Optimal fatigue life design of double row angular contact ball bearings by an adaptive RL-BFGS algorithm

An adaptive RL-BFGS (ARL-BFGS) algorithm was proposed for fatigue life design to speed up the convergence and obtain the global optimal solution under the circumstances of fewer optimization times. Fatigue life is one of the most essential criteria for the optimal design of double row angular contact ball bearings. The contact angle was selected as a design parameter besides the basic geometric parameters. The design constraints considering the manufacturing and mounting situations were processed by a penalty function. Three different constraint non-linear optimization models were established for the optimal dynamic and static load capacity, and their weighted form. The bearing model 3210 was optimized successfully to prove the correctness and effectiveness of the proposed algorithm. The overall performance of the ARL-BFGS algorithm was checked by the comparative experiments of different optimization methods and different bearing models. The result showed that the dynamic load capacity and static load capacity of the optimized bearing series 32 are approximately 60% and 30% higher than the standard value in Rolling Bearing Handbook by using the ARL-BFGS algorithm, respectively. The weighted form of the dynamic and static load capacity was also optimized to provide more selection for designers.