Using a Simple Soil Spring Model and Support Vector Machine to Determine Bridge Scour Depth and Bridge Safety

Scouring around bridge piers is an important safety issue of bridge management since it can lead to bridge slant and collapse. Therefore, bridge scour depth is an important index to determine a bridge's safety condition. However, directly measuring the scour depth is difficult due to the water surrounds bridge piers. Several researchers have discovered the bridge natural frequency, a proxy measurement that can be used to determine the scour depth of the bridge pier. In this study, two finite-element models including the complicated finite-element model and the simple soil spring model are introduced to compute the natural frequency of a bridge and the result is verified with field experiments. In addition, a series of simulations is then implemented to analyze the factors impacting scouring depth. Seven factors related to bridge substructure and soil property are identified. A support vector machine (SVM) is further employed to understand how those factors impact the safety of the bridge. The results show that by hybridizing the simple soil spring model and SVM, the bridge's safety level can be efficiently established. (C) 2015 American Society of Civil Engineers.