Fluid Services and Heat Transfer

Equipment
ME5512/ME5545
Carola König
Reminder to start lecture recording!
Week start date ME5512 Chapters Comments

18 Fri, 19 January 2024 1) Basic Aspects of Fluid Dynamics

2) Fluid Friction
19 Fri, 26 January 2024 3) Design of Ventilation Systems Equal P drop method
Velocity method
20 Fri, 2 February 2024 3) Design of Ventilation Systems Balanced capacity method
Static regain method
21 Fri, 9 February 2024 4) Fan Characteristics Performance laws
Fan performance
22 Fri, 16 February 2024 4) Fan Characteristics Fan matching
The basic ideas in fluid mechanics

1) Stress

2) Stress in a moving fluid: viscosity

3) Hydrostatic pressure/head

4) Hydrodynamics

5) Energy conservation of a moving fluid

6) Bernoulli’s theorem

7) Flow with friction losses or energy dissipation

8) Flow with energy addition or extraction

9) Fluid power
 Stress

• Stress: force per unit area F/A, unit: N/m2

• It is a vector with a normal component Fn and tangential component Ft
• For fluid flow:
– the normal stress is the pressure: acts equally in all directions and is
independent of the orientation of the surface;
– the tangential stress is analogous to the frictional force Ff.
It resists the relative motion of neighbouring layers of fluid.
Viscosity: The characteristic property of a fluid

Viscosity, m, is the property of a fluid, due to cohesion and

interaction between molecules, which offers resistance to sheer
deformation. Different fluids deform at different rates under the
same shear stress. Fluid with a high viscosity such as syrup,
deforms more slowly than fluid with a low viscosity such as
water.

shear stress 
m= = [Ns/m = kg/ms]
2

rate of strain du dy
Viscosity: The characteristic property of a fluid

V
Moving plate

h Fluid du dy = V h

Fixed Plate

y
Velocity
Gradient

du / dy
u(y)

Fixed Plate
 Kinematic viscosity
dynamic viscosity m
= = [m2/s]
fluid density 

Table 1.1
Representative Viscosity Coefficients

Fluid State Dynamic Viscosity Density Kinematic

Ns/m2 kg/m3 Viscosity
(10 −6 ) m2/s (10 −6 )

Air 0°C, 1 bar 17 1.293 13.15

Air 20°C, 1 bar 17.9 1.225 14.6
Air 102 °C, 1 bar 21.8 0.941 23.2
Water 0 °C, 0.01 bar 1780 1002 1.78
Water 20 °C, 0.04 bar 1010 1001 1.01
Water 100 °C, 1.01 bar 283 961 0.29
 Hydrostatics
All forces due to the pressure of the fluid are normal to the surfaces on which they act,
although pressure inside a fluid may vary from point to point.
Variation of fluid pressure under gravity.
F = patmos  A
Free surface
Plane (1)

Column of Weight
h
cross-section A

Plane (2)

p2gauge F = pabs  A
h=
g
or
p2gauge = gh
h = hydrostatic head.
 Manometers
Devices that contain a liquid column whose weight is
balanced against a pressure differential.

Table 1.2
Densities of Manometer Fluids
in Atmospheric Conditions

Liquid Density [kg/m3]

Water 1001

Mercury 13550

Alcohol 790

p = gh [Pa]
Manometers
Hydrodynamics

FLOW TYPES

1) Uniform and steady

2) Non-uniform and steady
3) Uniform and unsteady
4) Non-uniform and unsteady

Steady flow is present when the various

parameters at any point do not change with
time.

If, at a particular instant, the various quantities

do not change from point to point over a
specified region, then the flow is said to be
uniform over that region.
 u
1) Uniform Flow

u is not a function of y
y

2) Non-uniform flow u(y)

u is a function of y, i.e. u = f ( y)
 Laminar and Turbulent Flow
Laminar Case Path 1

Path 2
Starting
Point
Transition

Turbulent Case Path 1

Path 2
Starting
Point
 Laminar and Turbulent Flow
Laminar flow
a fluid flows in parallel layers, with no disruption between the
layers

Turbulent flow
a flow regime characterised by chaotic, stochastic property
changes
Laminar and Turbulent Flow
 Laminar and Turbulent Flow
Leonardo Da Vinci (1452 – 1519)
 Laminar and Turbulent Flow
Osbourne Reynolds (1842 – 1912)
 Dynamics – Reynolds number
The Reynolds number, Re, is usually used to identify and predict
different flow regimes, such as laminar or turbulent flow

𝜌 𝑢 𝐿 𝑖𝑛𝑒𝑟𝑡𝑖𝑎𝑙 𝑓𝑜𝑟𝑐𝑒 𝑀/𝐿3 𝐿/𝑇 [𝐿]

Re= = =1
𝜇 𝑣𝑖𝑠𝑐𝑜𝑢𝑠 𝑓𝑜𝑟𝑐𝑒 [𝑀/(𝐿𝑇)
]
where L is a characteristic length scale.

For flow in a pipe for instance, the characteristic length is the

pipe diameter.

As Re has no units it is a non-dimensional number.

Re => 10,000
 One-dimensional flow
•Useful for analysing flow in pipes and ducts
•Assumes flow in a duct can be represented by a single velocity.

Actual velocity
y u(y) distribution

One – dimensional
uu
approximation

In all following computations we shall assume we are dealing with the mean velocity
over the pipe cross section (ū), which we shall write simply as u.
 Continuity Equation
For steady flow
For a simple pipe, of varying cross section, application of
the continuity equation gives:

Rate at which mass Rate at which mass

=
enters a control leaves a control
volume volume

Plane 1 Plane 2
A1 A2
u2
u1
u(x), A(x)
y
x
 ( x)u ( x) A( x) = constant
•
1u1 A1 = 2u2 A2 = const. = m [kg / s]
u1 A1 = u2 A2 = constant = Q
 Continuity Equation
(3)
(1) (2)

1u1 A1 = 2u2 A2 + 3u3 A3

If fluid is incompressible u1 A1 = u2 A2 + u3 A3
Q1 = Q 2 + Q 3
Energy conservation for a Moving Fluid
Energy can not be created nor destroyed
– 1st Law of Thermodynamics

Q + W = E [J]

Energy E = Pressure Energy + Potential Energy + Kinetic Energy

Assuming Q=0 and dividing by mass, a change in energy is related to a change in work:

e = w [J/kg]

e2 − e1 = w [J/kg]
1→ 2

u2 p
Energy/unit mass e = + + gz [J/kg]
 2
Energy conservation for a Moving Fluid
 p2 u 22   p1 u12  w1− 2 [m]
 + + z 2 −  + + z1 =

 2 g 2 g   1  g 2 g  g

Assumption of uniform pipe flow:

u(y) u
y

u( y)   u
 p2 u 22   p1 u12  w1− 2
 +  + z −
2  +  + z1 =

 2 g
2
2 g 
  1 g
1
2 g  g
Energy Equation Pressure Velocity
p2 u2

p1 u1 Elevation
z2
Pump /
z1
Turbine
z=0

 p2 u 22   p1 u12  w1− 2
 +  + z −
2  +  + z1 =

 2 g
2
2 g 
  1 g
1
2 g  g

where  is the kinetic energy coefficient

w
1→ 2 is the head equivalent of work done
g
w
1→ 2 >0 implies work done on the fluid (e.g. pump)
g
w
1→ 2 <0 implies work done by the fluid. (e.g. turbine)
g
 Bernoulli’s Equation
w , then the steady flow energy equation
If 1→2 =0
g reduces to Pressure p2 Velocity
u2

 p u12   p u 22  p1 u1 Elevation

 1
+ 1 + z1  =  2
+2 + z2  z2

 1 g 2g    2 g 2g  z1
Datum of elevation
z=0

If 1 =  2 =1 and 1 =  2, then this reduces to Bernoulli’s equation:

 p1 u12   p2 u22 
 + + z1 =  + + z 2
 g 2 g   g 2 g 

or

p u2
+ + z = constant
g 2 g
 . of Bernoulli’s equation:
Limitations on the use

1) No heat transfer, q = 0
2) No external (or body) forces other than gravity are present
3) There is no energy transfer by external devices such as pumps, fans,
turbines etc. (w = 0 )
4) There are no viscous or turbulent friction losses
5) The internal energy of the fluid remains constant
6) The fluid density remains constant (incompressible flow)
7) The flow is steady

These restrictions are seldom completely satisfied:

Restrictions 4 and 7 cause the biggest difficulties in fluid services problems,
since frictional losses almost always occur in pipe flow, and unsteady regions
of turbulence are usually present.
PF1.6 A non-technical friend says that he has heard that, contrary to
what he would expect, the pressure falls as the flow through a
channel squeezes through a contraction. Explain in simple terms
why this does occur in reality.

