Booming Noise Analysis in a Passenger Car Using a Hybrid-Integrated Approach

SAE TECHNICAL PAPER SERIES 2000-01-0723 Booming Noise Analysis in a Passenger Car Using a Hybrid-Integrated Approach Doo-Ho Lee Kookmin University Woo-Seok Hwang Taegu University Myeong-Eop Kim Korea Advanced Institute of Science and Technology SAE 2000 World Congress Detroit, Michigan March 6-9, 2000 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 Sections 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 2000 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

2000-01-0723 Booming Noise Analysis in a Passenger Car Using a Hybrid-Integrated Approach Doo-Ho Lee Kookmin University Woo-Seok Hwang Taegu University Myeong-Eop Kim Korea Advanced Institute of Science and Technology Copyright 2000 Society of Automotive Engineers, Inc. ABSTRACT A hybrid-integrated approach is presented to analyze the structure-borne booming noise in a passenger car. We identify the critical noise transfer path from the engine to the target by the transfer path analysis. However, it does not give the answer for why the noise transfer function is so high at that path. Therefore, an integrated approach which applies the analysis tools systematically is presented. The running mode analysis gives us the operating motion of each component in the body structure. However, there is no evidence that the components that vibrate severely are the sources of this problem. The modal characteristics from the structural modal test enable us to describe the real motion of the body completely in terms of the structural modes. Similarly, the acoustic modal characteristics from the acoustic modal analysis describe the fundamental behavior of the cabin cavity. The introduction of the experimental running mode data of the structure to the acoustic finite element model makes the hybrid analysis possible. Through the structural and the acoustic modal analysis, we verify the mode that contributes to the booming noise. The panel contribution analysis points out the components that have the greatest influence on the booming noise. The modification of the body based on the results of the hybrid-integrated approach results in the great decrease of the noise level. INTRODUCTION In a passenger car, the NVH performance is one of the most important factors in determining the quality of a vehicle. Since a booming noise caused by the structural vibration is observed frequently, particular attention is paid to reduce the booming noise through the whole car development. The booming noise in the prototype stage can cause a severe problem in the program. Therefore, a systematic approach is needed to solve the booming noise problems effectively. In this paper, we investigate the low frequency booming noise in a passenger car with the experimental and the numerical approach. The experimental techniques include the structural modal analysis, the running mode analysis and the transfer path analysis. The numerical models for acoustic cavity and the experimental vibration data of the body are combined for the panel contribution analysis. The simultaneous applications of these methods identify the mechanism of the structure-borne noise for the complete vehicles accurately. In addition, the results give us very useful information for the design modifications. THE HYBRID-INTEGRATED APPROACH From the beginning of design stage, the design engineers pay much attention to improve the NVH performanace of a car. However, the body of a car is made of many beaded thin shells, which are welded or bolted to each other. Furthermore, there are a lot of melting sheets and various trims on the body. These make the numerical analysis limited tool although the numerical analysis is the most useful one in the early design stage. Since we cannot remove the NVH problems perfectly in the design stage, the booming noise can occur during the evaluation of a prototype vehicle. To tackle the booming noise problems in the prototype stage, there are many tools such as Transfer Path Analysis(TPA ) Running Mode Analysis(RMA) Structural Modal Analysis(SMA) 1

Acoustic Modal Analysis(AMA) Panel Contribution Analysis(PCA) Figure 1. Schematic diagram of the integrated approach The independent results of each method give us much valuable information for the problem. However, the information is not sufficient if only one of those is available. For example, the RMA can indicate which panels are vibrating severely. However, we can not conclude that those panels are the main source of the interior noise. Similarly the PCA can rank the panels according to their contribution to the noise but can not identify the noise transfer path or the structural modes in resonance. The insufficient information on the booming noise phenomenon may lead to some conclusions based on the engineering intuition. This may result in the misunderstanding of the phenomenon or the inefficient loop of trial-and-errors. The trial-and-errors in the prototype stage can cause a delay of the schedule. To avoid the possibility of the delay and to understand the booming noise problem in a vehicle completely, we should use the above mentioned tools systematically as shown in Figure 1. Only the complete understanding of dynamics can extract the optimal design modification for the booming noise and lead to the successful development of a high quality car. It is very difficult to represent the vibro-acoustic behaviors of the full vehicle by the numerical method due to the complex behavior of the fully trimmed body. On the contrary, the experimental techniques describe the structural behavior of the real vehicle effectively. However, the high cost of the prototype cars and the limited time for the evaluation is a huge obstacle for the experimental approaches. Fortunately, it is easy and accurate to describe the acoustic cavity of the car by the numerical method. Therefore, the combination of the experimental model of the structure and the numerical model of the cabin cavity gives an accurate result. We will show that the hybrid approach, which combines the test data with numerical data, is very useful. BOOMING NOISE ANALYSIS BY THE HYBRID- INTEGRATED APPROACH A booming noise is observed at around 1800 during the evaluation test of a prototype passenger car. The car is a mid-size sedan equipped with an inline 4-cylinder, 2- liter engine and an automatic transmission. After several preliminary tests, the noise is categorized as a structureborne noise excited by the explosion in the engine. To analyze the problem, the hybrid-integrated approach is suggested. Figure 2. main muffler first cross member engine center member submuffler subframe Selected paths for the TPA TRANSFER PATH ANALYSIS The transfer path analysis investigates the contribution of each path for the vibration/noise transfer from the sources to the targets. For the engine booming noise problem, the source is the vibration of the engine. And the target is the sound pressure level at the ear position of a passenger. The noise transfer paths are all the connection points between the engine side and the body side. There are three paths along the coordinate directions at each point in Figure 2. The transfer path analysis identifies the most critical path that transfers the engine vibration to the interior noise. The total interior noise is the sum of the contributions through each transfer path as follow[1] where ( P F ) P = Pi = Fi i P : total internal noise P i : internal noise transferred through i-th path F i : operating force at i-th path : vibro-acoustic transfer function at i-th path P F i 2

