1

Dimensional Analysis
and Similitude
 Why dimensional analysis?
2

At what speed dinosaurs may have been able to run? Data

available on these creatures are in the fossil record. These are its
average leg length L and stride s. Could these data be used to
extract the dinosaurs’ speed? Comparing data on L and s and
the speed V of quadrupeds (e.g., horses, dogs) and bipeds (e.g.,
humans) does not indicate a pattern, unless dimensional analysis
is used to plot as follows. Plot the dimensionless quantity V2/gL
against the dimensionless ratio s/L. When this is done, magically
the data for most animals fall approximately on one curve!
Hence, the running behavior of most animals can be obtained
from the graph. The dinosaurs’ value of s/L allows a corres-
ponding value of V2/gL to be interpolated from the curve, leading
to an estimate for V of dinosaurs.

Professor Alexander McNeil of a University in Scotland (2011)

3

Dimensional analysis is the study of finding the fewest

possible combinations of the flow and geometric
parameters involved in fluid flows in dimensionless form.
(It is in the interest of saving time and money.)

Similitude is the study of predicting prototype conditions

from model observations. It involves the use of the
dimensionless parameters obtained in dimensional
analysis.
4

The objectives of dimensional analysis are:

 To understand dimensions, units, and dimensional
homogeneity
 To understand benefits of dimensional analysis
 To know how to use the method of repeating variables
 To understand the concept of similarity and how to
apply it to experimental modeling
 Dimensions and Units
5

• Dimension: Measure of a physical quantity, e.g.,

mass, length, time
• Units: Assignment of a number to a dimension, e.g.,
(kg), (m), (sec)
• Primary Dimensions:
1. Mass m (kg)
2. Length L (m)
3. Time t (sec)
4. Temperature T (K)
• Non-primary dimensions can be formed by a
combination of the primary dimensions
Example: Velocity = Length/Time = L/t
 Dimensional Homogeneity
6

 Law of dimensional homogeneity (DH): every term in an

equation must have the same dimensions
 Example: Bernoulli equation

• p = force/area=(mass x length/time2)/length2 = m/Lt2

• V2 = mass/length3 x (length/time)2 = m/Lt2
• gz = mass/length3 x length/time2 x length =m/Lt2
 Nondimensionalization of Equations
7

 In a dimensionally homogeneous equation, if

each term in the equation is divided by a
collection of variables and constants that have
the same dimensions, the equation is rendered
non-dimensional
 In the process of nondimensionalizing an
equation, nondimensional parameters often
appear, e.g., Reynolds number and Froude
number
8

 To nondimensionalize an equation, the first step

is to list the primary dimensions of all variables
and constants, e.g.

p = m/Lt2 = m/L3 V = L/t

g = L/t2 z=L

Next, select scaling parameters. Here L0, U0 .

9

 By inspection, nondimensionalize all variables with

scaling parameters
z  z
p  p V  V Lo
ρU o2 Uo

 Substitute p, V, z into dimensional Bernoulli equation

p  1 V 2  gz  C
2

U o2 p  1 U o2V   gLo z  C
2
2
10

 Divide by U02

 2 gLo 
p  V 
1 z  C / U o2
2 U o2

 2 1  
p  V 
1 z C
2 Fr2

where Fr  Uo / gLo is the Froud number

11

Advantages from nondimensionalization

Increases insight about key parameters

Decreases number of parameters in the problem
•Easier understanding
•Fewer experiments/simulations required
Extrapolation of results to untested conditions
 Problem # 1
12

The equation describing motion of fluid in a pipe due to an

applied pressure gradient, when flow starts from rest, is

Using the average velocity V, pressure drop Δp, pipe

length L, and diameter D, nondimensionalize this
equation and obtain the dimensionless groups that
characterize this flow.
 Problem # 2
13

A 2D steady flow in a viscous liquid in absence of body

force is described by the Continuity and Navier-Stokes
equations as follows.

Nondimensionalize this equation using a length scale L

and a velocity scale V. Obtain the dimensionless
groups that characterize this flow.
 Problem # 3
14

Water at 20°C flows through a long, straight pipe. The

pressure drop is measured along a section of the pipe of
length L=1.3 m as a function of average velocity V through
the pipe as in the table. The inner diameter of the pipe is
D=10.4 cm. (a) Plot the Euler number as a function of the
Reynolds number. Has the experiment been run at high
enough speeds to achieve Reynolds number indepen-
dence? (b) Extrapolate the experimental data to predict the
pressure drop at an average speed of 80 m/s.

Contd.
15
0.310
V(m/s) P(N/m2)
0.5 77
1 306
0.308
2 1218
4 4865
Eu (=p/V )
2
6 10920
0.306 Independent of Re
8 19440
10 30340
15 68330
20 121400 0.304
25 189800
30 273200
35 372100
0.302
40 485586 0 1 2 3 4 5
45 614900 -6
50 758700 10 Re (=VD/)
 Dimensional Analysis and Similarity
16

 Nondimensionalization of an equation is useful

only when the equation is known!
 In many real-world flows, the equations are
either unknown or too difficult to solve.
• Experimentation is the only method of obtaining
reliable information
• In most experiments, geometrically-scaled models
(reduced scale) are used (for saving time and money).
• Experimental conditions and results must be properly
scaled so that results are meaningful for the prototype
(full-scale).
17

