Building coupling loss measurement and prediction due to train-induced vertical vibrations

Train-induced vibration prediction of the over-track buildings requires acceptable both efficiency and accuracy in the preliminary construction stage. To ensure accuracy, soil-structure interaction for vibration prediction is nonnegligible, leading to the coupling loss between soil and the building foundation. In this paper, train-induced vibration measurements were conducted in Wuhan, China. A total of 75 train passby events were recorded on the ground soil and building columns at the free field, a 1-story building, and a 4-story building. The transfer function of ground soil at the free field and building columns confirmed that the appearance of the building would restrain the vibration on the ground. Besides, a hybrid approach that combined the Bayesian Neural Network and the impedance model was proposed to predict vibrations inside a building by considering the coupling loss of soil and building structure. The vibration response of the building can be predicted before construction once the ground vibration at the free field is measured.