Answers are grounded in this paper and cite their sources.
Multiscale chains for architectured materials routinely close the gap between a resolved microstructure and a structural model with an additive correction to one stiffness component, and routinely estimate that correction by a parallel mixture of the load-bearing phase. We resolve that mesoscale geometrically and show the estimate is wrong in sign. The system is the elytral cuticle of Phloeodes diabolicus. We build its trabecular cell from measured morphology as a signed-distance field and solve it with a Galerkin FFT homogeniser. The measured hemolymph-space fraction of 11% is not an input but the equation that fixes the column diameter, which leaves the band height as the only free morphological quantity. Measured against dense cuticle occupying the same thickness — what the chain assumes when it ignores the mesoscale — the architecture removes 57% of the through-thickness stiffness, so the additive term is GPa where the parallel estimate gives GPa. The sign holds over the whole admissible range. The mechanism is read off the solved field rather than asserted. The band is a sandwich core whose dense faces flex across the unsupported span: it closes 30 times more at mid-span than over a column, where the closure at mid-span is 5.72 times the macroscopic one and over a column only 0.19 times it, and the apparent through-thickness stiffness of the band - extracted from the cell by removing the dense layers in series - is 20 times below the mixture value for the band. The break-even column fraction follows at 0.785 against the 0.160 the parallel argument predicts, unreachable at the measured porosity. 22 verification gates support the computation; one exposed a silent 2.6 per cent defect in FFT projectors on even grids, and propagating the measured thickness scatter cannot change the sign. For this architecture the additive scalar is not a material constant: no scale separation exists in the direction that matters.
No takes yet. Share an insight, caveat, or question.
SUAREZ et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: