Chapter 2

Dimensional Analysis and Similitude Analysis

2.1 Introduction

Dimensional analysis is a formal procedure of grouping a number of physical

variables that participate in a physical phenomenon, into a minimum set of non-
dimensional groups, such that these groups contain all the physical variables at least
once and form a relationship among them. Buckingham’s π theorem is used to obtain
the minimum number of non-dimensional groups.

2.2 Buckingham’s π theorem:

It states that, if a physical phenomenon involves ‘m’ number of physical variables

which can be described using ‘n’ number of fundamental dimensions, then it is
possible to have (m - n) minimum number of non-dimensional groups to describe the
physical phenomenon.

The Buckingham’s π theorem can be explained with the help of a turbine as shown in
Figure 2.1 The turbine blade rotates at a speed of N rpm developing a power of P KW
from a static head ‘h’ m of fluid having dynamic viscosity  N-s/m2 and a flow
discharge of Q m3/s.

Power (P) H = gh
Speed (N)

H - Available static Head

 - Viscosity of or the potential energy
the fluid input to the turbomachine

Discharge (Q)

Figure 2.1 Schematic of a turbine

Step 1: Identify the physical variables that are into the physical process
Step 2: Prepare a table of dimensions.

1
Dr K V Karanth, MIT Manipal
  kgm 
 2 m m2
H = Specific energy =
Energy mgh
= = gh
Nm
= gh  s 
= gh 2
mass m kg kg s
Energy mgh mgh Nm
Also = = =h = hm
weight weight mg N
N
m2 =  kgm / s m =  kg
2 2
du 
 =μ or μ =
dy du / dy m / s du / dt 1/ s du / dt m s
m

TABLE OF DIMENSIONS
No. Variable Symbol Unit of Fundamental Remark
Name Measurements dimension
Geometric
1 Diameter D m L
Variable
RPM (rev/min or Kinematic
2 Speed N T-1
rad/sec) Variable
Dynamic
3 Density ρ Kg/m3 ML-3
Variable
Head
J/kg = N-m/kg Performance
4 (specific H L2T-2
m2/s2. Variable
energy)
Discharge Performance
5 Q m3/sec L3/T-1
(flow rate) Variable
Performance
6 Power P N-m/s = J/s ML2T-3
Variable
Physical
Dynamic
7 µ N-s/m2 = kg/m-s ML-1T-1 property of the
Viscosity
fluid

Step 3: Determine the minimum number of dimensionless groups that can be

formed from the given set of variables.

Thus, using Buckingham’s π theorem, the minimum number of non-dimensional

groups = (m - n)
Where, m = the number of physical variables involved = 7
n = the number of fundamental dimensions involved = 3
Therefore the minimum number of π terms (non-dimensional number) = 7 – 3 = 4

Step 4: Determination of the π terms

π1 = M0L0T0 = [D]x [N]y [ρ]z [Q]
π2 = M0L0T0 = [D]x [N]y [ρ]z [H]
π3 = M0L0T0 = [D]x [N]y [ρ]z [P]
π4 = M0L0T0 = [D]x [N]y [ρ]z [µ]
Applying dimensional homogeneity principle to each π term, consider π1;
π1 = Dx Ny ρz Q

2
Dr K V Karanth, MIT Manipal
 M0L0T0 = (Lx) (T-1)y (ML-3)z (L3T-1)
= (M)z (L)x – 3z + 3 (T)-y -1.
Now,
Balancing the indices, by applying dimensional homogeneity,
M :0=z →z=0
L : 0 = x – 3z + 3 = x + 3 → x = -3
T : 0 = -y – 1 →y = -1

Therefore π1= D-3N-1ρ0.Q

Q
=
ND 3
This non-dimensional number is called capacity coefficient or flow coefficient.

