Aerosol Transmission of COVID-19 and other Airborne Diseases in office environments using Computational Fluid Dynamic Modeling and Machine Learning

The COVID-19 pandemic has shown the world how quickly airborne diseases can spread and the lasting impact they can have. Computational fluid dynamic (CFD) models and simulations and machine learning (ML) are powerful tools that allow engineers to create models to predict and advance tools to fight these airborne diseases. The research in this thesis studied the effects of heating, air conditioning and ventilation (HVAC) strategies in small office spaces. A novel methodology was developed to utilize ML, CFD and parallel computing by utilizing the user defined function (UDF) tool of ANSYS Fluent. It was shown that the resulting risk models were quick and effective at predicting high risk areas using spatial data or predicting regions of high risk over time. Future research will refine this method by creating higher fidelity ML models and investigating a wider range of input and output parameters.