Crack Resistance Ability Enhanced By Biomimetic Laser Shock Peening Technology And Residual Fatigue Life Prediction Of Aircraft Skin

Based on the laser shock peening technology,this paper takes plant leaves as the biomimetic object to develop a biomimetic laser shock peening technology suitable for aircraft skin,so as to enhance the anti-crack resistance ability and improve the residual fatigue life of aircraft skin.The feature parameters of 210 Cherry leaves are extracted and the results of Lilliefors tests show that the feature parameters obey the normal distribution.Based on the feature parameters,60 biomimetic structures are designed by central composite design method and random normal distribution method.Then,the numerical model of double-sided biomimetic laser shock peening（BLSP）is developed by ABAQUS/Explicit analysis module and ABAQUS/Standard analysis module.The crack propagation simulation results of biomimetic specimens show that the biomimetic structures retard the crack propagation and improve the residual fatigue life of the samples.Biomimetic specimens of M19,R23,M15 and R10 have the best crack arrest ability.It is found that there is great residual compressive stress on the upper and lower surfaces of the biomimetic specimens.In-depth residual stress of the s specimens presents compressive residual stress at both surfaces and tensile residual stress in the middle.The residual compressive stress generated by BLSP balances part of the tensile load applied to the samples,which decreases the stress intensity factors.In addition,the asymmetric distribution of residual stress facilitates crack turning during the crack propagation process of samples.Therefore,the absolute length of the crack propagation path is extended,which improves the residual fatigue life of samples.10 million biomimetic structures are generated by random normal distribution method.Based on the trained BP artificial neural network linking with Monte Carlo method,the optimal structure is predicted to be 0.3399（D₁）,0.4157（D₂）,0.5573（D₃）,0.2681（D₄）and 34.9911（Angle）,and the predicted residual fatigue life of the structure is 104887（Cycles）.The residual fatigue life of the optimal biomimetic specimens evaluated by Paris formula is 108426（Cycles）.The relative error between the predicted value and the evaluated value is 3.264%,indicating that the established ANN is well-trained.