A metaheuristic-based comparative structure for solving discrete space mechanical engineering problem Academic Article in Scopus uri icon

abstract

  • Composite materials have become widespread in various industries due to their exceptional properties of strength and flexibility, which creates an entirely new area of design opportunities. However, optimizing structures containing elements made of composite material is a complicated challenge in mechanical engineering due to the natural characteristics of the material. Especially, the way that two different laminates connect together needs meticulous attention. Bolt-nut joints are one such fasteners, characterized by the high concentration of stresses and skewed stress distribution along the bolt plane. To avoid mentioned problems in bolt-nuts, adhesive-bonded joints are commonly used in composite structures. But these joints are potentially vulnerable to other defects like delamination on free ends that reduction of its risk is the core of this paper. Most traditional optimization methods, such as finite element analysis or experimental approaches are characterized by numerous variables and restrictions, and complex relations described by controlling equations. So, it is crucial to seek more powerful and sound alternatives such as metaheuristic optimization techniques which can yield a reliable solution to challenging problems in a reasonable amount of time. In this study, the performance of eight well-known metaheuristic algorithms in the optimization of two distinct multilayer adhesively-bond joints is analyzed for the first time to tackle the strength against delamination which is one of the major concerns in the design of composite material structures. The performance of metaheuristic algorithms is also evaluated using two non-parametric tests of Friedman and Wilcoxon signed rank as well as interval plots. According to the findings, the three algorithms namely the Simulated Annealing, Harmony Search, and Particle Swarm Optimization offer the most reliable performance for finding the solution. Harris Hawks Optimization, Genetic Algorithm, and Bees Algorithm, on the other hand, have the worst performance in solving such problems. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

publication date

  • January 1, 2024