|
SS22 WS22 SS23 WS23 SS24 WS24 Guidelines for Master Students
Materials Properties from Microstructural based Experimental-Computational Synergy Video: https://www.dropbox.com/s/c9dq24zw6c1kt26/Ukrainczyk_uCTSeniorScientistPosition.mp4?dl=0 The porous microstructure of a cement-based material cover multiple length scales, ranging from nanometer-sized gel pores inside hydration products, via micrometer-sized capillary pores in-between the reactive cement grains, towards milimeter-sized air bubbles. A numerical scheme is used by means of either generating a porous network from advanced 3D numerical simulations (e.g. Hymostruc), or, by 3D sampling a porous network inside real cementitious samples, using experimental imaging via X-ray computed tomography. The algorithm starts with a discretization of the microstructure into a regular 3D mesh (stack of bitmaps), where each voxel in the mesh is assigned to a single phase. An assembled system of up to 1 billion equations are solved implicitly by a conjugate gradient algorithm employing e.g. an in-House developed OpenMP (shared memory) parallelized C-code. A novel method is proposed that mitigates resolution problems for numerical methods to calculate the effective properties of the multi-scale porous cementitious materials. The method refines initial mesh resolutions and includes the detailed sub-voxel into the transport calculation with a high accuracy and efficiency. The proposed method is validated against the effective diffusivity of a simple periodic arrangement of mono-sized particles, as well as against a number of complex 3D virtual cement paste microstructures. The method leads to very accurate results, significantly reduces computational efforts, and describes the porous network in a high level of detail. A new approach to calculate the effective diffusivity of cement pastes by considering a linear variation of the pore phase diffusivity with a grayscale level of the pore voxel that corresponds to the experimentally scanned CT image and that represents an intermediate level of diffusivity. |