Physics-Informed multiscale modeling of shock-to-detonation transition in heterogeneous energetic materials for robust performance under extreme environments

Energetic materials (EMs) are widely used in civil and military operations, where their detonation behavior must be carefully controlled to ensure reliability while preventing accidental ignition. A primary pathway to detonation is shock-to-detonation transition (SDT), a process in which a strong shock wave interacts with the material’s microstructure (the tiny internal features visible under a high-resolution microscope), initiates chemical reactions, and grows into a full detonation wave. In extreme environments, EMs can experience repeated pressure cycles that change their microstructure and potentially alter their detonation behavior. Engineers therefore need a way to understand how evolving microstructure influences detonation behavior under these demanding conditions.

Traditionally, SDT is simulated using simplified mathematical models that must be calibrated using expensive and time-consuming experiments. These models typically work only for the specific material and microstructure tested. As a result, they cannot reliably predict behavior when the microstructure changes over time, as may occur in extreme environments. This research develops a new simulation framework called MEASURE that removes the need for experimental calibration. Instead, MEASURE uses readily obtainable images of a material’s microstructure and applies a machine learning model to predict how shock waves will interact with it. This approach eliminates the need for experimental calibration and creates a direct link between microstructure and detonation behavior. Ultimately, the three objectives of this thesis establish, validate, and apply MEASURE as a tool for understanding how energetic materials behave in extreme environments.