Enhancing high-speed EDM performance of hybrid aluminium matrix composite by genetic algorithm integrated neural network optimization

Autor: Muhammad Asad Ali, Nadeem Ahmad Mufti, Muhammad Sana, Mehdi Tlija, Muhammad Umar Farooq, Rodolfo Haber
Jazyk: angličtina
Rok vydání: 2024
Předmět:
Zdroj: Journal of Materials Research and Technology, Vol 31, Iss , Pp 4113-4127 (2024)
Druh dokumentu: article
ISSN: 2238-7854
DOI: 10.1016/j.jmrt.2024.07.077
Popis: Hybrid aluminium matrix composites (HAMCs) are highly valued in manufacturing sectors but are difficult to machine conventionally due to reinforcements' inherent hardness and abrasiveness. This research finds high-speed wire electric discharge machining (WEDM) to be a potent solution for machining stir-squeeze-casted HAMC (AA2024 with ceramic nanoparticles: Al2O3, SiC, Si3N4, BN) and creating complex profiles with superior erosion characteristics. The erosion performance has been assessed in terms of material removal rate (MRR) for different profiles (plane, angular, and curve), and wire wear ratio (WWR) by employing machining variables, i.e., pulse voltage (PV), pulse current (PI), wire feed rate (WFR), pulse (P), and drum speed (DS). Results revealed that MRR_curve has highest MRR (37.84 mm3/min), followed by MRR_angular (36.07 mm3/min) and MRR_plane (34.40 mm3/min). The lowest WWR (0.0094%) was achieved at lower magnitudes of machining variables. The microscopic observations unveil shallow craters, minute-sized melt redeposits, and micro-pores on machined surface under the conditions of PV = 90 V, PI = 2 A, WFR = 13 m/min, P = 20 mu, and DS = 40 Hz. Optimization using non-dominated-sorting-genetic algorithm (NSGA-II) resulted in significant enhancements of 75.37, 73.90, and 76.01% in MRR_plane, MRR_angular, and MRR_curve, respectively, and a depreciation of 16.50% in WWR.
Databáze: Directory of Open Access Journals