The effect of relative air humidity on the evaporation timescales of a human sneeze.

Abstract

The present paper investigates droplet and aerosol emission from the human respiratory function by numerical and experimental methods, which is analyzed at the worst-case scenario, a violent sneeze without a face covering. The research findings develop the understanding of airborne disease transmission relevant to COVID-19, its recent variants, and other airborne pathogens. A human sneeze is studied using a multiphase Computational Fluid Dynamics (CFD) model using detached eddy simulation coupled to the emission of droplets that break up, evaporate, and disperse. The model provides one of the first experimental benchmarks of CFD predictions of a human sneeze event. The experiments optically capture aerosols and droplets and are processed to provide spatiotemporal data to validate the CFD model. Under the context of large random uncertainty, the studies indicate the reasonable correlation of CFD prediction with experimental measurements using velocity profiles and exposure levels, indicating that the model captures the salient details relevant to pathogen dispersion. Second, the CFD model was extended to study the effect of relative humidity with respect to the Wells curve, providing additional insight into the complexities of evaporation and sedimentation characteristics in the context of turbulent and elevated humidity conditions associated with the sneeze. The CFD results indicated correlation with the Wells curve with additional insight into features, leading to non-conservative aspects associated with increased suspension time. These factors are found to be associated with the combination of evaporation and fluid-structure-induced suspension. This effect is studied for various ambient air humidity levels and peaks for lower humidity levels, indicating that the Wells curve may need a buffer in dry climates. Specifically, we find that the increased risk in dry climates may be up to 50% higher than would be predicted using the underlying assumptions in Wells’ model.