Characterization and Validation of Acoustic Cavities of Automotive Vehicles

Size: px

Start display at page:

Download "Characterization and Validation of Acoustic Cavities of Automotive Vehicles"

Emerald George
5 years ago
Views:

1 Characterization and Validation of Acoustic Cavities of Automotive Vehicles John G. Cherng and Gang Yin R. B. Bonhard Mark French Mechanical Engineering Department Ford Motor Company Robert Bosch Corporation University of Michigan-Dearborn 0901 Oakwood Boulevard 3800 Hills Tech Drive 4901 Evergreen Rd. MD 99 Farmington Hills, MI Dearborn, MI 4816 Dearborn, MI 4816 U.S.A U.S.A. U.S.A. ABSTRACT Passenger compartment acoustic cavity resonances present challenges for NVH engineers. They can be either excited by powertrain or by road inputs. Their coupling with these inputs can cause discomfort and unacceptable sound pressure levels for the driver and passengers. A systematic study of acoustic cavity resonances has been conducted by using a simple analytical expression and the finite element analysis. Both analytical approaches are validated with experimental testing which were performed by using a representative test cabin. For the first several acoustic modes, two analytical approaches yield acceptable approximations with the experimental results. INTRODUCTION The issue of vehicle interior noise is becoming increasingly important as secondary characteristics such as vibration and sound levels are replacing primary ones such as reliability in vehicle purchasing decisions. Interior noise can come from many different sources, but a common one is cabin boom. These phenomena can be generally described as a low frequency (less than 00 Hz) acoustic resonance that is driven by the structure. Structural inputs usually come from either the powertrain or structural response from road inputs. Similar studies can be found in various publications [Ref. 1-5]. Matsuyama and Maruysma investigated the cab boom in a mini van with a 4 cylinder engine which had an unbalanced second order inertia imbalance. They looked at the body transfer function by the Reciprocal Method. This technique was used to determined mode shapes of the structure and led to structural changes. ee and Hwang discussed how the testing procedures can be integrated to solve interior noise problems. They used several analytical methods including Transfer Path Analysis, Running Modal Analysis, Structural Modal Analysis, Acoustic Modal Analysis, and Panel Contribution Analysis. Their work integrated these methods to successfully resolve the 1800 rpm moan issue. Kashani and Orzechowski focused on the development of an active boom noise damping system. The system uses both acoustic and panel vibration to activate the speaker to reduce the boom levels in the vehicle. The speaker is controlled to provide similar response as a Helmholtz resonators tuned to different frequencies. Chen and Ebbitt identified sound absorption characteristics of different seat covers. Their work showed that the seat can have meaningful sound absorption in the boom frequency range (00 Hz and below). The objective of this study is to characterize both quantitatively and qualitatively the behavior of the acoustic cavity in vehicles of the boom frequencies (under 00 Hz). We found that a simple analytical calculation showed reasonable correlation with a detailed finite element analysis and with the experimental results. Therefore, this simple expression can be used as a design tool in the early stage of vehicle development. ANAYTICA RESPONSE ANAYSIS At preliminary design level, there is often little need for precise calculations. Fast approximations are much more useful since they can indicate fundamental trends in vehicle performance and suggest, to a first order, how different concepts might perform. For these reasons, approximate analytical solutions are often used to determine acoustic cavity resonances. The familiar wave equation can be solved for a rectangular cavity with appropriate boundary conditions [6] to yield c l m n ω = π f freq. = + + x y (1) z Similarly, the equation for an open ended cavity is 90

