Random Distribution of Subsoil Damping Ratio and Interval Estimation of Ground Vibration

The damping ratio of site soil is one of the important factors in affecting the propagation and attenuation of ground vibration. Herein, a method is proposed to obtain the random distribution of subsoil damping ratio by field experiment and Bayesian theory. An on-site experiment was carried out and the experimental attenuation curve of the surface wave has been obtained by combining the half power bandwidth method in frequency-wavenumber domain. A priori probability distribution model and the likelihood function of soil damping ratio with depth is established based on the experimental curves. Then, the posterior probability distribution model is further obtained by means of the Monte Carlo Markov Chain-Metropolis–Hastings algorithm, which can be input to Thin-layered soil model to accomplish the interval estimation of ground vibration. Compared with the traditional deterministic prediction, the proposed probability prediction method is more reasonable and has more application value in predicting the environmental vibration induced by rail traffic.