Prediction Of Hardness And Wear Behaviour Of Friction Stir Processed Cast A319 Aluminum Alloys Using Machine Learning Technique

F. S. Al-Enzi, Faisal; S. Mohammed, S.

doi:10.21608/erjsh.2020.228172

Prediction Of Hardness And Wear Behaviour Of Friction Stir Processed Cast A319 Aluminum Alloys Using Machine Learning Technique

Document Type : Research articles

Authors

¹ Mechanical Engineering Department, Faculty of Engineering at Shoubra, Benha University, Cairo, Egypt

² Mechanical Engineering Department, Faculty of Engineering at Shoubra, Benha University, Cairo, Egypt.

10.21608/erjsh.2020.228172

Abstract

In the present investigation, the influences of the friction stir processing (FSP) on the microstructure, hardness and wear behaviour of cast A319 Al alloy were investigated. The influence of the FSP process parameters, namely, the tool rotational and traverse speeds as well as the number of processing passes on the aforementioned characteristics were evaluated. Machine learning (ML) artificial intelligence (AI) technique was used to develop models predict the hardness and wear rate of the friction stir (FS) processed A319 Al alloy. The results revealed that FSP significantly improved the microstructure, hardness and wear resistance of the as-cast A319 Al alloy. FSP eliminated the structural defects such as porosities and cavities found in the as-cast alloy. Also, the coarse Si particles found in the as-cast alloy were totally vanished and replaced with high density and fine Si particles which are more uniformly distributed in the FS processed zones. Such microstructural modifications resulted in an enhancement in the hardness and wear rate of the FS processed A319 Al alloy when compared with the as-cast alloy. The developed ML models showed high accuracy and can be used successfully to predict the hardness and wear rate of the FS processed A319 Al alloy. The models exhibited mean absolute error percentage (MAEP) of about 2.15%, and 2.3% for the hardness and wear rate, respectively.

Keywords