Modeling the mechanical performance of natural fiber composites

Abstract: Due to environmental concerns the interest in use of renewable and recyclable materials has dramatically increased over the recent years. Wood and other lignocellulosic fiber reinforced polymers have large potential as structural materials due to the high specific stiffness, high specific strength and high aspect ratio of the fibers. Composites made from wood fiber mats from paper production are also interesting from an economical point of view. In present time the limited use of cellulosic fiber composites in structural design is predominantly associated with disadvantages such as dimensional instability in humid environments, lack of well defined fiber properties and the fibers low ability to adhere to common matrix materials for efficient stress transfer. A better understanding of dimensional stability and both long term and short term mechanical performance of cellulose fiber composites is necessary if these materials are to reach their full potential. The objective of the work presented in this thesis is twofold: (i) to present material models and suitable data reduction methodology with the ambition to characterize these materials very complex time dependent behavior (Paper A and B) and (ii) to develop micromechanical models that can be used in parametric studies of fiber properties and their influence on composite properties (Paper C-E). In Paper A the nonlinear viscoelastic behavior of flax/polypropylene composites was characterized using different forms of the creep compliance. The viscoplastic behavior was described using a nonlinear function with respect to time and stress. In Paper B hemp/lignin composites were characterized in terms of nonlinear viscoelastic behavior using Prony series form of creep compliance. The viscoplastic behavior was described using the same nonlinear function as in Paper A. The presented material model also included a stiffness degradation function based on previous strain history. An incremental form of the constitutive model was used to simulate the material behavior in loading and unloading ramps and validated through experiments. In Paper C the effect of wood fiber anisotropy and their geometrical features on wood fiber composite stiffness was analyzed. An analytical model for an N-phased concentric cylinder assembly with orthotropic properties of constituents was developed and used. The model is a straightforward generalization of Hashin's concentric cylinder assembly model and Christensen's generalized self-consistent approach. In Paper D the same concentric cylinder model was used and extended to include also free hygroexpansion terms in the elastic stress-strain relationship. The hygroelastic properties on three levels were calculated. Using material data for the wood polymers available from literature the swelling characteristics on the (i) ultrastructural level, i.e. the microfibril unit cell was determined; (ii) the hygroexpansion coefficients of the fiber cell wall layers were determined and finally (iii) the hygroexpansion coefficients of an aligned wood fiber composite were calculated. In Paper E the influence of helical fiber structure on composite properties was evaluated. The fibers helical structure leads to an extension-twist coupling and thus a free fiber will deform axially and also rotate upon loading in longitudinal fiber axis direction. Within the composite the fiber rotation will be restricted however. Therefore, the decision was to compare the elastic properties in two extreme cases on both fiber- and composite level: (i) free rotation and (ii) no rotation of the layers in the cylinder assembly.