SAE TECHNICAL

PAPER SERIES 1999-01-1800

Application of Computer Simulation Using FEM

and Experimental Techniques for the Reduction
of Noise in Air Cooled Engine and Crankcase
Cover of Motorcycle
A. R. Askhedkar, R. R. Askhedkar and P. R. Sajanpawar
The Automotive Research Association of India

Reprinted From: Proceedings of the 1999 Noise and Vibration Conference

(P-342)

Noise and Vibration Conference & Exposition

Traverse City, Michigan
May 17-20, 1999

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760
The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the
paper may be made for personal or internal use of specific clients. This consent is given on the condition,
however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc.
Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sec-
tions 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as
copying for general distribution, for advertising or promotional purposes, for creating new collective works,
or for resale.

SAE routinely stocks printed papers for a period of three years following date of publication. Direct your
orders to SAE Customer Sales and Satisfaction Department.

Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.

To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
other works, contact the SAE Publications Group.

All SAE papers, standards, and selected

books are abstracted and indexed in the
Global Mobility Database

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written
permission of the publisher.

ISSN 0148-7191
Copyright 1999 Society of Automotive Engineers, Inc.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely
responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in
SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group.

Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300
word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA
 1999-01-1800

Application of Computer Simulation Using FEM and

Experimental Techniques for the Reduction of Noise in Air
Cooled Engine and Crankcase Cover of Motorcycle

A. R. Askhedkar, R. R. Askhedkar and P. R. Sajanpawar

The Automotive Research Association of India

Copyright © 1999 Society of Automotive Engineers, Inc.

ABSTRACT mission from the combustion chamber to the external

structure consists of cranktrain, main bearing and bulk
Measurement of sound intensity techniques has very heads. The vibration characteristics of engine structure
good application in the source identification of a particu- and the cranktrain structure are important parameters
lar noise character. It has been applied effectively along contributing to the engine surface noise. [1,2] The small
with modal analysis and FE experimental excitation tech- gasoline engine of a motorcycle had a problem that it pro-
niques to find out root cause of a particular noise charac- duced a peculiar noise which was annoying and affecting
ter in small gasoline engine. A FEM shell model was the noise quality. In order to remove that annoying noise
used to make cylinder block and cylinder head model. character it was necessary to identify the source of that
FEM simulation was carried out which matched with noise.
experimental results. It helped to remove the noise char-
The noise intensity technique has been used effectively
acter from engine.
for identifying the source of noise. [3,4]. Modal analysis
The other part of the paper describes the noise reduction and harmonic analysis techniques also have been used
of the crankcase cover used for the same motorcycle. It to find out the root cause of the particular noise character
houses crankcase as well as two speed gearbox. The in small gasoline engine. FEM simulation was carried out
methodology involves very effective combination of which matched with experimental results. These tech-
experimental harmonic analysis, FE model with the shell niques helped to remove the particular noise character
element for the 3 piece crankcase cover, and experimen- from the engine.
tal measurements. A particular sequence of this experi-
In most of the motorcycles the crankcase and the gear-
mental techniques along with computer simulation
box are in one housing. The crankcase cover of this
techniques gives extremely good results. It has an effect
motorcycle houses two speed gear box. The crankshaft
on the passby noise reduction as well as the noise char-
vibrations as well as the vibrations of the gears are trans-
acter elimination of the transmission cover.
mitted to the crankcase cover via bearings which ulti-
The paper describes the results carried on one of the mately results in the structure related noise [5,6]. This
motorcycle crankcase structure, the combination of the crankcase cover noise plays important role in pass by
techniques for NVH control in the motorcycle which has noise level of the vehicle and also in the overall noise
brought out tremendous improvements in the noise char- character of the vehicle. [7] In order to reduce the pass
acter by fine tuning the structure. by noise and change noise character it was necessary to
reduce the noise emitted by the crankcase cover. In this
INTRODUCTION paper the experimental techniques combined with the
analytical techniques have been used very effectively for
In view of overall vehicle acceptance the NVH behaviour noise reduction of crankcase assembly.
of engines and power train is important. The engine
noise consists of number of different components such NOISE REDUCTION OF AIR COOLED GASOLINE
as exhaust and intake noise, noise emitted by engine sur- ENGINE
face, noise of the accessories and that emitted by parts
connected with the engine. A control of engine surface When the motorcycle is operated in idling condition, the
noise is in most cases more difficult than the control of various sources such as inlet, exhaust, transmission and
other noise components. [1] The noise can be reduced engine contribute to the overall noise. This particular
by treatment of individual components or by modifying engine had a peculiar ringing noise character. The objec-
radiating surfaces. The important path of vibration trans- tive was to identify & eliminate the ringing noise character

