Module: Further Hydraulics - Dimensional Analysis

01/09/2021

 Dimensional Analysis
 Buckingham’s Pi Theorem
 Similitude

DIMENSIONAL ANALYSIS
Reference: Fluid Mechanics, J.F Douglas, J.M Gasiorek, J.A Swaffield, L. Black. Part III
Youtubevideo by Dr Benjamin Drew: https://www.youtube.com/watch?v=dofAaelv4yA
01-Sep-21
5
01-Sep-21

INTRODUCTION
•The application of fluid mechanics in design, perhaps more than most engineering
subjects, relies on the use of empirical results built up from an extensivebody of
experimental research.
•In many areas empirical data are supplied in the form of tablesand charts that the
designer may apply directly.
•To make experimental work less tedious, by replacing the relationship between any
two variables by generalised groupings.
•Dimensional analysis is a tool that allows such simplifications to be made.

HISTORICAL CONTEXT
•John Smeaton (1724-1792) first used scale models for systematic experimentation.
•William Froude (1810-1871) first proposed laws for estimating ship hull drag from model
tests.
•Aimee Vaschy, Lord Rayleigh, D. Riabouchinsky, E. Buckingham all made significant
contributions to dimensional analysis and similitude.
•Jean B. J. Fourier (1768-1830) first formulated a theory of dimensional analysis.
•Osborne Reynolds (1842-1912) first used dimensionless parameters to analyse
experimental results.
•Moritz Weber (1871-1951) assigned the name Reynolds numberand Froude number.
DIMENSIONS AND UNITS
•Dimension: A measure of a physical quantity (without numerical values).
•Unit: A way to assign a number to that dimension.
•There are seven primary dimensions :
1.Mass m (kg)
2.Length L (m)
3.Time t (sec)
4.Temperature T (K)
5.Current I (A)
6.Amount of Light C (cd)
7.Amount of matter N (mol)

•All non-primary dimensions (derived dimensions) can be formed by some combination of the seven
primary dimensions.
•e.g.
•{Velocity} = {Length/Time} = {L/t}
•{Force} = {Mass Length/Time} = {mL/t2}
•Density is denoted by mass per unit volume = m/L3

DIMENSIONAL REASONING, HOMOGENEITY AND DIMENSIONLESS

GROUPS
•Dimensional reasoning is based on the proposition that, for an equation to be true, then both sides of the equation must be
numerically and dimensionally identical.
•To take a simple example, the expression x + y = z when x = 1, y = 2 and z = 3 is clearly numerically true but only if the dimensions
of x, yand z are identical. Thus

1 elephant + 2 aeroplanes = 3 days

is clearly nonsense but1 metre + 2 metre = 3 metre
is wholly accurate.
•An equation is only dimensionally homogeneous if all the terms have the same dimensions. In general any equation of the form

𝒂𝟏𝒎𝟏𝒃𝟏𝒏𝟏𝒄𝟏𝒑𝟏+𝒂𝟏𝒎𝟐𝒃𝟏𝒏𝟐𝒄𝟏𝒑𝟐+…+=𝑿

will be physically true if, in addition to being numerically correct, the terms are dimensionally the same so that :

𝒂𝟏𝒎𝟏𝒃𝟏𝒏𝟏𝒄𝟏𝒑𝟏=𝒂𝟏𝒎𝟐𝒃𝟏𝒏𝟐𝒄𝟏𝒑𝟐=𝑿,

N.B: Square brackets [ ] should be read as “dimensionally”.

DIMENSIONAL REASONING, HOMOGENEITY AND DIMENSIONLESS
GROUPS
•The dimensions of the group are determined by the normal rules ofalgebra: for example, as
Newton’s second law states that Force = Mass ×Acceleration it follows that the dimensions of
force are given by [MLT-2].
•It follows from the identification of the dimensions of a group that the ratio oftwo parameters or
two dimensionally identical groups of variables will yield a dimensionless group. For example, the
ratio of a vehicle’s speed to that of sound yields V/c; dimensionally [V ]/[c] = [L0]/[T0] = [1].
•Similarly, strain is defined as Extension/ Original length or [L]/[L] = [1], a non-dimensional ratio.
•Dimensional homogeneity may be used to check the accuracy of any equation,remembered
or derived. This is an important tool and should be used as a matterof course.

