Proceedings of PVP 2007

ASME Pressure Vessels and Piping 2007 / Creep 8 Conference

July 22-26, 2007 San Antonio, Texas, USA

Draft PVP2007-726157

AN ACOUSTIC CRITERION FOR THE WHISTLING OF ORIFICES IN PIPES

P. Moussou* Ph. Testud Y. Auregan A. Hirschberg

LaMSID, UMR CNRS LaMSID, UMR CNRS Laboratoire d’Acoustique Fluid Dynamics Lab.,
EDF 2832, France EDF 2832, France de l’Université du Maine, Technische Universiteit
France Eindhoven, The
Netherlands

ABSTRACT orifice with a pipe upstream and an open air termination

Whistling due to vortex shedding with lock-in has been downstream has been studied by Anderson [3, 4], and some
extensively studied in the case of cylinders in cross-flows, of industrial studies involving cavitation have been reported [5, 6,
flow separation above cavities and of shear layers with flow 7, 8, 9].
impingement feedback. Less attention has been given to A better understanding of the whistling of orifices in the
pressure drop devices in piping systems, which are known to presence of acoustic feedback can be obtained experimentally
generate high noise levels due to single tones in gas systems, by testing an orifice in anechoic conditions, and evaluating its
and even in water systems. response to pressure waves. Recently, an instability criterion
Based on recent works of Auregan et Starobinski (1999), was proposed by Auregan and Starobinski [10], involving
an experimental criterion is proposed to evaluate the whistling acoustic waves of low or high order, and taking into account
ability of a pressure drop device in the presence of plane waves the influence of the flow velocity on propagation [11]. The
acoustic feedback. The idea of the criterion can be summarized purpose of the present study is to apply this criterion to orifices
as follows: if for a given combination of incident pressure at low Mach numbers, in the framework of plane acoustic
waves, the amount of acoustic power scattered is higher than waves.
the incident one, the pressure drop device behaves as an
acoustic amplifier, so that whistling can occur if the adequate
acoustic boundary conditions are met. The main interest of this EXPERIMENTAL PROCEDURE
criterion is that it depends only on the acoustic scattering Considered as a ‘black box’ inside an acoustic system, an
matrix of the device, and not on the arrangement of the orifice subjected to incident pressure waves can be described
surrounding pipe. by its plane wave scattering matrix [12], according to:
Results obtained in an air test rig with an inner diameter of
3 cm, a Mach number varying from 10-3 to 10-1 and a Reynolds  p down
+
 T + R −  pup 
+
 =  , (1)
number varying from 103 to 105 are reported for single hole   R +
 pup

−
 T −  p down
− 

orifices. Basing the Strouhal number on the thickness of the
orifice and on the average velocity through the hole, thin single
the complex coefficients T± and R± being functions of the
hole orifices with sharp angles appear to whistle in a range of
frequency and of the flow velocity, and the notations being
Strouhal numbers close to .2. Furthermore, it is shown that a
indicated in Fig. 1. The present study requires the determination
thin orifice with a downstream bevel is prone to whistling,
of these coefficients in order to express a whistling criterion.
whereas the same orifice with the bevel upstream cannot
whistle.
pup+ pdown
+

INTRODUCTION U
Vortex shedding with lock-in has been extensively studied pup- p-
down
in the past decades for cylinders submitted to cross-flow, but
the whistling of orifices in pipes due to acoustic feedback has Figure 1: sketch of the incident and scattered acoustic
pressure waves
been less studied except in overviews [1, 2]. The case of an

