| description abstract | Sliding bearings are a vulnerable component of bridges. Bearing wear will affect the free expansion of the bridge structure and produce greater temperature stress, resulting in damage to the main girder and other components of the bridge. In the process of bridge operation, the timely detection of bearing wear is very important for ensuring structural safety. Therefore, this paper proposes a wear detection method for bridge sliding bearings based on displacement amplitude by eliminating the effects of daily temperature variations. First, the bearing mechanical behavior under temperature effects and the correlation between temperature and bearing friction force are analyzed. The bearing displacement hysteresis model under temperature effects is established, and the variation in thermally induced bearing displacement after wear is obtained. Then, the correlation between temperature and thermally induced bearing displacement is analyzed and, based on the particle swarm optimization (PSO) algorithm and long short-term memory (LSTM) neural network, a multivariate temperature–displacement correlation model is established to achieve the accurate prediction of thermally induced bearing displacement and the elimination of temperature effects. According to the variation in the thermally induced displacement of the bearing after wear, the indicator of thermally induced displacement amplitude errors (TDAE) is proposed, and the cumulative sum (CUSUM) control chart is used to detect the bearing anomaly. Finally, a long-span bridge is analyzed as an example. The results show that the proposed TDAE detection indicator can effectively reflect the sliding friction force of the bridge bearing, and the proposed detection method can accurately detect the sliding bearing wear, which can provide effective information for the bridge caretakers to monitor the occurrence of bearing wear and replace the bearing slide plate in a timely manner. | |
