Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

www.springerlink.com/content/1738-494x

DOI 10.1007/s12206-011-0706-3

A general structural design methodology for multi-hole orifices and its

experimental application
Tianyi Zhao*, Jili Zhang and Liangdong Ma
School of Civil Engineering, Faculty of Infrastructure Engineering, Dalian University of Technology, Dalian, China
(Manuscript Received August 20, 2010; Revised March 7, 2011; Accepted May 27, 2011)
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Abstract
Compared to single-hole orifices (SOs), multi-hole orifices (MOs) have smaller orifice sizes and various patterns of orifice distribution.
The geometric description of MOs is more complex, increasing the difficulty of MO structural design. Therefore, it is worthwhile to investigate the key factors affecting MO throttle or flow control characteristics and to develop a general MO design method. This work
presents a practical geometric design methodology for MOs and applies this procedure in throttle experiments. To describe the MO geometry in detail, the methodology first introduces a comprehensive set of geometric architectures involving orifice arrangement criteria
and geometric parameters such as the total orifice number, n; the orifice distribution density, Dd; and the equivalent diameter ratio, EDR.
Then, a series of throttle tests in water flow are conducted to investigate the effect of various geometric features on the pressure loss characteristics of MOs. Finally, a simple model to calculate the pressure loss coefficient of MOs is presented.
Keywords: Multi-hole orifice; Pressure loss coefficient; Structural design methodology; Throttle test
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

1. Introduction
Orifice plates are widely used as throttle fittings in refrigeration and heat pump systems. Fig. 1 shows an electric expansion valve that uses circular orifices as throttle devices.
The orifices are installed at the bottom of the expansion valve.
Liquid refrigerant flows through the orifices and flashes to gas
because of the pressure drop created by the orifices. Refrigerant mass flow is regulated by the motor-driven screw stem.
Fig. 2 shows an integrated evaporator condenser in which
the condenser and evaporator are separated by the thermal
baffle and the orifice plate that throttle the refrigerant drawn
from the condenser into the evaporator. The orifice can also
regulate refrigerant mass flow when the difference between
the evaporate and condensate pressures changes.
Fig. 3 describes the configuration of a two-stage MO fitting
applied in some refrigeration systems.
As Fig. 3(a) shows, the flow of liquid refrigerant through
the primary orifice occurs without a phase change because the
refrigerant generally has a higher entry pressure in the full
load condition. After the primary throttle process, the refrigerant passes through the secondary MO and flashes to gas.
In Fig. 3(b), the phase change of the liquid refrigerant oc

This paper was recommended for publication in revised form by Associate Editor
Simon Song
*
Corresponding author. Tel.: +86 411 84707753, Fax.: +86 411 84707753
E-mail address: sebastian.zhao@gmail.com
KSME & Springer 2011

Fig. 1. Typical electric expansion valve.

curs through the primary MO because the refrigerant experiences a pressure decrease in the partial load condition. As a
consequence, the gas-liquid refrigerant flows across the secondary MO, and the gas lock phenomenon occurs at the micro-size perforated holes, which decreases the refrigerant mass
flow entering the evaporator and induces a flow adaptation
effect on the throttle fitting.
The removable design shown in Fig. 3(c) allows the pres-

2238

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

(a) Full load condition

Fig. 2. Integrated evaporator condenser.