1 Copyright © 2007 by ASME


Figure 2: experimental set-up
Test rig description
The test rig of the Laboratoire d’Acoustique de
l’Université du Maine (LAUM) is designed to determine the
scattering matrix of passive acoustic devices. The test section
consists out of a straight pipe with an inner diameter equal to
30 mm and a total length of 6 m. A constant air flow is
generated by a compressor Aerzen Delta blower GM10S (#1 in
Fig. 2). The flow rate is measured with a flow meter ITT
Barton 7402 (#2 in Fig. 2) with a measurement range from 0.03
to 0.157 m3.s-1. Quasi-anechoic terminations are arranged
upstream and downstream (#3 and #9 in Fig. 2) of the
measurement area; these terminations are made of a perforated
tube and covered with textile. Two loudspeakers are located
upstream (#4) and downstream (#8) of the orifice. They
generate an acoustical pressure up to 160 SPL in the frequency
range 400-4000 Hz.
Acoustic pressures are measured upstream and
downstream of the orifice (located at point 6 in Fig. 2) by a
couple of 4 microphones B&K 4938 ¼’’ with Nexus 2690
amplifiers , located respectively at points 5 and 7 in Fig. 2. The
distances between sensors are respectively 63.5 mm, 211.5 mm
and 700 mm in order to optimize the identification of
propagating pressure waves.
The duct is 2 m long upstream of the first series of
microphones, so that the flow is fully developed. The
temperature is measured with 2 sensors on both sides of the
orifice, so that the speed of sound can be estimated. Its value
was equal to about 345 m/s in all experiments.
1.0
0.8
0.6
d/D
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6
t/D
Figure 3: dimensionless features of the sharp angle
orifices tested
2 Copyright © 2007 by ASME
Tested orifices
The tested orifices exhibit a centered single hole with neat
sharp angle edges. Their thicknesses t and hole diameters d are
given in Fig. 3. For the sake of illustration, one orifice with a
bevel was tested as well.
Scattering matrix determination
The scattering matrix coefficients are automatically
determined with the help of a computer system, details can be
found in [13]. During the experiments, the loudspeaker sound
level was kept sufficiently low to ensure that the results were
not depending on it.
The sound generated by the loudspeakers of the test rig can
be fairly described by harmonic pressure waves proportional to
exp(jωt ± kx). It is worth mentioning the fact that in practical
applications, the incident pressure waves are generated by
external broadband noise sources of stochastic nature [14, 15],
and the formalism of Power Spectrum Density should be used
instead. This is yet of no practical consequence, because the
replacement of for instance pup+ by the cross-spectrum of pup+
with some reference pressure would leave the scattering matrix
unaltered. For the sake of simplicity, the harmonic formalism is
hence used in the frame of the present study.
Dimensionless representation of the results
The choice was made to plot the results using ft/Ud as a
dimensionless frequency, where t is the thickness of the orifice
and Ud is the flow velocity in the hole. The other dimensionless
numbers of the experiments are the Mach number U/c,
expressed in the framework of the study as a function of the
pipe velocity U, the ratio of hole diameter to pipe diameter d/D,
and the Reynolds number UD/ν , with self-evident notations.
LOW MACH NUMBER DESCRIPTION OF ORIFICES
Reduction of the scattering matrix to a unique
coefficient
Strictly speaking, the four coefficients T± and R± are
independent and need be determined separately. However, as
the size of the orifice is small compared to the wavelength, and
as the Mach number is low, one expects the acoustic velocity
(p+ - p- )/ρc to have the same value upstream and downstream,
which brings out :
T+ = 1 – R+ and T- = 1 – R- (2)
A more elaborate condition was tested by making equal the
upstream and downstream mass flows. As this procedure
introduces correction terms proportionnal to the Mach number
which did not alter significantly the results of the present study,
the simpler equation (2) is used instead.
 0
Solving equations (2) and (3) brings out a scattering matrix
1.2
with the following expression

1  p down
+
  T 1 − T  pup 
+
-0.2  =  , (4)
 pup
−  1 − T T  p down
− 
0.8    
-0.4
where the complex coefficient T alone describes the passive
0.6 behavior of an orifice. Such a description reasonably holds in
-0.6 the low Mach number range, as illustrated in Fig. 4.
0.4 Discrepancies appear for higher Mach numbers, as shown in
|T| from R+ Arg(T) from R+ Fig. 5. Assuming these discrepancies to be of little consequence
|T| from R- -0.8 Arg(T) from R- to the whistling criterion, use is made in the framework of the
0.2
|T| from T+ Arg(T) from T+ present study of a unique coefficient T, obtained by averaging
|T| from T- Arg(T) from T- the estimations from the four coefficients of the scattering
0 -1
matrix.
0 2500 5000 0 2500 5000 As can be seen in Fig. 4 and 5, the Mach number U/c has
Frequency (Hz) Frequency (Hz)
an influence upon the transmission coefficient T. Dimensional
Figure 4: derivations of the transmission coefficient T analysis can be used to determine explicitly this influence. Let
from the scattering matrix coefficients, for a sharp the pressure drop law of the orifice be written
edge orifice with t/D = 0.16, d/D = 0.63 and U/c = 0.035
∆p = ½ β ρU2, (5)

