FRICTION LOSSES IN

PIPE FITTINGS AND VALVES

A THESIS
Submitted in partial fulfillment
of the requirements for the Degree
of Master of Science in Chemical Engineering

by
Allen N. Smith

Georgia School of Technology

Atlanta, Georgia
1943

<;;
y ft"

*ty\
Vm
'

FRICTION LOSSES IN

KI

i4

PIPE FITTINGS AND VALVES

Approved:

Date Approved by Chairman ^ 4

ZlZf/3

ACKNOWLEDGMENTS
On the completion of this work I wish to express my
appreciation to Dr. Jesse W. Mason, who suggested and directed
this problem.

I also wish to thank Mr. C A. Mayes and Mr,

R. F. Waters for their assistance in construction of the equipment

TABLE OP CONTENTS
>
Acknowledgments

....,

List of Tables

*........

Introduction *

Literature Survey

1
*...

Purpose

.
,

Theory

1
^.....

Equipment

Comparison with Published Values

36
38

37
37

Suggestions for Further Work and Improvement

,

11
16

Discussion
Curves

6
6

..,

Procedure

BIBLIOGRAPHY

ii
iv

List of Figures

Conclusions .

PAGE

39

40

APPENDIX
Sample Calculations

41

iv

LIST OF TABLES

TABLE

I.
II

PAGE

Straight Pipe

18

Forty-five degree Elbow

20

III. Glob Valve

IV*
V.
VI,
VII,
VIII.

22

Ninety Degree Elbow

Tee (Oil Entering Branch)

26

Gate Valve

28

Tee (Oil Leaving Branch) ,

Cross

Comparison of Values

30
32

IX. Y
X.

24

34

36

LIST OF FIGURES

FIGURE

PAGE

1. Friction Loss Set-up

2.

12

Test Section

3. Centrifugal Pump and Tank .....

14
......

14

15

5. Friction Loss Chart of Straight Pipe and 45 Elbow

19

6. Friction Loss Chart of Globe Valve and 90 Elbow ....

21

7. Friction Loss Chart of Tee and Gate Valve

23

8. Friction Loss Chart of a Tee

25

9. Friction Loss Chart of a Cross

27

4.

Straight Pipe Section

10.

Friction Loss Chart of a Y

29

11.

Graphical Determination of Equivalent Length

31

12.

Friction Loss Chart of a Cross

33

13. Friction Loss Chart of a Y

35

FRICTION LOSSES IN
s

PIPE FITTINGS AND VALVES

INTRODUCTION
Literature Survey
"When a fluid flows through a pipe line consisting of straight
pipe and fittings, there is a definite loss of pressure due to friction,
This loss of head is often considerable and has been investigated many
time s
Prof, F, E. Giesecke (1) was the first to investigate the friction
loss caused by a fitting. He developed certain exponential formulae for
the friction loss of head of water flowing through standard American pipes
and fittings. Briefly summarized, the method of test was to allow water
to flow from one tank to another, with and without the given pipes and
fittings in the connecting line. In this manner, elbows of one-half inch
to three inches nominal size pipe diameters were tested.
D. E. Foster (2) prepared a table tabulating friction losses for
various type fittings ranging from one-half inch to twelve inches in dia~
meter. Foster*s work consisted of the rearrangement of a formula developed

(1) Giesecke, F. E., "Formulas Deduced for Friction Loss in Water

Pipes and Fittings**, Engineering News Record, 79: 469: 1917,
(2) Foster, D. E., "Effect of Fittings on Flow of Fluids",
Mechanical Engineering, 42: 616: 1920,

2
by Meier (3). Giesecke criticized Foster's values, saying that they were
inaccurate and based on mistaken assumptions. He stated that Meier's
factor of resistance "r"., was not a constant for each type of. fitting,
but varied with the diameter, and that Foster assumed a constant value;
thus Foster's values could not be applied to every diameter. Subsequent
investigations by Giesecke confirmed the fact that Foster's calculated
values were not checked by experimental evidence.
Until 1920 comparatively little reliable experimental work had
been performed on the flow of very viscous liquids through commercial
pipes and practically none had touched on the frictional resistance of
fittings in viscous flow. Wilson, McAdams, and Seltzer (4) conducted
many experiments on viscous flow through elbows and the determination
of correction factors for the pressure drop around bends. Elbows of
three sizes, one-inch, two-inch, and four-inch were tested. A manometer measured the pressure drop around the bends, which in each case
consisted of two standard ninety degree elbows connected by a close
nipple with straight pipe leading to the manometer connections which
were forty diameters on each side of the elbows. The one-inch elbows
were separated by a spacer one foot, eight and three-eighths inches in
length, the two inch elbows by a spacer one foot, one inch in length
and the four inch elbows by a spacer one and one-quarter inches in length

(3) Meier, K., The Mechanics of Heating and Ventilating.

(4) Vifilson, McAdams, and Seltzer, "The Flow of Fluids Through
Commercial Pipe Lines11, Journal of Industrial and Engineering Chemistry,
14: 105-19: 1922.

The pressure drop through the straight pipe between manometer connections
(excluding the two elbows) was calculated from the data secured at the
same time on straight pipe alone, and the amount was subtracted from the
total pressure drop around the bend.

This value divided by two gave the

pressure drop for one standard ninety degree elbow*

The accuracy of these

determinations of the pressure drop through elbows is doubtful due to the

extremely short length of the spacer.
Two elbows separated by a spacer according to Corp and Hartwell (5)
may be considered independently if the spacer is at least twenty diameters
in length. When the spacer is shorter, however, the combined loss is less
than the sum of the two elbows plus the friction of the spacer.

It indi-

cates that whenever two elbows are necessary in a pipe line, they should
be installed as close together as possible. Corp and Hartwell also studied the losses in head for U, S, and twisted S pipe bends.