1
and reduce the overall noise. Due to the presence of var- SOURCE IDENTIFICATION USING INTENSITY TECH-
ious sources, it was very difficult to detect the ringing NIQUE – One of the important steps in any noise control
noise of particular character and also the surface which program is the identification of major noise sources. The
was radiating that noise. Hence the isolation of the other sound intensity measurement technique using two
sources from the engine noise was essential. The trans- closely spaced microphones and dual channel FFT Ana-
mission was removed, the exhaust and inlet noise were lyzer has made the task of noise source location easier.
isolated by proper treatment, thus only the engine noise Sound intensity in the frequency domain is obtained from
was present. the imaginary part of the cross spectrum between two
microphones [3,4,8] The intensity in their direction is cal-
EXPERIMENTAL NOISE AND VIBRATION ANALYSIS – culated as,
The noise and vibration signals were recorded at various
locations on cylinder head and cylinder block figure 1. (1)

where ω is the angular frequency and Im {G12} is the

imaginary part of the cross spectrum between two micro-
phone signals.
Two methods are used for source location, namely
source ranking and intensity mapping. In this paper the
intensity mapping has been used for source identifica-
tion. This intensity mapping can be performed in real-
time or off line. Real time intensity mapping utilizes the
cosine characteristics of the intensity probe and can be
divided in two methods ‘the max search’ and ‘the null
search’. In max search, the probe is held normal to the
engine and moved close to the surface along different
Figure 1. The locations on the cylinder head and block components of it, while display screen is watched for high
where noise and vibration signals were intensity level areas. The null search method works on
recorded. the sharp minimum in the directional characteristics of
the probe at 90° position. In that case the probe is held
The noise signals were recorded in hot and cold condi- parallel to the engine surface and the rapid change of
tions of the engines. Figure 2 represents the typical intensity from +ve to -ve or vice versa locates the source
noise spectra in hot and cold condition of the engine. The when the probe is moved along the surface of the engine.
results show the predominant peak around 3.1KHz. The
results indicate that the noise became predominant in the
hot condition in the frequency range between 3 KHz to 4
KHz.

Figure 3. Locations on the engine where intensity was

recorded.

The sound intensity analysis was carried out to identify

the source of noise. The intensity probe was moved par-
allel to the engine from head to the crankcase and the
results were recorded at five locations (Figure 3).
Figure 2. The noise spectras in the hot and cold The results reveal that there is a change of intensity from
conditions. +ve to -ve at 3100Hz at cylinder head location, which
indicate that fin is the source of noise (Figure 4).

2
 The finite element model of cylinder head was prepared
using shell 63, four noded quadrilateral shell element
having thickness as a variable option. It has six degrees
of freedom at each node (translation along x, y, z and
rotation about x, y,z).
The material properties used are modulus of elasticity
= 0.675 x 108 kg mm/s 2 . mm2 and density = 2700 x
10 -9 kg/mm 3. The complete model consists of 2500
nodes and 2300 elements.
For finding out natural frequencies and modeshapes 40
(a) (Location 1) master degrees of freedom are given with operating
boundary conditions. The frequencies obtained are
given in Table 1.

Table 1. Frequencies of unmodified cylinder head.

Mode No. Frequency (Hz)
1 3000.7
2 3849.2
3 4352.5
4 5260.2
5 5457.7
The results of the mode shape reveal that fins have one
of the frequency around 3100Hz. (Figure 5)
(b) (Location 2)
HARMONIC ANALYSIS – The generalized equation of
motion for a structural system can be written as
Figure 4. Intensity plots of engine with unmodified head
at location 1 & 2.
(6)

