A novel numerical implementation of electrochemical-thermal battery model for electrified powertrains with conserved spherical diffusion and high efficiency

Abstract

The performance of batteries in electrified powertrain systems is highly influenced by mass diffusion and electrochemistry which are often ignored in the simulation of these systems due to the lack of a conserved, efficient, and integrable battery model. Therefore, this work numerically implements an electrochemical-thermal battery model with conserved numerical schemes and efficient numerical methods which include Jacobian-based and Jacobian-Free Newton Krylov (JFNK) solvers. The performance of the developed model is evaluated by simulating measurements of a LiFePO4 battery under constant discharge rates and Urban Dynamometer Driving Schedule (UDDS), as well as by a detailed comparison with existing battery models. The comparison highlights two features of our model: (a) negligible mass imbalances in the spherical diffusion modelling, which are five orders of magnitude smaller than those from a recent battery model in the literature; (b) efficient modelling of real-world driving cycles with the computational time two orders of magnitude shorter than that of the literature model. These advanced features indicate that our model can be applied in both fundamental electrochemical-thermal studies of lithium-ion battery and detailed simulations of electrified powertrains as an accurate and efficient sub-model.