Optimization Study On Technical Process Of Selective Laser Sintering Based On Simulation

Selective Laser Sintering (SLS) is a Rapid Prototyping (RP) technology that developed rapidly in recent two or three decades, and becomes an important part of concurrent engineering and reverse engineering. With the development of the SLS technology, higher accuracy and strength of the SLS parts are needed. Thus researches on the SLS process, accuracy and tensile strength of the sintering parts have great theoretical and practical significance.Based on a series of experiments, the SLS process parameters have been simulated and optimized using Neural Network and Genetic Algorithms in this paper. The main works in this dissertation are as follows:Firstly, experiments process of sintering polystyrene(PS) materials using the SLS technology are studied. The laser power, scANNing speed, hatch spacing, layer thickness as well as temperature of powder are taken as the main process parameters that affect the quality of sintering parts. Meanwhile, sintering forming stability is considered during sintering. And the shrinkage, warpage and tensile strength of the sintering parts are studied, the experimental datas that used for simulation and optimization of the BP neural network model and genetic algorithm are obtainedSecondly, shrinkage BP Neural Network model, warpage BP Neural Network model and tensile strength BP Neural Network model are established, which take the sintering process parameters as inputs,and shrinkage, warpage and tensile strength as outputs respectively. Then the neural network model is trained and tested using the experimental data. Some simulations are executed by the neural network model. Experimental results show that the Neural Network model can simulate the quantitative relationship between SLS process parameters and shrinkage, warpage , tensile strength of experiment parts. The process parameters are optimized by Genetic Algorithm and BP Neural Network model that established beforehand. The optimal process parameter combination is obtained as follows: the temperature of powder through is 94℃, the laser power is 14W, the scANNing speed is 1700mm/s, layer thickness is 0.16mm and the hatch spacing is 0.13mm.The results show that the optimal solutions of SLS processing parameters which have been optimized by Genetic Algorithm are effective.Lastly, effects on the sintering quality of parts influenced and interact influenced by a single process parameter and double ones are simulated and analysed by BP neural network model respectively. Effects on the sintering quality of experiment parts are evaluated using the prediction capability of neural networks. The results show that the laser power influences the shrinkage of the parts most followed by the hatch spacing and layer thickness. As to the warpage of the parts, the laser power and temperature of powder have more influences while the scANNing speed has less influence. For the tensile strength of parts, the laser power influences it most meanwhile the temperature of powder least.