Instable natural convection case
 

Hi,

I am busy for a test case using buoyantBoussinesqPimpleFOAM.
It is case with natural convection. The 2D geometry exist out of a square domain, with in the middle a cilinder. The cilinder has a temperature of 400K and the surrounding field has a temperature of 300K. I made the sides and bottom pressure inlets and the top a pressure outlet. I used buoyantBoussinesqPimpleFoam, because the case is lightly compressible and transient. When I am post processing the case I see that the heat that is rising is instable. Some kind of velocity waves comes in the domain from the outside. This is probably due to not matching density with my boundary temperature, or something. Any body ideas? How can I fix this? Define a rhoInf? How?

I will put my boundary conditions below, to see what I have done.

Alphat:
{
p_out
{
type zeroGradient;
}

p_in
{
type zeroGradient;
}

sym_back
{
type empty;
}

sym_front
{
type empty;
}

wall
{
type alphatWallFunction;
value uniform 0;
}
}

Epsilon:
{
p_out
{
type zeroGradient;
}
p_in
{
type zeroGradient;
}
sym_back
{
type empty;
}
sym_front
{
type empty;
}
wall
{
type epsilonWallFunction;
Cmu 0.09;
kappa 0.41;
E 9.8;
value uniform 0.01;
}
}

K:

internalField uniform 0.1;

boundaryField
{
p_out
{
type zeroGradient;
}
p_in
{
type zeroGradient;
}
sym_back
{
type empty;
}
sym_front
{
type empty;
}
wall
{
type kqRWallFunction;
value uniform 0.1;
}
}

Kappat:

internalField uniform 0;

boundaryField
{
p_out
{
type zeroGradient;
}

p_in
{
type zeroGradient;
}

sym_back
{
type empty;
}

sym_front
{
type empty;
}

wall
{
type kappatJayatillekeWallFunction;
Prt 0.85;
value uniform 0;
}
}

P:
internalField uniform 1e5;

boundaryField
{
p_out
{
type calculated;
value $internalField;
}

p_in
{
type calculated;
value $internalField;
}

sym_back
{
type empty;
}

sym_front
{
type empty;
}

wall
{
type calculated;
value $internalField;
}
}

p_rgh:


internalField uniform 1e5;

boundaryField
{
p_out
{
type buoyantPressure;
value uniform 1e5;
}

p_in
{
type buoyantPressure;
value uniform 1e5;
}

sym_back
{
type empty;
}

sym_front
{
type empty;
}

wall
{
type buoyantPressure;
value uniform 1e5;
}
}

T:

internalField uniform 300;

boundaryField
{
p_out
{
type zeroGradient;
}

p_in
{
type fixedValue;
value uniform 300;
}

sym_back
{
type empty;
}

sym_front
{
type empty;
}

wall
{
type fixedValue;
value uniform 400;
}
}

U:
internalField uniform (0 0 0);

boundaryField
{
p_out
{
type zeroGradient;
}
p_in
{
type pressureInletVelocity;
value uniform (0 0 0);
}
sym_back
{
type empty;
}
sym_front
{
type empty;
}
wall
{
type fixedValue;
value uniform (0 0 0);
}
}

I hope somebody can help me out. I can send the case file to email or something.

With kind regards

Peter Rupert

ideas?
 

Somebody can help me out?

Hi Peter

I am also having problems similar to yours which involves a buoyant natural convection case. What I am trying to simulate is essentially the natural updraft of buoyant warm air in chimney. I have tried various combinations of boundary conditions, buoyantSimpleFoam, buoyantPimplefoam, etc, however the velocities either get unreleastically large or the solution will not converge.

I think it would be very helpful if someone would be willing to post an example of a natural convection case similar to the hot room tutorial, but which also includes an inlet / outlet.

Eric

Hello Peter,

It is a little bit late and I am also a beginner in OpenFOAM, but I think, the boundary conditions for the out- and inflow of p_rgh are not very good. The boundary type "buoyant pressure" sets the right pressure gradient and I think, it is not recommended for the outflow.

Maybe you have already found it, but the following threads may help you:

http://www.cfd-online.com/Forums/ope...servation.html

http://www.cfd-online.com/Forums/ope...onditions.html

I hope this helps!

Best regards,

Alex

Natural Convection
 

He Alex,

Thanks for your reply.
I have seen the 'Temperature Inlet/Outlet thread.
My p_rgh is indeed wrong and I changed it. I got some better flow pattern, however my velocity is way to high, let say around 120m/s. It should be around 0.5 or something.

Greetings

Peter

Hi Peter,

maybe your pressure is too high and you get trouble with your U boundary condition for the inlet. buoyantBoussinesqPimpleFoam uses the pressure divided through the density. Look at the tutorial: the dimensions of the pressure are [0 2 -2 0 0 0 0]. I think you wanted the "normal" atmosphere pressure around 1e5.
Maybe that is the problem?

Regards,

Alex