2 c l m n f freq. x y ω = π = + + () z Where x is the length, y is the width and z is the height of a rectangular box. In addition, l is the x-mode number, m is the y-mode number, and n is the z-mode number. These equations were applied to a simple test cabin and various vehicle cabins. The results of a pick-up and a passenger van are presented below. MID-SIZE 4-DOOR PICKUP The mid-size 4-door pick-up cavity configuration yields excellent correlations between analytical calculations and finite element analysis. The reasonable uniform cavity sides, no large corrugations, and parallel surfaces of a pick-up are the characteristics of a rectangular box of which the analytical equation is derived. Figure 1 presents the first six pressure mode shapes from the FE model. These mode shapes provide valuable information for analyzing the cabin acoustic performance. For example, mode #1 (69.9 Hz) and mode #5 (138 Hz) present an issue for the back seat passengers. Both mode # (99.9 Hz) and mode #3 (13.4 Hz) will provide high sound pressure levels for the rear passengers ears on the side of the cavity especially, if the passenger is a short person. Mode # will also present the same concern for the passenger in the front. Mode #4 (134.9 Hz) will be an issue for tall people in the front seat. Therefore, the sound levels are highly geometric specific which in conjunction with the typical occupant in certain seats must be taken into consideration when evaluating acoustic acceptability. Mode Hz Mode Hz Mode 99.9 Hz Mode Hz Mode Hz Mode Hz Figure 1. First Six Modes of a Pickup. Table 1. Analytical and FE Predictions for a Pickup Mode Mode Shape Analytical Calculation FEA Model 1st mode 1, 0, Hz 69.9 Hz nd mode 0, 1, Hz 99.9 Hz 3rd mode 1, 1, Hz 13.4 Hz 4th mode 0, 0, Hz Hz 5th mode, 0, Hz Hz 6th mode 1, 0, Hz Hz 7th mode 0, 1, Hz Hz 8th mode, 1, Hz Hz 9th mode 1, 1, Hz Hz 10th mode, 1, Hz Hz FU SIZE VAN Full size van booms have long been an industry wide issue. The prime reason is that the fundamental rear suspension pinion pitch frequency aligns very closely with the 1st longitudinal cavity resonance mode. This pinion pitch mode can easily be excited by road inputs and 1st order drive line imbalance in normal vehicle cruising conditions. A full size van has an ideal cavity shape for predicting acoustic cavity resonances. The van body makes nearly a perfect rectangular acoustic cavity. An excellent correlation between analytical calculation and FE prediction was found. Up to 117 Hz the analytical calculations only vary 4% from FEA results. Figure shows the first six mode shapes from the full size van FE model. In reviewing the distribution of the sound pressure levels for the various modes, it can be seen that only mode #1 (38. Hz) and mode #3 (8.8 Hz) may create high sound pressure levels for the front seat passengers. These frequencies are very easily excited by 1st and nd order drive line inputs in the normal cruising speeds of the vehicle. In reviewing the FEA mode shapes we also see occupants in the mid-row seats could have a moderate sound pressure level at 74 Hz while the rear seat passenger will have high sound levels with the mode #1 at 38. Hz. Table presents the comparisons between the analytical calculations and the FE mode. A better correlation than the pickup case is obtained. For mode #5, we have grouped FEA frequencies Hz, Hz., and Hz together since they all exhibit the same fundamental modal shape with slight differences due to minor local cavity geometry differences. Table 1 presents the comparisons between the analytical calculations and FE analysis of first ten modes. Good agreements are found between two approaches, especially, in the low frequencies up to 138 Hz. This implies that small deviations in geometry between a rectangular box and the FE model are not so significant in predicting lower acoustic cavity resonances. 91

Mode 1 38. Hz Mode 74.0 Hz Mode 3 8.8 Hz Mode 4 91.1 Hz The test cabin has no acoustic and vibration treatments except baking mastic damper on some of panels.

It was found that the felt performs a slightly better than other absorption material in the frequency range above 500 Hz.

3 Mode Hz Mode 74.0 Hz Mode Hz Mode Hz The test cabin has no acoustic and vibration treatments except baking mastic damper on some of panels. In order to understand the effects of sound absorption on acoustic cavity resonances, three floor treatments were used, i.e. a fiber felt of 19 mm thick with 140 mm x 1134 mm in size, a closed cell foam of 10 mm thick with 1194 mm x 1194 mm in size and an open cell foam of 5 mm thick with 953 mm x 987mm in size. It was found that the felt performs a slightly better than other absorption material in the frequency range above 500 Hz. Figure 4 presents the comparison between no floor covering and the floor covered with 19 mm felt. Mode Hz Mode Hz Mode Hz Mode Hz In acoustic boom ranges under 00 Hz the absorption material makes little or no difference. Above 00 Hz the floor absorption material makes a significant difference since the higher frequency/shorter wavelength sound waves are more easily absorbed by the material. There is another phenomenon happening at the higher frequencies beside the overall reduction of sound pressure. The peak to peak variation is also greatly Figure. First Six Modes of a Full Size Van Table. Analytical and FE Predictions Full Size Van Mode Mode Shape Analytical Calculation FEA Model 1st mode 1, 0, Hz 38. Hz nd mode, 0, Hz 74.0 Hz 3rd mode 0, 1, Hz 8.8 Hz 4th mode 1, 1, Hz 91.1 Hz 5th mode 0, 0, Hz Hz, Hz, Hz 6th mode 0, 0, Hz Hz EXPERIMENTA VAIDATION In order to validate both analytical and FE predictions as well as to determine the sensitivities of an acoustic cavity, a vehicle cavity cabin was used as shown in Figure 3. This cabin is divided into two cavities by a partition wall. The front cavity is covered by an inclined panel to simulate the engine compartment and the rear cavity has one glass window plus a removable door to simulate the passenger compartment. reduced. Figure 4. Comparison between no floor covering and covered with 19 mm felt. EFFECTIVENESS OF CAR SEATS Car seats provide significant absorption in the vehicle cavity. Two car seats of a sub-compact vehicle were used to determine their effectiveness to the acoustic cavity resonances. Two orientations of the car seat were investigated as show in Figure 5. Figure 3. Test Cabin EFFECTIVENESS OF ABSORPTION MATERIA Figure 5. Seat Orientations in the Test Cabin Figure 6 presents the comparison between no car seats and two car seats placed side by side up to 800 Hz. It is 9

found out that even above 100 Hz, significant noise reduction was observed. At the higher frequencies 300 800 Hz, up to 15 0 db reduction in sound pressure level was achieved.

