Topology Optimization Method For Bridge Design Considering Boundary Design Optimization

Topology optimization is regarded as the most general type of structural optimization technique.In recent years,benefiting from the steady progress of computer science,topology optimization as a design technique has found its way into a wide range of real-world applications,especially in the fields of mechanical and aerospace engineering.However,its application in civil engineering is still relatively limited.In this thesis,the topology optimization of bridges considering moving loads and boundary design optimization is researched.The dynamic characteristics of moving loads and boundary design(such as bridge aspect ratio,deck location,and the distribution of supports)are all worthy factors to consider when one designs a bridge using topology optimization.In order to address the need for topology optimization of moving loads,a topology optimization methodology based on the equivalent static load method is introduced to find the optimal layout of a structure under moving loads.In an effort to optimize the boundary design(bridge aspect ratio,deck location,and the distribution of supports),optimization frameworks based on MMA(Method of Moving Asymptotes)and GA(Genetic Algorithm)are proposed successively.The conclusions are as follows:(1)A topology optimization methodology based on the equivalent static load method is proposed to realize topology optimization of structures under moving loads.A novel load processing technique,the symmetrical loading method,is proposed to neutralize the undesired directional influence of moving loads.The equivalent static load method is applied to convert dynamic problems into static problems.And the famous SIMP method is used for topology optimization.In order to demonstrate the effectiveness of the proposed method and investigate the relation between the optimal layout of a structure and its boundary design,several numerical examples with different boundary designs are given.(2)MMA(Method of Moving Asymptotes)is employed as the optimizer to formulate a topology optimization framework that considers the boundary design optimization.The proposed framework treats the supports location as design variables in addition to the density variables and optimizes them with MMA.The optimized support variables are then relocated in the design domain by a coordination interoperation scheme to achieve simultaneous boundary design optimization.Various numerical examples have shown that the proposed algorithm produces not only the optimal material distribution but also the optimized boundary design.(3)A boundary design optimization framework based on genetic algorism(GA)is proposed.The GA-based framework accepts the aspect ratio of the bridge,the location of the bridge deck,and the support distribution as design variables.The advantage of the GA-based approach is that it’s a global search algorithm,and it does not require any derivative information.Numerical examples concern sign-span bridge designs are provided,the result shows the GA-based framework indeed produces highly efficient results.