You are on page 1of 11

Computational Fluid Dynamics-

Based Numerical Analysis


of Acoustic Attenuation and Flow
Resistance Characteristics
Chen Liu
School of Power and Energy Engineering,
of Perforated Tube Silencers
Harbin Engineering University,
The 3D time-domain computational fluid dynamics (CFD) approach is used to calculate
No. 145 Nantong Street, Nangang District,
the acoustic attenuation performance of perforated tube silencers without and with
Harbin, Heilongjiang 150001, China
flow. For the crossflow perforated tube silencer and straight-through perforated tube
silencers, the transmission loss predictions agree well with the experimental measure-
Zhenlin Ji1 ments available in the literature. Then, the 3D time-domain CFD approach is employed
School of Power and Energy Engineering, to investigate the effects of flow velocity and temperature on the acoustic attenuation
Harbin Engineering University, performance of perforated tube silencers. The numerical results demonstrated that the
No. 145 Nantong Street, Nangang District, transmission loss is increased at most frequencies for the crossflow perforated tube
Harbin, Heilongjiang 150001, China silencer as the air flow increases, while the air flow has little influence on the acoustic
e-mail: zhenlinji@yahoo.com attenuation in the plane wave range and increases the acoustic attenuation at higher
frequencies for the straight-through perforated tube silencers. Increasing the air
temperature shifts the transmission loss curve to higher frequency and lowers the
resonance peaks somewhat. The pressure drops of perforated tube silencers are pre-
dicted by performing the 3D steady flow computation using CFD. The pressure drop of
the crossflow perforated tube silencer is much higher than those of the straight-through
perforated tube silencer at the same flow conditions, and the pressure drop of the
straight-through perforated tube silencer increases gradually as the porosity increases.
[DOI: 10.1115/1.4026137]

Keywords: perforated tube silencer, time-domain CFD approach, acoustic attenuation


performance, pressure drop, air flow, 3D analysis

1 Introduction expression of perforate impedance, which was applied to the eval-


uations of transmission loss for the perforated tube silencers with-
In order to reduce the exhaust noise of internal combustion
out and with mean flow by using the segmentation approach. With
engines, silencers are extensively used in exhaust systems. The
the development of the finite element method (FEM) and the
acoustic attenuation performance and aerodynamic performance
boundary element method, these methods have been widely used
are the two important indices of silencers and are often contradic-
to calculate and analyze the acoustic attenuation performance of
tory each other. A reasonably good design of a silencer needs the
perforated tube silencers with the help of perforate impedance
balance of both indices.
models [2–9]. However, the aforementioned works have been per-
For the acoustic attenuation prediction and analysis of silencers,
formed with steady medium or mean flow, and the effect of com-
there are two kinds of methods available: the frequency-domain
plex gas flow on the acoustic attenuation behaviors of the
approach and the time-domain approach. The frequency-domain
perforated tube silencers could not be included in the computa-
approach based on the linear acoustic model solves the linear
tion. In addition, the formula of perforation impedance may not
wave equation in the frequency-domain and then determines the
be obtained accurately in the presence of complex gas flow,
acoustic attenuation performance of the silencer. The time-
which also has a direct influence on the acoustic attenuation pre-
domain approach based on the nonlinear fluid dynamics model
diction of perforated tube silencers. In view of the above disad-
solves the fundamental balance equations of mass, momentum,
vantages, the 3D time-domain CFD approach is developed to
and energy and the ideal gas equation of state by using the compu-
predict the acoustic attenuation performance of perforated tube
tational fluid dynamics method. By performing the unsteady flow
silencers. In this approach, the perforation may be simulated
computation, the time-histories of pressures can be acquired, and
directly, and the effects of complex gas flow and medium
the pressure signals in the time-domain are converted to the
viscosity on the sound propagation and attenuation in the silencer
frequency-domain by using the fast Fourier transform (FFT), and
may be included. Therefore, accurate results may be acquired by
then the acoustic attenuation performance of the silencer may be
this approach.
determined. To date, the studies on the acoustic attenuation char-
Morel et al. [10] employed the software WAVE based on the one-
acteristics of perforated tube silencers have focused on the
dimensional nonlinear fluid dynamic model to evaluate the trans-
frequency-domain methods. Sullivan [1] derived the classical
mission loss of concentric tube resonators, and their predictions
show a reasonably good agreement with the one-dimensional fre-
1
Corresponding author. quency-domain approach and experiment in the range of plane
Contributed by the Noise Control and Acoustics Division of ASME for
publication in the JOURNAL OF VIBRATION AND ACOUSTICS. Manuscript received August
wave. Then, the influence of internal geometry on the acoustic
6, 2012; final manuscript received November 14, 2013; published online December attenuation performance of the silencers was discussed in detail.
24, 2013. Assoc. Editor: Lonny Thompson. A similar simulation code GT-Power was used to build two

