3 Corner Boundary Surface...help please

free edge controlled by a sketch.

prefered) to the adjacent surfaces.

always get a "pinched" surface at one of the corners which means I cant thicken the resultant

less than R=0.090" yet the surface created gets pinched to a curvature of R=0.002" at one of the

Use a Fill surface instead.

Or fiddle with the curve influence settings, but the fill is the better idea.

Daisy.

I'll look at it or just post the boundary (copy surface, three surfaces, export as a parasolid). But, yeah, 3 adjacent edges don't always work well so, you might get lucky but.., somehow,..making it 4 or more is desirable to resolve this. If you're fixed on the need to have just 3 sides,.. you'll have to look at the adjacent boundaries (as you probably have tried) to see what tweaks/cheats/curves,... can be added to help with the corner(s).

=2E.

is a free edge controlled by a sketch.

prefered) to the adjacent surfaces.

but always get a "pinched" surface at one of the corners which means I cant = thicken the resultant

of less than R=3D0.090" yet the surface created gets pinched to a curvature= of R=3D0.002" at one of the

I Second trying the Fill surface. Just be sure to turn off 'optimise surface'. If you eave it on, Fill makes a Loft. Turning it off unleashes a powerful patch algorithm that was brought over from Catia that will give you an 'un-pinched' four-sided surface to fill your opening. Let us know how it goes (what you do is very dependent on surrounding geometry, but Fill is better about 90% of the time) Ed

Thanks you all, especially Paul who showed me how to do it. It was eventually done by making 2 surfaces, one a boundary and the final one a patch. This surfacing game can make you pull your hair out sometimes, that tiny surface took longer than all the other surfacing on this project. You live and hopefully learn :-)

--

--------------------------------- --- -- - Posted with NewsLeecher v3.8 Final Web @

I'm glad that Paul helped. The sentence above struck me as needing another comment - the tiny surfaces almost always take longer. When doing a project my modeling plan includes trying for nice, 'roomy' surfaces with a lot of space to resolve themselves. When doing Loft, Fill, or Boundary, you are giving SWx some inputs then telling it to fill in the rest. If there is any innacuracy/ inconsistency at all in the inputs (and for just about everybody that I have seen there usually is something off, even if its a few decimal points to the right of the 'optimal' input) SWx needs space to smooth the innacuracies out. In tight spots everything has to be spot-on or it will fail, fail, fail.

Keeping away from small patches or three sided patches is really important. I will often use a trim or split line to cut a big, rectangular hole around final junctures and use Loft/Fill/Boundary on the new, manufactured, artificially large hole to avoid the issues of smaller or three sided holes.

Ed

In one of the SWW classes I took it was stated that if you have a surface that goes to a single point, it's called a singularity point and the solver has to make some compromises to make that work over a 4 sided surface. try to keep surfaces to at least 4 sides.

WT

I'm glad it helped. These cap/patches always need a few things to get it to resolve nicely, otherwise, they get all squirrely. BTW, in this case, the boundary is actually 2 edges and adding split lines to the 2 edges, (which make it 4 edges), helps the fill surface resolve cleaner. (also,.. in this case,.. Optimize (on/off....with tangent or curvature) resolves the same, no loft).

=2E.

urface took longer than all the other surfacing on this project.