Lattice Models for the Prediction of Load-Induced Failure and Damage in Wood

Abstract

Lattice models, which are essentially networks of springs having variable strength and stiffness, are used to predict the response of structural softwood to various mechanical loadings. The use of lattice models is motivated by the desire to mimic the morphology of wood, and thus realistically predict both load-displacement response and experimentally observed damage patterns that are not well captured by conventional continuum-based models. The goal of the work described in this paper was to evaluate suitability of lattice models for predicting several different failure modes of structural softwood, and to establish a methodology for matching model parameters to measured material properties. Using least squares techniques, the lattice model member properties were determined to give optimal predictions of bulk elastic constants and expected load-displacement response to failure under perpendicular-to-grain tension, parallel-to-grain tension, and shear. The lattice models were shown to accurately capture load-displacement response, including experimentally observed strain softening under perpendicular-to-grain tension. In addition, the lattice models successfully predicted damage localization, including crack bridging and microcracking around the critical crack. Mesh size effects were also examined, and while the lattice models are dependent on mesh size, it is demonstrated that for a fixed cell aspect ratio, a simple scaling of the lattice model strength and stiffness properties to account for cell size can yield reasonable predictions of ultimate capacity.