Journal of Vibration and Acoustics Copyright V


C 2014 by ASME APRIL 2014, Vol. 136 / 021006-1

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Fig. 2 Straight-through perforated tube silencer
Fig. 1 Crossflow perforated tube silencer

coupled models of the engine and intake/exhaust elements by


Morel et al. [11], and the instantaneous pressure and sound pres-
sure level in several exhaust mufflers were calculated with these
coupled models. Comparisons to engine data demonstrated that
the simulated exhaust pressure dynamics reasonably match the
experiment measurements. Selamet et al. [12] proposed a one- Fig. 3 Scheme for transmission loss calculation of a silencer
dimensional time-domain technique to analyze the acoustic
attenuation behavior of several basic silencing devices, and their five perforation patterns by experiments and CFD simulations,
predictions show good agreements with the experimental data. and the overall trend of the numerical results matched well with
Later, this approach was extended to calculate the transmission measurements.
loss of multipass perforated tube silencers by Dickey et al., and The objectives of this study are then: (1) to develop a 3D time-
good correlations between the numerical results and experiments domain CFD approach to determine the transmission loss of cross-
are obtained in the plane wave range [13]. Middelberg et al. [14] flow perforated tube silencer and straight-through perforated tube
developed a two-dimensional axisymmetric CFD method to calcu- silencers without and with air flow, (2) to examine the effects of
late the transmission loss of circular expansion chamber mufflers, flow velocity and temperature on the acoustic attenuation per-
and the predictions agree well with the available experimental formance of the perforated tube silencers, (3) to calculate and ana-
data in the literature. Broatch et al. [15] employed the 3D time-do- lyze the pressure drops of the perforated tube silencers, and (4) to
main approach to calculate the acoustic attenuation characteristics introduce a trade-off analysis between the acoustic attenuation
of a simple expansion chamber and a reversing chamber muffler, performance and aerodynamic performance of the perforated tube
and their time-domain numerical results agree fairly well with the silencers.
measured data and FEM predictions in the absence of air flow. The crossflow perforated tube silencer used in Ref. [1] and
However, the calculated transmission loss at low Mach number three straight-through perforated tube silencers used in Ref. [19]
mean flow (M ¼ 0.07) is not validated with experiment. The trans- will be considered in the present work. The crossflow perforated
mission loss of straight-through perforated tube reactive silencers tube silencer is shown in Fig. 1, and the dimensions are:
and resonators was evaluated by using the 3D time-domain CFD D ¼ 101.6 mm for the inner diameter of chamber, d ¼ 49.3 mm for
approach in Ref. [16], and the predictions show the excellent the inner diameter of perforated tube, l1 ¼ l2 ¼ 128.6 mm for the
agreements with the published experimental data in the absence of lengths of inlet and outlet perforated tubes. The wall thickness of
air flow. However, for the cases with mean flow, the accuracy of the perforated tube is 0.81 mm, and each perforated tube is drilled
the 3D time-domain CFD approach is not perfect enough. In order 160 holes with 2.49 mm diameter resulting in the porosity of
to reduce the computational cost, the nonreflecting boundary con- 3.9%. Figure 2 illustrates the configuration of the straight-through
dition (NRBC) may be applied at the inlet and outlet of the com- perforated tube silencer, which has three different perforation pat-
putational model. Reference [17] demonstrated that the NRBCs terns. The dimensions are: D ¼ 110 mm for the inner diameter of
available in software FLUENT may produce some spurious reflec- expansion chamber, l ¼ 200 mm for the length of expansion cham-
tions. Then, the improved NRBCs of the inlet and outlet were ber, d ¼ 32 mm for the inner diameter of perforated tube,
incorporated in the combined 1D-3D simulations to evaluate the tw ¼ 2 mm for the thickness of perforated tube, the diameter of
transmission loss of a rectangular muffler with stationary medium, hole dh ¼ 4 mm and porosity r ¼ 4.7% for silencer S1, dh ¼ 6 mm
and good correlations between the predictions and experiments and r ¼ 9.0% for silencer S2, and dh ¼ 8 mm and r ¼ 14.7% for
are obvious throughout the frequency range of interest. However, silencer S3.
there are also some tiny spurious reflections under the condition
of the improved NRBCs. Therefore, the present work does not
resort to the NRBC in order to exclude the effect of spurious 2 Computational Methods
reflections.
The aerodynamic performance index specifies the maximum 2.1 Acoustic Attenuation Calculation. The model for trans-
allowable pressure drop through the silencer. It is essential to the mission loss calculation by using the 3D time-domain CFD
engine performance that the installed exhaust system does not approach is shown in Fig. 3. In the present work, the commercial
exceed the manufacturer’s maximum exhaust backpressure limit. CFD software FLUENT is used as the computational tool, which is
The pressure drop of exhaust system includes losses due to piping, based on the finite volume method. Under the air flow condition,
silencer, and termination. High backpressure can cause a decrease the procedures of the 3D time-domain CFD approach are as fol-
in engine efficiency or increase in fuel consumption, overheating, lows. Firstly, the inlet and outlet of the silencer are connected
and may result in a complete shutdown of the engine, potentially with long tubes, and two pressure monitoring points are placed in
causing significant damage. Traditional head loss calculations uti- the appropriate location of the upstream and downstream tubes.
lizing standardized coefficients for sudden contraction and expan- The mass flow boundary condition is applied at the inlet of the
sion of fluids can be used to approximate the pressure drop upstream tube, and the steady calculation is performed by FLUENT.
through a silencer. More complex silencer internal structures After the completion of the steady calculation, the flow distribu-
should be analyzed using CFD where traditional empirical calcu- tion in the computational model could be obtained, which is used
lations or assumptions may lead to inaccurate results. Lee and Ih as the initial values for two unsteady computations in the next.
[18] investigated the pressure drop of concentric silencers with Then, an impulse signal (acoustic excitation) superimposed on a