IX1.7 Just inside a fire hose the gauge pressure is 4 bar. Estimate the
velocity in the jet of diameter 50 mm just outside the nozzle,
taking the inside hose diameter to be 100 mm. How high do you
think the jet from the hose might rise, if the hose is pointed
vertically upwards?

RX1.8 A fountain in the reception area of a building periodically spurts

water into the air. The outlet nozzle is circular and has a diameter
of 5 cm. If the water leaves the nozzle at a velocity of 10 m/s
determine,

(a) The maximum height that the water will reach.

(b) The diameter of the jet at half its maximum height.

[Ans: a) h=5.1m; b) ] d = 0.0595 m

 Flow with friction losses or energy
dissipation

“Head loss”, parameter that encompasses losses due to friction and turbulence

 p   p 
 1 u12   2 u22 
 + 1 + z1  −  + 2 + z2  = hL  0 [m]
 1 g 2g    2 g 2g 

IX1.9 (a) Water discharges from a tank via a pipe which runs out
horizontally from the bottom of the tank. If the water in the
tank is 20m deep, and the head loss in the pipe is known to
be 10m, calculate the velocity of the water on exit from the
pipe.

(b) If the end of the pipe is now placed at a distance of 10m

below the bottom of the tank, and the head loss in the pipe
is now given as 12m, calculate the new velocity of the water
on exit from the pipe.
Solution to IX1.9

(1)

20m

(2)

 p   p 
 1 u12   2 u22  b) h = 30 m
a)  +  1 + z1 −  +  2 + z 2  = hL and hL = 12 m
 1 g 2g    2 g 2g 

Assume turbulent flow: =1 u2 = 2  9.81(30 − 12 ) = 18 .79 m/s

Solve for u2=……

u 2 = 2 g (h − hL )

u 2 = 2  9.81(20 − 10 ) = 14 m/s
 Total head line
(no friction)
2
u
2g Energy grade line
p u2
=h
Total Head h g 2g
z1+h Hydraulic
p Grade line
g p u2
g 2g

z1 z1 z1 z1=z2

z=0
datum
Plane 1 Plane 2

u2
Total head line
2g
p (no friction)
=h
Total Head h g
z1+h u2 Energy grade line
p
2g
g

p Hydraulic
g Grade line

u2
z1 z1 2g

z2
z=0
Plane 1 datum
Plane 2
IX1.10 Water is supplied to a building from a tank (in which the depth is 1.5 m)
on the 10th storey. Consider the flow from a tap located 1 m above the
floor of the 5th storey, whose floor level is 15 m below the 10th storey
level. The water emerges from the tap with a velocity of 5 m/s, in a jet
of diameter 1 cm.

(a) What is the head loss in the piping system conveying water from
the tank to the outlet from the tap?

(b) A 26-mm diameter pipe section leads to the tap and starts 0.5 m
below it. The gauge pressure measured at the beginning of this
pipe section is equivalent to a head of 5 m of water. What is the
head loss in this section?

(1)
1.5m

(2)
15m Tap
1m
 (1)
1.5m

(2)
15m Tap
1m

 p   p 
 1 u2   u2 
 + 1 1 + z1  −  2 +  2 2 + z2  = hL
 1 g 2g    2 g 2g 

u22
z1 − z 2 − = hL
2g
52
hL = 15.5 − = 14.23m
2  9.81

b) (2)
 p3 u32   p2 u22 
(3)  +  + z 3 −  +  + z2  = hL
 3 g    2 g
3 2
2g 2g 

 2  3  1
 p3 z3 − z2 = −0.5 m
= 5m and
g
(2)
u32 u22
(3) hL = 5 − 0.5 + −
2g 2g

u22  u32 
hL = 4.5 − 1 − 
2 g  u22 

d 24
Utilise continuity: u3 A3 = u2 A2 → u =u 4
2
3
2
2
d3

u22  d 24 
hL = 4.5 − 1 − 
2 g  d 34 

52  10 4 
hL = 4.5 − 1 −  = 3.25 m
2  9.81  26 
 4
Flow with energy addition or extraction
 p1 u12   p2 u22  w
 +  + z1 −  +  + z 2  = hL −
 1 g   2 g
1 2
2g 2g  g

