matlab的plsm脚本,基于参数化水平集的多孔结构拓扑优化方法研究

【Abstract】 As the leapfrog development of the new generation equipment,multi-functional structures with both load-bearing abilities and functional characteristics are increasingly crucial.The most distinguishing features of these structures are multi-scale,multi-objective and multi-physics,etc.One of the typical representatives is the periodic cellular structure.Periodic cellular structures have very complicated topology-functionality coupling rules.Traditional design tools are no longer suitable for the design of cellular structures.Meanwhile,due to the small scale of the cellular structure,the design complexity of the cellular structure is rising sharply.Recently,as the development of structural optimization technology,numerical computation method,and computer technology,structural topology optimization technology is gradually proved to be an important design method for the periodic cellular structures.As for the many challenges in designing periodic cellular structures,this dissertation focuses on the research of topology optimization method for periodic cellular structures based on parametric level set method(PLSM).Firstly,a stiffness topology optimization method is presented for periodic cellular structures based on PLSM.The geometric description method based on local level set function(LLSF)is proposed to describe the geometry of cellular structures.A number of LLSFs are defined within the global design domain for the geometric description of the cellular structures.To reduce the computational costs,the macrostructure is physically partitioned into a number of scale-related substructures with substructuring.Solving a fully linear system is divided into solving a condensed linear system and the recovery of substructural displacement fields.A topology optimization mathematical model of stiffness maximization for periodic cellular structures is established.Due to the consistent geometric and physic model,only a representative substructure is optimized for the global periodic structures during the optimization process.The finite element analysis(FEA)efficiency,as well as the topology optimization efficiency of cellular structures,are therefore improved significantly in the case of a very large number of unit cells or mesh density.Secondly,the eigenvalue topology optimization method for periodic cellular structures based on PLSM is researched.We study the substructure method with dynamic condensation which is slightly different from the static condensation.The macrostructure is partitioned into a number of substructures as did previously.The number degree of freedoms(DOFs)of the substructure is dynamically condensed into a smaller one.A number of LLSFs are also applied to implicitly describe the substructures,avoiding a massive data global level set function(LSF)definition for the whole periodic cellular structure.The mathematical model of eigenvalue topology optimization for periodic cellular structures is established.The objective is the maximization of the weighted sum of specific eigenvalues.The constraint is the volume fraction of the substructure.The scale effects of the eigenvalue problem of cellular structures are also investigated.This method is valid for the eigenvalue topology optimization of periodic cellular structures with improved efficiency.Thirdly,we propose a topology optimization method for hierarchical cellular structures based on PLSM.The optimization of cellular structures is divided into two hierarchies.In macrostructure hierarchy,the PLSM is applied to optimize the substructure distribution in macroscale.In mesoscale substructure hierarchy,the PLSM is used to optimize the topology of the substructure.To avoid scale separation,the substructure method with static condensation is applied to bridge the macrostructure and the mesoscale substructures.Besides,a semi ersatz material(SEM)model is proposed to approximately describe those substructures cut by the zero contour of macroscale LSF.Owing to the scale non-separation nature of the substructure method,the proposed method avoids the connectivity problem between unit cells in the hierarchical topology optimization method when considering multiple types of substructures.Fourthly,a PLSM based multiscale topology optimization method for shell-infill cellular structures is proposed.Shell-infill cellular structures are generalized sandwich structures.Based on the advantages of LSFs,we utilize two level sets of a single macro LSF to describe the distribution domain of shell and infill.Another meso LSF is used to describe the topology of infill microstructures.A multiscale concurrent topology optimization model is established to obtain the optimal shell and infill structure simultaneously.In order to guarantee the uniform thickness of a shell,a reinitialization scheme is applied to the macro LSF for the signed distance property.The homogenization method is used to evaluate the effective elasticity matrix of the infill microstructure,bridging the macrostructure and microstructures.This method is shown to be effective for the design two and three-dimensional shell-infill cellular structures.Fifthly,the PLSM topology optimization platform is implemented by combing MATLAB and ANSYS software.The PLSM topology optimization functional module is written in MATLAB language.With the aid of user-programmable subroutines,functions,and other developing tools provided by ANSYS,it is able to access the FEA solver as well as results that are repeatedly called during topology optimization.The effectiveness of the developed PLSM platform and the reliability of the method are then verified by a subsystem structure and a sandwich sturcutre of satellites.Developing the PLSM topology optimization platform is the first integration of PLSM on commercial FEA software,which will promote the industrial application of PLSM.Finally,the dissertation concludes the main research achievements and innovations.Future research that worth further study are then outlined.