Downloaded from SAE International by David Neihguk, Friday, March 09, 2018

Acoustic Analysis of a Tractor Muffler 2017-01-1791

Published 06/05/2017

David Neihguk and Shreyas Fulkar

Mahindra & Mahindra, Ltd.

CITATION: Neihguk, D. and Fulkar, S., "Acoustic Analysis of a Tractor Muffler," SAE Technical Paper 2017-01-1791, 2017,
doi:10.4271/2017-01-1791.

Copyright © 2017 SAE International

Abstract With the increasing availability of powerful computational resources,

more and more focus is now being given to bring the power of
Parametric model of a production hybrid (made up of reactive and
simulation to the hands of designers and engineers [5]. This paradigm
dissipative elements) muffler for tractor engine is developed to
shift in the way simulations are performed gives rise to the
compute the acoustic Transmission Loss (TL). The objective is to
Appfication of simulation in which simulation workflow, expert
simplify complex muffler acoustic simulations without any loss of
knowledge capture and reuse are combined in an application package.
accuracy, robustness and usability so that it is accessible to all
product development engineers and designers. The parametric model
The usability and accessibility of the app is enhanced by a user-
is a 3D Finite Element Method (FEM) based built in COMSOL
friendly GUI and web-based deployment of the Apps, respectively.
model builder which is then converted into a user-friendly
The backbone of the App is the acoustic module of COMSOL
application (App) using COMSOL App builder. The uniqueness of
Multiphysics [6] which is a 3D FEM based acoustic simulation tool.
the App lies in its ability to handle not only wide range of parametric
The framework for creating applications is integrated in the
variations but also variations in the physics and boundary conditions.
COMSOL workflow with direct compatibility of publishing the app
This enables designers to explore various design options in the early
through web-based platform called COMSOL Server.
design phase without the need to have deep expertise in a specific
simulation tool nor in numerical acoustic modeling. The accuracy of
Traditional 3D FEM acoustic analysis of muffler is a tedious process
the model is firstly verified with the results obtained from published
as it involves creating and cleaning the geometry, meshing, solving
literature. The validated models are then extended to the current
post processing and interpretation of the results. Every design
application. Finally, parametric studies are conducted to obtain the
iterations necessarily call for a modified CAD geometry and a
effect of various design parameters and the results are presented as
re-mesh to assess the effect of the changes and the cycle continues.
design guidelines which enables efficient and cost effective design of
mufflers for tractor engines.
The App based approach is inspired by the need to improve
productivity by eliminating repetitive work during the product life
Introduction cycle. The App serves as a good and reliable repository of the best
practices as it is developed to be used as a front-loading design
Noise reduction and sound quality of agricultural tractors have
calculation tool which speaks the language of the designers but thinks
become a concern for tractor manufacturers. The driver ear noise level
and works like the expert. In this way, the effect of parameters can be
and pass by noise level are the two regulatory requirements that need
quickly assessed and analyzed.
to be met by the product. The European regulatory requirements
specified 86 dB (A) and 89 dB (A) as the limit for driver ear noise
The work presented here describes the development of an intuitive
level and pass-by noise level, respectively. Compared to the current
user-friendly application that helps the users to get the most out of a
driver ear noise level limit of 94 dB (A) in India, the European limit is
parameterized muffler model.
8 dB (A) lower [1,2]. It is expected that similar reduction in pass-by
noise limit will be implemented in India sooner or later. The
implications of this will require better design of mufflers since exhaust Muffler Geometry
noise is identified as one of the significant noise sources. Interestingly,
The production (baseline) muffler consists of four chambers as shown
the NSI study conducted by Yadav et al., [3,4] also reveals the exhaust
in the CAD image of Fig. 1. The first and third chambers are acoustic
noise source to be the highest ranked noise source.
cavity, the second chamber is an annular extended inlet extended
outlet (EIEO) type with perforations in the inlet pipe (d1) and annular
 Downloaded from SAE International by David Neihguk, Friday, March 09, 2018
pipe (d3), the fourth chamber is an annular concentric tube resonator
with three interacting ducts, d2, d4 and d5. The outermost annular duct
of the second and fourth chamber are filled with absorptive materials
as shown in Fig. 2 in the hybrid configuration. The total length of the
muffler is 521 mm. The overall Ltotal/d5 ratio is 3.
Figure 1. CAD model of the baseline muffler
Figure 2. Schematic sketch of the baseline muffler hybrid configuration
There are three baffles, one in the upstream side (upstream baffle),
one in the middle (middle baffle) and one in the downstream side
(downstream baffle). Then baffles consists of a rigid section and a
perforated section as shown in Fig. 3. In the baseline configuration,
the downstream baffle is impervious, as a result the third and fourth
chambers are connected only through the outlet pipe, d2.
Figure 3. Schematic sketch of the baseline muffler reactive configuration
The geometrical details of the baseline configurations for the hybrid
and reactive cases are given in the Appendix.
3D FEM Analysis
The muffler is divided into different domains which are then
discretized into small finite elements. By using the acoustic pressure
as the independent variable, the wave propagation is solved in the
frequency domain using the time harmonic pressure acoustic mode in
COMSOL [6]. The governing equation is the modified version of the
3D Helmholtz equation:
(1)
where ρ, c and ω are the density, speed of sound and angular
frequency respectively. The following boundary conditions are used
to simulate the system [6]:
1. At the solid boundaries, which include the outer walls of the
muffler, the baffles between the chambers and the walls of the
inner pipes, sound hard (wall) boundary conditions are used:
(2)
here n is the unit normal vector.
2. At the inlet boundary, a combination of an incoming and
outgoing plane waves is assumed:
(3)
where p0 = 1 Pa is the applied pressure at entrance to the inlet
pipe and i is the imaginary unit.
3. At the outlet boundary, an outgoing plane wave radiation
condition with a vanishing reflection coefficient (anechoic
termination) for normally incident waves is used:
(4)
4. At the perforated pipes and baffles, the acoustic impedance for
the reactive muffler configuration of Figs. 3 and 4 in stationary
medium is modeled by using Elnady et al.’s [8] expression. For
the hybrid muffler configuration of Figs. 2 and 5 in stationary
medium, the Kirby and Cummings’[9] expression has been
used. The present study uses the texturized fiber glass roving
manufactured by the Silentex™ process and the Delany and
Bazley model [10] of complex characteristic impedance and
wave number extended to low frequencies by Mechel [11]. The
Transmission Loss is computed at a temperature of 305.54 K.
The acoustic impedance of perforates is implemented by applying
a semitransparent boundary at the contours of the perforated area.
This is conveniently achieved by using the predefined Interior
Perforated Plate boundary condition in COMSOL [6] which is a
function of the hole diameter and pipe thickness.
5. The dissipative effect of glasswool is modelled using the
poroacoustics physics wherein the flow resistivity (fr) is
obtained as a function of the filling density (fd) as given below:
(5)
Figure 4. COMSOL model of tractor muffler - reactive configuration
 Downloaded from SAE International by David Neihguk, Friday, March 09, 2018

