately Olivier made an error in this video: the final expression is

to the students who double-checked and reported the problem!

Shear force on a board (turbulent part)

In the example above, what is the shear force on the turbulent layer part
of the board? And what would be the power lost to friction on the entire
board?

210
See this solution worked out step by step on YouTube
https://youtu.be/6i_yu1BKkVY (CC-by Olivier Cleynen)

211
212
 Problem sheet 10: Flow near walls
last edited July 9, 2020
by Olivier Cleynen — https://fluidmech.ninja/

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

In boundary layer flow, we assume that transition occurs at [Re]𝑥 & 5 ⋅ 105 .
The wall shear coefficient 𝑐𝑓 , a function of distance 𝑥,
is defined based on the free-stream flow velocity 𝑈 :
𝜏wall
𝑐𝑓(𝑥) ≡ 1
𝜌𝑈 2
(10/6)
2

Exact solutions to the laminar boundary layer along a smooth surface yield:

𝛿 4,91 𝛿∗ 1,72
=√ =√
𝑥 𝑥
(10/16)
[Re]𝑥 [Re]𝑥
𝛿 ∗∗
0,664 0,664
=√ 𝑐𝑓(𝑥) = √
𝑥
(10/18)
[Re]𝑥 [Re]𝑥

Solutions to the turbulent boundary layer along a smooth surface yield the following
time-averaged characteristics:

𝛿 0,16 𝛿∗ 0,02
≈ ≈
𝑥 [Re] 17 𝑥 [Re] 71
(10/22)
𝑥 𝑥
𝛿 ∗∗ 0,016 0,027
≈ 𝑐𝑓(𝑥) ≈
𝑥 1 1 (10/24)
[Re] 7 𝑥 [Re]𝑥7

Figure 10.11 quantifies the viscosity of various fluids as a function of temperature.

213
 2×10−2 2.4×10−5

10−2 2.2×10−5
9×10−3
8×10−3
7×10−3 ⟵ Crude Oil
6×10−3 Air ⟶
5×10−3
2×10−5
4×10−3
3×10−3
Viscosity 𝜇 of liquids in Pa s

Viscosity 𝜇 of gases in Pa s
2×10−3 1.8×10−5
CO2 ⟶

10−3 1.6×10−5
9×10−4
8×10−4
7×10−4
6×10−4
5×10−4
1.4×10−5
4×10−4 ⟵ Water
3×10−4

2×10−4 1.2×10−5

10−4 10−5
−20 0 20 40 60 80 100 120
Temperature 𝑇 in degree Celsius (◦C)

Figure 10.11: The viscosity of four fluids (crude oil, water, air, and CO2 ) as a function of tempera-
ture. The scale for liquids is logarithmic and displayed on the left; the scale for gases is linear and
displayed on the right.
Figure reproduced from figure 5.6 p. 99; CC-by by Arjun Neyyathala & Olivier Cleynen

10.1 Water and air flow

A flat plate of length 0,3 m is placed parallel to a uniform flow with speed 0,3 m s−1 . How
White [22] E7.2

thick can the boundary layer become:

10.1.1. if the fluid is air at 1 bar and 20 °C?

10.1.2. if the fluid is water at 20 °C?

10.2 Boundary layer sketches CC-0 Olivier Cleynen

A thin and long horizontal plate is moved horizontally through a stationary fluid.

10.2.1. Sketch the velocity profile of the fluid:

• at the leading edge;

214 • at a point where the boundary layer is laminar;
 • and at a point further downstream where the boundary layer is turbulent.

10.2.2. Draw a few streamlines, indicate the boundary layer thickness 𝛿, and the
displacement thickness 𝛿 ∗ .

10.2.3. Explain shortly (e.g. in 30 words or less) how the transition to turbulent regime
can be triggered.

10.2.4. Explain shortly (e.g. in 30 words or less) how the transition to turbulent regime
could instead be delayed.

10.3 Shear force due to boundary layer

A thin and smooth plate of dimensions 0,5 × 3 m is placed with a zero angle of attack in a
White [22] E7.3

flow incoming at 1,25 m s−1 , as shown in fig. 10.12.

Figure 10.12: A thin plate positioned parallel to an incoming uniform flow. Two configurations
are studied in this exercise.
Figure CC-0 Olivier Cleynen

10.3.1. What is the shear force exerted on the top surface of the plate for each of

𝜇air = 1,5 ⋅ 10−5 Pa s?

the two configurations shown in fig. 10.12, when the fluid is air of viscosity

10.3.2. What are the shear forces when the fluid is water of viscosity 𝜇water = 1 ⋅ 10−3 Pa s?

with an angle of 20° relative to the flow?

10.3.3. [difficult question] How would these shear efforts evolve if the plate was tilted

10.4 Wright Flyer I CC-0 Olivier Cleynen

a biplane with a 12 m wingspan (fig. 10.13). It had two wings of chord length 1,98 m
The Wright Flyer I, the first airplane capable of sustained controlled flight (1903), was

fly at very low angles of attack. Its flight speed was approximately 40 km h−1 .
stacked one on top of the other. The wing profile was extremely thin and it could only

10.4.1. If the flow over the wings can be treated as if they were flat plates, what is the
power necessary to compensate the shear exerted by the airflow on the wings
during flight?

10.4.2. Which other forms of drag would also be found on the aircraft? (give a brief
answer, e.g. in 30 words or less)

215
 Figure 10.13: The Wright Flyer I, first modern airplane in history. Built with meticulous care and
impeccable engineering methodology by two bicycle makers, it made history in December 1903.
Photo by Orville Wright, 1908 (public domain)

10.5 Power lost to shear on an airliner fuselageCC-0 Olivier Cleynen

An Airbus A340-600 (fig. 10.14) is cruising at [Ma] = 0,82 at an altitude of 10 000 m (where
the air has viscosity 1,457 ⋅ 10−5 N s m−2 , temperature 220 K, density 0,4 kg m−3 ).

Figure 10.14: The Airbus A340-600, a large airliner first flown in 2001.
Photo CC-by-sa by Iberia Airlines (retouched)

The cylindrical part of the fuselage has diameter 5,6 m and length 65 m.

10.5.1. What is approximately the maximum boundary layer thickness around the
fuselage?

10.5.2. What is approximately the average shear applying on the fuselage skin?

10.5.3. Estimate the power dissipated to friction on the cylindrical part of the fuselage.

10.5.4. In practice, in which circumstances could flow separation occur on the fuselage
skin? (give a brief answer, e.g. in 30 words or less)

10.6 Laminar wing profile

non-examinable. Based on a diagram from Bertin et al. 2010 [24]
The characteristics of a so-called “laminar” wing profile are compared in figs. 10.15

On the graph representing the pressure coefficient 𝐶𝑝 ≡

to 10.17 with those of an ordinary profile.
𝑝−𝑝∞
2 𝜌𝑉
1 2, identify the curve corre-
sponding to each profile.
What advantages and disadvantages does the laminar wing profile have, and how can
they be explained? In which applications will it be most useful?

216
Figure 10.15: Comparison of the thickness distribution of two uncambered wing profiles: an
ordinary medium-speed naca 0009 profile, and a “laminar” naca 66-009 profile.
Figure © Bertin & Cummings 2010 [24]

coefficient 𝐶𝑝 ≡ 𝑝−𝑝 2 ) as a function of distance 𝑥 (non-dimensionalized with the chord 𝑐) over

Figure 10.16: Static pressure distribution (represented as a the local non-dimensional pressure
∞
2 𝜌𝑉
1

the surface of the two airfoils shown in fig. 10.15.

Figure © Bertin & Cummings 2010 [24]

217
 Figure 10.17: Values of the section drag coefficient 𝐶𝑑 ≡ 𝑑
2 𝑐𝜌𝑉
1 2 as a function of the section lift
coefficient 𝐶𝑙 ≡ 𝑙
2 𝑐𝜌𝑉
1 2 for both airfoils presented in fig. 10.15.
Figure © Bertin & Cummings 2010 [24], based on data by Abott & Von Doenhoff 1949 [1]

10.7 Separation mechanism non-examinable, CC-0 Olivier Cleynen

Sketch the velocity profile of a laminar or turbulent boundary layer shortly upstream of,
and at a separation point.
The two equations below describe flow in laminar boundary layer:

𝜕𝑢 𝜕𝑢 d𝑈 𝜇 𝜕 2 𝑢
𝑢 +𝑣 =𝑈 +
𝜕𝑥 𝜕𝑦 d𝑥 𝜌 (𝜕𝑦)2
(10/12)
𝜕𝑢 𝜕𝑣
+ =0
𝜕𝑥 𝜕𝑦
(10/13)

Identify these two equations, list the conditions in which they apply, and explain shortly
(e.g. in 30 words or less) why a boundary layer cannot separate when a favorable pressure
gradient is applied along the wall.

218
Answers
10.1 1) At trailing edge [Re]𝑥 = 5 348 thus the layer is laminar everywhere. 𝛿 will grow
from 0 to 2,01 cm (eq. 10/15 p. 205);
2) For water: 𝛿trailing edge = 4,91 mm.

𝑦-direction is greatly exaggerated, and that the outer velocity 𝑈 is identical for
10.2 1) See fig. 10.6 p. 206. At the leading-edge the velocity is uniform. Note that the

both regimes;
2) See fig. 10.4 p. 202. Note that streamlines penetrate the boundary layer;
3) and 4) See §10.4 p. 205.

10.3 𝑥transition, air = 4,898 m and 𝑥transition, water = 0,4 m. In a laminar boundary layer, insert-
√ √ 𝑥transition
𝐹𝜏 = 0,664𝐿𝑈 1,5 𝜌𝜇 [ 𝑥 ]0
ing equation 10/18 into equation 10/6 into equation 10/5 yields
.

𝐹𝜏 = 0,01575𝐿𝜌 7 𝑈 7 𝜇 7 [𝑥 7 ]𝑥
6 𝑥trailing edge
In a turbulent boundary layer, we use equation 10/24 instead and get
6 13 1
. These expressions allow the calculation of the
transition
forces below, for the top surface of the plate:

1) (air) First case: 𝐹 = 3,445 ⋅ 10−3 N; second case 𝐹 = 8,438 ⋅ 10−3 N (who would
have thought eh?);

2) (water) First case: 𝐹 = 3,7849 N; second case 𝐹 = 2,7156 N.

10.4 Using the expressions developed in exercise 10.3, 𝑊̇ friction ≈ 255 W.

10.5 1) 𝑥transition = 7,47 cm (the laminar part is negligible). With the equations developed
in exercise 7.3, we get 𝐹 = 24,979 kN and 𝑊̇ = 6,09 MW. Quite a jump from the
Wright Flyer I!

2) When the longitudinal pressure gradient is zero, the boundary layer cannot
separate. Thus separation from the fuselage skin can only happen if the fuselage is
flown at an angle relative to the flight direction (e.g. during a low-speed maneuver).

219
220
 Fluid Dynamics
Chapter 11 – Large- and small-scale flows
last edited June 26, 2019
by Olivier Cleynen — https://fluidmech.ninja/

11.1 Motivation 221

11.2 Flow at large scales 221
11.2.1 Problem statement 221
11.2.2 Investigation of inviscid flows 222
11.3 Plotting velocity with functions 223
11.3.1 Kinematics that fit conservation laws 223
11.3.2 Strengths and weaknesses of potential flow 224
11.3.3 Superposition: the lifting cylinder 225
11.3.4 Circulating cylinder 227
11.3.5 Modeling lift with circulation 229
11.4 Flow at very small scales 232
11.5 Problems 235

11.1 Motivation
This exploratory chapter is not a critical component of fluid dynamics; in-
stead, it is meant as a brief overview of two extreme cases: flows for which
viscous effects are negligible, and flows for which they are dominant. This
exploration should allow us to answer two questions:

• How can we model large-scale flows?

• How can we model small-scale flows?

11.2 Flow at large scales

11.2.1 Problem statement

a representative length 𝐿 is very large. In particular, when 𝐿 is large enough,

In this section, we are interested in flow at very large scales: those for which

where the Reynolds number [Re] ≡ 𝜌𝑉 𝐿/𝜇 is very large.

the influence of viscosity is reduced. Formally, this corresponds to the case

To examine the mechanics of such a flow, we turn to our beloved non-

dimensional Navier-Stokes equation for incompressible flow derived as
eq. 8/13 p. 164,

𝜕 𝑉⃗ ∗ 1 1 ∗2 ⃗ ∗
+ [1] 𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ = ⃗ ∗𝑝∗ +
⃗∗ − [Eu] ∇
2 𝑔
⃗ 𝑉
∇
𝜕𝑡 ∗
[St]
[Fr] [Re]
(11/1)

We saw in chapter 8 (Engineering models) that we could compare the relative

weight of terms: when the Reynolds number [Re] is very large, the last term

221
 becomes negligible relative to the other four. Thus, our governing equation
can be reduced as follows:
𝜕 𝑉⃗ ∗ 1
+ [1] 𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ ≈ ⃗ ∗𝑝∗
𝑔⃗∗ − [Eu] ∇
𝜕𝑡 ∗
[Fr]2
[St] (11/2)

Now, converting eq. 11/2 back to dimensional terms, the governing momen-
tum equation for large-scale flow becomes:

D𝑉⃗
𝜌 ⃗𝑝
= 𝜌 𝑔⃗ − ∇
D𝑡
(11/3)

We see that with the starting proposition that [Re] was large, we have re-
moved altogether the viscous (last) term from the Navier-Stokes equation.
Flows governed by this equation are called inviscid flows. Equation 11/3 is
named the Euler equation; it stipulates that the acceleration field is driven
only by gravity and by the pressure field.

11.2.2 Investigation of inviscid flows

From what we studied in chapter 8 (Engineering models), we recognize immedi-
ately that flows governed by eq. 11/3 are troublesome: the absence of viscous
effects facilitates the occurrence of turbulence and makes for much more
chaotic behaviors. Although the removal of shear from the Navier-Stokes
equation simplifies the governing equation, the solutions to this new equation
become even harder to find and describe.
What can be done in the other two branches of fluid mechanics?

• Large-scale flows are difficult to investigate experimentally. As we

have seen in chapter 8 (Engineering models), scaling down a flow (e.g. so it
may fit inside a laboratory) while maintaining constant [Re] requires
increasing velocity by a corresponding factor.

• Large-scale flows are also difficult to investigate numerically. At high

[Re], the occurrence of turbulence makes for either an exponential
3,5
increase in computing power (we saw in chapter 9 that direct numerical
simulation computing power increases with [Re] ), or for increased
reliance on hard-to-calibrate turbulence models (in Reynolds-averaged
simulations).

All three branches of fluid mechanics, therefore, struggle with large-scale

flows, because of turbulence.
In spite of this, large-scale flows are undeniably important. In chapter 10 we
were able to understand and describe fluid flow very close to walls. Now we
wish to be able to to the same for large structures, for example, in order to
describe the broad patterns of fluid flow (in particular, pressure distribution)
in the wake of an aircraft, around a wind turbine, or within a hurricane. It is
clear that we have no hope of accounting easily for turbulence, but we can
at least describe the main features of such flows by restricting ourselves to
laminar cases. In the following sections, we will model such laminar solutions
directly, based on intuition and observation, and make sure that they match
the condition described by eq. 11/3 above.

222
11.3 Plotting velocity with functions

11.3.1 Kinematics that fit conservation laws

For simple flow structures, the velocity field can be simply described based
on observation and intuition. This is done for example in exercise 11.6, where
we reconstruct the flow field within and around a tornado using very simple,
almost primitive, kinematics.
Without so much as a small increase in complexity, this approach becomes
untenable. In the last exercises of chapter 6 (ex. 6.8 & 6.9 p. 133) we have
seen that it is easy to propose a velocity field that does not satisfy (is not
a solution of) either the mass balance equation or the momentum balance
equation. For example, if one considers two of the tornado flows mentioned
above together, the velocity field cannot be described easily anymore.
One approach has been developed in the 17th century to overcome this
problem. It consists in finding a family of flows, all steady, that always satisfy
the conservation equations. Those flows can then be added to one another
to produce new flows which satisfy the balance equations. Such flows are
called potential flows.
Two conditions need to be fulfilled for this approach to work:

there must correspond a single value of 𝑢, of 𝑣 and of 𝑤 at each of

1. The velocity field must always be describable with a function; that is,

the coordinates 𝑥𝑖 , 𝑦𝑖 , or 𝑧𝑖 in space. This means in practice that we

cannot account for flows which “curl up” on themselves, occasionally
recirculating back on their path.

⃗ ⋅ 𝑉⃗ = 0 (eq. 6/35 p. 123) is

is achieved if the continuity equation ∇
2. The velocity field must conserve mass. In incompressible flow, this

respected.

With potential flow, these two conditions are addressed as follows:

1. We restrict ourselves to irrotational flows, those in which the curl of

velocity (see Appendix A3 p. 250) is always null:

⃗ × 𝑉⃗ = 0⃗
∇ (11/4)
by definition, for an irrotational flow.

function 𝜙 (pronounced “phi” and named potential function) of which

It can be shown that flows are irrotational when there exists a scalar

the gradient is the velocity vector field:

⃗ 𝜙 ≡ 𝑉⃗
∇ (11/5)

In the case of two-dimensional flow, this translates as:

𝜕𝜙
≡𝑢
𝜕𝑥
𝜕𝜙
(11/6)

≡𝑣
𝜕𝑦
(11/7)

When plotted out, lines of constant 𝜙 (named equipotential lines) are

always perpendicular to the streamlines of the flow. 223
 2. The continuity equation is satisfied by referring to stream functions. It

vector field function 𝜓⃗ (pronounced “psi” and named stream function)

can be shown that the divergent of velocity is null when there exists a

of which the curl is the velocity vector field:

⃗ × 𝜓⃗ = 𝑉⃗
∇ (11/8)

In the case of two-dimensional flow, 𝜓 is a scalar field and eq. 11/8

translates as:
𝜕𝜓
≡𝑢
𝜕𝑦
(11/9)
𝜕𝜓
− ≡𝑣
𝜕𝑥
(11/10)

When plotted out, lines of of constant 𝜓 value are streamlines – in other

words, as they travel along, fluid particles follow paths of constant 𝜓
value.

In summary, we have shifted the problem from looking for 𝑢 and 𝑣, to looking
for 𝜓 and 𝜙. The existence of such functions ensures that flows can be added
and subtracted from one another yet will always result in mass-conserving,
mathematically-describable flows. If such two functions are known, then the
velocity components can be obtained (recovered) easily either in Cartesian
coordinates,
𝜕𝜙 𝜕𝜓
𝑢= =
𝜕𝑥 𝜕𝑦
(11/11)
𝜕𝜙 𝜕𝜓
𝑣= =−
𝜕𝑦 𝜕𝑥
(11/12)

or angular coordinates:
𝜕𝜙 1 𝜕𝜓
𝑣𝑟 = =
𝜕𝑟 𝑟 𝜕𝜃
1 𝜕𝜙 𝜕𝜓
(11/13)

𝑣𝜃 = =−
𝑟 𝜕𝜃 𝜕𝑟
(11/14)

11.3.2 Strengths and weaknesses of potential flow

The potential flow methodology allows us to find solutions to the Euler
equation: flows in which the Reynolds number is high enough that viscosity
has no significant role anymore. Potential flows are the simplest solutions
that we are able to come up with. They are:

• strictly steady;

• inviscid;

• incompressible;

• devoid of energy transfers;

• two-dimensional (in the scope of this course at least).

This is quite convenient for the academician, who recognizes immediately

224 four of the five criteria which we set forth in chapter 2 (Analysis of existing
 for using the Bernoulli equation. Along a streamline,
flows with one dimension)
the fifth condition is met, and Euler’s equation reduces to eq. 2/20 (p. 42),
reproduced here:
𝑝1 1 2 𝑝2 1 2
+ 𝑉1 + 𝑔𝑧1 = + 𝑉 + 𝑔𝑧2 = cst.
𝜌 2 𝜌 2 2
(11/15)

along a streamline in a steady incompressible inviscid flow.

and so it follows that if the solution to a potential flow is known, the pressure
is known everywhere, and the forces due to pressure can be calculated with
relative ease.
Nevertheless, from a science and engineering point of view, potential flows
have only limited value, because they are entirely unable to account for
turbulence, which we have seen is an integral feature of high-[Re] flows.
We should therefore use them only with great caution. Potential flows help
us model large-scale structures with very little computational cost, but this
comes with strong limitations.