This investi-

gation disclosed that fifty per cent or more of the loss occasioned by an
elbow occurs in the straight pipe below the fittings. Further they determined that losses due to the presence of bends in a pipe line are of two
different origins; first, those due to eddying and disturbances of normal
velocity distribution by virtue of the curvature or change of direction
of flow; secondly, losses due to increased friction and interference because of diameter changes, differences of interior surfaces, unevenness
of pipe connections, and changes in shape or form of passage.

(5) Corp and Hartwell, "Experiments on Loss of Head in U, S,

and Twisted S Bends", University of Wisconsin Engineering Experiment
Station, Series Number 66.

4
Corp and Ruble (6) conducted experiments on the loss of head in
valves and pipes of one-half inch to twelve inches in diameter. Valves,
both globe and gate, were investigated for friotional losses a,t various
openings.
vestigated.

The effect of piezometer rings in a test line also was inTheir results indicated that the loss of head due to valves

and other fittings occurs in part within the valve or fitting and in part
as an added loss in the pipe line downstream where normal flow has been
disturbed.

Measurement of the loss of head where the downstream piezo-

meter is attached too near the valve will give a loss in excess of that
actually produced. From twenty to twenty-five pipe diameters beyondthe
valve will probably give the best position for the downstream piezometer
opening. According to their paper, globe valves offer from fifteen to
forty times the resistance of gate valves for the same size pipe. This
ratio increases with the increase in the size of valves.

The length of

straight pipe of the valve size which will produce the same loss of head
varies from three-quarters of a foot to four feet for fully open gate
valves and from twenty to thirty-five feet for fully open globe valves.
Additional work in the determination of loss of head in standard
elbows and tees was conducted at Lafayette College by L. Perry (7)
Standard one-inch, one and one-half inch, two and one-half inch, and
three inch elbows and tees were used in the four different size pipe
lines.

The discharge was measured by permitting the water to flow into

(6) Corp and Ruble, "Experiments on Loss of Head in Valves and

Pipes of in. to 12 in. Diameter", Mechanical Engineering, 45: 250-51:
1923.
(7) Perry, L., "Tests of Loss of Head in Standard Elbows and
Tees", Engineering News-Record, 92: 940: 1924.

5
a carefully calibrated tank and the velocity calculated.

Overflows were

used to maintain a constant head on the pipe containing the fitting, a

constant head being necessary to maintain a constant velocity1.

The lost

head was measured in each case by making piezometric connections to differential gages. Particular care was taken to have these connections
just flush with the inside of the pipe. These connections were made four
to seven inches above and six to eight inches below the fitting.

The

piezometric connections to the one-inch pipe were made directly to a tee

screwed to a nipple of proper length and in turn screwed into the fitting,
The additional loss of head due to the extra tees to which the piezometric
tubes were connected was determined and proper allowances made. Due to
the proximity of the piezometric connections to the test fitting, Perry's
results do not appear to be very accurate,
Bruins, Othmer, James, and Berman (8) conducted an investigation
to study the relation of the loss of head to velocity at several viscosities in various fittings. Only one-half inch streamline fittings
were tested.

The testing equipment consisted of ten test lines contain-

ing nine different fittings and one of straight pipe alone.

drops were obtained for varying rates of flow.

The pressure

It was concluded that

the pressure loss due to friction in fittings is dependent upon the velocity of flow. Due to the fact that the manometer connections were placed
too close to the test fitting, it is highly probable that these results
are inaccurate,

(8) Bruins, Othmer, James, and Berman, "Friction of Fluids in

Solder-Type Fittings", Transactions of the American Institute of Chemical
Engineers, 36: 721-37: 1940,

6
Purpose
The object of the experimental work described herein was to fill
in the gaps in the existing data on the flow of fluids in pipes of commercial size and roughness in the viscous flow region and in the critical
region between viscous and turbulent flow; and to determine with reasonable accuracy correction factors for the pressure drop through fittings
and valves.
This subject is very important for almost every engineer, operating man and designer alike, has frequent need to determine in advance
the frictional pressure drop that is to be expected when some fluid with
which he is concerned is to be forced or drawn through a pipe line containing various fittings and valves,
From the preceding survey it is noted that, in spite of the fact
that a great deal of work has been conducted on this phase of fluid flow,
there is a considerable variation in the results*

Therefore this investi-

gation was undertaken to check this previous work and establish the correct results.
Since straight pipe data is correlated with Reynolds number it
seemed logical to present losses in fittings in this manner. Therefore,
in design of piping installations the friction loss values in equivalent
feet of various type fittings can be conveniently read from a graph once
the Reynolds number is known*
A series of manometer connections was constructed on each side
of the fitting in order to obtain an accurate measurement of the total
loss caused by the fitting. Five feet of pipe separated the last

manometer connections from any other fittings to prevent a disturbance

in the test line which would render the results inaccurate.

It was be-

lieved that a more reliable determination could be made in this way.

8
Theory
The friction loss of a fluid flowing through a pipe is but a
special case of a general law of the resistance between a solid, and
fluid in relative motion.

If a solid body of any desired shape be

immersed in a fluid stream and the velocity of the fluid past the body
is small in comparison with the velocity of sound, it hes been found
experimentally that the resisting force depends only on the roughness,
size, and shape of the solid and on the velocity, density, and viscosity of the fluid.

Through a consideration of the dimensions of

these quantities it can be shown that?

P/A

p u % *' ( W ; )

d)

where
F * the total resisting force,
A the area of the body.
u the velocity of the fluid past the body,
p - the density of the fluid.
D = the diameter,
jx the viscosity of the fluid*
*

g * the acceleration of gravity,

<)>' some function whose precise form must be determined for
each specific case.

In the particular case of a fluid flowing through a circular

pipe of length L, the total force resisting the flow must equal the
product of the area of contact between the fluid and the pipe wall,

9
and the F/A of equation (l). The pressure drop will equal this product
divided by the cross-sectional area of the pipe, since pressure is
measured in force per unit area.

