Multi-scale Hybrid Uncertainty Propagation Analysis And Reliability Optimization Of Torsion Beam CFRP Bea

Carbon fiber reinforced plastic(CFRP)is widely used as one of the main lightweight materials in key automotive components.The twist beam is the main load-bearing components of automobiles,and lightweight design of them can improve vehicle dynamics and fuel economy.Carbon fiber reinforced plastic have significant multiscale characteristics.Based on the multiscale analysis method,this paper analyzes the cross-scale uncertainty propagation and reliability optimization of the carbon fiber reinforced plastic crossbeam of the twist beam,providing a theoretical basis and technical support for the design and development of carbon fiber reinforced plastic automotive components,and promoting the application of lightweight and high-strength carbon fiber reinforced plastic in the automotive field.Firstly,based on the multi-scale asymptotic homogenization theory and periodic boundary conditions,the microscopic fiber bundle unit cell model is established,and the elastic properties of the fiber bundle are predicted by the formula method and the finite element method respectively.Based on the real fiber bundle structure,a mesoscopic plain woven CFRP unit cell model was established to predict the macroscopic equivalent elastic modulus of CFRP.The samples were made by vacuum assisted resin infusion for tensile,compression and shear test to verify the accuracy of the multi-scale analysis method.Then the finite element model of steel twist beam is established to analyze the stiffness,strength and modal of twist beam.Based on the principle of equal stiffness,the steel crossbeam of twist beam is replaced by CFRP crossbeam,and the section of CFRP crossbeam is improved.Considering the influence of moment of inertia,the U-shaped section crossbeam is replaced by S-shaped section crossbeam.The S-shaped CFRP crossbeam sample was made and the test bench was built for stiffness and modal tests.The accuracy of the finite element model of the S-shaped CFRP crossbeam was verified by the test results.Based on this,the ply stacking sequence of the S-shaped CFRP crossbeam is optimized considering the ply stacking process constraints,and the optimal stacking scheme of the S-shaped CFRP crossbeam is obtained.On this basis,the probability model is used to characterize the random uncertainty parameters,and the interval model is used to characterize the cognitive uncertainty parameters.The uncertainty propagation analysis of CFRP structure from low-order scale to high-order scale is realized by Monte Carlo simulations.Then,the Monte Carlo simulation method based on approximate model technology and the statistical solution method based on sampling method are used to analyze the cross-scale random-cognitive hybrid uncertainty propagation of the CFRP crossbeam,and the influence of multi-scale uncertain parameters on structural performance is studied.Finally,a multi-scale reliability optimization method suitable for plain woven CFRP structures is proposed.The third-order response surface approximation model is established.The minimum mass,maximum torsional stiffness and the maximum first-order natural frequency are taken as the optimization objectives.Considering the influence of maximum stress,bending stiffness and reliability,the multi-scale reliability optimization of the CFRP crossbeam is carried out based on NSGA-II algorithm.The Pareto front optimal solution is mined by the entropy weighted TOPSIS method.The optimization results show that,compared with the steel crossbeam,under the premise of satisfying the higher reliability,the mass of the optimized CFRP crossbeam is reduced by 46.96 %,and other properties are improved.The integrated design of material-structure-performance is realized.