Discharge Coefficients Through

Perforated Plates
P. A. KOLODZIE, JR., and MATTHEW VAN WINKLE
University of Texas, Austin, Texas

The design variables that affect the pressure drop across dry perforated plates were
determined and correlated. Average orifice coefficients were calculated for the perforated
plates by using a modified form of the single-orifice flow equation.
The variables which affected the orifice coefficients were found to be the hole diameter,
hole pitch, plate thickness, fraction of the plate covered by the perforated area, and a
Reynolds number based on the hole diameter. The orifice coefficients have been correlated
with these variables in dimensionless groupings.
The correlation presented covers a practicable range of variables for which the pressure
drop may be predicted in design.

which affected the pressure drop across

dry perforated plates.
PREVIOUS INVESTIGATIONS

Some data on dry-perforated-plate pressure drop have already been reported by

various authors (9, 4 to 8) in connection
with other distillation studies. The results
obtained by some of the investigators are
plotted in Figure 1. Table 1 is a compilation

W I C E PLATE

Fig. 1. Some results of previous investigators.

I n the past few years there has been an

increase in industrial use of perforatedplate distillation columns, as for many
services the perforated-plate column has
been found equivalent to, or better than,
the generally used bubble-cap column.
Since the perforated-plate column has
come into general use only recently,
however, there is still a lack of reliable
design information.
One of the important design considerations in a distillation column is the
pressure drop through the column. The
major portion of the pressure drop is that
through the plates. For design purposes
the pressure drop across or through a
plate has generally been considered to be
a combination of two effects: (1) the
pressure drop caused by the liquid head
on the plate and (2) the pressure drop
caused by the discharge of the vapor
through the dry plate. This study was
concerned with the second of these two
effects. An investigation was undertaken
to determine and correlate those variables
P. A. Kolodaie, Jr., is at present with Humble
Oil and Refining Company, Houston, Texas.

Vol. 3, No. 3

Fig. 2. Single-orifice plate in a circular conduit.

TABLE
1
PLATE
DATAOF PREVIOUS
INVESTIGATORS
Plate

Investigator

Hole diam.,
in.

1
2
3
4
5
6
7

0.125
0.1875
0.1875
0.1875
0.1875
0.1875
0.25

8
9
10
11
12
13
14
15
16
17
18
19
20
21

22

A.1.Ch.E. Journal

Free area,
% column

T/d
ratio

P/d
ratio

0.7*
12.7*
3.2*
5.7*
5.7*
5.7*
5.7*

2.00
1.33
1.33
1.33
2.00
0.67
1.00

4 . 0Ot
2.67t
5.33t
4 . OOt
4.00t
4.00t
4.00t

0.038
0.059
0.139
0.250
0.312
0.083
0.177
0.235
0.376
0.039
0.062
0.144
0.249
0.309

4.3
3.9
4.5
4.5
4.1
10.5
12.0
11.9
11.5
18.5
17.1
17.8
17.4
16.2

0.764
0.492
0.209
0.115
0.093
0.350
0.164
0.123
0.077
0.743
0.468
0.201
0.117
0.094

3.30$
3.70t
3.47t
3.48t
3.481
3.27t
2.12t
2.12t
2.16t
1.62$
1.77t
1.74t
1.75t
1.82t

0.125

6.7

0.496

3.00t

Page 305

TABLE
1.-Continued

of the physical data for the perforated

PLATE
OATAOF PREVIOUS
INVESTIGATORS
Plate

Investigator

Hole diam.,
in.

yo column

Free area,

T/d
ratio

Pld
ratio

23
24
25

7 (Lewish)

0.039
0.125
0.033

21.8
23.2
4.82

0.570
0.147
0.498

2.04t
1.98t
4.331

26

7 (Plank and Pike)

0.059

20.0

0.361

2.13t

27

7 (Plank)

0.059

20.0

0,361

2.13t

28
29
30

7 (Bagnoli)

