Oil transport within the piston ring pack plays a major role in oil consumption and therefore it’s necessary to develop a thorough understanding of the relevant mechanisms. However, this is not an easy task as the physics involved spans across multiple length and time scales. Engine experiments are often time-consuming and only provide qualitative but not quantitative understanding of the multiple physics involved, and many of the physics cannot be measured easily. The digital twin model is developed as a complement to engine experiments by serving as a tool for quick oil transport simulation and oil consumption prediction given a certain design. The overall framework consists of multiple sub-models for individual physics that can directly or indirectly influence oil transport:
- Finite element model for the structural dynamics of the piston rings
- Linear mass conservation model for zonal pressure gradient
- Hybrid machine learning and analytical models for complex gas flow patterns in the piston ring pack
- Two-phase gas and oil interaction model for oil redistribution and oil transport
- Piston ring pumping model
- Piston ring rotation model
- Liner evaporation model
By decoupling the various physics while carefully ensuring the coupling between the submodels, the digital twin model strikes a good balance between physical accuracy and computational efficiency. Compared to experiments, it offers more insights into how each mechanisms contribute to the overall oil transport pattern. The ultimate goal of the digital twin is to be used for optimizing the design of a piston ring pack such that proper lubrication and low oil consumption can be achieved.
Example study:
- Top left: piston ring conformability and ring-liner contact force
- Top right: CFD on vortex pattern around top ring gap during engine blowby
- Bottom right: machine learning framework for flow field prediction
