Error occurred while adjusting topology to create cohesive cells

PyLith
1

Dear everyone,

I am sorry if I am back again with the topology error. My previous problem Unable to find final slip is gone but then it is back to the topology error.
I have tried to simulate both step02 and step03 3D subduction. ( json file for both attached below ).
json file.tar.gz (26.4 KB)

I have different geometry (smaller than current geometry) and they worked well. I extended the geometry and starting to have this topology problem again. Previously, I can fix the problem by changing the suitable mesh size (for previous geometry was 5km. Higher than 5km, it didn’t work). I have tried also to remove the fault edge for the slab as suggested in the previous topic Values in UniformDB not specified - #10 by willic3 and changed the mesh size (smaller to coarser size). But it doesn’t work as well. ( I also attached the mesh below).
mesh_tet_alvina.exo.tar.gz - Google Drive

It seems that the problem appears each time it reads the boundary at z negative without the slab. Then the topology error at slab_top shows up. I have checked the boundary condition during the mesh generation but there is no problem with that.

Is there any way to fix this?

Thank you in advance.
Alvina KK

2

Your mesh has some local regions with very fine discretization. I suspect that this is causing the topology problem. Start with a relatively uniform resolution mesh. If you want higher resolution near the fault, then slowly add that in after the uniform resolution mesh works.

mesh
mesh739×619 278 KB

3

Dear Brad,

Thank you for your reply. I have defined the size for the surface only with a uniform size. But coreform(trelis) seems to create finer mesh at several areas.

I tried to use the uniform mesh size for all the volume larger than 10km and had the same topology issue. I have tried also using a uniform mesh size ~5km but Pylith stops working and closes the session by itself. It seems due to the big size of the memory. That is why I try to vary the mesh size.

Alvina KK

4

Dear Brad,

I did a smoothing for the topography before creating the surface in Coreform and the topology issue has gone.
It seems that the finer detail of the topography/bathymetry resulting the topology issue.

Thank you in advance.
Alvina KK