Analysis of tensile strength on friction stir welded Al 6061 composite reinforced with B₄C and Cr₂O₃ using RSM and ANN

Abstract

Aluminum (Al) alloys are reinforced with carbides and oxides to enhance their properties. Al composites are developed to meet current automotive, shipbuilding, and aviation requirements. In the current study, aluminum 6061 is reinforced with B4C and Cr2O3 separately to fabricate Al6061 + B4C and Al 6061+Cr2O3 aluminum metal matrix composites (Al MMC). The Al composites were fabricated by stir casting with a wt % in steps of 2%, 4%, and 6%. Joining of Al MMC is essential to develop valuable components. The developed composites were welded using friction stir welding (FSW). FSW is recognized and widely used for joining Al MMC due to premium weld quality with minimum defects. The present study aims to analyze the effect of process parameters and predictive accuracy of the artificial neural network (ANN) and response surface methodology (RSM). The parameters selected for the study are tool rotational speed, tool travel speed, and reinforcement wt %. The FSW was performed based on the experimental design developed by the design expert software. Through RSM analysis, it was found that both the independent factors (tool rotational and tool transverse speed) and the interaction of factors jointly contribute to the FSW joint properties. The higher ultimate strength of 139 MPa and lower tensile strength of 48 MPa are found. As the tool travel speed increase from 20 to 25 mm\min, ultimate tensile strength increase about 59%. The average accuracy of RSM was 98.26 and of ANN was 94.86.

Comparative analysis of sliding wear behaviour of A356 nanocomposites using response surface methodology and artificial neural network approaches

“…From this observation it is quite clear that the values predicted using ANN model are far more accurate than those predicted by RSM. The observation is well in line those reported by Tyagi et al [29] and Alam et al [31] on better predictive capability of ANN on wear behaviour of Al composites. Although under certain conditions RSM model did showed better predictive capability when compared with ANN model [27] however in most cases the ANN showed high reliability.…”

“…In the last decade lot of effort has been devoted towards using optimization tool like Taguchi's experimental design method, response surface methodology (RSM) and soft computing tool such as artificial neural network (ANN) for prediction of tribological properties of Al nanocomposites [26][27][28][29]. Although these tools were extensively used for prediction of tool wear and surface finish of metallic or composite materials subjected to machining but extending them to optimization of wear parameters of nanocomposites is relatively new.…”

Synergistic effects of cold rolling and age hardening on the hardness and tensile characteristics of AA6061 hybrid composites

“…A widely employed 6xxx series aluminium alloy AA6061 [24,25] with a density of 2.7 g cm −13 was used as the matrix in the study. The spectroscopic analysis of the alloy revealed the presence of 0.73, and 0.95 weight percentages (wt%) of Si, and Mg respectively, which are the major alloying elements and are within the permissible limits according to the standard AA6061.…”

Multiobjective optimization of process parameters of AZ91D/AgNPs/TiO2 composite fabricated by friction stir processing using response surface methodology and desirability

“…TOPSIS approach was used to optimize hardness and impact energy simultaneously (Suganeswaran et al , 2023). Uday et al used artificial neural network in conjugation with RSM for getting superlative output (Uday and Rajamurugan 2023a, 2023b).…”