Estimation of Machining Performance in Wire EDM of Aluminum Silicon Nitride Composite an Experimental Analysis and ANN Modeling

¹ College of Engineering, Department of Chemical Engineering, University of Bahrain, Bahrain
² Department of Mechatronics Engineering, KCG College of Technology, Chennai 600097, Tamil Nadu, India.
³ Department of Mechanical Engineering, CVR College of Engineering, Ibrahimpatnam, Telangana 501510
⁴ Department of Civil, KCG College of Technology, Karapakkam, Chennai 600 097, India
⁵ Department of Research and Innovation, Saveetha School of Engineering, SIMATS, Chennai, Tamil Nadu 602105, India.
⁶ Department of Mechanical Engineering, Government College of Engineering, Keonjhar, Odisha, 758002.

^* Corresponding author: malansari.uob@gmail.com

Abstract

The primary objective of the current research is to optimize machining performance in Al 7010 alloyreinforced with silicon nitride nanoparticles. This has been accomplished through a combination ofexperimental analysis and predictive modeling methodologies. Initially, composite materials were createdusing stir casting, and varied percentages of silicon nitride were incorporated into the material to supplementits mechanical properties. Wire Electrical Discharge Machining was performed using different parameters suchas Pulse On Time , Pulse Off Time , and Current , and a range of these parameters was defined according tolevels . Material Removal Rate and Surface Roughness were chosen as the machining responses and indicatedhigh sensitivity to variations in chosen parameters. Each response was thoroughly investigated and detectedusing these responses before establishing the optimized levels. Taguchi design of experiments and signal-tonoiseratio were two common techniques used to investigate parameter interactions, and they were also used todetermine the optimum combinations for both the parameters for optimizing MRR and minimizing SR.Moreover, an Artificial Neural Network (ANN) model was also established to foresee the response readingswith great precision and predict the parameter effect to enhance further predictive modeling capabilities inmachining. The present research optimization results indicated that the maximum MRR is obtained at Pulse OnTime , Pulse Off Time , and Current levels, while the minimum SR is obtained at Pulse On Time , Pulse OffTime , and Current levels. These findings provide promising avenues of research in the field of aerospace,indicating the possibility of machining components with superior machinability and mechanical strength.Furthermore, the predicting ability of an ANN model helps in obtaining the insights to engineers to optimizetheir process by gaining information about performance and material response.