(9)

*>

A P f F/A(4 LwiyfTD ) * 4 pu2L/gD . $' (Dup/Ji)

(2)

APfD/u2Lp = $ (Dup/^u) =AH f D/u 2 L

(3)

or

where
AP^ s pressure drop due to friction in pounds per square foot*
F/A * resisting force in foot-pounds per square foot of contact
area.
L length of pipe in feet.
D inside diameter of pipe in feet.
p density of fluid in pounds per cubic foot.
u a average velocity of fluid in feet per second.
u viscosity of fluid in English units.
g acceleration of gravity.
A L

loss in head due to friction in feet.

Certain general principles with regard to the flow of fluids

are, of course, generally recognized.

Thus it is well known that

there are two general types of motion for fluids - usually called

(9) Badger and McCabe, Elements of Chemical Engineering, 1935,

p. 35.

10
viscous and turbulent flow. Viscous flow is characteristic of low velocities, small pipes or very viscous liquids. In this form of motion, all
of the particles move in lines parallel to the movement of th mass as a
whole.

In the case of every fluid, as the velocity is increased a point

is eventually reached where a rapid transition takes place from viscous

flow to an eddying, mixing type of motion known as turbulent flow.

The

transition velocity is called the critical velocity.

The Reynolds number (Dup/^u), indicated on the previous pages is
of extreme importance in hydrodynamic discussions, because it correlates
the various factors affecting flow. When the Reynolds number is less
than 2100 in straight pipe, the flow is always viscous, and when above
4000, the flow is always turbulent.
known as the critical region.
in this paper is n Re M .

The range between these values is

The symbol for the Reynolds number used

This investigation was concerned with both the

viscous and critical regions in determining the friction losses in pipe

fittings and valves.

11
EQUIPMENT

The equipment used in the study of friction losses in standard

one-inch galvanized iron pipe fittings consisted essentially of the
following:

a centrifugal pump, a constant head supply tank, a cali-

brated orifice for determining the rate of flow through the pipe and
fittings, a pipe line containing the test fitting, a stretch of straight
pipe, valves for controlling the rates of flow and a manometer for measuring the pressure drop across the lines and fittings. Mercury thermometers were provided to indicate the temperatures of the oil at entrance
and exit,
Only standard one-inch galvanized iron pipe fittings and one-inch
brass globe and gate valves were tested. The fittings tested were:
(1) a ninety degree elbow,
(2) a forty-five degree elbow,
(3) a tee, (oil entering branch),
(4) a tee, (oil leaving branch).
(5) a Y, (oil leaving branch).
(6) a cross, (oil leaving branch).
No data was found in the literature on fittings, (5) and (6),
The test line in which the fitting was tested consisted of manometer connections at one foot intervals for five feet on each side of
the fitting being tested.

The straight pipe was continued about five

feet on each side of the manometer connections to eliminate all effect

of bends. This portion of the test line is illustrated in Figure 2.
By manipulating the proper needle valves, the friction loss between

GcCKv^A Jch^NFRICTION

LOW

P*re:4W^~Vj

or

I acHN&L&frY

EAj>zR>ME*r*L

LAYOUT

^u^Mnrfc-; fc'f; /y/* SMITH J

13
any two points on opposite sides of the test fitting may be determined.
Mercury manometers were used both for velocity determinations and for
measurement of the loss in head due to the test fitting";
In figure 4 is shown the entire section of straight pipe. This
pipe is 78.5 feet in length and consists of six sections of pipe connected by four couplings with a return bend. Since the friction loss
through a coupling is extremely small, an error would not be introduced
by assuming it to be negligible.
The piezometer connections were constructed as follows: at four
equally spaced points on the pipe, one-sixteenth inch holes were drilled
completely through the pipe, A detailed drawing of this construction
appears in Figure 1. After this drilling, some burrs remained on the
inside of the pipe. These burrs were removed by a specially machined
reamer in order to prevent any disturbance to the fluid flow.
A lead to the manometer was constructed by the use of a piece
of on and one-fourth inch pipe one inch in length as a collar around
the drilled holes. A small groove was cut on the inside of the collar
to allow the fluid to flow freely from all four holes and then a hole
one-eighth of an inch in diameter was bored in the collar so as to be
exactly in line with the groove. The collar was brazed to the pipe
and a petcock was inserted into the collar in order that manometer
readings could be taken at the various points without breaking any
connections.

In this manner there was obtained a satisfactory method

of determining the pressure drop at any point along the pipe without
causing any disturbance in the fluid flow.

14

KS
^ H W)\
1

MT^^^MMIM^^^M

Figure 2

Figure 3#

ij^l^BH
'

Test Section

Centrifugal Pump and Tank

15

Figure 4.

Straight Pipe Section

16

PROCEDURE
A series of runs was made at various oil temperatures varying
from low to high rates of flow and vice versa.

The oil was pumped

from the constant head tank through the lines until the entire system
was brought to a state of equilibrium.

The temperature of the oil was

recorded and the velocity obtained from the reading of the manometer
attached to the orifice.

The cock valves at the manometer and pipe

were then opened to flush out the air. All valves were then closed.
Referring to Figure 2, the pair of needle valves marked 1-1
were opened and the manometer reading recorded.

It is obvious that

this value represents the friction loss encountered in the fitting

plus two feet of straight pipe. The losses across 2-2, 3-3, 4-4, and
5-5 were obtained in the same manner.

These losses were determined

at various velocities.
The results obtained were converted into friction loss in feet
of oil and plotted against the feet of pipe.

The resulting straight

line was extrapolated past the zero point until it intersected the
negative abscissa.

The point of intersection gave the equivalent

length of the fitting in feet directly.

The zero point or intersec-

tion of the line with the ordinate gave the friction loss in feet of
oil caused by the fitting.

This method is depicted in the appendix.