 p1 u12   p2 u22 
 + 1 + z1  −  + 2 + z2  = hL − hW
 1 g 2g   2 g 2g 

Sign Convention for Work Transfer

Sign of Device
Work done Pump or fan: energy of downstream
on the fluid hW>0 fluid raised above that upstream
Work done Turbine or hydraulic motor: energy
by the fluid hW<0 taken from the fluid to give a reduced
downstream value
IX1.11 A pump is used to raise water from a tank (water depth 3 m) at
ground level to a tank (water depth 1.5 m) located on the 10th
storey of a building, whose floor level is 30 m above the ground.
The energy loss within the pipe leading from pump to the upper
tank can be estimated as equivalent to a head of 10 m, and that in
the supply line to the pump as 1 m. What rise in head must be
provided by the pump?
 IX1.11 A pump is used to raise water from a tank (water depth 3 m) at
ground level to a tank (water depth 1.5 m) located on the 10th
storey of a building, whose floor level is 30 m above the ground.
The energy loss within the pipe leading from pump to the upper
tank can be estimated as equivalent to a head of 10 m, and that in
the supply line to the pump as 1 m. What rise in head must be
provided by the pump?

(2)

1.5m

(1)
Pump 30m
3m

 p1 u2   p u2 
 +  1 1 + z1  −  2 +  2 2 + z 2  = hL − hW [m]
 1 g 2g   2 g 2g 

p1=p2=0 Pa u1=u2=0m/s hL total=10m+1m=11m

hW = ( z 2 − z1 ) + hL = 39.5m
 Fluid Power
Let us consider a simple high-pressure hydraulic system, where changes in both kinetic
energy and potential energy are insignificant compared to pressure changes. If a piston
moves at velocity u with a cylinder of cross section A, driven by force F giving rise to a
pressure differential Δp across the piston, we can calculate the power flowing to the fluid as

F
P = F  u = u  A = uA  p = Q p [W]
A

- Changes may occur in any of pressure, kinetic and potential energies.

- Losses may occur in the energy-transfer process, so that we must distinguish between
the reduction in the fluid's energy and the rate at which useful work flows from the
device.

- Energy may pass from the fluid (as in a turbine) rather than to the fluid (as in a pump),
as considered above.
1 2
pT = p + u + gy [N/m2]
2
 1 2
pT = p + u + gy [N/m2]
2

The rate at which work is done on the flowing fluid is then found to be

(
WP = Q pT2 − pT1 ) [W]

We have to distinguish between:

1) the rate Wp at which work is done on the fluid to increase its total
mechanical energy, and
2) the rate P at which energy is transferred to the pump or fan.

This difference can be expressed:

WP = P − D with D  0 where D is Dissipation rate

WP P − D
P = =  1 [dimensionless]
P P
The pressure differential can also be expressed as a change in head across the machine

pT = ghW [N/m 2 ], or  p = m

Q  ghW [W]
T

the rate at which mechanical energy is transferred to the flow is given by either:
i) the product of the volume flux and the rise in total pressure, or
ii) the product of the mass flux and the rise in total head (× g)

P =
(
Q pT2 − pT1 ) = m gh W
[dimensionless]
P P
These simple but very useful results describe the basic performance of the pumps and fans
that have many applications to building services. Parallel results can be developed for
devices - turbines and motors - that extract energy from a flowing fluid.

For a turbine
P P
t =  =
Q( pT − pT ) m ghW
P P
t = = 1 and
Wt P − D 2 1

Thus, the power transfer in or out of a device m ghw Q ghw

may be written in terms of hw using equation P= or P=
 
so that
RX1.12 Consider again the pump whose operation was described in IX1.11,
but this time it is required a volumetric flow rate of 20 litres per
second.

a) What is the rate at which mechanical energy is supplied to

the water passing through the pump?

b) If the hydraulic efficiency of the pump is 77 per cent and the

electrical efficiency of its motor is 83 per cent, what electrical
power does it require?

[Ans: a) P=7.75 kW; b) 12.13 kW]

RX1.13 Building services engineers are to install an outside pool with a

number of water fountains for a prestigious new office
development.

It is intended to supply the water to the fountain jets from a tank

35m above ground level. It is estimated that with a nozzle
diameter of 70 mm the exit velocity will be 25 m/s.

Determine
(a) The energy loss in the nozzle and pipes.
(b) The efficiency of the power transmission.

