[anon_q]

Hello everyone, I have a simple question about lift and drag calculation in OpenFOAM. Obviously, in a simulation with no moving mesh it is straightforward to specify LiftDir and DragDir (lift direction vector and drag direction vector, respectively). But in case of sliding mesh, that is, those directions are changing with respect to time, so How can we specify the lift and drag directions in this case?

Update: I want to clarify the question because someone got it wrong. What I mean exactly is: in case of say rotating airfoil about some axis, horizontal axis wind turbine, vertical axis wind turbine ...etc) obviously the body is moving so in this case how to specify the lift and drag directions?

[piu58]

Originally Posted by Evren Linda

Hello everyone, I have a simple question about lift and drag calculation in OpenFOAM. Obviously, in a simulation with no moving mesh it is straightforward to specify LiftDir and DragDir (lift direction vector and drag direction vector, respectively). But in case of sliding mesh, that is, those directions are changing with respect to time, so How can we specify the lift and drag directions in this case?

The direction of the force has nothing to do with the mesh. You may give it with respect to your body (which is a solid one, I assume) or with respect to the external flow.

[anon_q]

No please I guess you can understand what I mean. Of course I mean a body within a dynamic mesh (rotation, oscillation,...etc). In my example I've a rotating airfoil about some axis. Can you imagine lift and drag for no object??!!!

[wyldckat]

Greetings to all,

I've arrived at this thread, coming from here: https://github.com/blueCFD/Core/issues/98 - from which the following additional questions+information are relevant:

If a have a simulation with dynamic mesh because the body is also moving, How can I set the liftDir and dragDir directions? Can OpenFOAM detect the direction automatically ? or do I need to create some scripts to deal with that (e.g Python)?

OK, if the body is moving with the mesh, and if coding isn't possible, then currently the only possibility is calculate the forces after the simulation is over.

This way it's possible to change the lift and drag directions for each already recorded time step and use a function object entry in accordance to each time step. The annoying part is that the lift and drag directions have to be known for each time step and manually defined...

@Evren, a few questions:

1. Any chance you know of a tutorial case in OpenFOAM that we can use as an example to test this?
2. In your case, is the body rigid? Or does its surface also change?

Best regards,
Bruno

[anon_q]

Thank you very much dear Bruno,
For an example let's say we want to calculate force coefficients (Cl, and Cd) for the following simulation of vertical axis wind turbine:
https://www.youtube.com/watch?v=6o4L...ature=youtu.be
The surface is not changing but the mesh is composed of a rotating zone and a fixed zone

[wyldckat]

So that means that the positions are predefined and can be calculated directly based on the rotation speed?

My concern is if the rotation/movement speed was meant to change with the flow itself...

[anon_q]

In my case the rotation speed is constant, the position of the airfoils as a function of time are known.

[wyldckat]

Quick answer: Unfortunately I didn't have more time beyond the following attached example:

1. The attached case "movingCone.tar.gz" is based on the tutorial case "incompressible/pimpleDyMFoam/movingCone".
2. I used OpenFOAM 5.x.
3. To run the case (after unpacking), simply run:
   

   ./Allrun

4. The new line of interest for you is that I ran the post-processing step like this:
   

   pimpleDyMFoam -postProcess -dict system/functionsDict

   make sure you update to your own solver, whichever it is (I don't remember and I didn't check).
5. Study the following two files:
   1. "system/functionsDict" - this has the calculation of "forceCoeffs" for each specific time step.
   2. "system/forceCoeffs" - this file has the common configuration for the object being monitored.
6. The idea is that you should have one file "system/forceCoeffs" for each blade.
7. And for each time step, have an entry in "system/functionsDict", where each entry defines a new position and orientation, while including the "system/forceCoeffs" file for each entry.

I will likely only be able to give any additional answers around Thursday (holiday here in Portugal).