Zirconium oxide nanoparticle-reinforced aluminium alloy (AA7075) matrix composites via hot extrusion and equal channel angular pressing

Direct solid-state process such as hot extrusion and equal channel angular pressing (ECAP) are alternative and efficient solid-state processes for recycling aluminium alloy AA7075 scrap. These processes utilize less energy and are eco-friendly. Ceramic particles such as zirconium oxide (ZrO2) have favourable mechanical and electrical behaviours, good wear resistance and a wide bandgap. Therefore, ZrO2 is suggested as reinforcement in the production of aluminium alloy AA7075 matrix composites (AMCs). Aluminium alloy AA7075 recycling have limitations on achieving good mechanical and physical properties and the products of the direct recycling process are still struggling with parameters optimization. Moreover, the combination of hot extrusion and ECAP metal forming has gained acceptability, but there are extreme challenges through the quality issues and enhanced composite alloy with a cost-effective. This study investigated and optimized through the response surface methodology (RSM) the effect of the volume fraction (VF), preheating temperature (T), and preheating time (t) on the mechanical and physical properties of the AA7075-ZrO2 composite produced by hot extrusion. Additionally, the effect of heat treatment (T6) on the optimal sample was investigated. In addition, examine the elemental components the ECAP process. Moreover, developed a machine learning model based on extra trees (ET) to predict the properties and optimise the parameters. Each parameter was evaluated at varying magnitudes, i.e., 450, 500, and 550 °C for T; 1, 2, and 3 h for t, and 1, 3, and 5 % for VF. The effects of the process variables on the responses were examined using the factorial design with centre point analysis. A total of 28 experimental runs were performed through the hot extrusion process. The optimum sample was heat treated to investigate the effect on ultimate tensile strength (UTS), compressive test, microhardness, and density before and after the heat treatment condition as well as after ECAP. The recorded datasets were used for training and testing of Artificial Intelligence (AI) models were executed using machine learning methods. The AI models applied in this study was Extra Trees (ET). T and VF were crucial for attaining the maximum tensile strength 490 MPa, was attained at 550 °C, 1.58 h, and 1 vol% ZrO2 with a microhardness 95.2 HV, compressive strength 545 MPa and density of 2.89 g/cm3. Also, the hot extrusion and ECAP followed by heat treatment strengthened the microhardness by 64%, compressive strength by 17% and density by 3%. The results exhibited that the preheating temperature and volume fraction are the most important factor that was needed to be controlled to obtain the optimum UTS and microhardness. Preheating time has a big effect on density. The accuracy of mechanical and physical properties (ultimate tensile strength (UTS), microhardness and density) prediction of AI models along with RSM model. The obtained results revealed that the extra trees (ET) model showed outstanding performance amongst the model for training, testing, and overall datasets with coefficient of correlation (R2), mean absolute error (MAE) and mean squared error (MSE) value of 97.6, 10.8 and 2.32, respectively. The impact of hot extrusion parameters and ECAP followed by heat treatment on the average grain sizes and microstructural analysis of the recycled samples were equally investigated and discussed in detail. Thus, it concluded that the ZrO2, ECAP and heat treatment have a significant effect on recycled AA7075 chips. Ideationally, the ET machine learning model can minimize the experimental complexities, time, and expense in the manufacture.

Text FK 2022 101 IR.pdf Download (1MB)

Actions (login required)

View Item