Digital Analysis of Fibers and Binder Content of Four Carbon Paper GDLs
PDF tutorial
In 3D image data from µCT or FIB/SEM scans, fibers and binders often have identical gray values, making automatic separation based solely on brightness virtually impossible. However, a differentiated analysis can be achieved by using artificial intelligence to analyze the shape and geometry of the components.
The FiberFind module of the GeoDict software allows segmented scans of fiber composite materials to be analyzed by using neural networks to reliably distinguish between fibers and binders. Since version 2021, the GeoDict-AI module has enabled these networks to be created, parameterized, and applied directly in GeoDict.
In this PDF tutorial you will learn step-by-step:
- Image import & segmentation: Preparation of a 3D scan of a gas diffusion layer made of Toray carbon paper using ImportGeo-Vol.
- Binder recognition with AI: Separation of binders and fibers using a neural network.
- Fiber identification: Automatic recognition and analysis of individual fibers with another AI model in FiberFind.
- Post-processing results: Visualization and evaluation of the analyzed fiber structure for further processing or documentation.
The sample presented is commercially available Toray Paper TGP-H-030 and is used solely to illustrate the workflow – not for material evaluation.
The tutorial is intended for users who want to perform precise structural analyses of fiber composites – e.g., in the field of fuel cells, filter media, or composite materials.
Forgot your password?
Please enter your username or email address. Instructions for resetting the password will be immediately emailed to you.
Not registered yet? Register for free and enjoy all the benefits.
