Scaling Laws for the Aeroacoustics of High Speed Trains

Abstract

The present study focuses on the scaling of the main aeroacoustic sound sources of a high speed train. The investigations are carried out by measuring in a wind tunnel by means of microphone array technique. The experiments using a 1:25 Inter City Express 3 model were conducted in two different wind tunnels: The Aeroacoustic Wind tunnel (AWB) of the German Aerospace Center (DLR) in Brunswick, which provides nearly perfect acoustically conditions, and in the Cryogenic wind tunnel (DNW-KKK) of the DNW (German - Dutch wind tunnels) in Cologne, which allows measurements at higher Reynolds numbers. Two different sources of sound with different characteristics are identified from the measurements in the AWB at Reynolds numbers up to Re = 0.46 mio: The aeroacoustic noise from the bogie section is dominant for frequencies f < 6 kHz and can be characterised by cavity mode excitation. The shape of the spectrum is not velocity dependent within the investigated velocity range. The pantograph is the dominant source of sound above $f = 6$ kHz. It is more tonal noise and the frequency peaks show a strong velocity dependence. These two sources of sound demonstrate that it is not possible to formulate one scaling law for all sources of sound of a train. It is essential to treat every single source separately and to t-ake their characteristics into account. In order to investigate the noise generated at the bogies at higher Reynolds numbers up to 3.70 mio a second measurement campaign has been conducted in the cryogenic wind tunnel. By cooling down from T = 300 K to T = 100 K one changes the physical properties of the fluid and the Reynolds number increases by a factor of ~5. Besides, this wind tunnel admits to vary the Mach and Reynolds numbers independently. Drawback of this facility is that it is not optimised for aeroacoustic experiments and reflexions as well as the high background noise level can disturb the results. The Mach number is kept constant and a Reynolds number variation is realised by varying of the temperature. The first bogie shows only a weak Reynolds number dependence. The deviations are within the measurement accuracy. To go more into detail further investigations with an improved setup are planed.