Three-dimensional multiphase simulation and multi-objective optimization of PEM fuel cells degradation under automotive cyclic loads

• Published in: Energy conversion and management 2021-03, Vol.231, p.113837, Article 113837
• Main Authors: Moein-Jahromi, M., Kermani, M.J.
• Format: Article
• Language: English
• Subjects: Automobiles; Automotive fuels; Commercialization; Computer applications; Conservation equations; Cost function; Cycles; Cyclic loads; Dynamic loads; Electric vehicles; Electrochemical surface area loss; Fuel cell degradation; Fuel cells; Fuel technology; Load cycling; Mathematical models; Multiphase; Multiple objective analysis; Operating temperature; Optimization; Parameters; Performance degradation; Performance optimization; Power density loss; Proton exchange membrane fuel cells; Simulation
• ISSN: 0196-8904; 1879-2227
• DOI: 10.1016/j.enconman.2021.113837

•New combination of 3D-CFD simulation and degradation model for PEMFC is developed.•A new set of single/multi-objective optimization is planned through four scenarios.•The degradation rate is minimized as well as the initial performance is maximized.•Temperature is the most effective parameter on aging rate and initial performance.•The optimization results show an improvement of 36% in the final power density. One of the remaining bottlenecks of PEM fuel cell vehicle commercialization as a probable alternative to conventional vehicles is the performance degradation during dynamic loads. Herein, an innovative coupling is presented between a three-dimensional, multiphase computational fluid dynamic simulation with eight conservation equations and a novel degradation model to predict the performance loss of PEM fuel cell under vehicular load cycling. Moreover, a multi-objective optimization problem with four different scenarios is also planned for the first time as the other novelty, to minimize the cell power density loss as well as maximize the initial cell power density, in order to find the optimum value of some operating and structural parameters. The model predicts the power density degradation rate of about 0.001627 kW cycle−1 (equivalent to 4.067 kW m−2 loss after 2500 cycles) which is in good agreement with experimental data. The results reveal that the operating temperature is the most influential parameter with rank 1 for the cost functions. The optimization results also show a considerable enhancement of about 36.9% in the final power density after load cycling compared to the base case conditions by fine-tuning five operating and structural parameters.