These valves determined by the intersection of the negative abscissa

were plotted against Reynolds number. These results were checked by
converting the feet of oil into equivalent lengths through the use of

the straight pipe data on Figure 5.

The graphical and calculated re-

sults are included in the tabulated data. The most convenient method
of expressing the frictional resistance due to an elbow of given si2e
is in terms of the equivalent length of straight pipe which will give
the same resistance to the flow of a given fluid.

It is customary to

express the equivalent length in terms of pipe diameters rather than

feet so that the figure will be substantially the same for all sizes
of pipes.

C 0 - J < 1 0 > 0 i 0 i C n 0 l O l C n ^ ^ ^ W 0 ) M X ^ M H H H G O O O J 0

c * o a c o c o c o r o M M C o r o c o r o M M i ^ M H M M i J M O O O O O O
N K O 0 ) ^ ^ C n ^ W N H O 3 C D - 4 0 } O l ^ 0 ) N H t D W C B C 0 a g

co^roi-0503->o-<o>ci^^04<>iroroi--H-'--'M
, ^ , , ^ .
o o o 5 o o c n o o n > O W M o o < C D O J O > o i ^ o B M a o > c m o <

o o o o oo o o o o o o o o o o o o o o o o o o o o o o
o o o o o o o o o o o o o o o o o o o o p

o o o o o
O O O O O O O H H H N t O M M ^ 5 ^ 5
t o CO W
o o o o o H<o fc3
1
j
eo tf* o> co en
- 3 k O - 3 t o w C D h - r o < o o c o<f k
o rt f^i -ccnncon> i- ^< ii o
^o
c
o5i- oarf^i- roo><JoiO>
o
o o o o
OCO^HM^OJCDaCOMW^SODtD
|_j l_i j _ l !_ |_i

t o w >P> >> en

20

TABLE II
45 Elbow

Oil Temp.
OF
95
98
102
104
106
110
115
120
124
127
130
132
135
138
140
143
146
150
153
156
160
163
166
172
175
178
182
185
188
191
194
198
200
203
206
208
210

Velocity
ft/sec

Reynolds No.

Re

Equivalent
Graph.

Length
Calc.

Head Loss
AH f

1.60
1.70
1.79
1.84
1.92
2.15
2.30
2.42
2.51
2.57
2.60
2.65
2.74
2.82
2.89
2.94
3.00
3.12
3.20
3.28
3.32
3.36
3.44
3.50
3.54
3.56
3.58
3.59
3.62
3.63
3.64
3.66
3.68
3.70
3.71
3.72
3.73

81
95
114
127
142
180
227
278
331
375
407
442
500
560
595
659
729
835
920
1,015
1,150
1,215
1,295
1,470
1,580
1,700
1,820
1,905
2,020
2,140
2,200
2,450
2,540
2,670
2,830
2,920
3,060

0.10
0.14
0.21
0.22
025
0.38
0.41
0.50
0.61
0.70
0.76
0.82
0.91
0.94
0.97
1.12
1.18
1.22
1.40
1.42
1.43
1.48
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49
1.49

0.10
0.15
0.20
0.21
0.22
0.29
0.33
0.40
0.49
0.61
0.67
0.75
0.87
0.90
0.93
1.07
1.11
1.25
1.34
1.46
1.47
1.47
1.47
1.48
1.48
1.48
1.50
1.51
1.51
1.51
1.52
1.54
1.55
1.55
1.55
1.55
1.55

0.05
0.08
0.10
0.10
0.10
0.13
0.15
0.16
0.18
0.21
0.22
0.24
0.25
0.25
0.25
0.27
0.27
0.27
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28
0.28

TABLE I I I
Glob Valve
>.

Oil Temp.
OF
92
95
98
102
105
108
112
115
120
124
127
130
133
136
140
143
146
150
153
156
160
163
166
172
175
178
180
183
187
190
193
197
200
205

Velocity
ft/sec

1.27
1.47
1.58
1.72
1.81
1.93
2.08
2.15
2.30
2.42
2.50
2.53
2.62
2.68
2.76
2.86
2.90
2.99
3.06
3.10
3.17
3.24
3.26
3.34
3.36
3.40
3.41
3.44
3.46
3.48
3.48
3.54
3.56
3.60

Reynolds No,
Re
57
74
89
110
128
150
185
212
263
319
365
400
456
500
570
642
705
802
880
960
1,064
1,175
1,230
1,410
1,502
1,616
1,670
1,770
1,900
2,038
2,143
2,320
2,450
2,680

Equivalent
Graph.

Length
Calc.

Head Loss
A%

1.48
1.48
1.54
1.60
1.73
2.08
2.50
2.90
3.52
3.70
4.00
4.40
., 5.70
6.28
6.65
7.60
8.06
8.40
8.70
9.00
9.32
9.32
9.32
9.41
9.41
9.41
9.41
9.41
9.41
9.41
9.41
9.41
9.41
9.41

1.25
1.26
1.33
1.36
1.51
1.76
2.20
2.62
3.09
3.60
3.95
4.40
5.48
6.00
6.70
7.60
8.12
8.99
9.80
10.61
10,69
10.69
10.69
10.90
11.00
11.00
11.02
11.02
11.08
11.40
11.50
11.29
11.38
11.40

0.60
0.60
0.61
0.S4
0.70
0.79
0.94
1.05
1.19
1.30
1.34
1.40
1.60
1.66
1.72
1.83
1.86
1.90
1.90
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95
1.95

o
hJOOOOUD^CDOoooGo^^^050ior><^cricni^i^^03wrororoi-JH-,^-jo050<o
O 0 3 0 ) M O ( M O - 3 i f > H 0 ) M O ( 3 > O ^ i f , O 0 ) W O 0 5 O 0 ) W O 0 ) 0 l O ( O ( O

o
Hrj

H(-3

<*

O ^ M W O J C l J O J 0 1 W O J O J C l ^ M M O ^ O ^ O ] M O ^ O ^ M 0 1 0 ^ ^ 3 ^ 3 N N N N N N N H H H
0 3 ^ s i ^ j ^ - > 3 - J O ) c n a o j c n o i ^ i f ! > o a M N H O o o s o ) a j o i t ^ o 3 C c o o a ^ j ^
^ O t O ^ C D O > M H H a H W O ( N ( D O H D * > S ( a ' 5 t O N i ^ - 3 H H O H H O ) | - ' 0 ) N