Similarly; π2 = Dx Ny ρz H
M0L0T0 = (L) x (T-1)y (ML-3)z (L2T-2)
= (M)z (L)x – 3z + 2 (T)-y -2.
Balancing indices,
z = 0, x = -2, y = -2

Therefore, π2 = D-2 N-2 ρ0. H

H
π2 = 2 2
DN
This non-dimensional number is called ‘Head coefficient’

Again, π3 = Dx Ny ρz p
M0L0T0 = (L) x (T-1)y (ML-3)z (ML-2T-3)
= (M)z + 1 (L)x – 3z + 2 (T)-y -3.
Balancing indices;
z = -1; x = -5; y = -3
P
π3 = 3 5
N D
This non-dimensional number is called ‘Power coefficient’.

π4 = Dx Ny ρz µ
= (M)z + 1 (L)x – 3z - 1 (T)-y -1.
On balancing,
x = -2; z = -1; y = -1
Therefore π4 = D-2 N-1 ρ-1 µ

=
 ND 2
1
=
Re
This characteristic number represents the inverse Reynolds number.

2.3 Affinity laws of turbo machine:

3
Dr K V Karanth, MIT Manipal
There are two laws of affinity:
a) The law of corresponding speed ratio and
b) The law of corresponding diameter ratio.

2.31 The law of Corresponding Speed Ratio;

This law states that “for the given turbo machine operating at two different
speeds, the ratio of the discharge is directly proportional to the ratio of speeds,
the ratio of head is proportional to the square of the speed ratio and the ratio of
power is proportional to the cube of the speed ratio”.

Proof:-
Let 1 and 2 subscripts indicate the two operating conditions for the given turbo
machine.
2 3
Q N H N  P N 
i.e. 1 = 1 , 1 =  1  and 1 = 1 
Q2 N2 H2  N2  P2  N2 
H2 (Outlet Pressure)
H 2( Delivery head)

N (speed in rpm)
Q (flow rate)

P1 (Inlet Pressure) Power Input (P)

H 1( Suction head)

)
Figure2.2 flow inside a centrifugal pump
Q
(i) We have, π1 = = Capacity coefficient (or discharge coefficient)
ND 3
Therefore the capacity coefficient corresponding to operating conditions 1 and 2,
is given as
Q1 Q
3
= 2 3 (Because non-dimensional numbers are unique)
N1D1 N2D2
For a given machine, D1 = D2.
Q N
Therefore 1 = 1 i.e. if the speed is doubled, the discharge also is doubled.
Q2 N 2
H
(ii) We have, π2 = Head Coefficient = 2 2
DN
Therefore the head coefficient corresponding to operating conditions 1 and
2, can be written as

4
Dr K V Karanth, MIT Manipal
 H1 H
2 2
= 22 2
D1 N1 D2 N2
2
H1  N1 
=  i.e. the ratio of head is proportional to the square of the speed
H2  N 2 
ratio
P
(ii) We have π3 = Power coefficient = 3 5
ND ρ
Power coefficient corresponding to operating condition 1 and 2 can be
written as
P1 P
3 5
= 3 25
N1 D1 ρ1 N2 D2 ρ2
For a given machine, D1 = D2 and if incompressible ρ1 = ρ2.
3
P N 
Therefore 1 =  1  or the ratio of power is proportional to the cube of the
P2  N2 
speed ratio”.

2.32 The Law of Corresponding Diameter Ratio

This law states that – “for any two geometrically similar turbo machines running at
the same speed, the ratio of head is proportional to the square of the diameter ratio,
the ratio of discharge is proportional to the cube of the diameter ratio and the ratio
of power is proportional to the fifth power of diameter ratio”.
2 3 5
H D  Q D  P D 
i.e. 1 =  1  , 1 =  1  and 1 =  1 
H2  D2  Q2  D2  P2  D2 

Proof:
H
a) We have π2 =
N 2D 2
For a set of geometrically similar turbo machines operating at constant speed (rated
speed N1 = N2 = N)
H1 H
2 2
= 22 2
N1 D1 N2 D2
2
H D 
N1 = N2, Therefore 1 =  1 
H2  D2 

Q
b) We have π1 =
ND 3
For a set of geometrically similar turbo machine operating at same speeds,
Q1 Q
3
= 23
N1D1 N2D2
3
Q D 
Therefore 1 =  1 
Q 2  D2 

5
Dr K V Karanth, MIT Manipal
 p
c)  3 =
 N 3D 5
For a set of geometrically similar turbo machine operating at same speeds and
density being unchanged,
p1 p2
=
 N1 D1  N23D 52
3 5

5
P D 
Therefore 1 =  1 
P2  D2 

Remember:
• The law of corresponding speed is used to estimate the operating
performance of a given turbo machine at different speeds.
• The law of corresponding diameter is used to develop a prototype from a
model.

