PhD vacancy (4 years) on composite geometry reconstruction from CT images
This project is a large collaborative project between Ghent University and several industrial partners about X-ray imaging of materials and the related post-processing of the tomographic data to extract information about geometry, damage, moisture diffusion, etc. in materials.
For the PhD position at UGent-MMS, the particular interest is to extract the internal geometry of the composite from the X-ray data and build a finite element model of the reconstructed geometry.
Almost all composite materials for engineering applications are built up from a polymer matrix, reinforced with carbon or glass fibres. Those carbon or glass fibres are embedded in the polymer matrix in the form of weaves, knits, braids, etc. Besides that, the compaction pressure during the composite manufacturing makes that the different layers of the composite are slightly interpenetrating and the fibre reinforcement is deformed. As a result, the final product has a complex micro-structure, with a three-dimensional distribution of the reinforcement architecture. Nowadays, micro-tomography is very popular to inspect the internal geometry of composites, but automatic extraction of the fibre bundle positions is very difficult, because the X-ray contrast between e.g. carbon fibre and epoxy resin is very poor, and also the direction of the fibres is very hard to distinguish. Therefore, the conversion from the X-ray images to a finite element model of the composite geometry is still manual in many cases.
In this PhD, it is the purpose to automate the complete conversion from CT-images to CAD to Finite Element model, by use of advanced image processing, virtual noise estimation and enhancement of the X-ray contrast. This PhD will be in close collaboration with the colleagues from X-ray imaging (UGCT), image processing (TELIN-IPI), geology and WoodLab, and companies such as Siemens, Agfa, Unilin, Wienerberger,…
Only candidates with a Master degree should apply. The candidate should have a strong interest in image processing and mathematics and preferably have some knowledge about composites.
To postulate -> follow this link