The correlation between multilevel micro-nano structures and thermal conductivity of nanoporous phenolic composites reinforced by needled fiber preforms

Nanoporous phenolic composites (NPC) are the most promising ablative thermal protection materials, but accurately modeling and effective thermal conductivity (λeff) prediction remains a significant challenge due to their intricate micro-nano structures. Herein, a modified generation method, which incorporates structural parameters from micro-CT, is proposed to develop precise NPC structural models for assessing the correlation between micro-nano structures and λeff. The Lattice Boltzmann method and Fourier's law are employed to predict their λeff, with only 8% discrepancy from experiment. Results from phenolic matrix reveals that once the porosity of matrix exceeds 60%, any further increases have a limited effect on λeff due to the Knudsen effect at nanoscale. Additionally, fiber structure analysis indicates that an increase in yarn height or width results in a decrease in λeff, yet if product of the two is maintained, λeff remains unchanged as the yarn cross-sectional area remains constant. Furthermore, investigation into stacked preforms demonstrates that λeff increases linearly with the number of fabric layers at a constant felt density, but decreases at a constant preform density owing to densely packed yarns. These results reveal the correlation between multilevel micro-nano structures and thermal conductivity of NPC, providing important guidance for optimizing the NPC thermal insulation performance.