PALIVA 5 (2013), 1, P.

21 - 25

Measuring a Discharge Coefficient of an Orifice for an Unsteady Compressible Flow

MEASURING A DISCHARGE COEFFICIENT OF AN ORIFICE FOR AN UNSTEADY

COMPRESSIBLE FLOW
Ondej Novk, Vclav Koza
ICT Prague, Department of Gas, Coke and Air Protection, [email protected]
There are many applications where a compressible gas flows through an orifice or a nozzle. This complex process has been examined for long time. Nevertheless current one or two dimensional mathematic and
semi-empirical theories that predict a flow of a compressible gas through an orifice during an unsteady discharge from a vessel are only partially successful. It is impossible to treat all factors that might affect such a
discharge. An experimental examination of a compressible gas flow through an orifice is still important part
for evaluation and prediction of a discharge coefficient. This paper brings an idea for setup an apparatus on
which a broad range of properties of an orifice can be tested. Several groups of properties were tested on different types of orifices and the results are presented.
Key words: discharge coefficient, compressible flow, unsteady flow, discharge from a vessel
Received 6. 2. 2013, accepted 5. 3. 2013

1. Introduction
A flow of a gas through an orifice can be characterized as steady or unsteady by mass flow rate through the
orifice. If mass flow rate is constant during the examined process then flow is steady.
There are many applications where a gas flows
through an orifice or a nozzle. One of the applications is
a discharge of a gas from a vessel through an orifice.
The discharge is treated as complex process that is characterized by properties of the gas, flow regime, properties of the orifice, properties of the vessel and properties
of surrounding space. From a practical approach this
complex system of properties could be simplified to one
coefficient that could characterize difference between
real and theoretic discharge. The discharge coefficient is
usually treated as a parameter of an orifice that affects
the area of the orifice for further computation [1].
It is impossible to treat all the factors that might influence practical discharge process. Even for relatively
simple case of steady flow through an orifice several
mathematical and semi-empirical solutions were obtained for well-defined sharp-edged orifices [2-4].
Discharge process is influenced even by shape and
width of the orifice as stated before. Consequently even
quality and shape of the upstream edge of the orifice has
some influence on the discharge coefficient [5-7].
Thermodynamic parameters of the gas inside vessel are changing during the discharge process. The discharge process then does not depend only on thermodynamic properties in any instant but even on history of
the discharge as stated in [8].
Temperature of the gas in the vessel during the discharge from the vessel is bounded on one side by temperature computed from adiabatic expansion and on the
other side by the constant temperature of the gas at the
beginning of the discharge process. In addition, thermal
conditions in the vessel are changed not only by discharge process itself but even by heat transfer from
space surrounding the vessel. For long term discharges

this could be major influence that affects temperature of

the vessel wall but even thermodynamic properties of
the gas in the vessel at any instant of the process [9].
Consequently computation of the heat transfer
from the surrounding space into the gas in the vessel is
not a trivial problem. In this case convection is the main
heat transfer process. Due to dynamic changes of the
temperature of the gas inside the vessel the heat transfer
is unsteady. For larger vessels natural convection takes
main role during heat transfer and problem of boundary
layer must be solved. The natural convection causes
movement of gas layers inside the vessel so the temperature gradient in the gas changes not only with the time
but even with the position in the vessel [10-12].
In this paper we propose measurements of discharge coefficient for several types of orifices with
different shapes and other properties. In addition we
have measured special type of orifices that extends
measuring of two dimensional orifices to orifices with a
three dimensional lift that can be found for example
during pipeline accidents. We use a theoretic approach
to determine temperature of a gas inside a vessel during
a discharge process.

2. Experimental part
2.1. Theory
Suppose a vessel of a volume V [m3] full of a gas
characterized by a molar weight M [kg.mol-1] and an
isentropic coefficient [-]. Thermodynamic state of the
gas is defined by a thermodynamic temperature T []
and an absolute pressure P [Pa]. Properties of the gas
inside the vessel are without subscript.
Space that is surrounding the vessel is filled by air
with temperature Tatm and absolute pressure Patm. Properties of the surrounding gas have subscript atm.
Let us define critical pressure ratio [-]

P
1 1
crit

Patm 2

21

(1)

PALIVA 5 (2013), 1, P. 21 - 25

Measuring a Discharge Coefficient of an Orifice for an Unsteady Compressible Flow

There is an orifice with an area S [m2] connected to

a vessel. When a gas is flowing through the orifice then
a mass flow rate of the gas is

m Gtheor S

(2)
where
-1
m [kg s ] is a mass flow rate,
Gtheor [kg.m-2.s-1] is a theoretic density of a mass
flow rate and
S [m2] is an area of the orifice.

We prepared three-dimensional orifices for testing.

Model of such orifice is in Figure 1. We called this type
of orifices as lifted. The orifice consists of a
two-dimensional orifice and a three-dimensional lifted
flap, which may change effective area of the orifice and
consequently discharge coefficient. Area of the lifted
orifice, which is used for calculation, is the free area of
the two-dimensional part of the orifice.

Suppose that following properties are constant during the experiment:

the temperature of air surrounding the vessel,
the pressure of the air surrounding the vessel,
the volume of the vessel and
the molar weight of the gas in the vessel.
Influence of compressibility is omitted for our experiment and compressibility is treated as equal to one.
From thermodynamic analysis of the adiabatic flow
we get equation for theoretical density of mass flow rate
for adiabatic flow at any instant as
1/
Gteor (t ) Patm
P(t )( 1)/ 2

M
1 RT (t )

1 Patm
P(t )

Figure 1 Model of a lifted orifice

Tab. 1 Properties of tested orifices

(3)

P
when
and
Patm
1

Gteor (t ) P(t )

M
2 1 (4)

RT (t ) 1

otherwise.
R [J.mol-1.K-1] is the universal gas constant equals
to R=8.314 J.mol-1.K-1.
The mass in the vessel at any instant can be calculated from the state equation of the ideal gas
P(t )VM
m(t )
RT (t )
(5)
where V [m3] is a volume of a vessel.
As a consequence of equation 2 and 5 we get an
experimental density of a mass flow rate for measured
values P(t) and T(t)

Label

Shape

1_1
1_2
1_3
1_4
1_5
1_6
1_7
2_1
2_2
2_3
2_4
2_5
2_6
2_7

Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Circular
Narrow
rectangular slot
Narrow
rectangular slot
Square
Square
Square
Square
Triangular
Triangular
Triangular
Triangular

S1

Gexp (t ) m(t )S

(6)
Consequently a discharge coefficient [-] is defined by
G (t )
(7)
(t ) exp
Gtheor (t )

S2
C1.1
C2.1
C2.1
C2.3
T1.1
T2.1
T2.2
T2.3

2.2. Orifices
For the purpose of our experiment we developed
and tested several types of orifices. They differ in shape,
area and thickness to test broad range of parameters
during experiments.

22

S
Thickness Angle of
2
[mm]
lift []
[mm ]
3.27
0.6
7.79
0.6
15.34
0.6
28.27
0.6
62.77
0.6
110.47
0.6
260.16
0.6
3.70
10.0
7.40
10.0
15.00
10.0
26.79
10.0
64.33
10.0
111.59
10.0
262.73
10.0
5.16

1.6

18.95

1.6

28.62
113.42
93.72
84.14
22.45
99.39
90.90
94.69

1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6

33
52
33
43

PALIVA 5 (2013), 1, P. 21 - 25

Measuring a Discharge Coefficient of an Orifice for an Unsteady Compressible Flow

2.3. Apparatus
We built an apparatus for measuring a discharge
coefficient during an unsteady discharge. The apparatus
consists of a main pressure vessel, a flange for a tested
orifice and a pressure and temperature sensor. Volume
of the main pressure vessel is 0.16 m3. Maximum operating overpressure is 1 MPa.
We use naked thermocouples for measuring temperature inside and outside of the vessel. Thermocouples are calibrated online by Pt100 sensor at the beginning of each experiment.
Gas is supplied from a high pressure cylinder in
order to guarantee the same quality of the gas during
experiments. Because temperature drops subzero rapidly
during experiments low humidity of supplied gas is also
very important.
Data from temperature and pressure sensors are acquired via Advantech 16 bit AD convertor and with
timestamp synchronously logged by computer.

