FLOW-INDUCED NOISE PREDICTIONS OF AN AUTOMOTIVE ALTERNATOR USING A LATTICE BOLTZMANN METHOD

The increasing demand for comfort and quietness from the automotive industry transforms the acoustics performances of subsystems as a critical input for the selection of a specific design. Among this market, rotating systems noise takes a growing importance and automotive alternators are strongly impacted by this aspect. Alternators contain many different types of rotating parts such as cooling fans and claw poles and their corresponding low-induced noise contributions and interaction mechanisms driving the noise generation have to be assessed as early as possible in the product development process. Experimental methods have been historically used to identify and reduce the most obvious phenomena at the origin of the broadband and tonal contents of the noise. Considering the complexity of this device, it appears practically more and more difficult to understand the involved mechanisms and to identify and treat the remaining aeroacoustics sources. The use of digital solutions to simulate the corresponding flow-induced noise contributions and to provide an insight on the noise generation mechanisms represents an alternative to this experimental approach. Furthermore, numerical approach allows a broader design space exploration, where experimental testing can sometime be limited by other constraints, such as mechanical, thermal and electromagnetic aspects. Another advantage of using CAE method is to reduce the product development cycle and the number of expensive prototypes. The highly detailed geometry features and the constrained environment of alternators however represented a real challenge for computational aeroacoustics solutions. In this paper, an unsteady and compressible computational approach based on the Lattice Boltzmann Method (LBM) is used to simultaneously predict the 3-D turbulent flow and the corresponding acoustic field of an automotive alternator. The complete rotor-stator model including all geometrical details and the truly rotating geometry is simulated. Numerical and experimental far field sound pressure levels and acoustic power comparisons are presented. Additional transient and spectral flow analysis are performed to diagnose flow-induced noise problems and to provide a better understanding of the aeroacoustics sources.