Finite element analysis on aerodynamic noise of the high-speed train head based on computational fluid dynamics and acoustic analogy methods

摘要

As train speed increases, the aerodynamic noise gradually exceeds the wheel-rail noise and becomes the main noise source of high-speed trains. The aerodynamic noise affects the interior environment of a train in two ways: turbulent wall pressure fluctuation (TWPF) and acoustic wall pressure fluctuation (AWPF). In this paper, a hybrid aerodynamic model for noise analysis of high-speed trains based on the computational fluid dynamics (CFD) method and the Lighthill's acoustic analogy (LAA) algorithm, so-called CFD-LAA method, is established and solved by using the finite-element acoustic simulator of Actran software. The noise contributions of both the TWPF and the AWPF are examined and discussed. The method presented in this paper may be a valuable reference for revealing the generation mechanism of aerodynamic noises and evaluating their effects on a train interior noise. In recent years, with the development of the high-speed rail industry, train speed has been increased rapidly. Accordingly, the air pressure fluctuation act on the surface of a train body significantly increases, which makes the aerodynamic noise becomes one of the important elements in high-speed train noise. When the running speed of a train reaches 300 km/h, the generated aerodynamic noise exceeds the wheel-rail noise and becomes the main noise component of the high-speed train. Thus, to develop a high-speed train, it is both necessary and meaningful to investigate and control the aerodynamic noise. It has been found that, for the streamlined train in common use, the maximum sound power level (SPL) occurs on the head, followed by the tail, and the middle of the train has the minimum value of SPL. The main areas of sound source appear in the local places where air separates much easily and turbulent motion is more intense. Therefore, optimizing the shapes of train head and body and reducing the airflow disturbance caused by the concave and/or convex objects are the effective methods to reduce the aerodynamic noise. In the initial studies on the generation mechanism of aerodynamic noise, researchers thought that the interior radiated noise of a train was generated by the wall pressure fluctuations, due to the air turbulences created by the wind on the surfaces of the train body and windows. Later, some research found that the turbulence itself can be regarded as a kind of sound source. The turbulences outside of the train body have been classified into two sound sources, i.e., turbulent wall pressure fluctuation (TWPF) and acoustic wall pressure fluctuation (AWPF). Their combination is defined as the real aerodynamic noise radiated to the train cabin. Based on the above findings, to analyze the generating mechanism of aerodynamic noise, a hybrid model, so-called CFD-CAA method, for noise analysis of high-speed train is proposed in this paper. The computational fluid dynamics (CFD) and the Lighthill acoustic analogy algorithm (LAA) are used to simulate the aerodynamic noises generated from the TWPF and AWPF noise sources outside of the train head respectively. And the A-weighted sound pressure levels (SPLs) at the position of train driver's ear are calculated to estimate the contribution rate of the two noise sources. The works done in this paper is instructive to engineers involved in the aerodynamic analysis and shape design of high-speed trains.