MILD STEEL.

The continuous advancement in manufacturing and fabrication industries has led to the
development of numerous welding techniques designed to achieve high quality joints with superior mechanical performance. Among these, Tungsten Inert Gas (TIG) welding, also referred to as Gas Tungsten Arc Welding (GTAW), stands out for its precision, versatility, and ability to produce defect free welds on a variety of metals. The process is particularly effective for mild steel, which is extensively used in structural, automotive, and construction applications due to its good formability, weldability, and moderate strength (Singh & Sharma, 2020). However, achieving optimal weld quality in TIG welding is challenging because the mechanical properties of the welded joint depend on several interacting parameters, including welding current, welding voltage, gas flow rate, and welding speed. These parameters collectively determine the heat input, cooling rate, and solidification behaviour of the weld pool, which in turn influence responses such as hardness, tensile strength, yield strength, elongation, shear strength, and impact energy (Kumar & Yadav, 2018). Selecting the wrong combination of these parameters may result in weld defects, reduced mechanical strength, and inconsistent joint performance.

The application of metaheuristic approach in optimising some welding parameters in TIG welding of mild steel is presented in this work. The study aims to optimise arc efficiency (AE) and thermal efficiency (TE) of the TIG welding process by applying metaheuristic optimisation techniques (MTOs), specifically the Genetic Algorithm (GA) and Particle Swarm Optimisation (PSO). The investigation focuses on identifying the optimal combination of welding parameters current, voltage, and gas flow rate that maximise process efficiency while maintaining physical validity and conformity with established TIG welding standards. The research methodology involved implementing mathematical models of arc and thermal efficiency based on the Goldak double-ellipsoidal heat source model. These models were coded and executed using MATLAB R2024b, where GA and PSO algorithms were used independently to optimise the input parameters within defined physical ranges obtained from validated literature. Simulation runs recorded iteration-wise outputs for each parameter, allowing convergence analysis and comparative assessment between both algorithms in terms of solution quality and computational performance. The results revealed that for arc efficiency, GA achieved optimum values at 75.59A, 14.80V, and
11.27L/min, yielding an AE of 0.81, while PSO attained optimal conditions at 63.16 A, 15.57 V, and 6.97 L/min with an AE of 0.97. For thermal efficiency, GA recorded optimum values at 67.26A, 17.21V, and 13.69L/min giving TE of 0.89, whereas PSO produced 92.09A, 18.82V, and 8.47L/min resulting in TE of 0.99. The optimised efficiency values were validated with literature and found to be in close agreement with the established efficiency range for TIG welding (0.36– 0.90), confirming the reliability of the metaheuristic approach.