The operating forces are the multiples of the dynamic stiffnesses of mounts and the relative displacements before and after the mounts. The direct method such as the impact hammer testing or the reciprocal method[2] gives us the vibro-acoustic transfer function from the path point to the target. Pa db Figure 3. Measured Caculated_total Path 5(vertical) Synthesized sound pressures by the TPA We measure the second order operational sound pressures and accelerations on the chassis dynamometer. The impact hammer test at each path point to each direction on the trimmed body gives us the vibro-acoustic transfer functions. During the measurement, the power train, the subframe and the muffler are disconnected completely to avoid the coupling to each other. The dynamic stiffnesses of the engine mounts and rubber elements are measured on an elastomer tester. Figure 3 shows the comparison between the measured sound pressure level and the summed one of the contributions through each path. The synthesized interior sound pressure level shows good agreement with the measured one around 1800. As shown in Figure 3, the connection point 5 in vertical direction has the largest contribution to the noise around 1800. The high contribution of this path results from the high level of the operating force and the vibro-acoustic transfer function in the range as shown in Figure 4. Actually, the FRF from the point 5 to the interior noise in the vertical direction has higher level than those from the other points by 6 to. RUNNING MODE ANALYSIS We identify the critical noise transfer path through the transfer path analysis. At around 1800, the noise transfer function of that critical path is higher than those of the other paths. However, we do not know why the noise transfer function is so high yet. Since the booming noise is a structureborne phenomenon, investigating the vibration of the structure during operation may give the answer. We can understand the structural behavior of the body during the operation by the running mode analysis. Someone calls it the operational deflection shape (ODS) analysis[3]. Through the analysis, we can visualize the motion of the body at any interested driving condition. Figure 5 shows a test set-up for the running mode analysis of the body. An accelerometer attached at the engine block generates a reference signal. It works as a reference for the phase calculation from the other accelerometers even though they are not measured simultaneously. Therefore, we can see the relative motion of the whole body. Figure 6 shows a geometric mesh for the running mode analysis. The second harmonic components of the accelerations to the crankshaft revolutions are picked up through the order tracking analysis. The total degrees of freedom for the acceleration measurements are 1536. Figure 7 shows the operational deflection shape at around 1800 as an example. Pre-am plifier Reference Accelerom eter Figure 5. HP VXI Front-end ENGINE Tachometer Lms CADA-X S/W Microphone Test set-up for the running mode analysis of the body Vibro-acoustic FRF Operating force Pa/N db 10 N Force, N Figure 4. 20 40 60 80 100 120 Hz Operational data at the path 5 in the vertical direction Figure 6. Test geometry for the running mode analysis 3

