3D Modeling of Microstructures with Stochastic Geometry
In mineral processing and metallurgy, already comprehensive thermodynamic knowledge makes chemical reactions relatively well predictable. On the other hand, mechanical and physico-chemical separation processes, such as grinding, flotation or magnetic separation, can often only be reconstructed retrospectively from experiments.
Amongst others, particle microstructures which are responsible for particle interactions are yet not fully understood. Analysis methods such as the tomography, which are used to characterize the particle structure, do not achieve sufficient phase contrast in order to obtain precise mineralogical information. Other methods, such as the Mineral Liberation Analysis (MLA) or hyperspectral imaging, only produce 2D images that provide ony limited information about the true 3D material properties.
In order to investigate the influence of microstructures quantitatively and to be able to combine 2D measurements with 3D properties, the department develops 3D microstructure models but also data evaluation tools and statistical estimation theories, which aim to model particle-based processes.
Simulated mckrostructur: 1.Stochastic microstructur model (left); 2.Inner micro structur consisting of polyhedral cells (center); 3.Mosaic devided in 2D sections for computing geometric parameters as well as the display of fractured particles for the processing simulation (right), Photo: HZDR/ Peter Menzel
Selected Publications
- Teichmann, J., Menzel, P., Heinig, T., van den Boogaart, K. G
"Modeling and fitting mineral microstructures by multinary random fields.", Beitrag zu Proceedings, 18th Annual Conference IAMG, 2017
- Teichmann, J., van den Boogaart, K.G.
"Efficient simulation of stationary multivariate Gaussian random fields with given cross-covariance ", Applied Mathematics, 2016
DOI-Link: 10.4236/am.2016.717174
- Baaske, M.; Ballani, F.; van den Boogaart, K. G.
"A quasi-likelihood approach to parameter estimation for simulatable statistical models", Image Analysis & Stereology, 2014
DOI-Link: 10.5566/ias.v33.p107-119
