Prediction of exterior sound field of an automotive for airborne excitation and transmission to the interior of the vehicle

Transcription

1 Prediction of exterior sound field of an automotive for airborne excitation and transmission to the interior of the vehicle Arnaud Caillet 1 ; Antoine Guellec 2 ; Ludovic Dejaeger 3 ; Arthur Henry 1 1 ESI GmbH, Germany 2 Audi AG, Germany 3 Faurecia, France ABSTRACT In the automotive industry, several airborne noise sources are highly perceived by the car passengers. The perceived quality of the vehicle depends how these noises are transmitted to the interior of the vehicle. The most dominant airborne sources are: the engine, the tires and the exhaust. Simulation models can be created to represent the exterior sound pressure level around the vehicle for a given airborne source. An extensive measurement campaign has been done in order to collect informations to validate BEM simulation models using point acoustic sources at several locations. In this paper, the results of the simulation model will be compared to the measurements. Keywords: Automotive, Exterior sound field, Interior vehicle acoustics, BEM 1. INTRODUCTION This paper presents the analysis made on behalf of Audi AG by Faurecia and ESI Group to investigate better ways how to predict exterior sound pressure field due to airborne excitation and apply this exterior sound field to models used to predict the sound pressure level within the vehicle. An extensive measurement campaign has been completed by Faurecia in order to generate validation data for the simulation models. Monopole sources have been located at the typical location of airborne noise sources (engine, Front wheel house, rear wheelhouse, exhaust). This measurement campaign will be detailed within the first part. In the second part of the paper, the simulations run by ESI Group will be explained. A BEM model has been prepared in order to simulate the exterior pressure sound field due to airborne excitations. The results obtained from this external BEM model will be used later on as input for other models. 2. Measurements in semi-anechoic chamber In this study, Faurecia was in charge to perform the experimental measurements in order to validate the simulation models. In this stage of the study, the exterior sound pressure level around the vehicle and the noise transmitted inside the vehicle have to be characterized using an external airborne monopole source. 2.1 Room description The measurement campaign was conducted in the semi-anechoic chamber of the Faurecia Acoustic TechCenter in Mouzon, France. This chamber, classified as class 1 Room according to ISO3745, has a length of 11m, a width of 9m for a height of 6m. This chamber has cutting frequency of 68Hz, which offers optimum conditions to measure sound pressure level outside the vehicle. For measurements in 1 Arnaud.caillet@esi-group.com 2 Antoine.guellec@audi.de 3 Ludovic.dejaeger@faurecia.com 7419

2 vehicle operational conditions, this chamber is equipped with a 4x4 chassis dyno (Figure 1). 2.2 Monopole source Figure 1 Semi-anechoic chamber at Faurecia Acoustic TechCenter The exterior airborne source excitation was provided by a volume velocity source. Figure 2 Noise of used volume velocity source FRF are defined in the p/q sense, where p is the sound pressure level measured at a response point and Q the volume velocity of the monopole source. The volume velocity inter-spectrum is determined using the microphone doublet placed in the noise [1] (Figure 2). While CQQ is the volume velocity source inter-spectrum, and CQp the cross-spectrum between the source and the reference signal, the FRF Response microphone / volume velocity are then computed following equation 1: CQp HQp (1) CQQ The sound directivity of this source has been characterized (Figure 3) and could be considered as a monopole source up to the third octave 2 khz. Higher in frequency range, a preponderance has been observed in the directivity in the front of the acoustic source (Figure 4). 7420

3 Figure 3 Source directivity measurement in free-field room Figure 4 Sound Pressure Level directivity diagram for 6 different third octave frequencies 2.3 Measurement Configuration For this specific study, Audi has provided a car without mechanical parts (no engine, no transmission), but with complete interior trim. The sound pressure level characterization inside the vehicle was made for 2 car configurations: - Body In Blue: without interior trim, but with all opening mounted (doors, tail gate) and all holes, from pass-through, sealed; - TrimBody: with interior trim, with all opening mounted (doors, tail gate) and all holes sealed; 7421

4 Figure 5 Studied vehicle in semi-anechoic chamber The car body was putted on air cushion in order to benefit of free conditions for other stages of measurements with structure borne excitation (Figure 5). The acoustic source was placed one after the other in 4 different positions (Figure 6). Figure 6 Monopole source positions In order to measure the sound field all around the vehicle, and ensure the precise location of each microphone relative to the vehicle (essential to allow a good correlation between simulation results and measurement results), different racquets have been built and positioned. For each position of the racket, this one has been defined in the FE model (Figure 7). Microphones were placed at least 10 cm from the body. 868 points have been defined around the vehicle. With 20 free-field microphones 7422