1.6 where ∆p is the difference of the steady pressures upstream and

0
downstream, β is a dimensionless coefficient function of t/D
1.4 and d/D [16, 17] and ρ is the fluid density. In harmonic regime,
-0.2 the acoustic pressures can be related to the acoustic velocity
1.2 uacou by an equation of the same type
-0.4
1 pup – pdown = B ρU uacou, (6)
0.8 -0.6 where the acoustic velocity is assumed to have the same value
upstream and downstream, where B is a non dimensional
0.6 function depending on the frequency, and where ρ and U are
-0.8
|T| from R+ Arg(T) from R+ needed by dimension analysis. In quasi-steady regime [18], Eq.
0.4
|T| from R- Arg(T) from R- (6) can be obtained by linearizing Eq. (5), which indicates that
-1 B is equal to β for low frequencies. In the low Mach number
0.2 |T| from T+ Arg(T) from T+
|T| from T- Arg(T) from T- range, one expects B not to depend on the fluid compressibility,
0 -1.2 and not to depend on the Reynolds number either. Replacing p
0 2000 4000 0 2000 4000 by p+ + p- and uacou by (p+ - p-)/ρc in (6), one gets
Frequency (Hz) Frequency (Hz)
Figure 5: derivations of the transmission coefficient T + − 2 − BU / c + −
p down − pup = ( p up − p down ). (7)
from the scattering matrix coefficients, for a sharp 2 + BU / c
edge orifice with t/D = 0.16, d/D = 0.63 and U/c = 0.088
Identifying (7) with the second equation of (9), and
expressing B as a function of a dimensionless frequency, of t/D
A second condition upon the coefficients of the scattering and of d/D, one gets
matrix can be obtained by demanding a couple of incident 1
pressure waves with equal signs and amplitudes not to interact T= , (8)
with the flow inside the orifice, because a uniform variation of U  ft t d 
1 + B , , 
the pressure has no effect if the flow is incompressible: c  U d D D 

1  T + R − 1 a non-trivial result which makes possible the collapse of

  =  +   . (3) experimental data at different flow regimes. This scaling law
1  R T − 1 works reasonably well, as illustrated in Fig. 6 where
estimations of B at different Mach numbers are gathered.

3 Copyright © 2007 by ASME

 40 Abs (B) 1.8
Abs (2T - 1)
1.6
30
1.4

Mach = 0.035 1.2

20
Mach = 0.042
Mach = 0.052 1
Mach = 0.062
0.8
Whistling
10 Mach = 0.066 frequency
Mach = 0.075
0.6 range
Mach = 0.088
0
0.4 Mach = 0.035 Mach = 0.042
0 0.1 0.2 0.3 0.4 0.5 Mach = 0.052 Mach = 0.062
Strouhal ft /U d 0.2 Mach = 0.066 Mach = 0.075
Mach = 0.088
0
2.5 Arg (B) 0 0.1 0.2 0.3 0.4 0.5

