Comparison of Aerodynamic Performance of Moving Train Model at Bridge–tunnel Section in Wind Tunnel with or Without Tunnel Portal
Dehui Ouyang,Weichao Yang,E. Deng,Youwu Wang,Xuhui He,Linbo Tang
DOI: https://doi.org/10.1016/j.tust.2023.105030
IF: 6.9
2023-01-01
Tunnelling and Underground Space Technology
Abstract:The moving train model test for wind tunnels is an advanced and effective research method for studying the aerodynamic effect of trains passing through tunnel portals in crosswind environments. However, the modelling of tunnel portal structures is cumbersome, and its influence on test results remains unclear. Here, the influence law of the presence and absence of a tunnel portal on the aerodynamic performance of a train was investigated via a wind tunnel test and an improved delayed eddy simulation turbulence model with "mosaic" mesh tech-nology. First, the difference between train locomotion on the windward and leeward side lines with and without a portal was compared from the perspectives of aerodynamic load amplitude, power spectral density (PSD), and standard deviation. Second, the influence of vehicle speed on the amplitude of the train's aerodynamic load was studied, and the recommended critical speed was determined. Finally, the applicability of the moving model test results under different scale ratios (Reynolds numbers) was explored. Key results show that the aerodynamic load amplitude, PSD peak and standard deviation of the train on the leeward side line are generally higher than those on the windward side line. As indicated by the lift force in the 'EX' process without the tunnel portal, the amplitude on the leeward side line is 1.73 time the corresponding values on the windward side line. At speeds of 6, 12 and 18 m/s, when the moving train model speed reaches 12 m/s, the influence of the tunnel portal on the aerodynamic load amplitude can be ignored. The results imply a serious distortion of the aerodynamic load amplitude for the 1:20 model scale, whereas it appears to be appropriate at the 1:16.8 and 1:10 scales. The model test with a Reynolds number of 2.67 x 105 offers an ideal reference value for real trains.