Url https://www.cimne.com/sgp/rtd/Project.aspx?id=847
LogoEntFinanc LogoPrograma
Project title Advanced Multi-scAle moDEling of coupled mass transport for improving water management in fUel cellS
Reference PGC2018-101655-B-I00
Principal investigator Pavel RYZHAKOV - pryzhakov@cimne.upc.edu
Riccardo ROSSI - rrossi@cimne.upc.edu
Start date 01/01/2019 End date 31/12/2021
Coordinator CIMNE
Consortium members
Program P.E. Generación de Conocimiento y Fortalecimiento Científico y Tecnológico del Sistema de I+D+i Call Proyectos de I+D+i Generación de Conocimiento 2018
Subprogram Proyectos de I+D: Generación de Conocimiento Category Nacional
Funding body(ies) MCIU Grant 107.690,01 €
Abstract Polymer Electrolyte Fuel Cells (PEFCs) are a promising versatile energy-conversion technology suitable for applications ranging from mobile phones to emergency power supplies. Most importantly, PEFC is one of the leading candidates to replace internal combustion engines for vehicles. PEFCs fueled with hydrogen are ecological (the by-product is water) and their efficiency is higher than that of high- temperature combustion devices. Nevertheless, high cost and limited durability still preclude the PEFC from large-scale commercialization. One of the main factors for efficiency losses and premature degradation of PEFCs is the excess of liquid water in the cell. By-product water must be transported from the reaction site through the diffusion layer and evacuated in the gas channels by the airflow. Excessive water accumulation diminishes the concentration of reactants at the reaction site, while complete flooding can cause permanent degradation of the cell. Therefore, predicting and controlling coupled liquid-gas transport is a decisive factor for improving fuel cells' design. Reducing corresponding losses would permit the cells to be operated at high voltage and high current enabling an increase of specific power density, reducing the size (and thereby the cost) of the device. The objective of this project is the development of a multi-scale numerical tool for analysis of two-phase (liquid-gas) transport in PEFCs. To achieve this goal several ingredients will be developed and coupled. First, a robust an efficient model for macroscopic problem of liquid interacting with the airflow in gas channels will be developed. This will be done in the embedded finite element Eulerian-Lagrangian framework. The model will account for contact of the liquid phase with the gas channel walls of different roughness and wetting characteristics, as well as the topological changes. Novel algorithms will be developed for the efficient numerical solution of coupled governing equations. Second, a model for analysis of the microscopic flow through the diffusion media will be developed together with a series of numerical small-scale experiments. These will be done using high- resolution images provided by our partner institution, EDSL, Canada. The results will serve for estimating the effective transport properties of the diffusion media using machine learning techniques. Ultimately, the macroscopic model for the gas channel and the diffusion media will be coupled. A methodology for translating the results obtained from the diffusion media simulation into the boundary conditions (sizes of the liquid injection sites and injection rates) for the gas channel problem will be established. Thus, a multi-scale transient model for the two-phase transport problem in the fuel cell will be obtained. It will be implemented as an open source code, facilitating its dissemination and use by scientific and industrial community. The model will allow assessing fuel cell designs with respect to losses and failure risk due to two-phase transport. It will provide liquid-gas flow patterns and water distributions as a function of a) operation conditions (water production rates, airflow velocity) b) flow channel geometry c) diffusion media design d) material properties. Successful execution of the project will result in a novel virtual design tool for fuel cells analysts and will define a new step in multi-phase modeling research.