11.3.3 Superposition: the lifting cylinder

It is possible to describe a handful of basic potential flows called elementary
flows as fundamental ingredients that can be added to one another to create
more complex and interesting flows. Without going into much detail, the
most relevant elementary flows are:

• Uniform longitudinal flow,

𝜙 = 𝑉 𝑟 cos 𝜃
𝜓 = 𝑉 𝑟 sin 𝜃
(11/16)
(11/17)

of a (positive or negative) volume flow rate ̇ from a single point in

• Sources and sinks (fig. 11.1) which are associated with the appearance

̇
the flow:

𝜙= ln 𝑟
2𝜋
̇
(11/18)

𝜓= 𝜃
2𝜋
(11/19)

Figure 11.1: Concept of a source inside a two-dimensional potential flow. A sink

would display exactly opposed velocities.
Figure CC-by-sa Commons User:Nicoguaro

225
 velocity 𝑣𝜃 = 𝑓 (𝑟, 𝜃) on the flow, in addition to which there may exist a
• Irrotational vortices, rotational patterns which impart a rotational

radial component 𝑣𝑟 :
Γ
𝜙= 𝜃
2𝜋
Γ
(11/20)

𝜓 = − ln 𝑟
2𝜋
(11/21)

in which Γ (termed circulation) is a constant proportional to the strength of

the vortex.

• Doublets, which consist in a source and a sink of equal volume flow

rate positioned extremely close one to another (fig. 11.2):
cos 𝜃
𝜙=𝐾
𝑟
sin 𝜃
(11/22)

𝜓 = −𝐾
𝑟
(11/23)
̇
in which 𝐾 is a constant proportional to the source/sink volume flow rate .

It was found in the 17th Century that combining a doublet with uniform flow
resulted in flow patterns that imiated “perfect” flow around a cylinder: a
flow where the fluid flows smoothly and steadily everywhere (fig. 11.3). The
stream function of that flow is:
𝑅2
𝜓 = 𝑈∞ sin 𝜃 𝑟 −
( 𝑟 )
(11/24)

This stream function allows us to describe the velocity everywhere:

1 𝜕𝜓 𝑅2
𝑣𝑟 = = 𝑈∞ cos 𝜃 1 − 2
𝑟 𝜕𝜃 ( 𝑟 )
(11/25)

𝜕𝜓 𝑅2
𝑣𝜃 = − = −𝑈∞ sin 𝜃 1 + 2
𝜕𝑟 ( 𝑟 )
(11/26)

Indeed, along the cylinder wall, 𝑟 = 𝑅 and

We can even calculate the lift and drag applying on the cylinder surface.

𝑣𝑟 |𝑟=𝑅 = 0
𝑣𝜃 |𝑟=𝑅 = −2𝑈∞ sin 𝜃
(11/27)
(11/28)

Since the Bernoulli equation can be applied along any streamline in this
(steady, constant-energy, inviscid, incompressible) flow, we can express the

Figure 11.2: Left: a simple uniform steady flow; Right: a doublet, the result of a
source and a sink brought very close one to another
Figure CC-0 Olivier Cleynen
226
Figure 11.3: The addition of a doublet and a uniform flow produces streamlines for
an (idealized) flow around a cylinder.
Figure CC-by-sa by Commons User:Kraaiennest

pressure 𝑝s on the cylinder surface as a function of 𝜃:

1 1
𝑝∞ + 𝜌𝑈∞2 = 𝑝s + 𝜌𝑣𝜃2
2 2
1
𝑝s (𝜃) = 𝑝∞ + 𝜌 (𝑈∞2 − 𝑣𝜃2 )
2
(11/29)

fluid on the cylinder per unit width 𝐿, in each of the two directions 𝑥 and 𝑦:
Now, a relatively simple integration gives us the net forces exerted by the

𝐹net,𝑥 2𝜋
= −∫ 𝑝s cos 𝜃 𝑅 d𝜃 = 0
𝐿
(11/30)
0
𝐹net,𝑦 2𝜋
= −∫ 𝑝s sin 𝜃 𝑅 d𝜃 = 0
𝐿
(11/31)
0

The results are interesting, and at the time they were obtained by their author,
Jean le Rond D’Alembert, were devastating: both lift and drag are zero. This
inability to reproduce the well-known phenomena of drag is often called the
d’Alembert paradox.
To find out why the solution is not realistic, we can plot the resulting surce
pressure distribution graphically, and compared to experimental measure-
ments: this is done in fig. 11.4. Good agreement is obtained on the leading
edge of the cylinder; but as the pressure gradient becomes unfavorable, in
practice the boundary layer separates –a phenomenon that cannot be de-
scribed with inviscid flow— and a low-pressure area forms on the downstream
side of the cylinder.

227
 Figure 11.4: Pressure distribution (relative to the far-flow pressure) on the surface of
a cylinder, with flow from left to right. On the left is the potential flow case, purely
symmetrical. On the right (in blue) is a measurement made at a high Reynolds
number. Boundary layer separation occurs on the second half of the cylinder, which
prevents the recovery of leading-edge pressure values, and increases drag.
Figure CC-by-sa Commons User:BoH & Olivier Cleynen

11.3.4 Circulating cylinder

cylinder flow above if an irrotational vortex of stream function 𝜓 = − 2𝜋Γ ln 𝑟

An extremely interesting “hack” can be implemented with the potential

(eq. 11/21) is added to it. The overall flow field becomes:

𝑅2 Γ
𝜓 = 𝑈∞ sin 𝜃 𝑟 − − ln 𝑟
( 𝑟 ) 2𝜋
(11/32)
Video: watch the Brazilian foot-
ball team show their French
counterparts how circulation (in-
duced through friction by ball ro-
With this function, several key characteristics of the flow field can be obtained.
tation) is associated to dynamic The first is the velocity field at the cylinder surface:

𝑣𝑟 |𝑟=𝑅 = 0
lift on a circular body

Γ
by TF1, 1997 (styl)
(11/33)
𝑣𝜃 |𝑟=𝑅 = −2𝑈∞ sin 𝜃 +
https://youtu.be/oGeMZ3t8jn4

2𝜋𝑅
(11/34)

and we immediately notice that the velocity distribution is no longer sym-

metrical with respect to the horizontal axis (fig. 11.5): the fluid is deflected,
and so there will be a net force on the cylinder.
This time, the net pressure forces on the cylinder have changed:

𝐹net,𝑥 2𝜋
= −∫ 𝑝s cos 𝜃 𝑅 d𝜃 = 0
𝐿
(11/35)
0
𝐹net,𝑦 2𝜋
= −∫ 𝑝s sin 𝜃 𝑅 d𝜃 = −𝜌 𝑈∞ Γ
𝐿
(11/36)
0

flow— but that lift occurs which is proportional to the free-stream velocity 𝑈
We thus find out that the drag is once again zero —as for any potential

and to the circulation Γ.

In practice, such a flow can be generated by spinning a cylindrical object
in a uniform flow. A lateral force is then obtained, which can be used as a
propulsive or sustaining force. Several boats and even an aircraft have been
used in practice to demonstrate this principle. Naturally, flow separation
from the cylinder profile and the high shear efforts generated on the surface
228
Figure 11.5: The addition of an irrotational vortex on top of the cylinder flow
is developed, which depends directly on the circulation Γ.
described in fig. 11.3 distorts the flow field and it becomes asymmetrical: a lift force

Figure © White 2008 [22]

cause real flows to differ from the ideal case described here, and it turns out
that rotating cylinders are a horribly uneconomical and unpractical way of
generating lift.

11.3.5 Modeling lift with circulation

Fluid flow around cylinders may have little appeal for the modern student
of fluid mechanics, but the methodology above has been taken much further.
With further mathematical manipulation called conformal mapping, potential
flow can be used to described flow around geometrical shapes such as airfoils
(fig. 11.6). Because the flow around such streamlined bodies usually does
not feature boundary layer separation, the predicted flow fields everywhere Video: acting on swimming pool
water with a round plate sheds
except in the close vicinity of the solid surface are accurately predicted. a half-circular vortex that is ex-
tremely stable and can be inter-
There again, it is observed that regardless of the constructed geometry, no lift acted with quite easily. Such sta-
ble laminar structures are excel-
can be modeled unless circulation is also added within the flow. The amount lent candidates for analysis us-
ing potential flow.
of circulation needed so that results may correspond to experimental obser- by Physics Girl (Dianna Cowern) (styl)
https://youtu.be/pnbJEg9r1o8
vations is found by increasing it progressively until the the rear stagnation
point reaches the rear trailing edge of the airfoil, a condition known as the
Kutta condition. Regardless of the amount of circulation added, potential
flow remains entirely reversible, both in a kinematic and a thermodynamic
sense, thus, care must be taken in the problem setup to make sure the model
is realistic (fig. 11.7).
It is then observed in general that any dynamic lift generation can be modeled
as the superposition of a free-stream flow and a circulation effect (fig. 11.8).
With such a tool, potential flow becomes an extremely useful tool, mathemat-
ically and computationally inexpensive, in order to model and understand the
cause and effect of dynamic lift in fluid mechanics. In particular, it has been
paramount in the description of aerodynamic lift distribution over aircraft
wing surfaces (fig. 11.9), with a concept called the Lifting-line theory.

229
 Figure 11.6: Potential flow around an airfoil without (top) and with (bottom) circu-
can only result in a net vertical force if an irrotational vortex (with circulation Γ) is
lation. Much like potential flow around a cylinder, potential flow around an airfoil
added on top of the flow. Only one value for Γ will generate a realistic flow, with the
rear stagnation point coinciding with the trailing edge, a occurrence named Kutta
condition.
Figure CC-0 Olivier Cleynen

Figure 11.7: Potential flow allows all velocities to be inverted without any change in
the flow geometry. Here the flow around an airfoil is reversed, displaying unphysical
behavior.
Figure CC-0 Olivier Cleynen

Figure 11.8: A numerical model of flow around an airfoil. In the left figure, the
velocity vectors are represented relative to a stationary background. In the right
figure, the velocity of the free-stream flow has been subtracted from each vector,
bringing the circulation phenomenon into evidence.
Figures 1 & 2 CC-by-sa by en:Wikipedia User:Crowsnest

230
Figure 11.9: From top to bottom, the lift distribution over the wings of a glider is
modeled with increasingly complex (and accurate) lift and circulation distributions
along the span. The Lifting-line theory is a method associating each element of
lift with a certain amount of circulation. The effect of each span-wise change of
circulation is then mapped onto the flow field as a trailing vortex.
Figures 1, 2 & 3 CC-by-sa Olivier Cleynen

231
 11.4 Flow at very small scales
At the complete opposite of the spectrum, we find flow at very small scales:

flows the representative length 𝐿 is extremely small, which makes for small
flows around bacteria, dust particles, and inside very small ducts. In those

values of the Reynolds number. Such flows are termed creeping or Stokes
flows. What are their main characteristics?
Looking back once again at the non-dimensional Navier-Stokes equation for
incompressible flow derived as eq. 8/13 p. 164,

𝜕 𝑉⃗ ∗ ⃗∗ ⋅ ∇ 1 1 ∗2 ⃗ ∗
+ [1] 𝑉 ⃗ ∗ 𝑉⃗ ∗ = ⃗
𝑔 ∗
− ⃗ ∗𝑝∗ +
∇ ⃗ 𝑉
∇
𝜕𝑡 ∗ 2
[St] [Eu] (11/37)
[Fr] [Re]

⃗ ∗2 𝑉⃗ ∗ then becomes
smaller than 1. The relative weight of the term (1/[Re]) ∇
we see that creeping flow will occur when the Reynolds number is much

In addition to cases where [Re] ≪ 1, we focus our interest on flows for which:
overwhelming.

• gravitational effects have negligible influence over the velocity field;

• the characteristic frequency is extremely low (quasi-steady flow).

With these characteristics, the terms associated with the [St] (Strouhal) and
[Fr] (Froude) numbers become very small with respect to the other terms, and
our non-dimensionalized Navier-Stokes equation (eq. 11/37) is approximately

1 ∗2 ⃗ ∗
reduced to:

0⃗ ≈ −[Eu] ∇
⃗ ∗𝑝∗ + ⃗ 𝑉
∇ (11/38)
[Re]

We can now come back to dimensionalized equations, concluding that for a

fluid flow dominated by viscosity, the pressure and velocity fields are linked
together by the approximate relation:

∇ ⃗ 2 𝑉⃗
⃗ 𝑝 = 𝜇∇ (11/39)

In this type of flow, the pressure field is entirely dictated by the Laplacian

which the representative length 𝐿 is very small, spend their lives in such
of velocity, and the fluid density has no importance. Micro-organisms, for

flows (fig. 11.10). At the human scale, we can visualize the effects of these
flows by moving an object slowly in highly-viscous fluids (e.g. a spoon in
honey), or by swimming in a pool filled with plastic balls. The inertial effects
are almost inexistent, drag is extremely important, and the object geometry
has comparatively small influence.
In 1851, George Gabriel Stokes worked through equation 11/39 for flow

allowed him to show that the drag 𝐹D sphere applying on a sphere of diameter
around a sphere, and obtained an analytical solution for the flow field. This

𝐷 in creeping flow (fig. 11.11) is:

𝐹D sphere = 3𝜋𝜇𝑈∞ 𝐷 (11/40)

232
Reynolds numbers, since their scale 𝐿 is very small. For them, viscosity effects
Figure 11.10: Micro-organisms carry themselves through fluids at extremely low
dominate inertial effects.
Photo by Yutaka Tsutsumi, M.D., Fujita Health University School of Medicine

Inserting this equation 11/40 into the definition of the drag coefficient 𝐶𝐹 D ≡
𝐹D / 12 𝜌𝑆frontal 𝑈∞2 (from eq. 8/15 p. 169) then yields:

𝐹D sphere 24𝜇 24
𝐶𝐹 D = = =
1
𝜌𝑈∞2 𝜋4 𝐷 2 𝜌 𝑈∞ 𝐷 [Re]𝐷
(11/41)
2

These equations are specific to flow around spheres, but the trends they
describe apply well to most bodies evolving in highly-viscous flows, such
as dust or liquid particles traveling through the atmosphere. Drag is only
proportional to the speed (as opposed to low-viscosity flows in which it
grows with velocity squared), and it does not depend on fluid density.

Figure 11.11: Flow at very low Reynolds numbers around a sphere. In this regime,
the drag force is proportional to the velocity.
Figure CC-by-sa by Olivier Cleynen & Commons User:Kraaiennest
233
234
 Problem sheet 11: Large- and small-scale flows
last edited June 26, 2019
by Olivier Cleynen — https://fluidmech.ninja/

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 °C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Except otherwise indicated, assume that:

Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1

In a highly-viscous (creeping) steady flow, the drag 𝐹D exerted on a spherical body of

diameter 𝐷 at by flow at velocity 𝑈∞ is quantified as:

𝐹D sphere = 3𝜋𝜇𝑈∞ 𝐷 (11/40)

11.1 Volcanic ash from the Eyjafjallajökull Çengel & al. [25] E10.2
In 2010, a volcano with a complicated name and unpredictable mood decided to ground

We consider a microscopic ash particle released at very high altitude (−50 °C, 0,55 bar,
the entire European airline industry for five days.

1,474 ⋅ 10−5 N s m−2 ). We model it as a sphere with 50 µm diameter. The density of

volcanic ash is 1 240 kg m−3 .

11.1.1. What is the terminal velocity of the particle?

11.1.2. Will this terminal velocity increase or decrease as the particle progresses towards
the ground? (briefly justify your answer, e.g. in 30 words or less)

11.2 Water drop Çengel & al. [25] 10-21

water drop with diameter 42,4 µm is falling through air at 25 °C and 1 bar.
A rainy day provides yet another opportunity for exploring fluid dynamics (fig. 11.12). A

11.2.1. Which terminal velocity will it reach?

11.2.2. Which velocity will it reach once its diameter will have doubled?

235
 fall. When their diameter is lower than 2 mm, water drops are approximately spherical (B). As
Figure 11.12: A sketched diagram showing the geometry of water drops of various sizes in free
they grow beyond this size, their shape changes and they eventually break-up (C-E). They never
display the “classical” shape displayed in A, which is caused only by surface tension effects when
they drip from solid surfaces.
Figure CC-by-sa by Ryan Wilson

11.3 Idealized flow over a hangar roof based on White [22] P8.54
Certain flows in which both compressibility and viscosity effects are negligible can be
described using the potential flow assumption (the hypothesis that the flow is everywhere
irrotational). If we compute the two-dimensional laminar steady fluid flow around a
cylinder profile, we obtain the velocities in polar coordinates as:

𝑅2
𝑣𝑟 = 𝑉∞ cos 𝜃 1 − 2
( 𝑟 )
(11/25)

𝑅2
𝑣𝜃 = −𝑉∞ sin 𝜃 1 + 2
( 𝑟 )
(11/26)

where the origin (𝑟 = 0) is at the center of the cylinder profile;

𝜃 is measured relative to the free-stream velocity vector;
𝑉∞ is the incoming free-stream velocity;
and 𝑅 is the (fixed) cylinder radius.

In this exercise, we study the air flow over a hangar roof with this model. We use the
equations above to describe the air velocity everywhere, pretending the as the wind
blows about a large semi-cylindrical solid structure — an idealized description of an

Wind with a nearly-uniform velocity 𝑈∞ = 100 km h−1 is blowing across a 50 m-long

otherwise complex flow.

is 𝑅 = 20 m.
hangar with a semi-cylindrical geometry, as shown in fig. 11.13. The radius of the hangar

236
Figure 11.13: A semi-cylindrical hangar roof. Wind with uniform velocity 𝑈 flows perpendicular
to the cylinder axis.
Figure CC-0 Olivier Cleynen

11.3.1. Starting from eqs. 11/25 and 11/26, show that the pressure 𝑝𝑠 on the surface on

1
the roof is distributed as:

𝑝𝑠 = 𝑝∞ + 𝜌 (𝑉∞2 − 4𝑉∞2 sin2 𝜃 )

2
(11/42)

11.3.2. The pressure inside the hangar is set to 𝑝∞ . What is the total lift force on the
hangar?

(a couple of hints to help with the algebra: ∫ sin 𝑥 d𝑥 = − cos 𝑥+𝑘 and ∫ sin3 𝑥 d𝑥 =
(see also problem 4.6 p. 87)

1
3
cos3 𝑥 − cos 𝑥 + 𝑘).

11.3.3. At which position on the roof is the 𝑝𝑠 = 𝑝∞ ?

11.3.4. Describe briefly (e.g. in 30 words or less) two reasons why the results above
would not correspond to reality.

11.4 Cabling of the Wright Flyer derived from Munson & al. [29] 9.106
The Wright Flyer I, the first powered and controlled aircraft in history, was subjected
to multiple types of drag. We have already studied viscous friction on its thin wings in
exercise 7.4. The data in figure 11.14 provides the opportunity to quantify drag due to

A network of metal cables with diameter 1,27 mm criss-crossed the aircraft in order to
pressure.

which came at 40 km h−1 . The total cable length was approximately 60 m.

provide structural rigidity. The cables were positioned perpendicularly to the air flow,

What was the drag generated by the cables?

237
 Figure 11.14: Experimental measurements of the drag coefficient applying to a cylinder and to a
sphere as a function of the diameter-based Reynolds number [Re]𝐷 , shown together with schematic
depictions of the flow around the cylinder. By convention, the drag coefficient 𝐶D ≡ 𝐶𝐹 D ≡ 1 𝜌𝑆𝑈
𝐹D
2

(eq. 8/15 p. 169) compares the drag force 𝐹D with the frontal area 𝑆.
2 ∞

Both figures © from Munson & al.[29]

238
11.5 Ping pong ball Munson & al. [29] E9.16
A series of experiments is conducted in a wind tunnel on a large cast iron ball with a
smooth surface; the results are shown in fig. 11.15. These measurement data are used to

the ball to 2,7 g and its diameter to 40 mm.

predict the behavior of a ping pong ball. Table tennis regulations constrain the mass of

11.5.1. Is it possible for a ball thrown at a speed of 50 km h−1 to have a perfectly

horizontal trajectory?

11.5.2. If so, what would be its deceleration?

11.5.3. How would the drag and lift applying on the ball evolve if the air viscosity was
progressively decreased to zero?

Figure 11.15: Experimental measurements of the lift and drag coefficients applying on a rotating
sphere in an steady uniform flow.
Figure © from Munson & al.[29]
239
 11.6 Flow field of a tornado Çengel & al. [25] E9-5, E9-14 & E10-3
In this problem, we attempt to model a very large-scale flow: that of a tornado (fig. 11.16).
We begin by pretending the tornado is one perfectly straight, stationary structure. We
divide the flow into two regions: a core cylinder that rotates almost like a solid body,
and an outer region where flow spins in an irrotational matter. This model is called the
Rankine vortex (displayed in fig. 11.17) and is used widely as a simple, first approximation
to model flows as large as a hurricane and as small as turbulence-induced vortices.

Figure 11.16: Photo of an approaching tornado in Manitoba, Canada

Photo CC-by-sa by Commons User:Grhu

Figure 11.17: Modeled angular velocity in a vortex, according to the Rankine vortex model
Figure CC-by-sa by en:Wikipedia User:Justin1569

as being steady, two-dimensional (neglecting any movement in the vertical, 𝑧-direction),

We are first interested in the outer region of the tornado flow field. We model the flow

and having a rotational velocity 𝑣𝜃 such that:

Γ
𝑣𝜃 =
2𝜋𝑟
(11/43)

in which Γ is the circulation (measured in s−1 ) and remains constant and uniform.