[Ans: 30.85kJ; b)  = 91 % ]
 Flow measurements

- long-established techniques of flow measurement involve disturbing the

flow in some controlled fashion and then measuring a pressure
differential that is generated (illustrating the convertibility between
kinetic energy and pressure energy)
- modern measuring devices utilise thermistors and other sensing
elements whose resistance is temperature-dependent and therefore
dependent on the velocity of the heat-removing flow (more expensive
but allowing for smaller devices
 The Pitot Tube
To sensor
Pitot Tube

Total Pressure p0

Velocity u

1 2
p0 = p s + u
where:
2
p0 is the pitot or stagnation pressure
ps is the static pressure which would be measured by an instrument aligned
normal to the fluid flow
 is the fluid density (assumed to be constant)
u is the mean fluid velocity
u 2 2 is the dynamic pressure, the increase in pressure generated by the
deceleration of the fluid to stagnation conditions
The Pitot-Static Tube

Pitot –Static Tube

To sensors

Static Pressure
Tapping
Total Pressure p0

Velocity u
Venturi and Orifice Meters
p1 − p2 =
1
2
(
 u 22 − u12 )
Venturi Tube

A1
A2

p2
p1

Orifice Meter

A1 A2

p1 p2
 with u1 A1 = u2 A2

2( p1 − p2 )
Q = Cd A2u2 = Cd A2
 A22 
 1 − 2 
 A1 

Where Cd is the discharge coefficient which account for dissipative losses due
to friction and turbulence and non-parallel flow

For Venturi meters Cd ranges from 0.95 to 0.99

For Orifice meters Cd  0.6
 Other Velocity and Flow Meters
• Elbow or bend meters
• Rotameter
• Propeller-type meters

• Hot-wire anemometer
• Thermistor-based velocity meters
• Laser-Doppler anemometer
• (Particle image velocimetry)
 IX1.14 A pitot tube is aligned with the centre-line of a duct of cross
section 300 mm by 200 mm, and points upstream into an air flow
whose pressure and temperature are close to the normal
atmospheric levels. The pressure differential (measured between
the pitot tube and a small hole flush in the duct wall at the section
at which the tube is located) is 12 mm of water gauge. Estimate
the rate at which air flows through the duct.

p =  w gh = 1000  9.81  0.012 = 118 Pa

with  air = 1.2 kg/m 3

2 p 2  118
uc = = = 14 m/s
 air 1.2

Assuming that the average velocity is 90% of the velocity measured

Q = kuc A = 0.90  14  0.20  0.30  60 = 45 .4 m 3 / min

RX1.4 At a section of a circular air-conditioning duct at which the
diameter is 50 cm, the average velocity of the air flow is 5 m/s.

(a) Estimate the volumetric flow rate and mass flow rate through
the duct.

(b) Estimate the velocity at a second circular section of diameter

35 cm, carrying the same flow.

[Assume air = 1.22 kg/m3]

[Ans: a) 1.2 kg/s; b) 10.2 m/s]

RX1.5 Air flows into a tee junction in a ventilation system (see Figure
(2) (3)
RX1.5). The inlet diameter is 30 mm and the inlet velocity is 7 m/s. (1)
The diameter of branch 2 is 25 mm and the velocity is 5 m/s. If the
diameter of branch 3 is 15 mm, find the velocity of the water in
branch 3.

[Ans: 14.11 m/s]

RX1.8 A fountain in the reception area of a building periodically spurts

water into the air. The outlet nozzle is circular and has a diameter
of 5 cm. If the water leaves the nozzle at a velocity of 10 m/s
determine,

(a) The maximum height that the water will reach.

(b) The diameter of the jet at half its maximum height.

[Ans: a) h=5.1m; b) ] d = 0.0595 m

RX1.15 A light aeroplane flies at 360 km/hr at an altitude of 3000 m,
where the pressure is 0.70 bar. What dynamic pressure will be
recorded by a pitot-static tube fixed to the wing? To answer this
question, you will presumably use a result derived from the
constant-density form of Bernoulli's equation. Why are you
justified in doing so?

[To answer this question look up values for  on P24 of the

thermodynamic tables]
[Ans: ] Pdynamic = 4.55 kPa

PF1.16 Explain what is meant by the terms static pressure and stagnation
pressure.

PF1.17 What forms of mechanical energy can be distinguished for a flowing

fluid? Why can they be described in terms of specific energies,
pressures and gravitational heads? List the name that have been
introduced in this chapter for the various forms of the energy of a
flowing fluid.

PF1.18 Why is the calibration coefficient for an orifice meter much smaller
than that for a Venturi meter?

RX1.19 A Venturi meter is used to measure the flow rate of water for
supply to a water tank. If the pipe diameter is 95 mm. Determine
the Venturi throat diameter if a u-tube manometer connected
across it indicates a hydrostatic head of 200 mm and the
volumetric flow rate of the water is 0.10 m3/s. The coefficient of
discharge for the meter is 1.00.

[Ans: ] d = 0.0945 m
 Pipe Friction

We shall now attempt to quantify the effects of pipe friction so we can include
these in our energy equation.

