Easy surface but...

[ilmegor]

I know it may sound ridiculous, but I cannot make a surface within a closed, continuous (C1), non planar curve.


If I use the boundary command it says lines are tangent (make it non-tangent!)


I'd like to select the boundary and...make it. Is It possible? What command and procedure shall I use. I have WFII.


Thanks for your help.


Ilmegor

 

[Zaki]

Hello
Can u send me the sketch of part or part that I want to see what to want to make???
Zaki
[email protected]

 

[ilmegor]

Thanks for replying.


Say, for example I want to "cover" the slot in the image with a surface. I'd find "natural" to select the 4 edges and make a boundary... but it does not work because they are tangent! Of course there are other ways...but it sounds strange we cannot do it "directy" with one command.


ThanksView attachment 940

 

[Moroso]

If that hole is contained in 1 surface you could click on the surface, do a ctrl c & v to copy then exclude the hole from the surface.



Edited by: Moroso

 

[ilmegor]

thanks.





This is a simpliefied case thou. The more generic question would be:


how can I make a surface with non-planar but tangent edges of whatever shape?

 

[jeff4136]

Some basic understanding of NURBS surfaces might help. All NURBS surfaces have four edges. The closer the edge intersections (and the U, V spans) are to being at right angles, the happier the surface will be. Most programs have trouble creating and manipulating (offsets, for example) NURBS surfaces with singularities (zero length edges) and degenerate edges (the edges 'degenerate' into a single edge; e.g. they are tangent to one another).

So, in some circumstances you'll want to create analytic surfaces (say, a revolved arc) when a singularity is needed or a four sided surface (by any means) and trim if degenerate edges are needed. (I'll usually do that for singularities, also.)

 

[ilmegor]

Thanks Jeff,


This helps! Now I understand why I have problem with those surfaces.


if you have time, can you explain a little more in detail what you would do in case you have these singularities. Thanks a lot.

 

[jeff4136]

Well, I have only a wispy grasp of the theory and probably a little less of practical application, but ...

Singularities (three sided surface with one zero length edge or two sided surfaces with two zero length edges); Pro/E seems to handle a lot of them pretty well. Depends on boundary shapes, whether or not you are forcing G1 or G2 continuity, etc. Sometimes they will wrinkle or crease going into the "points".

Degenerate edges (where the edges aren't G0 at the intersections); I don't think Pro/E will even create them (gives you the "edges are tangent" error) unless you help it out a little by connecting vertices with some cross curves to break up the bounded region.

If I'm concerned with surface quality I'll try to trim existing surfaces to get a four sided boundary. Otherwise I'll go for a quick and dirty solution. I think if you'll merge the surfaces in the first group of patches and offset a relatively large distance (1.5 or so) you'll see that the quilt can be offset without failure but a quilt from the surfaces in the second group will fail. Doing an offset analysis on the patch surfaces will show why (look closely at the singular ends).

2005-06-09_060657_wf2_patch_examples.prt.zip

 

[ilmegor]

Thanks Jeff,


That's makes it more clear to me!!


This problem is solved!.......another one will come up though


Thanks again.