CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   Setting BC parameter values for a buoyantBoussinesq run (

NJG March 5, 2013 12:19

Setting BC parameter values for a buoyantBoussinesq run
Hello Community,

I need some advice on setting certain parameters for a run case I am working on. Let me give some background first then I will go into the specific parameters.

The case I am modeling is a room empty except for a single occupant in the middle breathing in and out. The space has no ventilation. I am most interested in the distribution of CO2 concentration in the room over time as a result of this breathing.

To study this I have modified the buoyantBoussinesqPimpleFoam solver to also include mole fraction of CO2, molecular diffusivity, and have changed effective density as far as body forces are concerned to change both from temperature changes and mole fraction changes (i.e. buoyant forcing both from temperature and composition changes). I have also created a new wall function for determining the turbulent molecular diffusivity near the wall.

I have built my domain as a simple rectangular space with the person modeled as a pillar in the center of the room that goes to roughly human height (it does not extended all the way to the “ceiling”). The man-pillar has a small mouth sized room inlet at face height (modeling breathing out warm CO2 rich air) and for now I have a room outlet on the top of his head to capture the breathing in.

The parameters I am trying to determine at this point are k, epsilon, nut, alphat, Dxt, and kappat (Dxt is my added turbulent molecular diffusivity term). Particularly I am interested in how to set these at my inlet and outlet, but I am also interested in double checking the values for my walls, ceiling, and floor. For the most part I have used the buoyantBoussinesqPimpleFoam tutorial case “hotRoom” as a guide to settings BCs for the the walls, ceiling, and floor for now, but would like a deeper understanding of this. Presently, I have “zeroGradient” BCs for all of these at my outlet and some “fixedValue, uniform” BC for each of these at my inlet.

So, my questions then are:

1) Does my basic approach to my inlet and outlet seem correct (fixed value and zero gradient)?

2) If so, I am struggling a bit to determine what those fixed values should be. Any suggestions?

3) If not, what should I consider instead?

4) For k, epsilon, alphat, kappat, and Dxt (for walls ceiling and floor) my approach has been to set the internal field to a uniform value (0 for alphat, kappat, and Dxt, 0.1 for k, and 0.01 for epsilon) initially, and then set my walls, ceiling, and floor to the relevant wall functions. These wall functions take input parameters themselves (alphat just a “value”, kappat a prandtl number and a value, and Dxt a schmidt number and a value). The “values” were set to uniform values the same as their initial field. The prandtl number and schmidt number I used 0.9 and 0.7 respectively. I would appreciate any thoughts or feedback on any of this, or tips on how to determine this all. All these “values” and numbers were chosen based on other tutorials I have seen so I could use a more rigorous basis for these.

Thanks for your time everyone! I appreciate any thoughts or feedback you can give.


All times are GMT -4. The time now is 04:52.