In dynamics, Newton’s second law provides the basis for the definition of derived dimensions,
e.g.
Force = Mass ×Acceleration
= Mass ×VelocityTime
= [M][LT−1]/[T]
so that
[Force] = [MLT−2]
Because Newton’s second law links force to mass and acceleration and hence to distance and
time, any variable such as work, power, viscosity, etc. can be expressed in terms that eventually
reach an expression of Newton’s second law.
For example:

•Pressure or stress = Force/Area = Mass ×Acceleration/Area,so that [Pressure or stress] = [M] [LT -
2]/[L-2] = [ML-1T-1]Density = Mass/Volumeso that [Density] = [M]/[L3] = [ML-3]

•Power = Rate of doing work= Force ×Distance/Time= Mass ×Acceleration ×Distance/Time so

that [Power] = [M] [LT-2] [L]/[T] = [ML2T-3]

•Similarly, viscosity, µ, may be defined in terms of shear stress, τ, and velocity gradient as τ = µ
𝜕𝑢𝜕𝑦, so that :[Viscosity] = [Stress]/[Velocity gradient] = [ML -1T-2]/[LT-1/L] = [ML-1T-1]
DIMENSIONAL HOMOGENEITY
Verify the dimensional homogeneity of the Bernoulli Equation

TABLE: FLOW CHARACTERISTICS, UNITS AND DIMENSIONS

Characteristics Unit (SI) Dimension (MLT) Dimension (FLT)

Length m L
Area m2 L2
Volume m3 L3
Time s T
Velocity m/s L/T
Acceleration m/s2 L/T2
Discharge m3/s L3/T
Mass Kg M FL-1T2

Forces N (Kg.m/s2) MLT-2 F

Pressure Pa (N/m2) ML-1T-2 FL-2
Energy J (Nm) ML2T-2 FL
QUANTITIES, DEFINING EQUATION & DIMENSIONAL MLT SYSTEM

NON-DIMENSIONALISATIONOF EQUATIONS
•Non-dimensional equation:If we divide each term in an equation by a collection of
variables and constants whose product has those same dimensions, the equation is
rendered non-dimensional.
A non-dimensionalisedform of the Bernoulli equation is formed by dividing each additive term by a
pressure (here we use P ). Each resulting term is dimensionless (dimensions of {1}).

•Most of these non-dimensional ratios are named after a notable scientist or engineer
(e.g. the Reynolds number and the Froude number, Mach number etc.)

METHODS OF DIMENSIONAL ANALYSIS

•In a general unsteady fluid flow problem with a free surface, the scaling
parameters include a characteristic length L, a characteristic velocity V, a
characteristic frequency f, and a reference pressure difference P0 -P .
•Non-dimensionalisationof the differential equations of fluid flow produces four
dimensionless parameters: the Reynolds number Re, Froude number Fr,
Strouhalnumber St, and Euler number Eu
COMMON NON-DIMENSIONAL PARAMETERS
In Fluid Mechanics,
•the Reynolds number(Re)is a dimensionless number that gives a measure of the ratio of inertial
forces to viscous forces.
•The Froude number(Fr)is a dimensionless number defined as the ratio of a body's inertia to
gravitational forces . In fluid mechanics, the Froude number is used to determine the resistance
of a partially submerged object tmoving through water, and permits the comparison of objects
of different sizes.
•The Strouhal number(St) is a dimensionless number describing oscillating flow mechanisms.
•The Euler number (Eu) is a dimensionless number used in fluid flow calculations. It expresses the
relationship between alocal pressure drop over a restriction and the kinetic energy per
volume,and is used to characterize losses in the flow,where a perfect frictionless flow
corresponds to an Euler number of 1.

COMMON DIMENSIONLESS GROUPS IN FLUID MECHANICS

N.B These groups will be used a lot in the chapter on SIMILITUDE.