Figure 7. Operational deflection shape at around 1800 The components that vibrate severely are the first cross member, the subframe, the roof, the rear windshield, the doors, the floor, the parcel shelf, the tire well and the center member. The center member is not shown in Figure 7 because the deflection is too large compared to those of the others. We can assume that some components among those mentioned above cause the high interior sound pressure. The following analyses verify why those components move so much and whether they are related to the booming noise. EXPERIMENTAL STRUCTURAL MODAL ANALYSIS The structural modal analysis leads us to understand the dynamic characteristic of the full vehicle. Based on the modal parameters such as the modal frequencies, the modal vectors and the modal damping estimations, an arbitrary structure motion can be decomposed into the linear combination of the structural modes[4]. This characteristic enables us to describe the real motion of the vehicle completely in terms of the modal characteristics. We select the same geometry points of the running mode analysis shown in Figure 6 for the modal test. There are two electrodynamic exciters at the front and the rear sides in the vertical direction. They shake the body in the burst random patterns. The modal parameters are estimated from the FRF's at every measuring point using Lms Cada-X Modal Analysis module[5]. Table 1 contains the summary of the primary modes around the concerned frequency. Using the extracted modal parameters, we can decompose the motions from the running mode analysis into the linear combinations of the structural modes. There are the modal decomposition weighting factors normalized to maximum one in Table 1. The greatest contribution to the running mode of Figure 7 comes from the 65 Hz mode, which is shown in Figure 8. The motions of the roof and the doors are very similar in Figure 7 and 8. Table 1. Mode No Structural modes of the body Modal Frequency (Hz) Modal Coefficient* (%) Mode Shape Description 1 54.3 46.6 Subframe 2 58.2 23.1 Parcel Shelf 3 61.5 29.3 Trunk Lid 4 65.1 100. Rear Roof, Door 5 68.6 40.8 Center Member 6 72.3 5.5 Floor, Door, Roof 7 79.5 33.7 Roof Center (*): Percentage ratios to maximum coefficient when the running mode decomposed Figure 8. Sturctural mode shape (65Hz) NUMERICAL ACOUSTIC MODAL ANALYSIS Since the booming noise is a vibro-acoustic phenomenon, the acoustic modes of cavity are very important. The numerical tools such as the finite element method or the boundary element method are very useful in the hybridintegrated approach, since they can model and analyze the internal cavity of the vehicle easily. We can combine the numerical cavity model with the experimentally measured body motions or the modal characteristic of the full vehicle. This hybrid approach makes the accurate and efficient analysis possible. The finite element analysis of the cavity model gives us the information regarding the characteristic of the cavity. There is an acoustic finite element model in Figure 9, and this model has 4762 solid elements. Since our concerned frequency range upto 200 Hz is relatively low, we use a 4

very simple model. Many detailed shapes of the cavity are simplified. The small holes of the parcel shelf connect the trunk space with the cabin cavity. The correlation of the numerical results with the experimental ones decide the equivalent size of the hole in the model. 4.0x10-3 Mode 1 Mode 2 3.0x10-3 Mode 3 Mode 4 Mode 5 2.0x10-3 1.0x10-3 0.0 Figure 12. Mode participation factors Figure 9. Table 2. Mode No Finite element model of the cavity Acoustic modes of the cavity Freq. (Hz) Mode Shape Description 1 22.7 Longitudinal mode by the trunk 2 77.6 Longitudinal mode of the cabin 3 100. Lateral mode of the cabin 4 116. Lateral mode of the trunk 5 127. 2 nd longitudinal mode of the cabin Figure 10. The fundamental acoustic mode shapes: the longitudinal(78hz) and lateral(100hz) modes Pa db Measured FE model Figure 11. Interior noise calculated by the finite element model Table 2 shows the acoustic modes calculated by the finite element analysis using SYSNOISE software[6]. There are the fundamental mode shapes in Figure 10. The modal frequencies of the fundamental modes are much higher than the concerned frequency. Therefore, the booming is not from the resonance of the cavity with the excitation. It is the forced response by the engine excitation. To know the participation factors of each acoustic mode to the interior noise, we introduce the running mode data of the body to the acoustic finite element analysis. The acceleration data from the running mode analysis are converted to the velocity boundary conditions of the finite element model of the cavity. To obtain more realistic results, the impedances of the surface material of the trim parts are added to the finite element model as the boundary conditions. We measure the impedance of the surface material by the impedance tube method. The modal superposition method gives the frequency responses due to the vibrations of the body and the interior noise. Figure 11 shows that the interior noises from the numerical analysis and the experiments are very similar. There are the mode participation factors in Figure 12. Note that the 78Hz mode has the greatest influence on the interior noise from 1500 to 2500. This means that the pressure response in the cabin around 1800 will be more sensitive to the change of the front or the rear part panel velocity than that of the central part. PANEL CONTIRIBUTION ANALYSIS We have identified the operational deflection shape of the body and the dynamic characteristic of the body and the cavity. The interior sound pressure is related to the motion of the structure. And there are some components that vibrate a lot during the operation. The panel contribution analysis[7,8] will derive the information that verifies the relation of those results from the previous analysis. 5

