Intelligent Analysis And Design Of Mechanical,Thermal And Electrical Properties Of Graphene-modified 3D Braided Composite

Graphene-modified epoxy 3D braided composites are a novel form of fiberreinforced structural-functional composites.With 3D multi-directional braided fabrics as reinforcements,it not only has good structural integrity and strong load-bearing capacity,but also fully utilizes the excellent mechanical-thermo-electrical properties of GNPs(Graphene nanoplatelets)modified epoxy to achieve the comprehensive performance improvement.It has become one of the research frontiers in the development of structuralfunctional integrated composites.This thesis takes GNPs modified epoxy 3D braided composites as the research object,adopts multi-scale analysis methods to study the mechanical-thermo-electrical properties,and realizes the intelligent analysis and design of material performance.The main contents include:1)In view of the problems such as high modeling difficulty and low calculation efficiency of high volume fraction GNPs modified epoxy,a 2D randomly distributed GNPs modified epoxy unit cell model is proposed.A prediction model for the mechanicalthermal equivalent performance of GNPs modified epoxy based on 2D periodic unit cell is constructed.The influence laws of GNPs volume fraction,aspect ratio and unit cell size on the mechanical-thermal equivalent performance are studied in detail.At the same time,using analytical methods,the equivalent conductivity of GNPs modified epoxy composites and its influence laws are studied.2)Based on the analytical prediction model of mechanical-thermo-electrical properties of GNPs modified epoxy-based yarn bundles and the equivalent matrix of GNPs modified epoxy,finite element analysis models of mechanical-thermo-electrical properties of 3D braided mesoscopic internal cells and surface cells are established.Then,a prediction model for the mechanical-thermo-electrical properties of 3D braided composites based on macroscopic global cells is proposed.The influence laws of critical process parameters such as braiding angle,yarn volume fraction and GNPs mass fraction on the equivalent mechanical-thermo-electrical properties of materials are studied in detail.3)A BP(Back Propagation)neural network model for predicting the mechanicalthermo-electrical performance of 3D braided composites is constructed.Combined with genetic algorithm,an optimization model for the mechanical-thermo-electrical performance of 3D braided composites is established.This provides an important idea for the intelligent analysis and design of mechanical-thermo-electrical performance of GNPs modified epoxy 3D braided composites.The analysis shows that the finite element method based on 2D randomly distributed GNPs modified epoxy unit cell model can effectively characterize the mechanicalthermal equivalent performance with high computational efficiency.At the same time,based on the two-level unit cell finite element analysis method of internal unit cell,surface unit cell and global unit cell,it effectively considers the "shell-core" structural characteristics of 3D braided composite wall panels.Combined with intelligent design and analysis model,it significantly improves the prediction accuracy and efficiency of mechanical-thermo-electrical performance of GNPs modified epoxy-based 3D braided composites.The above work lays a solid theoretical foundation for further studying the comprehensive performance of GNPs modified epoxy-based 3D braided composites.