Modeling of Load-Structure-Damage-Performance Linkages in Heterogeneous Energetic Materials

Energetic Materials (EMs) can accumulate microstructural damage in varied environments in which they are deployed, due to potentially large variations in temperature and load cycles. This damage may result in off-design performance or even failure of shock-to-detonation transition. To increase the robust performance of EMs, the load-structure-damage linkage must be predicted, and the effect of damage on performance must be minimized. A tool to accomplish the latter is presented here, describing an approach to rapidly estimate performance given the imaged microstructure of a damaged or undamaged EM. Specifically, the morphological features of the imaged microstructure are extracted, and the ignition and growth rates of each feature are estimated for a given impact strength using a surrogate model. These estimates are then used to form an aggregate sensitivity prediction. The surrogate model is trained using an ensemble of single-defect simulations produced using the Eularian code SCIMITAR3D. We hypothesize that the presented tool will provide rapid and accurate sensitivity predictions for both damaged and undamaged microstructures. When combined with a load-structure-damage linkage, this will provide a powerful tool to ensure robust performance of EMs in extreme environments such as hypersonic operational conditions.