[kaaja]

Hi,
Does anybody know how interDyMFoam check's for convergence?
Does it use the moving part of the mesh only?
If not, it seems to be a challange to get convergence, since the other parts of the mesh changes size all the time.

[Tobi]

Hi, please be more precise. In Foam we move the mesh in each time step according to the motion given by the used library. After that, we calculate all other quantities till they are converged (if you set-up your solver correctly). The introduced mesh fluxes are taken into account.

[kaaja]

Thanks for replying!

I have a mesh of a water tank, with a boat on top.
The boat starts from the left part of the mesh and moves to the right.
Same type of movement as in the movingCone-tutorial.
In the beginning there are many small cells to the left of the boat.
So there will be e.g. many velocity vectors for a small area on the left side.
As the boat moves, the area between the left boundary of the mesh and the boat will be covered by the same cells as in the beginning, but with each cell covering a larger space.
How will openFoam compare the velocities for the area to the left when the cells are not covering the same space as the mesh moves?
The point location would no be identical anymore.

[Tobi]

Sorry I don't get your point.

[kaaja]

How are U at two different time steps compared?
Is the coordinate system relative to the moving object, so U is defined as the position relative to the moving object at all times?

[kaaja]

I think one thing I did not understand, is that interDyMFoam is a transient solver. I was thinking of it as a steady state solver.

[jessica_ict]

Originally Posted by Tobi

Hi, please be more precise. In Foam we move the mesh in each time step according to the motion given by the used library. After that, we calculate all other quantities till they are converged (if you set-up your solver correctly). The introduced mesh fluxes are taken into account.

Hi,
Could I ask a question?
how to check the convergence of body motion?