CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Pointwise & Gridgen (https://www.cfd-online.com/Forums/pointwise/)
-   -   Bad Single Cell Projection on a Quilt! (https://www.cfd-online.com/Forums/pointwise/236716-bad-single-cell-projection-quilt.html)

pdp.aero June 11, 2021 13:52

Bad Single Cell Projection on a Quilt!
 
2 Attachment(s)
Hi there,

Dear Pointwise's folks,

Please help me figure out this issue because I am tired and out of ideas.

I have a structured surface mesh which is very very fine and I am projecting the surface mesh on a quilt before or after assembling a block in a very classical old fashioned way by just selecting 6 surface meshes.

The problem is that only one cell on the surface mesh jumps or stretches in a very bad way at some point far away from its original position over the plane and creates a series of negative volume cells.

Please check attached pics.

This happens only if I project the surface mesh. If I don't, there'll be no problem.

I used closest point + project points like this:

pw::Entity project -type ClosestPoint -interior [lindex $bdom_lastb($i) 1] [lindex $brackets $i]

Do you have any ideas how to project this domain on the quilt without making that particular cell stretches in that atypical way?

This is a very very fine grid I have 886,495,296 cells only on this bracket.

Cheers,
Pay

pdp.aero June 22, 2021 09:13

Quote:

Originally Posted by pdp.aero (Post 805882)
Hi there,

Dear Pointwise's folks,

Please help me figure out this issue because I am tired and out of ideas.

I have a structured surface mesh which is very very fine and I am projecting the surface mesh on a quilt before or after assembling a block in a very classical old fashioned way by just selecting 6 surface meshes.

The problem is that only one cell on the surface mesh jumps or stretches in a very bad way at some point far away from its original position over the plane and creates a series of negative volume cells.

Please check attached pics.

This happens only if I project the surface mesh. If I don't, there'll be no problem.

I used closest point + project points like this:

pw::Entity project -type ClosestPoint -interior [lindex $bdom_lastb($i) 1] [lindex $brackets $i]

Do you have any ideas how to project this domain on the quilt without making that particular cell stretches in that atypical way?

This is a very very fine grid I have 886,495,296 cells only on this bracket.

Cheers,
Pay

Okay.

I solved this.

I am not sure why Pointwise didn't like the way I was assembling my blocks and projecting domains over the quilts. But I rewrite that whole part of script in a different way and now there is no problem.

I dunno if this problem happened because I had a very fine surface mesh being project on a quilt, or if there was a bug in my script, or it was something to do with the CAD file. All I know, I haven't change the CAD and got it work by re-writing the script in a different way and assembling smaller blocks each time.

I got to generation total: 3,267,183,584 cells on my desktop with 64 GiB RAM.


All times are GMT -4. The time now is 07:14.