FUNDAMENTALS OF

FLUID MECHANICS
Chapter 7 Dimensional Analysis
Modeling, and Similitude

1
MAIN TOPICS
 Dimensional
Di i l Analysis
A l i
 Buckingham Pi Theorem
 Determination of Pi Terms
 Comments about Dimensional Analysis
 Common Dimensionless Groups in Fluid Mechanics
 Correlation of Experimental
p Data
 Modeling and Similitude
 Typical Model Studies
 Similitude Based on Governing Differential Equation

2
Dimensional Analysis 1/4
 A typical fluid mechanics problem in which
experimentation is required consider the steady flow of an
incompressible Newtonian fluid through a long, smooth
smooth--
walled, horizontal, circular pipe.
pipe.
 An important characteristic of this system,
system, which would
be interest to an engineer designing a pipeline, is the
pressure drop per unit length that develops along the pipe
as a result of friction.

3
Dimensional Analysis 2/4
 The first step in the planning of an experiment to study
this problem would be to decide the factors, or variables,
variables,
that will have an effect on the pressure drop.
drop.
 Pressure dropp pper unit length
g
p  f ( D ,  ,  ,V )

 Pressure drop per unit length depends on FOUR variables:

sphere size (D); speed (V); fluid density (ρ); fluid viscosity
(m)

4
Dimensional Analysis 3/4
 To perform the experiments in a meaningful and
systematic manner, it would be necessary to change the
variable, such as the velocity, which holding all other
constant, and measure the corresponding pressure drop.
 Difficulty to determine the functional relationship between
the pressure drop and the various facts that influence it.

5
Series of Tests

6
Dimensional Analysis 4/4
 Fortunately, there is a much simpler approach to the
problem that will eliminate the difficulties described
above.
Collectingg these variables into two non
non--dimensional
combinations of the variables (called dimensionless
product or dimensionless groups
groups))
 Only one dependent and one
Dp   VD  independent variable
    Easy to set up experiments to
V 2    determine dependency
 Easy to present results (one graph)
7
Plot of Pressure Drop Data Using …
Dp  L( F / L3 )
  F0 L0T 0
V 2 ( FL 4 T 2 )( FT 1 ) 2

VD ( FL4 T 2 )( LT 1 )( L) 0 0 0
 2
 F LT
 ( FL T )

dimensionless product or
dimensionless groups

8
Buckingham Pi Theorem 1/5
 A fundamental question we must answer is how many
dimensionless products are required to replace the original list of
variables ?
 The answer to this question is supplied by the basic theorem of
dimensional analysis that states
If an equation involving k variables is dimensionally
homogeneous it can be reduced to a relationship among
homogeneous,
k-r independent dimensionless products, where r is the
minimum number of reference dimensions required to
describe the variables.
Pi terms
Buckingham Pi Theorem 9
Buckingham Pi Theorem 2/5

 Given a physical problem in which the dependent variable

is a function of kk--1 independent variables
variables..

u 1  f ( u 2 , u 3 ,....., u k )
Mathematically, we can express the functional relationship
in the equivalent form
g ( u 1 , u 2 , u 3 ,,.....,, u k )  0
where g is an unspecified function, different from f.

10
Buckingham Pi Theorem 3/5
 The Buckingham Pi theorem states that: Given a relation
among k variables of the form
g ( u 1 , u 2 , u 3 ,....., u k )  0
 The
Th k variables
i bl may be b grouped d iinto
t k-r independent
i d d t
dimensionless products,
products, or Π terms, expressible in
f ti l form
functional f by
b
 1  (  2 ,  3 , , , ,  k  r )
or (  1 ,  2 ,  3 , , , ,  k  r )  0
r ?? Π??

11
Buckingham Pi Theorem 4/5
 The
Th number b off r is
i usually,
ll but
b not always,
l equall to the
h
minimum number of independent dimensions required to
specify the dimensions of all the parameters.
parameters. Usually the
reference dimensions required to describe the variables
will be the basic dimensions M,
M LL, and T or F,
F L,
L and T. T.
 Th
This
is theorem does not predict the particular functional
form of  or  . The functional relation amongg the
independent dimensionless products Π must be determined
experimentally..
experimentally
 The k-k-r dimensionless Π terms obtained from the
procedure are independent.
12
Buckingham Pi Theorem 5/5

 A Π term is not independent if it can be obtained from a

product or quotient of the other dimensionless products of
the problem. For example, if
3/ 4
21 1
5  or 6 
 2 3  32

then neither Π5 nor Π6 is independent of the other

dimensionless products or dimensionless groups.
groups.

