Neural Network Signal Processing in Spark Assisted Chemical Engraving (SACE) Micromachining

Abstract

Spark Assisted Chemical Engraving (SACE) is an emerging micromanufacturing technology of mainly non-conductive materials like glass and ceramic. The micromachining happens due to high temperature etching in electrolytic solution by electrochemical discharges which are generated through a tool-electrode across a gas film surrounding it. The gas film shall be present so that discharges, which are the heat source, can be generated hence causing local machining of the substrate. Studies have shown that the gas film breaks and reforms every few milliseconds depending on several factors, some of which are not known or are unclearly understood. Investigation of the gas film formation, its characteristics and the factors that affect its stability could lead to enhancing the SACE machining performance. In this work an algorithm based on Artificial Neural Networks (ANN) is developed to accurately estimate the gas film formation time. The method shown is a comprehensive one that can be applied to various machining conditions of the SACE process. To our best knowledge, few attempts have been done in the field of SACE signal processing and this work is the first study where ANN is used for gas film parameters calculation.