Muffler design, development and validation for pdf

Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 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 ABSTRACT 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 onedimensional 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 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-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 onedimensional 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 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) Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 Figure Properly designed muffler -five criteria (3) 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 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 The diameter of hole to be drilled or punched on pipe is calculated by a equation as given below: (5) STEP 3: MUFFLER VOLUME CALCULATION Porosity: Porosity, σ is given by Based on experience and theory of acoustics for muffler design for various engines, the following equation is used Volume of muffler (Vm): [13] (6) [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 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 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 dimensional but 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: There should not be any leakage of gas from one chamber to another Full welding is better than stitch welding Acoustic performance of extruded tubes with perforations is better than the tubes that are made out of perforated and welded sheets CEW or ERW tubes are the common materials used Either of Crimping or full welding of jacket can be used 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 to 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 Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 Figure Muffler design methodology - schematic representation STEP 1: BENCHMARKING STEP 2: TARGET FREQUENCIES As per design methodology we benchmarked same kind of engine models to set the target of TL of muffler To find fundamental frequency Cylinder firing rate: CFR to be calculated as per equation (2), 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 Engine firing rate: EFR to be calculated as per equation (3), The first harmonics are to be suppressed as higher order has very little effect on noise The diameter of the holes drilled should suppress these frequencies 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 5: VIRTUAL SIMULATION All three concept of muffler designed as per above steps are tested for flow analysis using CFD simulation tool 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 Design table (output results) Figure CFD concept 0A 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, expansions chambers are suggested for noise attenuation (two baffles play a major role in back pressure and acoustic advantages) Figure CFD concept 0B 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] Figure CFD concept 0C Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 Figure Comparison virtual acoustic results 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, and 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 concepts were manufactured by taking care all things from step 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 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 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 Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 Figure Flow diagram for experimental set up TL measurement Figure Comparison between experimental results Figure Comparison analysis and experimental concept 0A Figure 10 Comparison analysis and experimental concept 0B Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 REFERENCES Stewart, G W “Acoustic wave's filters,” Physics Review 20, 528-551(1922) 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) Igarashi, J and Toyama, M., Fundamental of acoustical silencers (I), Aeronautical Research Institute, University of Tokyo, Report no.339, 223-241 (1958) Figure 11 Comparison analysis and experimental concept 0C Comparing virtual acoustic results as per figure 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 Igarashi, J and Toyama, M., Fundamental of acoustical silencers (III), Aeronautical Research Institute, University of Tokyo, Report no.351, 17-31 (1960) 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) 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) 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) 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) 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) This paper emphasizes on importance of design methodology - a practical approach from the concept design to proto manufacturing 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 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 11 Erilksson, L J and Thawani, R T., “Theory and Practice in Exhaust System Design,” SAE Technical Paper 850989, 1985, doi:10.4271/850989 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 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 n No of cylinder Shital Shah is currently working in the product development division of Volvo India Pvt Ltd He can be contacted at d 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 Diameter of muffler d1 Diameter of hole D gangadhar.gs@volvo.com www.volvotrucks.com DEFINITIONS/ABBREVIATIONS CFD ID of cylinder bore L Stroke Length Computational Fluid Dynamics n1 Area Ratio N Expansion Ratio ƒ Cylinder firing rate C Engine firing rate S Open area ratio N transmission loss ƒF unit area of surface, m2 L Volume of Muffler m Volume of Muffler with factor mm Vf AR ER CFR EFR Aop TL A V Vm Vs Swept volume No of holes / row Max RPM frequency Hz Pitch between hole cross-sectional area of duct, m engine speed rpm firing frequency length of duct m expansion ratio Volume factor for muffler Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 fc cut off frequency P Engine power, hp σ Porosity la Length at which perforation starts on pipe lb Length at which perforation ends on pipe CEW Cold drawn electric resistance welded ERW Electric resistance welded 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 This 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 ISSN 0148-7191 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 SAE Customer Service: Tel: 877-606-7323 (inside USA and Canada) Tel: 724-776-4970 (outside USA) Fax: 724-776-0790 Email: CustomerService@sae.org SAE Web Address: http://www.sae.org Printed in USA [...]... transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE ISSN 0148-7191 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 SAE Customer Service: Tel: 877-606-7323 (inside USA and Canada) Tel: 724-776-4970...Downloaded from SAE International by Univ of California Berkeley, Tuesday, July 26, 2016 fc cut off frequency P Engine power, hp σ Porosity la Length at which perforation starts on pipe lb Length at which perforation ends on pipe CEW Cold drawn electric resistance welded ERW Electric resistance welded The Engineering Meetings Board has approved this paper for publication It has successfully completed