A two-stage multiobjective optimization algorithm for porous air bearing design. (January 2016)

• Record Type: Journal Article
• Authors: Wang, Nenzi; Chen, Hsin-Yi
• Is Part Of: Tribology international. Volume 93 Part A(2016)
• Page Start: 355
• Page End: 363
• Publication Date: 2016-01
• Subjects: Multiobjective optimization -- Air bearing design -- Porous bearing -- Group inching fortification method; Tribology -- Periodicals; 621.89
• DOI: 10.1016/j.triboint.2015.09.045
• Languages: English
• ISSNs: 0301-679X

Abstract

Abstract: Many porous bearing or tribological component designs require extensive numerical analyses which usually is followed by trial testing of prototypes before the final production. If chosen properly tribological models and optimization methods can be a practical and effective tool for the design tasks. This study presents a multiobjective optimization algorithm, two-stage group inching fortification (GIF) method, to solve a porous air bearing design. A comparison of the proposed approach with a genetic algorithm (GA) and hyper-cube dividing method (HDM) is conducted. The results show that the Pareto solution set obtained by the first- and two-stage GIF methods have a wider spread of Pareto front with a reduced number of objective-function calls than by using the GA or HDM. Highlights: Many tribological component designs require an efficient algorithm for solving problems with more than one objective. This study presents a two-stage multiobjective optimization method, which is straightforward, effective, and easy to implement. A comparison of the proposed approach with a genetic algorithm and hyper-cube dividing method is conducted to validate the claim. A better quality of Pareto front in terms of a broader spread was obtained in this study.