2.4 Specific speed of a turbo machine:

Specific speeds of a turbo machine are of two types:
• Specific speed of a pump (Power absorbing machine)
• Specific speed of a turbine (Power operating machine)

2.41 Specific speed of a pump:

Specific speed of a pump is defined as “the speed of a geometrically similar turbo
machine, operating on unit discharge and producing unit head” and is given by
N Q 
Ns (pump) =  3  where N → rpm; Q → m3/s; h → m
 4 
 h 
Proof:-
Q H gh
We have π1 = 3
and π2 = 2 2 = 2 2
ND ND ND
1/2 Q 1/2 g3/4 h3/4
Now, π1 =
3/4
and π2 = 3/2 3/2
N1/2 D3/2 N D
π1/2 Q 1/2 N3/2 D3/2
1
Dividing 3/4 = πnew = 1/2 3/2 
π2 N D g3/4 h3/4
1/2 N
Πp = Q 3/4
g h3/4
 N Q 
πnew =  3/4 3/4 
g h 
 Omitting g¾, we have the SI unit standard definition of specific
speed.

6
Dr K V Karanth, MIT Manipal
 N Q 
Therefore, Ns =  3 
 4 
 h 
This is called dimensional specific speed.

Let ns be the non-dimensional specific speed

n Q n Q 
p = =  3/4  N → rps; Q → m3/s; H → N-m/kg = J/kg
( gh)
3/4
H 

 Specific speed is a unique characteristic number and it defines an

entire class of turbo-machine for a given range of specific speed

If Q = unity, H= Unity then, p is = N, thus it satisfies the definition of specific speed.

2.42 Specific speed of a turbine

Specific speed of a turbine is defined as- “the speed of a geometrically similar
turbine which delivers unit power against unit head on which it works”.

N P 
NT =  5 / 4  where N → rpm; P → Power in KW; h → Head in m.
h 
i.e. when P = unity = 1 kW and h = 1m, we have, Ns = N

This is a characteristic number which is used to classify turbines. The specific speed
criterion is used in design, evaluation and prediction of the performance of turbines.

Proof:
H gh
We have, π2 = 2
= 2 2
2
ND ND
5/4  g 5/4 h5/4 
π
Therefore 2 =  N5/2 D5/2  ------ (2.1)
 
P
Also, π3 =
ρN 3D 5
P
Therefore π1/2
3 = ------ (2.2)
N D ρ1/2 3/2 5/2

Eliminating diameter D between eq.2.1 and eq.2.2, we get

P
π31/2 3/2 5/2 1/2
N D ρ  N P 
5/4
=π= 5/4 5/4
=  1/2 5/4 5/4 
π2 g h ρ g h 
5/2 5/2
N D

7
Dr K V Karanth, MIT Manipal
For hydraulic turbine, since ρ is a constant and g = 9.81 m/s2, it is industrial practice
to define specific speed of a turbine as
N P 
NT =  5/4 
 h 

Non-dimensional specific speed for a turbine is given by,

 n P  2 N
π = T =  1/2 5/4  Where n = rad/s or rps
ρ H  60
P is in W
ρ is in kg/m3 and H is in N.m/kg = m2/s2.

