Muffler design, development and validation for pdf

Sreenath and Munjal gave expression for attenuation of mufflers using the transfer matrix approach Muffler Design, Development and Validation for Acoustic Advantages on Automotive Vehicl

Noise pollution has become one of the major environmental

concerns in present era With the ever tightening laws and

increasingly straight regulations for controlling noise

pollution of automotive vehicles, mufflers are important part

of engine system and commonly used in exhaust system to

minimize noise caused by exhaust gases Design of mufflers

is a complex function that affects the noise characteristics and

fuel efficiency of the vehicle Traditionally, muffler design

has been an iterative process by trial and error method

However theories and science that has undergone

development in recent years has given a way for an engineer

to cut short number of iterations In today's competitive world

market, it is important for a company to shorten product

development cycle time and thereby cost

The objective of this paper is to propose a practical approach

to design, develop and validate muffler practically which will

give advantage over conventional method This paper also

emphasis on how modern CAE tools could be leveraged for

optimizing overall system design balancing conflicts like

noise and back pressure The project is considered for

validation plan on real time vehicle application to realize

objective of design as future scope of work

INTRODUCTION

Since invention of internal combustion engine in latter part of

nineteenth century, noise created by it has been a constant

source of trouble to environment Significantly, exhaust noise

in terms of pressure is about 10 times more than all other

noises (structural noise) combined So problems of reducing

engine noise consistently, mainly in attenuating exhaust noise

are a challenge

The design of mufflers has been a topic of great interest for many years and hence a great deal of understanding has been gained Most of the advances in theory of acoustic filters and exhaust mufflers have come about in last four decades Hence good design of muffler should give best noise reduction and offer optimum backpressure for engine Moreover, for a given internal configuration mufflers have to work for a broad range of engine speed

Usually when mufflers are designed by well established numerical techniques like boundary element method or finite element method, numerical model generation is time consuming often limiting user to try various other possible design alternates The process might be lengthy and laborious

as it involves more iteration with different prototypes The process discussed in this paper will help in overcoming above mentioned drawbacks

BRIEF OVERVIEW OF EARLIER WORK

Mufflers have been developed over last ninety years based on electro- acoustic analogies and experimental trial and error Many years ago Stewart used electro - acoustic analogies in deriving basic theory and design of acoustic filters [1] Later Davis et al published results of a systematic study on mufflers [2] They used travelling wave solutions of one-dimensional wave equation and assumption that acoustic pressure ρ and acoustic volume velocity υ are continuous at changes in cross sectional area An important step forward in analysis of acoustical performance of mufflers is application

of two- port network theory with use of four -pole parameters Igarashi and his colleagues calculated transmission characteristics of mufflers using equivalent electrical circuits [3-4] Sreenath and Munjal gave expression for attenuation of mufflers using the transfer matrix approach

Muffler Design, Development and Validation for

Acoustic Advantages on Automotive Vehicles

2011-01-2215

Published 09/13/2011

Shitalkumar Ramesh Shah and Gangadhar GS

Volvo India Pvt Ltd

Copyright © 2011 SAE International doi:10.4271/2011-01-2215

Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016

[5] The expression they developed was based on velocity

ration concept Later, Munjal modified this approach to

include the convective effects due to flow [6] Young and

Crocker used the finite element method to predict four-pole

parameters and then transmission loss of complex shaped

mufflers for case of no flow [7]

Ying-chang, Long-Jyi used optimized approach of maximal

STL and muffler dimension under space constraints

throughout the graphic analysis as well as computer aided

numerical assessment [8] Middlberg,J.M and Barber T.J

present different configurations of simple expansion chamber

mufflers, including extended inlet or outlet pipes and baffles

have been modeled numerically using CFD (Fluent −12) in

order to determine their acoustic response [9] However, most

of the research studies based on formulation of mathematical

equation and trial and error method T Rose and Jebasinski J

describes the application of a “Design of

Experiment”-Method (Taguchi-Experiment”-Method) to reduce the tail-pipe noise of an