See Slide: Dimensionless Numbers
METHODS OF DIMENSIONAL ANALYSIS
•If the number of variables involved in a physical phenomenon are known, then the relation
among the variables can be determined by thefollowing two methods;• Rayleigh’s Method•
Buckingham’s π-Theorem

•Rayleigh’s Method:It is used for determining expression for a variable (dependent)

whichdepends upon maximum three to four variables (Independent) only. If the number of
independent variables are more than 4, it then becomes very tedious to obtain expression for
dependent variable.

•Buckingham’s π-Theorem

•Since Rayleigh’s Method becomes laborious if variables are more than fundamental dimensions
(MLT), the difficulty is overcome by using Buckingham’s π-Theorem. This course will deal only with
this latter method.

BUCKINGHAM’S Π-THEOREM
•Such a technique known as Buckingham π method was developed in the early years
of the twentieth century.
•The initial step in the application of the Buckingham π method is to list thevariables
considered to be significant and to form a matrix with their dimensions.
•The Buckingham π method states that:

The number of dimensionless groups arising from a particular matrix formed from n
variables in m dimensions is n -r, where r is the largest non-zero determinant that can be
formed from the matrix, and therefore the equation relating the variables will be of the
form

φ(π1, π 2, π 3,..., πn-r) = 0

BUCKINGHAM’S Π THEOREM
•While this form of Buckingham’s theorem is correct, it is often simpler, in thetreatment of
fluid conditions that only involve the dimensions of mass, length and time, to state that
the number of dimensionless groups formed from n variables in three dimensions is n -m,
or n –3.
•While this is also correct and widely quoted, care must be taken in applying this rule
outside the confines of a strictly three-dimensional problem (e.g. in thermodynamics,
heat flow can be considered as another dimension).
BUCKINGHAM’S ΠTHEOREM (EXPLAINED)
•How many dimensionless products are required to replace the original list of variables? “If an
equation involving k variables is dimensionally homogeneous, it can be reduced to a
relationship among k –r independent dimensionless products, where m is the minimum number
of reference dimensions required to describe the variables.”

•The dimensionless products are frequently referred to as “pi terms”.

u1= f(u2,u3, … uk)

Π1= Φ(Π1Π2…Πk-r)
•The required number of pi terms is fewer than the number of originalvariables by r, where r is
determined by the minimum number ofreference dimensions required to describe the original list
of variables.(MLT, FLT )

BUCKINGHAM’S ΠTHEOREM (DETERMINATION OF PI TERMS)

•Step 1:
•List all the variables that are involved in the problem and count their total number.e.g.
Geometry of the system (such as pipe diameter)Fluid properties (ρ, μ)External effects (driving
pressure, V)It is important that all variables be independent.
•Step 2:
•Express each of the variables in terms of basic dimensions (MLT, FLT). 𝐹=𝑚𝑎=𝑀𝐿𝑇−2

𝑀=𝐹𝐿𝑇−2
∴𝜌=𝑀𝐿−3=𝐹𝐿−4𝑇2

BUCKINGHAM’S ΠTHEOREM (DETERMINATION OF PI TERMS)

•Step 3:
•Determine the required member of pi terms.From Buckingham pi theorem:k variablesr reference
dimensions (M, L, T, or θ)∴k –r independent dimensionless Πgroups
•Step 4:
•Select a number of repeating variables, where the number required is equal to the number of
reference dimensions.

N.B:
1. Each repeating variable must be dimensionally independent of the others.

2. Do not choose the dependent variable (e.g., Δp) as one of the repeating variables.

BUCKINGHAM’S ΠTHEOREM (DETERMINATION OF PI TERMS)

•Step 5:
•Form a “pi form” by multiplying one of the non-repeating variables by the product of the repeating
variables, each raised to an exponent that will make the combination dimensionless.

