Predictive modeling and multiobjective optimization of diamond turning process of single-crystal silicon using RSM and desirability function approach

In ultra-high precision machining (UHPM) of Si optical lenses, the prime task is the realization of nanometric surface finish within the reasonable tool wear rates. Thus, study of the influence of different cutting parameters like cutting speed, feed rate and depth of cut on the machinability characteristics is essential. This paper therefore aimed at developing predictive models for surface roughness (R-a) and tool wear and subsequent determination of optimal cutting conditions for the attainment of minimum R-a and tool wear as well as maximum material removal rate (MRR). Desirability function approach was used for the optimization based on response surface methodology (RSM). Also, ANOVA was used to find the statistical significance of the cutting parameters on the two responses. For R-a, the results showed that feed is the most influential factor with higher percentage contribution. The quadratic term of feed (f(2)) is the second most influential factor on R-a followed by speed. Depth is statistically insignificant. Concerning the tool wear, feed also has greater influence than the other factors, albeit, the percentage contributions are almost comparable. The speed-feed interaction in the tool wear is stronger than that of the R-a. The combined optimization of the cutting conditions, leading to minimum R-a and tool wear with maximum MRR, revealed the optimum cutting parameters to be as follows: v=679.97rpm, f=2.00mm/min and d=10.00 mu m. This combination is suitable for quality manufacturing and smaller MRR. The equivalent optimum output are as follows: R-a=2.099nm and tool wear=0.665 mu m.