Discussion:
2d tri vs. quad for element based rigid surface
minki
2011-05-09 08:12:33 UTC
Permalink
Dear all,

Does the surface using 2d tri give different results compared to that using 2d quad, in projecting the element based rigid surface into deformable strucutre?

Regards,Minki
Dave Lindeman
2011-05-09 14:21:11 UTC
Permalink
I assume you're talking about R3D3 vs. R3D4? Yes, you'll get somewhat
different results since the surface (if it's curved) will be represented
differently. The difference should depend on the mesh refinement level
and the degree of curvature.

Regards,
Dave

Dave Lindeman
Lead Research Specialist
3M Company
3M Center 235-3F-08
St. Paul, MN 55144
651-733-6383
Post by minki
Dear all,
Does the surface using 2d tri give different results compared to that
using 2d quad, in projecting the element based rigid surface into
deformable strucutre?
Regards,Minki
------------------------------------
minki
2011-05-10 03:44:40 UTC
Permalink
Hi Dave,

If the 2d tri represent better curvature for my complex geometry (assuming it acts as rigid body), then it is good to use 2d tri rather then 2d quad ?

Regards,
Minki
Post by Dave Lindeman
I assume you're talking about R3D3 vs. R3D4? Yes, you'll get somewhat
different results since the surface (if it's curved) will be represented
differently. The difference should depend on the mesh refinement level
and the degree of curvature.
Regards,
Dave
Dave Lindeman
Lead Research Specialist
3M Company
3M Center 235-3F-08
St. Paul, MN 55144
651-733-6383
Post by minki
Dear all,
Does the surface using 2d tri give different results compared to that
using 2d quad, in projecting the element based rigid surface into
deformable strucutre?
Regards,Minki
Dave Lindeman
2011-05-12 14:39:00 UTC
Permalink
I wouldn't necessarily say that either representation is "better". The
triangles will at least have a unique normal, whereas the quadrilaterals
could be warped. The contact algorithm will normally smooth the
surface, so there probably won't be too much difference.

Regards,
Dave

Dave Lindeman
Lead Research Specialist
3M Company
3M Center 235-3F-08
St. Paul, MN 55144
651-733-6383
Post by minki
Hi Dave,
If the 2d tri represent better curvature for my complex geometry
(assuming it acts as rigid body), then it is good to use 2d tri rather
then 2d quad ?
Regards,
Minki
I assume you're talking about R3D3 vs. R3D4? Yes, you'll get somewhat
different results since the surface (if it's curved) will be represented
differently. The difference should depend on the mesh refinement level
and the degree of curvature.
Regards,
Dave
Dave Lindeman
Lead Research Specialist
3M Company
3M Center 235-3F-08
St. Paul, MN 55144
651-733-6383
Post by minki
Dear all,
Does the surface using 2d tri give different results compared to that
using 2d quad, in projecting the element based rigid surface into
deformable strucutre?
Regards,Minki
------------------------------------

Loading...