exhaust system, the Taguchi-Method was applied to three

different exhaust systems[10] Several geometric parameters,

such as pipe diameter and length or muffler volume, have

been changed at each exhaust system

For the prediction of TL acoustic chambers in mufflers, an

approach that has had wide acceptance is based on acoustic

filter theory However, acoustic filter theory does not provide

a complete theoretical explanation of the action of acoustic

chambers, since when an acoustic filter passes a certain upper

frequency limit; it ceases to behave in the way predicted by

the simple plane-wave theory Moreover, in many cases when

the chamber configuration is complicated, the simple

one-dimensional theories cannot give a complete explanation of

the chamber behavior because of the complex wave

phenomena which exist However, most of the research

studies based on formulation of mathematical equation and

trial and error method

SCOPE OF THE WORK

We must understand that complexity of muffler design is in

incorporating noise reduction on meeting the required

(legislative) targets, without affecting engine performance,

package protected to the available space and ultimately

maintaining reasonable cost targets and production time This

defines scope of our work and intent is to develop a practical

approach to design muffler device which will have following

advantages:

i A brief background on evaluation of muffler concept

design for proto type and validation with new approach

ii A methodology has been developed for optimum design

stages and less cost for muffler design by balancing various

parameters

iii A practical tool to estimate quality of muffler design,

which is used for concept selection and filter out best concept

proposal at initial phase of design

iv A practical approach for muffler design to optimize

product development time and cost by balancing conflicting requirements like noise and back pressure

Design approach by modern CAE tools is used for optimizing overall system design to choose the best concept CFD analysis of muffler included in scope is to verify flow restrictions (engine-back pressure) and TL characteristics is computed by acoustic analysis using 3D analysis LMS virtual lab for noise prediction To establish a design methodology to make design process simpler and less time consuming by making use of acoustic theories [11, 12] and experience in short practical approach to get better design

A PRACTICAL APPROACH -DESIGN METHODOLOGY

The properly designed muffler for any particular application should satisfy often - conflicting demands of at least five criteria simultaneously

Muffler design methodology for a given vehicle involves 7 steps Following are broad steps followed to arrive at a good design of muffler making use of practical experimental data

STEP 1: TARGET SETTING AND BENCHMARKING

The first step is to measure the unmuffled noise spectrum of engine for vehicle under consideration This spectrum is recorded at 1m or 0.5 m and 45degree from exhaust outlet by replacing muffler with normal exhaust pipe The major frequencies are noted down and these become input to design The peaks in the unmuffled spectrum are major contributors

to vehicle exhaust noise and these are needs to attenuate to achieve required noise level Target values are set based on benchmarking (pass-by results) and vehicle class requirements Back pressure values are defined based on engine requirements

STEP 2: TARGET FREQUENCIES

After recording unmuffled noise spectrum, one needs to calculate target frequencies to give more concentration for higher TL Target frequencies are calculated by using engine data as follows,

Theoretical computation The exhaust tones are calculated using following formulae:

[13] (1)

(2)

The exhaust spectrum is measured with suitable

instrumentation for identification of these CFR and EFR An

order tracking analysis can be done to identify order of noise

data

By comparing theoretical target frequencies and frequencies

found out from analysis study, dominant frequencies are

identified

STEP 3: MUFFLER VOLUME

CALCULATION

Based on experience and theory of acoustics for muffler

design for various engines, the following equation is used

Volume of muffler (Vm):

[13] (4)

Now design has to be verified for packaging space that can be

made available for the muffler

STEP 4: INTERNAL CONFIGURATION

AND CONCEPT DESIGN

Based on benchmarking TL and target frequencies, few

concepts of internal configurations for muffler that meets

packaging dimension within volume mentioned above has to

be portrayed

Best possible configuration among the concepts as designed above is selected based on the best acoustic performance (maximum TL) and backpressure (least)

Perforations: Perforated pipe forms an important acoustic element of muffler, which is tuned in line with the problematic frequencies identified in step 2 The diameter of hole to be drilled or punched on pipe is calculated by a equation as given below:

(5)

Porosity: Porosity, σ is given by

[13] (6)

It is important to note that lesser the porosity is more restriction and hence more will be backpressure

The open area ratio Aop is given by,

(7)

At this stage, diameter of hole to be drilled, pitch, number of holes per row and number of rows for each pattern of holes is decided The distance at which perforation starts and at which the perforation ends is also decided Thus, the design of the perforated tube for individual hole patterns is finalized

Figure 1 Properly designed muffler -five criteria.

Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016

Design based on the above criterion is carried forward for

virtual simulations

STEP 5: VIRTUAL SIMULATION

Based on above mentioned approach, different concepts will

be arrived with optimum combinations of different elements

inside volume of silencer Finalized concepts will be verified

virtually using CAE simulation software's towards

achievement of TL and back pressure

CFD ANALYSIS

When steady air flow passes through mufflers, steady

pressure drop is developed which is related to flow and

geometry of air passages Pressure drop in an exhaust muffler

plays an important role for the design and development of

mufflers

Prediction of pressure drop will be very useful for design and

development of muffler To predict pressure drop associated

with the steady flow through muffler, CFD has been

developed over last two decades So the flow prediction can

be made reliable

TRANSMISSION LOSS ANALYSIS

Prediction of transmission loss virtually is an important

analysis required for development of muffler at an initial

design stage There are different software packages available

in market for predicting TL We have used LMS virtual lab

for TL predictions

Limitations of the CAE tools also has to be taken care, as the

co-relation at higher frequencies is difficult since the plane

wave theory holds good only up to 3000 Hz beyond which

wave is no more 2 dimensional but 3 dimensional for which

computations are far complex to match practical results

Hence need of research to blend both strengths of CAE and a

practical result obtained by a practical approach or

methodology is required After completion of simulation best

three concepts will (with less back pressure and higher TL)

be taken forward for prototype manufacturing to check for

TL and back pressure physically

STEP 6: PROTOTYPE

MANUFACTURING

All above stages combined with packaging of engine evolve

design of prototype muffler and can be taken up for

manufacturing

Following are some of important manufacturing

considerations summarized based on experience:

1 There should not be any leakage of gas from one chamber

to another

2 Full welding is better than stitch welding.

3 Acoustic performance of extruded tubes with perforations

is better than the tubes that are made out of perforated and welded sheets

4 CEW or ERW tubes are the common materials used.

5 Either of Crimping or full welding of jacket can be used.

6 Either of flanged or flared tubes can be used as end

connections of the muffler

However, with leakage point of view, flanged connections are better But at the same time, this adds to weight and cost of the exhaust system Bearing all above in mind, a physical prototype is made in such a way that there will not be any tooling investment for prototype

STEP 7: EXPERIMENTAL TESTING AND DESIGN FINALIZATION

The experimental determination of backpressure on engine and TL using two source method for different concepts are verified The prototypes of all concepts that are made at above step are tested for TL to verify target value

TL is difference in sound pressure level between incident wave entering and transmitted wave exciting muffler when muffler termination is anechoic, TL is a property of muffler only In this work an attempt has been made to experimentally measure TL by actually using the experimental set-up Two source techniques gives good results for measurement of TL at the different sound frequencies Also absence of anechoic termination, the decomposition method is found to be ineffective Therefore

we will be using two source methods in calculating TL

TL values obtained from simulations are compared with experiments At the same time if performance of muffler is found to be satisfactory as per engine noise requirement, then the above captured data becomes the input for further backpressure reduction The iteration is continued usually 2

to 3 times to achieve an optimum balance between noise requirement and target of least backpressure and best fuel efficiency

CASE STUDY

DEFINING THE CASE

We have attempted application of these design methodology

on LC diesel engine vehicle Initial we got only basic engine details and uncut layout of vehicle for designing exhaust system

STEP 1: BENCHMARKING

As per design methodology we benchmarked same kind of

engine models to set the target of TL of muffler

Bore diameter (D) = 80 mm, Stroke length (L) = 98 mm,

Number of cylinders (n) = 4, Engine power (P) = 65 hp, Max

RPM (N) = 3500 rpm, Allowable back pressure = 12″ of

H2O, TL noise target (muffler) = 20 dBA

STEP 2: TARGET FREQUENCIES

To find fundamental frequency Cylinder firing rate: CFR to be calculated as per equation (2), Engine firing rate: EFR to be calculated as per equation (3), The first 4 harmonics are to be suppressed as higher order has very little effect on noise The diameter of the holes drilled should suppress these frequencies

Figure 2 Muffler design methodology - schematic representation.

Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016

STEP 3: MUFFLER VOLUME

CALCULATION

Volume swept of each cylinder calculated as per equation (4)

STEP 4: INTERNAL CONFIGURATION

OF MUFFLER AND CONCEPT DESIGN

Diameter of muffler calculated as per equation (5), (6)

Open area and porosity for each muffler concept calculated

by using equation (6) and (7)

Design outputs

Table 1 Design table (output results).

Below mentioned are few important design guidelines

captured from theory of acoustics and experience while

designing the muffler

a Extended inlet and outlet diameter will be minimum 60 to

70 mm for better attenuation results

b Inlet and outlet are introduced 180 degree reversal to

increase the acoustic performance

c From benchmark and theory, 3 expansions chambers are

suggested for noise attenuation (two baffles play a major role

in back pressure and acoustic advantages)

d Hole perforations choose to match frequency that needs to

be suppressed based on CFR and EFR calculations

e la and lb is 15 to 20 mm as per theory of acoustic for good

acoustic performance

f Primary criterion for choosing diameter of the hole is

based on the first four CFR and EFR harmonics

We have made design of concept 01, 02, and 03 with double

expansion chamber Expansion chambers are made of

unequal length in two parts Elliptical chamber is used as we

have advantage of space and better attenuation To get more

attenuation effect inlet and outlet tubes are extended in the

chamber [10, 11]

STEP 5: VIRTUAL SIMULATION

All three concept of muffler designed as per above steps are tested for flow analysis using CFD simulation tool

Figure 3 CFD concept 0A.

Figure 4 CFD concept 0B.

Figure 5 CFD concept 0C.

Assumption and boundary conditions

• Flow is considered to be steady

• Air is considered as medium for computations

• Turbulent flow is considered (K-ε Model)

• Inlet is considered as mass flow boundary condition with

320 kg/hr flow rate

• Inlet temperature of fluid is 520 degree C

• Outlet considered as pressure outlet opened to atmosphere

Figure 2, 3 and 4 shows flow through concept 0A, 0B and 0C

mufflers CFD results shows that concept 0A is good for back

pressure as it provides less back pressure compared to

concept 0B and concept 0C

VIRTUAL ANALYSIS

The mean flow performance of three mufflers considered in

acoustic analysis has been assessed Assumption and

boundary conditions

• Sound termination is anechoic

• Perforated tubes are simulated using Sullivan-Crocker and

Mechel's relation

• Linear steps for analysis is 10 Hz in the frequency range of

10 - 2000 Hz

• Holes on tubes with zigzag pattern are modeled as parallel

pattern

• Embossing on inlet and outlet end cover of muffler is

neglected

As per virtual analysis TL results concept 0A is having better

TL compared to concept 0B and concept 0C at critical

frequencies (20 - 500 Hz or for initial four frequencies) as per figure 06

After completion of simulation we selected best three concepts (with minimum back pressure and higher TL) taken forward for prototype manufacturing to experimental validations of TL and back pressure

STEP 6: PROTOTYPE MANUFACTURING

Final 3 concepts were manufactured by taking care all things from step 6 of design methodology and in such a way that there will not be any tooling investment for prototype

STEP 7: EXPERIMENTAL TESTING AND FINALIZATION OF DESIGN

As explained in step 7 of design methodology we have used two source method for TL experimental validations for proto type manufactured mufflers Flow diagram for the experimental set up of TL measurement is as shown in figure

6 Also all three concepts are tested on engine experimentally for back pressure measurement

OBSERVATIONS

As per figure 8, TL of concept 0A is better compared to concept 0B and 0C at critical frequencies

Figure 6 Comparison virtual acoustic results.

Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016

Figure 9 Comparison analysis and experimental concept

0A.

Figure 10 Comparison analysis and experimental

concept 0B.

Figure 7 Flow diagram for experimental set up TL measurement.

Figure 8 Comparison between experimental results.

Figure 11 Comparison analysis and experimental

concept 0C.

