Process structure property correlation and multi response optimization of friction stir welded copper joints

Friction stir welding (FSW) of copper is challenging due to its high thermal conductivity, which complicates heat management and joint integrity. This study investigates the process–structure–property relationship in FSW of 5 mm thick commercially pure copper plates using a conventional vertical milling machine as a cost-effective alternative to dedicated FSW systems. A Taguchi L9 orthogonal array was employed to systematically analyze the effects of tool pin profile (cylindrical, threaded cylindrical, square), rotational speed (910–1700 rpm), and welding speed (16–44 mm/min) on tensile strength and hardness. Multi-response optimization using signal-to-noise (S/N) ratios, analysis of variance (ANOVA), and Grey Relational Analysis (GRA) quantified parameter significance and identified optimal combinations. Welding speed dominates tensile strength (76.90% contribution, p < 0.001) due to its direct control over heat input and material flow, followed by rotational speed (21.95%), with tool profile having a minor effect (1.15%). A maximum tensile strength of 350.14 MPa (~ 87% of base material) was achieved using the threaded cylindrical tool profile at 1700 RPM and 16 mm/min (S6). In contrast, tool geometry primarily governs hardness (84.67% contribution, p < 0.001) through its role in strain distribution and dynamic recrystallization, with a maximum hardness of 94 HRE obtained using the square pin profile at 910 RPM and 44 mm/min (S7). Microstructural analysis reveals significant grain refinement in the stir zone, with average grain sizes of 3.2–6.8 µm representing a 73–87% reduction compared to the base metal (25.4 µm). Quantitative SEM/EDS analysis confirms chemical homogeneity of the joints (99.6 ± 0.2 wt.% Cu) with no detectable tool contamination, indicating that the mechanical property variations arise from thermo-mechanical processing rather than compositional differences. Grey Relational Analysis identifies the threaded tool profile with 1700 RPM and 16 mm/min as the optimal multi-response condition (GRG = 0.706), confirmed by confirmation tests with errors of 0.66% and 0.39% for tensile strength and hardness, respectively. These findings establish a clear mechanistic basis for parameter selection in copper FSW and validate the feasibility of using conventional milling equipment, offering a practical and low-cost experimental route for FSW of copper in manufacturing and prototyping applications.