Modeling of Granular Structures


GrainGeo creates digital 3D models of ceramics, sintered materials, grain packings or powders. The starting point for modelling are user-defined parameter such as known grain size distribution, pore size distribution and grain shapes. The detailed digital 3D microstructure model allows a close visual analysis otherwise only possible from a µCT-scan.

By changing the parameters underlying the model, new material structures are designed and their material properties are computed. New designs and existing products are compared and performance is optimized.

Examles of Applications

Modeling of:

  • Packed beds of spheres (and other objects) 
  • Sintered ceramics
  • Catalyst layer and microporous layer in PEM fuel cells
  • Lithium ion cathodes and other battery materials
  • Rocks like sandstone
  • Sphere packings with very high packing density

Examples of Applications

GrainGeo Features

Various states of the sintering process are modeled using morphological operations and starting from random distributions, stock piles, and packings of grains with varying shapes and shape distributions.

The resulting structure models range from powders and sintered materials, with open and closed porosity, to sintered composites without porosity. The granular objects are spheres, ellipsoids, general polyhedra with randomly distributed parameters and even fibers.

Sphere packings with complex size distributions and high packing densities, very close to the theoretical limit, are modelled with GrainGeo, and binder might be added to these structure models.

Additional modules needed?

  • The GeoDict Base package is needed for basic functionality.
  • GrainFind: computes the statistical properties on the µCT scan and thus enables to realistically model an existing material with GrainGeo
  • ExportGeo-CAD: Export of 3D microstructures modeled with GrainGeo as surface triangulation (*.slt) for CAD programs
  • ExportGeo-Abaqus und ExportGeo-Fluent: Export of 3D microstructures modeled with GrainGeo as simulations with Fluent / Abaqus