Reynolds-averaged Navier–Stokes and stress-blended eddy simulation applications toward a realistic cough jet simulations in different oral configurations

The influence of turbulence modeling and oral cavity geometry on cough jet predictions was investigated using Computational Fluid and Particle Dynamics. Three oral geometries were examined: two realistic geometries with teeth details and different mouth openings, plus a simplified circular opening. The Reynolds-Averaged Navier–Stokes Shear Stress Transport (SST) k–ω model and hybrid Stress-Blended Eddy Simulation (SBES) model were compared to assess droplet dispersion predictions. The SST k–ω model overpredicts droplet diffusion, producing approximately 24% greater lateral dispersion than SBES simulations. SBES resolved coherent vortex structures critical to realistic particle transport in cough clouds, balancing computational cost with accuracy. A realistic narrow oral opening generated concentrated jets with rapid dispersion at face level, while the wide cavity produced downward droplet trajectories. The simplified circular mouth produced unrealistic symmetric jets that overestimated spatial spread. The narrow oral cavity produced rapid expansion: the cough cloud volume increased from 0.0227 m3 at 0.5 s to 0.0657 m3 at 2.0 s while droplet number density decreased from 40.77 to 11.9 n/cm3. The wide oral cavity showed slower cloud growth and rapid droplet number density reduction due to fewer ejected particles. The simplified model significantly overestimated concentrations initially, with a 99.9% reduction in droplet density within the first second. Droplet vaporization analysis showed rapid diameter reduction within 0.14 s, highlighting how water fraction affects droplet lifetime. Both anatomical realism and turbulence modeling influenced respiratory flow simulations and accurate infection risk prediction.