Professional Documents
Culture Documents
Peng Wang
University of Kentucky
Xin Hua
Faurecia Emissions Control Technologies
CITATION: Wang, P., Hua, X., Wu, T., and Herrin, D., "Methods for Determining Muffler Transmission Loss in Octave Bands," SAE
Technical Paper 2016-01-1317, 2016, doi:10.4271/2016-01-1317.
Copyright © 2016 SAE International
Abstract inlet duct at all frequencies. While the first two methods still rely on
the plane-wave assumption in the inlet and outlet ducts, the direct
Insertion loss in one-third or octave bands is widely used in industry
conversion method can be used for large silencers at higher
to assess the performance of large silencers and mufflers. However,
frequencies when higher-order cross modes propagate along with the
there is no standard procedure for determining the transmission loss
plane waves in the inlet and outlet ducts.
in one-third or octave bands using measured data or simulation. In
this paper, assuming that the source is broadband, three different
Two test cases are provided in this paper to demonstrate the three
approaches to convert the narrowband transmission loss data into
methods. The first test case is a simple expansion chamber at
one-third and octave bands are investigated. Each method is
frequencies below the plane-wave cutoff of the inlet and outlet ducts.
described in detail. To validate the three different approaches,
Below the cutoff, all three methods should agree with one another if
narrowband transmission loss data of a simple expansion chamber
the incident wave amplitude is assumed to be a constant across the
and a large bar silencer is converted into one-third and octave bands,
frequency range. The second test case is a large bar silencer for
and results obtained from the three approaches are demonstrated to
broadband noise reduction. The plane-wave cutoff of this particular
agree well with one another.
bar silencer example is at 612 Hz only, while the TL analysis is
carried out up to 8000 Hz. The direct conversion method is still
Introduction applicable above the plane-wave cutoff.
Measurements of insertion loss (IL) and noise reduction (NR) for
large mufflers and silencers are normally performed in one-third or Octave and One-Third Octave Bands
octave bands. On the other hand, transmission loss (TL) must be first
Each octave band filter has a fixed center frequency and is twice as
determined in narrow bands because TL computation requires
wide as the one before it, as shown in Fig. 1. In other words, the
additional post-processing of the measured or computed sound
bands are related by the following relationship
pressure data. In many cases, it is still desirable to convert the
narrow-band TL into one-third or octave bands in order to compare to
the measured IL or NR. However, there has been no standard
procedure for determining the TL in one-third or octave bands using
(1)
measured data or simulation. In this paper, we compare three
different methods for determining TL in one-third or octave bands. where fu and fu are the upper limit and lower limit frequency of the
They are (i) the wave decomposition method, (ii) the equivalent IL band, respectively.
method, and (iii) the direct conversion method, respectively. All three
methods assume a constant amplitude incident sound wave in the
Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018
(2)
(3)
(4)
(5)
(6)
Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018
(7)
where S1 and S2 are the cross-sectional areas of the inlet and outlet,
respectively. (b).
Wi and Wt in narrow bands are then integrated (or simply summed) Figure 4. (a) Typical duct silencing system; (b) Electric analog
over each band to get the TL in octave bands:
po can also be expressed in terms of [T], ZS, ZT and ps by solving the
equations above,
(8)
(13)
Method 2: Equivalent Insertion Loss Method
This method assumes that the four-pole matrix of the muffler, as Thus, insertion loss (IL) can be calculated by
shown in Fig. 3, has been either measured or calculated in narrow
bands. However, it should be noted that the four-pole matrix does not
exist above the plane-wave cutoff of the inlet and outlet ducts.
(14)
Figure 3. Four-pole matrix of the muffler To compute the TL in octave bands, the sound pressures at the outlets
of the two cases shown in Fig. 5 are computed. Since both the source
The narrow-band TL can be calculated by the four-pole parameters,
and termination are anechoic, the computed IL is the same as the TL.
(9)
(12)
(15)
(16)
(a).
Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018
Assume ps is constant across the frequency range, and then convert p1 The second test case is a large round bar silencer shown in Fig. 8.
and p2 to octave-band sound pressure levels SPL1 and SPL2, First, the narrow-band TL is calculated by the impedance-to-
respectively. It can be proved that a constant ps with an anechoic scattering method [3] in conjunction with the substructuring BEM
source is equivalent to a constant incident wave. Then, the equivalent [4]. The bar silencer is made of polyester (flow resistivity=16,000
TL can be calculated by Rayl/m) and is covered by a 30% open perforated facing sheet and
two rigid end panels.
(17)
1. Set the incident sound power Wi=1 in Eq. (8) at all frequencies.
2. Calculate the corresponding transmitted sound power Wt at each Figure 8. Dimensions of the bar silencer; L=6 m.
frequency from the narrow-band TL.
As demonstrated in Fig. 9, only three small sections need to be
3. Convert both Wi and Wt into octave bands by doing simple modeled in the BEM: the beginning section (substructure 1), the
integration (or summation). middle-section template (substructure 2 template), and the tail section
4. Calculate the TL from the octave-band Wi and Wt. (substructure 3). The majority of the middle section does not need to
be meshed because the impedance matrix of the template can be
successively repeated along the bar silencer. An impedance matrix
Results and Discussion synthesis procedure [4] is then performed to combine all the
The first test case is the simple expansion shown in Fig. 6. One-third substructure impedance matrices into a resultant impedance matrix.
octave band TL curves using the three different methods presented in
this paper are compared to the narrow-band TL in Fig. 7. The
frequency stepping used in the narrow-band calculation is 10 Hz. It is
seen from Fig. 7 that all three methods produce the same TL.
Conclusions
In this paper, three different methods for converting the narrow-band
TL into octave bands are presented. All three methods are based on
the same assumption that the incident wave amplitude is constant
across the frequency range. The direct conversion method is the
easiest to implement and can be extended to high frequencies when
the plane-wave assumption is not valid anymore. Although the direct
conversion approach may have been used by many noise control
engineers, a standard is still lacking at this point.
References
1. Ver, I. and Beranek, L., "Noise and Vibration Control
Engineering." Principles and Applications, Second Edition,
ISBN 471449423 (2008): 15-18.
2. Wu, T. W., and Wan, G. C. "Muffler performance studies using
a direct mixed-body boundary element method and a three-point
method for evaluating transmission loss." Journal of Vibration
and Acoustics 118, no. 3 (1996): 479-484.
3. Wang, P and Wu, T. W. "Impedance-to-scattering matrix method
for silencer analysis." in INTER-NOISE and NOISE-CON
Congress and Conference Proceedings, 248, 453-460, (2014).
4. Lou, G, Wu, T. W., and Cheng, C. Y. R. “Boundary element
analysis of packed silencers with a substructuring technique,”
Engineering Analysis with Boundary Elements, 27, 643-653
(2003).
Contact Information
T. W. Wu
Department of Mechanical Engineering, University of Kentucky
Lexington, KY 40506, USA
Acknowledgment
The work was partially supported by the Vibro-Acoustics
Consortium.
The Engineering Meetings Board has approved this paper for publication. It has successfully completed SAE’s peer review process under the supervision of the session organizer. The process
requires a minimum of three (3) reviews by industry experts.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or
otherwise, without the prior written permission of SAE International.
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE International. The author is solely responsible for the content of the paper.
ISSN 0148-7191
http://papers.sae.org/2016-01-1317