Viscous Flow in Pipes

Fluid Mechanics
 Types of Engineering Problems
• How big does the pipe have to be to carry a
flow of x m3/s?
• What will the pressure in the water
distribution system be when a fire hydrant is
open?
• Can we increase the flow in this old pipe by
adding a smooth liner?
 Viscous Flow in Pipes: Overview
• Boundary Layer Development
• Turbulence
• Velocity Distributions
• Energy Losses
– Major
– Minor
• Solution Techniques
 Laminar and Turbulent Flows

• Reynolds apparatus

VD inertia
Re  
 damping

Transition at Re of 2000
 Boundary layer growth: Transition
length
What does the water near the pipeline wall experience?
_________________________
Drag or shear
Why does the water in the center of the pipeline speed
Conservation of mass
up? _________________________
Non-Uniform Flow

v v v
 Entrance Region Length
le le le
 f  Re   0.06 Re  4.4  Re 
1/ 6

D D D

Distance
for
velocity
profile to
develop

Shear in the
entrance region vs laminar turbulent
shear in long pipes?
 Velocity Distributions

• Turbulence causes transfer of momentum

from center of pipe to fluid closer to the pipe
wall.
• Mixing of fluid (transfer of momentum)
causes the central region of the pipe to have
relatively _______velocity (compared to
uniform
laminar flow)
• Close to the pipe wall, eddies are smaller
(size proportional to distance to the
boundary)
 Log Law for Turbulent, Established Flow,
Velocity Profiles
u yu*
 2.5ln  5.0 Dimensional analysis and measurements
u*  yu*
Valid for  20

0 Turbulence produced by shear!
u* 
 Shear velocity Velocity of large eddies
rough smooth
h d Force balance
0  f
4l
ghf d
u* 
4l
y
f
u*  V
8 u/u
max
 Pipe Flow: The Problem
• We have the control volume energy equation
for pipe flow
• We need to be able to predict the head loss
term.
• We will use the results we obtained using
dimensional analysis
Viscous Flow: Dimensional Analysis

• Remember dimensional analysis?

D  VD  2p
C p  f  , Re 

Where Re  and C p 
l D   V 2

• Two important parameters!

– Re - Laminar or Turbulent
– /D - Rough or Smooth
• Flow geometry
in a bounded region (pipes, rivers): find Cp
– internal _______________________________
– external flow around an immersed object : find Cd
_______________________________
 Pipe Flow Energy Losses

D 
f   C p   f  , Re 
 
Dimensional Analysis
 L D   ghl  p   g z
 2p 2 ghl
Cp   ghl  p Cp  More general
V 2 V2
2 ghf D Assume horizontal flow
f 2
V L
L V2 Always true (laminar or turbulent)
hf  f
D 2g Darcy-Weisbach equation
u*2 L u*2
f=8 2 hf  8
V D 2g
 Friction Factor : Major losses
• Laminar flow
• Turbulent (Smooth, Transition, Rough)
• Colebrook Formula
• Moody diagram
• Swamee-Jain
 Laminar Flow Friction Factor

 gD 2 hl
V Hagen-Poiseuille
32 L
32 LV  D 4  ghl
hf  hf  V Q
 gD 2 128 l
L V2
hf  f Darcy-Weisbach
D 2g

32LV LV2
f
gD 2
D 2g f independent of roughness!
64 64
f  -1 on log-log plot
Slope of ___
VD Re
 Turbulent Flow: hf  f
L V2
D 2g
Smooth, Rough, Transition
• Hydraulically smooth 1  Re f 
 2 log  
pipe law (von Karman, f  2.51 
1930)
• Rough pipe law (von
1  3.7 D 
Karman, 1930)  2 log  
f   
• Transition function for
both smooth and rough
1  D 2.51 
pipe laws (Colebrook)  2 log   
f  3.7 Re f 
f
u*  V
8 (used to draw the Moody diagram)
 Moody Diagram
0.10
0.08
D
f   C p 
0.05
0.04
 l  0.06 0.03
0.05 0.02
friction factor

0.04
0.015
0.01

0.008

0.03
0.006
0.004
D
laminar
0.002

0.02 0.001
0.0008
0.0004
0.0002
0.0001
0.00005
0.01 smooth

1E+03 1E+04 1E+05 1E+06 1E+07 1E+08

Re
 Swamee-Jain
• 1976 0.25
f 2
• limitations    5.74  
– /D < 2 x 10-2 log  3.7 D  Re0.9   no f
  
– Re >3 x 103
– less than 3% deviation
from results obtained Q   D 5 / 2 ghf   L 
log   2.51 
with Moody diagram 2 L  3.7 D 2 ghf D 3 

