PEMFC

The challenge: Increasing the lifetime of a cell and reducing material costs

The proton exchange membrane fuel cell (PEMFC) is the beacon of hope for the mobility of the future.

Valuable properties of fuel cells, such as the high efficiency, compared to combustion engines, as well as the compact design or its good cold-start behavior, make a widespread application of the low-temperature fuel cell highly desirable in local passenger transport or in shipping, for example. The PEMFC is also interesting for self-sufficient electronic systems or for combined heat and power generation.

The greatest challenges in the development of fuel cells materials nowadays are in the handling of

  • gas permeability and water management
  • thermal and electrical conductivity
  • the operation at elevated temperature (350 K)
  • mechanical stress and aging

together with the reduction of material costs.

Highlights of the GeoDict solution at a glance

  • AI-assisted identification of binders and individual fibers and particles in 3D-image data 
  • Generation of realistic statistical Digital Twins of the materials at the microscale
  • Determination of important geometric and physical parameters
  • Quantitative and high-performance simulation of physical properties 
  • Automated parameter studies for the design of new fuel cell materials

Please note that after activating the video, data will be transmitted to YouTube. 
More Information

Digital development of PEMFC materials with GeoDict

Whether for GDL, MPL, CAT, PEM, BPP, or even a complete cell: Analyze, understand, create and optimize the material at the microscale. Math2Market brings to you the digitalization of Research and Revelopment of fuel cell materials with the GeoDict software.

Image Processing & Image Analysis

3D image data can be imported, segmented, and processed with GeoDict. The latest AI technology, based on artificial neural networks (ANN), enables the identification of all material constituents and all objects. Individual fibers can thus be recognized, analyzed, and characterized.


Analysis & Characterization

GeoDict determines the defining geometric and physical parameters of real or artificially created microstructures: porosity, grain and pore size distribution, active surface area, tortuosity, Gurley value, etc.


Modeling & Design

Through the parameters obtained in the in-depth analysis of constituents and objects, the structure generators in GeoDict create realistic Digital Twins of the materials. Random generators are used to create statistical variance and thus, to minimize statistical errors in the simulations. The Statistical Digital Twin is then created using the GeoDict material database to assign the corresponding material properties to all constituent materials.


Simulation & Prediction

GeoDict includes several high-performance, validated numerical solvers to simulate the relative and effective physical properties of a real or artificial microstructures. Among others: diffusivity, permeability, thermal and electrical conductivity, two-phase flow, capillary pressure curve, mechanical deformation, particle advection and diffusion, etc.


Automation & Interfaces

GeoPy, the Python-based automation in GeoDict enables the coupling of simulations and work steps for extensive parameter studies. Even complex issues are so digitally processed, such as the optimization of transport properties under mechanical compression.


The GeoDict solution for fuel cell Research & Development

The software package consists of all necessary GeoDict modules for the research and development of fuel cell materials, including GeoDict Base.

Module Recommendations

Image Processing & Image AnalysisImportGeo-Vol     
Characterization & AnalysisGrainFind-AIFiberFind-AIPoroDict + MatDict   
Modeling & DesignGrainGeoFiberGeoWeaveGeo   
Simulation & PredictionDiffuDictConductoDictFlowDictElastoDictAddiDictSatuDict

The modules that best fit your needs depend on your specific application in fuel cell Research & Development. Contact us for more information on your project.