sure loss coefficient of the integrated throttle fitting to be adjusted by changing the position of the movable MO. This is
another novel way to realize adaptive flow regulation.
According to the above illustration, MOs appear to have
various potential applications because of their convenience
and low cost, but more work is needed to realize these applications. The uncertain orifice arrangement and complex geometric features of MOs complicate their structural description and
design, greatly restricting the application of MOs. To solve
this problem, a general MO structural design methodology
needs to be investigated, and a simple MO pressure loss
model needs to be further studied.
With particular reference to orifice geometry and pressure
loss characteristics, in his early study, Anderson [1] related the
dependence of the primary frequency on differential pressure
to the orifice geometry with a circular-orifice number. The
acoustic effects of cavitation were discussed by testing a circular-centered single-hole orifice and a multi-hole orifice by P.
Testud et al. [2]. The pressure loss coefficients of squareedged orifices and perforated plates were determined by Guohui Gan et al. [3]. Some researchers have studied two-phase
flows: M. Fossa et al. [4] investigated two-phase flow pressure
drop and void fraction profiles with respect to the effect of the
area contraction ratios and orifice thicknesses. G. Kojasoy
conducted a similar study on two-phase pressure drops in
multiple thick and thin orifice plates [5]. Saadawi et al. conducted experimental investigations of two-phase flows across
orifices in large diameter pipes [6].
As shown by the above literature review, few studies have
investigated the structural design issue of orifices, especially
MOs. Even in the few studies involving MOs, a comprehensive set of geometric orifice parameters and their influences
on the MO pressure loss coefficient were not sufficiently addressed.
This paper presents a novel and general methodology for
MO geometric design, and this method was applied in a series
of throttle tests under single-phase conditions. A simple pres-

(b) Partial load condition

(c) Removable design

Fig. 3. Two-stage multi-hole orifice throttle device.

sure loss coefficient model of test MOs is provided based on

the design methodology and test results.

2. Methodology conception
MO geometry and refrigerant properties are both known to
affect the pressure drop across MOs under two-phase conditions. Generally, a large body of test data is needed to determine the correlation between these two characteristics and the
MO pressure loss. For convenience, this paper emphasizes the
MO geometry effect and leaves the influence of the refrigerant

2239

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

primary center circle

primary center circle

centered orifice

secondary center circle

secondary center circle

(a) 9-hole MO without a centered orifice

(b) 9-hole MO with a centered orifice

D c2

d h23

dh

01

12

dh

D c1

dh
13

d h02

D c2=2D c1
(c) 13-hole MO with a circular arrangement

d h10=d 12=d h13 d h20=dh23

(d) 13-hole MO with a rectangular arrangement

Fig. 4. Description of the orifice arrangement.

flashing process on MO throttle capability for consideration

by future studies. Therefore, the throttle tests in this work
were conducted under water flow.
In present work, the presented methodology mainly consists
of the definition of the MO geometric architecture and twostage throttle experiments. The primary experimental stage
investigated how the different geometric features mentioned in
MO geometric architecture affect the MO pressure loss coefficient. The secondary experimental stage constructed a model
of the MO pressure loss coefficient based on the dominant
geometric parameter extraction study conducted in the primary experimental stage.

3. MO geometric architecture
3.1 Orifice arrangement criteria
The random number and arrangement of orifices leads to
the complexity of MO geometry. To avoid this complexity,
the following criteria were defined for all MOs discussed in
the present work.
(1) All holes in each MO have a uniform size and are manufactured using the same technique. The orifice plate thickness
was 2 mm. Detailed craft specifications were according to
ISO5167-1 [9].

(2) Each orifice center was located in a set of concentric

circles defined as center circles, as shown in Fig. 4(a), (b).
(3) Orifice arrangement should be as symmetrical as possible, and the spacing among most orifice centers should be
equal.
(4) The total orifice number ranged from 3 to 13, and the
maximum number of center circles was 3.
The various patterns of orifice arrangement can be further
categorized as follows:
(1) MOs were prepared with or without a centered orifice,
as shown for typical 9-hole MOs in Fig. 4(a) and (b), respectively.
(2) Two different center circle arrangements were considered in the present work. The first mode is a circular arrangement, in which the center circle location was primarily fixed
according to the correlation Dcn=nDc1, resulting in symmetric
orifices located around these center circles as shown in Fig.
4(c).
The second mode is a rectangular arrangement with four
orifices at each center circle. This arrangement primarily considers the consistency of the spacing between each orifice
center, resulting in fixed center circle locations as shown in
Fig. 4(d).
The circular arrangement is better bedded than the rectan-

