A multi-scale computational model of respiratory droplet formation and dispersion

The need to study the mechanisms behind the formation of biological aerosols has become increasingly important as researchers continue to search for ways to limit the spread of COVID-19. Biological aerosols are formed from a continual mixing of the air that rapidly moves through the respiratory tract and oral cavity during breathing, speaking, coughing and sneezing with the saliva and phlegm that line the trachea and mouth. While the underlying processes that generate biological aerosols are quite complex, this study takes a simplified approach to modeling their formation by simulating the atomization of a thin liquid layer subjected to a high-speed shearing gas flow. Being able to properly reproduce the behavior of human saliva is challenging due to is inherent non-Newtonian behavior, as such this work explores various approaches to model the viscoelastic properties of saliva. To capture the formation and dispersion of the biological aerosol, a two-domain, multi-scale computational approach based on volume-of-fluid and Lagrangian droplet tracking is used to study two regions of interest: (1) a highly simplified oral cavity to observe the formation of fluid ligaments and sheets as well as the atomization of these liquid structures into droplets, and (2) a broader environment beyond the mouth where the ejected droplets are tracked in a Lagrangian fashion as they disperse. Results of the simulation allow for the observation of droplet formation and analysis of the size distribution of the generated droplets. The impact of non-Newtonian modeling on the break-up dynamics is explored.