Ensemble-based data assimilation for material model characterization in high-velocity impact
Published in International Journal of Impact Engineering, Volume 215, Article 105738, 2026
High-fidelity simulations for high-velocity impact rely on material models and parameters that are often calibrated through labor-intensive manual fitting. This work develops an ensemble-based data assimilation framework that integrates Smoothed Particle Hydrodynamics, the ensemble Kalman filter, and adaptive covariance inflation to automatically identify representative plasticity, fracture, and equation-of-state parameters from a single high-velocity impact test. Using synthetic back-face deflection data of an AZ31B magnesium plate, the study shows that the proposed approach can efficiently recover sensitive parameters, diagnose identifiability through ensemble statistics, and provide a robust alternative to traditional calibration workflows.
Recommended citation: Rong Jin, Guangyao Wang, and Xingsheng Sun (2026). "Ensemble-based data assimilation for material model characterization in high-velocity impact." International Journal of Impact Engineering, 215, 105738. DOI: 10.1016/j.ijimpeng.2026.105738
Download Paper
