Modelling of railway vibration in urban areas using a hybrid prediction method.

Increasing urban congestion and demand for mass transport is driving the development of new railways in urban areas. These railways generate ground-borne vibration, which can cause disturbance to nearby residents and interfere with sensitive equipment. A primary concern for engineers when designing a new railway is ensuring that vibration level limits are met. This requires accurate prediction methods that have been validated against experimental results. Current computational models for making railway-vibration predictions are subject to high levels of uncertainty and require detailed parameter inputs and simplifying modelling assumptions. These issues can be avoided by using a combination of field measurements and state-of-the-art numerical methods, known as a hybrid model. Little research into hybrid models has occurred to date, so this project will investigate the accuracy and applicability of using this modelling method to make predictions of railway induced vibration levels in buildings. The proposed research will examine how the vibration transfer between a railway track and a building can be divided into separate vibration terms. Each of these terms will be determined using either experimental or numerical methods, and then combined to form complete hybrid models for railway induced vibration in buildings. These models will be validated using experimental measurements, and the level of uncertainty present in hybrid model predictions will be investigated.