Intelligent self-evolving design method of a high-load-bearing hydrostatic oil groove

Hydrostatic oil grooves in friction pairs are responsible for guiding, storing, and distributing lubricating oil and are widely applied to ultrahigh-power hydraulic motors of tunnel boring machines, aerospace variable pumps, and hydrostatic precision guideways/spindles of high-end industrial mother machines. The traditional design methods for oil groove patterns highly rely on the designer’s experience and size optimization of preset shapes, making it difficult to achieve optimal lubrication. Therefore, this study proposes the intelligent self-evolving design method of oil grooves, in which the AI-assisted generalized pattern search algorithm (GPS+AI) is designed to make an oil groove pattern self-evolve toward maximizing the load-bearing capacity according to the friction pair’s contact force feedback from a lubrication model. The designed oil groove pattern is machined onto the piston of a hydraulic motor and is experimentally evaluated for its lubrication load-bearing capacity through a homemade quasiactual roller–piston pair testing rig. Comparing two traditional oil grooves, the new oil groove can reduce the friction torque (contact force) by a maximum of 88%, which is very significant for improving the efficiency and lifespan of ultrahigh power hydraulic motors (power > 10⁶ W), especially under the dual-carbon target.