Fraunhofer Institute for Industrial Mathematics ITWM

The microstructure accounts strongly for material properties like thermal conductivity, stiffness and strength, or acoustic absorption. 3D image analysis can provide geometric information to replace phenomenological constitutive laws for the macroscale by multiscale models taking into account both the constituent's bulk properties and their spatial structure. That way, the material's behaviour at the macroscopic scale can be predicted much more precisely.

For example, the mechanical stiffness of a glass fiber reinforced polymer differs strongly depending on whether it is loaded in fiber direction or perpendicular to the fibers. This anisotropic behaviour is captured by the fiber orientation tensor attainable from 3D images. Stiffness and strength computed as functions of the fiber orientation result in the full 4th order anisotropic stiffness tensor.

MAVI FeelMath combines image processing and material characterization. Using the established image processing techniques of MAVI, it enables you to preprocess, segment and characterize the 3D microstructure directly from micro-tomographic (µCT) reconstructions. Building on the thus created microstructure (voxel model), the robust and fast numerical solvers of FeelMath can be applied to compute thermal, elastic and acoustic material behavior and permeability of porous materials - without the need to switch to any other tool.

The following images are gained from a glass fiber-reinforced polymer specimen.

CT-reconstruction (3.5µm pixel spacing)
Segmentation of the fiber system
Von Mises stress under uniaxial loading