CIVL 3612/9612 FLUID MECHANICS, SEM 1, 2017

MAJOR AND MINOR LOSSES IN PIPE FLOWS

Phyo Paing Soe @ Chandan, Kumar

SID: 450426142
Email: pkum8611@uni.sydney.edu.au

ABSTRACT Next, water was pumped in and the manometer tubes were bled to
This report describes two experiments on the investigation of get rid of any air bubbles trapped inside which could alter the
pressure losses of fluid flow in pipes of different characteristics and measurements of the manometer. The zero positions of the
configurations. manometer levels were then set around the 250-mm mark. The stop
Experiment 1 showed that the pressure loss in pipe flows valve for pipe 1 was opened to allow water to flow through it. The
varied directly to the flow rate and the relative roughness of pipes. head values at the first and the last chambers of the pipe were read
The pressure loss varied inversely to the diameter of pipes. It was from the manometer tubes and recorded. This was done for a set of
also observed that the Blasius and the Prandtl-Colebrook formulae different flow rates. This entire process was then repeated for the
worked more accurately for fluid flows of higher Reynold’s other two remaining pipes (pipe 2 and pipe 3).
Numbers in predicting their theoretical coefficients of friction.
Experiment 2 showed that the pressure loss of fluid flow in 90°
pipe bends and angles are significantly higher than flows in straight Experiment 2: Minor Losses
pipes. In comparison, 90° bends caused lower friction losses than This experiment involved the use of pipe 6 of the machine
90° angles. which had two 90° bends at one location of the pipe and two 90°
angles at another.
To begin with, water temperature was measured. For this
INTRODUCTION experiment, the manometer was connected to all the eight
The aim of the experiments was to investigate the head measuring chambers of pipe 6. Next, the manometer tubes were
variations due to major and minor losses of fluid flow (water) in bled and the zero positions of the manometer were set around the
pipes of different characteristics and configurations. 400-mm mark. The stop valve for pipe 6 was then opened to allow
Most pipe systems consist of considerably more than straight the water to flow through. Finally, the head values at all the eight
pipes. These additional components add to the overall head loss measuring chambers were read and recorded from the manometer
of the system. Such losses are generally termed minor losses, with tubes. Like experiment 1, the head value readings were taken for a
the apparent implication being that the majority of the system loss set of different flow rates of water.
is associated with the friction in the straight portions of the pipes,
the major losses or local losses. (Kudela, 2015)
Experiment 1 involved investigating the major losses, head RESULTS
loss due to friction, and determining the experimental friction
factors of pipes of different materials and diameters for different
flow rates. Experiment 1: Major Losses
On the other hand, Experiment 2 explored the minor losses,
head loss due to pipe components, at different locations on a pipe
Pressure Loss against Flow Rate
at different flow rates. The pipe had a bend arrangement at one
9000
location of the pipe and an angle arrangement at another. Copper Pipe (d=26mm)
8000
Copper Pipe (d=20mm)
7000
EXPERIMENT DESCRIPTION
Pressure Loss Δp (Pa)

6000 Galvanised Steel Pipe

The G.U.N.T. HM 122 Pressure Loss in Pipes Trainer was the (d=16mm)
main equipment used in the experiments. It contained six different 5000
pipes, a flow-rate-controller and manometer tubes to measure the 4000
head values of flow at different locations along the pipes.
3000
2000
Experiment 1: Major Losses 1000
Experiment 1 consisted the use of the first three pipes of the
0
G.U.N.T. machine which had different diameters: 0 0.5 1 1.5 2 2.5
(1) Copper Pipe (D = 26mm)
Flow Rate Q (m3/hr)
(2) Copper Pipe (D = 20mm)
(3) Galvanised Steel Pipe (D = 16mm) Figure 1: Pressure losses in the three pipes at different flow
Firstly, the temperature of water which was to be pumped rates.
through the machine was measured. The manometer was The raw data and calculated results of this experiment are
connected to the first and the last measuring chambers of pipe 1. tabulated in table 1 up to table 5 in Appendix A. The distance, L,

