Meta-heuristic optimization of copper friction stir weldments

Abstract

This work focused on the optimization of process parameters, which may result in increasing mechanical properties of copper weldments. The different tool pin profiles such as plain taper cylindrical, taper cylindrical with threaded, triangular, square, pentagonal and hexagonal having constant shoulder diameters were used to fabricate the weldments. The experiments were conducted at different levels of tool rotational speed and weld speeds using six different tool pin profiles. The experimental results revealed that the defect free weldments could be obtained by using different tool pin profiles. From the investigation, it was found the weldments made by using a square (SQ) tool pin profile resulted in better mechanical properties compared to other tool pin profiles. Objective functions are developed for the mechanical properties in terms of input parameters. The input parameters of an SQ tool pin profile were optimized using a metaheuristic optimization based algorithm named teaching learning based optimization (TLBO) technique to improve mechanical properties. The TLBO suggests a combination of 900 rpm of tool rotation speed and 40 mm/min weld speed for better properties.

Study of optimum welding performance in friction stir welding of dissimilar Mg alloys using integrated RSM-TLBO algorithm

“…This technique provides enhanced desired responses at single set of parameters. 46 Effects of TRS, WS and SD on responses SR, GS, TS, MH and FS have been considered for study. Here, we have two types of objective: minimize SR and GS, and maximize TS, MH and FS.…”

“…Then optimized values obtained from single-objective optimization were used for developing regression function for multi-objective optimization and this function can be represented by equations (14) to (16) The objective function can be represented by Constraints TLBO algorithm code was run on MATLAB for obtaining optimum value of objective function. The values of function, SR, GS, TS, MH and FS at respective parameters (as per TLBO predicted value) were calculated using equations (7) to (11) 46 and listed in Supplemental Table 6. The obtained optimum value is listed in Table 4 and convergence curves are shown in Supplemental Figure 3.…”

“…TLBO algorithm code was run on MATLAB for obtaining optimum value of objective function. The values of function, SR, GS, TS, MH and FS at respective parameters (as per TLBO predicted value) were calculated using equations ( 7) to ( 11) 46 and listed in Supplemental Table 6. The obtained optimum value is listed in Table 4 and convergence curves are shown in Supplemental Figure 3.…”

Experimental and optimization studies of friction stir processed Cu-TiB2 surface composites

“…The mechanical characteristics of the FSP were enhanced by using conventional optimization methods such as like Taguchi technique, 19 design of experiments, 20,21 particle swarm optimization, 22 grey relation analysis, 23 genetic algorithm, 24 and teaching learning based optimization. 25,26 As the multiresponse optimization is confusing and complex, 27 a novel optimization technique is called GTUC. 28,29 In the present study, weights to the mechanical characteristics were estimated scientifically using the proposed optimization technique.…”