Discussion Forum

Hello all,

I am attempting to simulate the mixing of a dilute aerosol with atmospheric air in a tube in 2-D. The ID of the tube is constant at 1.7 cm. The inlet flow rate is 1 L/min and the outlet is 5 L/min. These flow rates correspond to Reynolds Numbers of 77 and 386 respectively. There is a square window cut in the tube where air is free to flow in. The aerosol is contained in the air flowing through the inlet. Video of the system functioning indicates that there exist eddies within the flow in the area of the window. Aerosol can also be seen leaving through the window. Because of this I believe that the flow regime within the region of the window is turbulent due to the cross flow velocity of the air entering through the window. I have had success in modelling the fluid flow in the laminar regime, although the results do not match with the observed results so I do not believe laminar flow is the correct physics model to use. I have attempted to use the k-epsilon and low Re k-epsilon models to model the flow but am unable to achieve a solution. I attempted to use the solution I obtained from the laminar flow as the initial conditions for my turbulent model with no success. I am very new to COMSOL and CFD in general and any advice on this subject is appreciated. I have attached the 2-D laminar model I made to demonstrate the geometry and boundary conditions.