1
 CIVL 3612/9612 FLUID MECHANICS, SEM 1, 2017

between the first and last chambers of the pipes was 2.2m. The Figure 4 illustrates the differential pressures relative to p1 along
head loss, ∆ℎ, along this length was calculated by finding the the length of the pipe. Sharp increases in negative differential
difference between the head values of the first and the last pressure can be seen in the flow between the chamber 3 and 4,
chambers. Sample calculations are shown in Appendix B. Figure 1 where the pipe bends were located. However, this is only true for
shows the pressure drops of the fluid flow in the three pipes at the highest three flow rates. An odd decrease in negative
different flow rates. differential pressure was seen for the lowest two flow rates, 0.4
The graphs show that the galvanized steel pipe (d=16mm) m3/hr and 0.8 m3/hr. In comparison, even sharper jumps in negative
exhibited the highest pressure loss. Among the copper pipes, the differential pressure can be seen between chamber 5 and 6 where
pipe with the smaller diameter (20mm against 26mm) displayed a the pipe angles were located. This was true for all the flow rates
higher pressure loss. Moreover, the graphs also convey that the except the first one, 0.4 m3/hr, where the pressure drop was not
pressure loss increased with increasing flow rate for all the pipes. very significant compared to the pressure drops between other
chambers.

Pressure Loss against Experimental Pipe Friction Factor

9000
Copper Pipe
8000 (d=26mm)
7000 Copper Pipe
Pressure Loss Δp (Pa)

(d=20mm)
6000

5000

4000

3000

2000

1000

0
0.00 0.05 0.10 0.15 0.20
Pipe friction coefficient fexp

Figure 2: Pipe friction factors against pressure loss for the three Figure 4: Differential pressure relative to p1 along the length of pipe
pipes at different flow rates. 6.
The jumps and declines are also evident in Figure 5 below
Figure 2 shows the experimental pipe friction coefficients of which plots differential pressure at the elbow locations against
the pipes at different flow rates. The experimental coefficient of flow rate. Overall, the pressure loss appeared to increase with the
friction was calculated by rearranging the D’Arcy-Weisbach increase in flow rates for both the pipe elbows. Furthermore, the
equation which can be seen in the equation 5 of Appendix B. The pipe angles displayed a greater pressure loss than the pipe bends
graphs show that the friction factor increased as the flow rate for all the tested flow rates.
increased. They also appear to indicate that the pressure loss
decreased with the increasing friction factors. Overall, Galvanised Differential Pressure against Flow Rate
steel pipe produced the highest friction factor followed by the 1.0
copper pipe (d=26mm) which was then followed by the smaller
0.5
copper pipe (d=20mm).
Differential Pressure (kPa)

Figure 3 shows a plot of experimental friction factors of fluid 0.0

flow in the three pipes on the Moody chart against their Reynolds 0 0.5 1 1.5 2 2.5
-0.5
numbers. The calculation for Reynolds number is shown in the
equation 4 of Appendix B. According to the Moody chart, the -1.0
flows in both the copper pipes were found to be smooth whereas -1.5
the flow in the galvanized steel pipe was found to be wholly
turbulent. -2.0

-2.5 2 Pipe Bends (dp4-dp3)

-3.0
Experiment 2: Minor Losses 2 Pipe Angles (dp6-dp5)
The raw data and the calculated results of this experiment are -3.5
tabulated in table 6 up to table 8 in Appendix A. Differential Flow Rate (m3/hr)
pressures at different locations of pipe 6 relative to the pressure at
location 1 were calculated. An example calculation is shown in
Appendix B. In addition, the pressure differences along the 90° Figure 5: Differential Pressure against flow rate for pipe elbow
locations
bends (between chamber 3 and 4) and the 90° angles (between
chamber 5 and 6) were also calculated.

2
 CIVL 3612/9612 FLUID MECHANICS, SEM 1, 2017

Moody Chart representing experimental results from flow experiment 1

Copper pipe (D=26mm)

Copper pipe (d=20mm)
Steel Pipe (d=16mm)

Figure 3: Moody chart representing the experimental friction factors and Reynolds numbers of flow in the three pipes from experiment 1.