2.43 Classification of turbo machines based on specific speed (Ns):

Dimensional specific speed (Ns) 10 30 250 850

Non-dimensional specific speed (ns ) 0.05 0.1 0.2 2.0 10 20

Pelton turbine Francis turbine Keplan turbine

Centrifugal pumps Axial flow pumps

Remember:
N Q N P
For dimensional specific speed Np = 3/4
, NT = 5/4
h h
Units: N → RPM; Q → m /S; h → m; P → kW.
3

n Q n P
For Non dimensional specific speed, p = , T = 1/2
( gh) ρ ( gh)
3/4 5/4

Units: n → rad/s; Q → m3/s; ρ → kg/m3; h → m; P → W; g → 9.81 m/s2.

Note:
Larger the diameter, larger is the friction loss.
High Reynolds number means high velocity.

2.44 Specific Speed of Pumps

Radial flow pump Mixed flow pump Axial flow pump

Radial exit Radial exit

Radial inlet Axial inlet

Medium specific speed,

Low specific speed, low discharge, Large High specific speed, large
medium discharge,
size machine and head output is high. discharge, compact in size
medium size and
and low head.
medium head.

2.45 Specific speed of turbines

8
Dr K V Karanth, MIT Manipal
 Specific speed in
[Link] Head in meters Type of turbine SI units

8.5 to 47
1 Above 300 Pelton turbine, Single or multiple jet
30 to 85
2 150 to 300 Pelton or Francis
85 to 188
3 60 to 150 Francis or Driaz (or Diagonal)
180 to 860
4 Less than 60 Kaplan, Propeller, Deiaz or Tubular

2.5 Unit Quantities

These parameters are obtained when the head on a given turbo machine ‘h’ is reduced
to unity. (i.e. h = 1)

2.51 Unit Discharge (Qu)

It is defined as- “the discharge obtained when the head on a given turbo machine
is reduced to unity.” i.e. when H = 1.0, Qu = Q
 Q 
Qu =  
 h
We have,
Q H
π1 = 3
and π2 = 2 2
ND ND
1/2
H
Therefore π1/22 =
ND
Q
π1 3  Q 
1/2
=  = ND 1/2
= 2  For a given machine, D = constant.
π2 H D H 
ND
 Q 
Therefore Qu =  
 h
2.52 Unit Speed
It is defined as- “the speed obtained when the head on a given turbo machine
is reduced to unity.” i.e. when H = 1.0, Nu = N
 N 
Nu =  
 h
H
We have, π2 = 2 2
ND
H
Therefore  = ,
ND
1 ND
π= =
π 2
H

9
Dr K V Karanth, MIT Manipal
 For a given machine, g and D are constants.
 N 
Therefore Nu =  
 h
2.53 Unit Power
It is defined as- “the power obtained when the head on a given turbo machine
is reduced to unity.” i.e. when H = 1, Pu = P
 P 
Pu =  3/2 
h 
We have,
H
π2 = 2 2 ,
ND
 H3/2 
Therefore π 3/2
2 =  N3D3  ---- (a)
 
P
π3 = ---- (b)
ρN 3D 5
P
3 ρN 3D 5 P
3/2
=  = 3/2
= 3/2 2
π2 H ρH D
3 3
ND
For a given machine, D an g are constants, and ρ for incompressible flow is also
constant.

 P 
Therefore Pu =  3/2 
h 
Unit quantity are used in the design, selection or estimation of a turbo machine and
also be used for performance prediction on different operating condition.

2.6 Non-dimensional analysis of compressible flow turbo machine

Blade
p
or vane
p 
02
01
D 01

Sonic velocity
=> P

. N
m


10
Dr K V Karanth, MIT Manipal
 Step-1: Identifying a list of physical variables involved in a compressible flow turbo
machine. → D, N,  , m, a0 , P,  and pressure p0
Step-2: Prepare a table of dimension.
P
From the list of variables, we find that pressure ratio 01 by itself is a non-dimensional
P02
number.
P 
Therefore π1 =  01  , by inspection.
 P02 
The remaining π terms are derived from Buckingham’s π theorem.
Variable Name Symbol Units Fundamental dimension Remark
Diameter D m L Geometric variable
Speed N RPM or RPS 1/T = T-1 Kinematic Variable
(rad/min)
(rad/sec)
Density ρ01 Kg/m3 ML-3 Dynamic Variable
Mass rate m Kg/sec MT-1 Performance
Variable
Sonic velocity a01 m/s LT-1 Flow variable
response for
compressibility
effect
Power P W=J/s = kgm2/s3 ML2T-3 Performance
variable
Dynamic µ N-s/m2 = kg/m-s ML-1T-1 Viscous friction
Viscosity flow property