11.6.1. The mass balance equation for incompressible flow (eq. 6/35 p. 123) is developed
in cylindrical coordinates as follows:
1 𝜕𝑟𝑣𝑟 1 𝜕𝑣𝜃 𝜕𝑣𝑧
+ + =0
𝑟 𝜕𝑟 𝑟 𝜕𝜃 𝜕𝑧
(11/44)

According to this mass balance equation, what form must the radial velocity 𝑣𝑟
240 take?
Among all the possibilities for 𝑣𝑟 , we choose the simplest form, so that from now on, we
model radial velocity as:

𝑣𝑟 = 0 (11/45)

11.6.2. The momentum balance equation for incompressible flow (eq. 6/42 p. 125) is
developed in cylindrical coordinates are as follows:

𝜕𝑣𝑟 𝜕𝑣𝑟 𝑣𝜃 𝜕𝑣𝑟 𝑣𝜃2 𝜕𝑣𝑟

𝜌 + 𝑣𝑟 + − + 𝑣𝑧
[ 𝜕𝑡 𝜕𝑟 𝑟 𝜕𝜃 𝑟 𝜕𝑧 ]
𝜕𝑝 1 𝜕 𝜕𝑣𝑟 𝑣𝑟 1 𝜕 2 𝑣𝑟 2 𝜕𝑣𝜃 𝜕 2 𝑣𝑟
= 𝜌𝑔𝑟 − +𝜇 𝑟 − 2 + 2 − +
𝜕𝑟 [ 𝑟 𝜕𝑟 ( 𝜕𝑟 ) 𝑟 𝑟 (𝜕𝜃)2 𝑟 2 𝜕𝜃 (𝜕𝑧)2 ]

𝜕𝑣𝜃 𝜕𝑣𝜃 𝑣𝜃 𝜕𝑣𝜃 𝑣𝑟 𝑣𝜃 𝜕𝑣𝜃

(11/46)
𝜌 + 𝑣𝑟 + + + 𝑣𝑧
[ 𝜕𝑡 𝜕𝑟 𝑟 𝜕𝜃 𝑟 𝜕𝑧 ]
1 𝜕𝑝 1 𝜕 𝜕𝑣𝜃 𝑣𝜃 1 𝜕 2 𝑣𝜃 2 𝜕𝑣𝑟 𝜕 2 𝑣𝜃
= 𝜌𝑔𝜃 − +𝜇 𝑟 − 2 + 2 + +
𝑟 𝜕𝜃 [ 𝑟 𝜕𝑟 ( 𝜕𝑟 ) 𝑟 𝑟 (𝜕𝜃)2 𝑟 2 𝜕𝜃 (𝜕𝑧)2 ]

𝜕𝑣𝑧 𝜕𝑣𝑧 𝑣𝜃 𝜕𝑣𝑟 𝜕𝑣𝑧

(11/47)
𝜌 + 𝑣𝑟 + + 𝑣𝑧
[ 𝜕𝑡 𝜕𝑟 𝑟 𝜕𝜃 𝜕𝑧 ]
𝜕𝑝 1 𝜕 𝜕𝑣𝑧 1 𝜕 2 𝑣𝑧 𝜕 2 𝑣𝑧
= 𝜌𝑔𝑧 − +𝜇 𝑟 + 2 +
𝜕𝑧 [ 𝑟 𝜕𝑟 ( 𝜕𝑟 ) 𝑟 (𝜕𝜃)2 (𝜕𝑧)2 ]
(11/48)

Starting from those equations, show that the pressure distribution in the outer
region of the tornado can be expressed as:
1 1
𝑝 = 𝑝∞ − 𝜌 Γ2 2
2 𝑟
(11/49)

where 𝑝∞ is the atmospheric pressure far away from the tornado.

We now turn to the core of the tornado, which we model as if it were a rotating solid (a
vortex core).

11.6.3. What is the radial velocity 𝑣𝜃 distribution?

11.6.4. What is the pressure field within the rotational core of the tornado?
(hint: you may start directly from an energy balance equation, eq. 2/18 p. 40,
without having to use the Navier-Stokes equations above).
11.6.5. Make a simple, qualitative sketch (i.e. without numerical data) of the pressure
as a function of radius throughout the entire tornado flow field.

be 50 m and the maximum wind velocity to be 180 km h−1 .

It is finally time to calibrate and exploit our model. We estimate the tornado diameter to

11.6.6. According to the model, what is the lowest pressure attained by the air?

50 km h−1 ?
11.6.7. According to the model, at what distance from the core are winds lower than

(curious students may play with the above model by adding a non-zero radial velocity,
and look up the phenomenon of vortex stretching) 241
 11.7 Lift on a symmetrical object non-examinable
Briefly explain (e.g. with answers 30 words or less) how lift can be generated on a sphere
or a cylinder,

• with differential control boundary layer control;

• with the effect of rotation.

Draw a few streamlines in a two-dimensional sketch of the phenomenon.

11.8 Air flow over a wing profile From Munson & al. [29] 9.109
The characteristics of a thin, flat-bottomed airfoil are examined by a group of students
in a wind tunnel. The first investigations focus on the boundary layer, and the research
group evaluate the boundary layer thickness and make sure that it is fully attached.

Measurements of the longitudinal speed 𝑢 just above the boundary layer on the top
Once this is done, the group proceeds with speed measurements all around the airfoil.

surface are tabulated below:

𝑥/𝑐 (%) 𝑦/𝑐 (%) 𝑢/𝑈

0 0 0
2,5 3,72 0,971
5 5,3 1,232
7,5 6,48 1,273
10 7,43 1,271
20 9,92 1,276
30 11,14 1,295
40 10,49 1,307
50 10,45 1,308
60 9,11 1,195
70 6,46 1,065
80 3,62 0,945
90 1,26 0,856
100 0 0,807

On the bottom surface, the speed is measured as being constant (𝑢 = 𝑈 ) to within

experimental error.
What is the lift coefficient of the airfoil?

242
Answers

to obtain 𝑈 = 𝑔𝜌sphere 18𝜇 = 0,1146 m s−1 : unbearably slow when you are stuck in an
11.1 1) At terminal velocity, the weight of the sphere equals the drag. This allows us
𝐷2

airport! With 𝑈 , check that the Reynolds number indeed corresponds to creeping
flow: [Re]𝐷 = 0,334.

11.2 Same as previous exercise: 𝑈1 = 4,578 ⋅ 10−2 m s−1 and 𝑈2 = 0,183 m s−1 , with Reynolds
numbers of 0,113 and 0,906 respectively (thus creeping flow hypothesis valid).

11.3 1) Integrate the vertical component of force due to pressure: 𝐹L roof = 1,575 MN.

11.4 A simple reading of fig. 11.14 gives 𝐹D = 6,9 N, 𝑊̇ = 76 W.

11.5 Yes — a reading of fig. 11.13 gives 𝜔 = 83 rev/s.

243
244
 Appendix
last edited February 12, 2021

A1 Notation 246
A2 Vector operations 247
A2.1 Vector dot product 247
A2.2 Vector cross product 247
A3 Field operators 250
A3.1 Gradient 250
A3.2 Divergent 250
A3.3 Advective 251
A3.4 Laplacian 251
A3.5 Curl 252
A4 Derivations of the Bernoulli equation 253
A4.1 The Bernoulli equation from the energy equation 253
A4.2 The Bernoulli equation from the integral momentum equation 253
A4.3 The Bernoulli equation from the Navier-Stokes equation 253
A5 Flow parameters as force ratios 256
A5.1 Acceleration vs. viscous forces: the Reynolds number 256
A5.2 Acceleration vs. gravity force: the Froude number 257
A5.3 Acceleration vs. elastic forces: the Mach number 257
A5.4 Other force ratios 258
A6 Details of the winter 2020-2021 final examination (updated Febru-
ary 2021) 259
A7 Example of previous examinations 261
A8 References 292
 A1 Notation
≡ By definition. The ≡ symbol sets the definition of the term on its left
(which does not depend on previous equations).

̇ (dot above symbol) Time rate: ̇ ≡ d𝑡d . For example, 𝑄̇ is the rate of heat
(in watts) representing a heat quantity 𝑄 (in joules) every second.

(bar above symbol) Time-average: 𝐴 ≡ avg(𝐴) = avg(𝐴 + 𝐴′ ). The prime

symbol indicates the instantaneous fluctuation around the average.

Δ Indicates a net difference between two values: (Δ𝑋 )A→B ≡ 𝑋B − 𝑋A . Can

be negative.

italics Physical properties (e.g. mass 𝑚, temperature 𝑇 ).

straight subscripts Points in space or in time (temperature 𝑇A at point A).

Subscripts “cst” indicate a constant property, “in” indicates “incoming”
and “out” is “outgoing”.
Subscript “av.” indicates “average”.

operators Differential d, partial differential 𝜕, finite differential 𝛿, total (alt.:

substantial) derivative D/D𝑡 (def. eq. 6/7 p. 115), exponential exp 𝑥 ≡ 𝑒 𝑥 ,
natural logarithm ln 𝑥 ≡ log𝑒 𝑥.

vectors Vectors are written with an arrow. Velocity is 𝑉⃗ ≡ (𝑢, 𝑣, 𝑤), alter-
natively written 𝑢𝑖 ≡ (𝑢, 𝑣, 𝑤). The norm of a vector 𝐴⃗ (positive or
⃗ its length (always positive) is ||𝐴||.
negative) is |𝐴|, ⃗

Dot product 𝐴⃗ ⋅ 𝐵⃗ (see §A2.1 p. 247);

vector calculus

Cross product 𝐴⃗ ∧ 𝐵⃗ (see §A2.2 p. 248);

Gradient ∇ ⃗ 𝐴 (def. eq. 4/11 p. 77, see also §A3.1 p. 250);
Divergent ∇ ⃗ ⋅ 𝐴⃗ (def. eq. 5/14 p. 95, see also §A3.2 p. 250);
Laplacian ∇ ⃗ 2 𝐴⃗ (def. eq. 6/38 p. 124, see also §A3.4 p. 251);
Curl ∇⃗ × 𝐴⃗ (def. eq. A/32 p. 252, see also §A3.5 p. 252).

units Units are typed in roman (normal) font and colored gray (1 kg). In
sentences units are fully-spelled and conjugated (one hundred watts).
The liter is noted L to increase readability (1 L ≡ 10−3 m3 ). Units in
equations are those from système international (si) unless otherwise
indicated.

ceded by a dot, integers are written in groups of three (1,234 ⋅ 103 = 1 234).
numbers The decimal separator is a comma, the decimal exponent is pre-

Numbers are rounded up as late as possible and never in series. Leading

and trailing zeros are never indicated.

246
A2 Vector operations
For a step-by-step revision of those notions and many more, written in a
progressive, nonjudgmental way, with plenty of worked-out exercises, you
can try John Bird’s Higher Engineering Mathematics [15].

A2.1 Vector dot product

The dot productw of two vectors is a number defined as:

𝑎⃗ ⋅ 𝑏⃗ = |𝑎| ⃗ cos 𝜃
⃗ |𝑏| (A/1)

⃗
where 𝜃 is the angle separating the two vectors 𝑎⃗ and 𝑏.

⃗
In this document, the dot product is always written with a median dot (𝑎⃗ ⋅ 𝑏),
but in other literature, it is sometimes written with the × symbol. Take care

⃗ 𝑏 , 𝑦𝑏 , 𝑧𝑏 }
not to confuse it with the vector cross product (see §A2.2 p. 248).
It can be shown that the dot product of two vectors 𝑎{𝑥 ⃗ 𝑎 , 𝑦𝑎 , 𝑧𝑎 } and 𝑏{𝑥

𝑎⃗ ⋅ 𝑏⃗ = 𝑥𝑎 𝑥𝑏 + 𝑦𝑎 𝑦𝑏 + 𝑧𝑎 𝑧𝑏
can be quantified as:
(A/2)

The dot product of two vectors is the same regardless of the order in which

𝑎⃗ ⋅ 𝑏⃗ = 𝑏⃗ ⋅ 𝑎⃗
they are multiplied:
(A/3)

Its is easily shown using eq. (A/1) that:

−𝑎 ⋅ →
(→
− ⃗
−𝑏) = −(𝑎⃗ ⋅ 𝑏) (A/4)

⃗
Figure A.1: Two vectors 𝑎⃗ and 𝑏. 247
 A2.2 Vector cross product
The cross productw of two vectors is a vector written as:

𝑎⃗ ∧ 𝑏⃗ = 𝑐⃗ (A/5)

The vector 𝑐⃗ is so that:

𝑐 = 𝑎 𝑏 sin 𝜃
its length is equal to
(A/6)

⃗
its direction is perpendicular to 𝑎⃗ and 𝑏;

its orientation is so that if 𝑏⃗ is positioned at the end of 𝑎,

⃗ then 𝑐⃗ points away
from a point from which the rotation generated 𝑏⃗ is in the clockwise
direction.

Describing 𝑐⃗ requires a third dimension, even if 𝑎⃗ et 𝑏⃗ have only two dimen-

In this document, the cross product is written with a wedge symbol (𝑎⃗ ∧ 𝑏) ⃗
sions.

but in the literature, it is often written with the symbol ×. Make sure you do

𝑐⃗ of two vectors 𝑎{𝑥

⃗ 𝑎 , 𝑦𝑎 , 𝑧𝑎 } et
not confuse it with the dot product (§A2.2 p. 248).

⃗ 𝑏 , 𝑦𝑏 , 𝑧𝑏 } is:
𝑏{𝑥
It can be shown that the cross product

| 𝑖⃗ 𝑗⃗ 𝑘⃗ ||
|
| |
𝑐⃗ = |𝑥𝑎 𝑦𝑎 𝑧𝑎 ||
|
|𝑥𝑏 𝑦𝑏 𝑧𝑏 ||
(A/7)
|
So that one obtains:
|𝑦 𝑧 | |𝑥 𝑧 | |𝑥 𝑦 |
𝑐⃗ = || 𝑎 𝑎 || 𝑖⃗ − || 𝑎 𝑎 || 𝑗⃗ + || 𝑎 𝑎 || 𝑘⃗
|𝑦𝑏 𝑧𝑏 | |𝑥𝑏 𝑧𝑏 | |𝑥𝑏 𝑦𝑏 |
(A/8)

= (𝑦𝑎 𝑧𝑏 − 𝑦𝑏 𝑧𝑎 )𝑖⃗ − (𝑥𝑎 𝑧𝑏 − 𝑥𝑏 𝑧𝑎 )𝑗⃗ + (𝑥𝑎 𝑦𝑏 − 𝑥𝑏 𝑦𝑎 )𝑘⃗ (A/9)

The two vectors 𝑎⃗ ∧ 𝑏⃗ and 𝑏⃗ ∧ 𝑎⃗ are pointing away one from the other (fig. A.3):

𝑏⃗ ∧ 𝑎⃗ = − (𝑎⃗ ∧ 𝑏)
⃗ (A/10)

Figure A.2: Two vectors 𝑎⃗ and 𝑏.⃗ The vector product 𝑎⃗ ∧ 𝑏⃗ has length the product of
the lengths 𝑏⟂ et 𝑎. In the case shown here, the vector 𝑐⃗ = 𝑎⃗ ∧ 𝑏⃗ is going into through
the document plane, going away from the reader.

248
Figure A.3: The vectors 𝑎⃗ ∧ 𝑏⃗ and 𝑏⃗ ∧ 𝑎⃗ have the same length but are pointing
directions opposite one from the other (the first away from the reader, and the other
towards the reader).

If any vector changes direction, the cross product also changes direction
→
− ⃗
𝑎⃗ ∧ −𝑏 = − (𝑎⃗ ∧ 𝑏)
(fig. A.4):
(A/11)

⃗ is pointing away from the vector 𝑎⃗ ∧ 𝑏.

Figure A.4: The vector 𝑎⃗ ∧ −𝑏 ⃗
249
 A3 Field operators
Four operators which apply on vector or scalar fields are important in fluid
mechanics: gradient, divergent, Laplacian and curl.

A3.1 Gradient

⃗ . It applies on a scalar field and produces a vector field. It is

written ∇
The mathematical operator gradientw (first introduced as eq. 4/11 p. 77) is

𝜕 𝜕 𝜕
defined as:

⃗ ≡ 𝑖⃗ + 𝑗⃗ + 𝑘⃗
∇
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/12)

𝜕𝐴 ⃗ 𝜕𝐴 ⃗ 𝜕𝐴 ⃗ ⎛⎜ ⎞
𝜕𝐴

⃗𝐴 ≡ ⎟
𝜕𝑥
∇ 𝑖+ 𝑗+ 𝑘= 𝜕𝐴
𝜕𝑥 𝜕𝑦 𝜕𝑧 ⎜ 𝜕𝑦
⎟
⎝ ⎠
(A/13)
𝜕𝐴
𝜕𝑧

⃗ 𝑝:
For example, the gradient of a pressure field is the vector field ∇

𝜕𝑝 ⃗ 𝜕𝑝 ⃗ 𝜕𝑝 ⃗ ⎛⎜ ⎞
𝜕𝑝

⃗𝑝 ≡ ⎟
𝜕𝑥
∇ 𝑖+ 𝑗+ 𝑘=
𝜕𝑝
𝜕𝑥 𝜕𝑦 𝜕𝑧 ⎜ 𝜕𝑦
⎟
⎝ ⎠
(A/14)
𝜕𝑝
𝜕𝑧

A3.2 Divergent

⃗ ⋅ and is defined as:

written ∇
The mathematical operator divergentw (first introduced as eq. 5/14 p. 95) is

𝜕 ⃗ 𝜕 ⃗ 𝜕 ⃗
⃗⋅ ≡
∇ 𝑖⋅ + 𝑗 ⋅ + 𝑘⋅
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/15)

When applied on a vector field, it produces a scalar field:

𝜕 ⃗ ⃗ 𝜕 ⃗ ⃗ 𝜕 ⃗ ⃗
⃗ ⋅ 𝐴⃗ ≡
∇ 𝑖⋅𝐴 + 𝑗⋅𝐴 + 𝑘⋅𝐴
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/16)
𝜕𝐴𝑥 𝜕𝐴𝑦 𝜕𝐴𝑧
= + +
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/17)

When applied on a 2nd order tensor field, it produces a vector field:

⎛ 𝜕𝐴𝑥𝑥
+ 𝜕𝑦𝑦𝑥 + 𝜕𝐴𝜕𝑧𝑧𝑥
𝜕𝐴
⎞ ⎛ ⃗ ⋅ 𝐴⃗𝑖𝑥
∇ ⎞
⎜ 𝜕𝑥
⎟ ⎜ ⎟
⃗ ⋅ 𝐴⃗𝑖𝑗 ≡ ⎜
∇
𝜕𝐴𝑥𝑦
+ 𝜕𝑦𝑦𝑦 + 𝜕𝑧𝑧𝑦
𝜕𝐴 𝜕𝐴
⎟=⎜ ⃗ ⋅ 𝐴⃗𝑖𝑦
∇ ⎟
⎜ ⎟ ⎜ ⎟
𝜕𝑥
+ 𝜕𝑦𝑦𝑧 + 𝜕𝐴𝜕𝑧𝑧𝑧 ⃗ ⋅ 𝐴⃗𝑖𝑧
∇
(A/18)
𝜕𝐴𝑥𝑧 𝜕𝐴
⎝ 𝜕𝑥 ⎠ ⎝ ⎠

⃗ ⋅ 𝑉⃗ :
For example, the divergent of a velocity field is the scalar field ∇
𝜕𝑢 𝜕𝑣 𝜕𝑤
⃗ ⋅ 𝑉⃗ ≡
∇ + +
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/19)

250
A3.3 Advective

The advective operator,w (𝑉⃗ ⋅ ∇

⃗ ) is defined as follows:

𝜕 𝜕 𝜕
𝑉⃗ ⋅ ∇
⃗≡𝑢 +𝑣 +𝑤
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/20)

Do not confuse the advective operator with the divergent of velocity, ∇ ⃗ ⋅ 𝑉⃗

The advective operator can be applied to a scalar field 𝐴:

(see Appendix A3.2 above, including eq. A/19), which is a scalar field.