ed Landrams equation system to compute temperature

inside the vessel [10]. The mathematical model of the
temperature inside the vessel was adjusted to correspond with experimental data.
Results of our experiments are grouped by dominating parameters of the tested orifices shape and area.
At first we tested circular orifices with different areas. Results for thin circular orifices are shown in Figure 3 and for thick circular orifices in Figure 4. For thin
orifices the discharge coefficient decreases rapidly with
decreasing pressure in the vessel in contrast to thick
orifices, which discharge coefficient is more stable
during the discharge. The discharge coefficient does not
depend heavily on the area of the orifice in both families of orifices. Orifices with the smallest area are the
only exception.

Figure 3 Dependency of the discharge coefficient on

Patm/P for a subset of thin circular orifices
Figure 2 Experimental apparatus
2.4. Setup of an experiment
A tested orifice was attached and sealed to the outflow flange. The main pressure vessel was filled by air
from high pressure cylinder up to overpressure about
0.85 MPa. Because pressure was reduced from highpressure in the HP gas cylinder to medium pressure
inside the vessel, temperature of the filled gas was lower
than surrounding temperature.
After an hour both temperatures equalized. Then
overpressure inside the vessel was examined and finely
adjusted to 0.85 MPa. Surrounding barometric pressure
was measured by an independent instrument and logged.
After computer program for data logging was started,
the outflow valve was opened and the discharge process
began.

3. Results
Even though we use naked thermocouples for
measuring of temperature, they are only in industrylevel quality and their time constant is too long for
measuring fast changes of the gas temperature inside the
vessel during some of our experiments. We implement-

Figure 4 Dependency of the discharge coefficient on

Patm/P for a subset of thick circular orifices
Discharge coefficients for square orifices are
shown in Figure 5 together with results for thin circular
orifices that have similar area. Orifice C1.1 is comparable to orifice 1_4 and C2.1 to 1_6. Graph shows that
discharge coefficient of circular and square orifices with
similar areas are closely related, even though there
might be differences for orifices with bigger areas.

23

PALIVA 5 (2013), 1, P. 21 - 25

Measuring a Discharge Coefficient of an Orifice for an Unsteady Compressible Flow

Figure 5 Dependency of the discharge coefficient on

Patm/P for square orifices and thin circular orifices with
comparable area
Discharge coefficients of narrow rectangular slots
are much higher than discharge coefficients for comparable circular orifices (Fig 6). The same phenomenon
can be more clearly observed for triangular orifices
(Fig 7). But it seems that the effect decreases with increasing area of the orifice.

Results for orifices with a lift are shown in Figure

8 for square orifices and in Figure 9 for triangular orifices. All lifted orifices were tested on both sides. So the
results show discharge coefficients for upstream position of the lift label Up and for downstream position of
the lift label Down. Upstream position of the orifice
means that the lifted part is oriented against the flow.
As is clearly visible from Figures 8 and 9 the position and the angle of the lift significantly affect value of
the discharge coefficient. Smaller angle of the lift reduces the discharge coefficient comparing to the discharge coefficient of the fully opened two-dimensional
orifice of the same shape and area. In contrast the angle
of the lifted part that is greater than some value can
increase discharge coefficient above the coefficient of
flat square shaped orifice with the same area. For the
triangular lifted orifices the discharge coefficients are
much lower than the discharge coefficient of the flat
triangular orifice with the same area. But higher lift
angles of tested square and triangle orifices are not the
same, so the threshold angle of the phenomenon mentioned above for square orifices may still exist even for
triangle orifices.