Note:
1. Here we use mass rate m instead of volume flow rate Q, as the fluid is
compressible and hence density keeps changing from inlet to exit. Hence it is
appropriate to use mass flow rate for compressible flow analysis.
2. The sonic velocity a0 is used as fundamental variable in compressible flows
where the flow velocities are generally large compared to incompressible flow.
Step-3: Construct the π terms using Buckingham's π terms.
The number of physical variables involved = m = 7
The number of fundamental dimensions involved = n = 3
Therefore, the minimum number of π terms that can be formed = (m - n) = 7 – 3 = 4
P
Let π2, π3, π4, and π5 be these non-dimensional parameter (π1 = 01 ) from inspection.
P02
Now, we have
π2 = M0L0T0 = [D]x [N]y [ρ01]z [a01]
π3 = M0L0T0 = [D]x [N]y [ρ01]z [ m ]
π4 = M0L0T0 = [D]x [N]y [ρ01]z [P]
π5 = M0L0T0 = [D]x [N]y [ρ01]z [µ]

11
Dr K V Karanth, MIT Manipal
Considering π2 term, we have
π2 = M0L0T0 = [L]x [T-1]y [ML-3]z [LT-1]
Therefore π2 = M0L0T0 = [M]z [L]x – 3z + 1 [T]-y-1
Apply the dimensional homogeneity principle, we have,
Balancing the indices of each fundamental dimension,
M = 0 = z => z = 0
L = 0 = x-3z + 1 => x = -1
T = 0 = - y – 1 => y = -1
a
Therefore π2 = D-1N-1ρ01 0
a01 = 01
ND
πDN
Since u = i.e u  N.D
60
We can write,
a   u 
π 2 =  01  => π2 =   . This is nothing but the Mach number of flow.
 u   a01 
Now, we have,
 u 
π2 = M =  
 a01 
 ND 
We know that, a01 =  RT01 , Therefore π2 = 
  RT 
 01 

For a given machine handling a specific fluid, D, R and  are constants and hence we
 N 
have π2  
 T 
. This is a characteristic number for any compressible flow turbo
 01 
machine and is called “speed parameter”. It is used in the performance estimation,
prediction and analysis of the turbo-machine..

If flow in high speed machine is considered, for π3 terms, we have

π3 = [D]-3 [N]-1 [ρ01]-1 [ m ]
π3 = M0L0T0 = [L]x [T-1]y [ML-3]z [MT-1]
= (M)z + 1 (L)x – 3z (T)-y-1.
M: 0 = z + 1 => z = -1
L: 0 = x – 3z => x = -3
T: 0 = -y – 1 => y = -1
Therefore π3 = [D]-3 [N]-1 [ρ01]-1 [ m ]
 m 
= 3
 ρ01ND 
Now, by equation of state for a gas (any compressible medium)
p01
= RT01
ρ01
p 
Therefore ρ01 =  01 
 RT01 

12
Dr K V Karanth, MIT Manipal
 m  1  1 
Therefore π3 = ×  
 p01   N×D   D2 
 
 RT01 
m  RT01 

 N×D 
π3 = × RT01
p01D2  
 m RT01   RT01 
=
 p01D2   N×D 
×
   
We have already shown that, for a given compressible flow machine handling a
specified fluid,
RT01 T N
 01 or speed parameter  =
ND N T01
 m RT0  1
Now let π3 = 
 p01 D 2  ( )
 
For a given compressible flow machine, D and R are constants.
 m T0 
The term,  = 
 p01 
is defined as “mass parameter”
 
π3 = f ( ,  )
The mass parameter is an important characteristic number for any compressible flow
machine. This is also used in the performance estimation, design and prediction of
performance at off-design conditions.
g
sin
os
cr
In

Surge line
( Increasing)
Unstable region
p p
02
01

p01 p02

Stable region 4 4

3 
2 3 
1
2


1
Choking limit
 
Fig 2.3 Performance curve of an air Fig. 2.4 Performance curve of a gas
compressor turbine

➔As can be seen from the above performance graph, there is clearly a marked surge
line, which represents the limit of stable operating range for a given pair of  and φ.