021006-2 / Vol. 136, APRIL 2014 Transactions of the ASME

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Fig. 5 (a) Time history of the pressure at downstream monitor-
Fig. 4 (a) Time history of the pressure at upstream monitoring
ing point with acoustic excitation, (b) time history of the pres-
point with acoustic excitation, (b) time history of the pressure at
sure at downstream monitoring point without acoustic
upstream monitoring point without acoustic excitation, and (c)
excitation, and (c) acoustic signals at downstream monitoring
acoustic signals at upstream monitoring point in time-domain
point in time-domain

constant mass flow is imposed at the inlet of the upstream tube, pressures may be obtained subsequently. Finally, the incident and
and the time histories of the pressures at the two monitoring points transmitted acoustic pressure signals in the time-domain are trans-
can be acquired through the unsteady CFD computation. Because formed into the frequency-domain by FFT, and then the transmis-
the pressures recorded at the two monitoring points are fluid pres- sion loss of the silencer could be determined by
sures including the acoustic signals, the unsteady CFD computa- "    #
tion needs to be run again without the acoustic excitation imposed Ai 1=2 pin 
at the tube inlet, and the time histories of pressures are recorded at TL ¼ 20 log10 p  (1)
the upstream and downstream monitoring points. The differences Ao tr

between the two unsteady computational results are the acoustic


pressures in the time-domain. After the interference signals where Pin and Ptr are the sound pressures of incident and transmit-
caused by wave reflections are removed from the pressure signals ted waves, respectively, and Ai and Ao are the cross-sectional areas
at the two monitoring points, incident and transmitted sound of the inlet and outlet of the silencer, respectively [20].

Journal of Vibration and Acoustics APRIL 2014, Vol. 136 / 021006-3

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Fig. 8 Comparison of transmission loss results of crossflow
perforated tube silencer (v 5 0 m/s, T 5 293 K)

Fig. 6 (a) The incident sound pressure in the frequency-


domain, and (b) the transmitted sound pressure in the fre-
quency-domain

Fig. 9 Comparison of transmission loss results of crossflow


perforated tube silencer (v 5 17 m/s, T 5 347 K)

Fig. 7 Scheme for pressure drop calculation of straight-


through perforated tube silencer

As the impulse signal arrives at the silencer, the reflected wave


is formed subsequently, and the transmitted wave slowly attenu-
ates to a steady state at the downstream monitoring point due to
the multiple reflections of the sound waves inside the silencer.
The transmitted signal recorded at the downstream monitoring
point may be interfered by the reflected signals from the tube ter-
mination. The incident signal recorded at the upstream monitoring Fig. 10 Effect of flow velocity on transmission loss of cross-
point also may be interfered by the reflected signals from the si- flow perforated tube silencer (T 5 347 K)
lencer. Therefore, the lengths of the upstream and downstream
tubes and the location of two monitoring points must be selected have attenuated to the stable conditions. Then, the incident and
reasonably so as to allow the reflected signals from the silencer transmitted signals can be isolated from the unwanted reflections
and downstream tube termination to arrive at the monitoring by a simple rectangular window function. It could be found from
points at a time after which the incident and transmitted signals many tentative calculations that to acquire reasonable predictions,

