Predictive analysis of surface roughness in argon-assisted EDM using semiempirical and ANN techniques

Abstract

This experimentation explores the utilization of argon-assisted electrical discharge machining (AAEDM) of high-carbon high-chromium die steel. High-pressure argon gas in conventional EDM was utilized to assess the surface roughness (SR). Analysis of variance was connected to decide the critical parameters influencing SR. In this study, a mathematical model has been instigated to get to know SR by using Buckingham pi-theorem while applying the AAEDM process. The fit summary confirmed that the quadratic model is statistically appropriate, and the lack-of-fit is insignificant. Root-mean-square error and absolute standard deviation, obtained through response surface method, were also used for developing the model and for its predicting abilities through ANN. The experiment and anticipated estimates of SR during the process, obtained by dimensional analysis and ANN, were found to be in accord with each other. However, the ANN technique proved to be more fitting to the response as compared to the dimensional analysis.

Identification of the effects and interactions of factors on the EDM process in order to model it using Taguchi method

“…Pulse time (Ton) [1]- [2]- [3]- [4]- [5]- [6]- [7]- [8]- [9]- [10]-[11]- [12]- [13]- [14]- [16] Rest time (Toff) [1]- [2]- [5]- [6]- [7]- [8]-[11]- [12]- [13]- [14] Discharge current (I) [1]- [3]- [4]- [5]- [6]- [7]- [8]- [9]- [10]-[11]- [12]- [14]- [16] Secondary Distance between electrode and work piece (GAP) [6] Duty cycle (DC) [2]- [10]- [12] ]- [16] Voltage (V) [2]- [3]- [4]- [8]- [10]- [12]- [13] Parameter (TUP) [9] Recently, there has been a sharp increase in the application of the experimental method based mainly on practical trials. Several experimental modeling techniques with varying degrees of complexity have been w...…”

“…TWR = b0 + b1 × I + b2 × Ton + b3 × I×Ton + b4 × Toff + b5 × I×Toff + b6 × GAP+ b7 × I×GAP + b8 × AUX + b9 × I×AUX+ b11 × I² + b22 × Ton² + b44 × Toff² + b66 × GAP² + b88 × AUX² (16) The statistical analysis leads us to the study table of the variance of Table 23. It mainly indicates that the determined model is very well fitted, since the sum of the squares of residuals is very small compared to the sum of the regression squares.…”

“…The material removal occurs by a series of successive electric discharges, separated from each other in time, in the main and secondary machining parameters have a major influence on the workpiece and the tool during machining [1]. Numerous studies have been carried out on the process of machining by spark erosion regardless of by experimental methods [1-2-3-4-5-6-7-8-9-10-11-12-13-14] or by numerical simulations using programs such as ABACUS [15] or MATLAB [16] for different types of material. Different types of EDM machining have been carried out to remove material from a part, mainly EDM (Sinking EDM) [1-5-6-8-9-10-11-12-17] in which an electrode in the shape to be machined is inserted into the work piece to etch hard materials, particularly wire EDM [3-4-7-13-14-18-19-20] where a moving conductor wire cuts a part along a ruled surface, especially Strip EDM [17] which uses a strip electrode to machine parts of hard materials.…”

Optimization of Electrical Discharge Machining Process by Metaheuristic Algorithms

“…The term "non-traditional machining" refers to the removal of surplus material utilizing a variety of methods, including electrical, chemical, thermal, and mechanical energy [3]. Electrically conductive materials with high levels of durability and hardness are often machined using the EDM process [1,4,5]. Besides, EDM processes are adaptable since they may continue to operate independently for extended periods [2,6].…”

“…This makes it necessary to carefully analyse the trade-off between surface finish and other machining performance outcomes. By understanding the relationship between the machining process parameters and Ra, manufacturers can optimize their processes and achieve the desired results [4,5]. Soft computing techniques such as fuzzy logic, genetic algorithms, and neural networks were used to overcome this challenge.…”