Numerical Simulation And Experimental Study On Ductile-to-brittle Transition In Micro-laser Assisted Ultra-precision Machining Of Single-crystal Silicon

Single-crystal silicon is widely used in the production of chips and aerospace optical components for its excellent performance.However,due to its high hardness and brittleness,the traditional methods are difficult to meet the processing requirements of high precision and low damage.In recent years,micro-laser assisted ultra-precision machining technology is developed,which can realize the ductile mode cutting of single-crystal silicon.And it provides a novel solution to solve the challenge of producing silicon components with high efficiency and high precision.The realization of ductile mode cutting of silicon is the key lying in this technology.However,the mechanism of ductile-to-brittle transition(DBT)of single-crystal silicon in micro-laser assisted machining has not been explored comprehensively.Also,the roles of laser parameters and tool geometry are vague,which limits the further investigation.In this paper,based on numerical simulation and experimental manner,the mechanism of DBT of single-crystal silicon is investigated in detail.The impacts of tool geometry and laser parameters on the DBT are researched.In addition,a prediction model of depth of cut(DOC)of DBT of silicon is established,which provides guidance for the optimization of silicon process parameters.The main contents of this paper are as follows:Firstly,a temperature field model is developed based on finite element method(FEM).The relationship between laser powers and temperatures of the contact area between tool and workpiece is obtained.Consequently,the ablation experiments are conducted to modify the model.The errors of relevant indexes are within 10%,which suggests that the actual temperature field can be partly reflect.Secondly,smoothed particle dynamics method(SPH)and coupled Eulerian-Lagrangian method(CEL)are used to simulate the process of micro-laser assisted ultra-precision machining of single crystal silicon.The mechanism of DBT of silicon is analyzed.During the process,the influence of temperature and tool geometry on the DBT is clarified.Moreover,by selecting appropriate temperature and tool geometry,the cutting force can be reduced by more than 50% and the DOC of DBT can be increased by more than 300%.Thirdly,grooving experiments,with the same configuration as the cutting simulations,prove that micro-assisted machining technology(μ-LAM)can effectively enhance the ductile response of single-crystal silicon.And it provides the favorable condition for the realization of ductile mode cutting under a larger DOC.In addition,based on the back propagation neural network(BP),the experimental values and the predict values of DOCs of DBT are quantitatively compared.The error of prediction value of SPH method is less than 5.9%,which verifies the accuracy of SPH cutting simulation model and its applicability in this research.Finally,based on the results of SPH simulation,a regression prediction model of DOC of DBT is established,applying BP neural network.And the model is optimized by multi population genetic algorithm(MPGA)and beetle antennae search without parameter tuning(BAS-WPT)to further enhance the prediction accuracy.The performance of the two optimization algorithms is compared.Also,the impacts of laser parameters and tool geometry on DOC of DBT are analyzed.A theoretical basis for effective parameter optimization can be obtained according to the conclusions.