13
Dr K V Karanth, MIT Manipal
 m T01  N  p
In the graph X – Axis →  =  = and Y – Axis → 02
p01  T  p01
 01 

➔ As can be seen from the above performance graph, there is clearly a marked choke
line, which represents the limit of the stable operating range for a given pair of  and
φ, for a turbine.

P
Now, π 4 =
ρ01 N3 D5
P
Therefore, π4 = and is called the “Power coefficient”.
ρ01 N3 D5
We know that P = m CpΔT0.
mCpΔT0 m
π4 = Mass rate =
(ρ01D ) (ND )(ND )2
2
ρ01 (ND )
v12 v2 v2
h01 = h1 + => Cp T01 = Cp T1 + 1 => T01 = T1 + 1
2 2 2Cp

2
m= ρAV , ρ01N3D5 = ρ01D2 (ND)(ND)2 = m(V )
 mc ΔT 
Therefore π 4 =  p 2 
 mv 
v12
We have, T01 = T1 + where; T01 is called the stagnation temperature or total
2cp
temperature.
 v2 
i.e. T01   
c 
 p
 ΔT 
Therefore π4 =   => Temperature ratio
 T01 
 μ  1  ρ01ND2 
Lastly,, π 5 =  2 
→ = → Reynolds number of flow.
 ρ01ND  π 5  μ 
 P   ΔT  
We can write;  01  ,    = f  ,  , Re 
 P02   T01  

2.7 Efficiencies of pump and turbine:-

14
Dr K V Karanth, MIT Manipal
 2.71 Efficiency of pump:

h
h = Manometric head M - Motor P - Pump

PI P0
M P

PI = Power Input
to the pump Energy flow diagram for a pump

Fug. 2.5 Pressure head from a pump

Power input to the pump = PI

(Useful) Power output of the pump = m g h = ρ Q g hmanometer
Power output ρQ ghmanometer
Now, ηPump = =
Power input PI
Assuming 100 % efficiency,
PI = ρQghmanometer.
Therefore, an estimate of the motor power required can be made by knowing
the kind of fluid to be lifted, the head and discharge of the pump.

2.72 Efficiency of a Turbine

Power input = m g havailable, where m =mass rate of flow (kg/s)
= ρ Q g havailable.

P0

h = Available Head (m)

Fug. 2.6 Power from a Pelton turbine

The power of pump and turbine are complimentary to each other

pump ➔Turbine
2πNT
Power output = Po = (kW)
60 ×1000
Power output Po
Therefore efficiency of turbine = =
Power input ρQghavailable
If efficiency of turbine = 100%, Po = ρQgh

15
Dr K V Karanth, MIT Manipal
2.8 The Effect of Friction and Reynolds number on Flow through a Turbo Machine

When the flow inside a turbo-machine is at high speeds, the flow tends to become
turbulent due to high velocity of the flow. Apart from this, there are dissipative losses
inside turbo machine such as losses due to viscous effect, roughness, leakage losses
etc. Hence a strict kinematic similarity between model and prototype cannot be
expected under such conditions though a geometric similarity exists between them.
The efficiency of the model and prototype may not correspond same values under
such circumstances. Hence there is a need to develop a relationship concerning the
loss of efficiency of model and prototype based on dissipative effect. Moody
suggested an empirical formula for relative effect of model and prototype as given
below.

2.9 Moody’s formula for Turbines:-

n
 1 - η1   D2 
  =   , where ‘n’ is an empirical constant and η1 and η2 are the model and
 1 - η2   D1 
prototype efficiencies respectively. D1 and D2 are the diameters of the model and
prototype.
n = 0.2 for most of the hydraulic turbines obtained by experimentation.
For pumps and compressors, Moody suggested a different empirical relationship given
by-
n
 1 - η1   Re2 
 =  Where Re1 and Re2 are the Reynolds number of the flow in
 1 - η2   Re1 
model and prototype respectively, the suggested n = 0.25 to 0.56 are obtained from
experiment.

