基于响应面法SLA光敏树脂3D打印工艺策略

3D Printing Process Strategy of SLA Photosensitive Resin based on Response Surface Method

综合考虑打印精度、试件强度及时间效率等指标,研究了立体光固化(SLA)光敏树脂的3D打印工艺,开发应用于不同需求的智能工艺策略包。通过对不同工艺参数的成型试件进行成型误差测量、拉伸强度测试、打印时间统计,基于响应面法以激光功率(设备最大激光功率百分数)、扫描速度、扫描间距三个关键的工艺参数为输入变量进行三因素三水平的Box-Behnken试验。采用逐步回归法剔除不显著的变量,拟合出二次回归方程,以便对试验结果进行多目标预测分析,从而得出三种不同的工艺策略。结果表明,在质量模式下(优先考虑成型误差和拉伸强度),工艺参数范围为激光功率33%~35%,扫描速度4300~4500 mm/s,扫描间距0.16~0.18 mm;在速度模式下(优先考虑打印时间),工艺参数为激光功率30%~32%,扫描速度6600~6750 mm/s,扫描间距0.18~0.2 mm;均衡模式下,工艺参数为激光功率35%~37.5%,扫描速度5100~6500 mm/s,扫描间距0.1~0.15 mm。

The 3D printing process of stereo lithography apparatus(SLA)photosensitive resin was studied with comprehensive consideration of printing accuracy,specimen strength and time efficiency,then an intelligent process strategy software package was developed for different requirements.The forming error measurement,tensile strength test and printing time statistics were carried out on the molding specimens with different process parameters.Based on the response surface method,three key process parameters,namely laser power(percentage of maximum laser power of equipment),scanning speed and scanning space,were used as input variables to carry out the box-Behnken test with three factors and three levels.The stepwise regression method was used to eliminate the insignificant variables,and the quadratic regression equation was fitted to make multi-objective predictive analysis of the test results,so as to obtain three different process strategy packages.The experimental results show that the process parameters ranges are 33%–35%of laser power,4300–4500 mm/s of scanning speed and 0.16–0.18 mm of scanning interval in the quality mode(forming error and tensile strength are preferred).In the speed mode(printing time is preferred),the process parameters are 30%–32%of laser power,6600–6750 mm/s of scanning speed and 0.18–0.2 mm of scanning interval.In the equalization mode,the process parameters are 35%–37.5%of laser power,5100–6500 mm/s of scanning speed and 0.1–0.15 mm of scanning interval.