Dissertation, RWTH Aachen University, 2019; Dissertation, RWTH Aachen University, 2019, The quantitative understanding of grain-scale deformation mechanisms in clay is important in hydrocarbon and water exploration as well as in the evaluation of clay formations as potential host rock for long-term underground repositories for high-level radioactive waste. Analyses on the nano-scale allow understanding the underlying microphysical processes and therefore form the basis for microphysics based constitutive laws for transport and deformation, which can be confidently extrapolated to conditions outside those in experiments. A microphysical understanding of the transport and deformation properties is important, because the properties of Boom Clay are known to be complex: deformation is anisotropic, after a certain amount of strain deformation tends to localize, and the transport properties are also expected to be dependent on deformation. In addition, one of the favourable properties of Boom Clay is that fractures have the tendency to self-seal, but a microphysical understanding of this process is so far not available. This study examines the development of microstructure during triaxial tests on Boom Clay freshly collected at HADES level with σ1 applied parallel and perpendicular to the bedding, to various total axial strains. We used a range of methods integrated in a multi-scale analysis. First, the samples were saturated and deformed in consolidated-undrained (CU) triaxial tests starting at 2.2 MPa effective stress, combined with in-situ micro computed tomography (μ-CT). The μ-CT data with a resolution of 13.5 μm/pixel were analysed by 3D digital image correlation (DIC) to compute the incremental displacement fields, and the evolution of the strain field in the sample. Deformed samples were slowly dried, sectioned, and the microstructure studied by optical and scanning electron microscopic (SEM) imaging with resolutions down to a few nanometres. The stress-strain curves our experiments are in good agreement with previous studies by Coll (2005); Sultan et al. (2010); Deng et al. (2011b); Bésuelle et al. (2014). The orientations of the shear zones (SZ) with respect to the shortening direction are 40 to 45°, in reasonable agreement with published values of friction angles. The behaviour is slightly anisotropic. The initial pore water pressure increase Δu of samples shortened perpendicular to the bedding (S⊥B) is higher than the values measured in samples deformed with stress parallel to the bedding (S ‖ B), which is in agreement with what is expected from microstructure. DIC analysis shows that the evolution of strain is also different in S⊥B and S ‖ B samples, although most samples localize the strain at about 2% axial shortening. From this point on, more and more of the axial shortening is taken up by movements along the SZ and the distributed strain in the sample decreases. Non-localized strain in samples S⊥B is highest (up to 3 %) in cone shaped zones close to the top and bottom of the samples. In samples S ‖ B, strain is more homogeneously distributed prior to strain localization, which occurs at slightly higher axial strain than in S⊥B. Because of the prominence of the evolving SZ, the shear strength at high shear strains of S⊥B and S ‖ B is similar. Based on the DIC maps, we selected representative regions with different styles of deformation for imaging at high resolution to understand the evolution of microstructure and porosity. We defined four different structural domains:1. OSZ: shear strain < 3 %, non-localized deformation and a microstructure comparable to undeformed samples;2. OSZ-HS: higher shear strain than in OSZ (≥ 3 %), but still non-localized deformation, in S⊥B: Microstructure comparable to undeformed samples, in S ‖ B: Microstructure characterized by numerous micro-kinks and -folds; in both, S⊥B and S k B, OSZ-HS are present in significant parts of the samples more so in S ‖ B;3.OSZ-TZ: S k B at both boundaries of the SZ, shear strain comparable to OSZ-HS (≥ 3 %), non-localized deformation with a strongly altered microstructure characterized by micro-kinks;4.ISZ: SZ with a shear strain between 5 and 50, strongly localized deformation with a shape preferred orientation (SPO) of elongated grains, reduction of porosity and pores parallel to the shearing direction. Evolved SZ in S ‖ B always have a kink-zone surrounding them. The thickness of SZ varies between 20 and 200 μm, increasing with increasing shear strain. The internal structure of ISZ is similar in S ‖ B and S⊥B. Microstructures show evidence for frictional/granular deformation mechanisms (grain rotation, grain sliding, pore collapse and reorientation, mica grain bending) and no evidences for cataclastic processes. This is in agreement with microstructure which contains about 20 vol.% silt-size hard grains (calcite, feldspar and quartz) embedded in a highly porous clay matrix. It is also generally observed that quartz and feldspar grains in SZ are much smaller than in the bulk of Boom Clay. This is not because they were fragmented to smaller pieces, but because the SZ develop in regions where are no large grains. This study provides a microstructural basis for the construction of a microphysics-based model for the deformation of Boom Clay, which forms the basis for a microphysics based constitutive law, which can be extrapolated to conditions outside those used in our experiments.