Optimal Design Methods For Laminated Composite Structures Based On Isogeometric Analysis

With high strength-to-weight and strength-to-weight ratios,excellent heat resistance,and other characteristics,laminated composite structures have been widely used in aerospace,medical,and other fields.Adjusting the layout of the fiber reinforced composites can further improve their mechanical property.Therefore,scholars attach great importance to the research on performance optimization of the design.The design optimization faces many challenges.First,tailoring the thicknesses of panels in composite structures may break the load transfer path and cause stress concentration.Second,the large number of design variables and manufacturing constraints on fiber layout increases the complexity of optimization problem.Third,as a spatial distribution optimization problem,a uniform dense design grid would incur a huge amount of computation,while a sparse grid would lose some design freedom.To address the above issues,this thesis presents the research work as follows:To guarantee the prediction precision for mechanical performance in the optimal design,this thesis constructs the static,free vibration,and buckling analysis models of laminated plates by using the isogeometric analysis approach,based on the equivalent single-layer plate theory.For the optimization of multiple interconnected straight-fiber laminates with variable thicknesses and blending constraints,this thesis proposes a stacking sequence design method based on multiple templates.The use of multiple templates extends the design space of traditional single-template based optimization methods,while the connection strengths between laminates are guaranteed.For the optimal design of curvlinear fiber pathes in variable-stiffness laminates,which is a nonconvex programming problem with a large number of design variables and complicated constraints,this thesis proposes a multi-stage optimization method.In the stages of lamination paramater optimization,fiber path fitting,and buckling load optimization,progressively refined formulations for design variables and their grids are adopted to avoid falling into local optima.For the optimal design of variable-stiffness laminates with a non-uniform stress distribution,this thesis proposes an adaptive variable-stiffness optimization method based on the truncated hierarchical B-Splines.The distribution of lamination parameters for a composite structure is locally refined in a hierarchical way during the multi-level optimization.The adaptive optimization method can reduce the number of design variables while maintaining the optimization accuracy.