Stochastic modeling of elastic behavior of fibrous ablators

Ablative materials in thermal protection systems (TPS) are used in aerospace applications to protect passengers and payload from critical degradation in high temperature conditions. The standard ablator, phenolic impregnated carbon ablator (PICA), is composed of carbon fibers embedded in a phenolic resin. Because the fibers are oriented randomly in the transverse plane, properties are not isotropic throughout the material. However, they are generally simplified to an imprecise macroscopic average that neglect local variations entirely. Here, we present a stochastic modeling approach that allows for assessment of properties as a function of local structure. The approach includes a computational toolkit for generating physically-motivated model representative volume elements (mRVEs), which are confirmed to exhibit transverse isotropy, just as PICA. We also present a methodology for computing the probability density functions of elastic properties for sets of mRVEs. Based on our results, local variation in fiber geometry within ablatives must be a major design consideration moving forward.