5 available, around 45 different setups were needed for each source location. Figure 7 On top: microphone positions defined in the vehicle model Pictures: different measurement setups. To measure the sound pressure level inside the vehicle 32 microphones position have been defined and also reported in the FE model (Figure 8). 7423

6 Figure 8 On top: microphone positions inside defined in the vehicle model Pictures: different measurement setups for both vehicle configurations. 2.4 Measurement results and post-processing FRF between each response microphone and source volume velocity were acquired up to 10 khz. The measurement results have been reduced in third octave frequency and the exterior sound field represented on cartography for each source location. Below are shown the representations for source location wheelhouse rear left at different third octave frequencies (Figure 9, 10, 11). Figure 9 Exterior sound field for source location rear left wheelhouse for 1/3 octave frequency 100 Hz and 1/3 octave frequency 200 Hz. 7424

7 Figure 10 Exterior sound field for source location rear left wheelhouse for 1/3 octave frequency 500 Hz and 1/3 octave frequency 1000 Hz. Figure 11 Exterior sound field for source location rear left wheelhouse for 1/3 octave frequency 2000 Hz and 1/3 octave frequency 8000 Hz. 3. Simulating the exterior sound pressure level using a BEM model The recent improvements on software and hardware sides allow now to compute BEM models up to high frequencies. In this part, the computation of the exterior sound pressure level will be presented using ESI Group commercial software VA One (2). The simulation has been done up to 1,6kHz. The mesh generated describes the waves with 6 elements at the maximum frequency of computation. The BEM mesh has been morphed to the shape of the vehicle. No coupling between the structure of the vehicle and the exterior sound field is accounted within this first model. The shape of the vehicle is rigid and the floor is modelled as an infinite rigid baffle. The exterior microphones arrays used within the measurement (Figure 7) are reproduced in the model in order to capture the exterior sound pressure level and compare the simulated pressure with the measured one. The air cushion boxes supporting the vehicle are also included within the model in order to account for the wave scattering around them. The sources are represented as monopoles and located at the exact same location as what was used during the measurement. The amplitude of the sources is normalized in order to compare directly the transfer functions. Figure 12 is illustrating the BEM model created to compute the exterior pressure sound field. 7425

8 Figure 12 BEM exterior sound field model The exterior sound pressure levels have been averaged for each microphone arrays used during the measurements. The model has been computed with a frequency step of 20Hz from 100Hz till 1600Hz. The results are displayed within figure 13 and 14 using an excitation located within the rear left wheel house. A good correlation between the measurement and the simulation can be observed at the different microphone arrays Figure 13 Underfloor microphone arrays measurement versus simulation for a monopolar excitation located at within the rear left wheelhouse 7426

9 Figure 14 Underfloor and side microphone arrays measurement versus simulation for a monopolar excitation located at within the rear left wheelhouse The pressure distribution can be also observed as contour plot and directly compared with the measured pressure maps displayed in paragraph 2.4. The Figure 14 shows the simulated pressure distribution at 100Hz and 1000Hz for the rear left wheelhouse excitation case. Figure 14 Simulated sound pressure distribution at 100Hz and 1000Hz for rear left Wheelhouse excitation 4. Conclusion This paper presented a comparison study of vibro-acoustics models for airborne source located at typical locations used in the automotive industry. The results show that the BEM method is reaching a satisfying level of accuracy to predict the exterior sound pressure level. This is a good basis to start a investigating the simulation methods to predict the interior sound pressure level. In that perspective, a fully coupled FEM/BEM model would be used for the lower frequencies and a SEA model for the higher frequencies. First results have already obtained with these methods. The models are currently being refined to improve the quality of the correlation between the simulation and the measurement. 7427

10 Acknowledgement The authors would like to thank Audi AG for supporting the analysis and K. Ejjed and A. Ben Abid from ESI Group for preparing and running the vibro-acoustics models. REFERENCES 1. Batel M., Hald J.. Brüel & Kjær Sound & Vibration Measurements A/S. Source/Path Contribution methodologies for airborne generated noises. 2 "VA One 2016 User s Guide",