Search Results

For most car manufacturers, aerodynamic noise is becoming the dominant high frequency noise source (≻500 Hz) at highway speeds. Design optimization and early detection of issues related to aeroacoustics remain an experimental art implying high cost prototypes, expensive wind tunnel sessions, and potentially late design changes. To reduce the associated costs as well as development times, there is strong motivation for the development of a reliable numerical prediction capability. This paper presents a computational approach that can be used to predict the vehicle interior noise from the greenhouse wind noise sources, during the early stages of the vehicle developmental process so that design changes can be made to improve the wind noise performance of the vehicle.

Vehicle wind noise is becoming increasingly important to customer satisfaction. Early wind noise assessment is critical to get things right during the early design phase. In this paper, SEA modeling technique is used to predict vehicle interior noise caused by the exterior turbulence. Measured surface turbulence pressures over vehicle greenhouse panels are applied as wind noise load. SEA representation of wind noise load case is investigated. It has been found that current SEA wind noise load case over-estimates at frequencies below window glass coincident frequency. A new concept of noise source pole index is introduced and a new wind noise load coupling has been developed. Comparison with vehicle wind tunnel measurements shows that the proposed load case significantly improved prediction accuracy.

Simulation of local flow structures in the A-pillar/side glass region of bluff SUV geometries, typical of Land Rover vehicles, presents a considerable challenge. Features such as relatively tight A-pillar radii and upright windscreens produce flows that are difficult to simulate. However, the usefulness of aerodynamics simulations in the early assessment of wind noise depends particularly on the local accuracy obtained in this region. This paper extends work previously published by the author(1) with additional data and analysis. An extended review of the relevant published literature is also provided. Then the degree to which a commercial Lattice-Boltzman solver (Exa PowerFLOW™) is currently able to capture both the local flow structure and surface pressure distribution (both time averaged and unsteady) is evaluated. Influential factors in the simulation are shown to be spatial resolution, turbulence and boundary layer modelling.