2240

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246
testing section
D/2
developing length 60D

orifice flow meter

DP transducer

00:00 kPa

testing plate

-12D -9D -6D

upstream pressure tap
p su

control valve

separator

20D 24D 28D

DP (D,D/2) downstream
pressure tap
p sd

developing length 50D

DP f

water tank

pump

Fig. 5. Schematic of the test rig.

coefficient, , is defined as

Ps
.
Pv

(1)

Relative relations can be expressed as

Ps = psu psd
1 2
v
2
4Qv
v=
D2
Pv =

Fig. 6. A portion of the test section.

gular arrangement, and the location of the center circle is not

restricted by the orifice spacing, while the rectangular arrangement has a more symmetric orifice distribution and the
orifice spacing controls the location of the center circle.
3.2 Key geometric parameters
According to the above definitions, the following three
geometric parameters were used to quantitatively characterize
the MO geometry.
(1) The total orifice number, n.
(2) The equivalent diameter ratio, EDR, which represents
the square root of the total open area ratio and is expressed as
n0.5d/D. EDR has the same geometric meaning as the diameter
ratio defined for SO specified in ISO5167.
(3) The orifice distribution density Dd, which is expressed
as dhmin/D, where dhmin represents the minimum spacing between the edges of the orifices located at adjacent center circles.
The above arrangement criteria and geometric parameters
can be used to explicitly describe the MO geometry.

4. Experimental set-up
4.1 Experimental theory
This paper adopts the pressure loss coefficient to represent
the local MO pressure drop characteristics. The pressure loss

Qv = kf

(2)
(3)
(4)

Pf

(5)

In Eq. (5), kf is a transformation factor provided by the

flowmeter manufacturer. Starting from Eqs. (1)-(5), can be
rewritten as in Eq. (6) in terms of psu, psd and Pf, which can
be directly experimentally measured.
=

2 D4
8kf 2 Pf

( psu psd )

(6)

4.2 Test rig

The experimental data were obtained from the test rig
shown in Fig. 5. The rig mainly consisted of a testing section,
an orifice flowmeter and pressure transducers. The testing
section was 6 m long. A transparent horizontal pipe with an
inner diameter, D, of 50 mm was chosen to investigate the
flow pattern. The test orifice plate was fixed on a test section
with a flange connection (shown in Fig. 6). Dematerialized tap
water was used in all experiments with a water velocity from
0.1 m/s to 1 m/s.
The test section was equipped with eight pressure taps. The
static pressure differences for calculations were obtained by
two mean values collected by tapping at distances of 6D, 9D
and 12D upstream and distances of 20D, 24D and 28D, aver-

2241

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

standard orifice

3-0-C

5-1-C

6-0-C (form 1)

6-0-C (form 2)

9-0-C

9-1-C

9-1-R

13-1-C

13-1-R

Fig. 7. Test plates employed in test 1.

aged separately. The test section was also equipped with two
pressure taps located at a distance of D upstream and D/2
downstream to compare the test results with the correlations in
ISO 5167. This test rig was used to obtain profiles of the pressure drop versus the mass of water flowing through the test
orifice plate, providing a comprehensive characterization of
the pressure loss of each test orifice.
4.3 Experimental error analysis
The relative error of the pressure loss coefficient, RE (),
was evaluated by Eq. (7) using the error combination law on
Eq. (6) and measuring the error of the relative testing instrument.

4.4 Experimental layout

is known to be dominated by the geometric features of
singularity in the square resistance region, including the orifice arrangement and the three geometric parameters discussed
above.
Therefore, the following test can be divided into the test of
the MO geometric features and the MO modeling test. The
primary testing stage analyzed the effect of the orifice arrangement and key geometric parameters on to simplify the
quantitative description of the MO geometry for modeling.
The secondary testing stage proposed a correlation based on
the test data.