2 Strouhal ft /U d
Figure 7: amplification factor of a sharp edge orifice
1.5 with t/D = 0.16 and d/D = 0.63
Mach = 0.035
Mach = 0.042
1 Let T be expressed as a function of B using Eq. (8). The
Mach = 0.052
Mach = 0.062 criterion |2T – 1| > 1 can be rewritten as
0.5 Mach = 0.066
Mach = 0.075 U U
Mach = 0.088 1− B > 1+ B ,
0 c c
0 0.1 0.2 0.3 0.4 0.5
Strouhal ft /U d a condition which is fulfilled when UB/c is closer to –1 than to
1. Such a whistling condition can be expressed under a simpler
Figure 6: collapse of B at different Mach numbers for form as
a sharp edge orifice with t/D = 0.16 and d/D = 0.63
U 
Real B  < 0 , (10)
c 
Whistling criterion for orifices in the low Mach
number range The physical meaning of this criterion can be highlighted
The eigenvalues of (4) can easily be found and one gets considering Eq. 6; in the frequency range where B is a real
+ − + − positive number, the acoustic pressure drop is in phase with the
p down + pup = pup + p down
. (9) acoustic velocity, and Eq. 6 describes a purely dissipative
+ − + − process. If B were a real negative number instead, acoustic
p down − pup = (2T − 1)( pup − p down )
energy would be generated by the orifice. The criterion (10) is
The first equation stands for the equality of the acoustic then related to the amount of acoustic energy generated or
velocity upstream and downstream, whereas the second one dissipated by the orifice [19], and it is similar to the notion of
shows that a couple of incident pressure waves with opposite negative damping in linear instability analysis. The major
values is amplified by a factor equal to 2T – 1. The whistling interest of the criterion (10) is that it can be expressed upon B
criterion can now be formulated. Considering the orifice as an only, as the Mach number is a positive number. An orifice
amplifier and the acoustic response of the surrounding pipe as a complying with Real(B) < 0 may theoretically whistle at any
feedback, instability occurs if the ‘open-loop’ gain is higher Mach number, but the higher the Mach number, the higher the
than unity, and if some phase condition is met. As the acoustic term |2T – 1|, and the higher the probability of whistling.
response of the surrounding system is lower than unity, a As an illustration of the equivalence of the criteria upon
necessary condition for whistling to occur is that in a given |2T – 1| and upon Real(B), the Figures 6 and 7 can be
frequency range, 2T – 1 has a modulus higher than unity. An compared: whistling may occur in the range of Strouhal
illustration of the criterion is given in Fig. 7 for an orifice prone numbers ranging from 0.2 to 0.35 because |2T – 1| is higher
to whistling at Strouhal numbers varying from 0.2 to 0.3. than unity (Fig. 6) and the argument of B is higher than π/2.

4 Copyright © 2007 by ASME

RESULTS FOR SHARP EDGE ORIFICES 2
All orifices exhibit a maximum of the argument of B in the Arg (B)
range of Strouhal numbers close to .2. Depending on the d/D 1.5
and the t/D ratios, this maximum may or may not be higher
than π/2, so that whistling is not possible in all cases. 1
Sometimes, a second maximum is observed for Strouhal
Mach = 0.007
numbers close to .6. Detailed results are reproduced in the 0.5
following figures. Some discrepancies are observed as regards Mach = 0.013
the collapse of the B curves at different Mach numbers, which 0
are probably due to the uncertainty of the velocity measurement
0 0.2 0.4 0.6 0.8 1 1.2 1.4
at low flow rates.
Strouhal ft /U d
Figure 10: argument of B for a sharp edge orifice with
Low d/D case t/d = 1 and d/D = 0.33
The first series of measurements deals with a d/D ratio
2 Arg (B)
equal to 0.33. The whistling ability appears to be maximum for
t/d close to 0.5. with a Strouhal number equal to 0.25. A second
maximum appears for a thicker orifice, at a Strouhal number 1.5
equal to 1.1, whereas the first maximum decreases below π/2,
preventing whistling at a Strouhal number close to 0.2. 1
Mach = 0.004
Mach = 0.007
0.5
2 Mach = 0.01
Arg (B)
1.8 0
1.6 0 0.5 1 1.5 2 2.5 3 3.5 4
1.4
1.2 Strouhal ft /U d
1
0.8
Figure 11: argument of B for a sharp edge orifice with
0.6 t/d = 1.5 and d/D = 0.33
0.4 Mach = 0.008
0.2 Mach = 0.012
0 Medium d/D case
0 0.1 0.2 0.3 0.4 0.5 For a d/D ratio equal to .66, whistling is again possible for
Strouhal ft /U d thin orifices at a Strouhal number in the 0.2-0.3 range. The
whistling ability vanishes at this Strouhal number for a thicker
Figure 8: argument of B for a sharp edge orifice with orifice, and it appears at a Strouhal number close to 0.7.
t/d = 0.27 and d/D = 0.34
1.8 Arg (B)
2.5 Arg (B) 1.6
1.4
2
1.2
1.5 1
Mach = 0.011
0.8
1 0.6 Mach = 0.021
0.4
0.5 Mach = 0.011 Mach = 0.033
0.2
Mach = 0.016
0 0
0 0.1 0.2 0.3 0.4 0.5 0 0.2 0.4 0.6 0.8 1
Strouhal ft /U d
Figure 9: argument of B for a sharp edge orifice with Figure 12: argument of B for a sharp edge orifice with
t/d = 0.5 and d/D = 0.33 t/d = 0.15 and d/D = 0.67

