Back to overview

Lecture

The potentialities in image processing by the example of the characterization and analysis of X-ray-µ-Tomography data of porous ceramics

Wednesday (07.10.2020)
13:20 - 13:40

Over the last decade micro-computed tomography (µCT) has established as an important method for characterization in material and engineering sciences as well as in fundamental research. By implementing this imaging technique, it is possible to obtain non-invasive high-resolution data for structural elucidation. In particular the terms of porous ceramics, which are composed of an over several orders of magnitude complex internal structure, the characterization places high requirements on the used measurement technology as well as on the digital image processing methods (DIP).

The downstreamed data processing of tomographic images consists of several steps such as image filtering, binarization and segmentation. An optimal accomplished image processing chain enables the analysis of porosity, pore and window size distribution and the analysis of the inner as well as of the outer surface. To ensure a reliable analysis, noise and measuring artefacts such as ring artefacts or beam hardening artefacts have to be eliminated by appropriate methods. Due to that the image quality can be improved insofar as the fulfilment of quantitative analysis are enabled.

A more sophisticated structure is represented by freeze casting ceramics. The special about the structure of freeze casted ceramic material is that they are composed of a highly branched, skeleton like, pore distribution. For that reason, to gain a reliable structure characterization, abstract and precise data are an indispensable presupposition.

An even more advanced application of µCT is the analysis of complex processes such as deep bed filtration. It is possible to obtain highly resolved images of the filter structure as well as on the particle deposition sites. By means of DIP typical process characteristics such as filter efficiency, particle deposition rate, deposition profiles (axial, radial) up to quantitative evaluation of particle deposition can be obtained. Furthermore, the topological information of filter structure and information on colloid depositions obtained by X-ray µ CT can be supplemented by the flow field obtained by means of magnetic resonance velocimetry (MRV). The combination of these two imaging methods allows conclusions to transport and retention mechanisms, which depend significantly on the local velocities.

Speaker:
Additional Authors:
  • Thomas Ilzig
    Technical University of Dresden
  • Dr. Gerd Mikolajczyk
  • Dr. Stefan Günther
    Technical University of Dresden
  • Prof. Dr. Stefan Odenbach
    Technical University of Dresden