where {F} is applied load vector and {x,}, {x
}, {
x
} are
FINITE ELEMENT MODAL ANALYSIS – Mechanical
systems are made up of structural elements such as
nodal displacement, velocity and acceleration vectors.
bars, beams, plates and shells. These elements have
continuously distributed mass and stiffness properties This harmonic analysis is carried out experimentally and
and expressing the kinetic and strain energies of ele- analytically. In the experimental harmonic analysis the
ments in terms of elemental degrees of freedom. [9] In structure is excited with known force and the responses
modal analysis we are basically interested in finding out at discrete points are measured. In the analytical method
eigen values and eigen vectors. Equations for free vibra- the finite element model of the structure with identical
tions for the assemblage is given by boundary conditions is excited with the same force and at
same location and the responses are recorded at loca-
(2) tions indicated in the experimental harmonic analysis.
The correlation is obtained in experimental and FEM
Here, as the modal displacements are time varying analytical results for acceptance of model.
assuming displacement vector {x} can be represented
assuming a simple harmonic motion with angular fre- The dynamic analysis of the cylinder head was carried
quency of motion ω rad/sec, as out. The dynamic excitation was simulated in the com-
puter and the responses were recorded at various loca-
{ x } = { X } sin ωt (3) tions as shown in figure 1. Figure 6 shows the typical
substituting this value in equation (3) we get, dynamic response at locations H 1 & H 7 on the head
respectively. The results show high amplitudes of vibra-
ω2 [M] {x} = [ K ] [ x ] (4) tion around 3100Hz. This is responsible for noise cre-
The above equation can be written in general form as ated by the typical fin structure excited by axial vibration
of cylinder block.
[K] {φ} = λ[M] {φ} (5)
equation (5) is the eigen value problem with λ = ω2 as a
eigen value and f as a eigen vector which can be solved
by using different numerical algorithms. The eigen val-
ues determine the natural frequencies while eigen vec-
tors give the mode shapes corresponding to each mode.

3
 (a)

(a) The mode shape of the cylinder head at 3086.0Hz.

(b)

Figure 6. Analytical harmonic responses of original

head.

(b) The mode shape of the cylinder head at 3145.75Hz.

Figure 5.

The experimental harmonic analysis of cylinder head

under above conditions was also carried out in which the
structure was excited with MB dynamics exciter. The
arrangement of experimental setup is shown in figure 7.
The results were recorded at the above mentioned loca-
tions. Figure 8 shows the typical harmonic response of
the structure at location H 1 & H 7 on the head respec-
tively.
A good correlation is obtained between experimental and
analytical results of harmonic analysis which validates
the FE model. Figure 7. Experimental setup for harmonic analysis.

4
 The modal analysis of modified head indicate that the
resonant frequency around 3100Hz was absent. The
harmonic analysis of the modified head was also simu-
lated on the computer.
The structure was excited with the same force and the
responses were recorded at the same locations as it was
done with unmodified head. The dynamic response
curves show that there is no predominant peak of vibra-
tion around 3100Hz (Figure 9)

(a)

(Location 1)

(b)

Figure 8. Experimental harmonic responses of original

head

ELIMINATION OF NOISE CHARACTER – The experi-

mental and analytical analysis carried out on cylinder
head indicate that typical noise is generated at the fin
level. To remove or eliminate the typical noise character,
it was decided to modify the fins in such a way that the
resonant frequency around 3100Hz should be elimi-
nated.
(Location 7)
The modifications were simulated on the FE model. The
modal analysis of modified head was carried out to deter-
Figure 9. Analytical harmonic responses of modified
mine the natural frequencies. Table 2 shows the natural
head.
frequencies of modified head.
The noise intensity analysis of engine with modified head
Table 2. Natural frequencies of modified cylinder head. was carried out to confirm the elimination of 3100Hz
Mode No. Frequency (Hz) peak. The results show that there is no change of inten-
sity from +ve to -ve at 3100Hz at location 2. (Figure 10)
1 2034.4
2 2889.8 The noise spectras of the engine with modified head and
3 3461.0 unmodified head are recorded which show the elimina-
4 4604.3 tion of 3100Hz noise peak from the noise signal of engine
5 4713.2 with modified head (Figure 11) The noise levels are
measured at 1m distance from engine. For unmodified

5
head the noise levels were 76.86 dBA and with modified meshing frequencies related to noise and vibration are
head the noise levels were reduced to 73 dBA. present in a gear train. The gear tooth irregularities such
as broken or chipped tooth produces high noise and
vibration levels at the narrow band peaks which are in
addition to the various gear meshing frequencies and
their associated harmonics. The vibrations of the shaft
due to misalignment in axial direction which results in
increased vibration at rotational frequency and first few
harmonics. All the vibrations are ultimately transferred to
the casing via bearing. So isolation of bearing is most
effective and direct approach in solving the structure born
noise problems. [14]
The crankshaft signals from engine inertia loads are
transferred to the outer casing via main bearing which
ultimately are transferred to the outer casing through
(Location 1) shaft via bearing housing which results in the structure
born noise. This has been shown in figure 12.
It was decided to fine tune the structure so that noise
related frequencies are eliminated. The methodology for
noise reduction of crankcase assembly is shown in the
figure 13.

(Location 2)

Figure 10. Intensity plots of engine with modified head at

Figure 12. Noise generation mechanism of crankcase
location 1 & 2.
assembly.

