Accelerat ing t he world's research.

Evaluation Of Friction Losses In Pipes

And Fittings Of Process Engineering
Plants
Zubaer Rajid

Related papers Download a PDF Pack of t he best relat ed papers 

FRICT ION IN FLUIDS

Emine Bozkuş

Int roduct ion t o Hydraulics and Review: Fundament als of Fluid Mechanics
Abdalla Arbab

LABORAT ORY REPORT HEAD LOSS IN PIPING SYST EM

EL CK
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 4, ISSUE 10, OCTOBER 2015 ISSN 2277-8616

Evaluation Of Friction Losses In Pipes And Fittings

Of Process Engineering Plants
F. W. Ntengwe, M. Chikwa, L. K. Witika

ABSTRACT: The impact of flow rate (Q) on the head loss (hL) in pipes and fittings was studied on different sizes or diameters (D) of pipes fitted with
o
gate valve, 45 and 90 bends using water as process fluid. Diameters of pipes ranged from 25 to 100 mm while the process fluid flow rates ranged fr om
0 to 50 m /h. The Darcy-Weisbach, Hazen-Williams and Poisselli’s methods were used to evaluate friction losses. The results showing increasing D of
3
o o
the pipe and decreasing the hL in the pipe line, gate valve, 45 and 90 elbow, entry and exits to pipes are presented. The results of increasing Q with
increasing exponential values of hL regardless of D of pipe also presented. Therefore, a number of choices can be made between transporting process
fluids using small D pipes (50>D>25 mm) and Reynolds (Re) numbers in the laminar region and large D pipes (100>D>50 mm) using Re numbers in
turbulent regions.

KEYWORDS: Friction-loss, Fittings, Flow-rates, Impact, Process, Engineering, Plants, Fluids-transport

————————————————————

1. INTRODUCTION due to faulty pumps or high friction losses in the delivery

Fluid flow in pipes is continuously impacted by the resistance system. This may be caused by pipe blockage or increased
to flow offered by the roughness of pipe at the walls based on roughness which may contribute to high friction losses or the
the law of similarity [1], [2]. Smooth pipes offer little or system has a high positive delivery head or insufficient net
negligible resistance to flow while rougher surfaces offer positive suction head at the pump. The purpose of the study
increasing resistance depending on the degree of was therefore to establish levels of friction losses in different
roughness.. Such resistance affects flow rate (Q) and sizes of pipes fitted with fittings; gate valve, 45 and 90o
velocity distribution of process fluid in the pipe [3]. The elbows and exit and entry structures in order to determine the
resistance increases for Q values in transition and turbulent head losses that contribute to the increase of equivalent
regions. Studies elsewhere have shown that high velocities lengths of pipes that increase the delivery heads and hence
produce high resistances to flow in pipes and hence hL problems of fluids delivery to desired destinations.
values for particular type of surface roughness [4]. Darcy-
Weisbach, Hazen-Williams, Moody and Fanning showed that 2. CONCEPTS OF FRICTION LOSS
for any flow of fluid in a pipe exhibiting some form of Friction and hL in pipes, bends and valves present a
roughness; head losses (hL) due to friction were produced [4]. challenge, as stated earlier, in the transportation of fluids in
Head losses due to gate valve, 45o and 90o elbows have also pipe-lines of process industry. While the transportation of
been dealt with by other authors [5]. Other losses include fluids is dependent on the quantity of fluid and capacity of the
entry and exit losses. Such losses are commonly added to pump and pipe line, properties of the fluids to be transported
the pipe line in question in order to determine the equivalent are important as these determine whether the fluid will flow
length (Le) of a pipe. In this paper hL were evaluated in five with little or high resistance. The choice of any pipe for a
poly-vinyl-chloride (PVC) pipes with different diameters (D) particular process fluid is based on the correlation between
but same length (L) using water as process fluid at 25oC. A the flow rate and cross-sectional area of pipe as given in
pumping source was used to produce different Qs in each (A1). The D of the pipe can then be determined from (A2) for

d 2
pipe and elevation effects were neglected meaning that the a particular maximum discharge.

