Sound Power Measurement in a Semi-Reverberant, Volume Deficient Chamber 2015-01-2359

Sound power can be determined using a variety of methods, but precision methods require the volume of the noise source to be less than 1% of the chamber volume leading to relatively large test chambers. Automotive torque converter performance and noise testing is completed in an enclosed metallic test fixture which inhibits the use of precision methods due to volume and space limitations. This paper describes a new method developed to accurately determine sound power of an automotive torque converter in a relatively small enclosure through characterization of the test environment. The test environment was characterized using two reference noise sources designed to represent torque converter noise output and physical geometry. Sound pressure levels of the sources were measured at multiple microphone locations and at three source amplitude levels to characterize the environment. Test results were analyzed statistically to determine the microphone positions that best represent the overall sound levels in the chamber. Optimum measurement positions were found to depend on source size but independent of source amplitude. Accuracy was determined based on the variance between the sound pressure levels at each microphone position. Sound power correction factors were found using sources of known sound power and the optimum microphone positions. The new method allows estimation of sound power of operational torque converters in a dynamometer test fixture.