Improving tensile strength of radial-additive friction stir repaired exceeded tolerance hole of 2024-T4 Al alloy by EHGWOA-BPNN

Abstract

The radial-additive friction stir repairing (R-AFSR) process was used to successfully repair the exceeded tolerance hole of 2024-T4 aluminum alloy, and the formation and tensile strength of repaired joint were investigated. The enhanced hybridizing grey wolf optimization algorithm (EHGWOA) was innovatively proposed and had the advantages of strong search ability and large convergence speed. The weights and thresholds of back propagation neural network (BPNN) were optimized by the EHGWOA to enhance its prediction accuracy. Then, the EHGWOA-BPNN system was used to predict the joint strength and optimize the process parameters combination of R-AFSR process. The results showed that the repaired joint had not only the stir zone (SZ) with the thickness almost equal to the plate thickness but also no kissing bond defect in the SZ, thereby leading to the high tensile strength of the repaired joint. The maximum tensile strength of 308 MPa was obtained under the process parameters optimized by the EHGWOA-BPNN system, which was 4.5% higher than the maximum strength before optimization. The R-AFSR has great prospects to repair the exceeded tolerance holes of aluminum alloys, and the EHGWOA-BPNN system can be used to maximize the strength of repaired joint.

Keywords

• Radial-additive friction stir repairing
• 2024-T4 aluminum alloy
• EHGWOA-BPNN system
• Tensile strength