open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am refining the mesh near the shoreline up to 20 m resolution, and left the general area with 1000 m resolution. The near shoreline mesh disclose from the boundary with the abrupt changes of resolution from 20 to 1000 m, and stretch them to the finer resolution. First, is the stretch mesh will increase the cost of computation significantly, if so, how to resolve this?

Thanks, Amyrhul

Hi Amyrhul,

This is due to a limitation in Blue Kenue.
The outline resampling algorithm doesn't respect the
"growth ratio" parameter. Work is currently underway
at NRC to address this issue.

In the meantime, the workaround that I use is to enable the
"Resample Lines Only" option to create an outline with the
desired density. I usually use a PointSet density object for
this because it gives a continuous, smooth virtual surface.

Then use this new outline to generate your mesh with the
"Resample Outline" option turned off.

Hope this helps...
Martin

Hi Martin,

Thank you for your reply. I need to reconfirm about the "PointSet density object". Is this the XYZ data, with Z stores the mesh density information? If so, I can convert the bathymetry data into it, but should the data overlaps and passing thought the shoreline to produce a smooth Resample Outline?

Thank you, Amyrhul

Hi Amyrhul,

Yes, you can supply a xyz data object (PointSet) where the Z gives mesh density (actually edgelength). This is used as a virtual surface by the mesher via Inverse Distance interpolation. You can use any xyz data object as long as the Z values are in units of "Edge length".
Because the interpolation is Inverse Distance, you don't have to extend the data outside of the outline.

Hope this helps... martin