Closing air curtains on aerosols

The US Army is building a large test facility to evaluate instruments for the optical detection of chemical and biological agents from 5km away.

The facility will disperse chemicals that optically simulate agents in a chamber to calibrate and test the instruments. The large test chamber is 15 x 18m² in cross-section and 135m long. An optical port 3m in diameter is located at each end of the chamber. Unfortunately, the optical ports cannot be conventional windows because unlimited optical access is required over a wide range of wavelengths. The simulants, while benign compared with the chemical and biological agents, must still be contained within the chamber and allowed to exit only through filters. Consequently, large air curtains have been put to use instead of windows. Battelle used computational fluid dynamics (CFD) and experiments to design the air curtain containment system. To represent one end of the test section, two large rooms are connected by a 9m long air curtain chamber, with a large port between each room and the air curtain chamber. One room acts as a dissemination chamber and is completely enclosed except for the port to the air curtain chamber. The opposite room has an opening to the ambient, simulating the optical port. The air curtain chamber has three air curtains, opposite each of which is a large suction blower, designed to pull in the air curtain flow and surrounding air, filter it, and discharge it. The experimental objectives were to demonstrate a five-order-of-magnitude concentration drop from the dissemination chamber to the ambient. The first experiments were promising because the required containment was achieved. However, smoke visualisation showed curtain oscillations when more than one curtain was operating, a result that FLUENT, the CFD software and consultancy service company, predicted accurately. Physically, the oscillations are caused by the instability that develops when the curtains interact with different surrounding pressure regions and with each other. When curtain #1 fluctuates toward curtain #2, the pressure rises between the curtains, and forces them apart. This is complicated by the colliding secondary flows from the air curtains as they fluctuate toward each other. Curtain #2 is prevented from swinging away from curtain #1 because ambient air is flowing hard into the #2 blower outlet. The pressure rise then pushes curtain #1 away. When it does, the pressure drops between the curtains, drawing the curtains together once again, and the oscillation continues. Based on the CFD predictions, Battelle modified the air curtain placement to create a stable vortex pattern between them. Both experiments and model predictions in the test facility verified the required contain-ment and stable operation. By using the unsteady modelling capabilities of FLUENT to offer insight into the physics of air curtain oscillations, a design that met performance targets was found cost effectively.