displacementLaplacian solver: move patch-nearest cells
 

Hi all,

I'm using mesh motion with the displacementLaplacian solver and a modified interDymFoam solver. During runtime, the points on the bottom boundary patch, i.e. in the boundaryField, are modified and used as a boundary condition for the whole mesh motion.

As my mesh somewhen gets distorted (I guess due to the fact that only the patch motion is given as a boundary condition), I'd like to move the cells (centroids) connected to the bottom patch as well, before the mesh motion solver is starting its job.

Is this possible, and if so, how? My problem is that these cells are of course in the internalField, whereas the motion of the other cells of the internalField is not known at this stage.

Arne

Hi Arnoldinho：
Do you solve the boundary motion problem? I want to realize that a boundary deformate due to the caculated result of the last time step, do you have some advice?
Thank you!

bojiezhang

Hi bojiezhang,

I'm sorry to say that I'm not using the mesh motion solvers any more. I never got good results due to distorted meshes and therefore wrote my own "mesh motion" using some distance functionalities to move inner mesh points in vertical direction. This has been implemented directly in the solver, using mesh.movePoints(newPoints).

Arne

Hi Arne,
I think we are doing the similar job. I want to model the
morphology of the terrain by the current scour. I hope you could help me. I want to ask some questions below:
(1) does motion mesh work for the morphology/scour?
(2) can you tell me something about disort mesh? which solver do you use and which type of dynamic mesh do you use?does it works well?
Thank you in advance!

bojiezhang

Hi bojiezhang,

if you want to model sediment transport and scour formation then yes, we are doing a similar thing. Concerning your questions:

1) Every mesh motion solver should to some extent work for this, as you have to calculate the bottom patch displacement/movement and use this as a boundary condition for the inner mesh displacement/movement (then e.g. using a motion solver).

2) For the inner mesh movement I so far tested almost all available solvers. laplaceFaceDecomposition did the only reasonable job, but as I am modeling scour around objects, it produced highly small cells and therefore the time step significantly decreased. It does not give you the full control on how the inner points are moved. Furthermore, it is really slow! All other solvers did not work, as inner mesh points did not move correctly, so sometimes inner points moved below bottom patch points, so the mesh got distorted.

Therefore I am, as said, directly calculating the inner points movement and therefore have full control over them.

I hope this answers your questions.

Arne