[jose_rodrig]

Hi

I am simulating a combustion chamber for flameless combustion with the outlet boundaries pretty close to the inlet boundaries.

When I ignite the mixture, the simulation diverges (the temperatures get higher and higher) as the hot air gets closer to the outlet. I checked on the solution right before the crash and I saw that the outlet temperature did not changed as the hot gases approached with temperatures 2000k higher.

1) I am setting zeroGradient in this boundary, which should allow the outlet temperature to raise with the outgoing gases. Am I right??

2) Also, a colleague suggested me use extrapolation to set the boundary values. How does OpenFoam do extrapolation for BCs?

3) What does $internalField do? For example in the dictionary

Outlet
{
    type    fixedValue;
    value   $internalField;
}

I am stuck in this for a while and I need to move forwards as this work is for my thesis and I dont have a lot of time now.
Thx

[fcollonv]

Hello Jose,

1) Indeed zeroGradient is the right boundary condition to use if you are sure that the flow can not enter by that boundary. In the latter case you can make use of inletOutlet.
But to have the temperature going outside your domain, you need a flow that is going outside:Could you provide the boundary conditions for the pressure and the velocity?
2) I don't understand what extrapolation means in that case. zeroGradient will copy the value of the nearest cell to the boundary face - is that the extrapolation your are speaking about??
3) $internalField means that the value use to set the internal mesh will be use as fixed value for the boundary condition.

Good luck

Frederic

[jose_rodrig]

Hi Frederic,

1) in that same boundary, I have pressure prescribed and velocity is also zeroGradient. I have checked the solution and the flow is, in fact, out flowing at that point.
2) Simply, I mean extrapolation by computing the boundary value with a simple extrapolation using the cells upstream the outlet ( as much cells as I prefer) I guess there is no other way to do this in OpenFOAM. So zeroGradient is the closest I can get.

Anyway, if it works like you say (copy the value of the neighbor cell) I dont understand why it diverges as the the hot gases come closer to the boundary.

[samegypt]

Originally Posted by jose_rodrig

Hi Frederic,

2) Simply, I mean extrapolation by computing the boundary value with a simple extrapolation using the cells upstream the outlet ( as much cells as I prefer) I guess there is no other way to do this in OpenFOAM. So zeroGradient is the closest I can get.

Hi,
I like to ask how do you define an extrapolation BC using OpenFoam? by using zeroGradient or is there another BC ??

regards,
Samir