Analysis and modeling of the effects of process parameters on specific cutting energy in Gas Metal Arc Welding (GMAW) using the artificial neural network meth

Rafiee, Muhammad Fadzil Azim (2024) Analysis and modeling of the effects of process parameters on specific cutting energy in Gas Metal Arc Welding (GMAW) using the artificial neural network meth. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (Full Text) Analysis and modeling of the effects of process parameters on specific cutting energy in Gas Metal Arc Welding (GMAW) using the artificial neural network meth.pdf - Submitted Version Download (2MB)

Abstract

The manufacturing industry plays a pivotal role in global economic expansion, its substantial energy consumption poses environmental challenges. In Malaysia, the manufacturing sector accounts for a substantial 79% of the country's total energy usage, emphasizing the urgency to develop sustainable practices. This study focuses on Gas Metal Arc Welding (GMAW), contributor to energy consumption in manufacturing. The research addresses the need to investigate the impact of process parameters on specific cutting energy. The research employs a Design of Experiment approach, using an Orthogonal Array L27 (34) to systematically collect data for GMAW process parameters using parameter settings such as welding voltage, wire feed rate, joint type, and material thickness. An Artificial Neural Network (ANN) model is developed using Anaconda Navigator for Python language. This model is evaluated through root mean square error (RMSE) and coefficient of determination (R2 score). The methodology integrates experimental design, advanced statistical techniques, and cutting-edge technology to enhance the accuracy of modelling. The study identifies wire feed rate as the most influential process parameter affecting energy consumption in GMAW. Interestingly, energy consumption in GMAW proves less sensitive to welding joint types. The optimal parameter settings identified were welding voltage at 19 V, wire feed rate at 4 m/min, joint type Tee, and material thickness at 4 mm. The application of an Artificial Neural Network model successfully captures the intricate relationship between process parameters and specific cutting energy in the GMAW process. This research contributes valuable insights for optimizing GMAW processes, promoting energy efficiency, and advancing sustainable manufacturing practices.

Actions (login required)

View Item