As a doctoral student, my fundamental research interest was in studying the phenomena and harnessing of fluid mechanics in engineered systems, utilizing both computational and experimental techniques. I also applied agent-based modeling for components of engineered systems that are not well described by equations, such as behavior of a human population.
The specific scenario I was studying was on the real-time detection and mitigation of hazardous contaminants in public spaces defined by a long conduit, such as an airport terminal. To do this, I built a physical prototype and combined it with a computational fluid dynamics (CFD) flow control model. The CFD flow control model is built using OpenFOAM opensource CFD software, and coupled with the DAKOTA optimization software for determining the optimal control strategies. The basis of physical prototype design was a blower wind-tunnel that provided uniform, unidirectional ambient flow at the beginning of a test section. The test section of this prototype was over twenty feet long, with a four square-feet cross-sectional area. The prototype had the capability of sensing (via digital camera) an injected ‘contaminant’ (propylene glycol smoke) and mitigating that contaminant using actuators (compressed air operated vacuum nozzles and air knives) that were operated by a set of pressure regulators and a programmable controller. Furthermore, an agent-based model was developed and coupled with the flow control model in order to simulate ‘agents’ (i.e. building occupants) as they evacuate a public space in which there exists a threatening, dynamic environment, and flow control decisions made around these dynamic agents.