Comparing virtual acoustic results as per figure 6 and

experimental results as per figure 8, TL of concept 0A is

suppressing initial four critical noise frequencies Also figure

9, 10 and 11 is comparison between analysis and

experimental results of frequency v/s Transmission loss for

muffler concept 0A, 0B and concept 0C The dotted vertical

lines represent EFR's and plot explains that TL curve of

proposed concept suppresses critical vertical peaks

The obtained result using experimental setup is compared

with theoretical result and virtual simulation results to find

out best concept for noise Also for back pressure CFD

results are compared with engine back pressure results Based

on this comparison concept 0A is selected for required back

pressure and noise (TL) as results are near provide targets

Further to test selected muffler (0A as per case study) on

vehicle to meet pass-by-noise requirements

CONCLUSIONS

This paper emphasizes on importance of design methodology

- a practical approach from the concept design to proto

manufacturing

and validation of exhaust muffler This design methodology

will help designers in understanding importance of each step

of designing in detail from concept level to validation level

This approach serves purpose of reducing number of

iterations, product development time and cost with better

design

Although practical approach has become an important tool in

making muffler design more of an art than a science, need for

design verification will always be necessary at end of each

step

REFERENCES

1 Stewart, G W “Acoustic wave's filters,” Physics Review

20, 528-551(1922)

2 Divis, D D Jr., Stokes, G M., Morse, D., and Stevens, G.

L., “Theoretical and Experimental Investigation of Muffler with Comments on Engine- Exhaust Muffler Design,” NACA

1192 (1954)

3 Igarashi, J and Toyama, M., Fundamental of acoustical

silencers (I), Aeronautical Research Institute, University of Tokyo, Report no.339, 223-241 (1958)

4 Igarashi, J and Toyama, M., Fundamental of acoustical

silencers (III), Aeronautical Research Institute, University of Tokyo, Report no.351, 17-31 (1960)

5 Munjal, M L., Sreenath, A V and Narasimhan, M V.,

“Velocity ratio in the analysis of linear dynamical system,” Journal of sound and Vibration 26, 173-191 (1970)

6 Munjal, M L., “Velocity ratio cum transfer matrix method

for the evaluation of muffler with neon flow,” Journal of sound and Vibration 39, 105-119 (1975)

7 Young, C I J and Crocker, M J., “Prediction to

transmission loss in mufflers by finite element method,” Journal of Acoustical society of America 57, 144-148 (1975)

8 Chang, Ying-Chun, Yeh, Long-Jyi, chiu, Min-Chie,

“Computer Aided Design on Single Expansion Muffler with Extended Tube under Space Constraints,” Journal of Science, 171-181 (2004)

9 Middelberg, J M., Barber, T J and Leong, T.J.,

“Computational fluid dynamics analysis of the acoustics performance of various simple expansion chamber mufflers,” Acoustics, 123-127 (2004)

10 Rose, T and Jebasinski, R., “Design of Experiment

-Application of a Statistical Evaluation Method to Optimize the Tailpipe Noise of an Exhaust System,” SAE Technical Paper 2003-01-1655, 2003, doi:10.4271/2003-01-1655

11 Erilksson, L J and Thawani, R T., “Theory and Practice

in Exhaust System Design,” SAE Technical Paper 850989,

1985, doi:10.4271/850989

12 Munjal, M L., “Acoustic of ducts and mufflers,” John

Weley and Sons (1987)

13 Bentley, Philip and Morrison, John C., “The Scientific

design of exhaust and intake system,” Robert Bentley, Inc Cambrige, Massachusetts (1971)

Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016

CONTACT INFORMATION

Shital Shah is currently working in the product development

division of Volvo India Pvt Ltd He can be contacted at

shital.shah@volvo.com

www.volvotrucks.com

Gangadhar GS is currently working in the product

development division of Volvo India Pvt Ltd He can be

contacted at

gangadhar.gs@volvo.com

www.volvotrucks.com

DEFINITIONS/ABBREVIATIONS

CFD

Computational Fluid Dynamics

AR

Area Ratio

ER

Expansion Ratio

CFR

Cylinder firing rate

EFR

Engine firing rate

Aop

Open area ratio

TL

transmission loss

A

unit area of surface, m2

V

Volume of Muffler

Vm

Volume of Muffler with factor mm

Vs

Swept volume

n

No of cylinder

d

Diameter of muffler

d1

Diameter of hole

D

ID of cylinder bore

L

Stroke Length

n1

No of holes / row

N

Max RPM

ƒ

frequency Hz

C

Pitch between hole

S

cross-sectional area of duct, m

N

engine speed rpm

ƒ F

firing frequency

L

length of duct m

m

expansion ratio

Vf

Volume factor for muffler