A 
pipe line was horizontal. Each loss due to pipe line, gate
Q
valve, 45 and 90o elbows and entry and exit losses were (A1)
correlated to the flow rate and D of pipe in order to establish 4 v
Q 
DC  
the relationships which could be useful to design engineers 0.5

v
and plant operators during design and plant operation. '
(A2)
Decisions could be made early about the size and type of
roughness of the pipe and appropriate Q of the fluid during
design or plant operation based on the correlations presented The hL in pipes can be estimated using Darcy-Weisbach or
in this paper. The challenge with the delivery of fluids is
Hazen-Williams formulas [6, 7] as given in (A3) and (A4) for
either non delivery or insufficient delivery to the desired given Q, L and D of pipe.

hL  C " L
destination. Oftentimes, it is either the insufficient pumping
Q2
_____________________ (A3)
D5
 10.77 L  Q 
hL  4.865  
1.852


F. W. Ntengwe*1, M. Chikwa*1 and L. K. Witika*2

 Chw 
*1 Copperbelt University, School of Mines and (A4)
D

Mineral Sciences, Box 21692, Kitwe, Zambia
*2 University of Zambia, School of Mines, Box
32379, Lusaka, Zambia These equations demonstrate that the hL is directly
proportional to the square of Q and inversely proportional to
the fifth power of diameter. If the D of pipe is changed, the hL
can be evaluated from (A5) provided that the pipe surface
330
IJSTR©2015
www.ijstr.org
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 4, ISSUE 10, OCTOBER 2015 ISSN 2277-8616

roughness and that viscosity of the fluid is the same. The

assumption is that the hL for the first pipe must be known.
D 
hL 2  hL1  1 
5

 D2 
(A5)

In this study the hL values due to each fitting and pipe line
were calculated separately and then correlated to the Q and
D of pipe as required. The head losses were then added to
determine the equivalent lengths of pipes in order to validate
Darcy-Weisbach method and others. If a pump is able to
produce the power that can overcome the losses then the
delivery problem is solved. However, the deliver is at the
expense of friction loss which takes up some energy power
resulting in reduced pumping efficiency.

3.0 EXPERIMENTAL
3.1 Process Evaluation
The process evaluation was carried out using water at 25oC
with ρ of 998.2 kg/m3 and µ of 1.003x10-3 Ns/m2. Four sizes
of pipes of different internal Ds were used as shown in Figure
A1 (a), (b), (c) and (d) and Table A1 with Ds from 25 to 100
mm. The fittings; Gate valve, 45 and 90o elbows were placed
in the pipe line in order to determine their contribution to hL
variation. A single value of Q was allowed into the pipe one
at a time and the characteristics of Q; v, ε/D and Re were
determined in order to establish the values of f which were
used to evaluate K1, hL and hLf using Darcy-Weisbach
formula [6] as given in (A6) and (A7) respectively.

hL  K 1v 2 (A6)

hLf  K 11v 2 (A7)

Moody method [6], [4] for determining the friction factor f from
relative roughness and Re was more convenient than models
developed elsewhere [8], [9], [10], [11]. Friction factor in
curved pipes have also been covered elsewhere [12].
Equation (A8) gave the total hL and Le of pipes. The values
of characteristics are shown in Table A1 for pipes P1 to P4
respectively.

hLt  hLfG  hLf 45  hLf 90  hLfE  hLfEx (A8)

3.2 Correlations
The values of Q were used to determine the correlations with
hL, entry and exit losses, losses in gate valve, 45 and 90o
elbows and the correlation between hL and D of pipe [6]. This
was done in order to establish whether there would be
deviations from Darcy-Weisbach and Hazen-Williams and
Hagen-Poisselli theories [4], [7].

