Multi-scale and nonlinear simulation methods for cellulose materials and applications to paper

Paper is a sustainable and versatile material, yet its mechanical behavior (how it deforms and fractures) is complex. This complexity arises because paper is not uniform; it is composed of countless microscopic cellulose fibers with varying orientations and bonding conditions. Understanding how these fibers interact to carry loads and resist fracture is essential for improving paper-based products and developing greener alternatives to plastics. This thesis presents an integrated experimental and modeling approach that connects the microscopic behavior of fibers with the overall mechanical response of paper and paperboard. At the microscale, specialized experiments and imaging techniques were used to observe how fibers stretch, bend, and separate during loading. These observations guided the development of a stochastic micromechanical model capable of describing the random fiber network and predicting its nonlinear response and fracture behavior. Building on this foundation, two continuum-scale models were proposed to predict the behavior of paper under complex loading conditions while preserving a clear link to the fiber-scale mechanisms. The same framework was extended to reinforced gummed papers, demonstrating that natural fibers such as flax can effectively replace synthetic glass fibers as sustainable reinforcements. Finally, experiments on paperboard with different fiber orientations confirmed the strong connection between material anisotropy and fracture direction, validating the predictive capability of the developed models