Consider a simple pipe:

L

Mean
Velocity u

Before we look at the he frictional head loss, h , we will do a little excursion into
L
dimensional analysis.
 Dimensional Analysis
Buckingham P Theorem
Consider pipe flow
Ø

L
 Dimensional Analysis
Buckingham P Theorem
Consider pipe flow
Ø

∆𝑃 = 𝑓 𝑢, 𝐿, 𝑑, 𝜌, 𝜇, 𝑘

L the length of the pipe

d the diameter of the pipe
u the mean velocity of the flow
m the dynamic viscosity of the fluid
 the density of the fluid
k the roughness height
 Dimensional Analysis
Buckingham P Theorem
∆𝑃 = 𝑓 𝑢, 𝐿, 𝑑, 𝜌, 𝜇, 𝑘
In a general form we can write this as

𝜋1 = 𝑓 𝜋2 , 𝜋3 , 𝜋4 , … , 𝜋𝑛

∆𝑃 = 𝑓1 𝑢, 𝐿, 𝑑, 𝜌, 𝜇, 𝑘

dependent variable independent variables

Then select 3 independent variables from your set such that the 3 basic dimensions L,
M, T are represented, e.g. u [m/s], d [m] and  [kg,m3].

Then use the 3 selected independent variables to non-dimensionalise all variables.

 Dimensional Analysis
Buckingham P Theorem

∆𝑃 = 𝑓1 𝑢, 𝐿, 𝑑, 𝜌, 𝜇, 𝑘

3 independent variables are e.g. u [m/s], d [m] and  [kg,m3]

Then non-dimensionalise all variables

∆𝑃 𝑢 𝐿 𝑑 𝜌 𝜇 𝑘
2
= 𝑓2 , , , , ,
𝜌𝑢 𝑢 𝑑 𝑑 𝜌 𝜌𝑢𝑑 𝑑

∆𝑃 𝐿 𝜇 𝑘
2
= 𝑓3 , ,
𝜌𝑢 𝑑 𝜌𝑢𝑑 𝑑
 Dimensional Analysis
Buckingham P Theorem

∆𝑃 = 𝜌𝑔ℎ𝐿
 Dimensional Analysis
Buckingham P Theorem

∆𝑃 = 𝑓1 𝑢, 𝐿, 𝑑, 𝜌, 𝜇, 𝑘

What happens if we choose different independent variables?

Let’s take d [m], u [m/s] and m [kg/ms]

Again, non-dimensionalise all variables with your selected set

∆𝑃𝑑 𝑢 𝐿 𝑑 𝜌𝑢𝑑 𝜇 𝑘
= 𝑓4 , , , , ,
𝜇𝑢 𝑢 𝑑 𝑑 𝜇 𝜇 𝑑

∆𝑃𝑑 ∆𝑃𝑑 ∆𝑃 𝑢 𝐿 𝑑 𝜌𝑢𝑑 𝜇 𝑘

= = 2 = 𝑓5 , , , , ,
𝜇𝑢 𝜌𝑢𝑑𝑢 𝜌𝑢 𝑢 𝑑 𝑑 𝜇 𝜇 𝑑
 Pipe Friction
We shall now attempt to quantify the effects of pipe friction so we can include
these in our energy equation.
Consider a simple pipe:
L

Mean
Velocity u

The frictional head loss, hL , will depend on the following

L the length of the pipe

d the diameter of the pipe
u the mean velocity of the flow
m the dynamic viscosity of the fluid
 the density of the fluid
k the roughness height
 Thus we have the following functional relationship:

hL = P (u, L, d , m ,  , k )

where P

is an arbitrary function.
At this point it is more convenient to work in terms of pressure drop along the
pipe, rather than head loss, i.e.
p =  (u, L, d , m ,  , k )

We must now establish a relationship between the pressure drop in

the pipe and the 6 parameters identified above.
The most convenient method for deriving a functional relationship is
to use dimensional analysis.

p1 u2 p u2
+  1 1 + z1 = 2 +  2 2 + z 2 + hL
Recall g 2g g 2g

hL =
1
( p1 − p2 ) for developed, turbulent flow with const. diameter
g

or ghL = p1 − p2 = p
 Dimensionless Group Description

p ghL Energy dissipation

= 2
u 2
u
L Pipe length to diameter ratio
d

k Ratio of wall roughness to pipe diameter:

d
“Relative roughness”
Ratio of viscous to inertial forces in the
ud
= Re fluid
m
Re = Reynolds number based on pipe
diameter.