3.3 Data Analysis

Exploration methods were used to analyze data in order
to establish correlations [13], [14], [15], [16]. Excel and
SPSS 11.5 were used as tools to analyze, evaluate and
interpret the data in order to draw meaningful conclusions.

331
IJSTR©2015
www.ijstr.org
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 4, ISSUE 10, OCTOBER 2015 ISSN 2277-8616

Fig. A1: The top view of the pipe line showing the fittings, entry and exit of the pipelines

Table A1: Characteristics of flow in pipes of different sizes

P1 = 25 mm pipe P2 = 50 mm pipe
Q v Re ε /D f v Re ε /D f
(m3/h) m/s - m-1 - m/s - m-1 -
5
10 5.50 1.4x10 5.9x10-5 0.017 1.37 1.0x105
2.95x10-5 0.020
5
20 10.9 2.8x10 5.9x10-5 0.015 2.74 1.4x105
2.95x10-5 0.017
5
30 16.4 4.1x10 5.9x10-5 0.015 4.12 2.1x105
2.95x10-5 0.016
5
40 21.9 5.5x10 5.9x10-5 0.014 5.48 2.8x105
2.95x10-5 0.014
5
50 27.4 6.9x10 5.9x10-5 0.014 6.85 3.5x105
2.95x10-5 0.015

P3 = 75 mm pipe P4 =100 mm pipe

Q v Re ε /D f v Re ε /D f
(m3/h) m/s - m-1 - m/s - m-1 -
4
10 0.61 4.6x10 1.97x10-5 0.015 0.34 3.40x104
1.47x10-5 0.016
4
20 1.22 9.2x10 1.97x10-5 0.016 0.69 6.97x104
1.47x10-5 0.017
5
30 1.83 1.4x10 1.97x10-5 0.017 1.03 1.04x105
1.47x10-5 0.018
5
40 2.44 1.8x10 1.97x10-5 0.018 1.37 1.38x105
1.47x10-5 0.019
5
50 3.05 2.3x10 1.97x10-5 0.021 1.71 1.73x105
1.47x10-5 0.022

332
IJSTR©2015
www.ijstr.org
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 4, ISSUE 10, OCTOBER 2015 ISSN 2277-8616

4.0 RESULTS
4.1 Impact of Flow Rate on Friction Losses in Fittings
The placement of a bend, gate valve, 45 and 90o elbow in a
pipe line introduces extra friction in addition to normal friction
due to the walls of the pipe. A gate valve provides friction to
the flow of the fluid in a pipe. The results in Figure A2 show
that the loss in Q due to the gate valve was the lowest (5 m)
and that due to exit loss was the highest (25 m) at Q of 40
m3/h. Similarly, the entry losses were lower (12.3 m) than
exit losses which were higher (24.5 m) at the same flow rate.
The hL due to 45o angle was higher (20.6 m) than that due to
90o angle which was lower (18.4 m) at the same flow rate.
However, the observed-general trend was that the hL
increased as the values of Q increased regardless of type of
bend and gate valve placed in the pipe line. Fig. A3: Impact of flow on friction loss in gate valve

4.2 Impact of Flow Rate on Friction Loss in Gate Valve

The hL due to gate valve increased from 0.3 - 7.3 m, 0.02 -
0.45 m, 0.001 - 0.09 m and 0.001 - 0.03 m for Q values of 10,
20, 30, 40 and 50 m3/h and D of pipes of 25, 50, 75 and 100
mm respectively (Figure A3). However, the gate valve losses
were higher in the 25 mm pipe and lowest in the 100 mm
pipe. This means that the gate valve hL decreased as the
pipe D increased regardless of the level of Q in the system.
It is important to note also that gate valve hL losses could not
be eliminated in anyway whether the pipe was rough or
smooth but could only be reduced by increasing the D of the
pipe.

