Comparing RANS with LES - AMR Study in Indoor Bio-aerosol Transmission

Slides from a presentation held at MULTIPHYSICS 2022, a conference organised by The International Society of Multiphysics, in Oslo, Norway, 15-16 December 2022. https://www.multiphysics.org/home.The study presents the multiphysics modeling of bio-aerosols generated from human respiratory impulses such as coughing, sneezing, breathing, and talking. These droplet-packed viruses exploit the influence of airflow for effortless transmission, rendering them a potential health risk to individuals in the same indoor environment. To gain insight into the transmission of these airborne particles, their accurate prediction is critical. Computational Fluid Dynamics (CFD) is a numerical method that can be utilized for this study, however, it's computational intensity in terms of accuracy relies on mesh. Thereby, optimizing mesh generation in the transient calculation of droplet transmission is essential to obtain more accurate results. This study investigates the application of Adaptive Mesh Refinement (AMR) and its integration using Reynolds Average Navier-Stokes (RANS) and Large Eddy Simulation (LES) models in the ANSYS Fluent CFD tool to validate an experimental and numerical study of indoor aerosol transmission. The simulation case setup takes after the experimental investigation of a simplified human body of 1.67m height, located at 1.64m in length, mid-width from the wall of a room of 5.4m (length) x 6.6m (width) x 3m (height) with four (4) air inlets on the roof and five (5) air outlets each on the lower side of both walls. The boundary conditions are set at; low, medium, and high, inlet air supply velocities of 0.16m/s, 0.28m/s, and 0.33m/s, respectively. A constant room temperature of 20⁰C is adopted. The exhaled particles in the computational ventilated room domain are injected using the Discrete Phase Model (DPM). Integrating the AMR with RANS and LES models, the meshes are refined in the region of interest as it successfully follows the main flow profile during transmission and establishes stable refined mesh around the dispersed particles. It is hypothesized that RANS with AMR methodology will be less computationally intensive; however, it will provide a reasonably accurate solution. We also intend to test LES with AMR and compare the computational intensity and accuracy of the results with the experiments. The presented study on the mesh optimization process is part of the more extensive investigation that aims to understand better the transmission modes of bio-aerosols (e.g., virus-laden droplets) in ventilated indoor environments