ghL L k ghL L  k
=  2  , Re,  =  3  Re, 
1 2
u d d 1 2 d 
u d
2 2

•From experiments: 2 ghL u 2 is directly proportional to L/d

•Dimensional analysis will not tell us the functional relationship, merely how to plan experiments.
 ghL L  k
=  3  Re,  (*)
1 2 d  d
u
2
The remaining functional relationship is re-written in terms of a new
parameter called the friction factor, f
 k
f = 3  Re, 
 d
 k
We shall define: 4 f =  3  Re, 
 d

 k
f 2 =  3  Re,  US
 d

 k
f1, the UK friction factor 4 f1 = 3  Re,  UK in (*)
 d

ghL L L u2
= 4 f1 or hL = 4 f1 (Darcy’s equation)
1 2 d
u d 2g
2
 k  k
from 4 f1 = 3  Re,  we redefine f1 =  4  Re, 
 d  d
Moody Chart
 Moody Chart
Exercise: Read of the friction factors for the following cases:
a) Re = 1 x 107, k/d = 0.01
b) Re = 3 x 106, k/d = 0.001
c) Re = 4 x 104, k/d = 0.002

K/d
Moody Chart

K/d
CE2702
IX2.1 Water flows through a circular galvanised steel pipe with a velocity
of 2.7 m/s. If the pipe has an internal diameter of 37.5 mm,
calculate the friction head loss in a 20 m section of pipe.
 Moody Chart
Worked example IX2.1: Water flows through a circular galvanised steel pipe with a
velocity of 2.7 m/s. If the pipe has an internal diameter of 37.5 mm, calculate the
friction head loss in a 20 m section of pipe (use 1.01x10-3 kg/ms for the viscosity).
 Moody Chart
Worked example IX2.1: Water flows through a circular galvanised steel pipe with a
velocity of 2.7 m/s. If the pipe has an internal diameter of 37.5 mm, calculate the
friction head loss in a 20 m section of pipe (use 1.01x10-3 kg/ms for the viscosity).

𝜌𝑢𝑑 1000 𝑥 2.7 𝑥 0.0375

First calculate Re = = −3 = 1x105
𝜇 1.01𝑥10

From Moody chart: k = 0.15mm (= 0.00015m)

Therefore k / d = 0.15 / 37.5 = 0.004

Then use Moody chart to obtain the friction factor f

Moody Chart

K/d
 Moody Chart
Worked example IX2.1: Water flows through a circular galvanised steel pipe with a
velocity of 2.7 m/s. If the pipe has an internal diameter of 37.5 mm, calculate the
friction head loss in a 20 m section of pipe (use 1.01x10-3 kg/ms for the viscosity).

𝐿 𝑢2
Apply Darcy’s equation ℎ𝐿 = 4 𝑓
𝑑 2𝑔

4 𝑥 0.0073 𝑥 20 𝑥 2.72
ℎ𝐿 = = 5.79m
37.5 𝑥 10−3 𝑥 2 𝑥 9.81
 Laminar Flow – Re < 2000
This represents only a small region of the Moody Chart.
When laminar flow is present:

16 L u2
f1 = in Darcy’s eq: hL = 4 f1
Re d 2g

Darcy’s Equation may now be written as

Lum
hL = 32
gd 2
Lum
p = 32 2
, where p = ghL
d
This is the Hagen-Poiseuille equation for laminar flow. ( )
Between Re = 2000 and Re = 4000 a transition occurs between laminar flow and
turbulent flow.
Turbulent Flow – Re > 4000
For values of Re above 4000, the flow is assumed to be turbulent, however we may
now distinguish between “smooth” pipes and “rough” pipes:
No pipe has completely smooth walls, however there is an area of the Moody chart
which we may simplify, since a straight line correlation between f1 and log(Re)
is observed at values of Re  105

.
4000 < Re < 105 - SMOOTH WALLED PIPES
In this region, we may write

0.079
f1 =
Re 0.25

•This is Blasius’ formulae and is valid only for “smooth” walled pipes.
•This equation must only be used if

Re  105

AND the relative roughness of the pipe walls is low enough to allow for the
assumption of smooth walls.
4000 < Re – ROUGH WALLED PIPES
This region covers the vast majority of the Moody Chart, and is governed
by the Colebrook-White formula:

1  2k 9.3 
= 3.48 − 1.74 ln  + 
f1  d Re f 1

•This equation is transcendental, so an iterative approach is required to calculate

f1 - best carried out on a computer.

