High-Frequency Time Domain Source Path Contribution: From Engine Test Bench Data to Cabin Interior Sounds

This work presents an application of airborne source path contribution analysis with emphasis on prediction of wideband sounds inside a cabin from measurements made around a stand-alone engine. The heart of the method is a time domain source path receiver technique wherein the engine surface is modeled as a number of source points. Nearfield microphone measurements and transfer functions are used to quantify the source strengths at these points. This acoustic engine model is then used in combination with source-to-receiver transfer functions to calculate sound levels at other positions, such as at the driver's ear position. When combining all the data, the in-cabin engine sound can be synthesized even before the engine is physically installed into the vehicle. The method has been validated using a powertrain structure artificially excited by several shakers playing band-limited noise so as to produce a complicated vibration pattern on the surface. First the excited structure is studied alone; next a vehicle cabin was lowered onto the structure without touching. As a result we can compare the combination of using only in-vehicle operating/transfer data or using powertrain only operating data and in-vehicle transfer data for synthesizing interior sounds. Very good agreement between the two procedures was obtained and comparable to the actual sound measured inside the cabin during operation. In addition to verifying the above procedure, the same near-field microphone setup, whether powertrain alone or with cab on top, can be used to assess the radiated sound power from the vibrating structure. The procedure is outlined and the obtained sound power spectra are validated against a standardized hemisphere sound power measurement showing very good agreement in general.