CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   BuoyantSimpleFoam Channel problem (

prashant24983 August 11, 2008 01:56

Hello, I do not think buoy

I do not think buoyantSimpleFoam is suitable for your application. Read more on boussinesqBuoyantFoam in the dev version. buoyantSimpleFoam uses the perfect gas equation to calculate the density of media, for your case with water, boussinesq approximation is most appropriate. enFOAM-1.4.1-dev/applications/solvers/heatTransfer/boussinesqBuoyantFoam/

pawan August 11, 2008 02:17

Helo,, use fixedFluxBuoyantP
use fixedFluxBuoyantPressure boundary conditon for pd .

tian August 11, 2008 10:10

Hi, thanks for the answer.

thanks for the answer. Prashant, I use air as fluid in the channel. The channel is open in the top and buttom (inlet and outlet). So, my task is to find out how many fluid is driven because of hot cylinder inside (buoyancy flow) near the bottom.

Yesterday, I copied "boussinesqBuoyantFoam" in OF 1.5 and wmake it. It is working. I also compiled "calcMassFlow" to see how many fluid goes through my inlet and outlet (3D case).

It is difficult to explain. So I found a small video and a picture. Maybe you can have a look and see:

I am not sure about the boundary conditions.

Inlet: I do not know about the velocity
Outlet: I do not know about the velocity
Hot pipe inside: Temperature constant
Air inside: Temperature at the beginning I know
wall: adiabatic

I hope you can understand my poor englisch :-)
I also will try out "fixedFluxBuoyanPressure"

Thanks for help


tian August 28, 2008 06:38

Hi, Until now I not find a

Until now I not find a solution for my problem. I am confused why "buoyantSimpleFoam" need at time 0 the values for "p" and "pd"? The version OF 1.4.1 only need p. How I have to setup a outlet and inlet at time 0? Velocity for inlet and zeroGradiet for p... Outlet zeroGradient for u and fixedValue for p but pd, I have not an idea about it... Can some body give me advice? Thanks a lot.


mkraposhin September 17, 2008 07:59

you can try fixedFluxBuoyantPr
you can try fixedFluxBuoyantPressure for p and
fluxCorrectedVelocity for U

i am also interested in natural convection simulation, can you post image of your study?

what BC's are using for turbulence model?

tian September 17, 2008 08:43

Hi, thanks Kraposhin, I can

thanks Kraposhin, I can send to your email some screenshot.

I change my solver to the boussinesqBouyantSimpleFoam Solver for my study. The application also use RAS turbulence model.


tian September 17, 2008 09:55

Hi, I tried the BC's fixedF

I tried the BC's fixedFluxBuoyantPressure with fluxCorrectedVelocity for U but there are some errors if I start the solver. Can you explain me more about how I can use this BC's?

Thanks a lot.

mkraposhin September 18, 2008 04:50

ok, my e-mail is mkraposhin@in
ok, my e-mail is

the difference between boussinesqBouyantSimpleFoam and buoyantSimpleFoam is only that boussinesqBouyantSimpleFoam uses simplified boussinesq assumption. Both solvers can be used for solving natural convection tasks. What RAS model are you using? How do you calculate initial turbulence parameters on boundaries and in internal domain?

about BC's
fixedFluxBuoyantPressure is used to setup full pressure field on boundaries (p+r*g*h)

fluxCorrectedVelocity can be used when pressure field is already known. Another way, i think, is to use pressureInletVelocity for inlet and zeroGradient for outlet.

sorry for bad English

All times are GMT -4. The time now is 10:39.