13
Determination of Pi Terms 1/12
 Several methods can be used to form the dimensionless
products, or pi term, that arise in a dimensional analysis.
 Regardless of the method to be used to determine the
dimensionless pproducts, one begins
g byy listingg all
(dimensional
dimensional)) variables that are known (or believed) to
affect the given flow phenomenon.
phenomenon.
 Eight steps listed below outline a recommended
procedure for determining the Π terms.

14
Determination of Pi Terms 2/12

Step 1 List all the variables

variables.. 1
List all the dimensional variables involved.
Keep the number of variables to a minimum,
minimum, so that we
can minimize the amount of laboratory work.work
All variables must be independent.
independent. For example, if the
cross--sectional area of a pipe is an important variable,
cross variable
either the area or the pipe diameter could be used, but
not both,
both since they are obviously not independent
independent..
Ex. γ=ρ×g, that is, γ,ρ, and g are not independent.
15
Determination of Pi Terms 3/12

Step 1 List all the variables. 2

Let k be the number of variables
variables..
Example: For pressure drop per unit length, k=5. (All
variables are p, D,
D ,, and V )
D,

p   f ( D, , , V )

16
Determination of Pi Terms 4/12

Step 2 Express each of the variables in terms of

basic dimensions and find the number of reference
dimensions..
dimensions
Select a set of fundamental (primary) dimensions
dimensions.
For example: MLT, or FLT.
Example: For pressure drop per unit length , we choose
FLT.
p   FL3 D  L   FL4 T 2
2 1
r=3
  FL T V  LT
17
Determination of Pi Terms 5/12

Step 3 Determine the required number of pi terms.

Let k be the number of variables in the problem.
p
Let r be the number of reference dimensions (primary
dimensions) required to describe these variables.
variables
The number of pi terms is kk--r
Example:
l For pressure drop
d per unit
i length
l h k=5,
k r = 3,
3
the number of pi terms is k-r=5
r=5--3=2
3=2..

18
Determination of Pi Terms 6/12
Step 4 Select
l a number
b off repeating
i variables,
variables
i bl ,
where the number required is equal to the number
of reference dimensions.
Select a set of r dimensional variables that includes all
the primary dimensions (repeating variables).
These repeating variables will all be combined with
each of the remaining parameters.
Example: For pressure drop per unit length ( r = 3)
select ρ , V, D. 3 4 2
p   FL D  L   FL T
  FL 2 T V  LT 1
19
Determination of Pi Terms 7/12
Step 5 Form a pii term by
b multiplying
l i l i one off the
h
nonrepeating variables by the product of the
repeating variables,
variables, each raised to an exponent that
will make the combination dimensionless. 1
Set up dimensional equations, combining the variables
selected in Step 4 with each of the other variables
(nonrepeating variables) in turn, to form dimensionless
groups or dimensionless product.
There will be k – r equations.
Example: For pressure drop per unit length
20
Determination of Pi Terms 8/12

Step 5 (Continued) 2
a b c
1  p  D V 
( FL3 )( L) a ( LT 1 ) b ( FL4 T 2 ) c  F 0 L0T 0
F :1  c  0
L : 3  a  b  4c  0
T :  b  2c  0 p  D
1  2
 a  1, b  2, c  1 V
21
Determination of Pi Terms 9/12

Step 6 Repeat Step 5 for each of the remaining

non--repeating
non p g variables.
a b c
 2  D V 
( FL2 T )( L) a ( LT 1 ) b ( FL4 T 2 ) c  F 0 L0T 0
F :1  c  0
L : 2  a  b  4c  0 
T : 1  b  2c  0 2 
 a  1, b  1, c  1
DV
22
Determination of Pi Terms 10/12

Step 7 Check all the resulting pi terms to make

sure they
y are dimensionless.
Check to see that each group obtained is dimensionless.
Example: For pressure drop per unit length .

p  D
1  2
 F 0 L0T 0  M 0 L0T 0
V
 0 0 0 0 0 0
2  F L T M L T
 
DV
DV
23
Determination of Pi Terms 11/12

Step 8 Express the final form as a relationship

amongg the pi
p terms,, and think about what is means.
Express the result of the dimensional analysis.
 1  (  2 ,  3 , , , ,  k  r )
Example:
p For pressure
p drop
p pper unit length
g .
p  D    Dimensional analysis will not provide
   the form of the function.
function The function
V 2  DV  can only be obtained from a suitable set
of experiments
experiments.
p .
24
Determination of Pi Terms 12/12
 The pi terms can be rearranged. For example, Π2, could be
expressed as
VD
2 


p  D  VD 
  
  
2
V

25
Example 7.1 Method of Repeating
Variables
 A thin
hi rectangular
l plate
l having
h i a widthid h w andd a height
h i h h is
i
located so that it is normal to a moving stream of fluid.
Assume that the drag
drag,, D,
D that the fluid exerts on the plate
is a function of w and h, the fluid viscosity, µ ,and ρ,
respectively and the velocity,
respectively, velocity V,
V of the fluid approaching
the plate. Determine a suitable set of pi terms to study this
problem experimentally.
p p y

