Efficient Evaluation of Sound Radiation of an Electric Motor Using Model Order Reduction
* Presenting author
Electric motors can suffer from severe noise radiation when the internal electromagnetic forces excite a natural vibration mode of the stator pack. These critical operation points can be identified from a Campbell diagram. The generation of the diagram based on numerical analyses is computationally very costly, in particular when the motor is to be analyzed in the mid-frequency range.The complex vibration behavior of the motor housing prohibits a straight-forward application of common model reduction processes. Therefore a new multiple-input-multiple-output model reduction process is developed using the moment matching method based on Krylov subspaces. The in-vacou surface velocity modes of the motor housing yield as input load patterns. A structure-preserving second-order Arnoldi algorithm is used for the calculation of the Krylov subspaces. The proposed process enables the direct evaluation of the radiated sound power for all considered frequency points within a frequency band. The reduction process, incorporating an adaptive selection of structural modes and an optimized parallelization of subspace calculations, allows for an efficient evaluation of the radiated sound power of the electric motor.