Multi Objective Design Optimization of Split Journal Bearing Using Genetic Algorithm

This paper deals with minimization objective of parameters like oil flow, temperature rise and power loss of split journal bearings used in sugar cane crushing mills of sugar manufacturing industry. Design variables like length to diameter ratio, lubricant viscosity, clearance, eccentricity ratio and journal speed are considered for design optimization. Hybrid solution scheme of journal bearings is used for constructing multi objective function. In the design of journal bearings l/d ratio, clearance, viscosity of lubricant, speed and eccentricity ratio are considered as designed variables. For design optimization, genetic algorithm with pareto optimality is applied. For multi objective optimization problems, pareto optimality eliminates complexity of selecting weighting factor of design variables. Results of pareto optimal front are tabulated and explained graphically in the paper. Design optimization results shows that optimum range clearance falls in between 208 microns to 269 microns and bounds for clearance is 200 microns to 300 microns. This justifies that larger clearance for cane mill journal bearings will reduce frictional power losses but will increase the oil flow rate. Viscosity of oil shows variation from 0.038 Pa-s to 0.040 Pa-s. A lower value of viscosity of oil is given preference for design considering minimum oil flow rate and minimum temperature rise. Other design variables like length to diameter ratio speed and eccentricity does not show any variations from lower bound to upper bound. Obtained results of optimization are the recommendations for selecting design variables of heavily loaded journal bearings of sugar cane mills