We may conclude that the influences on the resonant frequency shifts as well as the amplitude are mainly depending on the mass coupling effect between the traveling wave and Figure 6.

4 found out that even above 100 Hz, significant noise reduction was observed. At the higher frequencies Hz, up to 15 0 db reduction in sound pressure level was achieved. front to rear orientation has slightly more frequency shifts at fourth mode, which is a combination of longitudinal and lateral mode, than the side by side orientation. We may conclude that the influences on the resonant frequency shifts as well as the amplitude are mainly depending on the mass coupling effect between the traveling wave and Figure 6. Comparison between No Seat and Two Seats Side by Side (800 Hz) In order to further investigate the effectiveness of car seats in low frequency range, a frequency spectrum of 00 Hz was plotted as shown in Figure 7. It was discovered interestingly that not only the amplitudes of the resonances are substantially reduced but the resonant frequencies also shifted to the lower values. Acoustic Response (db) No Seats Two Seats Side by Side This lowering the resonant frequency may be due to the mass coupling between the air and the seats. Figure 7. Comparison between No Seat and Two Seats Side by Side (00 Hz) Figure 8 shows the comparison between two orientations of the two seats, i.e. first is side by side and second is one seat placed in front of the other. The side by side orientation has slightly lower amplitude than the front to rear situation. Regarding the resonant frequency shifts, it is found that the side by side orientation causes more frequency shift than front to rear orientation at first resonant frequency, which is a longitudinal mode. The second resonant frequency shifted more to the lower value in the front to rear orientation than the side by side orientation, since it is a lateral mode. The third resonant frequencies of two seat orientations about the same, because it is a top-to-bottom vertical mode. Again, the the medium. Figure 8. Comparison of Seat Placements Finite element model of the test cabin including both car seat orientations was developed. The seat was considered as a ten time heavier mass cavity than air cavity. Table 3 presents both experimental results and finite element model predications. The FE model predication correlates well with the test results. The seat element is connected to the adjoining air elements in the model. This finite element modeling assumption provides good model correlation thus supporting our theory that the mass of the seat is coupling with the air and thus lowering the frequency due to the addition of its greater mass. CONCUSIONS The following conclusions were obtained from this study: Based solely on the length, width, and height of a vehicle s acoustic cavity, the results from an analytical equation are typically 5-6 % different from those obtained by FEA for low frequencies. Below 00 Hz, floor absorption materials have little effect. Above 00 Hz, floor absorption materials can be very effective. Car seats appear to have significant influences on both the amplitude of the sound pressure level and the resonant frequency shifts. Mass coupling effect between the mass of air and the mass of car seats appears to cause the resonant frequency shifting in an acoustic cavity. REFERENCES 1. S. Matsuyama, S. Maruyama; Booming Noise Analysis Method Based on Acoustic Excitation Test, proceedings of SAE International Congress and Exhibition, Detroit, MI, February 3 6, P. ee, W. Hwang, M. Kim; Booming Noise Analysis in a Passenger Car Using a Hybrid-Integrated 93

5 Approach, proceedings of SAE World Congress, Detroit, March 6-9, P. Chen, G. Ebbitt; Noise Absorption of Automotive Seats, proceedings of SAE Noise and Vibration Conference and Exhibition, Traverse City, MI, May 17 19, Kashani, J. Orzechowski; Active Boom Noise Damping of Dodge Durango, SAE paper No , Maruyama, A. Hasegawa, Y. Hyoudou; Interior Noise Analysis Based on Acoustic Excitation Tests at ow- Frequency Range, International Congress and Exhibition, Detroit, MI February 3 6, D.M. Howard, J. Angus; Acoustics and Psychoacoustics, Butterworth-Heinemann, inacre House, Oxford, E. Kinsler, A.R. Frey, A.B. Coppens, J.V. Sanders; Fundamentals of Acoustics; John Wiley & Sons, New York, 000. Mode Number Mode Shape 1, 0, , 0 0, 0, 1 1, 1, 0 1, 0, 1 Test Results - no seats Hz 170 Hz 185 Hz 10 Hz 30 Hz - seats Hz 165 Hz 168 Hz 194 Hz 198 Hz side by side - seats, one 118 Hz 160 Hz 164 Hz 190 Hz 199 Hz in front of other FEA Results - no seats 118 Hz 160 Hz 167 Hz 199 Hz 05 Hz - seats 106 Hz 153 Hz 163 Hz 187 Hz 199 Hz side by side - seats. one in front of other 113 Hz 151 Hz 164 Hz 189 Hz 00 Hz Table 3. Experimental Test Results and FE Predictions for the Test Cabin 94

Car Cavity Acoustics using ANSYS

Car Cavity Acoustics using ANSYS Muthukrishnan A Assistant Consultant TATA Consultancy Services 185,Lloyds Road, Chennai- 600 086 INDIA Introduction The study of vehicle interior acoustics in the automotive