Figure 11. The noise spectras of engine with & without

modifications in the head.

NOISE REDUCTION OF CRANKCASE Figure 13. Methodology for noise reduction of crankcase
ASSEMBLY assembly for motorcycles.

The crankcase cover of this motorcycle houses two FINITE ELEMENT ANALYSIS – The crankcase of this
speed gear box. In the present gear box several gear motorcycle assembly consists of three parts. RH side,

6
middle and LH side casing. The model was prepared in
ANSYS using shell elements and solid elements were
used. The three casings of the crankcase are shown in
figure 14.

(Location 1)

Figure 14. The three casings of the crankcase.

Solid 45, eight noded, 3D isoparametric solid element is

used for solid portions. It has three degrees of freedom
at each node i.e. translation along x,y, & z direction. The
model of the casings are explained below. For preparing
the model shell 63 and solid 45 type elements are used.
Shell 63, four noded, quadrilateral shell element having
thickness as a variable option was used. It has six
degrees of freedom at each node i.e. translation along x,
y, z and rotations about x, y, z direction. Complete model (Location 6)
contains 6500 nodes and 6300 elements. Ten variable
constants are used for variable thicknesses in the model. Figure 15. Results of analytical harmonic analysis of
The material properties used are original crankcase.
Modulus of elasticity = 0.675 x 108 Kg.mm/s2.mm2 &
EXPERIMENTAL NOISE MEASUREMENT – The noise
Density = 2700 x 10 -9 kg/mm3
signals were recorded at 1m distance from the crankcase
under loaded and unloaded condition. In no load condi-
HARMONIC ANALYSIS – The harmonic analysis has
tion the vehicle was operated in II gear and the rear
been carried out to validate the finite element model. In
wheel is allowed to rotate freely. In loaded condition the
the analytical harmonic analysis, the structure was
vehicle was operated in same gear but the load was
excited at bearing location and the responses were taken
applied on rear wheel by giving torque which produces
at various locations on right hand side and left hand side
the same load as that in running condition.
casing. The typical results of harmonic analysis are
shown in Figure 15. The noise signal under loading condition is shown in Fig-
ure 17. The results of load and unloaded condition show
Also the experimental harmonic analysis has also been
that the noise levels are very high in case of loaded con-
carried out in which the crankcase assembly was excited
dition. It should be noted that the operating modal fre-
using MB dynamic exciter with the same force as given in
quencies of crankcase cover coincide with noise related
analytical analysis and the responses were recorded at
frequencies.
same locations as mentioned in analysis.
Figure 16 shows the typical results of harmonic analysis.
The results of experimental and analytical harmonic anal-
ysis were correlated and FE model was found to be accu-
rate enough to carry out the further work.

7
 Table 3. Natural frequencies of crankcase casing.
Mode No. Frequency (Hz)
1 3000.7
2 3849.2
3 4352.5
4 5260.2
5 5457.7

(Location 1)

(a) The mode shapes of crankcase at 3000.7Hz

(Location 6)

Figure 16. Results of experimental harmonic analysis of

original crankcase.

MODAL ANALYSIS – The modal analysis of crankcase

assembly had been carried out in order to determine the
natural frequencies and mode shapes of the structure.
The analytical results were determined using FEM model
using ANSYS software. The natural frequencies of the
casing are shown in table 3. Typical mode shapes are
(b) The mode shape of the crankcase at 3849.2Hz
shown in Figure 18. The nodal displacement plots are
shown in Figure 19
Figure 18.

REDUCTION OF CRANKCASE ASSEMBLY NOISE –

The crankcase assembly noise can be reduced by
proper design of casing in such a way that it should not
resonate in the entire operating range.
The nodal displacement plots, and the modeshapes indi-
cate that the right hand casing is responsible for emitting
noise around 3000Hz. The modifications were carried out
in the right hand side casing where the nodal displace-
ments are high. At first the modifications were made in
the crankcase assembly model to study the effect of
modifications on the natural frequencies and mode
shapes.

Figure 17. Noise spectras of crankcase with & without

load.

