Quantifying Covid-19 Transmission Risk Indoors using GPU Accelerated Spectrally Accurate Simulations of Aerosol Transport

One of the primary modes of transmission of the SARS-CoV-2 virus has been identified to be airborne, with human generated respiratory aerosols and droplets being the main carrier of the virus. This is not only applicable for SARS-CoV-2, but for all airborne diseases. The respiratory aerosols/droplets are generated during various expiratory events, like speaking and coughing; and understanding their dispersion will help design effective transmission mitigation strategies. To quantify this phenomena we conducted high-fidelity turbulence resolved multiphase simulations, where the aerosols were modeled using the Lagrangian particle tracking method. The model was implemented in the NekRS, the GPU accelerated version of the high-order spectral element based incompressible Navier-Stokes solver Nek5000 [1]. A fast and robust GPU-enabled interpolation utility for NekRS was utilized [2].

Simulations were conducted at different ventilation rates, for a room of size 3.35 m x 2.25 m x 4.25 m. We will present results that quantify the effect of aerosol size and initial position on ventilation induced dispersion and deposition pattern within the room. We also quantify the turbulence within the room, and how it might be affecting aerosol washout efficiency. We will also present speedup comparison between the Nek5000 and NekRS implementation of the aerosol transport models.



[1] https://nek5000.mcs.anl.gov/

[2] Lindquist, N., Min, M., & Fischer, P. (2021). Scalable Interpolation on GPUs for Thermal Fluids Applications (No. ANL-21/55). Argonne National Lab.(ANL), Argonne, IL (United States).