2.91 Derivation of Moody’s empirical formula for Hydraulic turbines:

1. The fluid friction losses are only due to the viscous effects i.e. due to viscosity
 fLV2 
of the fluid. If Δh is the loss of head, then, Δh =   , for such a case.
 2gD 
2. The friction coefficient ‘f’ is proportional to relative roughness of the flow
1
passage, where f  n where ‘n’ is an empirical constant.
D
3. There are no dissipative losses like leakage.

Let ‘H’ be the ideal head available for a turbo-machine and ΔH be the flow loss in it.
Actual Head
Therefore Efficiency= η =
Ideal Head
(h - Δh)  Δh 
= = 1- 
h  h 
 fLV2  1
From assumption 1, we have Δh =   and from assumption 2, we have f  n
 2gD  D

16
Dr K V Karanth, MIT Manipal
  k  2 
  Dn  LV   k  L   V 
2
Δh =     or Δh =  n    
 2gD   D  D   2g 
 
 Δh 
η= 1 -  
 h 
Now, let subscripts 1 and 2 represent model and prototype respectively.
 k   L   V2 
Therefore (1 - η1 ) =  n   1   1 
 D1   D1   2gH1 
 k   L   V2 
And (1 - η2 ) =  n   2   2 
 D2   D2   2gH2 
 L1   V1 
2

 
(1 - η1 )  D2   D1   2gh1 
n

Therefore =   
(1 - η2 )  D1   L2   V22 
 D   2gh 
 2  2 

Now, imposing a strict geometric similarity between the model and prototype,
L1 L 2
=
D1 D2
V12 V22
And for dynamic similarity, =
2gh1 2gh2
n
 1 - η1   D2 
Then,  =  where n is an empirical constant and is to be determined
 1 - η2   D1 
from experiment (n = 0.2)
Question:
What is the significance of specific speed as applicable to turbo machines? Explain
with neat sketches how the specific speed influences the shape and size of pump
impeller.

Guidelines for solving problems on similitude and dimensional analysis of

turbo machine:
1. Use specific speed criterion, when comparing the model and prototype.
2. When scale ratio (the diameter ratio) of the model and prototype is given, use
affinity law2. i.e. law of corresponding diameter ratios.
3. When solving for a different set of operating parameters for a given turbo
machine, use affinity law 1. i.e. law of corresponding speed ratios.
4. In general, any of the π terms like π1, π2, π3, π4, and π5 that are derived, can be
used between the model and the prototype or corresponding two operating
conditions of a given turbo machine,
i.e. (π1) model = (π2) prototype Etc. → for models prototype
or (π1) at operating point (A) = (π2) at operating point (B) etc. → for a given machine.

17
Dr K V Karanth, MIT Manipal
 Dimensional Analysis and Similitude Analysis
Tutorial 1

Problem 1: A radial flow hydraulic turbine is required to be designed to generate 30

MW under a head of 14 m at a Speed of 95 rpm. A geometrically similar Model with
an output of 40 kW and a Head of 5 m is to be tested under dynamically similar
conditions. At what speed the model is run? What is the scale ratio of the model to
the prototype and discharge through the model, if its efficiency is assumed to be
90%.

Problem2: A pump with an available driven speed of 800 RPM is required to

overcome a 2 m head, while pumping a discharge of 0.2 m3/s. What type of pump
is required and what is the power required?

Problem3: A reservoir has a head of 40 m and a channel leading from the dam
permits the flow rate of 35 m3/s. If the rotational speed of the rotor is 150 RPM,
what is the most suitable type of turbine required?

Problem 4: A customer approaches a salesman with a particular pump

requirement and is quoted for an axial flow pump of impeller diameter 150 mm.
running at a speed 900 RPM. The machine is said to deliver 0.3 m 3/s of water
against a head of 10 m at an efficiency of 85%. Are the claims of the salesman
realistic?