26
Example 7 1 Solution1/5
7.1
 Drag force on a PLATE
D  f ( w , h , ,  , V )
Step 1:List all the dimensional variables involved.
D h ρ,μ,V
D,w,h, V k=6k 6 dimensional
di i l parameters.
t
Step 2:Select primary dimensions M,L, and T. Express
each of the variables in terms of basic dimensions

D  MLT 2 w  L h  L
  ML1T 1   ML3 V  LT 1
27
Example 7 1 Solution2/5
7.1
Step 3: Determine the required number of pi terms.
k-r=6
r=6--3=3
w,V,.
Step 4:Select repeating variables w,V,
Step 5~6:combining the repeating variables with each
of the other variables in turn, to form dimensionless
groups or dimensionless products.
products

D  MLT 2 w  L h  L
  ML1T 1   ML3 V  LT 1
28
Example 7 1 Solution3/5
7.1
 1  Dw a V b  c  ( MLT  2 )( L ) a ( LT  1 ) b ( ML  3 ) c  M 0 L0 T 0
M :1  c  0
L : 1  a  b  3c  0 D
T : 2  b  0 1  2 2
 a   2 , b   2 , c   1
w V 

 2  hw a V b  c  L ( L ) a ( LT  1 ) b ( ML  3 ) c  M 0 L0 T 0
M :c  0
L : 1  a  b  3c  0 h
T:b0
2 
 a   1, b  0, c  0
w
29
Example 7 1 Solution4/5
7.1

 3  w a V b c  ( ML1T 1 )( L) a ( LT 1 ) b ( ML3 ) c  M 0 L0T 0

M :1  c  0
L : 1  a  b  3c  0 
T : 1  b  0 3 
 a  1, b  1, c  1
wV

30
Example 7 1 Solution5/5
7.1
Step 7: Check all the resulting pi terms to make sure
they are dimensionless.
Step 8: Express the final form as a relationship among
the ppi terms.
The functional relationship is

 1   (  2 ,  3 ), or
D  h  
  
 , 

 w wV  
2 2
w V 
31
Selection of Variables 1/4
 One
O off theh most iimportant, andd diffi
difficult,
l steps in
i applying
l i
dimensional analysis to any given problem is the selection
of the variables that are involved.
involved
 There is no simple procedure whereby the variable can be
easily identified.
identified Generally,
Generally one must rely on a good
understanding of the phenomenon involved and the
governing physical laws
laws..
 If extraneous variables are included, then too many pi
terms appear in the final solution,
solution and it may be difficult,
difficult
time consuming, and expensive to eliminate these
experimentally.
p y
32
Selection of Variables 2/4
 If important variables are omitted, then an incorrect result
will be obtained; and again, this may prove to be costly
and difficult to ascertain.
 Most engineering
g g pproblems involve certain simplifying
p y g
assumptions that have an influence on the variables to be
considered.
 Usually we wish to keep the problems as simple as
possible, perhaps even if some accuracy is sacrificed.
sacrificed.

33
Selection of Variables 3/4
 A suitable balance between simplicity and accuracy is an
goal.~~~~~ Enrico Fermi
desirable goal.~~~~~
 Variables can be classified into three general group:
Geometry
Geometry:: lengths and angles.
angles
Material Properties:
Properties: relate the external effects and the
responses.
responses
External Effects (Input & Output):
Output): produce, or tend to
produce,
d a change
h in
i theh system. Such
S h as fforce,
pressure, gravity,(<
gravity,(<-- Input) or velocity (<
(<-- Output)
Output)..
34
Selection of Variables 4-1/4
 Points should be considered in the selection of variables:
Clearlyy define the problem.
p What’s the main variable
of interest?
Consider the basic laws that govern the phenomenon.
Start the variable selection process by grouping the
variables into three broad classes
classes.

Continued

35
Selection of Variables 4-2/4
 Points should be considered in the selection of variables:
Consider other variables that mayy not fall into one the
three categories. For example, time and time dependent
variables.
Be sure to include all quantities that may be held
constant (e.g., g).
Make sure that all variables are independent
independent.. Look for
relationships among subsets of the variables.
variables

36
Determination of Reference Dimension 1/3

 When to determine the number of pi terms, it is important

to know how many reference dimensions are required to
describe the variables.
 In fluid mechanics, the required
q number of reference
dimensions is three, but in some problems only one or two
are required.
required.
 In some problems, we occasionally find the number of
reference dimensions needed to describe all variables is
smaller than the number of basic dimensions.
dimensions. Illustrated in
Example
a p e 7.2.
7. .
37
Example 7.2
7 2 Determination of Pi Terms

 An open, cylindrical tank having a diameter D is

supported around its bottom circumference and is filled to
a depth h with a liquid having a specific weight . The
deflection,  , of the center of the bottom is a
vertical deflection,
function of D, h, d, , and E, where d is the thickness of
the bottom and E is the modulus of elasticity of the bottom
material. Perform a dimensional analysis of this problem.

