Date of Award
Master of Science
Department of Engineering Physics
John W. McClory, PhD.
Multivariate statistical techniques have been applied in order to un-mix nuclear fallout debris chemical data. This information is critical to characterization of fallout particle formation following a nuclear detonation. Understanding the correlation between environmental precursors and actinide concentrations in post-detonation nuclear fallout material aids in understanding the physical and chemical processes that alter nuclear device signatures in the fireball. This research examines 123 nuclear fallout samples from a historical nuclear test. Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) are used to collect chemical compositions of the fallout samples. Principal component analysis (PCA) is applied in order to examine variation in elemental compositions and make estimates of environmental precursor compounds. Geological soil information and prior research are combined to develop a final estimate. Mineral precursor estimates are modeled using multivariate curve resolution-alternating least squares (MCR-ALS). This method was demonstrated as a useful tool in determining composition. The precursors exhibited on polished sample cross-sectional surfaces were spatially correlated with sample radioactivity using autoradiography imaging of the sample set. Results suggest that feldspar is correlated with a moderate level of radioactivity, samples with porous textures have unique compositions with a uniform surface radioactivity, and quartz is anti-correlated with radioactivity. Additionally, size, shape, and morphology each have a relationship with actinide concentration with large, homogeneous and spherical samples exhibiting the highest relative radioactivity.
DTIC Accession Number
Pitkins, Christopher R., "Improving Fallout Characterization by Using Multivariate Techniques to Determine Composition" (2018). Theses and Dissertations. 1755.