Figure 13 shows the procedure used in the hybrid approach for the panel contribution analysis. The boundary element model in Figure 14 is the same with the envelope of the acoustic finite element model except for the double nodes in the corner points. The calculated interior sound pressure is very similar to the measured one in Figure 15. There are the velocity boundary conditions and the corresponding surface sound pressures around 1800 in Figure 16 and 17, respectively. The deflections of the several regions are large in Figure 16. However, the pattern of surface sound pressure is very similar to the first longitudinal cabin cavity mode, which governs the sound pressure around 1800. Figure 16. Velocity boundary conditions at around 1800 Figure 13. Procedure of the panel contribution analysis using the hybrid approach Figure 17. Surface sound pressure at around 1800 Figure 14. A boundary element model of the cavity Measured BE Model imag. 4 10 1 0 2 9 6 3 8 0-1 Floor 0-2 RH FR Door 0-3 FR W/Shield 0-4 LH FR Door 0-5 RR W/Shield 0-6 LH RR Door 0-7 Roof 0-8 RH RR Door 0-9 P/Shelf 0-10 Trunk 0 real 7 5 Pa db Figure 15. Calculated sound pressure by the BE model Figure 18. Panel contributions for the noise at around 1800 Figure 18 represents the calculated panel contributions to the interior noise around 1800. The contributions of the rear windshield and the roof are very dominant. The comparison of the structural vibration in Figure 16 and the results of panel contribution shows that the local deformation of rear roof makes most of the booming noise. 6

STRUCTURE MODIFICATION AND VALIDATION The followings are the summary of the results from the hybrid-integrated analysis. The vibration energy of the engine around 1800 transfers to the body mainly through the first cross member. That energy makes the connection area between the roof and the rear windscreen vibrates severely. The motion of this area generates the high pressure in the interior cavity. Based on the summary, we reinforce the first cross member and the rear roof rail to improve the booming noise problem. Figure 19 shows the results of the reinforcement. The reduction of the noise level is 3 to 6 db around 1800. The reinforcement changes the noise transfer function from the first cross member to the interior noise as shown in Figure 20. It means that the interior noise level is lowered due to the reduction of the peak level of the noise transfer function by the modifications. Pa db Figure 19. Sound pressures after the structural modifications Pa/N db Figure 20. Noise transfer functions after the structural modifications CONCLUSIONS Initial Member modification Roof modification 20 40 60 80 100 120 Hz Initial Member modification Roof Modification A hybrid-integrated approach is presented to analyze the structure-borne booming noise in a passenger car. The conclusions from the analysis are as follows: The simultaneous applications of the analysis tools give us the accurate descriptions of the mechanism related to the booming noise. The analysis includes the transfer path analysis, the running mode analysis, the structural modal analysis, the acoustic modal analysis and the panel contribution analysis. The hybrid approach which combines the experimental data with the numerical model makes the accurate and efficient analysis possible. The numerical modeling of the trimmed body is not so easy to describe the acoustic response exactly. By using the experimental data of the structure as boundary conditions for the numerical acoustic analysis, the efficiency and the accuracy of the analysis improve very much. The modification of the body based on the results of the hybrid-integrated approach results in the decrease of the noise level. The hybrid-integrated approach is a very systematic and efficient one to solve the structure-borne noise problem in a passenger car. REFERENCES 1. T. Jee and Y.B. Choi, Transfer Path Analysis on the Passenger Car Interior Noise, J. of KSNVE, Vol. 9, No.1, pp.97~102,1999(in Korean). 2. P.J.G. Linden and M. Mantovani, The validity of Reciprocal Acoustic Transfer Function Measurements on Trucks for Pass-by Noise, Proceedings of INTERNOISE 96, pp. 2661~2666, 1996. 3. K. Wyckaert and H.V.D. Auweraer, Operational Analysis, Transfer Path Analysis, Modal Analysis: Tools to Understand Road Noise Problems in Cars, Proceedings of the SAE Noise and Vibration Conference, pp. 139~143, 1995. 4. W. Helen, S. Lammens and P. Sas, Modal Analysis Theory and Testing, Katholieke Universiteit Leuven, 1995. 5. Lms International, CADA-X User s Manual: Modal Analysis, 1997. 6. Numerical Integration Technology, SYSNOISE 5.3 User s Manual, 1996. 7. W. Hendricx, Y.B. Choi, S.W. Ha and H.K. Lee, Experimental Body Panel Contribution Analysis for Road Induced Interior Noise of a Passenger Car, Proceedings of the SAE Noise and Vibration Conference, pp. 351~356, 1997. 8. PJG Van der Linden and Ph Veret, Experimental Determination of Low Frequency Noise Contributions of Interior Vehicle Body Panels in Normal Operation, Proceedings of the SAE International Congress, pp. 61~66, 1996. 9. L. Gielen, PJG Van der Linden and R. Deges, Identification, Quantification and Reduction of Structural-Borne Road Noise in a Mid-Size Passenger Car, Proceedings of the SAE International Congress, pp. 67~74, 1996. 7