Thin liquid layer atomization: a simple numerical model of cough

Thin liquid sheets atomization in closed channels is one of the main physical processes related to aerosols production during coughing and sneezing, with applications to health related issues such as Covid-19 transmission. In this work, we study a simplified cough model that consists of a thin liquid layer subject to an impulsively started, rapid air stream in a closed rectangular channel. We run numerical simulations of this setup using the Volume-of-Fluid method with octree adaptive mesh refinement to capture local small-scale phenomena related to turbulence and breakup. The obtained droplet-size distributions confirm trends between a Log-Normal and a Pareto distribution, in agreement with a previous analysis of experimental results by one of the authors. The atomization mechanisms present in this problem are analogous to those observed on previous and current multiphase mixing layer experiments and simulations. No mechanism for a bimodal distribution, also sometimes observed, is evidenced in the current simulations. We also report the droplet velocity distribution, observing that smaller droplets reach the stream velocity before leaving the channel.