Research On Improved Bi-directional Evolutionary Structural Topology Optimization And Dynamic Optimization

Topology optimization is to seek the optimal materials distribution in design domain under given constraints and loads.The topology of the structure determines the main performance of the structure,so the topology optimization technology is important in conceptual design phase.From an economic perspective,it can reduce the cost and improve developing efficiency which is regarded as the highest return in optimization,and it has become a hot topic in structural optimization.This paper analyzes the research status of topology optimization,and study on BESO method which is one of the most important methods of topology optimization to improve its rationality.The improved BESO method is extended to the field of multi-objective optimization and dynamic optimization to make it more practical.Firstly,the basic principle of SIMP and BESO are introduced in the paper.The application of SIMP and BESO in the Michell structure and the MBB beam are discussed.The advantages and disadvantages of each method are acquired.Secondly,the reason of failure of BESO method in some structures is analyzed.Aiming at the errors in sensitivity calculation and poor searching ability of traditional BESO method,an improved BESO method is proposed,which is combination of BESO and SIMP,thus the accuracy of sensitivity can be improved by reducing the evolutionary step length of element density,the evolutionary step length can be determined by the influence of element density on the structure performance.Then the node sensitivity redistribution method is used to filter the sensitivity of element to control the numerical instabilities and improve the searching ability.The rationality and validity of the improved method is verified by some numerical examples.Thirdly,the improved BESO method is extended to the field of multi-objective optimization.The weighted compromise programming method is used to deal with the sensitivity of element,thus the influence of the load failure phenomenon can be reduced;to make compliance sensitivity and frequency sensitivity normalized and weighted to combine,the multi-objective sensitivity of element is obtained,then the optimization process is given.Through two examples,the results prove the applicability of the proposed method.Finally,based on the research above,the improved BESO method is utilized as the inner cycle iterative approch of Equivalent Static Loads method to solve the dynamic optimization problems.Several numerical examples are solved using the proposed method,the results contrast to the original Equivalent Static Loads method verify the effectiveness of the proposed method and prove that the proposed method can get better results with higher efficiency,thus improving the application in engineering.