𝑢𝑖, 𝑢1𝑎,𝑢2𝑏,𝑢3𝑐, where:

𝑢𝑖: non-repeating variable
𝑢1𝑎,𝑢2𝑏,𝑢3𝑐: Repeating variables
•Step 6:
•Repeat Step 5 for each of the remaining nonrepeating variables.
•Step 7:
•Check all the resulting pi terms to make sure they are dimensionless.

BUCKINGHAM’S ΠTHEOREM (DETERMINATION OF PI TERMS)

•Step 8:
•Express the final form as a relationship among the pi terms, and think about what it means :

Π1= Φ(Π1Π2…Πk-r)
•The actual functional relationship among the pi terms must be determined by experiments.

SUMMARY OF THE PROCEDURE FOR B’S PI THEOREM

PROPERTIES OF Π-TERMS
• Each π-term is dimensionless and is independent of system of units.
• Division or multiplication by a constant does not change the character of the π-
terms.
• Each π-term contains m+1 variables, where m is the number of fundamental
dimensions and also called repeating variable.
π -TERMS: METHODS OF SELECTION
•The number of repeating variables are equal to number of fundamental dimensions of the
problem. The choice of repeating variables is governed by following considerations;

• As far as possible, dependent variable should not be selected as repeating variable

• The repeating variables should be chosen in such a way that one variable contains geometric
property, other contains flow property and third contains fluid property
• The repeating variables selected should form a dimensionless group
• The repeating variables together must contain all three fundamental dimension i.e., MLT
• No two repeating variables should have the same dimensions.
Note: In most of fluid mechanics problems, the choice of repeating variables may be (i) d, v, ρ,
(ii) l, v, ρ or (iii) d, v, μ

APPLICATION OF BUCKINGHAM’S PI THEOREM --EXAMPLES

•P 1:

The thrust, F, of a propeller depends upon its diameter, d, speed of advance, v, revolutions per
second, N, the fluid density, ρ, and viscosity, µ.
Find an expression for F in terms of these quantities.
•Solution

The general relationship between dependant and independent variable may be expressed as: F
= φ(d, v, N, ρ, µ), which may be expressed as:
F = K(d m, vp, Nr, ρq, µs), whereK is a numerical constant.

EXAMPLE P1
•The dimensions of these variables may be conveniently expressed in the table below:

Hint: See table “Quantities, Defining Equation & Dimensional MLT System”

Alternatively, the dimensions of the variables involved may be tabulated as shown below:

No. Variable Symbol Dimension

1 Thrust
2 Diameter
3 Velocity
4 Viscosity
5 Density
6 Speed
F MLT-2
D L
V LT-1
μ ML-1T-1
Ρ ML-3
N T-1
EXAMPLE P1
•From either of the above tables, we can see the following:
•Number of variables n = 6
•Number of fundamental dimension m = 3
•Number of dimensionless πterm (n –m) = 3
•f1(π1, π 2, π 3)= 0 ………….(1)

•Selecting the repeating variable as d, V, ρ

•By using Buckingham’s πtheorem,

𝜋1=𝑑𝑎1𝑉𝑏1𝜌𝑐1𝐹
𝜋2=𝑑𝑎2𝑉𝑏2𝜌𝑐2𝑉
𝜋3=𝑑𝑎3𝑉𝑏3𝜌𝑐3μ

EXAMPLE P1
•Solve the π equation by the principle of dimension homogeneity:
•For π1term:

𝜋1=𝑑𝑎1.𝑉𝑏1.𝜌𝑐1.𝐹𝑀0𝐿0𝑇0=𝐿𝑎1.𝐿𝑇−1𝑏1.𝑀𝐿−3𝑐1.𝑀𝐿𝑇−2𝑀∶0=𝑐1+1∴𝑐1=−1𝐿∶0=𝑎1−3𝑐1+1+𝑏1∴𝑎1=−2𝑇∶0=−𝑏1−2∴𝑏1=−2
•Substituting the values of a1,b1,c1 in π1 equation:

