Use of a 'Hybrid' Empirical/Finite Element Approach for Predicting Groundborne Vibration from Rail Systems

摘要

The finite element method (FEM) is widely used in various engineering fields to model complex structures. Applying FEM to predict groundborne vibration from rail systems has been tested, but always has been limited by difficulty in characterizing soil properties. The empirical method developed in the 1980s (often referred to as the "FTA method") remains the standard approach used in North America to perform detailed predictions of groundborne vibration and to determine the need for vibration mitigation measures. The FTA method is based on in-situ vibration propagation tests and is almost always found to provide more accurate predictions than even the most complex computer models. This paper describes a "hybrid" approach that uses data from vibration propagation testing to refine input parameters for an FEM model. The basic steps are (1) perform a standard vibration propagation test, (2) develop an FEM model of the test configuration using estimated soil properties, (3) "tune" the properties of the FEM model to optimize the correlation with the test results, and (4) modify the FEM model to be a tunnel in place of a borehole. The accuracy of this hybrid approach and other approaches for using FEM models to improve predictions of groundborne vibration will be discussed.