5. Testing the effects of MO geometric features

2

AE(Pf ) AE(psu ) AE(psd )

+
+

+
Pf Ps Ps
RE( ) =
2
AE( D)
2
16
+ 4(RE so (Qv ))
D

5.1 Testing the effect of orifice arrangement (test 1)

(7)

where REso(Qv) refers to the relative error of Qv associated

with the flowmeter used for measurement. This term can be
neglected, as can AE(D)/D; the values of these terms are small
compared with the measurement error generated by the pressure transducers.
Suppose that the measurement errors of all transducers used
in the test were equal. The, the errors can all be expressed as
AE(Pf ) = AE(Psu ) = AE(Psd ) = AE(transducer) , and Eq. (7)
can be rewritten as
RE( ) = AE(transducer) (Pf ) 2 + 2(Ps ) 2 .

(8)

According to Eq. (8), the transducer accuracy should satisfy

the 0.1% requirement to ensure that the maximum RE() is
less than 5% under the worst-case condition.

This test aimed to determine the effects of different orifice

arrangements on the pressure loss coefficient. Ten test plates
with EDR (or DR)=0.4 were employed to perform this task,
and their descriptions and photos are shown in Fig. 7. For the
sake of brevity, the specific arrangement features of each MO
are described in a simple manner as A1-B1-C1, where A1
represents n; B1 equals 1 or 0 to indicate whether or not the
test plate has centered holes, respectively; and C1 is R or C
indicating a rectangular or circular arrangement, respectively.
Fig. 8 shows each value as a function of ReD . The results
show that for lower ReD, generally increases with ReD, while
the difference between the test plates values appears to be
much greater than that at higher ReD values. When ReD is near
10,000, there was a 47% deviation between the value obtained from SO and that from 9-1-C. The difference gradually
decreases to 17.7% when ReD increases to 15,000, and then
the difference remained stable for higher ReD values. Meanwhile, ReD=15,000 represents a cut-off point at which the
values started to deviated from the regression lines, which are

2242

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

represented as two continuous lines in Fig. 8.

Note that is usually greater for test plates with fewer orifices. This behavior is displayed in Fig. 9, which shows the
effect of orifice thickness evaluated in terms of s/d on the
mean values of . s/d represents the dimensionless orifice
thickness, where s is the orifice thickness (detailed specifica-

pressure loss coefficient

140

120

100

standard
3-0-C
5-1-C
6-0-C(1)
6-0-C(2)
9-0-C
9-1-C
9-1-R
13-1-C
13-1-R

80

60

40

10000

12500

15000

17500

20000

22500

5.2 Testing the effect of geometric parameters (test 2)

25000

Reynolds number ReD

Fig. 8. Effect of the Reynolds number on the measured pressure loss

coefficient (test 1).

with centered hole

without centered hole

pressure loss coefficient

125
120

6-0-C (form2)
115
110

rectangular arrangement

105

rectangular arrangement

100
95
0.10

0.15

0.20

0.25

0.30

0.35

0.40

dimensionless orifice thickness s/d

Fig. 9. Mean pressure loss coefficient vs. dimensionless orifice thickness (test 1).

3-0.2-0.02

3-0.3-0.04

6-0.2-0.08
Fig. 10. Test plates employed in test 2.

tions were according to ISO5167-1) and was 2 mm for all test

plates in this paper, and d is the restriction diameter of separated throttle holes.
According to Chisholm [7], all test plates here can be classified as thin plates due to their s/d values (s/d<0.5). generally
increases with decreasing s/d, with a 24% decrease in observed when s/d is increased from 0.1 to 0.38. The minimum
was attained with an s/d ratio of 0.3, at which undergoes a
sharp decrease. This behavior can be ascribed to the pattern of
orifice distribution, which induced a stronger decrease in to
overcome the influence of s/d. The data also reveal that various orifice arrangements result in different values at the
same s/d value, which is represented in Fig. 9 as three lines
with arrowheads.