𝜕𝐴 𝜕𝐴 𝜕𝐴
(𝑉⃗ ⋅ ∇
⃗ )𝐴 = 𝑢 +𝑣 +𝑤
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/21)

⃗
It can also be applied to a vector field 𝐴:

𝜕 𝐴⃗ 𝜕 𝐴⃗ 𝜕 𝐴⃗
(𝑉⃗ ⋅ ∇
⃗ )𝐴⃗ = 𝑢 +𝑣 +𝑤
𝜕𝑥 𝜕𝑦 𝜕𝑧
(A/22)

⎛ 𝑢 𝜕𝑥 + 𝑣 𝜕𝑦 + 𝑤 𝜕𝐴
𝜕𝐴𝑥 𝜕𝐴𝑥 ⎞
⎜ ⎟
𝑥
𝜕𝑧
=⎜ 𝑢 𝜕𝑥𝑦 + 𝑣 𝜕𝑦𝑦 + 𝑤 𝜕𝑧𝑦 ⎟
𝜕𝐴 𝜕𝐴 𝜕𝐴

⎜ 𝑢 𝜕𝐴 + 𝑣 𝜕𝐴 + 𝑤 𝜕𝐴 ⎟
(A/23)
⎝ ⎠
𝑧 𝑧 𝑧
𝜕𝑥 𝜕𝑦 𝜕𝑧

A3.4 Laplacian

⃗ 2 and defined as:

is written ∇
The mathematical operator Laplacianww (first introduced as eq. 6/38 p. 124)

⃗2 ≡ ∇
∇ ⃗ ⋅∇
⃗ (A/24)

When applied to a scalar field, it is equal to the divergent of the gradient of

the field, and produces a scalar field:

⃗ 2𝐴 ≡ ∇
∇ ⃗ ⋅∇⃗𝐴
𝜕 𝐴 𝜕 2𝐴 𝜕 2𝐴
(A/25)
2
= + +
(𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2
(A/26)

When applied to a vector field, the general expression uses the curl operator
(we never use this expression in this course), and produces a vector field:

⃗ 2 𝐴⃗ ≡ (∇
∇ ⃗ ) 𝐴⃗ − ∇
⃗ ⋅∇ ⃗ ⃗
⃗× ∇
( × 𝐴) (A/27)

In Cartesian coordinates, this simplifies as:

⎛ ⃗ 2 𝐴𝑥
∇ ⎞ ⃗ ⋅∇
⎛ ∇ ⃗ 𝐴𝑥 ⎞
∇𝐴=≡⎜
⃗ 2 ⃗ ⃗ 2 𝐴𝑦
∇ ⎟ = ⎜ ∇ ⃗ ⋅∇ ⃗ 𝐴𝑦 ⎟
⎜ ⎟ ⎜ ⎟
⎝ ⃗ 2 𝐴𝑧
∇ ⎠ ⃗ ⋅∇
⎝ ∇ ⃗ 𝐴𝑧 ⎠
(A/28)

⎛ 𝜕 2 𝐴𝑥
+ (𝜕𝑦) 2 + (𝜕𝑧)2 ⎞
𝜕 2 𝐴𝑥 𝜕 2 𝐴𝑥
⎜ (𝜕𝑥)2
𝜕2𝐴 ⎟
=⎜ 𝜕 2 𝐴𝑦
+ (𝜕𝑦)𝑦2 + (𝜕𝑧)𝑦2 ⎟
𝜕2𝐴

⎜ 𝜕 2 𝐴𝑧 ⎟
(𝜕𝑥)2
+ (𝜕𝑦) 2 + (𝜕𝑧)2
(A/29)
𝜕 2 𝐴𝑧 𝜕 2 𝐴𝑧
⎝ (𝜕𝑥)2 ⎠

251
 ⃗ 2 𝑉⃗ :
For example, the Laplacian of a velocity field is the vector field ∇

⎛ ⃗ 2𝑢
∇ ⎞ ⎛⎜ + + ⎞
𝜕2𝑢 𝜕2𝑢 𝜕2𝑢
(𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2
⎟
∇𝑉 ≡⎜
⃗ 2 ⃗ ⃗ 2𝑣
∇ ⎟=⎜ 𝜕2𝑣
+ 𝜕2𝑣
+ 𝜕2𝑣
⎟
⎜ ⎟ ⎜ ⎟
⃗ 2𝑤
(𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2
⎝ ∇ ⎠ ⎝ + +
(A/30)
𝜕2𝑤 𝜕2𝑤 𝜕2𝑤
(𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ⎠

A3.5 Curl

The mathematical operator curlw (sometimes named rotational) is written ∇ ⃗× .

It applies to a vector field and produces a vector field. It is defined as:
| 𝑖⃗ 𝑗⃗ 𝑘⃗ |
| |
⃗ × ≡ || 𝜕 𝜕 𝜕 ||
∇
| 𝜕𝑥 𝜕𝑦 𝜕𝑧 |
| |
(A/31)
| |
| 𝑖⃗ 𝑗⃗ 𝑘⃗ |
| | 𝜕𝐴𝑧 𝜕𝐴𝑦 ⃗ 𝜕𝐴𝑧 𝜕𝐴𝑥 ⃗ 𝜕𝐴𝑦 𝜕𝐴𝑥 ⃗
⃗ × 𝐴⃗ ≡ || 𝜕
∇ 𝜕 𝜕 |
| = − 𝑖 + − + 𝑗 + − 𝑘
| | ( 𝜕𝑦 𝜕𝑧 ) ( 𝜕𝑥 𝜕𝑧 ) ( 𝜕𝑥 𝜕𝑦 )
|𝐴𝑥 𝐴𝑦 𝐴𝑧 |
𝜕𝑥 𝜕𝑦 𝜕𝑧
| |
(A/32)

⃗ × 𝑉⃗ :
For example, the curl of velocity is the vector field ∇
| 𝑖⃗ 𝑗⃗ 𝑘⃗ ||
| 𝜕𝑤 𝜕𝑣 ⃗ 𝜕𝑤 𝜕𝑢 ⃗ 𝜕𝑣 𝜕𝑢 ⃗
⃗ × 𝑉⃗ = || 𝜕
∇ 𝜕 𝜕 |
| = − 𝑖 + − + 𝑗 + − 𝑘
| 𝜕𝑥 𝜕𝑧 | ( 𝜕𝑦 𝜕𝑧 ) ( 𝜕𝑥 𝜕𝑧 ) ( 𝜕𝑥 𝜕𝑦 )
|𝑢 𝑤 ||
𝜕𝑦
| 𝑣
(A/33)

252
A4 Derivations of the Bernoulli equation

A4.1 The Bernoulli equation from the energy equation

This is covered in section 2.6 p. 41.

A4.2 The Bernoulli equation from the integral momen-

tum equation
We begin with the integral linear momentum equation (eq. 3/9 p. 56):
d
𝐹⃗net = 𝜌 𝑉⃗ d + ∬ 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 ∭CV CS

streamline 𝑠 of the flow, this equation becomes:

When considering a fixed, infinitely short control volume along a known

d
d𝐹⃗pressure + d𝐹⃗shear + d𝐹⃗gravity = 𝜌 𝑉⃗ d + 𝜌𝑉𝐴 d𝑉⃗
d𝑡 ∭CV
along a streamline, where the velocity 𝑉⃗ is aligned (by definition) with the streamline.

Now, adding the restrictions of steady flow ( d/ d𝑡 = 0) and no friction ( d𝐹⃗shear =

0⃗), we already obtain:

d𝐹⃗pressure + d𝐹⃗gravity = 𝜌𝑉𝐴 d𝑉⃗

The projection of the net force due to gravity d𝐹⃗gravity on the streamline
segment d𝑠 has norm d𝐹⃗gravity ⋅ d⃗𝑠 = −𝑔𝜌𝐴 d𝑧, while the net force due to
pressure is aligned with the streamline and has norm d𝐹pressure,𝑠 = −𝐴 d𝑝.
Along this streamline, we thus have the following scalar equation, which we
integrate from points 1 to 2:

−𝐴 d𝑝 − 𝜌𝑔𝐴 d𝑧 = 𝜌𝑉𝐴 d𝑉
1
− d𝑝 − 𝑔 d𝑧 = 𝑉 d𝑉
𝜌
2
1 2 2
−∫ d𝑝 − ∫ 𝑔 d𝑧 = ∫ 𝑉 d𝑉
1 𝜌 1 1

transfer, where, therefore, the density 𝜌 is constant. In this way, we arrive to

The last obstacle is removed when we consider flows without heat or work

equation. 2/20 p. 42 again:

𝑝1 1 2 𝑝2 1 2
+ 𝑉1 + 𝑔𝑧1 = + 𝑉 + 𝑔𝑧2
𝜌 2 𝜌 2 2
1 2 1 2
(𝑝 + 2 𝜌𝑉 + 𝜌𝑔𝑧 )1 = (𝑝 + 2 𝜌𝑉 + 𝜌𝑔𝑧 )2

A4.3 The Bernoulli equation from the Navier-Stokes

equation
We start by following a particle along its path in an arbitrary flow, as dis-

its speed 𝑉 is not.

played in fig. A.5. The particle path is known (condition 5 in §2.6 p. 41), but
253
 travels from point 1 to point 2. An infinitesimal path segment is named d⃗𝑠.
Figure A.5: Different pathlines in an arbitrary flow. We follow one particle as it

Figure CC-0 Olivier Cleynen

(eq. 6/42 p. 125) onto an infinitesimal portion of trajectory d⃗𝑠. Once all terms
We are now going to project every component of the Navier-Stokes equation

have been projected, the Navier-Stokes equation becomes a scalar equation:

𝜕 𝑉⃗
𝜌 + 𝜌(𝑉⃗ ⋅ ∇
⃗ )𝑉⃗ = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
𝜕𝑡
𝜕 𝑉⃗
𝜌 ⋅ d⃗𝑠 + 𝜌(𝑉⃗ ⋅ ∇
⃗ )𝑉⃗ ⋅ d⃗𝑠 = 𝜌 𝑔⃗ ⋅ d⃗𝑠 − ∇ ⃗ 2 𝑉⃗ ⋅ d⃗𝑠
⃗ 𝑝 ⋅ d⃗𝑠 + 𝜇 ∇
𝜕𝑡

Because the velocity vector 𝑉⃗ of the particle, by definition, is always aligned

with the path, its projection is always equal to its norm: 𝑉⃗ ⋅ d⃗𝑠 = 𝑉 d𝑠. Also,
the downward gravity 𝑔 and the upward altitude 𝑧 have opposite signs, so
that 𝑔⃗ ⋅ d⃗𝑠 = −𝑔 d𝑧; we thus obtain:
𝜕𝑉 d𝑉 d𝑝
𝜌 d𝑠 + 𝜌 𝑉 d𝑠 = −𝜌𝑔 d𝑧 − ⃗ 2 𝑉⃗ ⋅ d⃗𝑠
d𝑠 + 𝜇 ∇
𝜕𝑡 d𝑠 d𝑠

When we restrict ourselves to steady flow (condition 1 in §2.6), the first

left-hand term vanishes. Neglecting losses to friction (condition 4) alleviates
us from the last right-hand term, and we obtain:
d𝑉 d𝑝
𝜌 𝑉 d𝑠 = −𝜌𝑔 d𝑧 − d𝑠
d𝑠 d𝑠
𝜌𝑉 d𝑉 = −𝜌𝑔 d𝑧 − d𝑝

This equation can then be integrated from point 1 to point 2 along the
pathline:
2 2 2
𝜌∫ 𝑉 d𝑉 = − ∫ 𝜌𝑔 d𝑧 − ∫ d𝑝
1 1 1

incompressible (condition 2), the density 𝜌 remains constant, so that we

When no work or heat transfer occurs (condition 3) and the flow remains

indeed have returned to eq. 2/20 p. 42:

1 1
Δ ( 𝑉 2 ) + 𝑔Δ𝑧 + Δ𝑝 = 0
2 𝜌
(A/34)
1 2 1 2
(𝑝 + 2 𝜌𝑉 + 𝜌𝑔𝑧 )1 = (𝑝 + 2 𝜌𝑉 + 𝜌𝑔𝑧 )2 (A/35)

254
Thus, we can see that if we follow a particle along its path, in a steady,
incompressible, frictionless flow with no heat or work transfer, its change in
kinetic energy is due only to the result of gravity and pressure, in accordance
with the Navier-Stokes equation.

255
 A5 Flow parameters as force ratios
This topic is well covered in Massey [6]

Instead of the mathematical approach covered in §8.2.2 p. 162, the concept of

flow parameter can be approached by comparing forces in fluid flows.
Fundamentally, understanding the movement of fluids requires applying
Newton’s second law of motion: the sum of forces which act upon a fluid
particle is equal to its mass times its acceleration. We have done this in
an aggregated manner with integral analysis (in chapter 3, eq. 3/9 p. 56),
and then in a precise and all-encompassing way with differential analysis
(in chapter 6, eq. 6/42 p. 125). With the latter method, we obtain complex
mathematics suitable for numerical implementation, but it remains difficult
to obtain rapidly a quantitative measure for what is happening in any given
flow.
In order to obtain this, an engineer or scientist can use force ratios. This
involves comparing the magnitude of a type of force (pressure, viscous,
gravity) either with another type of force, or with the mass-times-acceleration
which a fluid particle is subjected to as it travels. We are not interested in
the absolute value of the resulting ratios, but rather, in having a measure of
the parameters that influence them, and being able to compare them across
experiments.

A5.1 Acceleration vs. viscous forces:

the Reynolds number

acceleration. If a representative length for the particle is 𝐿, the particle

The net sum of forces acting on a particle is equal to its mass times its

mass grows proportionally to the product of its density 𝜌 and its volume 𝐿3 .
Meanwhile, its acceleration relates how much its velocity 𝑉 will change over
a time interval Δ𝑡: it may be expressed as a ratio Δ𝑉 /Δ𝑡. In turn, the time
interval Δ𝑡 may be expressed as the representative length 𝐿 divided by the
velocity 𝑉 , so that the acceleration may be represented as proportional to the
ratio 𝑉 Δ𝑉 /𝐿. Thus we obtain:
𝑉 Δ𝑉
|net force| = |mass × acceleration| ∼ 𝜌𝐿3
𝐿
⃗
|𝐹net | ∼ 𝜌𝐿 𝑉 Δ𝑉
2

to the shear effort and a representative acting surface 𝐿2 . The shear can
We now observe the viscous force acting on a particle: it is proportional

be modeled as proportional to the viscosity 𝜇 and the rate of strain, which

will grow proportionally to Δ𝑉 /𝐿. We thus obtain a crude measure for the
magnitude of the shear force:
Δ𝑉 2
|viscous force| ∼ 𝜇 𝐿
𝐿
|𝐹⃗viscous | ∼ 𝜇Δ𝑉 𝐿

The magnitude of the viscous force can now be compared to the net force:

|net force| 𝜌𝐿2 𝑉 Δ𝑉 𝜌𝑉 𝐿

∼ = = [Re]
|viscous force| 𝜇Δ𝑉 𝐿 𝜇
(A/36)

256
and we recognize the ratio as the Reynolds number (8/12 p. 164). We thus see
that the Reynolds number can be interpreted as the inverse of the influence
of viscosity. The larger [Re] is, and the smaller the influence of the viscous
forces will be on the trajectory of fluid particles.

A5.2 Acceleration vs. gravity force: the Froude number

The weight of a fluid particle is equal to its mass, which grows with 𝜌𝐿3 ,
multiplied by gravity 𝑔:

|weight force| = |𝐹⃗W | ∼ 𝜌𝐿3 𝑔

The magnitude of this force can now be compared to the net force:

|net force| 𝜌𝐿2 𝑉 2 𝑉 2

∼ = = [Fr]
2
|weight force| 𝜌𝐿 𝑔
3 𝐿𝑔
(A/37)

and here we recognize the square of the Froude number (8/11 p. 164). We thus
see that the Froude number can be interpreted as the inverse of the influence
of weight on the flow. The larger [Fr] is, and the smaller the influence of
gravity will be on the trajectory of fluid particles.

A5.3 Acceleration vs. elastic forces: the Mach number

In some flows called compressible flows the fluid can perform work on itself,
and and the fluid particles then store and retrieve energy in the form of
changes in their own volume. In such cases, fluid particles are subject to

resulting from this force as proportional to the bulk modulus of elasticity 𝐾

an elastic force in addition to the other forces. We can model the pressure

of the fluid (formally defined as 𝐾 ≡ 𝜌 𝜕𝑝/𝜕𝜌); the elastic force can therefore
be modeled as proportional to 𝐾 𝐿2 :

|elasticity force| = |𝐹⃗elastic | ∼ 𝐾 𝐿2

The magnitude of this force can now be compared to the net force:

|net force| 𝜌𝐿2 𝑉 2 𝜌𝑉 2

∼ =
|elasticity force| 𝐾 𝐿2 𝐾

because the value of 𝐾 in a given fluid varies considerably not only according
This ratio is known as the Cauchy number; it is not immediately useful

to temperature, but also according to the type of compression undergone by

the fluid: for example, it grows strongly during brutal compressions.
During isentropic compressions and expansions (isentropic meaning that the
process is fully reversible, i.e. without losses to friction, and adiabatic, i.e.

proportional to the square of the speed of sound 𝑐:

without heat transfer), it can be shown that the bulk modulus of elasticity is

𝐾 |reversible = 𝑐 2 𝜌 (A/38)

The Cauchy number calibrated for isentropic evolutions is then

|net force| 𝜌𝑉 2 𝑉 2
∼ = 2 = [Ma]
2
|elasticity force|reversible 𝐾 𝑐
(A/39)
257
 and here we recognize the square of the Mach number (1/10 p. 16). We thus
see that the Mach number can be interpreted as the influence of elasticity on
the flow. The larger [Ma] is, and the smaller the influence of elastic forces
will be on the trajectory of fluid particles.

A5.4 Other force ratios

The same method can be applied to reach the definitions for the Strouhal and
Euler numbers given in §8.2 p. 161. Other numbers can also be used which
relate forces that we have ignored in our study of fluid mechanics. For exam-
ple, the relative importance of surface tension forces or of electromagnetic
forces are quantified using similarly-constructed flow parameters.
In some applications featuring rotative motion, such as flows in centrifugal
pumps or planetary-scale atmospheric weather, it may be convenient to ap-
ply Newton’s second law in a rotating reference frame. This results in the
appearance of new reference-frame forces, such as the Coriolis or centrifugal
forces; their influence can then be studied using additional flow parameters.
In none of those cases can flow parameters give enough information to predict
solutions. They do, however, provide quantitative data to indicate which
forces are relevant in which places: this not only helps us understand the
mechanisms at work, but also distinguish the negligible from the influential,
a key characteristic of efficient scientific and engineering work.

258
A6 Details of the winter 2020-2021 final
examination (updated February 2021)
The final examination for this course in the winter semester 2020-2021 is an
open-book, take-home exam which will take place on February 18, 2021
from 14:00 to 16:00. The most important information is as follows:

• This is a take-home exam: you may consult your notes, books, use Video: The companion video to
software and resources offline or online. this exam briefing (2020)
by Olivier Cleynen (CC-by)
https://youtu.be/6vnK7-VsKXc
• You must take this exam alone: you cannot interact with anyone online
or offline during the exam.
• The exam lasts 90 minutes. Additionally, 30 minutes are provided for
the the scanning and upload of your answer.
• You will receive the assignment per email shortly before the start. You
must scan your answer with your student card on each page, and send
it as a single-PDF file to fluidmech@ovgu.de using your academic email
before the end of the exam.
• In the winter semester, this online examination grade is the only grade
you receive in this course.

You can download the template of the exam at

https://fluidmech.ninja/exams/exam_20210218_template.pdf.
The complete list of examinable problems (unchanged from the 2020 exam)
is as follows:

• 2.2 2.3 2.4 2.5 2.6

Video: How to survive the exam-
• 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 ination (2020)
by Olivier Cleynen (CC-by)

• 4.4 4.5 4.6 1 4.7 https://youtu.be/GD0l_5XqVoc

• 5.2 5.3 5.4 5.5 5.6

• 6.2 6.3 6.4 6.5 6.6 6.7 6.9
• 7.3 7.4 7.5 7.6 7.7 7.8
• 8.2 8.4 8.5
• Chapter 9 is not examinable this semester
• 10.2 10.3 10.4 10.5 10.6
• Chapter 11 is not examinable this semester

Three problems will be given, all mandatory. The first problem (10 pts is
exercise 6.2 p.131. The two other problems (45 pts each) are extracted from
the list above, and modified slightly. Typically, the input data is changed, as
well as the problem geometry. The method for solving the problems remains
the same.
A formula sheet is provided. It is the sum of the preambles of every problem
sheet in the lecture notes. It includes the Moody diagram and the viscosity
1 In exercise 4.6, only the calculation of the vertical force 𝐹top is examinable. 259
 diagram used in the problem sheets. You should definitely have a calculator
with you, to facilitate calculations.
The criteria for grading your answers are:

• You must show your work in all answers;

• Answers to questions starting with “show that” should be particularly

well-developed and continuous;

• Illegible or ambiguous answers are always discarded.

Examinations from previous years, and their full solution, are available on
the course website (https://fluidmech.ninja/). Since the course content has
changed over time, you might find a few differences:

• Problems involving calculating compressible air flow using tables are

no longer examinable;

• A problem involving a ball fountain (“Kugel fountain”) is no longer

examinable;

• Viscosity values were read in a different diagram, and may not match
values read in the 2020 viscosity diagram.

