In this study, trabecular architecture is considered as a cellular solid structure, including both intact and damaged bone models. ??Intact?? bone models were constructed based on ideal versions of 25, 60 and 80-year-old specimens with varying trabecular lengths and orientations to 5%, and 10% covariance of variation (COV). The models were also flipped between longer transverse and longer longitudinal trabeculae. With increasing COV of lengths and orientations of trabecular bone, the apparent modulus is linearly decreased, especially in the longer transverse trabeculae lengths. ??Damaged?? bone models were built from the 25 year old model at 5% COV of longer transverse trabeculae, and with removing trabeculae of 5% and 10% of trabecular volume in transverse and longitudinal directions, respectively, as well as in combination to total 10% and 15%. With increasing percent of trabeculae missing, the apparent modulus decreased, especially dramatically when removal was only in the transverse direction. The trabecular bone models were also connected to a cortical shell and it was found that the apparent modulus of an entire slice was increased in comparison to the modulus of trabecular bone alone. We concluded that the architecture of trabecular bone, especially both lengths and percent of trabecular missing in the longitudinal direction, significantly influences mechanical properties.

In this study, trabecular architecture is considered as a cellular solid structure, including both intact and damaged bone models. ??Intact?? bone models were constructed based on ideal versions of 25, 60 and 80-year-old specimens with varying trabecular lengths and orientations to 5%, and 10% covariance of variation (COV). The models were also flipped between longer transverse and longer longitudinal trabeculae. With increasing COV of lengths and orientations of trabecular bone, the apparent modulus is linearly decreased, especially in the longer transverse trabeculae lengths. ??Damaged?? bone models were built from the 25 year old model at 5% COV of longer transverse trabeculae, and with removing trabeculae of 5% and 10% of trabecular volume in transverse and longitudinal directions, respectively, as well as in combination to total 10% and 15%. With increasing percent of trabeculae missing, the apparent modulus decreased, especially dramatically when removal was only in the transverse direction. The trabecular bone models were also connected to a cortical shell and it was found that the apparent modulus of an entire slice was increased in comparison to the modulus of trabecular bone alone. We concluded that the architecture of trabecular bone, especially both lengths and percent of trabecular missing in the longitudinal direction, significantly influences mechanical properties.

<p>Transverse anisotropy is an important phenomenon of practical and scientific interest. Although the presence of ray tissue explains the high radial modulus in many hardwoods, experimental data in the literature shows that this is not the case for pine. It is possible that anisotropy in softwoods may be explained by the cellular structure and associated deformation mechanisms.</p><p>An experimental approach was developed by which local radial modulus in spruce was determined at sub-annual ring scale. Digital speckle photography (DSP) was used, and the density distribution was carefully characterized using x-ray densitometry and the SilviScan apparatus. A unique set of data was generated for radial modulus versus a wide range of densities. This was possible since earlywood density shows large density variations in spruce. Qualitative comparison was made between data and predictions from stretching and bending honeycomb models. The hypothesis for presence of cell wall stretching was supported by data.</p><p>A model for wood was therefore developed where both cell wall bending and stretching are included. The purpose was a model for predictions of softwood moduli over a wide range of densities. The relative importance of the deformation mechanisms was investigated in a parametric study. A two-phase model was developed and radial and tangential moduli were predicted. Comparison with experimental data showed good agreement considering the nature of the model (density is the only input parameter). Agreement is much better than for a regular honeycomb model. According to the model, cell wall bending dominates at both low and high densities during tangential loading. In radial loading, cell wall stretching dominates at higher densities.</p>

Transverse anisotropy is an important phenomenon of practical and scientific interest. Although the presence of ray tissue explains the high radial modulus in many hardwoods, experimental data in the literature shows that this is not the case for pine. It is possible that anisotropy in softwoods may be explained by the cellular structure and associated deformation mechanisms. An experimental approach was developed by which local radial modulus in spruce was determined at sub-annual ring scale. Digital speckle photography (DSP) was used, and the density distribution was carefully characterized using x-ray densitometry and the SilviScan apparatus. A unique set of data was generated for radial modulus versus a wide range of densities. This was possible since earlywood density shows large density variations in spruce. Qualitative comparison was made between data and predictions from stretching and bending honeycomb models. The hypothesis for presence of cell wall stretching was supported by data. A model for wood was therefore developed where both cell wall bending and stretching are included. The purpose was a model for predictions of softwood moduli over a wide range of densities. The relative importance of the deformation mechanisms was investigated in a parametric study. A two-phase model was developed and radial and tangential moduli were predicted. Comparison with experimental data showed good agreement considering the nature of the model (density is the only input parameter). Agreement is much better than for a regular honeycomb model. According to the model, cell wall bending dominates at both low and high densities during tangential loading. In radial loading, cell wall stretching dominates at higher densities. / QC 20101119