 Geometric Similarity - model must be the same

shape as the prototype. Each dimension must be
scaled by the same factor.
 Kinematic Similarity - velocity as any point in the
model must be proportional.
 Dynamic Similarity - all forces in the model flow,
scaled by a constant factor to corresponding forces
in the prototype flow.
 Complete Similarity - it is achieved only if above 3
similarities are met. This is not always possible, e.g.,
in river hydraulics models.
18

Geometric similarity : (1) Prototype , (2) Model

Kinematic similarity : (1) Prototype, (2) Model

19

 Complete similarity is ensured if all independent 

groups are same between the model and prototype.
 What is ?
• Greek letter  denotes a non-dimensional parameter,
e.g.,Reynolds number Re, Froude number Fr, Drag
coefficient, CD, etc.
 Generation of nondimensional
parameters 
20

Method of Repeating Variables (or Buckingham Pi

Theorem):
1. List the parameters in the problem and count their
total number n.
2. List the primary dimensions of each of the n
parameters
3. Set the reduction j as the number of primary
dimensions. Calculate k, the expected number of
's, k = n - j.
4. Choose j as repeating parameters.
21

5. Construct the k 's, and manipulate as necessary.

6. Write the final functional relationship and check
algebra.

Some common established nondimensional parameters ,

encountered in fluid mechanics and heat transfer, are in
Table 7.5 (Cengel & Cimbala)
22

Buckingham Pi Theorem
Buckingham Pi theorem states that a relationship can be
transformed between n parameters as

into a corresponding relationship between n-j indepen-

dent dimensionless Π parameters as
G(1, 2, ……, n-j)=0 or 1=G1(2, ……, n-j)

where j is the minimum number of primary dimensions

required to define the dimensions of all the parameters
q1, q2, . . . , qn.
 Illustration of Buckingham Pi Theorem
23

Friction in a Pipe

Friction on the inside wall of a

pipe. The shear stress w on
the pipe walls is a function of
average fluid speed V, average
wall roughness height , fluid
density , fluid viscosity µ, and
inside pipe diameter D.

Assumptions: 1) flow is fully developed, and 2) no

other parameters are significant in the problem.
24

The primary dimensions of each parameter are

The number of primary dimensions j =3. Expected number of ’s

is k = n-j = 6-3 =3.

3 repeating parameters since j = 3.

D and  cannot be choosen since their dimensions are identical,

and it would not be desirable to have µ or  appear in all the  ’s.
The best choice of repeating parameters is thus V, D, and .
25
26

Similarly, the two independent ’s are generated as

Final functional relationship is

27

Discussion
Applies to both laminar and fully turbulent pipe flow.
2 (i.e. /D) is not nearly as important in laminar pipe flow
…as in turbulent pipe flow.

Here is interesting connection between geometric

…similarity and dimensional analysis. Namely, it is
…necessary to match /D since it is an independent  in
…the problem.

From a different perspective, thinking of roughness as a

…geometric property, it is necessary to match /D to
…ensure geometric similarity between two pipes.
 Guidelines for choosing Repeating
parameters
28

(Although this is often the hardest or at least the most mysterious part)
1. Never pick the dependent variable. As, it may appear in all the 's.
2. Chosen repeating parameters must not by themselves be able to form a
dimensionless group. Otherwise, it would be impossible to generate the
rest of the 's.
3. Chosen repeating parameters must represent all the primary dimensions.
4. Never pick parameters that are already dimensionless.
5. Never pick two parameters with the same dimensions or with dimensions
that differ by only an exponent.
6. Choose dimensional constants over dimensional variables so that only one
 contains the dimensional variable.
7. Pick common parameters since they may appear in each of the 's.
8. Pick simple parameters over complex parameters.
24
 Experiment and Similarity
29

 One of the most useful applications of dimensional

analysis is in designing physical and/or numerical
experiments, and in reporting the results.

• Consider a problem with 5 parameters:

one dependent and 4 independent.
• Full test matrix with 5 data points for
each independent parameter would
require 54=625 experiments!!
• If we can reduce 5 parameters to 2 's,
the number of independent parameters
is reduced from 4 to 1, which results in
51=5 experiments vs. 625!!

29
30

Drag force is F = f(V,  , D), through dimensional analysis it

can be reduced to

30
 Incomplete Similarity
31

 For ship hydrodynamics, Fr similarity is maintained while Re

is allowed to be different. Why?
 Look at complete similarity:

 To match both Re and Fr, viscosity in the model test is a

function of scale ratio! This is not feasible.
 Problem # 4
32

The aerodynamic drag of a new sports car is to be

predicted at a speed of 80 km/h at an air temperature of
25°C. Automotive engineers build a one-fifth scale model
of the car to test in a wind tunnel. It is winter and the wind
tunnel is located in an unheated building; the temperature
of the wind tunnel air is only about 5°C. Determine how
fast the engineers should run the wind tunnel in order to
achieve similarity between the model and the prototype.
 Problem # 5
33

An airship is to operate at 20 m/s in air at standard

conditions. A model is constructed to 1/20-scale and tested
in a wind tunnel at the same air temperature to determine
drag. What criterion should be considered to obtain
dynamic similarity? If the model is tested at 75 m/s, what
pressure should be used in the wind tunnel? If the model
drag force is 250 N, what will be the drag of the prototype?
34