Affiliation:
1. Department of Mechatronics Engineering, Nelson Mandela University, Port Elizabeth, South Africa
Abstract
Due to increasing demand for high accuracy and high-quality surface finish in optical industry, contact lens manufacturing requires reliable models for predicting surface roughness (Ra) which plays a very important role in the optical manufacturing industry. In this study, a Nanoform 250 ultra-grind turning machine was used for machining, while cutting speed, feed rate, and the depth of cut (with values selected to cover a wide range based on the literature) were considered as the machining parameters for a diamond turned rigid polymethylmethacrylate (PMMA) contact lens polymer. Turning experiments were designed and conducted according to Box–Behnken design which is a response surface methodology technique. Fuzzy logic-based artificial intelligence method was employed to develop an electrostatic charge (ESC), Ra, and material removal rate (MRR) prediction models. The accuracy and predictive ability of the fuzzy logic model was then judged by considering an average percentage error between experimental values and fuzzy logic predictions. Further, a comparative evaluation of experiments and fuzzy logic approach showed that the average errors of ESC, Ra, and MRR using fuzzy logic system were in tandem with experimental results. Hence, the developed fuzzy logic rules can be effectively utilized to predict the ESC, Ra, and MRR of a rigid PMMA contact lens polymers in automated optical manufacturing environments for high accuracy and computational cost.
Subject
Condensed Matter Physics,Ceramics and Composites
Cited by
4 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献