Predictions of Aerodynamically Induced Wind Noise Around Ground Vehicles
Over the last decade wind noise has with few exceptions consistently generated a constant or even growing level of customer complaints to automotive companies world-wide. This can partly be explained by a relatively greater focus on engine, power-train and tire noise combined with a growing need to reduce weight in future cars.
One class of wind noise problems commonly referred to as air-rush noise relates to turbulent pressure fluctuations caused by separated or vortical flows, which is addressed in this thesis.
Sound generation and to some extent sound propagation are here evaluated from
incompressible flow fields, since the unsteady flow hydrodynamics at low Mach numbers is often the dominant sources of sound. A temporal form of Curle's equation is used in these evaluations.
It is shown that the noise level measured inside the compartment of a production vehicle has a dipole character. Accordingly,
sound generation is evaluated from the dipole terms in the acoustic
analogy and, for one specific case, the radiated sound is evaluated by the surface integral of the same
Three different cases are investigated, ranging from the laminar flow past an open two-dimensional cavity to the turbulent flow past a representative ground vehicle.
It is shown that the dipole terms in Curle's equation can with sufficient accuracy be predicted for the regions of interest even for the full vehicle case.
A potential problem is highlighted, which is that, even if the acoustic source magnitudes are correctly predicted in an incompressible field, the sound directivity may be erroneous due to distortions in the pressure field and small differences in the flow field for flows with few but dominant structures.
air rush noise
low Mach number