Prediction of Service Lives of Bridge Expansion Joints

This paper presents a service-life prediction model of expansion joints. Significant factors influencing the service lives of expansion joints were identified by statistical methods. Artificial neural network was implemented to establish the service-life prediction model of expansion joints. Taken finger plate joints for illustration, eight statistically significant factors influencing the service lives of finger plate joints are identified among twenty one factors studied. Through these eight factors, the service lives of expansion joints can be predicted by the established model. The training and testing errors indicate that the established artificial neural network model can provide accurate predictions which are essential information for maintenance strategies.