Study on prediction in far-field aerodynamic noise of long-marshalling high-speed train.

It is still difficult to conduct numerical calculation of the aerodynamic noise of full-scale, long-marshalling, high-speed trains. Based on the Lighthill acoustic analogy theory, the aerodynamic sound source of the high-speed train is equivalent to countless micro-vibrating sound sources. An acoustic radiation model of the dipole sound source of high-speed trains is established, and a method to predict the aerodynamic noise in the far field of long-marshalling high-speed trains is proposed. By this method, combined with numerical simulation technology, the flow field, noise source, and far-field noise characteristics of high-speed trains with different marshalling numbers are studied. The improved delayed detached eddy simulation method is used for flow field calculation, to obtain aerodynamic noise source information regarding the surface of high-speed trains. The numerical calculation method is verified by wind tunnel testing. The results show that the flow field and noise source characteristics of high-speed trains with different marshalling numbers are similar. The greater the length of the train body, the longer the trailing distance of the train wake, and the stronger of a surface noise source the tail car becomes. The spatial distribution characteristics of aerodynamic noise in the far field of high-speed trains do not change significantly with the length of the train body, but the magnitude of the sound pressure level will increase with the increase in length of the train body. The middle car body parts of high-speed trains with different marshalling numbers have similar noise distributions and sound pressure levels. Based on the noise calculation results of the 3-marshalling high-speed train, the far-field noise of the 5-marshalling and 8-marshalling train models is predicted and found to be in good agreement with the far-field noise of the actual train model. The differences in average sound pressure level are 1.01 dBA and 1.74 dBA, respectively. [ABSTRACT FROM AUTHOR]