Validation of 3D FEM COMSOL Model

To validate the model, the Concentric Tube Resonator (CTR) model
of Ref. [12] is reconstructed as shown in Fig. 7 for both reactive and
dissipative cases. The gas temperature is 32.57 degree Celsius. The
thickness of the inner pipe is 0.9 mm, diameter of perforated hole is
2.49 mm. The flow resistivity for filling density of 100 g/L and 200
g/L is 4896 Pa.s/m^2 and 17378 Pa.s/m^2, respectively. The model is
verified with six test cases as shown in Table 2.

Figure 5. COMSOL model of tractor muffler - hybrid configuration

Finally, the Transmission Loss (TL) is expressed in terms of the ratio

of the incident acoustic power at the inlet, Wincident using the applied
pressure, p0 and the transmitted acoustic power at the outlet, Figure 7. CTR model adapted from Ref [12]
Wtransmitted using the computed pressure, pc at the outlet as
Table 2. Parameters of CTR used for validation of the 3D FEM Model

(6)

where

The Transmission Loss results predicted by the 3D FEM model using

Here p0 is the pressure associated with the incident wave, and pc is
COMSOL is compared with the experimental results [12] with and
the pressure associated with the wave transmitted into an anechoic
without glasswool as shown in Fig. 8 and Fig. 9, respectively. Good
termination.
correlation can be observed between 3D FEM COMSOL and
experimental results. Hence, the results so obtained amply validates
Mesh Independent Study the 3D FEM Model.
Two mesh sizes with maximum mesh element of one-fifth and
one-tenth of the minimum wavelength and a minimum mesh element
of 11.2 mm and 5.6 mm are used as given in Table 1. As shown in
Fig. 6, it is found that the result is invariant of the two mesh sizes.
Consequently, the subsequent analysis is performed using Mesh 1.