38
Example 7 2 Solution1/3
7.2
 The vertical deflection

  f D, f , d, , E 
L
DL
For F
F F,L,T.
L T Pi terms=6-
terms=6
t 6-2=4
2 4 hL
For M,L,T Pi terms=6
terms=6--3=3 dL
  FL3 ,   ML 2 T  2
E  FL 2 , E  ML1T  2
39
Example 7 2 Solution2/3
7.2
 For F,L,T system, Pi terms=6-
terms=6-2=4
D and γ are selected as repeating variables
a1 b1 a 2 b2
1  D   2  hD 
 3  dD a 3  b3  4  ED a 4  b4

 h d E  h d E 
1  ,  2  ,  3  ,  4      , , 
D D D D D  D D D 
40
Example 7 2 Solution3/3
7.2
 For M,L,T system, Pi terms=6-
terms=6-3=3 ?
A closer look at the dimensions of the variables listed
reveal that “only two reference dimensions,
dimensions,”” L and
MT-2 are required.
required.

41
Determination of Reference Dimension 2/3

EXAMPLE h  f D,  ,  

MLT SYSTEM FLT SYSTEM

h D   h D  
M M F F
L L L L
2 2
LT T 2 L3 L

Pi term=4
term=4-3=1
3=1 Pi term=4-2=2
term 4 2 2

42
Determination of Reference Dimension 3/3
(option)
S Di
Set Dimensional
i l Matrix
M i
MLT SYSTEM FLT SYSTEM
h D   h D  
M 0 0 1 1 F 0 0 1 1
L 1 1 2 0 L 1 1  3 1
T 0 0 2 2 T 0 0 0 0
h
1 
Rank=2 Pi term=4
term=4--2=2 D

h   
2  2    2 
D  D D 
  43
Uniqueness of Pi Terms 1/4
 The Pi terms obtained depend on the somewhat arbitrary
selection of repeating variables.
variables. For example, in the
problem of studying the pressure drop in a pipe.
p   f ( D, , , V )
Selecting D,V, and  as p  D  VD 
repeating
i variables:
i bl  1  
V 2   
Selecting D,V, and  as p  D 2  VD 
repeating variables:   2  
V   
44
Uniqueness of Pi Terms 2/4

p  D  VD 
 1   Both are correct,, and both would
  
2
V
V
lead to the same final equation for
the ppressure drop.
p There is not a
p  D 2  VD  unique set of pi terms which
  2  
V    arises from a dimensional analysis
analysis.
y .
The functions Φ1 and Φ2 are will
be different because the dependent
p
pi terms are different for the two
relationships.
relationships.`
p `
45
Uniqueness of Pi Terms 3/4
EXAMPLE 1    2 ,  3 

' a b
Form a new pi term 2  23
'

1  1  2 ,  3   '
 2 2 , 2 
All are correct

46
Uniqueness of Pi Terms 4/4

p  D  VD  Selecting D,V, and  as

 1  
repeating variables:
V2
V   
p  D  VD 
2   1 
 

p  D VD p  D V
V 2
  
 
V 2  V p  D 2  VD 
  2  
V   
p  D 2  VD 
  2  
V   
47
Common Dimensionless Groups 1/2

 A list of variables that

commonlyy arise in fluid
mechanical problems.
 Possible to provide a
physical interpretation to
the dimensionless groups
which
hi h can be
b helpful
h l f l in
i
assessing their influence
in a particular application.
application

48
Reynolds Number 1/2
V V
R 
Re 
 
 In honor of Osborne ReReynolds
nolds (1842
(1842~1912),
1912) the British engineer
who first demonstrated that this combination of variables could be
used as a criterion to distinguish
g between laminar and turbulent flow.
 The Reynolds number is a measure of the ration of the inertia forces
to viscous forces.
forces.
 If the Reynolds number is small (Re<<1), this is an indication that
are dominant in the problem, and it may be
the viscous forces are
possible to neglect the inertial effects; that is, the density of the fluid
will no be an important variable.

49
Reynolds Number 2/2
 Flows
Fl with
i h very smallll Reynolds
R ld numbers
b are commonly l referred
f d to
as “creeping flows”.
 For large Reynolds number flow, flow, the viscous effects are small
relative to inertial effects and for these cases it may be possible to
neglect the effect of viscosity and consider the problem as one
i l i a “nonviscous”
involving “ i ” fluid.
fl id
 Flows with “large” Reynolds number generally are turbulent
turbulent.. Flows
in which the inertia forces are “small”
small compared with the viscous
forces are characteristically laminar flows.
flows.