021006-4 / Vol. 136, APRIL 2014 Transactions of the ASME

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


of holes are taken in the perforated region because of the stag-
gered distribution of perforations. For the straight-through perfo-
rated tube silencers, a column of holes are built in the
computational models in view of the uniform distribution of per-
forations. In the computational models, the mesh has been graded
such that regions of interest and large gradients, such as the center
of perforated tube and perforation region, have denser mesh. The
grid size of the expansion chamber increases slightly, and the
upstream and downstream tubes have less dense mesh.
For the sake of acquiring the initial distribution of flow field in
Fig. 11 The location of profile A1 in crossflow perforated tube the computational model, the steady flow calculation is carried
silencer out. The pressure-based implicit solver is employed to solve the
governing equations of continuity, momentum, and energy. The
pressure-velocity coupling scheme is the SIMPLEC algorithm,
the lengths of inlet and outlet tubes should exceed 15 times the and the second order scheme is chosen for the spatial discretiza-
length of the silencer. The choice for the location of two monitor- tion. The Realizable k-e turbulence model is used for closure. The
ing points follows the method of Singh and Katra [21]. detailed description of software FLUENT may be found in the FLU-
When the 3D time-domain CFD approach is employed to calcu- ENT user guide [22]. In the present study, the working medium is
late the transmission loss of perforated tube silencers with air air with the density satisfying the ideal gas law, and the other
flow, the grid size must be small enough to capture accurately the physical properties are determined by air temperature. The bound-
pressure disturbance propagation of a pulse superimposed on a ary conditions are set as:
high static pressure of the fluid medium, which could increase
considerably the grid number and time consumption. In order to
reduce the computational cost, the rotationally periodic boundary (1) The mass flow inlet boundary condition is defined at the
condition is used in the computational model. As a result, only inlet of the upstream tube.
one column or two columns of holes are built in the perforated (2) The pressure outlet boundary condition is specified as one
regions. For the crossflow perforated tube silencer, two columns standard atmospheric pressure.

Fig. 12 Velocity vectors on profile A1: (a) v 5 17 m/s, T 5 347 K and (b) v 5 34 m/s, T 5 347 K

Journal of Vibration and Acoustics APRIL 2014, Vol. 136 / 021006-5

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Fig. 13 Effect of air temperature on transmission loss of
crossflow perforated tube silencer (v 5 34 m/s)

(3) The walls are considered as stationary, adiabatic, and no-


slip.
(4) The rotationally periodic boundary condition is used.
After the convergence of steady calculation, the flow distribu-
tion is used as the initial values to carry on the two unsteady cal-
culation runs in the next. In the transient calculation, the second-
order implicit scheme is selected for the temporal discretization,
and the pressure-velocity coupling scheme is changed to the PISO
algorithm. When the acoustic excitation is enforced in the
unsteady calculation, an impulse signal superimposed on a con-
stant mass flow is imposed at the inlet of the upstream tube. A
half period of sinusoidal wave of the frequency 4000 Hz with
mass flow amplitude of 0.3 kg/(m2s) is used as the impulse signal.
After the impulse signal is completed, the tube inlet remains the
constant mass flow for the remainder of the computation. For the
second unsteady calculation without the acoustic excitation, the
boundary conditions are the same as that of steady flow
computation.
Figures 4 and 5 are the incident signal and transmitted signal at
the two monitoring points for the straight-through perforated tube
silencer S1 with Mach number M ¼ 0.1 and air temperature
T ¼ 288 K. When an acoustic excitation is imposed at the inlet of
the upstream tube, the time histories of the pressures at the two
monitoring points are depicted in Figs. 4(a) and 5(a). Figures 4(b)
and 5(b) illustrate the pressure time histories at the same monitor-
ing points without acoustic excitation. The differences between
the two computational results are the acoustic signals in the time-
domain, which are depicted in Figs. 4(c) and 5(c).
Attention must be paid that the time histories of the acoustic Fig. 14 Comparison of transmission loss results (M 5 0.1,
T 5 288 K): (a) silencer S1, (b) silencer S2, and (c) silencer S3
pressures to be Fourier transformed are not contaminated by
unwanted reflections. Therefore, the time histories of the acoustic
pressures at the two monitoring points need cutting off by rectan-
gular window function so as to obtain the isolated incident and
transmitted pulses. Then, the incident and transmitted pulse
signals in the time-domain are transformed into the acoustic model is split into several parts to generate the mesh individually.
pressures in frequency-domain by using the FFT as shown in The tetrahedral mesh with a grid size of 1.5 mm is applied at the
perforation region, and the rest of the computational model has a
Fig. 6.
In the present work, the time step is 4 ls, and the corresponding structured mesh of 1.5 mm.
The velocity inlet and pressure outlet boundary conditions are
sampling frequency is 250 kHz, which is much higher than the
maximum of the effective frequency in the acoustic excitation applied to the inlet and outlet of the computational model. The
signal, so that the sampling law is satisfied. pressure drops of perforated tube silencers with the flow velocity
10 m/s to 50 m/s will be calculated in Sec. 3, so the air used as the
working medium could be considered as incompressible. In the
2.2 Pressure Drop Calculation. Figure 7 illustrates the com- computations, the air temperature is 293 K, which determines the
putational model of the pressure drop for a straight-through perfo- physical properties of the working fluid. The other parameters
rated tube silencer. The lengths of both inlet and outlet tubes are used to calculate the pressure drop by using the software FLUENT
set as 100 mm, and a half model with symmetry boundary condi- are the same as those in the steady flow computation of the 3D
tion is built to calculate the pressure drop. The computational time-domain CFD approach that is introduced in Sec. 2.1.

