A novel model for vehicle/turnout nonlinear random vibration analysis based on full probability irregularity spectrum

The irregularity in the turnout is the primary source of excitation in vehicle/turnout interaction. By considering the amplitude, frequency, and probability characteristics of the irregularity, a full probability simulation method for modeling the turnout irregularity random field is proposed using Latin hypercube sampling and inverse Fourier transform. The probability density evolution method is then introduced to comprehensively reveal the random vibration characteristics of the vehicle/turnout system. The results demonstrate that the full probability irregularity spectrum can achieve a cumulative probability resolution of up to 1/100 with the amount of data being less than 1/10 of the measured data. The computational efficiency of PDEM is enhanced by 1 to 2 orders of magnitude compared to the Monte Carlo simulation while maintaining the same level of accuracy. The position of the heart rail is identified as having the most significant influence on the random contact behavior between the wheel and rail. The vertical acceleration of the vehicle is more sensitive to the heart rail position, whereas the lateral response is more influenced by the switch rail position.