H) <D

" \ 0
W O
<D H-

o d*<
Jtf
03Coror\yrorot\>cv3t\3i-'}-,i-Ji-JMi-JH-,-H
0 < O f f l ^ C n ^ M H O ^ O ( 0 ) ( J 1 l ^ W N H O < O C O J O J O ) 0 1 l ^ 0 3 M ^ 3 N ^ 5 H h j H
^ c n o o c * < o r o o ^ ^ c o w o i i ^ c p r o a } C n ^ c o ^ < ^ a > w ^ o a < i c a a > ^ i - - J c v - M M CO OJ
tO H-
O O O O O O O O O G O i O O O W M O ( O O C J l t O C O l ^ ^ ^ i f i > a ) ' , J i ^ M t O O

O
W !-
<D P09
SB*

o
frj
r o M M M M r o r o r o r o w r o t o r o w t o M c o r o M M M M M M M M O O O O O O O O O
^ w o i ^ c n c n wJ w ^ ,^ J ^ ^ ^ J^ ^ , W M M O C ^ 0 , 5 ^ w O O <
o c n ^ M W M M - j
O O O O O O O O l - M } - - ' l - J | - l - O O M ^ 0 0 0 3 l - 0 0 O 0 0 l - J i - i O O 1 P O J O O O - 0 .

MrororowwroroMrorotocoroMrororoMMMMi^MMOOOOOOOOOO
o^o^c^o^c^wwcncVi^Jktf'^Ik^wWM^to^w^MOto
OOOOOooitoc^030000toO^WM03CDCocno3MMO^<orocncoOiP

Q.O
P p.

^ <$

tr p
-

a>
S$
ct

f
o a>
P

tr

w
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
CO tO <0 to CO CO < D 5 D ^ - 3 - 5 ^ > J ^ - J ^ C 0 ) 0 > W a i H C D O ) 0 1 0 ) O M a ) G > 0

0 3 ^

<*s
WP

-b

IkBLE V
Tee (Oil entering branch)

Velocity
ft/sec

Equivalent
Graph.

Length
Calc.

Head Los
AH f

56
65
90
127
156
185
217
252
292
366
405
495
547
617
676
746
824
955
1,048
1,128
1,205
1,300
1,385

0.20
0.30
0.37
0.40
0.49
0.69
0.71
0.91
1.00
1.20
1.40
1.51
1.70
1.90
2.10
2.41
2.80
2.91
3.10
3.20
3.24
3.38
3.40

0.19
0.22
0.28
0.28
0.39
0.53
0.80
0.80
0.88
1.05
1.26
1.47
1.60
1.69
2.00
2.28
2.66
3.07
3.16
3.34
3.35
3.35
3.35

0.10
0.12
0.15
0.15
0.20
0.26
0.32
0.35
0.37
0.40
0.45
0.46
0.49
0.53
0.55
0.60
0.65
0.65
0.65
0.65
0.65
0.65
0.65

3.57

1,470

3.40

3.35

0.65

3.60
3.62
3.64
3.66
3.69
3.72
3.73
3.76
3.77
3.78
3.79
3.80
3.82
3.83
3.83
3.84

1,580
1,680
1,790
1,882
1,988
2,130
2^260
2,380
2,430
2,525
2,610
2,720
2,785
2,918
3,010
3,082

3.47
3.47
3.47
3.47
3.47
3.47
3.60
3.60
3.60
3.60
3.60
3.60
3.60
3.60
3.60
3.60

3.35
3.35
3.35
3.35
3.35
3.35
3.39
3.39
3.39
3.41
3.41
3.41
3.40
3.40
3.40
3.40

0.65
0.65
0.S5
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.65
0.85
0.65
0.65

1.31
1.45
1.71
1.98
2.11
2.22
2.32
2.42
2.52
2.66
2.72
2.84
2.92
3.00
3.08
3.15
3.22
3.32
3.38
3.42
3.43
3.48
3.54

Reynolds No.
Re

IC
CO

a Jw*
T3
OS
<D

<5

N O ^ O W O O W l O O O W W l O O O O O O O O O O O W y X D l O C t D t O t O ^ t D
W W W ^ ^ W i O C O t ^ a J O O O O r H i H i H i H r H H i ^ r H r H r H i H ^ H H r T i H r ^ i H r H r H
O P O O O O O O O O i H r ^ r H i ^ r H i H r ^ r ^ i H i H i H i H r H r H i H ^ r H r H r H r H i H r - * r H r -

O O P O O O O O O O O O O O O P O O O O O O O O P O O P O O O O O O

.S

to to ^

-P o
ttfiH
d

t~COCOCftOr^CVllOb-OSlO^OCVlDCrHTjCOa3#HtOCDCVie0^si*^^^'#'>l^C-O O O O H H H H H H W N W n n W ^ ^ ^ ^ l O l O l D t O t C i C O t O t D f f i t O W l O t O l f l

-3

O O P O O O O O O O O O O O O O O P O O O O O O O O O O O O O O O O

<D O

-P

<r> XJ

i-H

cri

P4
C*

u
o

>
H

o1
w

lO tO
tCOCftO^OJCsa^OtOCiH^OiHU>OOOiHC\l^CWCV3^C^C^t-^^t^C^^I>-t^
O O O H H H H H C \ J N N W 6 0 ^ < * s t < ^ ^ l O l O l O l O ) < O ^ O t O ) t O < C t O < O t O

