Visualisation of Roof Bar Noise Sources through the Use of Acoustic Beamforming and Computational Aeroacoustics

The reduction in wind noise is increasingly important to vehicle designers as overall vehicle refinement increases. Customers often fit accessories such as roof bars to vehicles, with the aerodynamic interaction of these components generating aeroacoustic noise sources. These are often tonal in nature and of particular annoyance to occupants. Sensors for automated driving fitted to future vehicles may also have a similar detrimental effect on vehicle refinement. Therefore, careful design of such components is important to minimise dissatisfaction. This paper presents the combined application of acoustic beamforming in a full-scale aeroacoustic wind tunnel and the use of a Lattice Boltzmann Method CFD code to characterise the aeroacoustic performance of a roof bar design when fitted to a production vehicle. The use of large microphone arrays within a wind tunnel allows the visualisation of noise sources generated by the flow over a vehicle, with correlation techniques utilised to identify those sources most relevant to occupants. The use of a validated CFD code allows further visualisation of the noise sources virtually, with additional benefits when used in advance of prototype parts being available. A series-production vehicle and roof bar paring were assessed, before deliberately inducing several aeroacoustic error states that may occur during roof bar development. These were analysed using visualisation techniques in both wind tunnel and CFD simulation. A range of aeroacoustic sources were identified, including tonal sources due to vortex shedding and acoustic resonance, as well as broadband sources from the interaction of the roof bars and mountings with flow structures formed over the vehicle. The paper concludes by highlighting key considerations when designing low-noise roof bars.