• easy to program for Colebrook

computer or calculator 0.04
use  4.75
 L  
5.2
1.25  LQ 
2
D  0.66     Q 
9.4
 
 gh
 f   ghf  
Each equation has two terms. Why?
 Pipe roughness
pipe material pipe roughness  (mm)
glass, drawn brass, copper 0.0015
commercial steel or wrought iron 0.045
asphalted cast iron 0.12 
d Must be
galvanized iron 0.15
dimensionless!
cast iron 0.26
concrete 0.18-0.6
rivet steel 0.9-9.0
corrugated metal 45
PVC 0.12
 Solution Techniques
findhead loss given (D, type of pipe, Q)
0.25 8 LQ 2
4Q f hf  f 2
Re     5.74  
2
D  g D 5
log  3.7 D  Re0.9  
  
find flow rate given (head, D, L, type of pipe)

 ghf   L 
Q D5 / 2 log   2.51 
2 L  3.7 D 2 ghf D 3 

find pipe size given (head, type of pipe,L, Q)

0.04
  LQ 2

4.75
 L  
5.2

D  0.66  
1.25
  Q 
9.4
 
 gh
 f   ghf  
 Example: Find a pipe diameter
• The marine pipeline for the Lake Source Cooling
project will be 3.1 km in length, carry a maximum flow
of 2 m3/s, and can withstand a maximum pressure
differential between the inside and outside of the pipe
of 28 kPa. The pipe roughness is 2 mm. What diameter
pipe should be used?
 Minor Losses: Expansions!

• We previously obtained losses through an

expansion using conservation of energy,
momentum, and mass
• Most minor losses can not be obtained
analytically, so they must be measured
• Minor losses are often expressed as a loss
V2
coefficient, K, times the velocity head. hex K
2g
High Re
 2p 2 ghex V2
C p  f  geometry, Re  Cp  Cp  hex  C p
V 2 V 2
2g Venturi
 2
V2
 Ain 
hex  in
1  
2g  Aout 
Sudden Contraction
EGL 2
 1  V2
HGL hc    1 2
1
C  2g
 c 
0.95

Ac 0.9

Cc 
0.85
Cc 0.8
0.75
A2 0.7
0.65
0.6
0 0.2 0.4 0.6 0.8 1
A2/A1
c 2
V1 V2

vena contracta
• Losses are reduced with a gradual contraction
• Equation has same form as expansion equation!
 Entrance Losses
V2
Losses can be he  K e
reduced by K e  1.0 2g
accelerating the flow Estimate based on
gradually and K e  0.5
contraction equations!
eliminating the
vena contracta
K e  0.04
 Head Loss in Bends
High pressure
• Head loss is a function high
of the ratio of the bend low Possible
radius to the pipe n separation
R from wall
diameter (R/D)
• Velocity distribution
2
V
D p    dn   z  C
 R
returns to normal Low pressure
several pipe diameters V2
hb  Kb
downstream 2g
Kb varies from 0.6 - 0.9
 Head Loss in Valves

• Function of valve type and What is V?

valve position V2
hv  K v
• The complex flow path through 2g
valves can result in high head 8Q 2
hv  K v
loss (of course, one of the g 2 D 4
purposes of a valve is to create
head loss when it is not fully
open)
Yes!
Can Kvbe greater than 1? ______
 Solution Techniques
• Neglect minor losses
• Equivalent pipe lengths
• Iterative Techniques
– Using Swamee-Jain equations for D and Q
– Using Swamee-Jain equations for head loss
– Assume a friction factor
• Pipe Network Software
 Solution Technique: Head Loss

• Can be solved explicitly

V2 8Q 2 K
hminor  K
2g
hminor 
g 2
 D4
0.25
4Q f  8 LQ 2
Re     5.74  
2
hf  f
 
D log  
 3.7 D Re 0.9  
g 2 D5


hl   hf   hminor
 Find D or Q
Solution Technique 1
• Assume all head loss is major head loss
• Calculate D or Q using Swamee-Jain equations
• Calculate minor losses
• Find new major losses by subtracting minor
8Q 2
hex  K
losses from total head loss g 2 D 4

hf  hl   hex
0.04
  LQ 
2
4.75
 L  
5.2

D  0.66  1.25    Q 
9.4
 
  ghf   ghf  
 ghf   L 
Q D5 / 2 log   2.51 
2 L  3.7 D 2 ghf D 3 
 Find D or Q
Solution Technique 2: Solver
• Iterative technique
• Solve these equations

0.25 8 LQ 2
4Q f 
Re      5.74 
2
hf  f
D  log 
 3.7 D Re
0.9 

g 2 D5

8Q 2 Use goal seek or Solver to

hminor  K
find discharge that makes the
g 2 D 4
calculated head loss equal
the given head loss.
hl   hf   hminor
Spreadsheet
 Find D or Q
Solution Technique 3: assume f
• The friction factor doesn’t vary greatly
• If Q is known assume f is 0.02, if D is known
 3.7 D 
assume rough pipe law 1
f
 2 log 
  


• Use Darcy Weisbach and minor loss

equations
• Solve for Q or D
• Calculate Re and /D
• Find new f on Moody diagram
• Iterate
 Example: Minor and Major Losses
• Find the maximum dependable flow between the reservoirs
for a water temperature range of 4ºC to 20ºC.