50
Correlation of Experimental Data
 Dimensional analysis only provides the dimensionless groups
describing the phenomenon, and not the specific relationship
between the groups.
groups.
 To determine this relationship, suitable experimental data must be
obtained.
obtained
 The degree of difficulty depends on the number of pi terms.

51
Problems with One Pi Term
 The functional relationship for one Pi term.
1  C
where C is a constant. The value of the constant must still be
d
determined
i d by
b experiment.
i

52
Example 7.3
7 3 Flow with Only One Pi Term

 Assume that the drag, D, acting on a spherical particle that falls very
slowly through a viscous fluid is a function of the particle diameter,
d the particle velocity,
d, velocity V,
V and the fluid viscosity,
viscosity μμ. Determine,
Determine
with the aid the dimensional analysis, how the drag depends on the
pparticle velocity.
y

53
Example 7
7.3
3 Solution
 The drag
D  F d  L   FL2T
D  f (d, V,  )   ML3 V  LT 1

D
1  C D  CVd
Vd
For a given particle and fluids, the drag varies
DV directly with the velocity

54
Problems with Two or More Pi Term 1/2

 Problems with two pi terms

1  (  2 ) An empirical relationship is
the functional relationship valid over the range of Π2.
among the variables can the
be determined by varying Π2
and measuring the
corresponding value of Π1.
The empirical
p equation
q Dangerous to
relating Π2 and Π1 by using extrapolate beyond
a standard curve
curve--fitting valid range

technique.
55
Problems with Two or More Pi Term 2/2

 Problems with three pi terms.

1   2 , 3  To determine a suitable empirical
p equation
q
relating the three pi terms.

To show data correlations on simple graphs.

Families curve of curves

56
Modeling and Similitude
To develop the procedures for designing
models so that the model and prototype
will behave in a similar fashion…….

57
 vs Prototype 1/2
Model vs.
 Model
M d l ? A modeld l is
i a representation
t ti off a physical
h i l systemt that
th t may
be used to predict the behavior of the system in some desired respect.
Mathematical or computer models may also conform to this
d fi iti our interest
definition, i t t will
ill be
b in
i physical
h i l model.
d l
 Prototype? The physical system for which the prediction are to be
made.
 Models that resemble the prototype but are generally of a different
size,, may involve different fluid, and often operate under different
size
conditions.
conditions
 Usually a model is smaller than the prototype.
prototype.
 Occasionally,y if the pprototype
yp is veryy small, it mayy be advantageous
g
to have a model that is larger than the prototype so that it can be
more easily studied. For example, large models have been used to
studyy the motion of red blood cells.
58
 vs Prototype 2/2
Model vs.
 With the successful development of a valid model,
model, it is possible to
predict the behavior of the prototype under a certain set of
conditions.
conditions
 There is an inherent danger in the use of models in that predictions
can be made that are in error and the error not detected until the
prototype is found not to perform as predicted.
 It is imperative that the model be properly designed and tested and
that the results be interpreted correctly.
correctly.

59
Similarity of Model and Prototype
 What
Wh conditions
di i must be
b met to ensure the
h similarity
i il i off model
d l andd
prototype??
prototype
 Geometric Similarity
Model and prototype have same shape
shape..
Linear dimensions on model and prototype correspond within
constant scale factor.
factor.
 Kinematic Similarity
Velocities at corresponding points on model and prototype differ
only by a constant scale factor.
factor.
 Dynamic
D i SiSimilarity
il it
Forces on model and prototype differ only by a constant scale
factor..
factor
60
Theory of Models 1/5
 The
Th theory
h off models
d l can be
b readily
dil developed
d l d by b using
i
the principles of dimensional analysis.
 For
F given
i problem
bl which
hi h can be
b described
d ib d in
i terms off a
set of pi terms as
Thi relationship
This l i hi can be b formulated
f l d
1  (  2 ,  3 , , ,  n ) with a knowledge of the general
nature of the physical phenomenon
and the variables involved.

This equation applies to any system that is

governed by
g y the same variables.
61
Theory of Models 2/5
 A similar relationship can be written for a model of this
prototype; that is,
 1m  (  2 m ,  3 m , , , ,  nm )
where
h th fform off th
the the ffunction
ti will
ill be
b the
th same as long
l
as the same phenomenon is involved in both the
prototype and the model.

