Optimum Design Of Steel Construction Of Large-span Spatial Lattice Shell Based On Discrete Variables

Reducing the structure's weight and designing the more lighting and steadying structure is an important research aim in the field of construction. Evidently, the structural optimum design is adopted and used by designers.The engineering structure design may be classed into the traditional design and the optimum design.In this thesis the mechanical analysis of a spatial reticulated steel truss is studied. The aim is to modify the original design of the steel truss by an optimum method so that the new scheme of the truss can save engineering cost. By optimization analysis, the construction's weight is reduced, while construction's security is satisfied.The main work of the present paper are summarized as follows:1,the large-scale software ANSYS was used to analyze the strength and stability of the steel truss. From the computed result, it is found that the stresses of some members in the primary design didn't satisfy the allowable stress.2,For the large-span spatial lattice shell structure, the areas of all steel member sections, which are all standard tube shapes, were taken as the design variables. The total weight of the structure was considered as the objective function and was optimized in accordance with the stress ratio method in the full stress method. The constrain conditions such as strength, rigid, stability, cross section dimensions of all members were also considered to satisfy the national design codes. By using the APDL language, the optimization design programs was written and combined with ANSYS. Therefore, the reliability of program is improved and the applicability of ANSYS is enlarged. The optimization design program is combined with ANSYS to obtain the result of the optimum design of large-span spatial lattice shell.3,Furthermore, applying the computed program, the spatial lattice structure was studied and analyzed when some of members had been removed. The weight of construction had been obviously reduced while the optimum structure keeps the same functions as the original requirement.