8
 Figure 20. The noise spectras of crankcase with &
without modifications.
Figure 19. Nodal displacement plots of crankcase around
3000Hz

Table 4 gives the natural frequencies of modified casing

which shows that the predominant frequencies which is
responsible for producing noise around 3000Hz are
absent. The nodal displacement values are reduced sig-
nificantly. So to study the effect of this modification on
the noise level, the actual modification were carried out in
the crankcase assembly.
The experimental noise analysis under loading condition
was carried out in the second gear. The noise was mea-
sured at same location as mentioned earlier. This result
reveals that substantial noise reduction is achieved in the
frequency range of 2 to 5KHz (Figure 20). The vibration
Figure 21. The vibration spectras of crankcase with &
measurements were also carried out on the original and
without modifications.
modified crankcase. (Figure 21) shows the vibration
results of crankcase with and without modifications. The
The crankcase assembly noise contributes significantly
vibration spectra reveal that substantial reduction is
to the passby noise level. To study the effect of this mod-
obtained in the vibration levels.
ification on the pass by noise level, the pass by noise
testing of the motorcycle with and without modification in
Table 4. Natural frequencies of modified crankcase the crankcase assembly had been carried out. The
casing. results show that upto 2 dB noise reduction is obtained
Mode No. Frequency (Hz) with the modifications in the crankcase assembly.
1 4333.9
2 4593.9
3 4726.8
4 5414.2
5 5486.6

9
CONCLUSIONS REFERENCES

The ringing noise is generated due to ring and port 1. Hans H.Priebch & Josef Affenzealler, “Structure born
groove assembly problem which generates signal in the noise prediction techniques”, SAE paper 900019,
axial direction and radiated through fins around 3100Hz. 1990.
2. Dejong R.G. and Parsons N.E., “High frequency
Finite element modal analysis of original head show that
vibration transmissions through the moving parts of
fins have natural frequency around 3100Hz and har- an engine”, SAE Paper 80045, 1980.
monic analysis show that the high level of fin vibration
around 3100Hz. 3. J.Y.Chung, J.Pope and D.A.Feldmair “Application of
acoustic intensity measurement to engine noise eval-
The modifications on the head shifts the natural fre- uation, SAE Diesel engine conference, 1979.
quency from 3100Hz and also very low levels of fin vibra- 4. N.Tondon “A brief summary of sound intensity mea-
tions around 3100Hz. surement and their application of noise control”,
applied acoustics, 28 (1989), 37-47.
Noise levels at 1m distance from engine indicate that lev-
els for modified head are lower by about 3.2dB than for 5. N.Lalor, J.D.Dixon & G.J.Stimpson, “Source identifi-
the unmodified head. cation of automotive gearbox noise”, C404/033,
IMechE Conference 1990.
The results of noise signal analysis show that number of 6. Terry A.Dunlap & William G.Haivorsen, “Transmis-
predominant peaks in the noise spectra of modified head sion noise reduction”, SAE Paper 720735, 1972.
are less as compared to unmodified head indicating that 7. K.Fujita, T.Abe and Y.Hori, “Simulation of accelera-
the quality of noise has been improved. tion passby noise considering the acoustic radiation
The noise is generated by gears and crankshaft which is characteristics of a vehicle body”, I.J. of a vehicle
transmitted to crankcase assembly via bearings is design Vol.8, Nos.4/5/ p.p. 514-525, 1987.
around 3 to 5 KHz. The modal analysis shows that the 8. F.J.Fahy, “Sound intensity” EL sevier applied science,
natural frequencies of the crankcase assembly lie New york, 1979.
between 3 to 5KHz. 9. O.C.Zienkiewicz, “The finite element method”, Mc
Grow Hill, Publication, New Delhi, 1979.
The nodal displacement plots obtained from the modal
analysis show that the right hand side casing have high 10. D.J.Ewins, “Modal Testing : Theory and Practice,”
Research studies press ltd., Tounton, England,
displacement values around 3000Hz indicating that it is
Revised Edition.
responsible for emitting noise. The modified crankcase
assembly gives substantial reduction in the noise levels. 11. ANSYS 5.1, Theoretical manual, ANSYS inc. Hous-
ton, USA.
The noise analysis results indicate that the predominant 12. ANSYS 5.1 Users Manual, Vol.I,II ANSYS Inc Hous-
peaks are absent between 3 to 5 KHz which were ton USA.
present in unmodified crankcase assembly.
13. L.E.Kinsler and A.R.Fray, “Fundamentals of Acous-
The pass by noise results shows that reduction upto 2 dB tics”, Willey Eastern Ltd., New Delhi 1987.
is obtained with the modified assembly. 14. M.P.Norton “Fundamentals of Noise and vibration
analysis for engineers”, Cambridge university press,
ACKNOWLEDGMENT Cambridge, 1989.

The authors are grateful to the Director of ARAI for givx-

ing opportunity to work on this project and for providing
required facility. We are deeply indebted to staff in our
Vehicle Engineering Department, ARAI for their assis-
tance in preparation of manuscript.

10