5 Copyright © 2007 by ASME

 2.5 Arg (B) 2 Arg (B)

2 1.5

1.5
1
Mach = 0.042
1 Mach = 0.035 Mach = 0.042
0.5
Mach = 0.052 Mach = 0.062 Mach = 0.071
0.5 Mach = 0.066 Mach = 0.075
Mach = 0.088 0
0 0 0.2 0.4 0.6 0.8 1
0 0.2 0.4 0.6 0.8 1 Strouhal ft /U d
Strouhal ft /U d Figure 16: argument of B for a sharp edge orifice with
Figure 13: argument of B for a sharp edge orifice with t/d = 0.2 and d/D = 0.83
t/d = 0.26 and d/D = 0.67 2 Arg (B)
2 Arg (B)
1.75 1.5
1.5
1.25 1

1 Mach = 0.033 Mach = 0.066

0.75 0.5 Mach = 0.038
Mach = 0.053
0.5 Mach = 0.084
0
0.25
Mach = 0.106 0 0.5 1 1.5 2
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 Strouhal ft /U d
Strouhal ft /U d Figure 17: argument of B for a sharp edge orifice with
t/d = 0.42 and d/D = 0.8
Figure 14: argument of B for a sharp edge orifice with
t/d = 0.4 and d/D = 0.67
dmax= 12 mm
2 Arg (B)
t = 5 mm
1.5
d = 10 mm
1 Mach = 0.033 Figure 18: features of the bevel-shaped orifice
Mach = 0.054
0.5 Mach = 0.074
Mach = 0.1
RESULTS FOR A BEVEL-SHAPED ORIFICE
0 Experience shows that vortex shedding with lock-in is
0 0.2 0.4 0.6 0.8 1 1.2 1.4 dependent on the location where the flow separates [20].
Strouhal ft /U d
Bevels are hence likely to enhance or to prevent whistling. An
attempt to reproduce this effect was made using a bevel-shaped
Figure 15: argument of B for a sharp edge orifice with orifice. The bevel has an angle equal to 45°, its thickness is
t/d = 0.5 and d/D = 0.67 t/D = 0.17, the hole diameters are d/D = 0.33 and dmax/D = 0.4 ,
as shown in Fig. 18.
The presence of the bevel upstream totally removes the
High d/D case whistling ability of the orifice as shown in Fig. 19. Whistling is
For larger holes, the argument of B becomes close to π/2 still possible with a bevel downstream, as shown in Fig. 20,
for all Stouhal numbers. Whistling is not likely to appear though the whistling ability seems slightly lower than in the
except in very reverberating conditions. case without bevel, as shown in Fig 9 where the thickness and
the hole diameter are the same as in Fig. 20.

6 Copyright © 2007 by ASME

 2 Arg (B) Reflection
1
0.9 coefficient
1.5 0.8
0.7
1 0.6
0.5
Mach = 0.0041 0.4
0.5
Mach = 0.0076 0.3
0.2 Upstream, Mach = 0.05
0 0.1 Downstream, Mach = 0.05
0 0.2 0.4 0.6 0.8 1 0
Strouhal ft /U d 0 1000 2000 3000 4000 5000
Frequency (Hz)
Figure 19: argument of B for the orifice with its bevel
upstream Figure 21: modulus of the upstream and downstream
reflection coefficient in the whistling configuration
2 Arg (B) 1E+0 Pressure
1E-1 PSD
1.5 (Pa²/Hz)
1E-2
152 Mach = 0.0046
Mach = 0.0054 1E-3
1 Mach = 0.0057
Mach = 0.0077
1E-4 Mach = 0.0067
Mach = 0.0117
0.5 1E-5
Mach = 0.0118
1E-6
0
0 0.2 0.4 0.6 0.8 1 1.2 1E-7