0.125
0.125
0.125

1.22
1.22
2.24

31
32
33
34
35
36
37
38
39
40

0,197
0.197
0.118
0.118
0.0787
0.0787
0.0393
0,0393
0.197
0.0787

1.04
4.59
1.01
4.51
1.30
4.70
1.24
4.64
9.05
1.45

41
42
43
44
45
46
47
48

0.125
0.125
0.250
0.250
0.500
0.500
0.375
0.250

7.50t
9.oot
6. oot
0.060
0.060
0.133
0.133
0.200
0.200
0.400
0,400
3.16
7.90

8 .00t
4 .O O t
s.oot
4 . OOt
8.001
4.00t
8.00t
3 . OOt
3.201
8 .OOt

1 .oo
1.oo
1.00
1.00
1.oo
1.oo
1.oo
1.00

2 oot
4.00t
4.OOt
2 . oot

18.8
4.9
5.4
19.0
4.9
21.5
5.1
9.5

plates used by the various investigators.

(Curves in Figure 1 correspond numerically
to plate numbers in Table 1.)
Baines and Peterson ( 3 ) made a partial
study of perforated barriers some years ago
as part of a n investigation on pressure
drops across screens. They predicted that
the discharge coefficients through the
barriers were a function of a Reynolds
number, solidity ratio, and perforation
pattern.
Mayfield et al. (8) reported on one of the
first investigations related to the design
of perforated-plate columns. A study of
dry-plate pressure drop was made on a
g-in.-diam. column using air. They concluded from their work that the discharge
coefficients through the plates were independent of plate thickness, drilling pattern,
and hole size. It was indicated that the
coefficient did not vary with the Reynolds
number over the ranges covered. Shortly
after the appearance of the work of Msyfield
et al. (8), Arnold et al. (2) reported a
similar study using air in 6- and 15-in.diam. columns. Arnold concluded that the
discharge coefficients were a function of
hole size, free area, and Reynolds number.
These authors reported that an attempt a t
a correlation by use of a modified Reynolds
number was unsuccessful and they proposed
the following correlation:

4.oot

2 . oot
4.00t
3.00t

*Estimated.
t'rriangular.
$Square.

where Ic, a dimensional constant, was

assumed to be dependent only on hole
diameter and free area.

Fig. 4. General layout of the apparatus.

-"I

D
P-HOLE PITCH
D-I.D. OF COLUMN
d - HOLE DIAMETER

Fig. 3. A perforated plate with the holes on an equilateral-triangle

pitch.

Page 3Eb

Fig. 5 . Detail of pressure-tap and flange arrangement.

A.1.Ch.E. Journal

September, 1957

IlS'DlAYl

HOLES

3l3f

MAY HOLES

Recently Hunt et al. (4)reported a study

on a 6-in.-diam. column using air, Freon 12,
carbon dioxide, argon, and methane. These
authors proposed a dry-plate correlation
based on the analogy between a plate perforation and a short tube with entrance
and exit losses. Hunt et d.suggested that
the discharge coefficients were probably

affected by the plate-thickness-to-ho 1

diameter ratio. I n their study, however
this ratio was limited to 1.0.
THEORETICAL CONSIDERATIONS

The number and nature of the variables

involved in a study of pressure drop

TABLE
2
PHYSICAL
DATAFOR PERFORATED
PLATES

VE" DIAM HOLES

5/32*' DIAM HOLES

Hole
diam.,
in.

T/d
ratio

No.
holes

Free area,
% column

Hole pitch to hole diameter = 2.0

%" M A Y HOLES

114" D U H HOLES

Fig. 6. Perforated-plate layouts for a pitchto-hole-diameter ratio of 2.0.

1/16" DIAY HOLES

116" DIAM HOLES

3/32" M Y . HOLES

5/32'' DlAH HOLES

nom

1f4" DIAM HOLES

Fig. 7. Perforated-plate layouts for a pitchto-hole-diameter ratio of 3.0.

