Extreme-Fast Charging Performance Optimization of Immersion-Cooled Battery Systems with Cylindrical Cells

Efficient and robust optimization frameworks are essential to develop and parametrize battery management system (BMS) controls algorithms. In such multi-physics application, the tradeoff between fast-charging performance and aging degradation needs to be solved while simultaneously preventing the onset of thermal runaway. To this end, a multi-objective optimization framework was developed for immersion-cooled battery systems that provides optimal charging rates and dielectric flowrates while minimizing aging and charging time objectives. The developed production-oriented framework consists of a fully coupled, lumped electro-thermal-aging model for cylindrical cells with core-to-surface and immersion-cooling heat transfer, the latter controlled by the dielectric fluid flowrate. The modeled core temperatures are inputs to a semi-empirical aging degradation model, in which a fast-aging solver computes the updated capacity and internal resistance over multiple timescales, which in turn affect the cell electrical response and Ohmic heat generation. All building-block models are validated using cell core/surface and fluid temperature measurements and cycle aging experiments of 21700 cells with Nickel-rich NCA chemistry. The multi-physics model is coupled to a multi-objective genetic algorithm (GA) optimizer with fast charging time taken from 0%-80% SOC and aging degradation objectives, and cell core temperatures taken as nonlinear constraints. We do not consider the cell temperature as a separate cost function since it is taken as a stress factor for the aging cost. The framework provides evolving Pareto fronts with State of Health (SOH)-dependent optimal charging current profiles and dielectric flowrates, providing a system-level controls optimality between the BMS and the thermal management unit (TMU).