유전알고리즘을 이용한 로터-베어링 시스템 최적설계 = Optimal Design of Rotor-Bearing System using Genetic Algorithm

다국어 초록 (Multilingual Abstract)

This paper presents a methodology for designing rotor–bearing systems that use dimensionless journal bearing data and a multiobjective genetic algorithm (MOGA). The novelty of this research lies in integrating preestablished dimensionless bearing data to enhance rotordynamic stability while minimizing power loss. We consider five types of journal bearing models in the design process: axially grooved, pressure dam, partial arc, fixed pad, and tilting pad. The preestablished bearing data include the bearing type, L/D ratio, number of pads, preload, and load direction. We calculate the dimensional values of the bearings, including the power loss and rotordynamic coefficients, by considering the rotor geometry, bearing load, operating conditions, and lubricant properties. We incorporate these coefficients into the rotordynamic analysis of the rotor–bearing system. We evaluate the rotordynamic stability based on the amplification factor, separation margin, unbalanced response, and logarithmic decrement. Finally, we define the optimization problem with two objective functions: improving rotordynamic stability and minimizing power loss. The resulting Pareto front reveals a trade-off between the two objectives. Despite exploring only 0.67% of the entire design space, the proposed MOGA-based approach effectively identifies optimal bearing configurations, significantly enhancing operational efficiency and system robustness.