Lewish ( 7 ) , reported a study of the

variables affecting pressure drop across
perforated plates, using water in a 1.5-in.diam. column. He also reported some work
done by Bagnoli, Plank, and Plank and
Pike. In general, Lewish verified the
findings of Arnold and in addition concluded
that a Reynolds number based on the
perforation diameter was a definite correlating factor. The variable of plate thickness
was not studied by Lewish. Kamei et al. (6)
reported that discharge coefficients for dry
perforated plates were a function of hole
diameter and plate thickness only he
tested a number of plates of different
design using air in a 5.12-in.-diam. column.

Vol. 3, No. 3

T/d
ratio

No.
holes

Free area

% column

Hole pitch to hole diameter = 3.0

0.0625
0.0938
0.125
0.1563
0.1875
0.250

1.30
0.86
0.65
0.52
0.43
0.33

331
151
85
55
37
19

14.50
14.75
14.75
14.93
14.45
13.20

0.0625
0.0938
0.125
0.1563
0.1875
0.250

1.30
0.86
0.65
0.52
0.43
0.33

207
93
52
33
19
13

9.08
9.08
9.03
8.96
7.42
9.03

0.0625
0.0938
0.125
0.125
0.125
0.125
0.1563
0.1875
0.250

2.00
1.33
1.00
1.00
1.00
1.00
0.80
0.67
0.50

331
151
91
85
61
37
55
37
19

14.50
14.75
15.80
14.75
10.60
6.42
14.93
14.45
13.20

0.0625
0.0938
0.125
0.125
0.125
0.125
0.1563
0.1875
0.250

2.00
1.33
1.00
1.00
1.00
1.00
0.80
0.67
0.50

207
93
52
37
31
19
33
19
13

9.08
9.08
9.03
6.42
5.38
3.30
8.96
7.42
9.03

0.0625
0.0938
0.125
0.1563
0.1876
0.250

3.00
2.00
1.50
1.20
1.00
0.75

331
151
85
55
37
19

14.50
14.75
14.75
14.93
14.45
13.20

0.0625
0.0938
0.125
0.1563
0.1875
0.250

3.00
2.00
1.50
1.20
1.00
0.75

207
93
52
33
19
13

9.08
9.08
9.03
8.96
7.42
9.03

0.0625
0.0938
0.125
0.1563
0.1875
0.250

4.00
2.67
2.00
1.60
1.33
1.OO

331
151
85
55
37
19

14.50
14.75
14.75
14.93
14.45
13.20

0.0625
0.0938
0.125
0.1563
0.1875
0.250

4.00
2.67
2.00
1.60
1.33
1.00

207
93
52
33
19
13

9.08
9.08
9.03
8.96
7.42
9.03

Hole pitch to -ole diameter = 4.0

3/16" OIAM

Hole
diam.,
in.

