Microphone Arrays an a Wind Tunnel Environment with a Hard Reflective Floor

The usage of a microphone array is in most cases determined for free field conditions, which does not hold for measurements in a wind tunnel. Whereas the walls can be assumed as sound-absorbing, this is not correct for the floor. Due to high aerodynamic requirements the floor in a wind tunnel is often a very flat surface which reflects the sound field. Without special computational consideration of the floor, this reflection could lead to a loss of dynamic or in case of using deconvolution algorithms to wrong source positions. To overcome this degradation, there is the theoretical approach of mirror-ground beamforming using virtual microphones. In the present paper the influence of the reflections on the quality of acoustic maps is shown. Furthermore the performance of arrays optimized for free field conditions is compared with arrays optimized for mirror-ground conditions (virtual microphones). By simulations it can be shown that using virtual microphones either leads to a gain of information or enables the reduction of the real microphones' number. In contrast, wind tunnel measurements suggest a drastic loss of coherence between the signals of source and reflection due to turbulent effects. Therefore, the advantages of mirror-ground beamforming are relativized.