Reynold’s Numbers increased. This is because the equations,

DISCUSSION Blasius and Prandtl-Colebrook (Appendix B: equation 6 and 7),
Error Analysis for the results of the experiments is shown in used to calculate fth work more accurately for turbulent flows or
Appendix C. for flows with higher Reynold’s Numbers.
• When the layers of fluid inside the boundary layer slide
over each other in a very disciplined way, the flow is
Main factors for pressure losses in pipes called laminar (Smits, 2018). Generally, laminar flows
Experiment 1 showed that the pressure drops varied directly have Reynold’s Numbers less than 2100.
to the flow rates for all three pipes tested. It also showed that the • In contrast, flows with disordered eddying motions are
pressure loss varied inversely to the pipe diameter as the smaller called turbulent flows. These flows generally have
copper pipe had higher pressure drops than the larger copper pipe. Reynold’s Numbers greater than 4000.
The experiment results fairly agreed to a couple of principles The more turbulent flows had closer predicted friction factor
identified in The Pumping Apparatus Driver/Operator Handbook values to the experimental values. Three assumptions were made
by IFSTA. The second principle in the Handbook mentions that in determining fth:
the friction loss is approximately proportional to the square of the • the flow was steady
fluid flow rate. The third principle states at a constant flow rate, • the pipe had a constant diameter (and)
the friction loss is inversely proportional to the fifth power of the • the flow was fully developed.
diameter of pipe ( Compartment Fire Behavior Training (CFBT) -
US LLC, 2009). This relationship is also evident in the D’Arcy-
Weisbach equation (Appendix B: equation 5). In comparison, the Pipe bends vs. pipe angles
galvanized steel pipe showed much higher friction losses for a Since the pipe bends in experiment 2 showed lower pressure
given flow rate. This was due to its much higher relative drops than the pipe angles for a range of flow rates, pipe bends
roughness compared to the other pipes. The relative roughness should be preferred as a fitting to minimise the friction losses.
values of the three pipes are shown in Table 1 of Appendix A.
Higher relative roughness corresponds to a higher friction factor
according to the Moody Chart or the Prandtl-Colebrook formula CONCLUSION
(Appendix B: equation 7). Consequently, according to the The results from the Major Loss experiment agreed to theory
D’Arcy-Weisbach equation, a higher friction factor corresponds especially to the D’Arcy-Weisbach equation:
to a higher head loss and hence, to a higher friction/pressure loss. • Pressure losses in pipe flow vary directly to the flow rate
Figure 2 appeared to indicate that the pressure loss decreased and the relative roughness.
with the increase in friction factor. However, it is not true because
• Pressure losses vary inversely to the diameter of the
other variables such as, flow rate and pipe diameter, were not pipe.
constant. Furthermore, the theoretical friction factors were found to be
more accurate predictions for flows with higher Reynold’s
Experimental and theoretical friction factors, fexp and fth Numbers.
Table 2 to 5 in Appendix A shows the experimental and On the other hand, Minor Loss experiment showed that
theoretical friction factors and the percentage difference, for all pressure losses of flow in 90° pipe elbows are much higher than
the pipes and flow rates. The difference in the two values was those in straight pipes. In addition, 90° pipe bends should be
very high for flows with small Reynold’s Numbers. It is evident preferred as a fitting compared to 90° angles as the latter causes
from the tables that the percentage difference decreased as higher pressure loss.

3
 CIVL 3612/9612 FLUID MECHANICS, SEM 1, 2017

REFERENCES Q Δp Re Re*ε/D fexp fth %Difference

Compartment Fire Behavior Training (CFBT) - US LLC. (2009). (m3/hr) (Pa) (fexp vs. fth)
Effective and Efficient Fire Streams: Part 3. Retrieved 0.4 538 7666 0.498 0.078 0.0338 130
April 26, 2018, from http://cfbt- 0.8 1174 15331 0.997 0.043 0.0284 50
us.com/wordpress/?m=200912#[.] 1.2 1810 22997 1.495 0.029 0.0257 14
Kudela, H. (2015, January 14). Hydraulic losses in pipes. 1.6 2935 30663 1.993 0.027 0.0239 12
2.0 4206 38328 2.491 0.025 0.0226 8.0
Retrieved from Research Gate:
https://www.researchgate.net/profile/Henryk_Kudela/p
ublication/265535267_Hydraulic_losses_in_pipes/links
/54b68cc30cf24eb34f6d2cfa/Hydraulic-losses-in- Table 4: Calculated results for pipe 3 (Galvanised Steel,
pipes.pdf?origin=publication_detail D=16mm, ɛ=0.1mm)
Smits, A. J. (2018). A Physical Introduction to Fluid Mechanics.
A.J. Smits. Retrieved Nov 3, 2017, from Q Δp Re Re*ε/D fexp fth %Difference
https://www.payscale.com/research/AU/Job=Civil_Eng (m3/hr) (Pa) (fexp vs. fth)
ineer/Salary
0.4 2201 9667 60.420 0.105 0.0319 229
0.6 4597 14501 90.630 0.098 0.0373 161
0.8 6211 19334 120.840 0.074 0.0363 105
APPENDIX A: RAW DATA AND RESULTS 1 8461 24168 151.050 0.065 0.0356 82

