[ChrisWe]

Hey FOAMERS,

i want to calculate the Dissipation in OpenFOAM in my LES (pimpleFoam). Therefore I have to calculate <du'_i/dx_j * du'_i/dx_j>.
This value has to be averaged during runtime and so my solver needs the actual UPrime2Mean values of every time step.

In OpenFOAM-2.2.0 it is very easy to read in the averaged velocity fields. You only have to add the following code into your createFields.H of your solver:

 volVectorField UM
  (
  IOobject
  (
     "UMean",
     runTime.timeName(),
     mesh,
     IOobject::READ_IF_PRESENT,
     IOobject::AUTO_WRITE
  ),
     mesh,
     dimensionedVector("UM",dimLength/dimTime, vector::zero)
  );

When you compare UMean and UM, they are exactly the same (maybe deltaT shifted) and you can use UM in your solver.

When you do exactly the same in OpenFOAM-2.2.2 or OpenFOAM-2.3.0, UM will not be filled with the UMean values and UM will not be written in your timefolder.
OpenFOAM brings the following warning:

--> FOAM Warning : 
    From function GeometricField<Type, PatchField, GeoMesh>::readIfPresent()
    in file /home/iagwenze/opt/OpenFOAM/of230/OpenFOAM-2.3.0/src/OpenFOAM/lnInclude/GeometricField.C at line 108
    read option IOobject::MUST_READ or MUST_READ_IF_MODIFIED suggests that a read constructor for field UMean would be more appropriate.

Is there anybody who knows, why the presented code is not working anymore in OpenFOAM-2.2.2 or OpenFOAM-2.3.0?
Is there another possibility to read in Mean values to your solver?

Cheers,
Christoph

[fedvasu]

Hi ChrisWe,

Did you solve this problem?

Or at least did you circumvent it?

Regards,

[ChrisWe]

Hi,

I did not solve it in an elegant way, but as you mentioned, I circumvented it.
I do not exactly know how I solved it right know (I am not at my PC), but I created a new IOobject and calculated the average with some additional lines in my solver. Something like:
U_avg = U_avg * (timeSteps-1)/timeSteps + U / timeSteps

For example: When I am in my 1001th iteration, I calculate:
U_avg = U_avg * 1000/1001 + U / 1001

I was not happy with this solution, but it works.

I hope this helps...

All the best,
Christoph

[fedvasu]

Originally Posted by ChrisWe

Hi,

I did not solve it in an elegant way, but as you mentioned, I circumvented it.
I do not exactly know how I solved it right know (I am not at my PC), but I created a new IOobject and calculated the average with some additional lines in my solver. Something like:
U_avg = U_avg * (timeSteps-1)/timeSteps + U / timeSteps

For example: When I am in my 1001th iteration, I calculate:
U_avg = U_avg * 1000/1001 + U / 1001

I was not happy with this solution, but it works.

I hope this helps...

All the best,
Christoph

Ahh.. I think that should work for me.. Yeah not elegant.

Thanks,

[benqing]

I did not solve it in an elegant way, but as you mentioned, I circumvented it.
I do not exactly know how I solved it right know (I am not at my PC), but I created a new IOobject and calculated the average with some additional lines in my solver. Something like:
U_avg = U_avg * (timeSteps-1)/timeSteps + U / timeSteps

For example: When I am in my 1001th iteration, I calculate:
U_avg = U_avg * 1000/1001 + U / 1001

hi ChrisWe and fedvasu,

i am a new foamer . i want to Read in "Mean" values during runTime.
did you slove this problem? Or would you mind posting your code ?

Regards,
benqing