Structural Behavior Analysis In Construction Phase Based On The Monitoring Data

During the construction process,there are complex linear adjustment of structure deformation,stress redistribution and other issues.Accurately grasping the changes of structural behavior in the process is an important prerequisite to ensure structural safety in the construction stage.Traditional construction analysis is difficult to reflect the local response state of joints based on macro element model.Meanwhile,the number of sensors is usually limited by the cost and the installation location is limited by steel joints sizes,it is difficult to fully perceive the overall response of structure base on health monitoring system.What’s more,owing to the inevitable difference between the finite element model and the actual structure,the results of the simulation analysis deviate from the real response frequently.Therefore,it is necessary to establish a multi-scale finite element model in the construction stage,and conduct the model updating based on monitoring data.Then,structure behavior change will be analyzed by updated model in the construction stage.Structure behavior in the construction phase is to be analyzed by multi-scale model and a modified structural deformation control method is proposed in the paper.Aiming at the problem that the macro element model is difficult to reflect the local response of the structure at the joint,a multi-scale model during construction stage is considered to be built.The overall structure is modeled by macro finite element and the local key joints are modeled by micro element,so that the overall and partial responses can be both obtained from the model.Multiscale model is an important way to reflect the condition of structure.Moreover,aiming at the problem of "dead" node drift when the "life and death" element method was adopted in construction simulation,a modified shape control method is proposed to constrain the key points of the structure in the construction process.At last,the multi-scale modeling method and the modified shape control method were applied in the case of a three stories steel truss structure to verify the effectiveness of these methods.A two stages structure model updating method during the construction process is proposed.In view of the problem of response error between construction simulation and actual structure,a model updating method based on monitoring data is proposed.In each construction step,the overall response of the structure and the response of the local key joints are corrected chronologically based on the monitoring data,so that all responses of the structure are consistent with the actual structure.For the modification of the local joint model,a test about welded joint is to be carried out to explore the main factors affecting the stress field of the joint.Finally,a three stories steel truss structure was taken as an example to analyze the selection of the material parameters,the solution of objective function,the model updating and the whole state behavior acquisition during the construction process for the purpose of proving the feasibility of the method in the paper.Actual structure response in construction phase is to be analyzed base on monitoring data.For Zhuhai Opera house,in view of the big shell structure health monitoring system,24 welded box shaped steel joints with strain sensors are selected,and the multi-scale model is established respectively.The two stages model updating method is applied in the project to update the finite element by utilizing the monitoring data.Through comparing the structure response before and after model updating,it is demonstrated that the effective model can be obtained by the method.What’s more,the response acquisition in structure joints where no sensor was attach to and the response prediction in the follow step are conducted base on updated model.By comparing the model response with the monitoring data,it is showed that the method proposed in the paper can be used in the real steel structure behavior analysis in construction phase effectively.