021006-6 / Vol. 136, APRIL 2014 Transactions of the ASME

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Fig. 15 Comparison of transmission loss results of silencer
S1 (M 5 0.2, T 5 288 K)

3 Results and Analyses


3.1 Acoustic Attenuation Performance. In order to validate
the accuracy of the 3D time-domain CFD approach, the published
experimental results of transmission loss for the crossflow perfo-
rated tube silencer in Ref. [1] and three straight-through perfo-
rated tube silencers in Ref. [19] mentioned in Sec. 1 are compared
with the present CFD numerical predictions in this subsection.
In the computational model of transmission loss for the cross-
flow perforated tube silencer, the lengths of upstream and down-
stream tubes are chosen as 4000 mm and 6000 mm, respectively,
the distance between the upstream monitoring point and silencer
is l1 ¼ 1350 mm, and the distance between the downstream moni-
toring point and silencer is l2 ¼ 850 mm. The inlet and outlet tubes
are discretized with a structured mesh of less than 4 mm, the si-
lencer is discretized with an unstructured mesh of less than 1 mm,
and the region near the perforations is refined further.
Figure 8 compares the transmission loss results of the crossflow
perforated tube silencer from the present 3D time-domain CFD
approach and experimental measurement for the case of flow ve-
locity v ¼ 0 m/s and air temperature T ¼ 293 K. The present pre-
dictions match closely the experimental results throughout the
frequency range of interest.
Figure 9 shows the transmission loss results from the present
time-domain CFD approach and experimental measurement for
the silencer with v ¼ 17 m/s and T ¼ 347 K, and good agreement
between the numerical results and measured data are observed.
Figure 10 compares the transmission loss predictions from the
3D time-domain CFD approach for the crossflow perforated tube Fig. 16 Effects of flow velocity on transmission loss of
silencer at different flow velocities with fixed temperature straight-through perforated tube silencers (T 5 347 K): (a) si-
T ¼ 347 K. It may be seen that the transmission loss is increased at lencer S1, (b) silencer S2, and (c) silencer S3
most frequencies as the mean flow velocity increases, which may
be explained in that the acoustic resistance of perforations
increases and the effective flow-path areas in the holes decrease as
the flow velocity increasing.
In order to examine the influence of air flow (especially in the
holes) on the acoustic behavior of the crossflow perforated tube si-
lencer, a profile A1 is built as shown in Fig. 11. Considering that
the pulse signal has negligible effect on the flow distribution
inside the silencer in the presence of air flow, the velocity vectors
on the profile A1 are attained approximately from the steady flow
computations (without the acoustic excitation). Figure 12 depicts
the flow velocity vectors on the profile A1 at v ¼ 17 and 34 m/s
with fixed air temperature T ¼ 347 K, respectively. The distribu-
tion of flow velocity is anisotropic and nonhomogeneous inside
the silencer. Therefore, it is not accurate enough to consider the
actual gas flow as mean flow. In addition, the flow velocity in the
holes increases as the flow velocity in the tube increases, which Fig. 17 The location of profile A2 in silencer S1

Journal of Vibration and Acoustics APRIL 2014, Vol. 136 / 021006-7

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Fig. 18 Velocity vectors on profile A2 for silencer S1: (a) v 5 18.64 m/s, T 5 347 K and (b)
v 5 37.28 m/s, T 5 347 K

leads to the increment of perforate acoustic resistance and the M ¼ 0.1 and air temperature T ¼ 288 K. It may be seen that the
improvement of acoustic attenuation performance of the silencer. present time-domain CFD predictions agree well with the meas-
The effect of air temperature on the acoustic attenuation per- urements throughout the frequency range of interest.
formance of the crossflow perforated tube silencer is examined The calculated and measured transmission loss for silencer S1
next by fixing all physical parameters. Figure 13 shows the trans- with M ¼ 0.2 and T ¼ 288 K is depicted in Fig. 15, which exhibits
mission loss predictions at three different air temperatures. It is an excellent agreement between them.
observed that the transmission loss curve is shifted to a higher fre- The effects of flow velocity on the transmission loss of the three
quency, and the resonance peaks are lowered as the temperature straight-through perforated tube silencers are shown in Fig. 16.
increases. The lowering of resonance peaks may be attributed to The air flow has little influence on the acoustic attenuation in the
the fact that the higher temperature enhances the air viscous plane wave range and leads to the irregular variation of the acous-
damping. tic attenuation after the first resonant peak in transmission loss
The acoustic attenuation characteristics of the aforementioned curve. Generally speaking, the air flow increases the acoustic
straight-through perforated tube silencers with three perforation attenuation of the straight-through perforated tube silencer at
patterns are considered next. In the computational models of higher frequencies mainly.
transmission loss for the three silencers, the lengths of upstream In order to analyze the effect of air flow on the acoustic behav-
and downstream tubes are chosen as 3000 mm and 5000 mm, ior of the straight-through perforated tube silencer, a profile A2 of
respectively. The distance between the upstream monitoring point the silencer S1 is created and shown in Fig. 17. Figure 18 shows
and silencer is l1 ¼ 900 mm, and the distance between the down- the velocity vectors at different flow velocities with fixed air tem-
stream monitoring point and silencer is l2 ¼ 500 mm. The inlet perature T ¼ 347 K, which are obtained from the steady flow com-
and outlet tubes are discretized with structured mesh of less than putations. As the flow velocity changes from 18.64 m/s to
2 mm, the silencers are discretized with unstructured mesh of less 37.28 m/s, the jet velocity through holes increases gradually; how-
than 0.8 mm, and the region near the perforation is refined further. ever, the flow velocity in the holes is much lower than that in the
Figure 14 compares the present time-domain CFD predictions tube. In addition, there is crossflow in the perforations, and the
and experimental results of transmission loss for the three direction and magnitude of crossflow velocity through perforates
straight-through perforated tube silencers with Mach number are different along the axial direction. In deriving the empirical

