Despite many attempts to model effective behavior, there is currently no recognized approach that can capture the most important aspects of deformation of 3D textile reinforcements during their processing, and predict efficiently both the macroscopic response of the textile structure in the dry state or as pre-impregnated from the behavior of the fibers or yarns at the smaller scales. Recent developments in multi-scale simulation and 3D imaging techniques in materials science, particularly X-ray microtomography combined with appropriate image analysis techniques, make it possible to finely analyze the micro-mechanisms of deformation at the level of interactions between fibers and to enrich the interpretation of micro or meso-mechanical tests, which opens up new ways for the exploration and the understanding of the phenomena occurring at this level, in particular for elaborating and identifying models at intermediate scales, essential for an in-depth prediction of macroscopic behavior. The general objective of the project is the development of constitutive laws with an enriched kinematics for dry and pre-impregnated 3D technical textile reinforcements at different scales, which integrate the geometry of the constituents identified by X microtomography, the rheology of fibers and yarns, geometric and structural nonlinearities, singular and dissipative phenomena related to the presence of defects, irregularities in behavior (contact, friction, microcracking), as well as a statistical variability of the geometry and mechanical properties of the yarns or fibers within the armor. These aspects are scientific locks that define the innovative nature of the project compared to the literature works. The project is multidisciplinary since it concerns the science of fibrous materials, the mechanics of discrete and continuous media, multiscale higher order homogenization methods, rheology, stochastic methods, microtomography image analysis techniques, and numerical methods.