Artificial neural network based modeling to predict micro-hardness during EDM of cryo-treated titanium alloys

In the present paper, an attempt have been made to predict the surface micro-hardness of cryogenically treated titanium alloys using powder mixed electric discharge machining (PMEDM) process with artificial neural network (ANN) approach and Taguchi methodology through transfer of various elements on machined components. The investigated control parameters were pulse off-time, peak current, pulseon time, manganese/tungsten powder suspended dielectric fluid, electrode material, workpiece material and cryogenic-treatment (shallow and deep) of both electrode and workpiece. A special orthogonal array having a mixture of two and three levels (L-18 Orthogonal Array) of Taguchi methodology was used to assign the parameters and conduct the experiments accordingly. Micro-hardness and surface properties were highly influenced by changing the value of peak current. More hardness of specimens was observed at higher current than low value. The presence of different elements of electrode material and powder mixed dielectric was noticed on the machined surface resulted in alter the surface properties. The ANN predicted values of micro-hardness and actual experimental values of micro-hardness were observed within the limit of the agreeable error. Copyright (C) 2021 Elsevier Ltd. All rights reserved.

Automated summary

In this paper, the surface micro-hardness of cryogenically treated titanium alloys was predicted using the PMEDM process, ANN approach, and Taguchi methodology. The control parameters, such as pulse off-time and peak current, were found to significantly influence the micro-hardness and surface properties. The ANN predicted values of micro-hardness were in good agreement with the actual experimental values.