Strouhal ft /U d 1E-8
1E-9
Figure 20: argument of B for the same orifice with its
bevel downstream 1E-10
0 1000 2000 3000 4000
Frequency (Hz)
Figure 22: pressure PSD generated by the orifice in
VALIDATION OF THE CRITERION
the lower range of Mach numbers
In order to validate the criterion, the orifice of Figure 9
was tested in reverberating conditions at different flow rates. 1E+0 Pressure
An expansion chamber was arranged 137 mm upstream of the 1E-1 PSD Mach = 0.0112
orifice, and an unflanged open pipe termination was arranged (Pa²/Hz) Mach = 0.0122
1E-2
270 mm downstream of the orifice, the corresponding Mach = 0.0131
boundary conditions being illustrated by the reflection 1E-3
coefficients in Fig. 21. 1E-4
Due to the length of the pipe from one end to the other, one
1E-5
expects the first natural frequency of the system to be close to
some 170 Hz, and the higher natural frequencies to be 1E-6
multiples of this frequency. Hence, in the range 1000-3000 Hz 1E-7
and for a Mach number varying from 0.0046 to 0.023, there
1E-8
should always exist an acoustic natural frequency such that the
Strouhal number is in the range 0.2-0.4, i.e., in the range where 1E-9
whistling is likely to occur (see Fig. 9). 1E-10
Whistling actually occurs in all cases, as shown in Fig. 22 0 1000 2000 3000 4000
to 25 where the pressure PSD downstream of the orifice is Frequency (Hz)
shown for different Mach numbers. The first natural frequency
of the pipe is 152 Hz, and whistling occurs at 1000 Hz, 1800 Figure 23: pressure PSD generated by the orifice in
Hz, 2200 Hz and 2900 Hz. the middle range of Mach numbers

7 Copyright © 2007 by ASME

1E+0 Pressure 1E+0 Pressure Mach = 0.0141
1E-1 PSD 1E-1 PSD
Mach = 0.016
(Pa²/Hz) (Pa²/Hz)
1E-2 1E-2 Mach = 0.0152
Mach = 0.017
1E-3 1E-3
1E-4
1E-4
1E-5
1E-5
1E-6
1E-6
1E-7
1E-7
1E-8
1E-8
1E-9
1E-9
1E-10
0 1000 2000 3000 4000 1E-10
Frequency (Hz) 0 1000 2000 3000 4000
Frequency (Hz)
Figure 24: pressure PSD generated by the orifice in
the middle range of Mach numbers Figure 27: whistling at 1800 Hz and 2900 Hz
1E+0 Pressure
At intermediate values of the Mach number, the whistling
1E-1 PSD
occurs alternatively at several frequencies, as illustrated in Fig.
1E-2 (Pa²/Hz)
Mach = 0.0198
26 and Fig. 27.
1E-3
Mach = 0.0209
1E-4 In order to summarize the results, the dimensionless r.m.s.
Mach = 0.0218
1E-5
values of the pressure peak are plotted as a function of the
Mach = 0.0227 Strouhal number in Fig. 28 for all flow conditions. It clearly
1E-6
appears that whistling occurs at a Strouhal number in the range
1E-7 0.2-0.4, consistent with the whistling criterion in Fig. 9. It
1E-8 should be noted that the increase in the 2300 Hz peak
1E-9 amplitude at a Strouhal number of .5 is actually due to an
harmonic of the 1000 Hz peak.
1E-10
0 1000 2000 3000 4000
Frequency (Hz)
p peak
Figure 25: pressure PSD generated by the orifice in 1.E-02
rho*c*U
the higher range of Mach numbers 1000 Hz peak
1E+0 Pressure Mach = 0.0077 1.E-03 1800 Hz peak
PSD Mach = 0.009 2300 HZ peak
1E-1
(Pa²/Hz) Mach = 0.0101 2900 Hz peak
1E-2 1.E-04

1E-3
1E-4 1.E-05

1E-5
1E-6 1.E-06

1E-7
1.E-07
1E-8
0 0.2 0.4 0.6 0.8
1E-9
Strouhal ft /U d
1E-10
0 1000 2000 3000 4000
Frequency (Hz) Figure 28: dimensionless peak amplitude vs. Strouhal
number in whistling conditions
Figure 26: whistling at 1000 Hz and 1800 Hz

