מרק לסט

Nuclear forensics depends on accurate identification of fissile isotopes for security and nonproliferation purposes. However, traditional fission track analysis (FTA) is largely manual, operator-dependent, and incapable of distinguishing isotopes based on track features. This research aims to modernize FTA by developing a physics-informed, automated framework that combines Monte Carlo simulations (GEANT4), synthetic data generation, and image-processing algorithms for isotope differentiation. Simulations were performed for ²³³U, ²³⁵U, and ²²⁹Th using a LEXAN® polycarbonate solid-state nuclear track detector (SSNTD) with Aerogel spacers of various densities to improve track resolution. Synthetic datasets were analyzed through the newly developed FTA-Finder application. Results indicate that a 50 μm aerogel spacer enhances the separation of light and heavy fission fragment clusters. Image-processing algorithms successfully differentiated isotopes in synthetic data by reconstructing isotope-specific track-length distributions and angular spreads. These findings demonstrate that morphology-based parameters can serve as isotopic fingerprints, paving the way for automated, reproducible, and high-accuracy nuclear forensic analysis. The framework's integration into practical safeguards investigations could improve the reliability and speed of fissile material identification worldwide.