Influence of Model Detail Level on the Prognosis Quality of the Perceived Sound of an Electrified Passenger Car Powertrain

Abstract

Abstract Complex and resource intensive multi-physical models which integrate the domains of electrotechnical, structural dynamics and acoustics are needed to predict the noise, vibration and harshness (NVH) behavior of an electric passenger car drive train. Because of the highly limited resources in product development process of cars, engineers request efficient models with a good balance between accuracy and model creation as well as calculation effort. Knowing about which components and effects influence the prognosis is essential to adjust this balance between accuracy and effort. Therefore, in this paper a method is presented to determine the influence of the model fidelity level on the prognosis quality of the perceived drive train-related airborne sound under the use of hearing tests to consider the human noise perception. The work is based on a high fidelity multi-physical NVH of a small, series-produced electric vehicle drivetrain which is validated in structure-and air-borne-noise against test bench measurements. The influence of model detail level on the prognosis quality is analyzed using the aforementioned method taking into account four exemplary variant calculations of the proposed model: (1) Power electronics harmonics vs. ideal sine electric machine current, (2) electric machine with vs. without production deviations, (3) modal damping parameters derived from component measurements vs. modal damping taken from literature, and (4) numeric vs. analytical sound calculation method.