Mechanical testing

Bones were kept hydrated in PBS for 24 hours prior to testing (placed in PBS after cutting and during scanning; only taken out of PBS immediately before testing) to prevent desiccation. Samples were tested using a single column tensile/compression testing machine (Zwick/Roell, Ulm, Germany) with a 1‐kN load cell (see Fig. ​Fig.11 C). Cores were tested in uniaxial unconfined compression between parallel aluminum plates at a rate of 2 mm/min until failure was observed. Force and displacement data recorded during the test were used together with sample dimensions to calculate apparent stress (σ, MPa) and strain (ε) for each sample. The apparent elastic modulus (Young's modulus, MPa) of cancellous cores was calculated from the slope of the linear portion of the stress–strain curve. Apparent failure strength (MPa) was deemed to be the maximum stress endured during the test. The area under the loading curve from 0.01 strain up to the strain at maximum stress endured was used to estimate the “energy to failure.” As samples failed by compressive crushing, no definitive fracture point could be identified, thus energy to fracture (a measure of toughness) could not be estimated. A measure of “post‐yield energy” was calculated from the area under a portion of the post‐yield curve (normalized by the strain experienced). This portion corresponded to the area under the loading curve for 0.004 to 0.01 strain directly after the maximum stress experienced during a test.