Probabilistic empirical model for train-induced vibrations Available to Purchase

A novel model for predicting railway-induced vibrations nuisance and structure-borne noise in nearby dwellings exposed to railway-traffic is described in this paper. The model is based on partitioning the transmission path from the railway to a building into different steps through the track system, ground, foundation and building – each characterised by a probabilistic approach towards adding frequency response functions obtained by statistical analysis of measurements obtained at various locations passed by the most common train types. The model output consists of an estimate of the acceleration level at a particular building floor, an estimate of the associated error and an estimate of the structure-borne noise level. The main new feature of the proposed approach in relation to existing empirical formulations is the probabilistic framework undertaken to estimate model errors as a function of the quality of the measured data and of the mathematical formulation developed to predict railway-induced vibration and noise levels. The paper first describes the general model formulation considered. Details of the measurement campaigns along with the associated data analyses for model development are then described. Finally, the mathematical formulation for calculating the model output is presented, and a comparison between model-based predictions and measurements is carried out to assess the performance of the model.