021006-8 / Vol. 136, APRIL 2014 Transactions of the ASME

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Table 1 Static pressure difference of straight-through perfo-
rated tube silencer (T 5 299 K)

Mach number Experiment Prediction in Ref. [18] Prediction


(Pa) (Pa) (Pa)

0.085 281.6 250.9 251


0.17 1074 870.7 878.8

Fig. 20 Pressure loss results of crossflow perforated tube si-


lencer (T 5 293 K)

Fig. 19 Effects of air temperature on transmission loss of


straight-through perforated tube silencers (M 5 0.1): (a) silencer Fig. 21 Comparison of pressure loss results of three straight-
S1, (b) silencer S2, and (c) silencer S3 through perforated tube silencers (T 5 293 K)

Figure 19 shows the transmission loss predictions of straight-


expression of perforated impedance under grazing flow, the cross- through perforated tube silencers from the present time-domain
flow velocity is usually assumed to be the same and very small at CFD approach with fixed Mach number M ¼ 0.1 and different air
all positions. The above observation illustrated that this assump- temperatures. It may be seen that, as the air temperature rises
tion may affect the computational accuracy of frequency-domain from 347 K to 747 K, the transmission loss curves of straight-
method with use of the empirical expression of perforated through perforated tube silencers move to a higher frequency.
impedance.
It may be seen that by comparing Figs. 10 and 16, the crossflow
perforated tube silencer is more sensitive to the flow velocity than 3.2 Aerodynamic Performance. In order to validate the ac-
the straight-through perforated tube silencer since the flow veloc- curacy of the pressure drop prediction of perforated tube silencers
ity in holes of the crossflow perforated tube silencer is much by using the CFD steady flow computation, the straight-through
higher than that of the straight-through perforated tube silencer as perforated tube silencer with uniform perforation in Ref. [18] is
illustrated in Figs. 12 and 18. The flow velocity may influence the considered. The scheme of this silencer is similar to that as shown
acoustic resistance significantly and then affect the acoustic in Fig. 2, and the dimensions are D ¼ 110 mm for the inner diame-
attenuation performance of the perforated tube silencer. ter of the expansion chamber, l ¼ 200 mm for the length of the

Journal of Vibration and Acoustics APRIL 2014, Vol. 136 / 021006-9

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


Fig. 22 Contours of velocity magnitude for crossflow perfo- Fig. 24 Contours of turbulent kinetic energy for crossflow per-
rated tube silencer (v 5 50 m/s, T 5 293 K) forated tube silencer (v 5 50 m/s, T 5 293 K)

used to measure the midstream flow velocity. The average value


of the Mach number in the test duct was assumed to be 85% of the
midstream Mach number. The discrepancies between the numeri-
cal predictions and experiments of static pressure difference may
be attributed to the errors involved in the measurement of mid-
stream flow velocity and the determination of average value of
Mach number in the test duct.
The pressure drops of crossflow perforated tube silencer and
three straight-through perforated tube silencers are examined
next, and the total pressure difference is used to describe the pres-
sure drop. Figures 20 and 21 depict the pressure drop predictions
of the crossflow perforated tube silencer and three straight-
through perforated tube silencers. It could be found that the pres-
sure drop increases with flow velocity, and there is a parabolic
relationship between them. The numerical predictions show that
the pressure drop of crossflow perforated tube silencer is much
higher than those of straight-through perforated tube silencers.
Because the baffle installed in the center of the perforated tube for
the crossflow perforated tube silencer causes the air flow to divert
into the chamber through the perforations and then back again
into the tube, the jet velocity in the perforations of the crossflow
perforated tube silencer is much higher than those of straight-
through perforated tube silencers, which may be found by com-
paring the Figs. 22 and 23. The abrupt change of flow velocity
inside the crossflow perforated tube silencer, particularly near the
perforation regions, increases the turbulent kinetic energy and
cause additional pressure drop, as shown in Fig. 24.
In addition, the porosity seems to have a modest influence on
the pressure drops of the straight-through perforated tube
silencers. Figure 21 illustrates that the pressure drop increases
gradually with the porosity of perforation. Increasing the open
area of perforations would have a negative influence on the uni-
formity of flow in the perforated tube. Under the action of pres-
sure difference between the chamber and perforated tube, the
higher porosity may lead to more airstream flowing into the cham-
ber from the holes and then moving out from the chamber to the
perforated tube and, subsequently, enhances the jet and turbulent
kinetic energy near the holes as shown in Figs. 23 and 25.

