Stochastic analysis of railway embankment with uncertain soil parameters using polynomial chaos expansion

This paper focuses on the stochastic response of railway embankments considering the uncertainties in soil cohesion and friction angle. The non-sampling stochastic method concerning generalised polynomial chaos (gPC) expansion was employed for the dynamic numerical simulation. The uncertain parameters, including soil cohesion and friction angle, were defined by the truncated gPC expansions. Furthermore, the system's response, namely, the displacement and acceleration of different embankment sections, was presented by the gPC expansion with unknown deterministic coefficients. The stochastic Galerkin projection was used to calculate a set of deterministic equations. The unknown gPC coefficients of the system's response were determined by a non-intrusive solution as a set of collocation points. In addition, the results of these analyses were compared with classical Monte Carlo simulations. It is essential to note that although only a few collocation points have been used, the results are in good agreement with the MC sampling method. One of the main objectives of this study is to demonstrate the accuracy of the results and the time efficiency of the proposed non-sampling method in quantifying the uncertainty of stochastic systems compared to the sampling procedure (e.g. Monte Carlo simulation).

Automated summary

This paper focuses on the stochastic response of railway embankments, considering the uncertainties in soil cohesion and friction angle. A non-sampling stochastic method, involving generalised polynomial chaos (gPC) expansion, is used for dynamic numerical simulation. The uncertain parameters, soil cohesion and friction angle, are defined using truncated gPC expansions. The accuracy and time efficiency of the proposed non-sampling method in quantifying uncertainty are demonstrated through comparisons with classical Monte Carlo simulations.