Innovative Production Machines and Systems Conference

Authors: Khalil Awan

Abstract

Knurling is a machining operation typically performed on a lathe machine. Knurling is a process by which a pattern is imposed on a work piece. Knurling is applied on work piece for good holding grip, for repairing or increasing the size of a component or for aesthetic reasons so that the product will look more attractive. A good quality knurl on a part usually depends on the skill of the worker or if it is made on CNC machine on the good programming of the machine. Unfortunately there are very few studies in literature which quantify the good or bad quality of a knurl. In this research different available approaches for measuring the quality of knurling are reviewed and it is suggested that knurling quality may best be defined by measuring it depth (i.e. depth of the knurl teeth). The research also focus on different parameters that influence the quality of knurl (like work piece material and diameter, tool approach and pitch). This research uses Design of Experiment (DoE) to study the impact of these parameters on the knurling quality. A factorial design 24 with 3 replicates is selected and experiments done. Using the results obtained by experiments, an Artificial Neural Network (ANN) is designed to predict the knurling quality. The ANN model’s results indicate that the ANN model predicts within 1% error of the actual values.