O O O P O O O O O O O O P O O O O O O P O O O O O O O O O O O O O O

55
M

x>

I-I

Q
&

I f l W O i O J N C O l O t O i O t O i l X O ^ i J l N ^ W l O N W O C O i a O O O l O O O l O l O O l O O
( 0 - * ^ U 5 ^ 0 ) H [ 0 0 ( J > N U 3 0 t O ( J > ' * G O < O M O ) 0 > < 0 ' * < O S W W s | t l O ( C O H t O O
iH i-H H H N N 6 0 C O t O ^ ' # l O t O t O C > - 0 0 ( J > O H C M W s | ( l O O t - 0 1 0 < M

H H H H H H H H H W N

Ptf

>

-P

H <D
O^ W

o \

t-

inrosHCVj^CNj^OOCSlOOsHC\ltOOOLOCNllOCvJ^s|<CVl^lOC\J^pU5PlOOsHlO^
C n O H N N t O f O O O O H H N W ^ ^ U H O N O C O O J O O H N C M W W ^ ^ ^ t O l O l O

cvjcvicvjcvicvjcvjcvacjcvjcvipjcvicvicocotocotoeotOftoeosototoco

+>

<D < M

>

Q*

S
Q>
e-

H O
H

Cn

S S 2 3

^ 2 5 5

0 6 0 ( D

l : s 0 e 0 ( C

o w ) O w c o o i o t o o t o t o o w ( o o w

C O O f t a i ^ O O O H H H O J N N W t O K ^ ^ ^ i O l O t O C O O t D S N N C n r o C O O C D
F^rHiHiHiHiHiHiHiHiHiH^iHiHiHiHiHiHi^iHi^i-4i-tr-{fHiHiHiHiH

30

TA.BLE VII
Tee ( O i l l e a v i n g branch)
Oil Temp.
op
88
90
96
99
105
108
110
112
116
118
121
125
128
132
135
139
142
145
150
154
156
160
165
170
173
175
180
183
185
186
187
191
197
200
202
205
207
208

Velocity
ft/sec
1.30
1.38
1.62
1.69
1.92
2.04
2.15
2.22
2.36
2.42
2.52
2.63
2.72
2.80
2.93
3.04
3.12
3.18
3.28
3.35
3.39
3.48
3.54
3.60
3.62
3.64
3,68
3.70
3.73
3.73
3.74
3.76
3.77
3.82
3.84
3.86
3.86
3.88

Reynolds No.
Re
51
59
85
99
138
150
161
200
246
266
306
357
408
466
534
625
685
753
880
980
1,058
1,168
1,322
1,452
1,550
1,630
1,808
1,905
1,980
2,005
2,060
2,240
2,500
2,630
2,750
2,878
2,970
3,040

Equivalent
Graph.

Length
Calc.

0.11
0.20
0.30
0.39
0.48
0.48
0.52
0.65
0.75
0.90
0.90
1.20
1.30
1.40
1.49
1.60
1.65
1.81
1.90
2.02
2.16
2.20
2.30
2.40
2.40
2,40
2.40
2.40
2.40
2.40
2.40
2.50
2.50
2.50
2.50
2.50
2.50
2.50

0.10
0.15
0.25
0.31
0.34
0.38
0.39
0.55
0.62
0.76
0.78
1.00
1.12
1.23
1.35
1.51
1.54
1.70
1.98
2.12
2.14
2.19
2.28
2.34
2.34
2.36
2.36
2.34
2.34
2.36
2.36
2.40
2.40
2.44
2.44
2.44
2.44
2.43

Head Loss
AHf
0.05
0.08
0.12
0.15
0.19
0.19
0.20
0.25
0.27
0.31
0.32
0.38
0.40
0.40
0.42
0.42
0.43
0.43
0.43
0.44
0.44
0.45
0.45
0.46
0.46
0.46
0.46
0.46
0.46
0.46
0.46
0.47
0.47
0.47
0.47
0.47
0.47
0.47

32

TA.BLE VIII
Cross

Oil Temp.
op
92
94
97
100
103
106
110
113
116
120
123
126
130
133
136
140
143
146
150
153
156
160
163
166
170
173
176
180
183
186
190
193
196
202

Velocity
f t/seo

Reynolds No.
Re

Equivalent
Graph.

Length
Calc.

1.36
1.39
1.54
1.71
1.84
2.00
2.13
2.26
2.35
2.44
2.53
2.60
2.69
2.78
2.86
2.94
3.02
3.06
3.15
3.22
3.24
3.32
3.38
3.42
3.46
3.48
3.50
3.53
3.55
3.60
3.62
3.63
3.64
3.64

62
68
83
104
124
146
176
209
242
280
324
364
425
483
530
608
677
744
846
925
1,000
1,113
1,222
1,285
1,395
1,488
1,594
1,730
1,850
1,940
2,120
2,240
2,342
2,660

0.20
0.26
0.30
0.40
0.49
0.59
0.65
0.70
0.77
0.88
0.91
1.01
1.10
' 1.19
1.39
1.50
1.61
1.72
2.03
2.17
2.40
2.50
2.50
2.55
2.69
2.69
2.69
2.69
2.69
2.69
2.72
2.72
2.72
2.72

0.21
0.22
0.29
0.32
0.39
0.43
0.47
0.55
0.61
0.69
0.77
0.87
0.96
1.04
1.23
1.40
1.49
1.64
1.99
2.14
2.48
2.51
2.51
2.51
2.58
2.58
2.58
2.60
2.60
2.60
2.60
2.60
2.60
2.61

Head Loss
AHf
0.10
0.10
0.13
0.15
0.19
0.21
0.22
0.25
0.25
0.28
0.30
0.32
0.32
0.32
0.37
0.38
0.40
0.42
0.45
0.45
0.46
0.46
0.46
0.46
0.47
0.47
0.47
0.47
0.47
0.47
0.47
0.47
0.47
0.47