In this test, test plates with fixed orifice arrangements such

as 3-0-C, 6-0-C (form 2) and 9-0-C were used to investigate
the influence of three geometric parameters on the pressure
loss coefficient. Nine test plates were selected with various
combinations of three different n values (n=3, 6, 9), three different EDR values (EDR=0.2, 0.3, 0.4) and three different Dd
values (Dd=0.02, 0.04, 0.08). For the sake of brevity, the plates
are referred to using the notation A2-B2-C2, where A2, B2
and C2 indicate the values of n, EDR and Dd, respectively. Fig.
10 shows images of and detailed information on these test
plates.
Fig. 11 shows the overall test results in terms of as a function of ReD. As can be observed, the experimental data can be
divided into three regions ranging from 1500 to 2000, 250 to
500 and 60 to 120. These regions are mainly characterized by
different EDR values. For higher EDR values, the data have a
lower dependence on ReD. Meanwhile, the other two geometric parameters appear to have extremely weak effects on the
pressure loss coefficient. These results demonstrate that EDR

3-0.4-0.08

9-0.4-0.02

6-0.3-0.02

9-0.2-0.04

6-0.4-0.04

9-0.3-0.08

2243

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

3-0.2-0.02
3-0.3-0.04
3-0.4-0.08
6-0.3-0.02
6-0.4-0.04
6-0.2-0.08
9-0.4-0.02
9-0.2-0.04
9-0.3-0.08

2000

pressure loss coefficient

2500

1500

d hmin

le
D

1000

D'

500

10000

12500

15000

17500

20000

22500

25000

Reynolds number ReD

Fig. 13. Configuration schematic of 6-0-C (form 2).

Fig. 11. Overall results of test 2.

1800

pressure loss coefficient

1500

D ' = 4.155d + 8.31le + 3.154d h min

( n)
( Dd)

EDR = 6

EDR
1200

Dd =

(10)

d
D

(11)

d hmin
.
D

(12)

900

Starting from Eqs. (9)-(12), the relationship between EDR,

Dd and le can be determined as

600

300

EDR < 0.59 1.86 Dd 4.9le D 1 .

3

0.04 0.08
Dd
geometric parameter

0.02

0.2

0.3
EDR

Fig. 12. Range values of the geometric parameters studied here.

has the dominant effect on pressure loss.

Fig. 12 presents more explicit evidence for the relationship
between geometric parameters and the loss coefficient using
the aid of range method. As can be observed, the range values
for n, Dd and EDR were 177, 137 and 1560, respectively,
demonstrating the overwhelming influence of EDR on , especially for test plates with EDR values ranging from 0.2 to
0.3, due to the high range value in this domain (nearly 1400).
These data allowed us to conclude that EDR is the dominant
geometric parameter because of its very strong influence on
the of MO, as seen in Fig. 10. Therefore, it is feasible to
adopt EDR as the single parameter affecting in a further
modeling study.

6. Modeling the MO pressure loss coefficient

6.1 Testing the accuracy of the test rig using SOs according
to ISO5167 (test 3)
Based on the investigations above, this test was designed to
model the relationship between and EDR using five 6-0-C
test plates (form 2). First, we analyzed the detailed geometry
of the MOs. As can be observed in Fig. 13,

Adopting a Dd value of 0.02, the EDR should be lower than

0.454 because of the test restrictions of D=50 mm and le=1
mm. According to the EDR range, ten test plates (shown in
Fig. 14) including five 6-0-C (form 2) MOs with EDR=0.25,
0.3, 0.35, 0.4 and 0.45 (denoted as M-EDR) and five SOs
(defined in ISO5167) with DR=0.25, 0.3, 0.35, 0.4 and 0.45
(denoted as S-DR) were chosen for this experiment. Because
EDR and DR have the same geometric meanings for MO and
SO, respectively, the results of this test can be compared to
available correlations proposed in ISO5167 to investigate the
reliability of the test data. The relevant equations can be generalized as:
2

