Classification and prediction of burr formation in micro drilling of ductile metals

Abstract

In the micro drilling of precision miniature holes, the formation of exit burrs is a topic of interest, especially for ductile materials. Because such burrs are difficult to remove, it is important to be able to predict various burr types and to employ burr minimisation schemes that consider burrs' micro-scale characteristics. In the present work, an artificial neural network (ANN) was used to predict the formation of burrs in the micro drilling of copper and brass, along with burr formation/optimisation analysis specialised for micro drills. The influence of cutting conditions, including cutting speed, feed and drill diameter, upon exit micro burr characteristics such as burr size and type was observed, analysed and classified. Based on the results, an empirical equation to predict micro burr height is proposed herein. The classification results were compared with conventional burr cases using burr control charts. Then, micro burr types were predicted by means of an ANN, using the influential parameters as input vectors. The usefulness of the proposed scheme was demonstrated by comparing the experimental and prediction/analysis results.