Computing mapping question in 3D simulation

Using preCICE
,
1

Hi everyone,

I find a strange question about mapping mesh, I tried to turn the perpendicular-flap tutorial from 2D to 3D. I change dimensions="3" and include all nodes of coupling surface in structure solver. I receive error as :
---[precice] e[0m Computing "partition-of-unity RBF" mapping from mesh "Fluid-Mesh-Centers" to mesh "Solid-Mesh" in "write" direction.
---[precice] e[31mERROR: e[0m Too many vertices have been filtered out.

I think problem is about mapping one array of data from OF to two array of data from structure solver. Because in FVM, force is computed in the center of mesh, so when there is only one layer of mesh in z-direction, OF will send one array of mesh data (mesh center) to precice. But in structure solver, data are all conserved in nodes, so even with one layer of mesh in z-direction, there are two array of mesh data (node points) in z-direction, as shown below.

图片
图片173×708 20.4 KB

In quasi-2D simulation, I convert the force from OF into two equal piece, and allocate them to each of these array (Although I don’t know if it is right choice). But in 3D simulation, how should I solve this problem?

Thank you in advance!

2

@DavidSCN The error originates from the partition of unity mapping.
It seems like the mapping doesn’t tolerate fully filtered-out partitions.
Why this happens can have many reasons though.

@Gong_CHEN You could try using the nearest-neighbour mapping. The preCICE output contains a summary on the distribution of distances. This can be helpful to get a better understanding of the problem.