(9)

Re
P( D , D / 2) = 0.5 1 D ( DR 4 1)
D C

(14)

where P(D,D/2) represents the pressure difference using the D

and D/2 pressure tapping measurements.
The SO discharge coefficient C can be calculated using the
Stolz equation (Eq. (15)) [11] and the Reader-Harris/Gallagher equation (Eq. (16)) [12] to determine the discharge coefficients according to ISO5167 (1991) [8] and ISO5167 (2003)
[9, 10], respectively.
C = 0.5959 + 0.0312 DR 2.1
0.0184 DR8 + 0.0029 DR 2.5 (106 / Re D )0.75
4 1

+0.039 L1DR (1 DR ) 0.0337 L2 DR

4

D' < D

(13)

0.4

(15)

2244

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

S-0.25

S-0.3

S-0.35

S-0.4

S-0.45

M-0.25

M-0.3

M-0.35

M-0.4

M-0.45

Fig. 14. Test plates employed in test 3 and test 4.

90000

75000

Stolz Eq.(15)
Reader-Harris
/Gallagher Eq.(16)

60000

P(D,D/2) (Pa)

ways over-estimates P(D,D/2) as calculated by the ReaderHarris/Gallagher equation, with 2%, 1.7%, 1.4%, 1.0% and
0.6% deviations for DR(EDR)=0.25, 0.3, 0.35, 0.4 and 0.45,
respectively.
For the same ReD, the P(D,D/2) value of 6-0-C (form 2) was
generally greater than that of SO, with a 6.5% deviation when
DR (EDR)=0.25. The difference in P(D,D/2) decreases with
increasing DR (EDR) values, with a 1.5% deviation as DR
(EDR) =0.4. This trend also applies to ReD, as when DR (EDR)
=0.3, the differences in P(D,D/2) were 0.9% and 3.8% at
ReD=9600 and ReD=19,200, respectively.

S-0.25
S-0.3
S-0.35
S-0.4
S-0.45
M-0.25
M-0.3
M-0.35
M-0.4
M-0.45

45000

30000

15000

10000

12500

15000

17500

20000

22500

25000

Reynolds number ReD

6.2 Modeling the MO pressure loss coefficient (test 4)

Fig. 15. Pressure difference between the D and D/2 pressure measurements as a function of the Reynolds number.

For the MOs presented in Fig. 14:

= Function(n, Dd , EDR ) .

C = 0.5961 + 0.0261DR 2 0.216 DR 8 +

0.000521(106 DR / Re D )0.7 + (0.0188 + 0.0063 A) DR 3.5 (106 / Re D )0.3
+(0.043 + 0.08e 10 L1 0.123e 7 L1 )(1 0.11A) DR 4 /(1 DR 4 )

(17)

According to the results of test 2, Eq. (17) can be simplified as

= Function( EDR) .

0.031( M 0.8M 1.1 ) DR1.3 + 0.011(0.75 DR )(2.8 D / 25.4)

(18)

(16)
where L1=1L2=0.47 for the D and D/2 pressure tapping
measurements, and A=(19000DR/ReD)0.8M=-2L2/(1-DR).
Fig. 15 shows P(D,D/2) as a function of ReD. The experimental data were compared with the Stolz equation and the
Reader-Harris/Gallagher equation.
As can be observed, near S-0.25, the experimental data are
fit well by the correlations involving both the Stolz and
Reader-Harris/Gallagher equations, especially for a higher
DR(EDR) or a lower ReD. These comparisons may explain the
reliability of this experimental set-up, and they indicate that
the model is sufficiently accurate for use in a further modeling
study on the value of 6-0-C plates (form 2).
Furthermore, it is worth noting that the Stolz equation al-

The model =Polynomial(EDR)(EDRC-1) was adopted for

the correlations. The results are shown in Fig. 16 with the
following model description:

( EDR ) = Pm ( EDR 4.448 1)

(19)

( DR) = Ps ( DR 4.187 1)

(20)

where
Pm = 160.325(71.467 EDR 4 100.300 EDR 3
+52.021EDR 2 11.801EDR + 1)
Ps = 150.848(74.679 DR 4 103.507 DR 3
+53.001DR 2 11.874 DR + 1).

