Diameter Error Prediction Using Fuzzy Logic for Cast Nylon 6 Turning Operation

A prediction model using fuzzy logic system for cast nylon 6 turning work piece diameter error forecasting was developed. The cutting parameters consisting of feed rate, depth of cut and cutting speed were determined as input variable of 75 fuzzy rules of input membership function. The diameter error values from 11 levels of fuzzy rules of output membership function were arranged as prediction outputs. Turning was carried out at 3 levels of turning speed and 5 levels feed rate then depths of cut were accomplished. Both experimental and predicted work piece diameter errors were used for validation. After that factors affecting on diameter error, precision and accuracy of the prediction values were statistically analyzed. The experimental results revealed that length of work piece and cutting speed did not effect on diameter error. However, depth of cut and feed rate affected on diameter error with total causal effect values at 0.6553 and 0.3085, respectively. Prediction of diameter error by the developed fuzzy system has an average absolute error at 0.535 μm and R2 at 0.988 compared with the experimental data. The precision of prediction was not significantly different from the experimental data at 95% of confidence interval. Therefore, the developed fuzzy system can predict diameter error in turning operation with high accuracy and precision.