Hole pitch to hc-2 diameter = 5.0

0.0625
0.0938
0.125
0.1563
0.1875
0.250

1.30
0.86
0.65
0.52
0.43
0.33

121
55
31
19
13
7

5.31
5.37
5.38
5.17
5.07
4.87

0.0625
0.0938
0.125
0.1563
0.1875
0.250

1.30
0.86
0.65
0.52
0.43
0.33

77
37
19
13
7
7

3.37
3.61
3.30
3.35
2.73
4.87

0.0625
0.0938
0.125
0.1563
0.1875
0.250

2.00
1.33
1.00
0.80
0.67
0.50

121
55
31
19
13
7

5.31
5.37
5.38
5.17
5.07
4.87

0.0625
0.0938
0.125
0.1563
0.1875
0.250

2.00
1.33
1.00
0.80
0.67
0.50

77
37
19
13
7
7

3.37
3.61
3.30
3.53
2.73
4.87

0.0625
0.0938
0.125
0.1563
0.1875
0.250

3.00
2.00
1.50
1.20
1.00
0.75

121
55
31
19
13
7

5.31
5.37
5.38
5.17
5.07
4.87

0.0625
0.0938
0.125
0.1563
0.1875
0.250

3.00
2.00
1.50
1.20
1.00
0.75

77
37
19
13
7
7

3,37
3.61
3.30
3.53
2.73
4.87

0.0625
0.0938
0.125
0.1563
0.1875
0.250

4.00
2.67
2.00
1.60
1.33
1.00

121
55
31
19
13
7

5.31
5.37
5.38
5.17
5.07
4.87

0.0625
0.0938
0.125
0.1563
0.1875
0.250

4.00
2.67
2.00
1.60
1.33
1.00

77
37
19
13
7
7

3.37
3.61
3.30
3.53
2.73
4.73

A.1.Ch.E. Journal

Page 307

.........

.........
........

~~.~.f.~.~.~.~.~.~.~.~.~l
\...

(>

12.*.-.*.~.*>,

.
, .......
.'-.-.-.r, .,,
....\

I . . .
\.

1/16" DIAM W E $

0
118" MAM HOLES

I/o

o\I

\o

01

3/52" UAM. WOLES

5/32. UAM. HOLES

-.G
' -!',

Fig. 8. Perforated-plate layouts for a pitchto-hole-diameter ratio of 4.0.

(T)
. . .>,
................
....................\I
',.'. . . . . . .
';' .'. . :
..?V
\-:z*'
4 . .

I..

MAXWUI

1I1..

FREE *RE*.

SQ IN.

Fig. 10. Correlation for obtaining maximum free area for circular plates with holes on
equilateral-triangular pitch.

' '

I n 6 " OlAM HOLES

5/32" DlAY. WOLES

Fig. 9. Perforated-plate layouts for a pitchto-hole-diameter ratio of 5.0.

across perforated plates make a rigorous

theoretical analysis almost impossible.
I n this study a semiempirical approach
was used. Discharge coefficients were
calculated for the perforated plates and
correlated against the physical characteristics of the fluid, column, and plate.
A typical perforated plate is shown in
Figure 3. It was assumed to be composed
of a number of individual orifices acting
independently and in parallel. Although
there is no theoretical basis for this
assumption, it appears logical and allows

Page 308

Fig. 11. Plot of orifice coefficient and Reynolds number for a pitch-to-hole-diameter
ratio of 2.0.

the application, in part, of single-orifice

relations. For a horizontal circular
conduit containing a single circular
orifice (Figure a), a generally accepted
flow equation for a compressible fluid is

For the perforated plate, it was assumed

that the fluid flowing distributed equally
among all holes. For this case the continuity equation can be written as
ADvo =

(NO.holes) (&)(v)
A.1.Ch.E. Journal

A ~ (3)
v

where vD = fluid velocity a t A D

v = fluid velocity a t Ad
Incorporating Equation (3) in the derivation of Equation (2) and altering the
notation will give

Equation (4) is the flow equation for

perforated plates analogous to Equation
(2) for single orifices.
A special A.S.M.E. report ( 1 ) has
shown that the single-orifice coefficient
can be correlated with the physical prop-

September, 1957

Pressure-tap location and shape of the

upstream orifice edge were held constant
in this investigation. The dimensionless
groups in Equation (6) were used as t h e
basis for correlating the experimental

data.
Y

O
-,.

om-

a-

EXPERIMENTAL WORK

060

005

080

075

a70

a65

I
0501

1ooo

I
2000

I
I

4WO

6000

8000

/
10,000

I
2qow

E W O S NUMBER

Fig. 13. Plot of orifice coefficient and Reynolds number for a pitch-to-hole--diameter
ratio of 4.0.

