[vinz]

Dear all,

I need to simulate the heat transfer around an object at a certain altitude where the pressure is about 5000Pa. The problem is that since my object is quite big, I also have a large domain which is a sphere of 2km radius.

I want my fluid to be driven by the heat transfer from the object wall to the fluid, so natural convection. The problem is that my simulation crash after one or two time step.

If I scale down the whole domain to a sphere of about 2m radius, the simulation is working fine.

The difference is mainly the large gradients in pressure that are present in the large domain (due to gravity) and which are much smaller in the small domain.
So I tried to initialize the p field with the expected pressure before starting the simulation, but it does not solve the problem.

Does someone have an idea on how to solve this issue?

Here are the boundary conditions I use:
- p:

internalField   uniform 5000;

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

    farfield
    {
        type            calculated;
        value           $internalField;
    }
}

- p_rgh:

internalField   uniform 5000;

boundaryField
{
    object
    {
        type            fixedFluxPressure;
        value           uniform 5000;
    }

    farfield
    {
        type            fixedValue;
        value           uniform 5000;
    }

}

- T:

internalField   uniform 220;

boundaryField
{
    object
    {
        type            fixedValue;
        value           uniform 290;
    }

    farfield
    {
        type            inletOutlet;
        inletValue     uniform 220;
        value           uniform 220;
    }
}

- U:

internalField   uniform (0 0 0);

boundaryField
{
    object
    {
        type            fixedValue;
        value          uniform (0 0 0);
    }

    farfield
    {
        type            inletOutlet;
        inletValue     uniform (0 0 0);
        value           uniform (0 0 0);
    }

}

[vinz]

Nobody has an idea on how to solve this issue?

Do you need more information?

I forgot to say that the mesh is ok according to chechMesh.