Prediction and sensitivity analysis of the pressure wave peak value induced by the high-speed train in the long tunnel under a high geothermal environment

Pressure changes outside a train can be transmitted into a carriage when the train is running in the tunnel, seriously affecting passenger comfort. In this paper, the polynomial response surface (PRS) is used to predict the value of the compression wave and obtain the predictive model equations. For the traveling pressure waves, many influencing factors contribute differently to the final pressure wave form and magnitude. In geothermal tunnels, the speed of a high-speed train (V), air temperature of the inner tunnel (T), and atmospheric pressure (P) are the three main influencing parameters. Using a PRS prediction model, the sensitivity of each influencing parameter is analyzed by the Sobol sensitivity method. The result shows that V has the greatest effect on the pressure peak value, and T has the least effect. The initial compressional waves and expansion waves are most and least sensitive to T, respectively. The coupling effect among parameters P, T, and V is relatively small. Using the sensitivity results of the parameters, targeted and reasonable parameter adjustment can effectively relieve the pressure inside and outside the train and improve passenger comfort in geothermal tunnels. These results provide important technical support for the mitigation of pressure waves.