I-

;
Y

u
Y

2000

4000

6000

BOW

10.000

20,000

REYNOLDS NUMBER

Fig. 14. Plot of orifice coefficient and Reynolds number for a pitch-to-hole- -diameter
ratio of 5.0.

erties of t h e conduit, fluid, and orifice

plate through use of certain dimensionless
groups. B y dimensional analysis,

Though not included i n t h e dimensional

analysis, the coefficient has also been
shown t o be a function of pressure-tap
location and shape of t h e upstream
orifice edge.
Based on experimental work and the
A.S.M.E. report ( I ) , the variables t h a t

Vol. 3, No. 3

were considered t o affect the orifice

coefficient for perforated plates were

1. Fluid velocity through t h e orifice v

2. Fluid density p
3. Fluid viscosity p
4. Hole diameter d
5. Hole pitch P
6. Free area of plate A ,

7. Maximum free area possible if total

area of t h e plate were perforated A ,
8. Thickness of t h e plate T
Consideration of these variables i n a
dimensional analysis gave

A.1.Ch.E. Journal

The experimental equipment (Figure 4)

was designed specifically to measure pressure drop across perforated plates a t known
flow rates. Air, the only fluid used, was
obtained from a 100 Ib./sq. in. gauge system
available in the laboratory. Two Fisher and
Porter Flowrators (tube sizes 5 and 8)
connected in parallel were used t o measure
the air flow rate. Depending on the flow
range required one or the other flowrator
was used. The temperature (Weston dialtype thermometer) and pressure (Meriam
standard 30-in. mercury manometer) were
measured at the entrance of each flowrator.
The perforated plates were flanged between
two sections of 3-in. I.D. Pyrex pipe. A
48-in. length of straight pipe was used as
the upstream approach to the plate to
avoid interference from upstream fittings
(9). Pressure taps were located 3 in. upstream and 3 in. downstream from the
plate. Details of the flange arrangement are
shown in Figure 5. Pressure drops acrose
the plates of less than 2 in. of water were
measured a i t h a Meriam draft gauge
calibrated in 0.01-in. divisions. A Meriam
standard 36-in. water manometer was used
to measure pressure drops exceeding 2 in.
of water.
A systematic method was used in making
the plates so that a practical range of
variables might be tested. Pitch-to-hole
diameter ratios of 2.0, 3.0, 4.0 and 5.0 were
selected with all plates being laid out on
an equilateral-triangle pattern. Hole sizes
of 1/16-, 3/32-, 1/8-, 5/32-, 3/16-, and
1/4diam. were used. All holes were drilled
on a vertical drill press with the plates
backed by wooden blocks to prevent excess
burring. After drilling, the plates were
drawn across No. 0 emery paper on a flat
surface to knock off the burrs. Thicknesses
of the plates were 0.081, 0.125, 0.1875, and
0.25 in. Plates with 0.125-, 0.1875-, 0.25-in.
thicknesses were made of red brass and the
0.081-in.-thick plates were made of yellow
brass. All plates were 3.78 in. in diameter.
Combinations of all variables were used
so that for each P/d ratio there were
twenty-four different plates. Six additional
plates were also used in a study of free-area
effects. The physical data on the plates are
listed in Table 2 . Actual layout patterns
for the plates are shown in Figures 6 to
9 and 20. I n laying out the hole patterns
on the plates, it was realized that the
spread of holes over the plate would
probably affect the orifice coefficient (shown
later); therefore, the plates were laid out so
that the holes covered a major portion of
the plate. In this manner the effect of
P/d, plate thickness, and hole diameter
could be studied on the same basis.
The maximum number of holes that can
be fully exposed on a circular plate depends
upon the plate diameter and the hole pitch.

Page 309

The free area will depend on the number and

size of the holes used. A correlation relating
plate diameter, hole pitch, and hole size
to maximum free area is presented in
Figure 10. The correlation was arrived at
by simple geometric relations and will hold
only for an equilateral-triangular hole
pattern.
Several checks were made at the outset
of the investigation to determine the degree
of reproducibility which could be expected
from the equipment and procedures used.
Several runs were made independently on
the same plate to check the consistenc.y
of the data. A comparison of the pressure
drops yielded a maximum difference of
only 3%. A check wits also made to determine how closely the plates themselves
could be reproduced. Identical plates were
tested and a comparison of the pressure
drops yielded a maximum difference of
only 2%. It was concluded, therefore, that
the methods used would produce plates
whirh would give comparable results.
Since the patterns on some of the plates
were not symmetrical, it was necessary to
determine the effect of the relative rotational position of the plates with respect
to the pressure taps. No effect on pressure
drop was found.