•Hence use the Moody chart in this region.

SUMMARY
FRICTIONAL LOSSES – Darcy’s Equation

LAMINAR FLOW –
Re < 2000
16
f1 =
Re
TURBULENT FLOW

4000 < Re < 105 - Smooth Walled Pipes

0.079
f1 =
Re 0.25

4000 < Re – Rough Walled Pipes

Moody Chart
RX2.2 The friction coefficient is estimated to be f1=0.004 for a straight
pipe of diameter 30 mm and length 30 m. Water flows through
the pipe at 2 litres per second.

a) Taking the pipe to be horizontal, find the pressure drop

across the pipe length.

[Ans: ] p = 64 kPa

IX2.3 Find the Reynolds number of the flow specified in RX2.2. For what
range of volumetric flow rates is it probable that laminar flow will
persist through the length of the pipe?
Sand-Grain Roughness

Representative Values of Roughness

Nature of Surface k in mm

Drawn glass or metal up to 0.0015

Commercial steel or iron 0.05 to 0.3

Concrete 0.3 to 3

Riveted steel 0.9 to 9

 Specific Losses

In addition to energy dissipation at the pipe walls, energy may also be

dissipated when complex (usually turbulent) flows occur.

This may include flow around or through devices such as

i) Valves
ii) Bends
iii) Elbows
iv) Entry
v) Exit
vi) Expansions or contractions

Since this energy dissipation occurs when turbulence is present we may

assume the head loss due to specific losses is given by hL  u 2
u2
or hL = K
2g
where K is the loss coefficient.
 Thus we may add specific losses to Darcy’s equation:

L u2 u2
hL = 4 f1 + K
d 2g 2g

u2  L 
or hL =
2 g 
4 f1
d
+  K 

where K represents the loss coefficients of all the components in the pipe network.

Values for K have been measured experimentally for a large number of

components.
 Loss Coefficients

>d/2 d

K=0.5 K=1

K=1

Abrupt expansion A1
u2
u1
A2

2
 A 
K = 1 − 1 
 A2 
Abrupt contraction

A2
u1
u2
A1

u22
hL = K
2g

where:

d 2 d1 0 0.2 0.4 0.6 0.8 1.0

K 0.5 0.45 0.38 0.28 0.14 0

Note: Loss coefficients apply only between entry and exit planes
Equivalent length
An alternative method of quantifying specific losses is to write them in terms of an
equivalent length of pipe – which will generate the same energy dissipation:

u2 Le u 2 Le … equivalent length
hL = K = 4 f1
2g d 2g

d
where Le = K
4 f1
The head loss is then given by

L u2 Le u 2
hL = 4 f1 + 4 f1
d 2g d 2g

hL = 4 f1
L + Le ] u 2
d 2g
Provided Le has been written in terms of a length of the pipe being studied.

If ΣL ΣLe = 4 then pipe friction is dominant (specific losses not significant)

Specific Losses in Pipe Flows
Distortion K Le/d

45° elbow 0.4 15

90° standard elbow 0.75 30

90° square elbow 1.5 60

Standard tee:
0.5 20
through crosspiece

Standard tee:
1.5 60
through branch

Gate valve:
0.2 8
fully open

Gate valve:
5.0 200
half open
Further Consideration of Specific Loss

Reattac hment

Separation Separation bubble

point w ith recirculation
IX2.4 Water flows from a tank via a pipe 20m long and 50 mm in diameter.
The pipe run changes direction three times through standard 90°
elbows. The friction coefficient is estimated to be f1 = 0.005.

a) Express the head loss for the pipe run as a multiple of the
kinetic head of the flow through it.

b) Check that the specific losses can, in this case, be properly

described as 'minor losses'.

c) Determine the total effective length of this pipe, with allowance

made for the equivalent length added by the specific losses.
Solution to IX2.4 (1)
Note: Assume h to be
h Negligible.

(2)

L u2 u2
hL = 4 f1 + K Darcy’s equation
d 2g 2g (Friction in the pipes and specific losses!)
Specific losses: - reservoir to pipe system
- 3X 90 degrees elbows  K = 3  0.75 + 0.5 = 2.75

u2  L  u2
a) hL =
2 g 
4 f 1
d
+  
K = 10 .75
2g

=8
b) Specific losses are ~1/3 of the frictional losses → marginal case

c) u2
From a) hL = 10.75
2g
u2 Le u 2
Equate to Darcy to obtain Le 10.75 = 4 f1
2g d 2g
d
Le = 10 .75 = 26 .875 m
4 f1