4.3 Impact of Flow Rate on Entry Losses

Like gate valve hL, entry losses are fixed by the size of entry
of a pipe. The smaller the D of entry of the pipe the higher
the entry losses and the opposite was true. The results in
Figure A4 show that the entry losses were highest (0.7 m) for Fig. A4: Impact of flow rate on friction loss in entry of pipe
the 25 mm pipe and lowest (0.027 m) for the 100 mm pipe at o
a Q value of 30 m3/h. Simmilarly, the trends at other Q 4.4 Impact of Flow Rate on Friction Loss in 45 Elbow
values indicated that the smaller the diameter of the pipe was A 45o elbow placed in the pipe line introduced a loss in fluid
the higher the entry loss. But it was also observed that when flow. The loss depended on the flow velocity, diameter of the
Q increased from 0 - 50 m3//h the entry losses increased from pipe and flow rate. When the flow increased from 30 - 50
0 - 1.92 m for the 25 mm pipe, 0 - 1.2 m for the 50 mm pipe, m3/h, the friction loss in each pipe increased from 1.2 - 3.2 m
0 - 0.26 m for the 75 mm pipe and 0 - 0.075 m for the 100 for the 25 mm pipe, 0.73 - 2.1 m for the 50 mm pipe, 0.2 -
mm pipe regardless of the size of pipe line. Therefore, entry 0.43 m for the 75 mm pipe and 0.02 - 0.05 m for the 100 mm
losses could only be decreased when the diameter of the pipe (Figure A5). It was also be observed that values for the
pipe increased. 25 mm pipe were higher than those for the 100 mm pipe
meaning that the increase in diameter of the pipe decreased
the losses in the 45o elbow. Therefore increasing the D of
the pipe reduced the hL in the 45o elbow but did not remove it
completely.
o
4.5 Impact of Flow Rate on Friction Loss in 90 Elbow
A 90o elbow placed in the pipe line introduced, also, a loss in
fluid flow. The loss depended on the flow velocity, D of the
pipe and Q similar to that in the 45o elbow. When Q
increased from 0 - 50 m3/h, the hL in each pipe increased
from 0 - 2.87 m for the 25 mm pipe, 0 - 1.8 m for the 50 mm
pipe, 0 - 0.40 m for the 75 mm pipe and 0 - 0.11 m for the
100 mm pipe (Figure A6). It was also observed that values
for the 25 mm pipe were higher than those for the 100 mm
pipe meaning that the increase in diameter of the pipe
decreased the losses in the 90o elbow. Therefore increasing
D of pipe reduced the hL in the 90o elbow, a concept that
Fig. A2: Impact of flow on the friction loss in fittings
333
IJSTR©2015
www.ijstr.org
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 4, ISSUE 10, OCTOBER 2015 ISSN 2277-8616

agrees with Darcy-Weisbach, Hazen-Williams and Hagen-

Poisselli principles [4], [7].

