Artificial Neural Network based Evaluation of Wire-EDM Process

摘要

Wire-EDM process is gaining a significant importance in modern manufacturing activity when a complicated shape is to be made on difficult to machine materials. This paper highlights the relevance of selection of machining parameters like discharge current and discharge time for better machining in wire-EDM. Machining has been carried out on different work materials of varying hardness like Mild steel, OHN steel and HCHCr steel with two types of tool wires, bare brass wire and zinc coated brass wire. An artificial neural network model has been developed to study the process parameters such as surface finish, material removal rale, wire wear rate and cutting speed. Experimental results and ANN computed values are observed to be compatible. SEM and EDXA results are discussed to study the characteristics of machined surface and worn out wire.