Supervised Adaptive Method for Robust and Fast Solving of Lubrication Problems

Xiaolong Zhang,Chao Zhang,Kou Du,Bing Xu,Junhui Zhang
DOI: https://doi.org/10.1016/j.triboint.2024.110191
IF: 5.62
2024-01-01
Tribology International
Abstract:The supervised adaptive method (SAM) is introduced to rapidly and robustly solve lubrication problems of conformal friction pairs, especially under extremely severe mixed lubrication with strong nonlinearity and ill-conditioning in equilibrium equations. First, the SAM adopts the rewritten form of equilibrium equations, forced decoupling, and error truncation strategies to deal with possible ill-conditioning of the Jacobian matrix. Then, an auto-tuning strategy is designed to self-adaptively adjust the stepsizes to cope with the high nonlinearity, which also fuses multiple strategies to supervise the iteration solution. The SAM is applied to a hydraulic motor and is compared with three existing well-evaluated algorithms. The speed and convergence domains of the SAM are 2.43 and 10 times those of the moderate-performing Dogleg method, respectively.
What problem does this paper attempt to address?