TO

m
o
i-l
'O ^<
W
ai
ft <q
W
.
P o
W H
S j
(D
t-4 o

^ U ) O O O H l O U 5 t O ( 0 0 > O W N ^ I O 3 ^ 0 0 0 > O O O O H H H H H H H H H H H
O O O O H H H H H H N W N W W W N C M N W M W W t O W t O l O n t O e O W W c O l

o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o

tOOU3tDCN3t~CN3C^lOOt*^OtOCOCftCO<DGOU3000^'#lOlOir)lOCOa>0>rHi
O H H H C M C M W C O ^ l O l O C D N O O a c n O N t O W ^ O t O C O t D t O W t D ^ O t D ^ t O W S t

o o o o o o o o o o o o o o o o

<D

rH J :
Ht

O H H W N ^ ^ l O > t D N N O ) O O H ( M t 0 ^ l j O W l O 0 5 < D t O < > < 0 ( 0 0 )

O O O O O O O O O O O O O O i H r H i H i H r H i H i H i H i H r H t H i H r - l i H i - l t H i H i H i H i

$
>

aS

-.
o

O*

o
S3?
0}
t*
H

Q>

rt

t N ( D D H H ^ 0 ) ^ i O l O O i O ^ H N O l O i O O H O O O i O, i O O O O O t 0 0
O 0 0 0 > G W U ) N y j l > W 0 0 C M l O H O N ^ ^ N H ( M C v 3 C 0 H H H s H H N H C M ^ l O W
iHiHrHrHev30atyjW'sl<'s4<LOt0<0^00CJ>OrHCVJ0J#lO>t>-0>OC\at0lOt0t
N N N N N N

&

>>
o
+>
>rt <D
O 0)
O V

+>

l-

<D <VH

cntoevj^o>oc^3>w^cvjtoc\3cvj0^^cNjc30stcoeococa^oOc\ivl<ooow
^ 3 t - N Q O O r l N ^ i n ) t O t > - 0 0 0 ) O O H W N W t 0 ^ ^ l O l O U ) 3 < 0 5 ) S t t
H H H H H N M N N N N C v l N N W t O t O W W W W t O W t O C O W t O W W t O W t O t O l

>

CU
6
o
>
SH

f*4

C M l O O O O ^ N O l O O ^ ^ O N W O W C O O l O W O W O O t O C O O W O O N W O N ^
0>0>0>OOOrHiHCvJcvjCVJtOtOtO'^sJ<^U>lOlrttOOCOt>-{>-ISCOOOOOi3>0>000
<^i^iHrHiHiHiHrHiHiHiHrHiHiHrHr-4rHi-lrHiHiHr-rHrHiHiHrHiHC\lC\IC

36

Comparison with Published Values

Table X

Fittings

Test Value s
(Diameters )
Re a 100 Re 2500

Published Values
(Diameters)
Turbulent
Flow
Ermenc (11) Perry (10)

Forty-five degree elbows,

1.9

18

15

19

Ninety degree elbows*

1.9

30

32

35

Tee (oil leaving branch),

3.5

29

60

46

Tee (oil entering branch).

3.2

39

90

46

Gate valve (full-open).

1.2

Globe valve (full-open).

15.5

132

300

240

Cross (off-run).

3.6

30

46

Y (off-run).

1.9

20

34

(10)

Perry, J . H., l o c . c i t .

(11) Ermenc, E . , " F r i c t i o n Losses i n Pipe F i t t i n g s and Valves 11 ,

Master's Thesis, Georgia School of Technology.

37

DISCUSSION

Curves
The curves obtained are very similar in shape for all the fittings
In the range of Reynolds numbers between 1100 and 3000, the equivalent
length is practically constant.
The forty-five degree elbow and Y curves are very similar and the
losses incurred in these fittings in the viscous flow region are approximately the same. The losses in the ninety degree elbow, cross and tee
where the fluid leaves the branch are approximately the seme over the
total range of Reynolds numbers. In the case of the tee where the fluid
enters the branch, the loss is considerably higher.

The loss through the

gate valve was extremely small, whereas the loss through the globe valve
was considerably higher than any of the fittings tested.

It is readily

seen from the graphs that the equivalent length is a function of Reynolds
number.
The manometer readings constituted the main source of error. This
was especially true in the determination of losses in the globe valve
where a variation in the manometer readings made an average reading necessary.

38
Suggestions for further work and improvement.
Tests should be made on different sized pipe fittings and the
friction losses correlated with the diameter. Additional fittings
should also be tested,
The investigation should extend farther into the critical region
of flow in order to obtain the losses in fittings within that region also

CONCLUSIONS
1

In the viscous flow region the friction losses incurred

in the ninety degree standard radius elbow, cross and tee in which
the fluid leaves the branch are approximately the same,
2.

The equivalent length increases with Reynolds number as

a straight line function.

It increases rather rapidly in the viscous

flow region, but remains approximately constant in lower range of the

critical region.

BIBLIOGRAPHY
Badger and McCabe, Elements of Chemical Engineering, 1935, p. 35.
Bruins, Othmer, James, and Berman, "Friction of Fluids in'Solder-Type
Fittings", Transactions of American Institute of Chemical Engineers, 36: 721-37: 1940,
Corp and Hartwell, "Experiments on Loss of Head in U, S, and Twisted S
Bends", University of Wisconsin Engineering Station Bulletin,
Series Number 66*
Corp and Ruble, "Experiments on Loss of Head in Valves and Pipes of J
inch, to 12 in. Diameter", Mechanical Engineering, 45: 250-51:
1923.
Ermenc, E., "Friction Losses in Pipe Fittings and Valves", Master*s
Thesis, Georgia School of Technology, 1942.
Foster, D. E., "Effect of Fittings on Flow of Fluids", Mechanical
Engineering, 42: 616: 1920.
Giesecke, F. E., "Formulas Deduced for Friction Loss in Water Pipes
and Fittings", Engineering News-Record, 79: 469: 1917.
Meier, K., The Mechanics of Heating and Ventilating*
Perry, J. H., Chemical Engineering Handbook., 1941, p. 825.
Perry, L., "Tests of Loss of Head in Standard Elbows and Tees",
Engineering News-Record, 92: 940: 1924.
Wilson, MoAdams and Seltzer, "The Flow of Fluids Through Commercial
Pipe Lines", Journal of Industrial and Engineering Chemistry,
14: 105-19: 1922.