Table 1. Mesh parameters

Figure 8. Transmission Loss of CTR without glasswool

Figure 9. Transmission Loss of CTR with glasswool

Figure 6. Transmission Loss of CTR without glasswool

 Downloaded from SAE International by David Neihguk, Friday, March 09, 2018

Application and GUI

The 3D FEM Model is converted into an App by creating an intuitive
GUI as shown in the Appendix. The GUI consists of inputs for
geometry properties, gas temperature, options for selection of
perforation types, check box for use of dissipative materials
(glasswool), user-defined frequency range and expected and past
computation time. Other features of the App includes the reset button,
manual and event based geometry update, buttons for documents and
geometry details, buttons for compute, plot and compare results,
buttons to download report and email report, export results and clear
solutions. On loading the application, the isoperimetric view of the
muffler is displayed with the default parameters. The display Figure 11. Transmission Loss of tractor muffler without glasswool
geometry and geometry details buttons display the 3D geometry and
2D schematic sketch of the muffler to aid the users in feeding the
Parametric Studies
input parameters as shown in Figs. 10 and 2, respectively.
Parametric studies are conducted to understand the sensitivity of
some parameters on the acoustic performance of the tractor muffler.
The effect of the following parameters are investigated in this study:

1. Effect of baffles
2. Effect of acoustic cavities
3. Effect of location of pipe perforations

Effect of Baffles
The muffler investigated here consists of three baffles, namely the
upstream baffle, middle baffle and downstream baffle. Perforated
baffles introduces acoustic interaction between the acoustic
pressures of the first, second and third chambers of the baseline
muffler configuration.

Four cases are investigated as shown in Table 3. The first case B1 is

Figure 10. 3D muffler geometry as displayed in the App
Incorrect data entry is prevented by using data validation which
prompts an error message if the input values are out of the intended
range. Checkboxes are provided to highlight the relevant boundaries
and domains. It may be noted that the icons in the toolbar are linked
to the events in the UI. The report and export result buttons are
activated only after the computation is successfully completed. The
advantage of this feature is that it makes the App self-explanatory so
that the usability is not compromised.
same as the baseline configuration in which the downstream baffle is
impervious or rigid. In the second case B2, all the baffles are
perforated. In the third case B3, the upstream and downstream baffles
are perforated but the middle baffle is impervious. Finally, in the last
case B4, only the upstream baffle is perforated. As shown in the
Appendix, combo boxes are provided for the selections of baffle types.
The nomenclature P stands for perforated baffle and R stands for the
rigid baffle.
Table 3. P=Perforated, R=Rigid