4.6 Impact of Flow Rate on Exit Losses

Like in bends and valves, exit losses increased as Q
increased (Figure A7). An increase from 0 - 50 m3/h
produced an increase in exit losses from 0 - 3.8 m for 25 mm
pipe, 0 - 2.4 m for the 50 mm pipe, 0 - 0.5 m for the 75 mm
pipe and 0 - 0.15 m for the 100 mm pipe. It could also be
observed that the 25 mm pipe produced highest exit losses
than those of 50 mm pipe which were higher than those of 75
mm pipe which were also higher than those produced by the
100 mm pipe. As an example, exit loss for the 25 mm pipe at
40 m3/h was 2.5 m which was higher than that of 50 mm pipe
(1.5 m) which was higher than that of 75 mm pipe (0.3 m)
which was higher than that of 100 mm pipe (0.1 m).
Therefore the exit losses were found to depend directly on
the Q and indirectly on the size of the pipe. This agrees well Fig. A7: Impact of flow rate on exit losses in pipes
with Darcy-Weisbach, Hazen-Williams and Hagen-Poisselli
principles [4]. 25 mm pipe, 0 – 1.12 m for the 50 mm pipe, 0 – 0.2 m for the
75 mm pipe and 0 – 0.04 m for the 100 mm pipe. It could also
4.7 Impact of Flow on Total Head Losses in Pipes be observed that the 25 mm pipe value was higher (33 m)
According to Darcy, the head loss is directly proportional to than that of the 50 mm pipe (1.12 m) which was higher than
the square of velocity and indirectly to the diameter of pipe. the one for the 75 mm pipe (0.2 m) which was higher than
The Q values and cross area of pipe were linked to the that for the 100 mm pipe (0.04 m) at 40m3/h flow rate.
velocity of fluid in the pipe. An increase in the Q produced an Therefore a choice of pipe roughness, Q and diameter of
increase in head loss. The results in Figure A8 showed that pipe was important in attaining the required operating
when Q increased from 40 m3/h, the hL increased from 0 – 33 conditions for the chosen pipe and process fluids.
m for the
4.8 Impact of diameter of pipe on equivalent length
The equivalent lengths of pipes hence total hL were found to
be inversely proportional to the diameter of pipes. This
statement agrees well with Darcy’s formula or method [4]. As
D increased the equivalent length (Le) decreased
correspondingly regardless of the level of flow rate. In this
regard, the 25 mm pipe produced highest Le and highest hL
followed by the 50 mm pipe. The 75 and 100 mm pipes
produced least Le and hL for all the levels of flow rates (Q10
– Q50 m3/h) (Figure A9). The total hL decreased as the D of
the pipe increased. As the D of pipe increased from 25 – 125
mm the hL decreased from 177 – 0 m for the 25 mm pipe, 30
-0 m for the 50 mm pipe, 1.8 – 0 m for the 75 mm pipe and
0.22 -0 m for the 100 mm pipe. It is therefore important to
understand that the transportation of fluids could be carried
out with minimal losses when pipe diameters greater than 75
Fig. A5: Impact of flow on friction loss in 45o elbow mm were used.

5. DISCUSSIONS
Every organization requires that the transportation of process
fluids be efficient, i.e., that the transporting system must not
provide excessive hindrance to the flow of fluid in the pipe.
Any fittings, bends and valves placed in the fluids-transport
system should also offer little resistance to flow of material in
order to maintain high efficiency in transportation and
minimize losses due to friction whether the fluid is a single,
double or triple phase system. For a single phase system,
such as water in this study, it was quite clear that D of pipe,
Q of fluid, type of pipe, bend or valve and the type of entry
and exit played a key role not only in determining the hL but
also contributed significantly to the Le of pipe [17], [18]. It
was observed further that the flow rate cannot be left out
during the design of pipe system because it is linked to the
area and hence the diameter of the pipe.
Fig. A6: Impact of flow on friction loss in 90o elbow
334
IJSTR©2015
www.ijstr.org
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 4, ISSUE 10, OCTOBER 2015 ISSN 2277-8616