PLATE TWCKNESS
HOLE DlAMETER

Fig. 15. Cross plot of Figures 11, 12, 13, and 14 at a Reynolds number of 2,000.

a95

0.90
0.85
I-

EXPERIMENTAL RESULTS

$ 0.60
?! 0.75
U

In the calculations involving the air

flow, the ideal-gas law was used. The air
was considered to be dry and to have
a molecular weight of 29.0. Orifice coeficients were calculated from Equation (4)
after certain simplifications had been
introduced. Since the pressure drop did
not exceed 15 in of water, an average air
density of 0.0731 lb./cu. ft. (temperature
= 95"F., pressure = 775 mm. Hg) was
used. For the small pressure drops that
were encountered, the expansion factor
was found to be nearly unity ( 1 1 ) .
Incompressible flow was assumed. The
average possible error was estimated to
be about 2V0 for the calculated coefficients.
In following through with the dimensional analysis, a Reynolds number based
on the hole diameter was calculated, an
average viscosity for air (temperature =
95F.) being obtained from Perry (10).
For each P/d ratio, the orifice coefficient
was plotted vs. the Reynolds number
(Figures 11 to 14) for various parameters
of T/d ratio. When all data points had
been plotted, the curves became obliterated by the large number of overlapping
points. To enhance the clarity of presentation the points were omitted from
the figures. The curves presented represent the best average of the points
(within 3% maximum deviation). I n
general, the Reynolds number covered a
range from 2,000 to 20,000. The dashed
lines in the figures represent extrapolation
beyond the experimental data. These
extrapolations were based on interpolation of the existing curves and the symmetry which existed for the various P/d
groupings. Beyond a Reynolds number of
4,000 the coefficients became constant.
Page 310

0.70

0.65

0.6C
(

04

0.6

08

LO

12

1.4

1.6

1.6

2.0

22

PLATE THCKNESS
HOLE D W T E R

Fig. 16. Cross plot of Figures 11, 12, 13, and 14 at a Reynolds number of 3,000.

IW

0
U
W

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

22

PLATE THCKNESS
HOLE DlWETER

Fig. 17. Cross plot of Figures 11, 12, 13, and 14 at a Reynolds number of 10,000.

The general shape of the curves representing the smaller T / d ratios is similar
to that of curves of the same type for
the thin-plate (T/d < 0.125) single
orifice.

A.1.Ch.E. Journal

Cross plots were made from Figures

11, 12, 13, and 14 t o determine the effect
of the P/d ratio. The resulting plots were
of coefficient vs. T/d ratio with the P/d
ratio as parameter. Cross plots were
September, 1957

Loo

090

080
I-

y
Y
y

b om

ow

020

025

040

OM

OW

0 50

030
033
HOLE CHAKTER
HOLE PITCH

Fig. 18. Cross plots of Figures 15, 16, and 17 at various plate-thickness-to-hole-diameter
ratios.

105

p i d RATIO OF

zn

P A RATIO OF 30

Fig. 20. Perforated-plate layouts for the

free-area study.