π1=𝑑1−2.𝑉−2.𝜌−1.𝐹=𝐹𝑑2𝑉2𝜌
For π2 term:
𝜋2=𝑑𝑎2.𝑉𝑏2.𝜌𝑐2.𝑁
𝑀0𝐿0𝑇0=𝐿𝑎2.𝐿𝑇−1𝑏2.𝑀𝐿−3𝑐2.𝑇−1
𝑀∶0=𝑐2 ∴𝑐2=0
𝐿∶0=𝑎2+𝑏2−3 𝑐2∴𝑎2=1
𝑇∶0=−𝑏2−1∴𝑏2=−1
π2=𝑑1.𝑉−1.𝜌0.𝑁=𝑁𝑑/𝑉

EXAMPLE P1
•For π3 term :
𝜋3=𝑑𝑎3.𝑉𝑏3.𝜌𝑐3.𝜇
𝑀0𝐿0𝑇0=𝐿𝑎3.𝐿𝑇−1𝑏3.𝑀𝐿−3𝑐3.𝑀𝐿−1𝑇−1
𝑀∶0=𝑐3+1∴𝑐3=−1
𝐿∶0=𝑎3+𝑏3−3𝑐3−1∴𝑎3=−1
𝑇∶0=−𝑏3−1∴𝑏3=−1
π3= 𝑑−1.𝑉−1.𝜌−1.μ=𝜇/𝑑𝑉𝜌
•So, substituting the values of 𝜋1,𝜋2𝑎𝑛𝑑𝜋3in equation (1),
EXAMPLE P2
•Determine the relationship among the parameters z, t, wo, zoand g of a sphere
falling in vacuum.

•Pretend that we do not know the equation related but only know the relation of
elevation z is a function of time t, initial vertical speed w0, initial elevation z 0, and
gravitational constant g.

EXAMPLE P2
STEP 1

STEP 2

n=5
EXAMPLE P2

The primary dimensions are [M], [L] and [t].

The number of primary dimensions in the problem are (L and t).

Step 3:
Then the number of pie’s predicted by the Buckingham Pi
theorem is

EXAMPLE P2
Need to choose two repeating parameters since j=2.
Therefore

Step4 :

Caution
1. Never choose dependent variable
2. Do not choose variables that can form dimensionless group
3. If there are three primary dimension available , must choose repeating
variables which include all three primary dimensions.
4. Don’t pick dimensionless variables. For example, radian or degree.
5. Never pick two variables with same dimensions or dimensions that differ
by only an exponent. For example, w0and g.
6. Pick common variables such as length, velocity, mass or density. Don’t
pick less common like viscosity or surface tension.
7. Always pick simple variables instead of complex variables such as
energy or pressure.
EXAMPLE 3
•P3: The resisting force Rof a supersonic plane during flight can be considered as
dependent upon the length of the aircraft l, velocity V, air viscosity µ, air density ρ, and
bulk modulus of air K. Express the functional relationship between the variables and the
resisting force.
•Solution:

Now each Pi-term is solved by the principle of dimensional homogeneity

This relationship enabled experiments to be designed, which led eventually to the Colebrook-White equation.
PROBLEM 1
•A thin rectangular plate having a width, w, and height, h, is locatedso that it is normal
to a moving stream of fluid. Assume the drag D, thatthe fluid exerts on the plate is a
function of w and h, the fluid viscosityand density µ, and ρ, respectively, and velocity V
of the fluidapproaching the plate. Determine a suitable set of pi terms to study
thisproblem experimentally.
PROBLEM 2
PROBLEM 2 –SOLUTION
EXERCISES 1

EXERCISES 2

EXERCISES 3
EXERCISES 4
A small sphere of density 𝜌𝑠and diameter D settles at a terminal velocity V in a liquid of density 𝜌𝑓and dynamic velocity 𝜇.
Gravity gis known to be a parameter. Express the functional relationships between these variables in a dimensionless form.

Hint:there are 6 variables,

3 basic dimensions
∴6-3 dimensionless terms
Select D, 𝜌𝑓, and g as the repeating variables
Obtain the indices of the 𝜋terms

𝑽𝒈𝑫=𝝋𝝆𝒔𝝆𝒕,𝝁𝝆𝒕𝑫𝒈𝑫