The type of bend or elbow to be placed in the pipe line in

It was also observed that different levels of flow of fluid order to develop low hL could depend on the data produced in
produced different levels of hL in pipes, gate valve, 45 and this paper. If a comparison between the 45o elbow and
90o elbows. The defect law applies only to systems that are 90oelbow is made, it can be observed that the 90o elbow
not cylindrical [19]. When Q was doubled the hL quadrupled produced high values of hL than the 45o elbow and therefore
[17], [20]. This was confirmed elsewhere [18]. It means the later would be the best choice [20]. Sinnott (2005) [18]
therefore that appropriate Q is required to be determined showed high values of velocity heads and hence equivalent
early in the design process so as to avoid high hL in pipes. lengths for the 90o elbow and low values for the 45o elbow
The results in this study provide a solution to the choice of thus confirming that the 45o elbow would be the best choice.
the level of Q for particular design of pipe. However, what would dictate is the type of change of
direction of pipe than the choice of elbow required in most
practical situations. It would therefore be important to make
the right choice of elbow for a particular pipe line or trade off
with a larger D pipe in order to lower the losses. Entry and
exit losses cannot be avoided because there is always an
entry and exit point for any process fluid transported through
pipe lines. In this study, entry losses were lower than exit
losses which agreed with the trend given by Sinnot (2005)
[18]. It was also observed that entry and exit losses in pipes
were high for small D pipes (25<D<50 mm) and less for large
D pipes (50<D<100 mm). If one would choose to use small
D pipes, one would require operating the transportation
system with Re numbers in the laminar region in order to
lower the hL due to entry and exit of pipes. It is also
important to note that there are several types of entry
geometry to pipes such as inward projecting, square-edged,
chamfered and rounded inlets with resistance coefficients (K)
Figure A8: Impact of flow rate on head loss in pipes of 1, 0.5, 0.25 and 0.04 respectively. This means that the
inward projecting inlet would produce higher levels of hL than
the square-edged inlet that was used in this study. The K for
the exit of pipe was unity since the pipes released water to a
large-space volume.

6. CONCLUSION
The hL decreased with the increase in D of pipes but
increased with the increase in velocity or Q regardless of size
of pipe. However, exit and entry losses increased with the
increase in Q and increased with the decrease in D of pipe.
The 25 mm pipe produced the highest level of hL than other
Ds of pipes. The fittings; gate valve and elbows were found
to contribute to the hL dependent on D of pipe and the Q
values. Therefore, engineers and operators can make
choices between D of pipe or fitting and Q in order to balance
the characteristics for optimum operating conditions.

Figure A9: Impact of diameter of pipe on head loss in pipe NOMENCLATURE

lines A Area of pipe, A=πd2/4, m2
C1 Constant, C1=(4/π)½
The other key characteristic is the D of the pipe. It has been C11 Constant, C11= 8f/ π2g
shown in this study that when the D of the pipe increased the Chw Hazen-Williams coefficient
hL decreased significantly. The pipes that were greater than D Diameter of pipe, m
75 mm D produced less hL than pipes less than 75 mm in D ε/d Pipe roughness, m-1
for high Reynolds numbers or turbulent flows. Small D pipes ε Absolute roughness, ε =1.56x10-6
(75>D>0) would require Q values with low Reynolds f Friction factor
numbers, i.e., laminar region. White (1929) [21] reported g Acceleration due to gravity, m/s2
laminar conditions for Re values up to 8000 meaning that the hL Head loss, m
introduction of a curved pipe or bend slows lowers the hLf Head loss due to fittings, m
velocity of process fluid. Therefore, a choice can be made k Constant for type of fitting
between one D of pipe and the other in order to obtain the K1 Constant, K1=fL/2dg, s2m-1
optimum D that would provide minimal hL in the pipe line, K11 Constant, K11=k/2g, s2m-1
bends and valves. L Length of pipe, m
Q Fluid flow rate, m3/h
Re Reynolds number, Re=vdρ/µ
335
IJSTR©2015
www.ijstr.org
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 4, ISSUE 10, OCTOBER 2015 ISSN 2277-8616

v Velocity of fluid, v=Q/A, m/s [14] G.E.P. Box, W.G. Hunter and J.S. Hunter, ―Statistics
ρ Density of fluid, kg/m3 for Experimenters: An Introduction to Design, Data
µ Fluid viscosity, mNs/m2 Analysis and Model Building‖, John Wiley & Sons,
New York, 1978.
REFERENCES
[1] T. Stanton and J. Pannell, ―Similarity of Motion in [15] J. Chambers, W. Cleveland, B. Kleiner and T. Tukey,
Relation to Surface Friction of Fluids‖, Phil. Trans. R. ―Graphical Methods for Data Analysis‖, Wadsworth.,
Soc., vol. 214, pp. 199, 1914. 1983.

