Application Research On Large-span Spatial Structure Optimization Design

In order to develop efficient large-scale civil engineering structures, the structural-form optimization design is becoming increasingly important. Especially for complex large-span spatial structures(LSSS), the beautiful forms and reasonable load paths have become an intense research topic in the engineering field. Structural morphology research is focuses on the relationship between the geometrical forms and structural performance, which provides a theoretical basis for the development of new LSSS using an inverse hanging method, improved evolutionary theory method, empirical method, and other methods. Inverse hanging method is a form generation method based on the mechanical balance principle for zero bending moment space structures and has played an important role in structural-form optimization design of many LSSS. Evolutionary structural optimization(ESO) algorithm is a relatively new optimization algorithm proposed in recent years, which can obtain the optimal topology configuration by removing inefficient elements without affecting the integrity of the structure. However, this algorit hm is not mature enough to be applied to LSSS. In this paper, a modified ESO method is applied to a large-span spatial structure for a structural-form design combined with the morphological knowledge and optimization design methods. Accordingly, valuable c onclusions are obtained by comparing the results of the ESO method and the inverse hanging method.In this thesis, the main research contents are as follows:Firstly, it summarizes the domestic and overseas general situation of structural-form optimization design of many LSSS, and expounds the relationship between structural form and structural optimization design.Secondly, the basic theoretical knowledge is reviewed on the structural-form optimization design with the inverse hanging method and the ESO method. After the comparison of two methods, the reasonability for structural-form optimization design of many LSSS is proved on the theoretical perspective.Thirdly, the analysis process based on two methods is studied about a large-span spatial structure for a structural-form design in the conception design phase, and some key problems in the optimization process are explained. Then, the feasibility and efficiency are proved for the ESO method applied to a structural-form design.The fourth, the static analysis for selective proposal is carried out by using the finite element software SAP2000. By comparison of the structural deformation, internal force and material utilization ratio, the essentiality and necessity of structural-form optimization design are pointed out in the conceptual design phase.Finally, on the basis of the static analysis, the dynamic characteristic analysis for modal analysis and seismic response spectrum analysis are also carried out. In the modal analysis, the dynamic characteristic complexity of LSSS is obtained by analyzing the vibration mode and frequency of the structure before 50 order vibration mode. In the seismic response spectrum analysis, by comparison of the displacement and internal force under one-dimensional seismic action and the most unfavorable loading condition, it is proved that the static load is a control effect in structural-form optimization design of a LSSS. Furthermore, the reasonability and validity for structural systems optimized are presented.