Experimental modeling and optimization of surface quality and thrust forces in drilling of high-strength Al 7075 alloy: CRITIC and meta-heuristic algorithms

• Published in: Journal of the Brazilian Society of Mechanical Sciences and Engineering 2021-05, Vol.43 (5), Article 244
• Main Authors: Patel, G. C. Manjunath, Jagadish
• Format: Article
• Language: English
• Subjects: Algorithms; Aluminum base alloys; Automotive engineering; Chips; Correlation analysis; Criteria; Cutting parameters; Cutting speed; Design parameters; Drilling; Engineering; Errors; Feed rate; Genetic algorithms; Heuristic methods; High strength alloys; Machine learning; Mechanical Engineering; Morphology; Optimization; Particle swarm optimization; Power consumption; Surface properties; Surface roughness; Technical Paper; Thrust
• ISSN: 1678-5878; 1806-3691
• DOI: 10.1007/s40430-021-02928-3

Al 7075 is a renowned high-strength engineering material used in automotive and aerospace applications, wherein many functional cylindrical parts are subjected to internal or external loads. Engineered parts with form errors (cylindricity CE and circularity error Ce) result in undesirable vibration and high deformation in rotating parts. In addition, reduced surface roughness (SR) and thrust forces (TF) are essential to limit the secondary process (namely, polishing) and power consumption. Experiments are performed based on central composite design considering drilling parameters (point angle, cutting speed, and feed rate) as inputs and output performances as CE, Ce, TF, and SR. It is noted that, except feed rate for Ce, all other parameters are found significant toward the output performance. Also, prediction accuracy with ten random experimental cases resulted with the percent error of 8.4% for SR, 5.41% for TF, 10.64% for Ce, and 10.35% for CE, respectively. Continuous ribbon-like chips at higher cutting speed, loose fragmented chips at higher feed rate, and increased arc length and radius at higher point angle were observed from the chip morphology analysis. Criteria importance through inter-criteria correlation (CRITIC) method applied to determine the weight fractions for Ce, CE, TF, and SR was found equal to 0.2802, 0.1991, 0.3293, and 0.1914, respectively. Four algorithms (genetic algorithm GA , particle swarm optimization PSO, teaching learning-based optimization TLBO , and JAYA algorithm) were applied to determine the optimal drilling conditions. JAYA algorithm determined optimized drilling conditions ensure predicted output values found close to experimental values with an acceptable percent error of 10.8% for Ce, 8.9% for CE, 6.73% for SR, and 3.51% for TF, respectively.