\:\

APPENDIX

It
t

Sample Calculations
A? AH(dx-d2)
AHj. = ^^(S.G^-S.G.g)
45 Elbow

Run

Connection
Number

Manometer
Right
Left
6.31
6.66
6,98
7.31
7.64

1-1
2-2
3-3
4-4
5-5

5.39
5.04
4.72
4.39
4.06

41%

AE

0.92
1.62
.2.26
2.92
3.58

r
0.775
1.36
1.90
2.46
3.01

0.86
1.46
2.08
2.62
3.20

0.72
1.22
1.75
2.20
2.69

u 2.60 ft./sec.
T 130F

II

6.28
6.88
6.89
7.16
7.45

1-1
2-2
3-3
4-4
5-5

5.42
5.12
4.81
4.54
4.25

u * 2.89 ft./sec.
T = 140P
Manometer used -- mercury-oil.
Inclination of manometer

sin 9 0.728

AH r = AR^ (S.G.X-S. G.2)sin 9

AH

r " -jg- (13.6-0.915)0.728

AHp 0.775 feet of oil

The above data, was plotted on the accompanying graph, from
which the follovdng results are obtained:
I
II
Equivalent Length (ft.)
0.76
0.87
4 H f (feet of oil)

0.22

All results were obtainBd by similar methods.

0.25

42
Sample Calculations (cont.)

The equivalent length values were calculated from

values obtained from Figure 5 on straight pipe.

Hf using the

Reynolds Number * D ^ P Re
F
D
u
p
;u

*
=

1.04n
2.60 ft./sec.
56.0 lb./cu.ft.
46.2 centipoises

Re m 1.04x2.60x56
1 2 x 4 6 ,, 2 x 0 . 0 0 0 6 7 2

Run I

Re * 4 0 7 . 0
AP f D

Figure 5

0, 0043

pu 2 L

0.0043 x 2.60 x 2.60 x 56 - 18.73

pAHf
e
jpr*
- "
- P^
I ;sure
fitting
L

"

56.0x0.22
18.73

ft. pipe
pressure

0#67

ft. pipe
_
fitting

ft

Calculation Method
L 0.67 feet of one-inch galvanized pipe
Graphical Method
L - 0.76 feet of one-inch galvanized pipe

Graph
Run II

Equivalent Length

0.97

Calculated
0.93

U L 9 P W ^ 1 ^ ' 0 W W * ^ W H W * ? T ? 3 W M * *n

1*^1 u n p i i j

IIIU

' ^ T ^ H ^r i

^
^rrfctttt;^
j-rrr-: !;-. n ' u i r

^Lir-'^
'M'-"-

. . . . i^i,- J - J - - J + J f J

i. iL ii- :xttltl - i n : ~*

t* ^T^i - " ' i r i ' v

; + 7 - *

" l~*-r

"> 'It- 1 i'v t -ivUi-i-t--"-^-w*M*"f"t

'

r.t&int'z-Ui.

1M6
' - ",-i'

r ' '

j^

fri:tr.fi

-" ' :

^Tr1'"r't'i

, T--"+
. . ^ ^ _ JJ - rT- T
" T- f ' r T
- 4- J -_ , -

_ i p _]

__ .

' T ' ' T JT X-T..""F " ""yur''

, . . . .

7
. ~ f: ' l i i r ; : r t rryji uyjii^-"^
p*:.^~^
-^I:~:iilH^^-f^
::ll".

::-'

'

L* ""tTTXT- ITTIT 7. " 1~

! - *-. _"T t ^ f t l " ' " 7 7 - " f " ' 1 " 4 1 *" ^- ': 7 - ^ 1 ~~! 1
:1
n
r i l^^w-^-' 1 ' * " j * **-f \~* ' t "! Y'- ' T' 1~*-f-+

y rx r^TTTrr*'

i, r * i I "
1 V| - k "j'

..i 1 4 * , j - "
' T ^ r t r ' j T1" y J-/-' -* *- r

D ~ " "J* i*i 7 - r y - i "- T *: - ' ' , - f - r r ' - i i

(^ *
J. r . . ^ ^ F--,
f
.Tr^^r.j.'
y / ? ^ : L r. *f?^?;^:_.yi li.-'^""-' - - - l rHii ^ :^1'l1V:::': ^ "
"iiLI:
'Lrifi?"

| .. - a r: ^

CTpITTZ-

; ~

. ^ J'L I-i- ["J*feL,"^i-T^!L'~r j . ' , ' ' ', -1+T3 H r '-*" V. 7.1
. ' J "-[ * 1 ' ^ p * " r "
** -

' *"

^ l U . : , : .

_ Li.-i'ij,,^. ^
- ': *

l^t>X
. r " . i ir_ ^ 7 ^ 1 " rT^rr
~" zj " J ^T hi j - j

- ] I TV
-- . - - -j l ^
I *T - - i. ^7-- LJ - -y T, . T""
L
- "l -- -
* --J

-j-j +4^
_ L i , . j ; ...>-. -^-J-n* L ; t ' i :
T
:
f
I.'l"
j
i.2
Ljl"i
C
i* ^^'''i^irt
T-t t " f - T * L S

:,: H ^ - t - M * -j
TJ

1 T J

i^3

f7

z&Pi

ifat^^r^'

KB*?'*'

^j- : to-&;jg

::-k&Bi,