(Invited) Open-Source Macro- and Meso-Scale Modeling of PEFCs

Since the early polymer electrolyte fuel cell (PEFC) mathematical modeling studies of Springer and Bernardi and Verbrugge [1], a variety of steady-state and transient PEFC models have emerged in the literature, see [2] and references within. Fuel cell modeling results have highlighted the importance of multi-dimensional effects (e.g., dead-zones under the land area [3]), coupled physical processes (e.g., heat-pipe effects [4]), liquid water accumulation, and interfacial effects (e.g., water/channel/porous media interactions [5]). Even though substantial progress has been made in developing meso- and macro-scale models, a generally accepted PEFC model does not yet exist, even for dry conditions. The absence of such model is due to: a) multiple scales under consideration, i.e., from nanometers in the catalyst layer to millimeters in the channel; b) the complex heterogeneous materials used; and, c) the coupled physical processes occurring inside the PEFC. The small number of experimental validation techniques used, and the limited number of comparative numerical modeling studies has also contributed to the disagreement between modeling approaches. A common platform for numerical modeling of fuel cells is essential in order to integrate new models, compare and assess the validity of existing ones on a one-to-one basis, and eventually, to perform fuel cell design and optimization studies. Our research group has developed an open-source mathematical modeling platform for the analysis of polymer electrolyte fuel cells, i.e., OpenFCST [6]. This modeling platform has been implemented using object-oriented concepts in order to develop a toolbox with multitude of physical models, such as multi-step reaction kinetics, gas, liquid and ion transport equations. It is also capable of handling various theoretical representations of the main layers in a fuel cells, for example, the catalyst layer can be analyzed with either a macro-homogeneous or an agglomerate model. Users can easily modify the mathematical formulation for an individual physical model while maintaining all other models and parameters the same. In this talk, the general architecture of OpenFCST and the new models implemented since its first release in 2013 [6] will be discussed. OpenFCST will be used to analyze several recently proposed mathematical models for fuel cells. Critical to improving PEFC mathematical models is the development of: 1) methodologies to estimate macro-scale reaction and transport properties based on meso- and micro-scale information of the heterogeneous materials, such as micro-scale imaging, electrode kinetic mechanisms, and pore-size distributions; 2) accurate transport models for multi-component gas transport, and multi-phase flow; and, 3) improved models for the interfaces between layers in the membrane electrode assembly (MEA) and also between the MEA and the channel. In this talk, new advances in micro-scale simulations using FIB-SEM data in OpenFCST that include voxel meshing and a morphological image opening algorithms will be presented and used to estimate effective transport properties of CL under dry and wet conditions [8]. A multi-dimensional, two-phase, non-isothermal membrane electrode assembly model in OpenFCST that includes a dual wettability pore-size distribution model will be described and compared to: a) experimental performance, water fluxes and temperature and saturation distributions, and b) to the common saturation-based, two-phase flow transport model in porous media. The pore-size distribution model is shown to be able to identify when liquid water would start leaving the cell, and possible discontinuities in the saturation profiles inside the MEA. A key issue in two-phase flow models is the channel/GDL interface. A two-phase flow finite element formulation that uses two independent meshes for air and liquid will also be described and used to study droplet dynamics in a channel [5]. References [1] T.E. Springer et al., J. Electrochem. Soc., 1991, 138 , 2334; D.M. Bernardi and M.W. Verbrugge, J. Electrochem. Soc., 1992, 139:2477. [2] A.Z. Weber et al., J. Electrochem. Soc, 2014, 161 (12), F1254-F1299; C.Y. Wang, Chem. Rev., 20 0 4, 104, 4727; and, A.Z. Weber and J. Newman, Chem Rev., 2004, 104(10), 4679-726. [3]A.A. Kulikovsky et al., J. Electrochem. Soc., 1999, 146(11) 3981-3991. [4] A.Z. Weber, Electrochimica Acta, 2008, 53(26), 7668–7674. [5] A. Jarauta et al., Journal of Power Sources, 2016, 323, 201-212. [6] M. Secanell et al., ECS Transactions, 201 4, 64 (3), 655-680 and www.openfcst.org. [7] M. Sabharwal et al., Fuel Cells, 2016, 1615-6854, http://dx.doi.org/10.1002/fuce.201600008 . [8] J. Zhou et al., J. Electrochem. Soc.,2016 (under review). Figure 1

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.