As candidates for tissue-independent phases of cortical and trabecular bone we consider (i) hydroxyapatite, (ii) collagen, (iii) ultrastructural water and non-collagenous organic matter, and (iv) marrow (water) filling the Haversian canals and the intertrabecular space. From experiments reported in the literature, we assign stiffness properties to these phases (experimental set I).
On the basis of these phase definitions, we develop, within the framework of continuum micromechanics, a two-step homogenization procedure: (i) at a length scale of 100–200 nm, hydroxyapatite (HA) crystals build up a crystal foam (“polycrystal”), and water and non-collagenous organic matter fill the intercrystalline space (homogenization step I); (ii) at the ultrastructural scale of mineralized tissues (5–10 microns), collagen assemblies composed of collagen molecules are embedded into the crystal foam, acting mechanically as cylindrical templates. At an enlarged material scale of 5–10 mm, the second homogenization step also accommodates the micropore space as cylindrical pore inclusions (Haversian and Volkmann canals, inter-trabecular space) that are suitable for both trabecular and cortical bone. The inputs for this micromechanical model are the tissue-specific volume fractions of HA, collagen, and of the micropore space. The outputs are the tissue-specific ultrastructural and microstructural (=macroscopic=apparent) elasticity tensors.
A second independent experimental set (composition data and experimental stiffness values) is employed to validate the proposed model. We report a small mean prediction error for the macroscopic stiffness values. The validation suggests that hydroxyapatite, collagen, and water are tissue-independent phases, which define, through their mechanical interaction, the elasticity of all bones, whether cortical or trabecular.