3.3 Design Considerations. The crossflow perforated tube si-


Fig. 23 Contours of velocity magnitude for straight-through lencer may achieve the very good acoustic attenuation perform-
perforated tube silencers (v 5 50 m/s, T 5 293 K): (a) silencer S1, ance in the presence of air flow. However, high flow velocity
(b) silencer S2, and (c) silencer S3
leads to much high pressure drop, which restricts the application
of this silencer in the case with high flow velocity. However, the
expansion chamber, d ¼ 32 mm for the inner diameter of the per- pressure drop of the crossflow perforated tube silencer may be
forated tube, tw ¼ 2 mm for the thickness of the perforated tube, reduced by increasing the open area of perforation, which may
the diameter of hole dh ¼ 4 mm, and porosity r ¼ 10.3%. The air lead to a lower acoustic attenuation in some level. Therefore, the
temperature in the test duct was T ¼ 299 K during experiments. accurate predictions of acoustic attenuation performance and aero-
Table 1 shows the measured and calculated results of the static dynamic performance are necessary for the silencer design to
pressure difference for the straight-through perforated tube si- meet the requirements for both indices.
lencer. The present CFD predictions agree well with the numerical The straight-through perforated tube silencer exhibits the dome
results in Ref. [18], but both of them are less than the measured attenuation behavior in the low frequency range and the axial
data. In the experiment, the pitot-tube and micromanometer were resonances in the high frequency range. The effect of porosity is

021006-10 / Vol. 136, APRIL 2014 Transactions of the ASME

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms


silencers could be obtained by using the 3D time-domain CFD
approach. The numerical results demonstrated that the flow velocity
and temperature may affect the acoustic attenuation characteristics
of perforated tube silencers. The transmission loss of the crossflow
perforated tube silencer is increased at most frequencies as the air
flow increases while the air flow has little influence on the acoustic
attenuation in the plane wave range and increases the acoustic
attenuation at higher frequencies for the straight-through perforated
tube silencer. Increasing the air temperature shifts the transmission
loss curve to higher frequency and lowers the resonance peaks
somewhat. The pressure drops of perforated tube silencers are cal-
culated by using the 3D CFD approach. The numerical results
revealed that the pressure drop of the crossflow perforated tube
silencer is much higher than those of the straight-through perforated
tube silencers under the same flow conditions. The pressure drop of
the straight-through perforated tube silencer increases gradually as
the porosity increasing.

Acknowledgment
This work has been supported by the Natural Science Founda-
tion of China (No. 11174065). The authors would like to thank
Professor Ih at The Korea Advanced Institute of Science and
Technology for providing their experimental results.