2245

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

Experimental data with standard orifice

Experimental data with 6-0-C(form2)
fitting curve for standard orifice
fitting curve for 6-0-C(form2)

pressure loss coefficient

1000

800

600

400

200

700

Experimental data
Correlation Eq.(19)

600

pressure loss coefficient

1200

500

400

300

200

100
0.20

0
0.20

0.25

0.30

0.35

DR(EDR)

0.40

0.45

M-0.37

0.30

0.35

0.40

0.45

0.50

EDR

Fig. 18. Model validation results.

Fig. 16. Fitting curves vs. experimental data.

M-0.27

0.25

0.50

M-0.33

M-0.43

Fig. 17. Test plates employed in test 5.

As shown in Fig. 16, the two curves for SO and 6-0-C (form
2) exhibit almost the same trend when DR (EDR) is greater
than 0.27. The value of 6-0-C (form 2) was slightly greater
than that of an SO when the DR (EDR) was lower than 0.27.
6.3 Validation of the model (test 5)
To verify Eq. (19), another four 6-0-C (form 2) plates with
EDR= 0.27, 0.33, 0.37 and 0.43 were tested, as shown in Fig.
17. The results are presented in Fig. 18. Note that Eq. (19)
agrees well with the experimental data obtained from the four
test plates when EDR ranged from 0.25 to 0.45.

7. Conclusion
This paper introduced an MO structural design methodology and implemented it experimentally. First, the MO structure was quantified through the geometric architecture. Then,
experiments were used to simplify and further develop the

correlation between MO geometric features and . The final

model is highly accurate, indicating that the methodology used
here is effective.
This study leads to the following conclusions:
(1) The values of MOs are generally higher for MOs with
fewer perforated holes. Even though all test plates can be classified as thin orifices, the s/d value is the dominant factor determining the MO value, and the orifice arrangement is also
an important aspect affecting the MO value.
(2) The orifice arrangement of MOs can be expressed by
three key parameters: the total orifice number, n; the equivalent diameter ratio, EDR; and the distribution density, Dd.
EDR has a dominant influence on the MO value compared
with the other two parameters.
(3) The geometric factors can be ranked according to decreasing effect on as EDR, n, Dd, s/d and orifice arrangement
form.
(4) The relationships between EDR and DR were used to
compare available correlations with the test data. This comparison revealed that the test apparatuses had an acceptable
accuracy, as the results fit with correlations using either the
Stolz or Reader-Harris/Gallagher discharge coefficient models.
(5) The model obtained in the present paper provides a reliable MO prediction with EDR values ranging from 0.25 to
0.45. The test procedures may also be applied to MOs with
other orifice arrangements.
(6) The design methodology and test procedure presented in
this work allow for the simplification of the structural design
of MOs. This work also provides a good reference for further
studies of MO throttle characteristics in two-phase flow systems.

Acknowledgment
The research presented in this paper was financially supported by the National Natural Science Foundation of China
(Grant No. 51008042, No. 51078053), the Fundamental Research Funds for the Central Universities of China (No.
DUT11ZD105) and by China Postdoctoral Science Foundation (No. 20100481235).

