Numerical Simulation And Process Optimization Of Selective Laser Sintering

Selective Laser Sintering (SLS) was a Rapid Prototyping (RP) technology that developedrapidly in the end of the1980s. It attracted more and more attention for its wide use ofmaterial, having no use clamping fixture, high material utilization and shorting the cycle ofdesign.With the development of the times, the requirements of the strength and accuracy ofparts fabricated by SLS were increasing.Therefore, it was meanful to study the SLS process.In this paper, employing304L stainless steel as material, the numerical simulationanalysis of thermal physical process of SLS was carried out by the finite element analysismethod. The prediction models of process parameters and performance parameters oftemperature field and stress field based on BP neural network have been established.Theprocess parameters have been optimized by genetic algorithms.The study contents and resultsof this dissertation were as follows:By means of the finite element software ANSYS, a three-dimensional finite elementmodel of SLS processing was built, in which allowed for heat conduction,heat radiation andheat convection, non-linear behavior of the thermophysical property parameters and latentheat due to phrase transformation from powder to liquid, the dynamically loading of themoving Gaussian laser thermal resource have been carried out using the APDL (ANSYSParametric Design Language).The temperature field, stress field and deformation of SLS havebeen analyzed by this model. The sintering simulation deformation results were consistentwith the experimental results.Based on a large numbers of simulations, the prediction models between the processparameters and the average maximum temperature, average maximum temperature gradient,maximum deformation and maximum equivalent residual stress in SLS sintering process havebeen established individually by BP neural network. The quantitative relations betweenlaser power, scanning speed, scanning spacing，thickness of the powder layer, preheattemperature, and the outputs were investigated. The models could be used to predict theaverage maximum temperature, the average maximum temperature gradient, the maximumdeformation and the maximum equivalent residual stress in SLS sintering process. It washelpful to choose the reasonable process parameters.The process parameters have been optimized by the BP neural network and geneticalgorithms.The optimized results were laser power p=567W, scanning speed v=11mm/s,scanning space l=1mm, the thickness of the powder layer d=0.17mm, preheat temperatureT=442K.The simulation results showed that the optimized process parameters were effected. Based on BP neural network prediction models, the impact of one and two processparameters on average maximum temperature, average maximum temperature gradient,maximum deformation and maximum equivalent residual stress in laser sintering processwere analyzed.