The prototype and the model must have

the same phenomenon.
phenomenon
62
Theory of Models 3/5
 Model design (the model is designed and operated)
conditions, also called similarity requirements or modeling
laws..
laws
 2   2m  3   3 m .....  n   nm
 The form of Φ is the same for model and prototype, it
f ll
follows that
h This is the desired prediction equation and
indicates that the measured of Π1m obtained
 1   1m with the model will be equal to the
corresponding Π1 for the prototype as long as
the other Π parameters are equal.
equal
63
Theory of Models 4/5 – Summary1
 The prototype and the model must have the same
phenomenon..
phenomenon

For prototype 1  (  2 ,  3 , , ,  n )

For model  1m  (  2 m ,  3 m , , , ,  nm )

64
Theory of Models 5/5 – Summary2
 The model is designed and operated under the following
conditions (called design conditions,
conditions, also called similarity
requirements or modeling laws)
 2   2m  3   3 m .....  n   nm
 The measured of Π1m obtained with the model will be
equall to
t the
th corresponding
di Π1 for
f ththe prototype.
t t

 1   1m Called pprediction equation

q

65
Theory of Models EXAMPLE 1
 Example:
E l Considering
C id i the h drag
d force
f on a sphere.
sphere
h .
F   VD 
F  f ( D , V , ,  )  f 1 
 

V 2 D 2   
The prototype and the model must have the same phenomenon.
Fm   m Vm D m  F   VD 
 f1    f1  
   prototype
2 2 m 2 2
 m Vm D m   V D

Design conditions.   VD 

  VD
 


 
  model    prototype

Then …  F 


 
F 

 V 2 D 2  2 2
  model  V D  prototype 66
Theory of Models EXAMPLE 2
 Example:
E l Determining
D i i the
h drag
d force
f on a thin
hi rectangular
l plate
l (w(w
× h in size)
D  w  Vw 
D  f w , h ,  ,  , V  w 2V 2
   ,
 h 


The prototype and the model must have the same phenomenon.
Dm  w m  m Vm w m  D  w  Vw 
   ,     , 
w m 2  m Vm 2  mh m  w 2V 2  h   prototype

wm w  V w  Vw
Design conditions. hm
 , m m m 
h m 

2 2
Then … D Dm  w     V 
  D        D m
 m
2 2
w V w m 2  m Vm 2  wm    Vm 
67
Example 7.5 Prediction of Prototype
Performance from Model Data 1/2
 A long structural component of a bridge has the cross section
shown in Figure E7.5. It is known that when a steady wind blows
past this type of bluff body,
body vortices may develop on the downwind
side that are shed in a regular fashion at some definite frequency.
Since these vortices can create harmful pperiodic forces actingg on the
structure, it is important to determine the shedding frequency. For
the specific structure of interest, D=0.2m, H=0.4m, and a
representative wind velocity 50km/hr. Standard air can be assumed.
The shedding frequency is to be determined through the use of a
small--scale model that is to be tested in a water tunnel.
small tunnel For the
model Dm=20mm and the water temperature is 20℃ 20℃.

68
Example 7.5 Prediction of Prototype
Performance from Model Data 2/2
Determine the model dimension, Hm,Hm, and the velocity at which the
test should be performed. If the shedding frequency ω for the model
is found to be 49.9Hz,
49 9Hz what is the corresponding frequency for the
prototype??
prototype
5 3
 For air at standard condition   1 .79  10 kg / m  s,   1 .23 kg / m
20℃,  water  1  10 3 kg / m  s,  water  998 kg / m 3
 For water at 20℃

69
Example 7 5 Solution1/4
7.5
 Step 1:List all the dimensional variables involved. ω
D,H,V,ρ,μ
D,H,V, ρ,μ k=6 dimensional variables.
 Step 2:Select primary dimensions F,L and T. List the
dimensions of all variables in terms of pprimaryy dimensions.
r=3 primary dimensions
1
T
 D  L H  L
V  LT 1   FL 4 T 2   ML  2 T

70
Example 7 5 Solution2/4
7.5
 Step 3: Determine the required number of pi terms.
k-r=6
r=6--3=3
 Step 4:Select repeating variables D,V, μ.
 Step 5~6:combining the repeating variables with each of
the other variables in turn, to form dimensionless groups.
a1 b1 c 1 D a2 b2 c2 H
 1  D V    2  HD V  
V D
 VD
 3   D a 3 V b3  c3 

71
Example 7 5 Solution3/4
7.5
D  H  VD 
 The functional relationship is   , 
V D  

Strouhal number

 The prototype and the model must have the same

phenomenon.
h
m D m  H m  m Vm D m 
  , 
Vm  Dm m 

72
Example 7 5 Solution4/4
7.5
 The Design conditions.
H Hm  VD  m V m D m
 
D Dm  m
Then
Then….
H m  D
Hm  D m  ...  60 mm Vm  V  ...  13 .9 m / s
D  m D m

V Dm
  m  ...  29 .0 Hz
Vm D

73
Model Scales
 The ratio of a model variable to the corresponding
prototype variable is called the scale for that variable.
1   
Length Scale  1m     1m  2 m
 2  2m 1 2