Water 25 m elevation difference in reservoir water levels

Reentrant pipes at reservoirs

Standard elbows
2500 m of 8” PVC pipe
Sudden contraction
Gate valve wide open
1500 m of 6” PVC pipe
Spreadsheet
Directions
 Example (Continued)
• What are the Reynolds numbers in the two
pipes?
90,000 & 125,000 /D= 0.0006, 0.0008
• Where are we on the Moody Diagram? 0.1

0.05
0.04
0.03
0.02

friction factor
0.015
0.01

• What is the effect of temperature?

0.008
0.006
0.004
laminar
0.002
0.001
0.0008

0.0004
0.0002
0.0001
0.00005
0.01 smooth

1E+03 1E+04 1E+05 1E+06 1E+07 1E+08

Re

• Why is the effect of temperature so small?

• What value of K would the valve have to
produce to reduce the discharge by 50%?
140
Spreadsheet
 Example (Continued)
• Were the minor losses negligible?
Yes
• Accuracy of head loss calculations?
5%
• What happens if the roughness increases by a
factor of 10? 0.1

f goes from 0.02 to 0.035

0.05
0.04
0.03
0.02
friction factor

0.015
0.01
0.008
0.006

• If you needed to increase the flow by 30%

0.004
laminar
0.002

0.001
0.0008
0.0004
0.0002
0.0001
0.00005
0.01 smooth

1E+03 1E+04 1E+05 1E+06 1E+07 1E+08

Re

what could you do?

Increase small pipe diameter
 Pipe Flow Summary (1)

• Shear increases _________

linearly with distance
from the center of the pipe (for both
laminar and turbulent flow)
• Laminar flow losses and velocity
distributions can be derived based on
momentum (Navier Stokes) and energy
conservation
• Turbulent flow losses and velocity
distributions require ___________
experimental results
 Pipe Flow Summary (2)
• Energy equation left us with the elusive head loss
term
• Dimensional analysis gave us the form of the head
loss term (pressure coefficient)
• Experiments gave us the relationship between the
pressure coefficient and the geometric parameters
and the Reynolds number (results summarized on
Moody diagram)
 Pipe Flow Summary (3)

• Dimensionally correct equations fit to the

empirical results can be incorporated into
computer or calculator solution techniques
• Minor losses are obtained from the pressure
coefficient based on the fact that the
pressure coefficient is _______
constantat high
Reynolds numbers
• Solutions for discharge or pipe diameter
often require iterative or computer solutions
 Pressure Coefficient for a Venturi
Meter
10
Cp

 2p
Cp 
V 2

1
1E+00 1E+01 1E+02 1E+03 1E+04 1E+05 1E+06

Re
Vl
Re 
0.1

0.05
0.04
0.03
0.02

friction factor
0.015
0.01
0.008
0.006
0.004
laminar
0.002

0.001
0.0008

0.0004
0.0002
0.0001
0.00005
0.01 smooth

1E+03 1E+04 1E+05 1E+06 1E+07 1E+08

Re
 Moody Diagram
0.1

 D 0.05
f  Cp  0.04
 l  0.03
0.02
friction factor

0.015
0.01

0.008
0.006
0.004
D
laminar
0.002
0.001
0.0008
0.0004
0.0002
0.0001
0.00005
0.01 smooth

1E+03 1E+04 1E+05 1E+06 1E+07 1E+08 Minor Losses

Re
 LSC Pipeline

z=0
cs1 cs2
0 Ignore minor losses
p1 V12 p2 V22 KE will be small
 z1  1   z2   2  hl
 2g  2g
Q  2m 3 / s
-2.85 m
28 kPa is equivalent to 2.85 m of water
0.04   106 m 2 / s
  LQ 2 
4.75
 L  
5.2

D  0.66  1.25   Q 9.4   L  3100m

 gh f  
 gh f  
        0.002m

V22 h f  2.85m
D  154
. m V  1.07 m / s  2  0.06 m
2g
 Directions
• Assume fully turbulent (rough pipe law)
– find f from Moody (or from von Karman)
• Find total head loss (draw control volume)
• Solve for Q using symbols (must include minor
losses) (no iteration required)

hl   hf   hminor Solution
0.1

Water
0.05
0.04
0.03
0.02
friction factor

0.015

Pipe roughness
0.01
0.008
0.006
0.004
laminar
0.002
0.001
0.0008

0.0004
0.0002
0.0001
0.00005
0.01 smooth

1E+03 1E+04 1E+05 1E+06 1E+07 1E+08

Re
 Find Q given pipe system
8Q 2 8 LQ 2
hminor  K hf  f
g D
2 4
g 2 D5
hl   hf   hminor

8Q 2  f L   K 
hl 
g 2 

 
 D 5 

  4
D



ghl
Q  Water

  L   K 
8   f 5     4 
  D  D 