Optimization of polyoxymethylene spur gear pair using meta-heuristic algorithms: A comparative study

Abstract

Design-optimization of power transmitted elements such as gears is a complicated process to accomplish as the mathematical model acquires to be in compliance with the optimization algorithm used. Utilization of technical standards and handbooks assisted engineers and designers with designing the gears, however the results were inefficient, thus meta-heuristic algorithms were approved for optimization, for example, genetic algorithm. In this paper, optimization of single stage spur gear pair made of polyoxymethylene material was carried out using genetic algorithm (GA), artificial bee colony (ABC), JAYA, grey wolf optimizer (GWO), and whale optimization algorithm (WOA) in MATLAB. Three single functions were used as objectives, those are, weight, power loss, and center distance. Module ( m), face width ( b), number of pinion teeth ( z1), and finally profile shifts ( x1, x2) served as design variables. Unlike steel gears which were constrained by handling root bending stress and flank pressure, more constraints were added to the problem with the increase of complexity such as root and flank temperature, tooth deformation, and wear abrasion. The results showed a significant decrease in all three objectives when optimizing each objective solely with variation of the variables. It is also observed that JAYA has the superiority in decreasing the three objectives in comparison to GA and ABC nevertheless, in agreement with GWO, and WOA.