Reduced radiation models of high-temperature porous and fibrous insulation compress the pore geometry into a few scalar optical properties. We test whether that description transfers between morphologies with different directional pore paths. Four digital microstructures (open-cell foam, random and aligned fiber networks, tortuous layered stack) are compared at matched solid volume fraction, material properties, and boundary conditions, with pore-scale surface-to-surface radiation as the common reference in prescribed-temperature and strictly coupled radiation-conduction calculations. At a representative condition (700 K mean temperature, ɛ = 0.6 , 10 mm slab), rearranging the same amount of solid changes the coupled heat flux by a factor of 2.62, and radiation carries about 14% to 74% of the total. Turning the transfer axis changes the flux by almost a factor of two for the layered structure and by about 40% for the aligned fiber network, against less than 10% for the near-isotropic foam and random fiber. An aligned and a random fiber geometry matched in solid volume fraction, interfacial area density, and mean beam length still swap their flux ordering between transfer axes. A transport correction lowers the median wall-flux error of the homogenized discrete ordinates model from about 28% to 12%, with about 8% for the near-isotropic structures but 23% and 62% for the aligned fiber and layered ones. Reduced-model accuracy thus depends jointly on pore morphology and on how the equivalent properties are obtained; within the structures examined, equivalent-medium prediction is adequate for near-isotropic pore spaces, whereas strongly directional ones require an angular representation of extinction.
No takes yet. Share an insight, caveat, or question.
Yan et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: