Identification of moving vehicle loads from bridge bending moments

Abstract

The axle loads identification of a moving vehicle from bridge bending moments is studied. The objectives are to investigate the effectiveness of two moving load identification methods which are (a) Method I: Constant magnitude of moving axle loads assumption using least-square optimization with influence line and (b) Method II: Time-varying magnitude of moving axle loads assumption using dynamic programming and regularization with updated static component technique (USC). In this dissertation, both small-scale and full-scale tests of moving vehicle on the bridge are performed besides the numerical study on computer simulation. The effectiveness of two identification methods is intensively evaluated under various passing conditions of vehicles. Moreover, the application of the B-WIM to monitor and record the actual truck information in the road network is conducted. Bridge live load models from design code such as the HS20-44 and design Thai truck are compared with this measured truck database from B-WIM using statistical and probabilistic approach. From the numerical study results, vehicle speed, surface roughness level and measurement error seem to have stronger effects on the weight estimation accuracy than other parameters. About 81 and 47 conditions of passing vehicles are conducted for small-scale and full-scale investigations, respectively. The obtained results indicate that the two methods can estimate the axle loads of vehicle under various speeds and weights regardless of its traveling paths. Comparing between the two methods, it is found that the identification accuracy obtained from the regularization with USC technique is much better. Moreover, the method also provides the identified dynamic axle loads which are very useful information for dynamic load assessment. Based on the full-scale tested results, the gross weight estimation error within ± 6 % can be achieved which using USC technique. For B-WIM application, approximately 5,049 heavy trucks are monitored and recorded. The results show that the mean value of truck weight data does not exceed the legal limit. However, it is found that there are certain amounts (more than 16%) of overloaded trucks traveling in the transportation network. By comparing the design Thai truck and HS20-44 based on the short to medium span bridges (5-30 m), it is observed that the design Thai truck load can conservatively be used with a live load factor of 2.17 for 75 years life time, while HS20-44 is not conservative. It is also found that the design Thai truck load seems to be more suitable than HS20-44 and its weight could be theoretically reduced by 10% for the existing live load factor based on the obtained truck records.