[PeterShi]

Dear all,

I am using rhoSimpleFoam to simulate the flow around NHLD-2D, which is a multi-element airfoil. The flow condition is Ma=0.197, Re=3.52e6, T=288, Pref=101755.6Pa, AOA=0 degree. And the version of my OpenFOAM is 4.1.

Now, the same error always occur, please see the picture attached below. And I have already tried many methods, but I still cannot get it done. I am not sure whether I set the wrong value, wrong schemes (stuffs in the fvScheme), or wrong solution method (stuffs in fvSolution).

As such, I upload my case folder, except the mesh file(it is too big, but if you need it, please let me know). Actually, the quality of my mesh is really good. It could not be caused by mesh.

As far as I am concerned, I think this error is related to thermodynamic parameters.

Please help me, it is a little bit urgent. Thanks in advance.

Best Regards,
Peter

[dybuk]

Hi PeterShi
Often on early iterations of compressible solvers temperature/energy equation produces unphysical values(negative temperature) if fields are not physicly initialized. Almost always "soft start" helps with this issue.
Try set inlet velocity conditions to uniformFixedValue and set small velocity on first iterations ramping to desired velocity in 100-500 iteration.
Something like this:

inlet
{
    type            uniformFixedValue;
    uniformValue    table
    (
        (0 (1 0 0)
        (200  (67.014 0 0)
    );
}

[PeterShi]

Originally Posted by dybuk

Hi PeterShi
Often on early iterations of compressible solvers temperature/energy equation produces unphysical values(negative temperature) if fields are not physicly initialized. Almost always "soft start" helps with this issue.
Try set inlet velocity conditions to uniformFixedValue and set small velocity on first iterations ramping to desired velocity in 100-500 iteration.
Something like this:

inlet
{
    type            uniformFixedValue;
    uniformValue    table
    (
        (0 (1 0 0)
        (200  (67.014 0 0)
    );
}

Hello Dybuk,

I copied your code and pasted it exactly as you showed. However, an error message came out:
--> FOAM FATAL IO ERROR:
"ill defined primitiveEntry starting at keyword 'uniformValue' on line 29 and ending at line 58"

file: /home/peter/OpenFOAM/Peter-4.1/2017/NHLP/rhoSimpleFoamrevised/3rrr/0/0/U at line 58.

From function void Foam:rimitiveEntry::readEntry(const Foam::dictionary&, Foam::Istream&)
in file db/dictionary/primitiveEntry/primitiveEntryIO.C at line 189.

FOAM exiting

I am unfamiliared with this type of bc, what is going wrong?

Many thanks.

Best,
Peter

[floquation]

There is a bracket mis-match.

[PeterShi]

Originally Posted by floquation

There is a bracket mis-match.

Hi Kevin,

What do you mean by that? Could you please give me more details?

Thanks in advance, and nice to meet another guy form Delft!

Best,
Peter

[PeterShi]

Hi,

I got it, the right form should be:
inlet
{
type uniformFixedValue;
uniformValue table
(
(0(1.0 0 0))
(200(67.0142924 0 0))
);
}

But the same error message as I showed at the very beginning came out again. And I tried other values instead of one, unfortunately, it still does not work out.

Any other advice?

Thanks.
Peter

[PeterShi]

Originally Posted by dybuk

Hi PeterShi
Often on early iterations of compressible solvers temperature/energy equation produces unphysical values(negative temperature) if fields are not physicly initialized. Almost always "soft start" helps with this issue.
Try set inlet velocity conditions to uniformFixedValue and set small velocity on first iterations ramping to desired velocity in 100-500 iteration.
Something like this:

inlet
{
    type            uniformFixedValue;
    uniformValue    table
    (
        (0 (1 0 0)
        (200  (67.014 0 0)
    );
}

Hello Dybuk,

Thanks. Do you have additional suggestions regarding to my case?

Best,
Peter

[floquation]

I don't know what is wrong with your case, but I do agree that dybuk's solution is a very plausible cause for an initial instability (assuming that your BC are not wrong etc.).
However, I have never worked with his method of achieving the ramping, and I'm not sure if it does the trick by just reading it. (For that purpose, I'd use groovyBC.)

I'd try the following:

1. Set a very low writeInterval.
2. In paraview, observe whether the inlet velocity does indeed ramp up, and not jump up after 200 iterations.

If yes (it does ramp), make the ramp slower.
If no (it jumps up), go for groovyBC for the same purpose.