Vm
Velocity Scale V 
V
m
Density Scale  


m
Viscosity Scale  

74
Distorted Models
 In
I many model d l studies,
di to achieve
hi dynamic
d i similarity
i il i
requires duplication of several dimensionless groups.
groups.
 In
I some cases, complete l dynamic
d i similarity
i il i between
b
model and prototype may not be attainable. If one or more
of the similarity requirements are not met,
met for example,
example
if  2   2 m , then it follows that the prediction equation
is  1   1m
 1   1m is not true; that is,
MODELS for which one or more of the similarity
requirements
i are not satisfied
i fi d are called
ll d
DISTORTED MODELS.
MODELS.
75
Distorted Models EXAMPLE
EXAMPLE--1 1/3

 Determine
D i the
h drag
d force
f on a surface
f ship,
hi completel
dynamic similarity requires that both Reynolds and Froude
numbers be duplicated between model and prototype.
prototype
Vm Vp
Fm 
Fr 1/ 2
 Fr
Fp F d numbers
Froude b
( g m ) ( g  p )1 / 2

Vm  m Vp  p
Re m   Re p  Reynolds numbers
m p
1/ 2
 To
T matchh Froude
F d numbersb Vm   m 


between model and prototype Vp   p 

76
Distorted Models EXAMPLE
EXAMPLE--1 2/3

 To match Reynolds numbers between model and prototype

1/ 2 3/ 2
m V 
 m m
Vm   m   m   m  Vm   m 
1/ 2
   
p Vp  p V p   p   p   p  Vp   p 


If m/ p equals
l 1/100(a
1/100( typical
t i l length
l th scalel for
f ship
hi
model tests) , then υm/υp must be 1/1000.
>>> The kinematic viscosity ratio required to
duplicate Reynolds numbers cannot be attained
attained..
77
Distorted Models EXAMPLE
EXAMPLE--1 3/3

 It is impossible in practice for this model/prototype scale

of 1/100 to satisfy both the Froude number and Reynolds
number criteria;
criteria; at best we will be able to satisfy only
one of them.
 If water is the only practical liquid for most model test of
free--surface flows, a full
free full--scale test is required to obtain
complete dynamic similarity.
similarity.

78
Typical Model Studies
 Flow through closed conduits.
 Flow around immersed bodies.
 Flow with a free surface.

79
Flow Through Closed Conduits 1/5
 This type of flow includes flow through pipes, valves,
fittings, and metering devices.
devices.
 The conduits are often circular,
circular, they could have other
shapes
p as well and may y contain expansions
p or contractions.
 Since there are no fluid interfaces or free surface, the
dominant forces are inertial and viscous forces so that the
Reynolds number is an important similarity parameter.
parameter.

80
Flow Through Closed Conduits 2/5
 For low Mach numbers (Ma<0.3), compressibility effects
are usually negligible for both the flow of liquids or gases.
 For flow in closed conduits at low Mach numbers, and
dependent
p pi
p term, such as pressure
p drop,
p can be expressed
p
as
  i  V 
term=   ,  ,  
Dependent pi term
 
Where  is some particular length of the system and i
represents a series of length terms, ε/  is the relative
roughness of the surface, and ρV/μ is the Reynolds number.

81
Flow Through Closed Conduits 3/5
 To meet the requirement of geometric similarity
 im  i m   im  m  m
      
m  m  i  

 To meet the requirement of Reynolds number

 m Vm  m V Vm  m   m 
    1
m  V  m  m  m

Vm 
If the same fluid is used, then   Vm  V /  
V m
82
Flow Through Closed Conduits 4/5
 The fluid velocity in the model will be larger than that in
the prototype for any length scale less than 1. Since length
scales are typically much less than unity.
 Reynolds
y number similarityy mayy be difficult to achieve
because of the large model velocities required.
required.
Vm 
  Vm  V /   Vm  V
V m

83
Flow Through Closed Conduits 5/5
 With these similarity requirements satisfied, it follows that
the dependent pi term will be equal in model and
prototype. For example,
p
D
Dependent
d t pii term
t 1 
V 2
 The prototype pressure drop
2
  V 
p    p m In general p  p m
m  Vm 
84
Flow Around Immersed Bodies 1/7
 This type of flow includes flow around aircraft,
automobiles, golf balls, and building.
building.
 For these problems, geometric and Reynolds number
similarityy is required.
q
 Since there are no fluid interfaces, surface tension is not
important. Also, gravity will not affect the flow pattern, so
the Froude number need not to be considered.
 For incompressible flow
flow, the Mach number can be omitted
omitted.

85
Flow Around Immersed Bodies 2/7
 A general formulation for these problems is
  i  V 
d pii term=  , ,
Dependent 
    
Where  is some characteristic length of the system and i
represents other pertinent lengths, ε/  is the relative roughness
of the surface, and ρV/μ is the Reynolds number.

