Modeling damage and its influence on material anisotropy and swelling behavior in soft composites

Soft materials, such as rubber-like elastomers and hydrogels, are playing increasingly important roles in emerging technologies, such as soft actuators, drug delivery systems, and flexible sensors. However, due to diverse internal microstructures and complex loading histories, these materials often exhibit initial anisotropy or damage-induced anisotropy. Given that soft materials span multiple scales, from micro- to meso- and macroscales, non-affine deformation mechanisms are commonly introduced in theoretical modeling. Despite significant progress, generalized and computationally efficient constitutive models capable of capturing anisotropic behaviors are still limited. (I) To address this gap, this dissertation first reviews the analytical network-averaging homogenization method, which introduces closed-form kinematic measures to represent the average behavior of fibers. The proposed model successfully captures anisotropic responses across a range of soft materials, verified through comparison with experimental data. (II) For engineering applications, reinforcement by fillers or additional networks are used to enhance mechanical properties but introduces anisotropic Mullins effects. Based on mesoscopic measures, an anisotropic damage model is proposed, idealizing the network as a continuous fiber distribution. Directional damage evolves based on an energy-based damage potential with a stored energy limit. The model accurately predicts uniaxial and biaxial responses. Additionally, to capture volumetric changes in materials like hydrogels, a compressible anisotropic model is formulated via a modified bulk modulus approach. (III) Hydrogels exhibit great promise in stimuli-responsive devices; however, the reswelling behavior following unavoidable internal damage remains poorly understood. Considering that hydrogels are inherently polymer-solvent mixtures, a novel theoretical framework is proposed by taking the taken-out state as the reference configuration. A damage-induced reswelling formulation is derived analytically. To further capture complex geometries and boundary conditions, a thermodynamically consistent model coupling finite deformation with fluid transport is established. The finite element implementation is carried out and several application scenarios are simulated to illustrate the significant influence of damage on the reswelling response of hydrogel devices. (IV) In summary, this dissertation develops a unified anisotropic damage model for soft materials based on analytical network-averaging, with applications extending from anisotropic stress-softening phenomena to damage-induced anisotropic reswelling. The proposed models are validated against experimental data, demonstrating the robustness, accuracy, and versatility of the proposed framework across various classes of soft materials.