Problem 5: In a proposed hydraulic power plant site, the available discharge in the
river is 70 m3/s under a net Head of 15 m. Assuming a turbine efficiency of 90 % and
the Speed as 100 RPM, compare the following two proposals, both of the same size
in so far as the number of turbines are concerned.
(i) Francis turbines with a Specific Speed of 150.
(ii) Kaplan Turbines with a Specific Speed of 350.

Problem 6: A single stage centrifugal pump lifts 50 liters/s of water at 1800 RPM.
The diameter of the impeller is 250 mm and power required to run the same is 20
kW. Determine diameter of each impeller of multistage pump which has
dynamically similar impellers and which delivers 75 liters/s of water running at 1200
RPM. Determine the number of stages required of the multistage pump if the total
head developed by this pump is 155 m.

Problem 7: An axial flow compressor is designed to run at 4200 rpm, when ambient
atmosphere conditions are 101.3 kPa and 150C with mass flow rate of 70 kg/s. On

18
Dr K V Karanth, MIT Manipal
the day when the performance characteristics are obtained the atmospheric
temperature is 300C. What is the correct speed at which the compressor must be
run? If the entry pressure is changed to 62 kPa, compute the mass flow rate obtained
in the test. Assume given conditions correspond to stagnation conditions.

Problem 8: A large centrifugal pump contains a liquid whose kinetic viscosity is 6

times that of water. The dimensionless specific speed of the pump is 0.183 and it is
to discharge 2 m3/s of liquid against a total head of 15 m. Determine the test speed
and head for a one quarter scale investigation model of the full size pump, if the
model uses water.

Problem 9: A Pelton turbine produces 10 MW of power while working under a head

of 500 m. The speed is 300 RPM. Assuming the turbine efficiency to be 80% compute
the values of Unit quantities. If the head on the turbine falls to 350 m find the new
discharge, power and speed for efficiency to remain same.

Problem 10: A multi stage Centrifugal Pump is required to lift 2 m3/ min of water
from a mine, the total head including friction being 180 m. If the speed of the pump
is 6000 RPM, find the least number of stages if the specific speed per stage is not to
be less than 100.

Problem 11: A multi stage centrifugal pump having six stages with 225 mm
diameters impellers develop a head of 120 m when running at 1500 RPM and
discharging 5.45 m3/min. Four geometrically similar stages 300 mm diameter is used
to build a multi stage pump, which is to run at 1000 RPM. Assuming that each stage
in each pump operates under dynamically similar conditions, find

(a) The quantity of discharge

(b) Head that will be developed by this pump.

Problem 12: The specifications for an axial flow coolant pump for a pressurized
nuclear reactor are:
1, Head = 85 m.
2. FloW rate =20,000 m3/ hr.
3. Speed = 1490 RPM.
4. Diameter 1200 mm.
5 Power = 5 MW.
The manufacturer plans to build the model. The test conditions are:
- Available electric power = 500 kW,
- Flow rate = 0.5 m3/ s of cold water.

19
Dr K V Karanth, MIT Manipal
 If the model and the prototype efficiencies are same, find the head, speed
and scale ratio of the model. Calculate the dimensionless specific speed of the
prototype

Problem 13: Two geometrically similar pumps are running at same speed of 1000 rpm. One
pump has an impeller diameter of 0.3 m and lifts water at the rate of 20 lit/s against a head
of 15 m. Determine the head and the impeller diameter of the other pump to deliver half
the discharge.

Problem 14: A small scale model of a hydraulic turbine runs at 360 rpm under a head
of 22 m and produces 10 kW output. By assuming turbine efficiency as 80%,
Determine: i) Unit discharge ii) Unit speed iii) Unit power. If a prototype turbine of
12 times the model size with its efficiency same as that of model is used, i) What is
the power output of the prototype? ii) What is the dimensionless specific speed if
the head available for the prototype is 250 m.

20
Dr K V Karanth, MIT Manipal