[2] R.L.Webb, R.G. Eckert and R.J. Goldstein, ―Heat [16] P. Velleman and D. Hoaglin, ―The ADCs of EDA:
Transfer and Friction in Tubes With Repeated-Rib Applications, Basics and Computing of Exploratory
Roughness‖, International Journal of Heat and Mass Data Analysis‖, Duxbury, 1981.
Transfer, vol. 14, no. 4, pp. 601-617, 1971.
[17] B.E. Larock, R.W. Jeppson and G.Z. Watters,
[3] R.R. Rothfus, C.C. Monrad and V.F. Senecal, ―Velocity ―Hydraulics of Pipeline Systems‖, CRC Press, LLC.,
Distribution and Fluid Friction in Smooth-Concentric No. 0-8493-1806-8, 2000.
Annuli‖, Ind. Eng. Chem., vol. 42, no. 12, pp. 2511-
2520, 1950. [18] R.K. Sinnott, ―Chemical Engineering Design‖, 4th Ed.,
Butterworth-Heinemann, Oxford, UK, vol.6, pp. 194-
[4] J.M. Couslson and J.M. Richardson, ―Chemical 207, 2005.
Engineering: Fluid Flow, Heat Transfer and Mass
Transfer‖, vol.1, Butterworth-Heinemann, Oxford, pp. [19] L.B. Ellis and P.N. Joubert, ―Turbulent shear flow in
58-72, 1999. curved duct‖, ―Journal of Fluid Mechanics‖, vol. 62, pp.
65-84, 1974.
[5] K. Sudo, M. Sumida and H. Hibara, ―Experimental
Investigation on Turbulent Flow in a Circular-Sectioned [20] M. Chikwa, ―Design and Analysis of Friction Loss in
90o Bend‖, Experiments in Fluids, vol. 25, pp. 42-49, Piping Systems‖. B. Eng Thesis, ―Unpublished‖,
1998. Copperbelt University, 2012.

[6] D.H. BACON and R.C. STEPHENS, ―Fluid Mechanics [21] C.M. White, ―Stream Line Flow Through Curved
for Technicians‖, TEC3/4, Butterworth & Co Ltd, Pipes‖, Proc. Roy. Soc. Vol. A123, pp. 645, 1929.
Essex, pp. 59-66, 1982.

[7] E.A. Mostafa, ―Correction Factor for Friction Head Loss

Through Lateral and Maniford‖, In: Eighth International
Water Technology Conference, IWTC8, Alexandria,
Egypt, pp. 735-749, 2004.

[8] H. ITO, ―Friction Factors for Turbulent Flow in Curved

Pipes‖, Trans. ASME, J. Basic Eng., vol. 81D, pp. 123-
134, 1959.

[9] R.P. Singh and P. Mishra, ―Friction Factor For

Newtonian and Non-Newtonian Fluid Flow in Curved
Pipes‖, J. Chem. Eng. Japan, vol. 13, pp. 275-280,
1980.

[10] S.K. Das, ―Water Flow Through Helical Coils in

Turbulent Conditions‖, Can J. Chem. Eng., vol. 71, pp.
971-973, 1993.

[11] P.I. Spedding, E. Benard and G.M. McNally, ―Fluid

Flow Through 90o Bends‖, Dev. Chem. Eng. Min.
Process, vol. 12, pp. 107-128, 2004.

[12] J. Lohrenz and F. Kurata, ―A Friction Factor Plot for

Smooth Circular Conduits, Concentric Annuli and
Parallel Plates‖, Ind. Eng. Chemistry, vol. 52, no. 8, pp.
703-706, 1960.

[13] J. Turkey, ―Exploratory Data Analysis‖, Addison-

Wesley, 1977.

336
IJSTR©2015
www.ijstr.org