Experiment 1: Major Losses Table 5: Densities and kinematic viscosities of water at different
temperatures

Table 1: Water temperature and raw manometer measurements Water Temperature, T Water Density, ρ Kinematic Viscosity, ν
(˚C) (kg/m3) (m2/s)
for pipes 1, 2 and 3.
20 998.2 1.00E-06
Copper Pipe Copper Pipe Galvanised 23.6 997.3 9.31E-07
(D=26mm) (D=20mm) Steel Pipe 24.0 997.2 9.23E-07
(ɛ=0.0013mm) (ɛ=0.0013mm) (D=16mm) 24.4 997.1 9.15E-07
(ɛ=0.1mm) 24.7 997.0 -
30 995.7 8.01 E-07
Water 23.6 24 24.4
Temperature, T
(˚C)

Relative 0.000050 0.000065 0.00625

Experiment 2: Minor Losses
roughness, 𝜺/D
Q (m3/hr) h1 h5 h1 h5 h1 h5
(mm) (mm) (mm) (mm) (mm) (mm)
Table 6: Raw manometer readings for pipe 6 (water temperature,
0.4 320 305 670 615 440 215 T=24.7°C)
0.6 - - - - 690 220
0.8 170 135 600 480 870 235 Q h1 h2 h3 h4 h5 h6 h7 h8
1.0 - - - - 980 115 (m3/hr (mm (mm (mm (mm (mm (mm (mm (mm
1.2 170 110 615 430 - - ) ) ) ) ) ) ) ) )
1.6 120 25 760 460 - - 0.4 465 455 445 470 425 400 380 375
2.0 205 65 925 495 - -
0.8 525 505 385 440 425 350 320 305

Table 2: Calculated results for pipe 1 (Copper, D=26mm, 1.2 615 585 550 465 435 315 275 240
ɛ=0.0013mm) 1.6 765 715 655 510 465 250 180 125

Q Δp Re Re*ε/D fexp fth %Difference 2 895 825 745 550 480 190 85 20
(m3/hr) (Pa) (fexp vs. fth)

0.4 147 5845 0.292 0.079 0.0362 120

0.8 342 11690 0.585 0.046 0.0304 52
1.2 587 17536 0.877 0.035 0.0275 28 Table 7: Pressure losses at different locations in pipe 6 relative to
1.6 929 23381 1.169 0.031 0.0256 23 p1 for different flow rates.
2.0 1370 29226 1.461 0.030 0.0242 23

Pressure Losses Along Pipe Relative to p1

Q Length Along Pipe (m)
Table 3: Calculated results for pipe 2 (Copper, D=20mm, (m3/hr 0.35 0.7 0.9 1.25 1.45 1.8 2.15
) Δp2-1 Δp3-1 Δp4-1 Δp5-1 Δp6-1 Δp7-1 Δp8-1
ɛ=0.0013mm)
(kPa) (kPa) (kPa) (kPa) (kPa) (kPa) (kPa)
0.4 -0.098 -0.196 0.049 -0.391 -0.636 -0.831 -0.880
0.8 -0.196 -1.369 -0.831 -0.978 -1.712 -2.005 -2.152

4
 CIVL 3612/9612 FLUID MECHANICS, SEM 1, 2017

1.2 -0.293 -0.636 -1.467 -1.761 -2.934 -3.325 -3.668

1.6 -0.489 -1.076 -2.494 -2.934 -5.037 -5.722 -6.260
2.0 -0.685 -1.467 -3.374 -4.059 -6.895 -7.922 -8.558
Experimental friction factor, fexp

ℎ𝑓 𝐷 2𝑔
Table 8: Differential pressures at the locations of pipe bends and 𝑓𝑒𝑥𝑝 = × (5)
𝐿 𝑉2
angles for different flow rates.
15 26
× 2 × 9.81
Differential Pressures for Pipe Bends and Pipe Angles 𝑓𝑒𝑥𝑝 = 1000 1000 × = 0.079
2.2 0.212
Flow rate, Q 2 Pipe Bends 2 Pipe Angles
(m3/hr)
Δp4-3 (kPa) Δp6-5 (kPa)
Theoretical friction factor, fth
0.4 0.2445 -0.2445 𝜀
0.8 0.5379 -0.7336 For 𝑅𝑒 ( ) < 65, fth was calculated using Blasius in the
𝐷
1.2 -0.8314 -1.1737 range 2320 < Re < 100000.
1.6 -1.4182 -2.1029
2.0 -1.9073 -2.8364
𝑓𝑡ℎ = 0.3164 𝑅𝑒 −0.25 (6)

𝑓𝑡ℎ = 0.3164 × 5845−0.25 = 0.0362

APPENDIX B: SAMPLE CALCULATIONS

𝜀
Experiment 1: Major Losses For 65 < 𝑅𝑒 ( ) < 1300, fth can be calculated
𝐷
iteratively using the Prandtl-Colebrook formula.