Transmission Loss of Baseline Muffler

The acoustic performance of the baseline tractor muffler is shown in
Figs. 11 for both reactive and hybrid configurations. The effect of
glasswool results into improved transmission loss at high frequencies
(900 Hz and above in this case), which agrees with expectations [12].
It is observed from Fig. 12 that for the reactive case, improvements in
low frequency performance is achieved at the cost of deterioration in
mid-frequency performance by making all the baffles to be perforated
which corresponds to the second case B2. Interestingly, the loss of
performance in the mid frequency range for the reactive configuration
is compensated by the addition of glasswool which corresponds to the
hybrid configuration of case B2. The improvements so obtained can be
attributed to the dissipative effects of glasswool as shown in Fig. 13.
 Downloaded from SAE International by David Neihguk, Friday, March 09, 2018
Figure 12. Effect of Baffle type on the Transmission Loss of tractor muffler in
reactive configuration.
Figure 13. Effect of Baffle type on the Transmission Loss of tractor muffler in
hybrid configuration.
Effect of Acoustic Cavities
The first and third chambers with length L1 and L7, respectively
comprises the acoustic cavities. The acoustic cavities are assumed to
have rigid end plates and thereby exhibits the quarter wave resonance
characteristics which is inversely proportional to the cavity length.
Four cases are identified as shown in Table 4. The first case AC1
corresponds to the baseline configurations. The second case AC2
have the first cavity length doubled and the second cavity length
reduced by a factor of 10. Similarly, the third case AC3 have the first
cavity length reduced by a factor of 10 and the second cavity length
doubled. Finally, the last case AC4 doubles both the length of the first
and second cavities. The cavity lengths for each case are used in the
length section of the GUI as shown in the Appendix.
Table 4. Variation of acoustic cavity lengths
As shown in Figs 14 and 15, low frequency performance
improvements are observed in Case AC2 and AC4 for both reactive
and hybrid configurations. In contrast, low frequency performance
deteriorations are observed in case AC3 for both reactive and hybrid
configurations. It can be inferred that the first cavity (first chamber)
have more influence on the TL performance.
Figure 14. Effect of Acoustic cavity lengths on the Transmission Loss of
tractor muffler in reactive configuration.
Figure 15. Effect of Acoustic cavity lengths on the Transmission Loss of
tractor muffler in hybrid configurations.
Effect Location of Pipe Perforations
Perforation on the pipe coupled to the perforations in the baffles
introduces multiple path for the acoustic waves to propagate. By
selecting the right location for perforations, the acoustic performance
of mufflers can be improved. To this end, two test cases are set up as
shown in Table. 5 with PP1 and PP2 where PP1 corresponds to the
baseline configuration. In PP2, the perforations of the inlet pipe at the
second chamber is replaced with a rigid surface as shown in Fig. 21.
The selected pipe can be visualized in the graphics window of the UI
by checking the checkboxes as shown in the Appendix.
Table 4. P=Perforated, R=Rigid, PP1 = Baseline
It can be observed from Figs. 16 and 17 that improvements in low
frequency performance is achieved by making the inlet pipe (Pipe 1)
fully rigid for both reactive and hybrid configurations. Again, this can
be attributed to the quarter wave resonance behavior of the extended
inlet length (L2 + L3 + L4) of the inlet pipe in the second chamber [13].
 Downloaded from SAE International by David Neihguk, Friday, March 09, 2018
Figure 16. Effect of placement of pipe perforations on the Transmission Loss
of tractor muffler in reactive configurations.
Figure 17. Effect of placement of pipe perforations on the Transmission Loss
of tractor muffler in hybrid configurations.
Summary/Conclusions
The following conclusions can be drawn from the preceding analysis:
1. To improve low frequency performance: make all baffles
perforated, increase length of first chamber and do not provide
perforations in the inlet pipe of second chamber.
2. To reduce low frequency performance: make the middle baffle
rigid or impervious, reduce length of first chamber.
3. The baseline configuration gives the best performance in mid-
frequency range (400 Hz to 1400 Hz approximately).
The parametric studies presented here represents only a few insights
that can be obtained from the Muffler App. As such, no attempt is
being made to present a comprehensive parametric study with regards
to the muffler under investigation. Rather, the emphasis is on the
creation of the model and its conversion into an App that simplifies
muffler acoustic simulation and extend the power of simulation to the
hands of designers and engineers. The creation of this Apps requires
the involvement of multiple resources like designers, analysts and
App creator to put together all the requirements and best practices
specific to the muffler design being Appified. Herein, lies the value
addition of the App.
References
1. http://eur-lex.europa.eu/eli/dir/1977/311/1982-12-21 (accessed
10th April 2017)
2. V, R. and Prakash, B., "Agricultural Tractor Noise Control,"
SAE Technical Paper 2013-01-2342, 2013, doi:10.4271/2013-
01-2342.
3. Yadav, P., Gaikwad, A., Badgujar, S., Surkutwar, Y. et al.,
"Noise Reduction on Agricultural Tractor," SAE Technical
Paper 2013-26-0103, 2013, doi:10.4271/2013-26-0103.
4. Rowley, D., "Control of Farm Tractor Intake and
Exhaust Noise," SAE Technical Paper 660598, 1966,
doi:10.4271/660598.
5. https://www.nafems.org/about/regional/americas/
events/2020vision/ (accessed 22nd March 2017)
6. COMSOL Multiphysics, Acoustic Module User’s Guide,
COMSOL 5.2a: January (2017).
7. COMSOL Multiphysics, Application Builder User’s Guide,
COMSOL 5.2a: January (2017).
8. Elnady T., Abom M. and Allam S., “Modelling perforates in
mufflers using two-ports”, ASME J. Vibr. Acoustics, 132(6),
1-11, (2010).
9. Kirby R. and Cummings A., “The impedance of perforated
plates subjected to grazing flow and backed by porous media”,
J. Sound Vibr., 217(4), 619-636, (1998).
10. Delany M.E. and Bazley E.N., “Acoustical properties of fibrous
absorbent materials”, Applied Acoustics, 3(2), 105-116, (1970).
11. Mechel F.P., “Extension to Low Frequencies of the Formulae
of Delany and Bazley for absorbing Materials.” Acta Acustica
united with Acustica 35, (3), 210-213, (1976).
12. Selamet, A., Lee, I., Ji, Z., and Huff, N., "Acoustic Attenuation
Performance of Perforated Absorbing Silencers," SAE Technical
Paper 2001-01-1435, 2001, doi:10.4271/2001-01-1435.
13. Munjal M.L., Acoustics of Ducts and Mufflers, Second Edition,
Wiley, Chichester, UK, (2014).
Contact Information
David Neihguk
neihguk.david@mahindra.com
Acknowledgments
The authors gladly acknowledge the valuable feedback and
suggestions given by the reviewers.
 Downloaded from SAE International by David Neihguk, Friday, March 09, 2018

APPENDIX
Appendix
GUI of the Muffler App

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/2017-01-1791