Infoscience

Conference paper

Multi-scale modeling methodology for computer aided design of a Solid Oxide fuel cell stack

This paper presents the modeling strategy developed for the design of a planar solid oxide fuel cell repeat element. Design goal are to reach good performance (power density) and reliability (ie. limit risk of failure and degradation). This challenging problem, involving multiple physical phenomena, different scales (from the pores of the catalyst to the interaction of the fuel cell with the system), is addressed with a multi-scale modeling approach. Handling the complexity of the problem (including kinetics, mass, species and energy balances for fluids and solid parts) and the complete geometry may lead to unmanageable models in terms of complexity and CPU time. Therefore, simplified models are needed to understand the behavior and give the trade-off between the conflictive design objectives. The general methodology is presented as well as the different models features. A model for the routine electrochemical experiment allows to identified parameters for the kinetics (which is a key parameter for behavior and performance prediction). A simplified 2D model for the repeat element allows to easily explore and compare different configurations, make sensitivity studies and optimize some key design parameters. Finally a more complete CFD-based model is used to validate the decisions and options identified with the simplified model. This approach makes possible to explore different configuration and use optimization tools for the design of a repeat element. As mesh generation is the main bottleneck in CFD modeling, this method uses this tool only at the end of the process to validate the previous models used and decisions, therefore only a couple of mesh have to be generated. One of the key problem is here to make sure the simplified model and CFD model give the same trends and comparable results. This approach can be extended to the interaction of the system with the stack.

Fulltext

  • There is no available fulltext. Please contact the lab or the authors.

Related material