You are welcome (and in fact encouraged!) to ask me questions of all sorts
about the exam. You may contact me as described in the introduction, page 7.
I wish you to have productive and joyful revisions!
Olivier
February 2021
(updated February 12, after cancellations following covid-19 restrictions)

260
A7 Example of previous examinations
The following pages present the final examinations for this course in 2020
and 2021, and their full solutions.

261
 Fluid Mechanics examination — July 11, 2019
Continuity equation for incompressible flow:
⃗ ⋅ 𝑉⃗ = 0
∇
Fluid Mechanics for Master Students
(7)
Duration: 2 h – Use of calculator is authorized; documents are not authorized.
Solve problem 1, plus three other problems among problems 2 to 6.
Navier-Stokes equation for incompressible flow:
D𝑉⃗
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
D𝑡
Except otherwise indicated, we assume that:
(8)
Fluids are Newtonian
The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 ◦C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
In a highly-viscous (creeping) steady flow, the drag 𝐹𝐷 exerted on a spherical body of
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1 diameter 𝐷 at by flow at velocity 𝑉∞
is quantified as:
𝐹𝐷sphere = 3𝜋𝜇𝑉∞ 𝐷 (9)
In cylindrical pipe flow, we assume the flow is always laminar for [Re]𝐷 . 2 300,
Balance of mass in a considered volume with steady flow:
0 = Σ [𝜌𝑉⟂ 𝐴]incoming + Σ [𝜌𝑉⟂ 𝐴]outgoing (1) and always turbulent for [Re]𝐷 & 4 000. The Darcy friction factor 𝑓 is defined as:
where 𝑉⟂ is negative inwards, positive outwards.
|Δ𝑝loss |
𝑓 ≡ 𝐿1
𝜌𝑉av.
2
(10)
𝐷2
The loss coefficient 𝐾𝐿 is defined as:
Balance of momentum in a considered volume with steady flow:
𝐹⃗net on fluid = Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ]incoming + Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ]outgoing |Δ𝑝loss |
𝐾𝐿 ≡
(2)
where 𝑉⟂ is negative inwards, positive outwards. 1
𝜌𝑉av.
2
(11)
2
Viscosities of various fluids are given in fig. 1. Pressure losses in cylindrical pipes
Balance of energy in a considered volume with steady flow:
𝑝 1
𝑄̇ net + 𝑊̇ shaft, net
can be calculated with the help of the Moody diagram presented in fig. 2 p.4.
= Σ 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]in
𝑝 1 2
+Σ 𝑚̇ 𝑖 + + 𝑉 + 𝑔𝑧
[ ( 𝜌 2 )]
(3) Non-dimensional incompressible Navier-Stokes equation:
𝜕 𝑉⃗ ∗ 1 1 ∗2 ⃗ ∗
where 𝑚̇ is negative inwards, positive outwards.
+ [1] 𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ = ⃗ ∗𝑝∗ +
𝑔⃗∗ − [Eu] ∇ ⃗ 𝑉
∇
out
𝜕𝑡 ∗ [Fr]2
[St] (12)
[Re]
in which [St] ≡ 𝑓𝑉𝐿 , [Eu] ≡ 𝑝𝜌0 −𝑝
𝑉2
∞
, [Fr] ≡ √𝑉 and [Re] ≡ 𝜌 𝑉𝜇 𝐿 .
𝑔𝐿
The force coefficient 𝐶𝐹 and power coefficient 𝐶P are defined as:
d
Mass balance through an arbitrary volume:
0 = 𝜌 d + ∬ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
𝑊̇
d𝑡 ∭CV 𝐶𝐹 ≡
𝐹
𝐶P ≡
(4)
1
𝜌𝑆𝑉 2 1
𝜌𝑆𝑉 3
CS (13)
2 2
d The speed of sound 𝑐 in air is modeled as:
Momentum balance through an arbitrary volume:
𝐹⃗net = 𝜌 𝑉⃗ d + ∬ 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴 √
d𝑡 ∭CV 𝑐 = 𝛾 𝑅𝑇
(5)
CS (14)
d
Angular momentum balance through an arbitrary volume:
𝑀⃗ net,X = 𝑟⃗X𝑚 ∧ 𝜌 𝑉⃗ d + ∬ 𝑟⃗X𝑚 ∧ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ 𝑉⃗ d𝐴
d𝑡 ∭CV
(6)
CS
1 2
 In boundary layer flow, we assume that transition occurs at [Re]𝑥 & 5 ⋅ 105 .
The wall shear coefficient 𝑐𝑓 , a function of distance 𝑥,
is defined based on the free-stream flow velocity 𝑈 :
𝜏wall
𝑐𝑓(𝑥) ≡ 1
𝜌𝑈 2
(15)
2
Exact solutions to the laminar boundary layer along a smooth surface yield:
𝛿 4,91 𝛿∗ 1,72
= √ = √
𝑥 𝑥
(16)
[Re]𝑥 [Re]𝑥
𝛿 ∗∗ 0,664 0,664
= √ 𝑐𝑓(𝑥) = √
𝑥
(17)
[Re]𝑥 [Re]𝑥
Solutions to the turbulent boundary layer along a smooth surface yield the following
time-averaged characteristics:
𝛿 0,16 𝛿∗ 0,02
≈ ≈
𝑥 1
𝑥 1 (18)
[Re]𝑥7 [Re]𝑥7
𝛿 ∗∗ 0,016 0,027
≈ 𝑐𝑓(𝑥) ≈
𝑥 1 1 (19)
[Re]𝑥7 [Re]𝑥7
In a highly-viscous (creeping) steady flow, the drag 𝐹D exerted on a spherical body of
diameter 𝐷 at by flow at velocity 𝑈∞ is quantified as:
𝐹D sphere = 3𝜋𝜇𝑈∞ 𝐷 (20)
Figure 1 – Viscosity of various fluids at a pressure of 1 bar (in practice viscosity is almost independent of
Figure 2 – A Moody diagram, which presents values for 𝑓 measured experimentally, as a function of the
diameter-based Reynolds number [Re]𝐷 , for different relative roughness values.
pressure).
Figure © White 2008
Diagram CC-by-sa S Beck and R Collins, University of Sheffield
3 4
 1 Governing equation
1.1. [5 pts] Write out equation (8), the Navier-Stokes equation for incompressible flow,
in its fully-developed form in three Cartesian coordinates.
1.2. [5 pts] Write out equation (7), the continuity equation for incompressible flow, in
its fully-developed form in three Cartesian coordinates.
2 Observation window in a water tank
A water tank used in a laboratory is filled with stationary water (fig. 3). A window is
installed on one of the walls of the canal, to enable observation. The window is hinged
The window has a height of 1,5 m and a width of 3,5 m. The walls of the tank are inclined
on its top face.
Solve problem 1,
and three other problems among problems 2 to 6. with an angle 𝜃 = 70° relative to horizontal.
The following marking guidelines will be used:
• Answers to questions starting with “show that” should be fully-developed and
continuous;
• In all other questions, the correct result with the correct unit is enough to obtain Figure 3 – A door installed on the wall of a water tank.
full points;
• Illegible or ambiguous answers are always discarded.
2.1. [15 pts] What is the magnitude of the net force applying on the tank window?
2.2. [10 pts] At what distance away from the hinge does this force apply?
Water is added to the tank, so that the water level increases.
2.3. [5 pts] How will the distance calculated above change as water is added? (briefly
justify your answer, e.g. in 30 words or less)
5 6
3 Piping leading to a turbine 4 Boundary layer on a flat plate
A thin and smooth plate with width 𝑊 = 0,6 m and length 𝐿 = 2 m is placed with a
zero angle of attack in atmospheric air flow incoming at 21 m s−1 , as shown in figure 5.
A pipe leads water from one reservoir to a turbine, which discharges into another
reservoir, as shown in figure 4.
We would like to study the shear exerted by the flow over the top surface of the plate.
Figure 5 – A thin plate positioned parallel to an incoming uniform flow.
Figure CC-0 o.c.
4.1. [5 pts] At what distance 𝑥tr. along the plate, approximately, will the boundary layer
transit and become turbulent?
4.2. [10 pts] Starting from equation (21), which quantifies the friction factor 𝑐𝑓 (see
definition 15) in a laminar boundary layer,
0,664
Figure 4 – Layout of the water pipe. For clarity, in this figure, the vertical scale is greatly exaggerated. In
𝑐𝑓(𝑥) = √
the vertical scale, the diameter of the pipe is also greatly exaggerated.
(21)
The pipe is made of coarse concrete (roughness 0,25 mm) and carries 800 L s−1 of water
[Re]𝑥
at 20 ◦C. It has a diameter of 1,1 m and features four elbow bends with sharp angles, each show that the shear force 𝐹𝜏 laminar exerted in the laminar section of the boundary
inducing a loss coefficient 𝐾𝐿 of 0,75. layer is:
√
𝐹𝜏 laminar = 0,664 𝜌𝜇 𝑈 2 𝑊 𝑥tr.2
1
3.1. [10 pts] Represent qualitatively (i.e. without numerical data) the pressure distribu-
3
(22)
4.3. [5 pts] What is the shear force exerted on the top surface of the plate by the laminar
tion along the length of the pipe, both when the turbine is shut down (without any
flow), and when it is operating.
3.2. [15 pts] What is the hydraulic power available to the turbine?
section of the boundary layer?
4.4. [5 pts] What is the shear force exerted on the top surface of the plate by the
The outlet tank on the right is very large, so that its water level does not vary. The source turbulent section of the boundary layer?
4.5. [5 pts] Would the boundary layer become thicker if the velocity was increased?
water tank on the left, however, sees its height decrease as the water is emptied through
the turbine. Ultimately, as the water level decreases, the water stops flowing entirely.
(briefly justify your answer, e.g. in 30 words or less).
3.3. [5 pts] When the water stops flowing, what will be the height of the water level in
the source tank on the left?
7 8
 The width of the profile (perpendicular to the flow, in the 𝑧-direction) is 70 cm. The water
has uniform temperature and density (20 ◦C, 999 kg m−3 ) and the pressure is uniform
5 Velocity measurements in a tunnel
across the measurement surface.
A group of students proceeds with speed measurements in a water tunnel. The objective
5.1. [20 pts] What is the drag force applying on the profile?
is to measure the drag applying on a an object with constant cross-section, positioned
across the tunnel test section (fig. 6).
5.2. [10 pts] If water was replaced with a fluid with higher viscosity, how would you
expect the drag force to change? (briefly justify your answer, e.g. in 30 words or
less)
6 Lift and drag on a rotating football
A group of fluid dynamicists investigates the air flow around a football. In particular,
flying through the air. The football has diameter 22 cm, a weight of 430 g; it is traveling
they are interested in the forces applying on the ball when it has been kicked and is
at 70 km h−1 .
completely across the tunnel (in the 𝑧-direction). The horizontal velocity distributions upstream and
Figure 6 – An object with constant cross-section positioned across a water tunnel. The object spans
sphere has a diameter of 1,1 m. Drag force measurements are carried out in the tunnel.
In order to observe the flow, they install a steel sphere in a wind tunnel (figure 7). The
downstream of the profile are also shown.
Figure CC-0 o.c.
Upstream of the object, the water flow velocity is uniform (𝑢1 = 𝑈 = 3,2 m s−1 ).
Downstream of the object, horizontal velocity measurements are made every 5 cm across
the flow; the following results are obtained:
vertical position 𝑦 (cm) horizontal speed 𝑢2 (m s−1 )
0 3,2
5 3,2
10 3,15
15 3,14
20 3,03
25 2,92
30 2,81
35 2,87
40 2,89 travels with speed 𝑉tunnel . Force measurements are carried out on the ball.
Figure 7 – A steel sphere positioned in a wind tunnel. The sphere is maintained stationary, while the air
45 2,97
50 3,19
Figure CC-0 o.c.
55 3,2
60 3,2
9 10
 6.1. [5 pts] What is the wind tunnel speed required, so that the flow around the real
football is reproduced around the sphere in the tunnel?
6.2. [5 pts] With the speed calculated above, by which factor should the drag force
measured in the wind tunnel be multiplied, in order to obtain the drag force on the
real football?
The fluid dynamicists now investigate the effect of spin on the ball. When the football
is rotated along a horizontal axis during travel, a lift force exerts laterally on the ball,
curving its trajectory. This is represented, from above, in figure 8.
In order to quantify this effect, the wind tunnel sphere is rotated in the wind tunnel, and
measurements are carried out; the results are plotted in figure 9.
Figure 9 – Experimental measurements of the lift and drag coefficients applying on a rotating sphere in an
steady uniform flow.
Figure © from Munson & al. 2013
6.3. [10 pts] How many rotations per second are required in order to generate a lift
force of 3,1 N on the real football when it travels?
6.4. [5 pts] What is then the corresponding drag force ?
6.5. [5 pts] Propose and quantify one possibility for the football player to double the
lift force applying on the ball.
Figure 8 – Trajectory of a rotating football in free flight, as seen from above. A lift force exerts towards the
left, and deviates the trajectory towards the left.
Figure CC-0 o.c.
11 12
 Solution: Fluid Mechanics examination — July 11, 2019 2 Observation window in a water tank
Fluid Mechanics for Master Students
fluidmech.ninja 2.1 Net force
We define coordinates 𝑟 and 𝑧, and the length 𝐿1 , as shown in the figure below.
1 Governing equation
1.1 N-S equation question
𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢 𝜕 2𝑢
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑥 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑥 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑝 𝜕 2𝑣 𝜕 2𝑣 𝜕 2𝑣
(1)
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑦 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑦 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑝 𝜕 2𝑤 𝜕 2𝑤 𝜕 2𝑤
(2)
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑧 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑧 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
(3)
The force on a small section of door with length d𝑟 and width 𝑊 is d𝐹 . On the complete
door, the force applying due to the net pressure 𝑝net of water and air is:
1.2 Continuity question
𝜕𝑢 𝜕𝑣 𝜕𝑤 𝑟=𝑅max
+ + = 0 𝐹net = ∫ d𝐹
𝜕𝑥 𝜕𝑦 𝜕𝑧
(4) (5)
𝑟=0
𝑟=𝑅max
= ∫ 𝑝net d𝑆 (6)
𝑟=0
𝑟=𝑅max
= ∫ 𝑝net 𝑊 d𝑟 (7)
𝑟=0
𝑟=𝑅max
= ∫ 𝜌 𝑔 𝑧 𝑊 d𝑟 (8)
𝑟=0
A coordinate transform is needed to solve the integral, expressing 𝑧 as a function of 𝑅.
This is obtained by geometry:
𝑧 = 𝑍min + 𝑟 sin 𝜃 (9)
1 2
Inserting eq. 9 into eq. 8, we continue with: 2.3 Change in distance
𝑟=𝑅max
𝐹net = ∫ 𝜌𝑔 (𝑍min + 𝑟 sin 𝜃) 𝑊 d𝑟 Increasing water height translates (only) in an increase of the value of 𝐿1 in the equations
above. Both 𝑀net and 𝐹net increase together with 𝐿1 (eqs. 14 & 22), so it is not immediately
(10)
𝑟=0
𝑟=𝑅max
= 𝜌𝑔𝑊∫ (𝑍min + 𝑟 sin 𝜃) d𝑟 (11) apparent how 𝑅𝐹 changes in eq. 26.
𝑟=0
𝑟=𝑅max
= 𝜌𝑔𝑊∫ (𝐿1 sin 𝜃 + 𝑟 sin 𝜃) d𝑟 (12) Several possibilities can be used to find the answer:
𝑟=0
1 𝑟=𝑅max
= 𝜌 𝑔 𝑊 sin 𝜃 [𝐿1 𝑟 + 𝑟 2 ] • Substituting 𝑥𝐿1 instead of 𝐿1 and comparing the two radiuses, one can write:
2 𝑟=0
(13)
= 𝜌 𝑔 𝑊 sin 𝜃 (𝐿1 𝑅max + 0,5𝑅max 2 )
𝑅𝐹 2 < 𝑅𝐹 1
(14)
= 103 × 9,81 × 3,5 × sin(70°) (0,8 × 1,5 + 0,5 × 1,52 )
(29)
(15)
= 75 014 N 1
𝑥𝐿1 𝑅max
2
+ 1 𝑅3 1
𝐿1 𝑅max + 3 𝑅max
2 1 3
2 3 max
< 2
(16)
𝐹net = 75,01 kN 𝑥𝐿1 𝑅max + 21 𝑅max
2 𝐿1 𝑅max + 21 𝑅max
2
(30)
(17)
1 < 𝑥 (31)
2.2 Distance from hinge
(it is even possible to show, using this equation 30, that 𝑅𝐹 tends towards 𝑅max /2 as
𝐿1 increases);
We first calculate the moment exerting about the hinge due to the net pressure of air and
water, using the same notation as above:
𝑟=𝑅max
𝑀net = ∫ 𝑟 d𝐹
• It is possible to observe graphically that the center of application of the force
(18)
𝑟=0
increase in 𝐿1 ;
moves closer to the hinge when the net pressure distribution is changed due to the
𝑟=𝑅max
= ∫ 𝑟𝜌 𝑔 𝑧 𝑊 d𝑟 (19)
𝑟=0
𝑟=𝑅max
= ∫ 𝑟𝜌 𝑔 (𝑍min + 𝑟 sin 𝜃) 𝑊 d𝑟 (20)
𝑟=0
𝑟=𝑅max
= 𝜌𝑔𝑊∫ 𝑟 (𝐿1 sin 𝜃 + 𝑟 sin 𝜃) d𝑟 (21)
𝑟=0
1 1 𝑟=𝑅max
= 𝜌 𝑔 𝑊 sin 𝜃 [ 𝐿1 𝑟 2 + 𝑟 3 ]
2 3 𝑟=0
1
(22)
= 103 × 9,81 × 3,5 × sin(70°) (0,5 × 0,8 × 1,52 + × 1,53 )
3 • It is also possible to calculate manually one or several new values for 𝑅𝐹 ;
(23)
= 65 335 N m (24)
𝑀net = 65,34 kN m (25) All those methods will provide some evidence that the distance 𝑅𝐹 will in fact decrease
when 𝐿1 is increased.
The distance away from the hinge 𝑅𝐹 is obtained by dividing the moment by the force:
𝑀net
𝑅 =
𝐹net
𝐹 (26)
75 014
=
65 335
(27)
𝑅𝐹 = 0,87 m (28)
3 4
 The Reynolds number is
3 Piping leading to a turbine
𝜌𝑉av. 𝐷
[Re]𝐷 =
𝜇
(36)
103 × 0,842 × 1,1
=
3.1 Pressure distribution
10−5
(37)
[Re]𝐷 = 9,251 ⋅ 105 (38)
The relative roughness is
𝜖 0,25 ⋅ 10−3
=
𝐷 1,1
𝜖
(39)
= 2,27 ⋅ 10−4
𝐷
(40)
With those values, the Moody diagram reads:
𝑓 = 0,0158 (41)
Finally, the wall friction losses along the pipe are calculated as:
1 2 𝐿
Δ𝑝𝑓 = −𝑓 𝜌𝑉av.
2 𝐷
(42)
1 4 ⋅ 103
= −0,0158 103 0,8422
2 1,1
(43)
3.2 Turbine power Δ𝑝𝑓 = −2,03 ⋅ 104 Pa (44)
We want to calculate three pressure drops: • Pressure drop due to losses in the four bends, Δ𝑝bends :
• Pressure drop due to wall friction losses along the pipe, Δ𝑝𝑓 : The average velocity 1 2
Δ𝑝bends = −4 × 𝐾𝐿 𝜌𝑉av.
2
1
(45)
= −4 × 0,75 × × 103 × 0,8422
2
in the pipe is
̇
(46)
𝑉 = Δ𝑝bends = −1,06 ⋅ 103 Pa
𝑆
(47)
̇
av. (32)
= 𝐷2 • Pressure drop due to hydrostatic pressure change across the turbine, Δ𝑝ℎ :
𝜋 4
(33)
0,8
=
𝜋 × 1,14
2
Δ𝑝ℎ = 𝜌𝑔(Δ𝑧)
(34)
𝑉av. = 0,842 m s−1
(48)
(35) = 103 × 9,81 [4 − (25 + 51)] (49)
Δ𝑝ℎ = −7,06 ⋅ 105 Pa (50)
5 6
Finally, the turbine hydraulic power is obtained as:
4 Boundary layer on a flat plate
̇
𝑊̇ turbine = (Δ𝑝turbine )
= ̇ (Δ𝑝ℎ − Δ𝑝𝑓 − Δ𝑝bends )
(51)
(52) 4.1 Transition point
= 0,8 × [−7,06 ⋅ 105 − (−2,03 ⋅ 104 ) − (−1,06 ⋅ 10−3 )] (53) The transition point occurs at [Re]𝑥 ≈ 5 ⋅ 105 . Solving for 𝑥tr. , we have:
= −5,479 ⋅ 105 W
𝜌𝑈 𝑥tr.
𝑊̇ turbine = −547,9 kW [Re]𝑥tr. =
(54)
𝜇
(56)
[Re]𝑥tr. 𝜇
(55)
𝑥tr. =
𝜌𝑈
(57)
5 ⋅ 105 × 1,5 ⋅ 10−5
=
3.3 Residual water height
1,225 × 21
(58)
𝑥tr. = 0,29 m
residual water height in the left tank will be 8 m.
The water will flow until air is entrained (“sucked”) into the pipe inlet. At this point, the (59)
4.2 Shear in laminar section
We start with the given equation and implement the definition (15) of the formula sheet,
as well as the definition of the distance-based Reynolds number [Re]𝑥 :
0,664
𝑐𝑓(𝑥) = √ (60)
[Re]𝑥
𝜏wall, laminar 0,664
= √
1
𝜌𝑈 2 𝜌𝑈 𝑥
(61)
2 𝜇
𝜌𝑈 𝑥 2 1 2
−1
𝜏wall, laminar = 0,664 𝜌𝑈
( 𝜇 ) 2
√
(62)
𝜏wall, laminar = 0,332 𝜌𝜇 𝑈 1,5 𝑥 − 2
1
(63)
The shear force is the integral of the shear with respect to area:
𝐹shear, laminar = ∫ 𝜏wall, laminar d𝑆 (64)
We split the total area covered by the laminar boundary layer in strips of width 𝑊 and
length d𝑥, with 𝑥 ranging from 0 (leading edge) to 𝑥tr. (where the laminar part of the
7 8
boundary ends), obtaining: 4.5 Thickness
𝑥=𝑥tr.
𝐹shear, laminar = ∫ 𝜏wall 𝑊 d𝑥 (65)
and 18 for 𝛿 (one for the laminar section, the other for the turbulent section). In both, the
𝑥=0
Models for the boundary layer thickness are given in the formula sheet as equations 16
𝑥=𝑥tr.
√
= ∫ 0,332 𝜌𝜇 𝑈 1,5 𝑥 − 2 𝑊 d𝑥
1
(66)
𝑥=0 Reynolds number [Re]𝑥 appears in the denominator (the lower part of the fraction). As
√ 𝑥tr.
𝑈 is increased, [Re]𝑥 will increase too, and consequently, the thickness of the boundary
= 0,332 𝜌𝜇 𝑈 1,5 𝑊 ∫ 𝑥 −0,5 d𝑥 (67)
0
√ 1
𝑥
= 0,332 𝜌𝜇 𝑈 1,5 𝑊 𝑥 −0,5+1
layer will decrease.
[ −0,5 + 1 ]0
tr.
√
(68)
= 0,332 𝜌𝜇 𝑈 1,5 𝑊 2 𝑥tr.0,5
√
(69)
𝐹𝜏 laminar = 0,664 𝜌𝜇 𝑈 2 𝑊 𝑥tr.2
3 1
(70)
4.3 Value of shear in the laminar section
We simply insert values into eq. 70:
√
𝐹𝜏 laminar = 0,664 1,225 × 1,5 ⋅ 10−5 21 2 × 0,6 × 0,292 2
3 1
(71)
𝐹𝜏 laminar = 0,0887 N (72)
4.4 Shear in turbulent section
The process is the same in the turbulent part of the layer, with 𝜏wall, turbulent derived from
equation (19) in the formula sheet:
1 𝜌𝑈
− 71
𝜏wall, turbulent = 0,027 𝜌𝑈 2 𝑥−7
( 𝜇 )
1
2
(73)
This is integrated with respect to area, with 𝑥 ranging from 𝑥tr. (where the turbulent
section begins) to 𝑥max (the trailing edge of the plate):
𝑥=𝑥max
𝐹shear, turbulent = ∫ 𝜏wall 𝑊 d𝑥 (74)
𝑥=𝑥tr.
𝑥max
= 0,0135 𝜌 7 𝑈 7 𝜇 7 𝑊 ∫ 𝑥 − 7 d𝑥
6 13 1 1
(75)
𝑥tr.
𝑥max
= 0,01575 𝜌 7 𝑈 7 𝜇 7 𝑊 [𝑥 7 ]𝑥tr.
6 13 1 6
(76)
= 0,01575 × 1,225 7 × 21 7 × (1,5 ⋅ 10−5 ) 7 × 0,6 × (2 7 − 0,292 7 )
6 13 1 6 6
(77)
𝐹shear, turbulent = 0,961 N (78)
9 10
 We insert the expression for ℎ1 obtained above in this last expression, continuing as:
5 Velocity measurements in a tunnel
1 ℎ2 ℎ2
−𝐹net = −𝜌𝐿 𝑢2(𝑦) d𝑦 + 𝜌𝐿 ∫ 𝑢2(𝑦) d𝑦
𝑈 ∫0
2
(87)
0
5.1 Drag force ℎ2
= 𝜌𝐿 ∫ (𝑢2(𝑦) − 𝑈1 𝑢2(𝑦) ) d𝑦
2
(88)
0
We build a control volume around the object:
Instead of a function 𝑢2 = 𝑓 (𝑦), we have discrete values. The integral is therefore
approximated as:
−𝐹net = 𝜌𝐿 ∑ (𝑢22 − 𝑈1 𝑢2 ) δ𝑦 (89)
𝑦
= 𝜌𝐿 ∑ [𝑢2 (𝑢2 − 𝑈1 )] δ𝑦 (90)
𝑦
−𝐹net = 𝜌𝐿 ∑ [3,2 (3,2 − 3,2)
𝑦
+3,2 (3,2 − 3,2)
+3,15 (3,15 − 3,2)
+3,14 (3,14 − 3,2)
+3,03 (3,03 − 3,2)
We use a mass balance equation to quantify the height ℎ1 of the inlet:
+2,92 (2,92 − 3,2)
d
0 = 𝜌 d + ∬ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴 +2,81 (2,81 − 3,2)
d𝑡 ∭CV
(79)
+2,87 (2,87 − 3,2)
= − ∬ 𝜌|𝑉in | d𝐴 + ∬ 𝜌|𝑉out | d𝐴
CS
+2,89 (2,89 − 3,2)
(80)
ℎ1 ℎ2
= −𝜌𝐿 ∫ 𝑈 d𝑦 + 𝜌𝐿 ∫ 𝑢2(𝑦) d𝑦 (81) +2,97 (2,97 − 3,2)
0 0
+3,19 (3,19 − 3,2)
ℎ2
= −𝜌𝐿𝑈 ℎ1 + 𝜌𝐿 ∫ 𝑢2(𝑦) d𝑦
+3,2 (3,2 − 3,2)
(82)
0
1 ℎ2
ℎ1 = 𝑢2(𝑦) d𝑦
𝑈 ∫0 +3,2 (3,2 − 3,2) ] δ𝑦
(83)
(91)
= 999 × 0,7 × (−5,3325) × 0,05 (92)
scalar equation in the 𝑥-direction: = −186,45 N
The drag force is quantified using a momentum balance equation, which reduces to a
(93)
𝐹net = 186,45 N
d
𝐹⃗net = 𝜌 𝑉⃗ d + ∬ 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 ∭CV