8 Copyright © 2007 by ASME

CONCLUSION [9] Testud, P., Moussou, P., Hirschberg, A. and Auregan, Y.,
An acoustic criterion for evaluating the whistling ability of 2007, “Noise generated by cavitating single-hole and
an orifice was proposed. It is found that sharp edges orifices multi-hole orifices in a water pipe”, Journal of Fluids and
with a thickness to hole diameter ratio in the range 0.2-0.5 are Structures, 23, 2, 163-189
prone to whistling, for Strouhal numbers close to 0.2. [10] Auregan, Y. & Starobinski, R., 1999, “Determination of
The criterion is validated by tests: in reverberating acoustic energy dissipation/production potentiality from
conditions, an orifice generates single tone noise in the the acoustic transfer of a multiport”, Acustica, 85, 788-
Strouhal number range corresponding to the criterion. 792
The main advantage of the criterion is that it does not
[11] Morfey, C., L., 1971, “Sound transmission and generation
depend on the acoustic boundary conditions of the test rig; the
in ducts with flows”, Journal of Sound and Vibration, 14,
determination of the scattering matrix in non-whistling
37-55
conditions makes the whistling ability of the orifice in an other
acoustic surrounding predictible. It can hence be used to design [12] Abom, M., 1991, “Measurement of the scattering matrix
whistle-free orifices. of acoustical two-ports”, Mechanical Systems and Signal
Further work is needed to test the criterion at other Processing 5, 89-104
Reynolds numbers, and with other fluids so as to validate the [13] Auregan, Y. & Leroux, M., 2003, “Failures in the discrete
independence of the criterion with the Mach number. Another models for flow duct with perforations : an experimental
consists in assembling several orifices to determine the investigation”, Journal of Sound and Vibration, 265, 109-
conditions where the whistling ability is enhanced. 121
[14] Gibert R.J., 1988, “Vibrations des structures - Interactions
avec les fluides - Sources d’excitation aléatoires”, ed.
Eyrolles, Paris (in French)
REFERENCES
[15] Au-Yang, M. K., 2001, “Flow-induced vibration of power
[1] Blake, W., K., Powell, A., 1986, “The development of and process plant components : a practical workbook”,
contemporary views of flow-tone generation”, Vol. 1, ASME Press, New York
Academic Press, Orlando
[16] Idelchik, I.,E., 1996, “Handbook of hydraulic resistance”,
[2] Rienstra, S. W., Hirschberg, 2003, A. , “An Introduction ASME Books, New York
to Acoustics”, revised version of IWDE 92-06, Eindhoven
University of Technology, the Netherlands [17] Blevins, R., D., 1992, “Applied Fluid Dynamics
Handbook”, Krieger Publishing Company, Malabar,
[3] Anderson, A. B. C., 1953, “A circular-orifice number Florida
describing dependency of primary Pfeifenton frequency
on differential pressure, gas density and orifice [18] Hofmans, G., Boot, R., Durrieu, P., Auregan, Y., 2001,
“Aeroacoustic response of a slit-shaped diaphragm in a
geometry”, J. Acoust. Soc. Am., 25, pp. 626-631.
pipe at low Helmholtz number, 1: quasi steady results” ,
[4] Anderson, A. B. C., 1954, “A jet-tone orifice number for Journal of Sound and Vibration , 244 (1), 35-56
orifices of small thickness-diameter ratio”, J. Acoust. Soc.
[19] Testud, P. 2006, “Aéroacoustique des diaphragmes en
Am., 25, pp. 21-25.
conduit: sifflement et cavitation”, Ph.D. Thesis, Ecole
[5] Sato, K., and Saito, Y., 2001, “Unstable Cavitation doctorale de l’Université du Maine, Le Mans, France
Behavior in a Circular-Cylindrical Orifice Flow”,
[20] Naudascher, E., & Rockwell, D., 1994, “Flow-induced
CAV2001 (4th Symposium on Cavitation).
vibrations, an engineering guide”, Balkema, Rotterdam
[6] Moussou, P., Caillaud, S., Villouvier, V., Archer, A.,
Boyer, A. Rechu, B. Benazet, B., 2003, “Vortex shedding
of a multi-hole orifice synchronized to an acoustic cavity
in a PWR piping system”, PVP-Vol. 465, Flow-induced
vibrations, pp. 161-168
[7] Janzen, V. P. & Smith, B., 2004, “Acoustic properties of
high energy orifice assemblies” in Proc. of the 8th int.
Conf. on Flow Induced Vibration, de Langre & Axisa ed. ,
Paris
[8] Testud, P., Hirschberg, A., Moussou, P. and Auregan, Y.,
2005, “Cavitating orifice: flow regime transitions and low
frequency sound production”, Paper #71232 in Proc. of
ASME PVP 2005 Conf., July 17-21, Denver

9 Copyright © 2007 by ASME