Uncertainty quantification of the advanced detonation shock dynamics model

Reliable prediction of detonation behavior in condensed-phase explosives requires models that offer quantitative robustness while maintaining computational efficiency for practically sized (centimeter-scale) systems. Advanced Detonation Shock Dynamics (ADSD) builds on the classical Detonation Shock Dynamics (DSD) framework by incorporating shock acceleration. In this work, we introduce methodological advances to ADSD to improve its ability to describe shock-to-detonation transition (SDT), diameter effect, and detonation failure with reactive-burn level accuracy at a substantially reduced computational cost. We then perform a comprehensive uncertainty quantification (UQ) analysis of the ADSD model parameters for PBX 9502 using experimental data from three experimental benchmarks: SDT, the diameter effect, and the shape effect. Statistical characterization is performed using UQ, yielding posterior parameter distributions that capture uncertainty and correlations; notably, many parameters showed low standard deviations, suggesting high confidence in their inferred values. The UQ-informed parameter sets are subsequently used to simulate corner-turning experiments, yielding predicted dead zones that show strong agreement with experimental observations. This combined optimization and Bayesian UQ framework provides the first systematic quantification of uncertainty in ADSD, enhancing confidence in its predictive capability and laying the groundwork for broader application of reduced-order detonation models in safety-critical engineering contexts.