Prediction of aerodynamic pass-by noise for a high-speed train pantograph by combining the wind tunnel model and the Farassat_1A formula

To effectively control noise emission from a high-speed train in early stage of design, it is necessary to predict the pass-by noise accurately. In the past, wind tunnel model was widely used to predict noise emission at a fixed far-field point, with the train remaining stationary and the wind blowing head-on at high speed. By doing so however, no Doppler effect is predicted. However, if the movement of the source considered, a sliding mesh or overset mesh will be used, which will generate the so-called interfaces between the sliding region and other stationary regions for flow field data exchange, and the existence of such interfaces may have a negative effect on calculation accuracy. Therefore, in this paper, a method is proposed combining the wind tunnel model and the Farassat_1A formula without using sliding mesh or overset mesh. In order to verify the effectiveness of the written program, a moving monopole and dipole source are given first. Based on the verified model, a typical high-speed train pantograph is studied. The subdomain method is used to obtain the sound source data, and then pass-by noise at different far-field observer points is calculated. The contribution of different pantograph components to the standard far-field noise is studied. Results show that due to the Doppler effect, the peak at the vortex shedding frequency of the pan-head is distributed over a frequency range.