I00

0
I

IL

075

070
02

04

06

08

1
2
14
RATE M l M N E S S
HOLE CHAMTER

10

16

18

20

22

24

Fig. 19. A correlation relating orifice coefficient for perforated plates with the physical
characteristics of the plate and gas.

taken a t Reynolds numbers of 2,000,

3,000, and 10,000 (Figures 15 to 17).
Since the coefficients are constant above
Reynolds numbers of 4,000, Figure 17
will apply for a Reynolds number range
from 4,000 to 20,000 and probably
beyond.
Upon inspection of Figures 15, 16,
and 17, a definite trend was noted between
the curves of the different P/d ratios.
A plot of coefficient vs. of d / P ratio on
log-log paper revealed a series of straight
parallel lines at the various T/d ratios
(Figure 18). These curves can be expressed by the relation

log C

log K

+ a log (g)

(7)

where K = constant (function of T/d)

The family of curves can be represented
bY
0.10

K(g)

Equation (8) can be rearranged to

Vol. 3, No. 3

Since the value of K is constant for any

T/d ratio, the curves in Figures 15, 16,
and 17 may be represented as one curve by
plotting K vs. T/d ratio. Figure 19 is a
plot of K vs. T/d a t Reynolds numbers of
2,000, 3,000, and 4,000 to 20,000.
The correlation presented in Figure 19
was derived by using perforated plates
with holes covering a majority of the
cross-sectional areas available. I n distillation practice, however, chord weirs and
downcomers limit the perforated area
available. Perforated area is the area
effectively covered with holes. I n order
to determine what effect the perforated
area would have, several plates with
different free-area patterns were tested.
The layouts were chosen to simulate
distillation-column plates (Figure 20).
The physical characteristics of these
plates are given in Table 2. The pressure
drops for these plates were found to be
independent of the rotational position of
the pattern in relation to the pressure
taps. Orifice coefficients for these plates
were plotted against the Reynolds
number with the free area as the resulting
parameter (Figure 21).
The fraction of perforated area on the
plates was expressed as a ratio of the

A.1.Ch.E. Journal

actual free area (A,) to the maximum

free area ( A , ) possible if the total area
of the plate were perforated at the same
P/d ratio and pattern. The maximum
free area was determined from the correlation presented in Figure 10. A t any
given A I / A t ratio the A I / A D ratio is a
function only of the P/d ratio. Figure 22
is a cross plot of Figure 21 at a Reynolds
number 5,000. The coefficients were
plotted vs. A,/A, ratio with the P / d
ratio as the parameter. The dashed lines
represent approximations based on a
single data point. The curves shown are
for a T/d ratio of 1.0, but they show the
trends that can probably be expected
at all T/d ratios. From Figure 22 i t was
concluded that the orifice coefficient is
not affected by the distribution of the
holes on a plate as long as the A , / A ,
ratio is greater than 0.60. Since the correlation in Figure 19 is based on A , / A ,
ratios of greater than 0.60, the correlation will apply between A,/A, ratios of
0.60 and 1.00. I n actual practice the
A,/A, ratio usually falls between 0.60
and 0.70.
APPLICATION OF CORRELATION

A generalized correlation is presented

in Figure 19 which can be used for the
design of perforated plates insofar as the
dry-plate pressure drop is concerned. The
correlation covers a wide range of practicable variables which are sufficient to
cover most design needs. Although the
correlation particularly applies to equilateral-triangular hole pattern, it may
prove to be applicable to other arrangements.
The correlation presented was based
only on air, the original supposition
being that the Reynolds number is the
criterion for dynamic similarity and

Page 311

0.90I

0 85
0 80
I-

1,

plates ay not have the same characteristics. Comparisons of these data and
those obtained from commercial types
of perforations (2, 8 ) have been shown
to be in good agreement.

075

NOTATl ON

070
090

A = cross-sectional area
A , = free area of holes on a plate
A , = maximum free area possible on a

085

plate

080

C
d

075

D
0 70
1000

2000

4000
REYNOLDS NUMBER

6000

8000

IQOOO

Fig. 21. Effect of free area on the orifice coefficient at a plate-thickness-to-hole-diameter

ratio of 1.0.

K
P
AP

T
v

Y
a
p

pD
p

= average orifice coefficient

= hole or orifice diameter
= inside diameter of column or

con-

duit
= gravity constant
= constant
= hole pitch
= pressure drop across a perforated
plate
= plate thickness
= average fluid velocity in the
orifice
= average expansion factor for the
fluid
= slope of curves
= fluid density in the orifice
= fluid density at upstream pressure
tap
= fluid viscosity in the orifice

Subscripts

D
1
2
3

ACTUAL FREE AREA

MAXIMUM FREE ARE&

Fig. 22. Cross plot of Figure 21 at a Reynolds number of 5,000.

