[Agavi]

Hi all,

Been stuck on this problem :

I'm trying to get the force distribution on a wind turbine blade. The total force and moment are calculated well using the "forces" function object from library "forces". However I need the spanwise distribution.

When I tried to use bins aligned with the blade using the "binField" function object from libraray "fieldFunctionObjects" I got a 'STEPPY', rugged distribution. This happens because I reckon 'binField' bins the cells based on some distance and integrates the force on those cells. However, my cells are not aligned with the separating surfaces of the bins and I think this is why I get a quite odd distribution. (see sketch attached)

I really need to get a clear picture of the integrated force in a single section, and therefore integrate within the right areas that might be cutting some of the faces. This is also why I wouldnt be able to use a facezone for this job.

What I decided to attempt is the following: (1) Make STL surfaces of the various blade sections (from the hub r~0 to r=Rmax radius). I can do this by chopping the initial blade using snappy and features and so on. (2) Then I'd want to map the "force" field from the "BLADE" patch to the section*.stl. (3) then I'd want to perform the "sum" operation using the "surfaceFieldValue" function object from "fieldFunctionObjects" library and operation "sum".

To check if steps (2) and (3) would work, if I could map a field from the patch to an stl and then perform summation on the stl, I use the blade.stl surface ( I know its not a section but I wouldnt go ahead doing step (1) before I'm sure steps (2) and (3) work here).

So basically I thought something like this should work :

forces1
{
type forces;
libs (forces);
writeControl timeStep;
writeInterval 1;
exectueControl timeStep;
executeInterval 1;
writeFields yes;
log yes;
porosity false;

region region0;
patches (BLADE);
pName p;
UName U;
rhoName rhoInf;
rho rhoInf;
rhoInf 1;
CofR (0 0 0); // Origin for moment calculations
}


totalForceInStl
{
type surfaceFieldValue;
libs (fieldFunctionObjects);

writeControl timeStep;
writeInterval 1;

writeFields true;
surfaceFormat none;

regionType sampledSurface;
name sampledTriSurface;

operation sum;
fields (force); % I've tried with U too just to see if the force field is not transferred but I get the same error.

sampledSurfaceDict
{
type sampledTriSurfaceMesh;
store true;
surface blade0.stl;
source boundaryFaces;
}

}

The last message from the solver is :

surfaceFieldValue totalForceInStl write:
total faces = 72368
total area = 0.045134908

(so it does run the FO and recognise the surface; the stl does have 72368 faces indeed and area is fine).

Sadly, right after, I get the following error:

Lists have different sizes: 0 != 6270

(I don't know what list has size 6270 in my simulation, the blade patch has many faces ~400,000!)

I'd expect to get a result that's the total force of the blade, that should also match the result of the 'forces' function object as a validation..

Any help would be much apppreciated

Many thanks

PS: This same thing is done in the sqareBend tutorial in of 2212, where UI3 is calculated on a sampled plane.

[Svetlana]

Hello,

Sorry, this is a pyvista suggested solution, not function object:

1. read openfoam case
2. read stl file
3. resample openfoam data to the stl
4. do the average or sum or whatever you wanted

1. https://docs.pyvista.org/examples/99.../openfoam.html
2. https://docs.pyvista.org/api/readers...ista.stlreader
3. https://docs.pyvista.org/api/core/_a...Filters.sample
4. numpy ?

[Agavi]

Thank you!! This looks like a great solution. Ended up following this tecplot tutorial: https://www.youtube.com/watch?v=dfpTj7hmPaE and got the smooth force distributions I needed.

Eventually should give a go to the pyvista solution so that I can keep my analysis independent of licensed software.

Many thanks