CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Freestream boundary condition (https://www.cfd-online.com/Forums/openfoam-solving/93093-freestream-boundary-condition.html)

 YJ Lee October 4, 2011 23:03

Freestream boundary condition

Dear FOAMers,

Hi, I'm pretty new to OpenFOAM and CFD. One question I have is regarding Freestream boundary condition - how exactly does it work?

Specifically, I tried to run an external aerodynamics case, using fixedValue for velocity, nut and nuTilda at outer boundary, zeroGradient for pressure at outer boundary; simpleFOAM, S-A turbulence. The results are rather unphysical and diverging.

However, following the airfoil example for simpleFoam and using freestream boundary condition, the simulation seems to be more reasonable and without divergence. My question is, how does freestream boundary condition differ from fixed values?

Thanks, FOAMers!

 fcollonv October 7, 2011 10:54

Hello Lee,

The freestream BC has the type inletOutlet meaning that it looks locally (for every face of the patch) at the mass flow rate. And if the flow is going outside the boundary will be locally zerogradient, if it is going inside the boundary will be locally fixedValue.
The freestreampressure BC is a zeroGradient BC but it fixes the flux on the boundary to be rho*Sf*freestreamValue.

Good luck

Frederic

 YJ Lee October 8, 2011 10:37

Thanks, Frederic! Think I get the point now...

 malaboss December 11, 2012 12:29

Hi,
Thank you for the explanation. If i am right the difference between a freestream BC and a fixedValue BC is that for fixed value there are constraints on each vector of the velocity field, while with freeStream we have a constraint only for the flux.

This is why the solution sounds more physical ?

 YJ Lee December 19, 2012 22:18

Hi, malaboss

Freestream BC is like a hybrid fixedValue and zeroGradient boundary condition. It behaves like a zeroGradient when fluid is flowing out of the boundary face, but behaves like a fixedValue when fluid is not flowing out.

So, instead of fixedValue that imposes its constant value regardless of situation, freestream is more flexible, doing whatever is more physically realistic, so to say.

 ziemowitzima October 8, 2014 07:05

Hi,
I am using currently freestream BC for the flow in the tunnel (channel).
Is it an appropriate BC for such flow ?

In short, there is large tunnel (3m in diameter) flow is from left to right (inlet, outlet with freestream BC) but inside the tunnel there is additional small inlet with some mass flow specified.

thanks

 ASo May 5, 2015 04:00

Hi everyone,

Is it possible to use the freestream BC (to simulate a external hydrodynamic case) in 3D.
In this case what is the good boundary condition to use for the frontAndBack ?

Thanks for answers

 ASo June 4, 2015 08:03

Quote:
 Originally Posted by fcollonv (Post 327067) Hello Lee, The freestream BC has the type inletOutlet meaning that it looks locally (for every face of the patch) at the mass flow rate. And if the flow is going outside the boundary will be locally zerogradient, if it is going inside the boundary will be locally fixedValue. The freestreampressure BC is a zeroGradient BC but it fixes the flux on the boundary to be rho*Sf*freestreamValue. Good luck Frederic
Hi all,

Could you explain me what is Sf.

In advance, many thanks

 lonelywing July 7, 2015 02:04

Quote:
 Originally Posted by ASo (Post 548870) Hi all, Could you explain me what is Sf. In advance, many thanks
As far as I know, Sf means the surface area of a face of the cell.

 salisadeghi December 19, 2016 10:08

3D airfoil mesh

Hi everyone, i import .msh file (3D Cgrid mesh for airfoil in gambit) in openfoam and i got this error: {illegal cell label -1 in neighbour addressing for face 0} is it all about boundary condition?
the main question is haw to define that in 3D gambit to not face this error?or is it possible to import just geometry and mesh without boundary and then define it just in openfoam?

thanks all of you in advance

 hfs July 19, 2017 04:36

Quote:
 Originally Posted by fcollonv (Post 327067) Hello Lee, The freestream BC has the type inletOutlet meaning that it looks locally (for every face of the patch) at the mass flow rate. And if the flow is going outside the boundary will be locally zerogradient, if it is going inside the boundary will be locally fixedValue. The freestreampressure BC is a zeroGradient BC but it fixes the flux on the boundary to be rho*Sf*freestreamValue. Good luck Frederic
Quote:
 Originally Posted by YJ Lee (Post 398354) Hi, malaboss Freestream BC is like a hybrid fixedValue and zeroGradient boundary condition. It behaves like a zeroGradient when fluid is flowing out of the boundary face, but behaves like a fixedValue when fluid is not flowing out. So, instead of fixedValue that imposes its constant value regardless of situation, freestream is more flexible, doing whatever is more physically realistic, so to say.

Thank you both for the explanation.
I want to use freestreampressure but combined with having the pressure prescribed as a value:
"Prescribed pressure; with allowed in/outflow reversal"
Is this possible in OpenFoam? Thanks,

PS: more details:
It is a wind engineering in-compressible flow simulation.
I have a prescribed inlet velocity BC (Fluctuating Inlet). We usually combine this with a zeroGradient Pressure BC on the inlet.
I want to have Inlet/Outlet condition on the Top, Sides and Outlet. However, a pressure value should be described on some boundary. Usually we use a fixedValue 0 for pressure. Is there a way to combine this with freestreampressure?

 All times are GMT -4. The time now is 14:49.