Head loss 2.51 0.27

𝑓𝑡ℎ = [2 𝑙𝑜𝑔10 ( + 𝐷 )]−2 (7)
𝑅𝑒 × √𝑓𝑡ℎ ( )
𝜀

∆ℎ = ℎ1 − ℎ5 (1)
The Blasius equation is first used to a value for an initial guess:
∆ℎ = 320 − 305 = 15
𝑓𝑡ℎ = 0.3164 × 14501−0.25 = 0.0288

Pressure difference Therefore, 𝑓𝑡ℎ = [2 𝑙𝑜𝑔10 (

2.51
+
0.27
)]−2 = 0.0379
16
14501 × √0.0288 ( )
0.1
∆𝒑 = 𝝆𝒈∆𝒉 (2)
2.51 0.27
15 Again, 𝑓𝑡ℎ = [2 𝑙𝑜𝑔10 ( + 16 )]−2 = 0.0373
14501 × √0.0379 ( )
∆𝑝 = 997.3 × 9.81 × = 147 𝑃𝑎 0.1
1000
2.51 0.27
Again, 𝑓𝑡ℎ = [2 𝑙𝑜𝑔10 ( + 16 )]−2 = 0.0373
14501 × √0.0373 ( )
Velocity 0.1

𝑓𝑡ℎ = 0.0373
𝑉 = 𝑄/𝐴 (3)
0.4
3600 Experiment 2: Minor Losses
𝑉= 0.0262
= 0.21 m/s
𝜋×
4

Pressure difference of measuring chamber 2 relative to

Reynolds Number chamber 1, Δp2-1

𝑅=
𝑉𝐷
(4) ∆𝑝2 − 1 = 𝜌𝑔∆ℎ2−1
𝜗

0.21 ×0.026
(455 − 465)
𝑅= = 5845 ∆𝑝2 − 1 = 997.0 × 9.81 × = −97.8 𝑃𝑎
9.31 × 10−7 1000

≈ −0.098 𝑘𝑃𝑎
Re (𝜺/D)

𝜀 0.0013 APPENDIX C: ERROR ANALYSIS INCLUDING SAMPLE

𝑅𝑒 ( ) = 5845 × = 0.292 CALCULATIONS
𝐷 26

5
 CIVL 3612/9612 FLUID MECHANICS, SEM 1, 2017

Error in head loss, ∆h

𝜀ℎ = 2.5 𝑚𝑚

Since ∆h = hm – hn,

2.5 2.5
𝜀∆ℎ = √( )2 + ( )2 = 0.0035 m ≈ 0.004 𝑚
1000 1000

Error in pressure loss, ∆𝒑

∆𝑝 = 𝜌𝑔∆ℎ

Since 𝜌 and g were given i.e., not measured,

𝜀∆ℎ
𝜀∆𝑝 = ∆𝑝 ×
∆ℎ

0.004
𝜀∆𝑝 = 147 × = 35 𝑃𝑎
0.015

Error in flow rate, Q

𝜀𝑄 = 0.1 𝑚3 /ℎ𝑟

Error in experimental friction factor, fexp

ℎ𝑓 𝐷 2𝑔𝐴2 𝜋 2 𝑔 ∆ℎ 𝐷5
𝑓𝑒𝑥𝑝 = × 2 =
𝐿 𝑄 8 𝐿 𝑄2

Since g, D and L were given i.e., not measured,

𝜀∆ℎ 2 𝜀𝑄
𝜀𝑓𝑒𝑥𝑝 = 𝑓𝑒𝑥𝑝 √( ) + (2 )2
∆ℎ 𝑄

0.004 2 0.1 2
𝜀𝑓𝑒𝑥𝑝 = 0.079 √( ) + (2 × ) ≈ 0.044
0.015 0.4

Error in Reynold’s Number, Re

𝑉𝐷 4𝑄
𝑅= =
𝜗 𝜋𝐷𝜗

Since D and 𝜗 were given i.e., not measured,

𝜀𝑄
𝜀𝑅 = 𝑅𝑒
𝑄

0.1
𝜀𝑅 = 5845 × = 1461
0.4