(94)
(84)
−𝐹net = − ∬ 𝜌|𝑉in |2 d𝐴 + ∬ 𝜌|𝑉out |2 d𝐴 𝐹net is the net force exerted on the fluid by the object. It is positive in the 𝑥-direction.
CS
(85)
ℎ2
= −𝜌𝐿ℎ1 𝑈12 + 𝜌𝐿 ∫ 𝑢2(𝑦) d𝑦 opposite direction (flow-wise direction): 𝐹drag = −186,45 N.
The drag force is the force exerted on the object on the fluid, and so is pointing in the
2
(86)
0
11 12
5.2 Dependence on viscosity 6 Lift and drag on a rotating football
Viscosity does not appear in equation (90) above. Nevertheless, an increase in viscosity
will translate into higher shear, and so it is likely that the object will affect a larger 6.1 Required wind tunnel speed
of 𝑢2 in the tabled measurement values). The expression for 𝐹drag will not change, but its
amount of fluid around itself. This will result in a larger velocity deficit (reduced values
The two flows will have identical behavior if the Reynolds numbers are equal. With 1
denoting the real football, and 2 denoting the wind tunnel sphere, we have:
value will increase.
[Re]1 = [Re]2
𝜌𝑉1 𝐷1 𝜌𝑉2 𝐷2
(95)
=
𝜇 𝜇
(96)
𝐷1
𝑉2 = 𝑉1
𝐷2
(97)
70 0,22
=
3,6 1,1
(98)
= 3,89 m s−1 (99)
𝑉2 = 14 km h−1 (100)
6.2 Ratio of forces
Since the two flows are dynamically similar, the force coefficients are the same. Consid-
ering the lift coefficients,
𝐶𝐿1 = 𝐶𝐿2
𝐿 𝐿2
(101)
= 1
1
1
𝜌𝑆 𝑉 2
𝜌𝑆 2 𝑉1
2
(102)
2 1 1 2
𝐿1 𝐿2
=
𝐷12 𝑉12 𝐷22 𝑉12
(103)
𝐿1 𝐷2 𝑉 2
= 1 1
𝐿2 𝐷22 𝑉22
(104)
𝑉2 𝑉2
= 22 12
𝑉1 𝑉2
(105)
𝐿1
= 1
𝐿2
(106)
So, the forces will be identical on both the wind tunnel model and the real football.
13 14
6.3 Rotation speed
The desired lift force is 𝐿1 = 3,1 N. This corresponds to a lift coefficient of:
𝐿1
𝐶𝐿1 = 1
𝜌𝑆1 𝑉12
(107)
2
𝐿1
=
1
𝜌𝜋 𝐷41 𝑉12
2 (108)
2
3,1
=
0,5 × 1,225 × 𝜋 × 0,222
× 19,442
(109)
4
𝐶𝐿1 = 0,352 (110)
Inputting this value in figure 9, one corresponding value of 𝜔𝐷/2𝑈 is 1,51. This allows
us to obtain a value for 𝜔 (other higher values also work):
𝜔1 𝐷1
= 1,51
2𝑈1
(111)
1,51 × 2𝑈1
𝜔1 =
𝐷1
(112)
1,51 × 2 × 19,44
=
0,22
(113)
= 266 rad s−1 (114)
𝜔1 = 42,5 rotations/s (115)
6.4 Drag force
The chosen value of 𝜔𝐷/2𝑈 corresponds to a drag coefficient reading of 0,56 in figure 9.
Inputting this in the definition for the drag coefficient, we can solve for the drag 𝐹𝐷 :
𝐹𝐷1
0,56 = 𝐶𝐷1 = 1
𝜌𝑆 1 𝑉1
2
(116)
2
1 𝐷2
𝐹𝐷1 = 0,56 × 𝜌𝜋 1 𝑉 2
2 4 1
(117)
0,222
= 0,56 × 0,5 × 1,225 × 𝜋 × 19,442
4
(118)
𝐹𝐷1 = 4,93 N (119)
6.5 Doubling of lift force
√ √
The force is “easily” doubled by multiplying the speed 𝑉1 by a factor 2: 𝑉3 = 2𝑉1 .
We obtain the same lift coefficient (𝐶𝐿3 = 𝐶𝐿1 = 0,352). The rotation speed has to be
adapted according to the expression 112: we obtain 𝜔3 = 60,1 rotations/s.
15
 Fluid dynamics examination — September 21, 2020 Shear force on a flat solid surface:
𝐹shear, direction 𝑖 = ∬ 𝜏direction 𝑖 d𝑆 (7)
Fluid Dynamics for Engineers by Olivier Cleynen 𝑆
Shear in the direction 𝑗, on a plane perpendicular to direction 𝑖:
Solve problem 1, plus three other problems among problems 2 to 6.
Duration: 2 h – Use of calculator is authorized; documents are not authorized.
𝜕𝑉𝑗
||⃗
𝜏𝑖𝑗 || = 𝜇
𝜕𝑖
(8)
Except otherwise indicated, assume that:
The atmosphere has 𝑝atm. = 1 bar; 𝜌atm. = 1,225 kg m−3 ; 𝑇atm. = 11,3 ◦C; 𝜇atm. = 1,5 ⋅ 10−5 Pa s
Air behaves as a perfect gas: 𝑅air =287 J kg−1 K−1 ; 𝛾air =1,4; 𝑐𝑝 air =1 005 J kg−1 K−1 ; 𝑐𝑣 air =718 J kg−1 K−1
Continuity equation for incompressible flow:
Liquid water is incompressible: 𝜌water = 1 000 kg m−3 , 𝑐𝑝 water = 4 180 J kg−1 K−1
⃗ ⋅ 𝑉⃗ = 0
∇ (9)
Navier-Stokes equation for incompressible flow:
D𝑉⃗
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
D𝑡
Balance of mass in a fixed control volume with steady flow:
0 = Σ [𝜌𝑉⟂ 𝐴]incoming + Σ [𝜌𝑉⟂ 𝐴]outgoing
(10)
(1)
where 𝑉⟂ is negative inwards, positive outwards.
In a highly-viscous (creeping) steady flow, the drag 𝐹𝐷 exerted on a spherical body of
diameter 𝐷 at by flow at velocity 𝑉∞ is quantified as:
Balance of momentum in a fixed control volume with steady flow:
𝐹⃗net on fluid = Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ] + Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ]
𝐹𝐷sphere = 3𝜋𝜇𝑉∞ 𝐷
(2)
(11)
where 𝑉⟂ is negative inwards, positive outwards.
incoming outgoing
Balance of energy in a fixed control volume with steady flow: In cylindrical pipe flow, we assume the flow is always laminar for [Re]𝐷 . 2 300,
𝑝 1 and always turbulent for [Re]𝐷 & 4 000. The Darcy friction factor 𝑓 is defined as:
𝑄̇ net + 𝑊̇ shaft, net = Σ 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
[ ( 𝜌 2 )]in |Δ𝑝loss |
𝑓 ≡
𝑝 1 𝐿1
𝜌𝑉av.
+Σ 𝑚̇ 𝑖 + + 𝑉 2 + 𝑔𝑧
2
(12)
[ ( 𝜌 )]out
𝐷2
2
The loss coefficient 𝐾𝐿 is defined as:
(3)
where 𝑚̇ is negative inwards, positive outwards.
|Δ𝑝loss |
𝐾𝐿 ≡ 1
𝜌𝑉av.
2
(13)
2
Viscosities of various fluids are given in fig. 1 p. 4. Pressure losses in cylindrical pipes
Mass balance through an arbitrary volume:
d
can be calculated with the help of the Moody diagram presented in fig. 2 p.5.
0 = 𝜌 d + ∬ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴
d𝑡 ∭CV
(4)
CS
Momentum balance through an arbitrary volume: The non-dimensional incompressible Navier-Stokes equation:
d 𝜕 𝑉⃗ ∗ 1 1 ∗2 ⃗ ∗
𝐹⃗net = 𝜌 𝑉⃗ d + ∬ 𝜌 𝑉⃗ (𝑉⃗rel ⋅ 𝑛)
⃗ d𝐴 + [1] 𝑉⃗ ∗ ⋅ ∇
⃗ ∗ 𝑉⃗ ∗ = ⃗ ∗𝑝∗ + ⃗ 𝑉
d𝑡 ∭CV 𝜕𝑡 ∗
𝑔⃗∗ − [Eu] ∇ ∇
(5)
[Fr]2
[St] (14)
CS [Re]
in which [St] ≡ 𝑓𝑉𝐿 , [Eu] ≡ 𝑝𝜌0 −𝑝
𝑉2
∞
, [Fr] ≡ √𝑉 and [Re] ≡ 𝜌 𝑉𝜇 𝐿 .
𝑔𝐿
Angular momentum balance through an arbitrary volume:
d
𝑀⃗ net,X = 𝑟⃗X𝑚 ∧ 𝜌 𝑉⃗ d + ∬ 𝑟⃗X𝑚 ∧ 𝜌 (𝑉⃗rel ⋅ 𝑛)
⃗ 𝑉⃗ d𝐴
d𝑡 ∭CV
(6)
CS
1 2
The force coefficient 𝐶𝐹 and power coefficient 𝐶P are defined as:
𝐹 𝑊̇
𝐶𝐹 ≡ 𝐶P ≡
1
𝜌𝑆𝑉 2 1
𝜌𝑆𝑉 3
(15)
2 2
The speed of sound 𝑐 in air is modeled as:
√
𝑐 = 𝛾 𝑅𝑇 (16)
2×10−2 2.4×10−5
In boundary layer flow, we assume that transition occurs at [Re]𝑥 ≈ 5 ⋅ 105 .
The wall shear coefficient 𝑐𝑓 , a function of distance 𝑥, is defined using the free-stream
flow velocity 𝑈 : 10−2 2.2×10−5
9×10−3
𝜏wall 8×10−3
𝑐𝑓(𝑥) ≡ 7×10−3 ⟵ Crude Oil
1
𝜌𝑈 2 6×10−3 Air ⟶
(17)
5×10−3
2
2×10−5
Exact solutions to the laminar boundary layer along a smooth surface yield: 4×10−3
𝛿 4,91 𝛿∗ 1,72 3×10−3

