Prediction of HAZ Maximum Hardness in Tig Welding Process by Using Response Surface Methodology Based on JWES ANN Model

  • Saleh Awad Algdary Department of Industrial and Manufacturing Systems Engineering, Faculty of Engineering, University of Benghazi, Benghazi, Libya
  • Nagwa Mejid Elsiti Department of Industrial and Manufacturing Systems Engineering, Faculty of Engineering, University of Benghazi, Benghazi, Libya
Published: 2025-06-11
Download PDF
56
24

Abstract

This study presents a developing modeling approach integrating Response Surface Methodology (RSM) with the JWES (ANN) model to predict the maximum hardness of the Heat Affected Zone (HAZ) in Tungsten Inert Gas (TIG) welding. Using EH36 TMCP steel as the base material, experimental design through Central Composite Design (CCD) was conducted by varying welding current, voltage, and velocity. The HAZ maximum hardness values were predicted via JWES ANN model and compared against those derived from RSM-based regression equations. The model demonstrated high accuracy, with a Predicted R² of 0.9821 and Adequate Precision ratio of 84.226. Statistical analysis confirmed the significance of key process parameters. The results validate the effectiveness of combining JWES ANN model predictions with RSM optimization in achieving accurate HAZ maximum hardness modeling.

Keywords: TIG welding HAZ hardness JWES ANN Response Surface Methodology Regression model Welding parameters

How to Cite

Prediction of HAZ Maximum Hardness in Tig Welding Process by Using Response Surface Methodology Based on JWES ANN Model. (2025). African Journal of Academic Publishing in Science and Technology (AJAPST), 1(2), 63-73. https://www.easrjournals.com/index.php/ajapst/article/view/42

Issue

Volume 1, No. 2, 2025

Section

Articles