=
=
=
=
=

orifice or hole
column or conduit
condition upstream from orifice
condition in the orifice
condition downstream from orifice

LITERATURE CITED

1. Am. SOC.Mech. Enrrrs. Suecial Com-

would generalize the correlation for other

gases. Data presented by Hunt et al. (4)
for air, carbon dioxide, methane, Freon
12, and argon were compared with values
predicted from this correlation and found
to have an average deviation of less than
5%. This seems to indicate that the correlation is applicable, in general, to
other gases.
Although the correlation is based on
data obtained from small-scale equipment, the dimensionless groups used in
the correlation should allow application
to larger columns. Scaling up of this type
has been indicated by Mayfield et al. (8).
Arnold et al. (2) also claim good agreement between 6 and 15-in.-diam. columns.
A comparison was made between this
correlation and the data presented by
other investigators using larger columns.
Data reported by Arnold et al. had an
average deviation of less than 3%.
Deviation from the data reported by
Mayfield et al. was as high as 33% in
one case, but generally the average mas
about 10%. Both Hunt and Mayfield
used 6-in. columns and Arnold used
6- and 15-in. columns. With some reservations, it seems reasonable that the
Page 312

correlation presented can be applied to

larger diameter columns.
Certain generalizations can be made
from the correlation about perforatedplate design. As a rule it is desirable to
maintain a low dry-plate pressure drop
across the plate. The following conditions
were found to give the maximum coefficients: (1) T/d ratio of 1.0 or greater,
(2) P/d ratio of 2.0 or less, and (3) A f / A t
ratio from 0.60 to 1.00. For these conditions the optimum Reynolds-number
range was found to be 4,000 to 20,000.
APPLICATION TO DESIGN

OF COLUMNS

Although a Reynolds-number range

from 2,000 to 20,000 has been covered in
this work, the majority of commercial
designs operate in a Reynolds-number
range from 3,500 to 15,000. I n this range
the orifice coefficient becomes independent of Reynolds number and so the
design engineers use of these data is
simplified.
It should be emphasized clearly that
the data and correlations presented here
were obtained from carefully finished
plates and that commercially available

A.1.Ch.E. Journal

mittee on Fluid Meters, Flhd Meters,

Part I, New York (1937).
2. Arnold, D. S., C. A. Plank, and E. M.
Schoenborn, Chem. Eng. Progr., 48,
633 (1952).
3. Baines, W. D., and E. G . Peterson,
Trans. Am. SOC.Mech. Engrs., 73, 467
(1951).
4. Hunt, C. dA., D. N. Hanson, and
C. R. Wilke, A.I.Ch.E. Journal, 1,
441 (1955).
5. Jones, J. B., and C. Pyle, Chem. Eng.
Progr., 51, 424 (1955).
6. Kamei, S., T. Takamatau, S. Miauno,

and Y. Tomisawa, Chem. Eng. (Japan),

18, No. 3, 108 (1954).
7. Lewish, W. T., Prog. Report No. 7 ,
Atomic Enerau Comm.. Microcard OR0
I

142 (1954).

8. Mayfield, F. D., W. L. Church, A. C.

Green, D. C. Lee, and R. W. Rassmussen, Ind. Eng. Chem., 44, 2238

(1952).

9. Perry, J. H., ed., Chemical Engineers

Handbook, 3 ed., p. 407, McGrawHill Book Company, Inc., New York
(1950).
10. Ibid., p. 371.
11. Stearns, R. F., R. J. Russell, R. M.
Jackson; and C. A. Larson, Flow

Measurements with Orifice Meters,

D. Van Nostrand Company, New
York (1951).
September, 1957