Viscosity 𝜇 of liquids in Pa s

Viscosity 𝜇 of gases in Pa s
= √ = √
𝑥 𝑥 1.8×10−5
2×10−3
(18)
[Re]𝑥 [Re]𝑥
𝛿 ∗∗ 0,664 0,664
= √ 𝑐𝑓(𝑥) = √ CO2 ⟶
𝑥
(19)
[Re]𝑥 [Re]𝑥
10−3 1.6×10−5
9×10−4
8×10−4
7×10−4
Solutions to the turbulent boundary layer along a smooth surface yield the following
time-averaged characteristics: 6×10−4
𝛿 0,16 𝛿∗ 0,02 5×10−4
≈ ≈ 1.4×10−5
𝑥 1
𝑥 1 (20) 4×10−4 ⟵ Water
3×10−4
[Re]𝑥7 [Re]𝑥7
𝛿 ∗∗ 0,016 0,027
≈ 𝑐𝑓(𝑥) ≈
𝑥 2×10−4 1.2×10−5
1 1 (21)
[Re]𝑥7 [Re]𝑥7
10−4 10−5
−20 0 20 40 60 80 100 120
Temperature 𝑇 in degree Celsius (◦C)
Figure 1 – The viscosity of four fluids (crude oil, water, air, and C02) as a function of temperature. The scale
for liquids is logarithmic and displayed on the left; the scale for gases is linear and displayed on the right.
Figure CC-by by Arjun Neyyathala & Olivier Cleynen
3 4
 Solve problem 1,
and three other problems among problems 2 to 6.
The following marking guidelines will be used:
• Answers to questions starting with “show that” should be fully-developed and
continuous;
• In all other questions, the correct result with the correct unit is enough to obtain
full points;
• Illegible or ambiguous answers are always discarded.
Figure 2 – A Moody diagram, which presents values for 𝑓 measured experimentally, as a function of the
diameter-based Reynolds number [Re]𝐷 , for different relative roughness values.
Diagram CC-by-sa S Beck and R Collins, University of Sheffield
5 6
1 Navier-Stokes equation 3 Governing equations
1.1. [5 pts] Write out equation (10), the Navier-Stokes equation for incompressible flow, We consider a two-dimensional fluid flow described with the following velocity field,
in its fully-developed form in three Cartesian coordinates. described in Cartesian coordinates 𝑥 and 𝑦 in m s−1 :
1.2. [5 pts] In which flow conditions does this equation apply? 𝑉⃗ = (2𝐴𝑥 − 𝐶)𝑖⃗ + (−2𝐴𝑦 − 𝐵𝑥)𝑗⃗
where 𝐴, 𝐵, and 𝐶 are all constants.
3.1. [5 pts] Show that this flow satisfies the continuity equation for incompressible flow
(equation 9).
3.2. [10 pts] What is the acceleration field corresponding to this flow?
3.3. [5 pts] What is the value of acceleration at a point of coordinates (3; 3)?
2 Pressure forces on the panels of a barge
3.4. [10 pts] Does a function exist to describe the pressure field of this flow, and if so,
A large barge is being built with the dimensions shown in figure 3. Once completed, it
will be floated in a lake. The bottom panel of the barge will then sit horizontally, 2 m
what is it?
below the surface of the water.
Figure 3 – Basic layout of a barge
2.1. [15 pts] What is the magnitude of the force resulting from pressure efforts on each
of the panels labeled A, B and C?
2.2. [5 pts] How would the force on panel C change if the angle 𝜃 was increased? (briefly
justify your answer, e.g. in 30 words or less)
2.3. [10 pts] What is the weight of the barge?
7 8
4 Pipe installation with a pump-turbine 5 Friction on a flat plate
Engineers in a chemical company would like to pump 50 L s−1 of water at 20 ◦C from A laboratory is developing a special oil with Newtonian fluid characteristics. An experi-
reservoir A to reservoir B in the installation drawn below in figure 4. The network is ment is set up to measure its viscosity. For this, a thin layer of oil is poured at the bottom
built with a cylindrical pipe made out of concrete (surface roughness 𝜖 = 0,25 mm), with of a tank with rectangular walls. A flat plate is moved perfectly horizontally, sitting flush
a diameter of 12 cm. The Y-junction induces a loss coefficient of 0,2. with the surface of the oil. A force measurement is carried out.
The plate has width 𝑊 = 40 cm and length 𝐿 = 60 cm. It is moved at 15 cm s−1 at the
surface of the oil, which is at a distance 𝐷 = 4 mm above the bottom of the tank.
Figure 4 – Basic layout of the piping network
The installation lengths are as follows:
𝐿1 = 15 m 𝐿5 = 16 m Figure 5 – Flat plate moved horizontally at the surface of a layer of oil, at the bottom of a tank
𝐿2 = 12 m 𝐿6 = 6 m
𝐿3 = 8 m 𝐿7 = 5 m
𝐿4 = 14 m
The flow below the plate is smooth, steady, and laminar, so that the velocity distribution
in the oil between the plate and the bottom of the tank is entirely uniform.
4.1. [10 pts] What is the hydraulic power that the pump must provide, in order to deliver
the required volume flow? 5.1. [10 pts] What is the relationship between the viscosity of the oil and the drag force
4.2. [10 pts] On a diagram, represent qualitatively (i.e. without numerical data) the
due to shear on the bottom side of the plate?
pressure distribution along the length of the pipe, indicating both the top and the The magnitude of the drag force resulting from the shear exerted by the oil is measured
bottom paths followed by the water. as 𝐹drag = 0,47 N.
4.3. [10 pts] If the water was to be transferred back from reservoir B to reservoir A, the
5.2. [5 pts] What is the viscosity of the oil?
pump/turbine device would be operated as a turbine. In that case, what would be
the hydraulic power available to the turbine?
9 10
On the top surface of the plate, in the air, there is enough space for a boundary layer to
6 Pickup truck with snow plow
develop. The scientists in the laboratory would like to check that the drag force generated
A pickup truck is equipped with a snow plow blade. It travels steadily at 30 km h−1 in
by this boundary layer does not influence the measurement. The air in the room has
properties 1 bar and 20 ◦C.
20 cm of snow with density 𝜌snow ground = 400 kg m−3 , with the blade angled at 𝛼 = 60°.
5.3. [10 pts] Starting with equation 19 for the friction coefficient in a laminar boundary
density 𝜌snow compacted = 600 kg m−3 at velocity 𝑉⃗snow/truck relative to the truck. The cross-
The snow is deflected along the blade, compacted by the movement, and it exits with
sectional area of the rejected snow is 𝐴outlet = 0,55 m2 , measured in a plane perpendicular
layer:
0,664 to 𝑉⃗snow/truck .
𝑐𝑓(𝑥) = √ (19)
[Re]𝑥
show that the relationship between the shear force on the top side of the plate and
the properties of the air is:
𝑝air 𝜇air 𝐿 2 32
1
𝐹top = 0,664𝑊 𝑉
( 𝑅𝑇air ) plate
(22)
5.4. [5 pts] What is the magnitude of the force exerted by the air on the top surface of
the plate?
Figure 6 – A pickup truck using a large blade to clear snow
Pickup drawing CC-0 by en:Wikipedia User:Wikideas1; diagram CC-0 Olivier Cleynen
6.1. [15 pts] What is the force exerted on the blade by the snow?
6.2. [5 pts] What is the power required for the truck to plow the snow?
6.3. [10 pts] How would the power change if the angle 𝛼 was reduced? (briefly justify
your answer, e.g. in 30 words or less)
11 12
 Solution: Fluid Dynamics examination — Sept. 21, 2020 2 Pressure forces on the panels of a barge
Fluid Dynamics for Engineers by Olivier Cleynen
https://fluidmech.ninja/ 2.1 Magnitude of the forces
We define coordinates 𝑟, 𝑟𝐶 , and 𝑧, as well as the lengths 𝐿𝐴 , 𝐿𝐵 and 𝐿𝐶 , as shown in the
1 Governing equation figure below.
1.1 N-S equation question
𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑝 𝜕 2𝑢 𝜕 2𝑢 𝜕 2𝑢
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑥 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑥 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑝 𝜕 2𝑣 𝜕 2𝑣 𝜕 2𝑣
(1)
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑦 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑦 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑝 𝜕 2𝑤 𝜕 2𝑤 𝜕 2𝑤
(2)
𝜌 +𝑢 +𝑣 +𝑤 = 𝜌𝑔𝑧 − +𝜇 + +
[ 𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 ] 𝜕𝑧 [ (𝜕𝑥)2 (𝜕𝑦)2 (𝜕𝑧)2 ]
(3)
1.2 Conditions for this equation On each wall, the net pressure applying is that due to water and air from both sides,
This equation applies to all incompressible flows of a Newtonian fluid. 𝑝net = 𝑝water + 𝑝atm. − 𝑝atm. (4)
= 𝜌𝑔𝑧 (5)
Wall A
The force on a small horizontal strip of panel with height d𝑟 and width 𝐿A is d𝐹A . On the
complete door, the force applying due to the net pressure 𝑝net of water and air is:
𝑟=𝑅max
𝐹netA = ∫ d𝐹A (6)
𝑟=0
𝑟=𝑅max
= ∫ 𝑝net d𝑆 (7)
𝑟=0
𝑟=𝑅max
= ∫ 𝑝net 𝐿A d𝑟 (8)
𝑟=0
𝑟=𝑅max
= ∫ 𝜌 𝑔 𝑧 𝐿A d𝑟 (9)
𝑟=0
1 2
A coordinate transform is needed to solve the integral, expressing depth 𝑧 as a function Wall C
of 𝑟. This is obtained by geometry: 𝐹net C is found with a similar calculation. This time, the relationship between 𝑟C and 𝑧 is
𝑧 = 𝑍max −𝑟
a little more complicated, and found by trigonometry:
(10)
𝑧 = 𝑍max − 𝑟C sin 𝜃 (26)
Inserting eq. 10 into eq. 9, we continue with:
𝑟=𝑅max
𝐹net A = ∫ 𝜌 𝑔 (𝑍max − 𝑟) 𝐿A d𝑟
Picking up eq. 9 applied to panel C and inserting eqs. 10 and 26, we get:
(11)
𝑟=0
𝑟C =𝑅Cmax
𝐹netC = ∫ 𝜌 𝑔 (𝑍max − 𝑟C sin 𝜃) 𝐿C d𝑟C
𝑟=𝑅max
= 𝜌 𝑔 𝐿A ∫ (𝑍max − 𝑟) d𝑟 (27)
𝑟C =0
(12)
𝑟=0
1 𝑟=𝑅max
𝑟C =𝑅Cmax
= 𝜌 𝑔 𝐿A [𝑍max 𝑟 − 𝑟 2 ] = 𝜌 𝑔 𝐿C ∫ (𝑍max − 𝑟C sin 𝜃) d𝑟C
2 𝑟=0
(13) (28)
𝑟C =0
= 𝜌 𝑔 𝐿A (𝑍max 𝑅max − 0,5𝑅max 2 ) 1 𝑟C =𝑅Cmax
= 𝜌 𝑔 𝐿C 𝑍max 𝑟C − sin 𝜃𝑟C2 ]
[ 2
(14)
= 103 × 9,81 × 20 × (2 × 2 − 0,5 × 22 )
(29)
𝑟C =0
1
= 𝜌 𝑔 𝐿C 𝑍max 𝑅Cmax − sin 𝜃𝑅 2 )
(15)
= 3,924 ⋅ 105 N ( 2
(30)
𝑍max 1 𝑍2
(16) Cmax
𝐹net A = 392,4 kN = 𝜌𝑔𝐿 𝑍 − sin 𝜃 max 2
( max sin 𝜃 2 (sin 𝜃) )
(17) (31)
1 1
C
= 𝜌 𝑔 𝐿C 𝑍2 − 𝑍2
sin 𝜃 ( max 2 max )
(32)
1 1 2
= 𝜌 𝑔 𝐿C 𝑍
Wall B
𝐹net B 𝐿B is a function of 𝑟, which can be sin 𝜃 2 max
(33)
1
is found with a similar calculation. This time,
= 103 × 9,81 × 6 × × 0,5 × 22
sin 30°
found by trigonometry: (34)
𝑟 = 2,3544 ⋅ 105 N
𝐿B =
(35)
tan 𝜃 𝐹net C = 235,4 kN
(18)
(36)
Picking up eq. 9 applied to panel B, and inserting eqs. 10 and 18, we get:
𝑟
2.2 Change in force with angle
𝑟=𝑅max
𝐹net B = ∫ 𝜌 𝑔 (𝑍max − 𝑟) d𝑟
tan 𝜃 If the angle 𝜃 was increased, the term 1/ sin 𝜃 in eq. 33 would decrease, and so the
(19)
𝑟=0
1 𝑟=𝑅max
= 𝜌𝑔 (𝑍max 𝑟 − 𝑟 2 ) d𝑟
tan 𝜃 ∫
(20) calculated force would decrease.
𝑟=0
1 1 1 𝑟=𝑅max
= 𝜌𝑔 𝑍 𝑟2 − 𝑟3
tan 𝜃 [ 2 3 ]𝑟=0
However, the barge would also sink in more deeply (because the immersed volume, which
(21)
1 1 1
max
is responsible for buoyancy, would have to remain constant). This would result in an
= 𝜌𝑔 𝑍max 𝑅max − 𝑅max 3 ) increase in 𝑍max , likely smaller than the decrease caused by the term 1/ sin 𝜃.
tan 𝜃 ( 2
2 2
3
(22)
1 23
= 103 × 9,81 × × 0,5 × 2 × 22 −
tan 30° ( 3)
(23)
= 2,266 ⋅ 104 N
2.3 Weight of the barge
(24)
𝐹net B = 22,66 kN (25) The weight of the barge is equal to the net force due to pressure on all panels. This can
be calculated either one of two ways:
• Using force components. The weight is equal to the force on the bottom panel,
𝐹D = 𝜌𝑔𝑍max 𝐿C 𝐿A , plus the vertical component of the force on panel C, 𝐹Cvertical =
3 4
 𝐹netC cos 𝜃. The sum is:
3 Governing equations
𝐹weight = 𝐹D + 𝐹Cvertical (37)
= 𝜌𝑔𝑍max 𝐿C 𝐿A + 𝐹netC cos 𝜃 (38) 3.1 Continuity equation
= 2,558 ⋅ 106 N (39)
𝐹weight = 2,558 MN
The incompressible continuity equation (eq. 9 in the formula sheet) translates in two
(40) dimensions as:
𝜕𝑢 𝜕𝑣
+ = 0
𝜕𝑥 𝜕𝑦
• Using the immersed volume and multiplying it by the density of water and gravity (47)
to obtain the buoyancy:
𝐹weight = immersed 𝜌𝑔
In this case, we have
1 𝜕𝑢 𝜕𝑣 𝜕(2𝐴𝑥 − 𝐶) 𝜕(−2𝐴𝑦 − 𝐵𝑥)
(41)
= 𝜌𝑔 (𝐿A + 𝐿Bmax ) 𝐿C 𝑍max + = +
2 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦
(42) (48)
1 𝑍max = 2𝐴 − 2𝐴
= 𝜌𝑔 𝐿A + 𝐿C 𝑍max
( 2 tan 𝜃 )
(49)
= 0 s−1
(43)
1 2
(50)
= 103 × 9,81 × (20 + ×6×2
2 tan 30° )
(44)
= 2,558 ⋅ 106 N (45) So the incompressible continuity equation is satisfied, and the flow is incompressible.
𝐹weight = 2,558 MN (46)
3.2 Acceleration field
The acceleration field 𝑎⃗ (in m s−2 ) is:
The mass of the barge is approximately 260 tons.
D𝑉⃗
𝑎⃗ =
D𝑡
(51)
𝜕 𝑉⃗
= + (𝑉⃗ ⋅ ∇
⃗ )𝑉⃗
𝜕𝑡
(52)
The 𝑥-component of 𝑎⃗ is
𝜕𝑢 𝜕𝑢 𝜕𝑢
𝑎𝑥 = +𝑢 +𝑣
𝜕𝑡 𝜕𝑥 𝜕𝑦
(53)
𝜕(2𝐴𝑥 − 𝐶) 𝜕(2𝐴𝑥 − 𝐶) 𝜕(2𝐴𝑥 − 𝐶)
= + (2𝐴𝑥 − 𝐶) + (−2𝐴𝑦 − 𝐵𝑥)
𝜕𝑡 𝜕𝑥 𝜕𝑦
(54)
= 0 + 2𝐴(2𝐴𝑥 − 𝐶) + 0 (55)
𝑎𝑥 = 4𝐴2 𝑥 − 2𝐴𝐶 (56)
5 6
The 𝑦-component of 𝑎⃗ is Likewise in the 𝑦-direction, we have
𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑝 𝜕 2𝑣
𝑎𝑦 = +𝑢 +𝑣 𝜌𝑎𝑦 = 𝜌𝑔𝑦 − +𝜇
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑦 (𝜕𝑦)2
(57) (70)
𝜕(−2𝐴𝑦 − 𝐵𝑥) 𝜕(−2𝐴𝑦 − 𝐵𝑥) 𝜕(−2𝐴𝑦 − 𝐵𝑥) 𝜕𝑝 𝜕 2 (−2𝐴𝑦 − 𝐵𝑥)
= + (2𝐴𝑥 − 𝐶) + (−2𝐴𝑦 − 𝐵𝑥) 𝜌 (4𝐴2 𝑦 + 𝐵𝐶 ) = 𝜌𝑔𝑦 − +𝜇
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑦 (𝜕𝑦)2
(58) (71)
= 0 − 𝐵(2𝐴𝑥 − 𝐶) − 2𝐴(−2𝐴𝑦 − 𝐵𝑥) 𝜕𝑝
𝜌 (4𝐴2 𝑦 + 𝐵𝐶 ) = 𝜌𝑔𝑦 − +0
𝜕𝑦
(59)
= −2𝐴𝐵𝑥 + 𝐵𝐶 + 4𝐴2 𝑦 + 2𝐴𝐵𝑥
(72)
𝜕𝑝
= 𝜌 (−4𝐴2 𝑦 − 𝐵𝐶 + 𝑔𝑦 )
(60)
𝑎𝑦 = 4𝐴2 𝑦 + 𝐵𝐶 (61) 𝜕𝑦
(73)
So, the acceleration field is, in m s−2 : We have:
𝜕 ( 𝜕𝑥 ) 𝜕 𝜕𝑦
(
𝜕𝑝
)
4𝐴2 𝑥 − 2𝐴𝐶
𝜕𝑝
𝑎⃗ = = 0 =
( 4𝐴2 𝑦 + 𝐵𝐶 ) 𝜕𝑦 𝜕𝑥
(62) (74)
which guarantees that there exists a function to describe 𝑝.
3.3 Acceleration at one point We proceed to integrate eq. 69 with respect to 𝑦:
At point with 𝑥 = 3 and 𝑦 = 3, we have 𝑝 = 𝜌 [−2𝐴2 𝑥 2 + (2𝐴𝐶 + 𝑔𝑥 ) 𝑥 ] + 𝑓(𝑦,𝑡) (75)
12𝐴2 − 2𝐴𝐶 with 𝑓 a function of 𝑦 and 𝑡 only.
𝑎⃗point = m s−2
( 12𝐴2 + 𝐵𝐶 )
(63)
To obtain function 𝑓 , we derivate this equation 75 with respect to 𝑦, and compare with
eq. 73:
𝜕𝑝 d (𝜌 [−2𝐴2 𝑥 2 + (2𝐴𝐶 + 𝑔𝑥 ) 𝑥 ])
3.4 Function for pressure
= + 𝑓(𝑦,𝑡)
′
= 𝜌 (−4𝐴2 𝑦 − 𝐵𝐶 + 𝑔𝑦 )
𝜕𝑦 d𝑦
(76)
0 + 𝑓(𝑦,𝑡) = 𝜌 (−4𝐴2 𝑦 − 𝐵𝐶 + 𝑔𝑦 )
We use the incompressible Navier-Stokes equation (eq. 10 of the equation sheet) to express
′
the acceleration field as a function of the gravity, pressure and shear terms: (77)
D𝑉⃗ We can now integrate this equation 77 with respect to 𝑦 to find the function 𝑓 :
𝜌 = 𝜌 𝑔⃗ − ∇ ⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
D𝑡
(64)
⃗ 2 𝑉⃗
⃗ 𝑝 + 𝜇∇
𝜌 𝑎⃗ = 𝜌 𝑔⃗ − ∇ (65) 𝑓(𝑦,𝑡) = 𝜌 [−2𝐴2 𝑦 2 + (−𝐵𝐶 + 𝑔𝑦 ) 𝑦 ] + 𝑖(𝑡) + 𝑝0 (78)
with 𝑖 a function of 𝑡 only,
Splitting that last equation into two components, we get, for the 𝑥-component: and 𝑝0 an arbitrary integration constant.
𝜕𝑝 𝜕 2𝑢 Inserting this equation 78 into eq. 75, we finally get the function for pressure (in Pa)
𝜌𝑎𝑥 = 𝜌𝑔𝑥 − +𝜇
𝜕𝑥 (𝜕𝑥)2
(66)
𝜕𝑝 𝜕 2 (2𝐴𝑥 − 𝐶)
corresponding to the given velocity field:
𝜌 (4𝐴2 𝑥 − 2𝐴𝐶 ) = 𝜌𝑔𝑥 − +𝜇
𝜕𝑥 (𝜕𝑥)2 𝑝 = 𝜌 [−2𝐴2 𝑥 2 + (2𝐴𝐶 + 𝑔𝑥 ) 𝑥 ] + 𝜌 [−2𝐴2 𝑦 2 + (−𝐵𝐶 + 𝑔𝑦 ) 𝑦 ] + 𝑖(𝑡) + 𝑝0
(67)
𝜕𝑝
𝜌 (4𝐴2 𝑥 − 2𝐴𝐶 ) = 𝜌𝑔𝑥 − +0
(79)
𝜕𝑥
(68) 𝑝 = 𝜌 [−2𝐴2 𝑥 2 − 2𝐴2 𝑦 2 + (2𝐴𝐶 + 𝑔𝑥 ) 𝑥 + (−𝐵𝐶 + 𝑔𝑦 ) 𝑦 ] + 𝑖(𝑡) + 𝑝0
𝜕𝑝
(80)
= 𝜌 (−4𝐴2 𝑥 + 2𝐴𝐶 + 𝑔𝑥 )
𝜕𝑥
(69)
All functions 𝑖 of time, and any initial value 𝑝0 , may be inserted in equation 80, and it
will still satisfy the incompressible Navier-Stokes equation.
7 8
 • Pressure drop due to wall friction losses in each of the two branches, Δ𝑝𝑓 2 : The
4 Pipe installation with a pump-turbine
average velocity in the pipe is
𝑉av.1
𝑉av.2 =
2
4.1 Hydraulic power of pump (95)
𝑉av.2 = 2,21 m s−1 (96)
We want to calculate four pressure differences:
• Pressure drop due to wall friction losses along the horizontal pipe, Δ𝑝𝑓 1 : The The Reynolds number is
[Re]𝐷 2 =
average velocity in the pipe is
[Re]𝐷 1
̇ 2
(97)
𝑉av.1 = [Re]𝐷 2 = 2,653 ⋅ 105
𝑆
(81)
̇
(98)
= 𝐷2
𝜋 4
(82)
0,05
With the same relative roughness, the Moody diagram reads:
=
𝜋 × 0,12 𝑓2 = 0,025
2 (83)
4
𝑉av.1 = 4,42 m s−1
(99)
(84)
Finally, the wall friction losses along the pipe are calculated as:
1 2 𝐿2
The Reynolds number is
𝜌𝑉av.1 𝐷 Δ𝑝𝑓 2 = −𝑓2 𝜌𝑉av.2
[Re]𝐷 1 = 2 𝐷
(100)
𝜇 12
= −0,025 × 0,5 × 103 × 2,212 ×
(85)
103 × 4,42 × 0,12 0,12
=
(101)
10−3 Δ𝑝𝑓 2 = −6,108 ⋅ 103 Pa
(86)
[Re]𝐷 1 = 5,305 ⋅ 105
(102)
= −0,06 bar
(87)
(103)
• Pressure drop due to losses in the junction, Δ𝑝junction :
The relative roughness is
𝜖 0,25 ⋅ 10−3 1 2
= Δ𝑝junction = 𝐾𝐿 𝜌𝑉av.1
𝐷 0,12 2
(88)
𝜖
(104)
= 2,08 ⋅ 10−3 1
𝐷 = 0,2 × × 103 × 4,4212
2
(89)
(105)
Δ𝑝junction = −1,954 ⋅ 103 Pa (106)
= −0,02 bar
With those values, the Moody diagram reads:
(107)
𝑓1 = 0,0241
• Pressure drop due to hydrostatic pressure change across the pump, Δ𝑝ℎ :
(90)
Δ𝑝ℎ = 𝜌𝑔(𝐿5 + 𝐿6 − 𝐿3 )
Finally, the wall friction losses along the pipe are calculated as:
1 2 𝐿1
(108)
Δ𝑝𝑓 1 = −𝑓1 𝜌𝑉av. = 103 × 9,81 × (16 + 6 − 8)
2 𝐷
(91) (109)
15 Δ𝑝ℎ = +1,373 ⋅ 105 Pa
= −0,0241 × 0,5 × 103 × 4,422 ×
0,12
(110)
= +1,37 bar
(92)
Δ𝑝𝑓 1 = −2,944 ⋅ 104 Pa
(111)
(93)
= −0,29 bar (94)
9 10
Finally, the pump hydraulic power is obtained as:
5 Friction on a flat plate
̇
𝑊̇ pump = (Δ𝑝pump ) (112)
= ̇ (Δ𝑝ℎ − Δ𝑝𝑓 1 − Δ𝑝𝑓 2 − Δ𝑝junction )
The flow can be represented as shown below:
(113)
= 0,05 × [+1,373 ⋅ 105 − (−2,944 ⋅ 104 ) − (−6,108 ⋅ 103 ) − (−1,954 ⋅ 103 )] (114)
= 0,05 × [+1,373 + 0,2944 + 0,0611 + 0,01954] × 105 (115)
= +8,742 ⋅ 103 W (116)
𝑊̇ pump = +8,74 kW (117)
4.2 Pressure distribution
5.1 Shear force on bottom side
In order to calculate shear on the bottom face, we express the horizontal fluid velocity
below the plate (relative to the plate) in a very simple velocity field:
𝑉𝑥 = 𝑢 = 0 + 𝑘𝑦 (124)
The constant 𝑘 is found using boundary conditions: at a distance 𝐷 away from the plate,
4.3 Hydraulic power of turbine the velocity of the fluid relative to the plate is 𝑉plate :
𝑢|@ y = D = 𝑉plate = 0 + 𝑘𝐷
𝑉plate
If the flow runs from B to A, the magnitude of the pressure differences remains the same. (125)
𝑘 =
𝐷
The hydrostatic pressure difference changes sign, but the pressure losses due to friction (126)
do not:
̇
𝑊̇ turbine = (Δ𝑝turbine ) (118) So the velocity field becomes:
= ̇ (Δ𝑝ℎturbine − Δ𝑝𝑓 1 − Δ𝑝𝑓 2 − Δ𝑝junction ) 1
𝑢 = 𝑉plate 𝑦
(119)
= ̇ (−Δ𝑝ℎpump − Δ𝑝𝑓 1 − Δ𝑝𝑓 2 − Δ𝑝junction ) 𝐷
(127)
(120)
= 0,05 × [−1,373 + 0,2944 + 0,0611 + 0,01954] × 105 (121)
= −4,992 ⋅ 103 W (122)
𝑊̇ turbine = −4,99 kW (123)
So, pumping the flow from A to B costs 8,7 kW of power, but turbining the flow back
down from B to A only generates 5 kW of power.
11 12
Now, the shear on the bottom plate is found through integration, starting with equation 7 reading the value 𝜇air = 1,85 ⋅ 10−5 Pa s in figure 1, we have:
𝜌 𝑈𝑥
[Re]𝑥tr. = air tr.
from the formula sheet:
𝜇air
(138)
𝐹bottom = ∫ d𝐹 = ∬ 𝜏plate d𝑆 [Re]𝑥tr. 𝜇air
𝑥tr. =
(128)
𝜌air 𝑈
𝑆
= ∬ 𝜏plate d𝑆
(139)
𝑝air [Re]𝑥tr. 𝜇air
=
(129)
𝑥=𝐿
𝑅𝑇air 𝑈
= ∫ 𝜏plate 𝑊 d𝑥
(140)
1 ⋅ 105 × 5 ⋅ 105 × 1,85 ⋅ 10−5
=
(130)
𝑥=0
287 × (273,15 + 20) × 0,15
(141)
Expressing the shear 𝜏plate as a function of the velocity field in a Newtonian fluid, we 𝑥tr. = 73,3 m (142)
continue to get:
Since 𝑥tr. > 𝐿, the boundary layer never transits and remains completely laminar.
𝑥=𝐿
𝜕𝑢 ||
𝐹bottom = ∫ 𝜇oil 𝑊 d𝑥
𝜕𝑦 ||@𝑦=0 To find an expression for 𝜏 , we start with the given equation and implement the defi-
(131)
𝑥=0
( 𝐷 plate 𝑦 ) ||
𝑥=𝐿 𝜕 1 𝑉
= 𝜇oil 𝑊 ∫ | d𝑥
𝜕𝑦 |
nition 15 of the formula sheet, as well as the definition of the distance-based Reynolds
|@𝑦=0
(132)
𝑥=0 number [Re]𝑥 :
𝑥=𝐿
1
= 𝜇oil 𝑊 ∫ ( 𝐷 𝑉plate )@𝑦=0 d𝑥 0,664
𝑐𝑓(𝑥) = √
(133)
𝑥=0
1
(143)
𝐹bottom = 𝜇oil 𝑊 𝑉plate 𝐿
[Re]𝑥
𝐷 𝜏 laminar 0,664
= √
(134)
1
𝜌 𝑈2
wall,
𝜌air 𝑈 𝑥
(144)
2 air
Rearranging this equation to solve for 𝜇, we get:
𝜇air
𝜌air 𝑈 𝑥 2 1
−1
𝐷 𝜏 = 0,664 𝜌 𝑈2
( 𝜇air ) 2 air
𝜇oil = 𝐹bottom
(145)
𝑊 𝑉plate 𝐿 √
wall, laminar
𝜏wall, laminar = 0,332 𝜌air 𝜇air 𝑈 1,5 𝑥 − 2
(135) 1
(146)
5.2 Viscosity of oil The shear force on the top surface is the integral of the shear with respect to area:
The viscosity of the oil is obtained by plugging in values in eq. 135: 𝐹top = ∫ 𝜏wall d𝑆 (147)
4 ⋅ 10−3 We split the total area covered by the laminar boundary layer in strips of width 𝑊 and
𝜇oil = 0,47 ×
0,4 × 0,15 × 0,6
length d𝑥, with 𝑥 ranging from 0 (leading edge) to 𝐿 (trailing edge), obtaining:
(136)
𝜇oil = 5,22 ⋅ 10−2 Pa s (137)
𝑥=𝐿
𝐹top = ∫ 𝜏wall 𝑊 d𝑥 (148)
𝑥=0
𝑥=𝐿
√
5.3 Shear force on top side
= ∫ 0,332 𝜌air 𝜇air 𝑈 1,5 𝑥 − 2 𝑊 d𝑥
1
(149)
𝑥=0
√ 𝐿
= 0,332 𝜌air 𝜇air 𝑈 1,5 𝑊 ∫ 𝑥 −0,5 d𝑥
A boundary layer develops on the top side, in air. That boundary layer would transit at
[Re]𝑥 ≈ 5 ⋅ 105 . Solving for 𝑥tr. with 𝑈 (the faraway velocity) being equal to 𝑉plate , and (150)
0
√ 1
𝐿
= 0,332 𝜌air 𝜇air 𝑈 1,5 𝑊 𝑥 −0,5+1
[ −0,5 + 1 ]0
(151)
√
= 0,332 𝜌air 𝜇air 𝑈 1,5 𝑊 2 𝐿0,5 (152)
13 14
 √
= 0,664 𝜌air 𝜇air 𝑈 2 𝑊 𝐿 2
3 1
(153)
6 Pickup truck with snow plow
= 0,664𝑊 (𝜌air 𝜇air 𝐿) 2 𝑈 2
1 3
(154)
𝑝air 𝜇air 𝐿 2 23
1
= 0,664𝑊 𝑈
( 𝑅𝑇air )
enters with velocity 𝑉⃗1 and leaves with velocity 𝑉⃗2 = 𝑉⃗out .
(155) We define a control volume moving together with the truck, as shown below. The snow
𝑝air 𝜇air 𝐿 2 23
1
𝐹top = 0,664𝑊 𝑉
( 𝑅𝑇air ) plate
(156)
5.4 Force exerted by the air
We simply insert values into eq. 156:
1 ⋅ 105 × 1,85 ⋅ 10−5 × 0,6
0,5
𝐹top = 0,664 × 0,4 × × 0,151,5
( 287 × (273,15 + 20) )
(157)
𝐹top = 5,6 ⋅ 10−5 N (158)
This force is is very much smaller than the force generated by the oil, so the air’s influence
in the viscosity measurement experiment is indeed very small.
6.1 Force on blade
̇ as well as for the two vectors 𝑉⃗1 and 𝑉⃗2 .
We want expressions for the mass flow 𝑚,
The mass flow is found using the inlet conditions:
𝑚̇ = 𝜌1 𝑉1 𝐴⟂1 (159)
= 𝜌𝑉𝑊
1 1 ℎsnow
30
(160)
= 400 × × 3,5 × 0,2
sweep
3,6
(161)
𝑚̇ = 2,333 ⋅ 103 kg s−1 (162)
The outlet velocity can then be obtained:
𝑚̇ = 𝜌2 𝑉2 𝐴⟂2 (163)
𝑚̇ = 𝜌2 𝑉2 𝐴out
𝑚̇
(164)
𝑉2 =
𝜌2 𝐴out
(165)
2,333 ⋅ 103
=
600 × 0,55
(166)
𝑉2 = 7,07 m s−1 (167)
15 16
The velocity vector 𝑉⃗2 is then expressed according to its two components: ⎛ +1,429 ⋅ 104 ⎞
⎜ ⎟
= ⎜ 0 ⎟N
𝑉2𝑥 = +𝑉2 sin 𝛼 ⎜ ⎟
(181)
(168)
⎝ −1,12 ⋅ 104 ⎠
= 7,07 × sin(60°) ⎛ +14,3 ⎞
⎜ ⎟
(169)
𝑉2𝑥 = +6,12 m s−1 𝐹⃗net = ⎜ 0 ⎟ kN
⎜ ⎟
(170) (182)
⎝ −11,2 ⎠
𝑉2𝑧 = +𝑉2 cos 𝛼 (171)
= 7,07 × cos(60°)
The force exerted on the blade by the snow is exactly the opposite, i.e.
⎛ −14,3 ⎞
(172)
𝑉2𝑧 = +3,54 m s−1 ⎜ ⎟
𝐹⃗snow on blade = −𝐹⃗net = ⎜ 0 ⎟ kN
(173)
⎜ ⎟
(183)
So, we have all the necessary information about the flow: ⎝ +11,2 ⎠
𝑚̇ = 2,333 ⋅ 103 kg s−1 (174) 6.2 Power required of truck
⎛ 0 ⎞
⎜ ⎟ The power 𝑊̇ truck provided by the truck to compensate for the force is
𝑉⃗1 = ⎜ 0 ⎟ m s−1
⎜ ⎟
(175)
⎝ +8,33 ⎠
𝑊̇ truck = 𝐹⃗net ⋅ 𝑉⃗truck on ground
⎛ +6,12 ⎞
(184)
⎜ ⎟ = −𝑉truck 𝐹net 𝑧
𝑉⃗2 = ⎜ 0 ⎟ m s−1
(185)
⎜ ⎟ = −8,333 × (−1,12 ⋅ 104 )
(176)
⎝ +3,54 ⎠
(186)
= 9,329 ⋅ 104 W (187)
Now, applying eq. 2 from the equation sheet, we can express the net force 𝐹⃗net on the 𝑊̇ truck = 93,3 kW (188)
snow:
𝐹⃗net = Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ] + Σ [𝜌𝑉⟂ 𝐴𝑉⃗ ]
(approximately 120 horsepower)
(177)
= 𝑚̇ (𝑉⃗2 − 𝑉⃗1 )
incoming outgoing
(178) 6.3 Power change with angle
⎛ 𝑉2𝑥 − 𝑉1𝑥 ⎞
⎜ ⎟
= 𝑚̇ ⎜ 𝑉2𝑦 − 𝑉1𝑦 ⎟
The power is quantified using equation 185, in which we include eq. 179, eq. 171, and
⎜ ⎟
(179)
⎝ 𝑉2𝑧 − 𝑉1𝑧 ⎠
then eq. 160:
⎛ (+6,12) − 0 ⎞ ̇
𝑊truck = −𝑉truck 𝐹net 𝑧
⎜ ⎟
(189)
= 2,333 ⋅ 103 ⎜ 0−0 ⎟ = −𝑉truck 𝑚̇ (𝑉2𝑧 − 𝑉1𝑧 )
⎜ ⎟
(180)
⎝ (+3,54) − (+8,33)
(190)
⎠ = −𝑉truck 𝑚̇ (𝑉2 cos 𝛼 − 𝑉1𝑧 )
𝑚̇
(191)
= −𝑉truck 𝑚̇ cos 𝛼 − 𝑉truck
( 𝜌2 𝐴out )
(192)
17 18
 𝜌 𝑉 𝑊sweep ℎsnow
= −𝑉truck 𝑚̇ cos 𝛼 − 𝑉truck
1 1
( 𝜌2 𝐴out )
(193)
𝜌1 𝑉truck 𝑊sweep ℎsnow
= −𝑉truck 𝜌𝑉truck ℎsnow 𝑊sweep cos 𝛼 − 𝑉truck
( 𝜌2 𝐴out )
(194)
𝜌1 𝑊sweep ℎsnow
= −𝑉truck
3
ℎsnow 𝑊sweep cos 𝛼 − 1
( 𝜌2 𝐴out )
(195)
In equation 195, decreasing 𝛼 would:
• increase the term cos 𝛼;
• decrease the effective width 𝑊sweep (unless a new, larger blade is used).
Both factors tend to decrease the magnitude of 𝑊̇ truck . Therefore, the power would
decrease, likely faster than 𝛼.
19
 A8 References

