Development of a Multi-physical 2-D model of PEM Fuel Cell for Real-Time Control

International audienceThis paper presents a computationally efficient 2-D steady-state model for fuel cell real-time control implementation. Both the fluid and electrochemical physical domains are considered in the proposed real-time model. The fuel cell under-rib convection is fully described by considering the geometry of serpentine channel. In addition, in order to solve the implicit activation voltage loss, and further explore the computational performance, three numerical root-searching algorithms: bisection, secant and Newton-Raphson methods are respectively applied to the proposed implicit iterative solver and compared. The preferred secant method has been proven to effectively improve both the efficiency and robustness performance of the proposed real-time fuel cell model. Moreover, a CFD-based COMSOL fuel cell model is used to validate the calculation accuracy. Furthermore, the practical feasibility of the presented real-time model has been verified using a RT-LAB simulator platform from Opal-RT