Mechanics of plant fiber-reinforced composite materials : from micro–meso modeling to interactive frameworks for material-informed design in early planning phases

Abstract

Plant fiber-reinforced polymer composites (biocomposites) offer renewable, lightweight alternatives for the built environment and can help reduce embodied carbon. Their structural adoption,however, remains limited by a combination of factors, including variability in constituent properties, fiber-matrix adhesion challenges, the breadth of possible material architectures, and a scarcity of harmonized test methods and structural design guidelines. In this context, this thesis addresses two complementary needs: (i) predictive, numerical modeling of stiffness, strength,and failure, and (ii) integration of such knowledge into early design tools that lower knowledge barriers for architects and engineers. Nonlinear finite element models at various scales are developed to capture the coupled failure mechanisms of biocomposites. A two-fiber unit cell incorporates matrix softening, interface failure, and fiber rupture, reproducing nonlinear responses reported in the literature and resolving interactions among damage mechanisms. Sensitivity studies highlight the decisive role of interface behavior, motivating a second study in which microdroplet tests are simulated across wide ranges of geometric/mechanical parameters. A neural-network surrogate reproduces full load-displacement curves and enables efficient parameter studies, yielding traction-separation parameters for the interface that transfer from micro- to mesoscale (demonstrated for a flaxpolypropylenecomposite). A sketch-to-analysis framework connects stylus-based 3D input to analysis-ready models and couples it with a micromechanics module that supplies elastic stiffness and limits for varied biocomposite materials and microstructures (e.g., fiber orientation and volume fraction), providing immediate feedback in early iterations. The workflow is expanded to extended reality, and complemented by haptic feedback: a collaborative robot with a touchable plate prop provides local stiffness responses so users can physically sense differences among material/structural choices. By coupling haptic and visual feedback, the system enhances communication of structural performance and lowers the knowledge barriers for engaging with novel materials. Within the broader materials-to-structures process chain, these contributions address a subset of issues—predictive mechanics and support for early, material-informed design decisions—to aid credible consideration of biocomposites in structural applications.