An investigation of the surface topography of wire tool and the modelling of wire electro-spark machined Ti-alloy using ANN

Abstract Titanium and its alloys are in demand in the field of biomedical and aerospace engineering because of its high strength-to-weight ratio and anti-corrosive properties. WESM emerged as a processing method as traditional methods could not process because of strain hardening and low material conductivity; however, wire tool consumption contributes to high costs while processing with WESM. The current Article investigates the effect of input factors on wire consumption and their significance level on wire consumption. 81 experimental runs were conducted while taking input factors Spark on, Spark off time, peak current and servo voltage at three levels, each using a full factorial approach. Experimental results were analysed using ANOVA and MLR (Multi-Linear regression). Moreover, ANN modelling is done to model the wire consumption using a 4-10-1 architecture that increased the coefficient of correlation (R 2 ) up to 0.991. Lastly, after processing, the wire tool surface is evaluated using a scanning electron microscope to assess the effect of various input energy levels and find input combinations for minimum wire consumption.