Quantifying Aerosol Deposition Under Non-standard Environmental Conditions: Combined Experimental Measurement and Computational Modeling Techniques

REN GARITY, James Henry, Moein Mousavi, Andrew Metcalf, Prasad Rangaraju, John Saylor, Clemson University

     Abstract Number: 413
     Working Group: Advancing Aerosol Science through Data Analysis Tools

Abstract
Currently, there remains a lack of data on aerosol deposition under “non-standard” environmental conditions, which include extremes of temperature gradients, humidity gradients, and/or changes in flow turbulence. This talk will address the use of data assimilation and surrogate modeling to evaluate aerosol deposition mechanisms under non-standard conditions, specifically in the case of a physical lab-scale model of a spent nuclear fuel dry storage cask (SNF DSC) system. As aerosol deposition in this system can result from a combination of non-standard environmental conditions, a combination of experimental measurement and computational modeling is being employed to quantify deposition and isolate the effects of specific mechanisms. The mechanisms of interest being investigated include thermophoresis, turbophoresis, turbulent dispersion, Saffman lift, and Stefan flow.

We will present an overview of the design of the lab-scale SNF DSC and physical deposition experiments, detailing our use of surrogate modeling through dimensional similitude used to replicate conditions in a full-scale cask. Additionally, we will display the use of computational modeling for evaluating correlations between environmental conditions and deposition. Finally, as our goal is to understand the causes of deposition in this system, we will discuss how statistical analyses are being employed to isolate effects of individual mechanisms, furthering understanding of which mechanisms dominate in non-standard systems.

Alone, either physical measurements or computational modeling place limitations on the conclusions that can be drawn regarding aerosol deposition in non-standard systems. Rather than making associations, we will discuss how our work aims to bridge the gap between experimental and computational results by using statistical comparisons as a data assimilation tool. Significance and effect size provide a way to determine the magnitude of interactions, while linear regressions allow for isolating effects of individual mechanisms. In this way, we aim to advance understanding of aerosol deposition and provide meaningful constraints on the findings.