Model of the National Bank of Commerce, San

Antonio, Texas, for measurement of peak, rms,
and mean pressure distributions. The model is
l
located
d in
i a long-test-section,
l i meteorological
l i l
wind tunnel. 86
Flow Around Immersed Bodies 3/7
 Frequently, the dependent variable of interest for this type
of problem is the drag, D, developed on the body.
The dependent pi terms would usually be expressed in the
form of a drag coefficient
D
CD 
1
V 2  2
2
 To meet the requirement of geometric similarity
 im  i m   im  m  m
      
m  m  i  
87
Flow Around Immersed Bodies 4/7
 To meet the requirement of Reynolds number similarity
 m Vm  m V Vm  m   m 
   
m  V  m  m  m
2 2
D Dm   V    
 D     D m
1 1 m  Vm   m 
V 2  2  m Vm 2  m 2
2 2
Vm 
The same fluid is used, then   Vm  V /  
V m

88
Flow Around Immersed Bodies 5/7
 The fluid velocity in the model will be larger than that in
the prototype for any length scale less than 1. Since length
scales are typically much less than unity.
 Reynolds
y number similarityy mayy be difficult to achieve
because of the large model velocities required.

89
Flow Around Immersed Bodies 6/7
 How to reduce the fluid velocity in  mVm  m V Vm  m 
  
the model ?  m  V  m

A different fluid is used in the model such that  m /   1

For example,
p , the ratio of the kinematic viscosityy of water to that
of air is approximately 1/10, so that if the prototype fluid were
air, test might be run on the model using water.
water.

Vm  V This would reduce the required model velocity, but

itt still
st mayay be difficult
d cu t to achieve
ac eve thet e necessary
ecessa y
velocity in a suitable test facility, such as a water
tunnel.
90
Flow Around Immersed Bodies 7/7

How to reduce the fluid velocity in the model ?

Same fluid with different density.. m>

An alternative way to reduce Vm is to

increase the air pressure in the tunnel so that
m>. The pressurized tunnels are obviously
complicated and expensive.

91
Similitude Based on Governing Differential
Equations 1/5

 For a steady incompressible two

two--dimensional flow of a
Newtonian fluid with constant viscosity.
The mass conservation equation is
u v
 0 Has dimensions of 1/time
1/time.
x y
The Navier
Navier--Stokes equations are
 u u  p  2u 2u 
  u v      2  
 x y  x  x  y 2  Has dimensions of
 
force/volume
 v v  p  2v 2v 
  u v     g    2  
 x y  y  x  y 2 

92
Similitude Based on Governing Differential
Equations 2/5
 Dimensionlize
Dimensionli e withith characteristic (standard) quantities
q antities
such that the quantities of dimensionless parameters are O
(1) x  x l  l u  Vu p  pp t  tt
(1). * * * * *

c c c c c

 How to non
non--dimensionalize these equations ?
x y u v p t
x*  y*  u*  v*  p*  t* 
c c Vc Vc pc tc

u Vc u *  2u   u  Vc  2 u *
    2
x  c x * x 2
x  x   c x *2
The mass conservation equation u  v 
 0
x  y
93
 Similitude Based on Governing Differential
Equations 3/5

The Navier-
Navier-Stokes equations R
Reynolds
ld number
b

  c  u  u 
u 
 p  p 
     2 
u  2  
u 
 
 
  
c
  
   2    2
u v
  
 
  
 2 
 c c
t V t x y  c 
V x  c c  x
V  y 
  c  v  v 
v 
 p  p 
 g        2 
v  2  
v 
  u

v 
  c
 2 
c 
  2  2 
 2
 Vc   Vc  c  x
  
 tcVc  t x y  Vc  y y 

Strouhal number Euler number Reciprocal of the sq

square
are
of the Froude number

94
Similitude Based on Governing Differential
Equations 4/5

 From these equations it follows that if two systems are

governed by these equations, then the solutions (in terms
of u*,v*,p*,x*,y*, and t*) will be the same if the four
parameters are equal for the two systems.
systems.
 The two systems will be dynamically similar. Of course,
boundary and initial conditions expressed in
dimensionless form must also be equal for the two systems,
systems,
and this will require complete geometric similarity.

95
Similitude Based on Governing Differential
Equations 5/5
 These
Th are the
h same similarity
i il i requirements
i that
h would ld be
b
determined by a dimensional analysis if the same variables
were considered.
considered These variables appear naturally in the
equations.
 All the common dimensionless groups that we previously
developed by using dimensional analysis appear in the
governing equations that describe fluid motion when these
equations are expressed in term of dimensionless variables.
 The use of governing equations to obtain similarity laws
provides an alternative to conventional dimensional
analysis.
analysis
y .
96