Investigations on wire spark discharge and MRR using neural network modeling

TiNiCo alloy is a smart material that can recover its original shape when the temperature is increased or decreased. Due to its unique class of properties, such kinds of materials are used in various industrial applications such as robotics, biomedical and aerospace industries. Conventional machining of these alloys is quite difficult hence non-conventional machining process are found to be more suitable for machining of these alloys. Wire sparks discharge machining process was chosen for the machining of Ti₅₀Ni₄₉Co₁ and Ti₅₀Ni₄₅Co₅ alloys. Pulse on the duration and servo voltage was selected as machining process parameters with their five levels and metal removal rate considered as machining response. The error analysis has been carried out through a comparison between experimental values and predicted values. An artificial neural network (ANN) was used as the prediction technique. Predicted values were found to be close to experimental values. 6% error for Ti₅₀Ni₄₉Co₁ and 5% error have been found in Ti₅₀Ni₄₅Co₅ alloy.