2246

T. Zhao et al. / Journal of Mechanical Science and Technology 25 (9) (2011) 2237~2246

Nomenclature-----------------------------------------------------------------------AE
C
D
D
Dc
Dd
DR
d
dhmin
EDR

n
Pv
P(D,D/2)
Pf
Ps
psd
psu
Qv
RE
s
v

: (Effective) work potential absolute error value

: Exergy discharge coefficient
: Energy of a system inner diameter of pipe, m
: External edge diameter, m
: Diameter of center circle, m
: Orifice distribution density
: Diameter ratio of single-hole orifice
: Orifice restriction diameter, m
: Minimum spacing between the edges of orifices
located at adjacent center circles, m
: Equivalent diameter ratio of multi-hole orifice le
difference between the orifice edge radius and the
orifice radius, m
: Total number of orifices
: Velocity pressure, Pa
: Pressure difference using D and D/2 tapping measurements, Pa
: Pressure difference across orifice determined using
a flowmeter, Pa
: Static pressure difference across the test plate, Pa
: Static pressure downstream, Pa
: Static pressure upstream, Pa
: Volume flow, m3/s
: Relative error value
: Orifice thickness, m
: Mean velocity, m/s
: Pressure loss coefficient
: Dynamic viscosity, Pas
: Water density, kg/m3

References
[1] A. B. C. Anderson, A circular-orifice number describing
dependency of primary frequency on differential pressure,
gas density and orifice geometry, Journal of Acoustical Society of America, 25 (1953) 626-631.
[2] P. Testud, P. Moussou, A. Hirschberg and Y. Auregan,
Noise generated by cavitating single-hole and multi-hole orifices in a water pipe, Journal of Fluids and Structures, 23 (2)
(2007) 163-189.
[3] G. H. Gan and S. B. Riffat, Pressure loss characteristics of
orifice and perforated plates, Experimental Thermal and
Fluid Science, 14 (2) (1997) 160-165.
[4] G. Kojasoy, P. Kwame-Mensah and C. T Chang, Two-phase
pressure drop in multiple thick and thin orifices plates,

Experimental Thermal and Fluid Science, 15 (4) (1997) 347358.

[5] M. Fossa and G. Guglielmini, Pressure drop and void fraction profiles during horizontal flow through thin and thick
orifices, Experimental Thermal and Fluid Science, 26 (5)
(2002) 513-523.
[6] A. A. Saadawi, E. Grattan and W. M. Dempster, Two-phase
loss in orifice plates and gate valves in large diameter pipes,
in: G. P. Celata, P. Di Marco, R. K. Shah (Eds.), 2nd Symp.
Two-Phase Flow Modeling and Experimentation. ETS, Pisa,
Italy (1999).
[7] D. Chisholm, Two-phase flow in pipelines and heat exchangers, Longman Group Ed, London (1983).
[8] International Organization for Standardization, Measurement
of fluid flow by means of pressure differential devicespart
1: orifice plates, nozzles and Venturi tubes inserted in circular cross-section conduits running full, ISO 5167-1. International Organization for Standardization, Geneva (1991).
[9] International Organization for Standardization, Measurement
of fluid flow by means of pressure differential devicespart
1: orifice plates, nozzles and Venturi tubes inserted in circular cross-section conduits running full, ISO 5167-1. International Organization for Standardization, Geneva (2003).
[10] International Organization for Standardization, Measurement of fluid flow by means of pressure differential
devicespart 2: orifice plates. International Organization for
Standardization, Geneva (2003).
[11] J. Stolz, A universal equation for the calculation of discharge
coefficients of orifice plates, In Flow Measurement of Fluids
(ed. H.H.Dijstelbergen and E.A.Spencer), North Holland,
Amsterdam (1978) 519-534.
[12] M. J. Reader-Harris, J. A. Sattary and E. P. Spearman, The
orifice plate discharge coefficient equation-further work.
Flow Meas. Instrum., 6 (2) (1995) 101-144.

Zhao Tianyi is a postdoctoral researcher

in HVAC department at Dalian University of Technology, Dalian, China, 2009.
He received his Ph.D degree in School of
Municipal & Environmental Engineering,
Harbin Institute of Technology, Harbin,
China, 2005-2009. His research interests
are optimal control of HVAC equipment
and system, and design and development of novel HVAC
equipment.