Figure 6 Dependency of the discharge coefficient on

Patm/P for narrow rectangular slots and thin circular
orifices with comparable area

Figure 8 Dependency of the discharge coefficient on

Patm/P for square orifices with a lift positioned upstream
and downstream and flat square orifice with comparable
area

Figure 7 Dependency of the discharge coefficient on

Patm/P for triangular orifices and thin circular orifices
with comparable area

Figure 9 Dependency of the discharge coefficient on

Patm/P for triangular orifices with a lift positioned upstream and downstream and flat triangular orifice with
comparable area

24

PALIVA 5 (2013), 1, P. 21 - 25

Measuring a Discharge Coefficient of an Orifice for an Unsteady Compressible Flow

Upstream and downstream position of the lift affects the discharge coefficient of the orifice. The discharge coefficient of the same lifted orifice is bigger in
downstream position then in upstream position. Same
effect was observed for both square and triangle orifices.
Deeper investigation of the lift angle effect on the
discharge coefficient will be the main task for the further research.

4. Conclusion
In this paper we proposed measurements of the
discharge coefficient for several types of orifices with
different shapes and other properties. In addition we
measured special type of orifices that extends measuring
of discharge coefficient of the two-dimensional orifices
to the measuring of the discharge coefficient of the
orifices with a three-dimensional lift. They can be found
on ruptured pipelines during pipeline accidents caused
by puncturing the pipeline from outer side. We used
Landrams theoretic approach to determine the temperature of the gas inside the vessel during measured discharge processes.

Acknowledgment
This work was realized with support of Tor
Kjeldby from Department of Energy and Process Engineering NTNU Trondheim.

Literature

7. Kayser, J. Shambaugh, R. Discharge coefficients

for compressible flow through smalldiameter orifices and convergent nozzles. Chemical Engineering
Science. 1991, 46, 7, s. 16971711.
8. Chang ,Y. F. Transient Effects in the Discharge of
Compressed Air from a Cylinder throught an Orifice. PhD. Thesis, University of Saskatchewan, 1968
9. Johnston, S. C. A Characterization of Unsteady Gas
Discharge from a Vessel, Dissertation, University of
California, Davis, 1974
10. Landram, C. S. Heat Transfer during Vessel Discharge:Mean and Fluctuating Gas Temperature.
Journal of Heat Transfer. 1973, 95, 1, s. 101106.
11. Haque, A. et al. Rapid depressurization of pressure
vessels. Journal of Loss Prevention in the Process
Industrie s. 1990, 3, 1, s. 4 7.
12. Xia, J. Smith, B. Yadigaroglu, G. A simplified
model for depressu-rization of gas-filled pressure
vessels. International Communications in Heat and
Mass Transfer. 1993, 20, 5, s. 653 664.
13. Novk J.: Termodynamick vlastnosti plyn. Praha :
VCHT, 2007.
14. Koza V., Novk O.: Men nestacionrnho niku
plynu, Sbornk 18. chemicko-technologick konference APROCHEM 2009, Milovy 20-22.4.2009, 2.
dl, s. 2356-2359, MPO program Tandem projekt
FT-TA2/005.
15. Koza V., Novk O.:Vliv sloen unikajcho plynu
na vtokov souinitel otvoru, Sbornk pednek
konference VOC 2009, Pardubice 18-19.6.2009,
s.57-60.

1. Lapple, C. E. Isothermal and Adiabatic Flow of

Compressible Fluids. AIChe. 1943, 39, s. 385428
2. Bragg, S. L. Effect of Compressibility on the Discharge Coefficient of Orifices and Convergent Nozzles. Journal of Mechanical Engineering Science.
1960, 2, 1, s. 3544.
3. Benson, R. S. Pool, D. The Compressible Flow
Discharge Coefficients for a Two-Dimensional Slit.
International Journal of Mechanical Sciences.1965b,
7, s. 337353.
4. Deckker, B. E. L. Chang, Y. F. Paper 7: An Investigation of Steady Compressible Flow through Thick
Orifices. Proceedings of the Institution of Mechanical Engineers, Conference Proceedings. 1965, 180,
10, s. 312323.
5. Deckker, B. E. L. Compressible Flow through
Square Edge Rectangular Orifices. Proceedings of
the Institution of Mechanical Engineers. 1978, 192,
1, s. 277288.
6. Deckker, B. E. L. Chang, Y. F. Paper 19: Slow
Transient Compressible Flow through Orifices. Proceedings of the Institution of Mechanical Engineers,Conference Proceedings. 1967, 182, 8,
s. 175183.

25