References
[1] Sullivan, J. W., 1979, “A Method for Modeling Perforated Tube Muffler Com-
ponents. II. Applications,” J. Acoust. Soc. Am., 66(3), pp. 779–788.
[2] Ross, D. F., 1981, “A Finite Element Analysis of Perforated Component Acous-
tic Systems,” J. Sound Vib., 79(1), pp. 133–143.
[3] Fang, Z., and Ji, Z. L., 2012, “Finite Element Analysis of Transversal Modes
and Acoustic Attenuation Characteristics of Perforated Tube Silencers,” Noise
Control Eng. J., 60(3), pp. 340–349.
[4] Wu, T. W., and Wan, G. C., 1996, “Muffler Performance Studies Using a Direct
Mixed-Body Boundary Element Method and a Three-Point Method for Evaluat-
ing Transmission Loss,” ASME J. Vibr. Acoust., 118(3), pp. 479–484.
[5] Wang, C. N., and Liao, C. Y., 1998, “Boundary Integral Equation Method for
Evaluating the Performance of Straight-Through Resonator With Mean Flow,”
J. Sound Vib., 216(2), pp. 281–294.
[6] Ji, Z. L., and Selamet, A., 2000, “Boundary Element Analysis of Three-Pass
Perforated Duct Mufflers,” Noise Control Eng. J., 48(5), pp. 151–156.
[7] Lee, I., and Selamet, A., 2006, “Impact of Perforation Impedance on the Trans-
Fig. 25 Contours of turbulent kinetic energy for straight- mission Loss of Reactive and Dissipative Silencers,” J. Acoust. Soc. Am.,
through perforated tube silencers (v 5 50 m/s, T 5 293 K): (a) si- 120(6), pp. 3706–3713.
[8] Ji, Z. L., 2010, “Boundary Element Acoustic Analysis of Hybrid Expansion
lencer S1, (b) silencer S2, and (c) silencer S3 Chamber Silencers With Perforated Facing,” Eng. Anal. Boundary Elem.,
34(7), pp. 690–696.
minimal on the lower frequency acoustic attenuation while the [9] Ji, Z. L., and Fang, Z., 2011, “On the Acoustic Impedance of Perforates and Its
acoustic behavior at higher frequencies is strongly dependent on the Application to Acoustic Attenuation Predictions for Perforated Tube Silencers,”
Inter-Noise 2011 Conference, Osaka, Japan, September 4–7.
porosity of perforation. Increasing the porosity extends the effective [10] Morel, T., Morel, J., and Blaser, D., 1991, “Fluid Dynamic and Acoustic Mod-
attenuation band to a higher frequency. Simultaneously, the pressure eling of Concentric-Tube Resonators/Silencers,” SAE Technical Paper 910072.
drop increases gradually with the porosity of perforation. When the [11] Morel, T., Silvestri, J., Goerg, K., and Jebasinski, R., 1999, “Modeling of
porosity exceeds 25%, the acoustic attenuation characteristic of the Engine Exhaust Acoustics,” SAE Technical Paper 1999-01-1665.
[12] Selamet, A., Dickey, N. S., and Novak, J. M., 1995, “A Time-Domain Computa-
straight-through perforated tube silencer is similar to that of a simple tional Simulation of Acoustic Silencers,” ASME J. Vibr. Acoust., 117, pp. 323–331.
expansion chamber. Therefore, the porosity may be increased mod- [13] Dickey, N. S., Selamet, A., and Novak, J. M., 1998, “Multi-Pass Perforated Tube
estly to improve the acoustic attenuation behavior of the straight- Silencers: A Computational Approach,” J. Sound Vib., 211(3), pp. 435–448.
through perforated tube silencer in the mid to high frequency range. [14] Middelberg, J. M., Barber, T. J., Leong, S. S., Byrne, K. P., and Leonardi, E.,
2004, “CFD Analysis of the Acoustic and Mean Flow Performance of Simple
Expansion Chamber Mufflers,” ASME Paper No. IMECE2004-61371.
4 Conclusions [15] Broatch, A., Margot, X., and Gil, A., 2005, “A CFD Approach to the Computa-
tion of the Acoustic Response of Exhaust Mufflers,” J. Comput. Acoust., 13(2),
The 3D time-domain CFD approach is employed to calculate the pp. 301–316.
acoustic attenuation performance of the crossflow perforated tube [16] Ji, Z. L., Xu, H. S., and Kang, Z. X., 2010, “Influence of Mean Flow on Acous-
tic Attenuation Performance of Straight-Through Perforated Tube Reactive
silencer and the straight-through perforated tube silencers without Silencers and Resonators,” Noise Control Eng. J., 58(1), pp. 12–17.
and with air flow, and the numerical predictions of transmission [17] Torregrosa, A. J., Fajardo, P., Gil, A., and Navarro, R., 2012, “Development of
loss of the silencers are compared with the experimental measure- Non-Reflecting Boundary Condition for Application in 3D Computational Fluid
ments available in the literature. The comparisons showed that the Dynamic Codes,” Eng. Appl. Comp. Fluid Mech., 6(3), pp. 447–460.
[18] Lee, S. H., and Ih, J. G., 2008, “Effect of Non-Uniform Perforation in the Long
present 3D time-domain CFD approach could predict accurately Concentric Resonator on Transmission Loss and Back Pressure,” J. Sound Vib.,
the acoustic attenuation performance of the perforated tube 311, pp. 280–296.
silencers without and with air flow. Compared with the frequency- [19] Lee, S. H., and Ih, J. G., 2003, “Empirical Model of the Acoustic Impedance of a
domain approach, the influences of complex gas flow and viscosity Circular Orifice in Grazing Mean Flow,” J. Acoust. Soc. Am., 114(1), pp. 98–113.
[20] Munjal, M. L., 1987, Acoustic of Ducts and Mufflers, Wiley, New York, Chap. 2.
on the sound propagation and attenuation in the silencers may be [21] Singh, R., and Katra, T., 1978, “Development of an Impulse Technique for
included in the time-domain computations. Therefore, the accurate Measurement of Muffler Characteristics,” J. Sound Vib., 56(2), pp. 279–298.
predictions of acoustic attenuation performance of perforated tube [22] Fluent, 2006, Fluent 6.3 User’s Guide, Fluent Inc, New York.

Journal of Vibration and Acoustics APRIL 2014, Vol. 136 / 021006-11

Downloaded From: http://vibrationacoustics.asmedigitalcollection.asme.org/ on 05/21/2015 Terms of Use: http://asme.org/terms

You might also like