Moving Load Identification Of Multi-girder Bridge Based On Finite Element Model Updating

The investigation of the traffic load are the basis of which the following studies such as the designed load, the time-variant bridge reliability analysis, the life expectation of the existing bridge, the ultimate capacity calculation and the random traffic flowing simulation are based. The identification of transverse location and vehicle weight is a vital element in the investigation of traffic loads. Usually, previous vehicle weights were collected by Weigh-in-Motion (WIM). However, this method has been limitated seriously as the lacking of complexity in installation and higher costs in maintenance. In this case, indirect identification has been rasied, which is used to identify the moving vehicle load by the dynamic response of bridge. But these methods, which have limitations in different degrees, are not suitable forä»–the traffic load identification with large volume of data. Therefore, a simple, rapid and effective approach of moving loads identification to determine the key parameters of large-scale traffic loads is very important.First of all, moving loads identification on the existing theories and methods were reviewed, the approach called Moving Load of Multi-girder Bridge based on Finite Element Model Updating is proposed on this basis. In this thesis, the process and key technologies of this approach is described in detail, then selecting a simply-supported voided slab bridge and a continuous T-typed bridge as an engineering example, moving loads are identified, and the accuracy and reliability of identification is verified, finally by the use of such identification method, classified by the vehicle type, the transverse location and the vehicle weight parameter corresponded with each running single vehicle in this investigation is obtained, on the basis of which, the statistical analysis is conducted. This article focused mostly on:1. Based on the calibrated FEM, the low pass filter and the vehicle-bridge interaction analysis program specialized for the multi-girder bridge, the moving loads identification method is applied for the multi-gider bridge is presented. Then, the process of moving loads identification is explained, the calibrated FEM and the low pass filter are elaborated, and the analytical methodology governing the vehicle-bridge interaction analysis for the multi-girder bridge is introduced, the related analytical modules is programmed.2. Taken a simply-supported voided slab bridge as an engineering example, firstly, a Butterworth low pass filter is designed for filtering dynamic displacement signals of 5650 vehicles moving across the bridge separately collected in the field for 24 hours, extreme values of static response are extracted, classification and statistics are performed in terms of vehicle type rigorously. Secondly, the reference finite element model, which coule reflect real state of bridge, is developed by updating finite element model based on static test for observed bridge. And then, the updated finite element model is taken as an input of self-developed multi-girder vehicle-bridge coupling vibration program, and curve of transverse location identification and curved surface of vehicle weight identification are calculated and plotted with the unit of vehicle types and driveways. According to extreme values of static response, moving loads are are identified with dividing vehicle type and driveway, and the accuracy of identified results is verified. Finally parameters of 5650 moving vehicles across the bridge collected in 24 hours are identified in bulk with the vehicle type and driveway classification, and results is analysed.3. Taken a continuous T-typed bridge with six spans as an engineering example, firstly, dynamic responses of 324 vehicles collected in two periods are filtered, extreme values of static response are extracted. Secondly, the updated finite element model is taken as an input of self-developed multi-girder vehicle-bridge coupling vibration program, curve of transverse location identification and curved surface of vehicle weight identification are calculated and plotted. And then, curves of transverse location identification for different vehicle type are contrasted, in order to verify feasibility of classification. According to extreme values of static response, moving loads are are identified, and the accuracy of identified results is verified. Finally parameters of all moving vehicles collected in two periods are identified in bulk with the vehicle type and driveway classification, then results is analysed. The applicability of identification is verifiedThrough the above process.