These notes are based on textbooks by White [22], Çengel & al.[25], Munson & al.[29], and de Nevers [17].

[1] Ira Herbert Abbott and Albert Edward Von Doenhoff. Theory of wing
sections, including a summary of airfoil data. Courier, 1959.
[2] George Keith Batchelor. An introduction to fluid dynamics. Cambridge
University Press, 1967. isbn: 9780521663960.
[3] Geoffrey Ingram Taylor. “Film notes for low-Reynolds-number flows”.
In: National Committee for Fluid Mechanics Films. Illustrated experi-
ments in fluid mechanics: the NCFMF book of film notes. MIT Press, 1967.
isbn: 9780262640121. url: http://web.mit.edu/hml/ncfmf/07LRNF.pdf.
[4] John Leask Lumley. “Film notes for Eulerian and Lagrangian descrip-
tions in fluid mechanics”. In: National Committee for Fluid Mechanics
Films. Illustrated experiments in fluid mechanics: the NCFMF book of
film notes. MIT Press, 1969. isbn: 9780262640121. url: http://web.mit.
edu/hml/ncfmf/01ELDFM.pdf.
[5] Hendrik Tennekes and John Leask Lumley. A first course in turbulence.
MIT press, 1972. isbn: 978-0262200196.
[6] Bernard Stanford Massey. Mechanics of fluids. 5th ed. Van Nostrand
Reinhold, 1983. isbn: 0442305524.
[7] Phillip M. Gerhart and Richard J. Gross. Fundamentals of fluid dynamics.
1st ed. Addison-Wesley, 1985. isbn: 0201114100.
[8] Hendrik Tennekes. The simple science of flight. From insects to jumbo
jets. 1st ed. MIT Press, 1992. isbn: 0262201054.
[9] John David Anderson. Computational fluid dynamics. McGraw-Hill,
1995. isbn: 0071132104.
[10] Paul A. Libby. An introduction to turbulence. CRC Press, 1996. isbn:
9781560321002.
[11] Jewel B. Barlow, William H. Jr. Rae, and Alan Pope. Low-speed wind
tunnel testing. 3rd ed. Wiley & Sons, 1999. isbn: 0471557749.
[12] Jean Mathieu and Julian Scott. An introduction to turbulent flow. Cam-
bridge University Press, 2000. isbn: 0521775388.
[13] Stephen B. Pope. Turbulent flows. Cambridge University Press, 2000.
isbn: 9780521177849.
[14] Peter S. Bernard and James M. Wallace. Turbulent flow: analysis, mea-
surement, and prediction. John Wiley & Sons, 2002. isbn: 9780471332190.
[15] John Bird. Higher engineering mathematics. 4th ed. Newnes, 2004. isbn:
9780750662666.
[16] Tuncer Cebeci. Analysis of turbulent flows. 2nd ed. Elsevier, 2004. isbn:
9780080443508.
[17] Noel de Nevers. Fluid mechanics for chemical engineers. 3rd ed. McGraw-
Hill, 2004. isbn: 9781259002380.
[18] Pierre Sagaut. Large eddy simulation for incompressible flows: an in-
troduction. Ed. by Charles Meneveau. 3rd ed. Springer, 2006. isbn:
9783540263449.
292
[19] James O. Wilkes. Fluid mechanics for chemical engineers. with microflu-
idics and CFD. 2nd ed. with contributions by S. G. Birmigham, B. J.
Kirby, COMSOL (FEMLAB), and C-Y. Cheng. Prentice Hall, 2006. isbn:
9780131482128.
[20] Henk Kaarle Versteeg and Weeratunge Malalasekera. An introduction to
computational fluid dynamics: the finite volume method. 2nd ed. Pearson
Education, 2007. isbn: 9780131274983.
[21] Ascher Shapiro and the Massachusetts Institute of Technology. Na-
tional Committee for Fluid Mechanics Films. 2008. url: http://web.mit.
edu/hml/ncfmf.html (visited on 09/23/2017).
[22] Frank M. White. Fluid mechanics. 7th ed. McGraw-Hill, 2008. isbn:
9780071311212.
[23] Hendrik Tennekes. The simple science of flight. From insects to jumbo
jets. 2nd ed. MIT Press, 2009. isbn: 9780262513135.
[24] J.J Bertin and R.M. Cummings. Aerodynamics for engineers. 5th ed.
Pearson/Prentice Hall, 2010. isbn: 978-0132272681.
[25] Yunus A. Çengel and John M. Cimbala. Fluid mechanics. Fundamentals
and applications. 2nd ed. McGraw-Hill, 2010. isbn: 9780070700345.
[26] Philipp Schlatter, Mattias Chevalier, Miloš Ilak, and Dan S. Henning-
son. The structure of a turbulent boundary layer studied by numerical
simulation. 2010. arXiv: 1010.4000 [physics.flu-dyn].
[27] Philipp Schlatter and Ramis Örlü. “Assessment of direct numerical
simulation data of turbulent boundary layers”. In: Journal of Fluid
Mechanics 659 (2010), p. 116. doi: 10.1017/S0022112010003113.
[28] J. H. Lee, Y. S. Kwon, N. Hutchins, and J. P. Monty. Spatially devel-
oping turbulent boundary layer on a flat plate. 2012. arXiv: 1210.3881
[physics.flu-dyn].
[29] Bruce R. Munson, Theodore H. Okiishi, Wade W. Huebsch, and Alric P.
Rothmayer. Fluid mechanics. 7th ed. Wiley, 2013. isbn: 9781118318676.
[30] Jiyuan Tu, Guan Heng Yeoh, and Chaoqun Liu. Computational fluid
dynamics: a practical approach. 2nd ed. Butterworth-Heinemann, 2013.
isbn: 9789382291787.
[31] Peter Alan Davidson. Turbulence: an introduction for scientists and
engineers. 2nd ed. Oxford University Press, 2015. isbn: 9780198779469.
[32] Michael Leschziner. Statistical turbulence modelling for fluid dynam-
ics demystified. An introductory text for graduate engineering students.
World Scientific, 2015. isbn: 9781783266630.
[33] Zhan Su, Jinbo Wang, Patrice Klein, Andrew F Thompson, and Dimitris
Menemenlis. “Ocean submesoscales as a key component of the global
heat budget”. In: Nature communications 9.1 (2018), p. 775. doi: 10 .
1038/s41467-018-02983-w.
[34] Olivier Cleynen. Git repository: Fluid dynamics. Under CC-by-nc license.
2020. url: https : / / framagit . org / olivier / fluidmech (visited on
04/20/2020).
[35] Olivier Cleynen, Germán Santa-Maria, Mathias Magdowski, and Do-
minique Thévenin. “Peer-graded individualized student homework in
a single-instructor undergraduate engineering course”. In: Research in
Learning Technology 28 (2020). doi: 10.25304/rlt.v28.2339.
293