I cannot get the filled surface tool to fill an elliptical aperture with two (elliptical) guide curves. It works if I select one of the guide curves but fails when I select the second. This is not a real part, just a test.
SW 2011, SP3
I had to make a soap bar mold cavity recently that had a similar shape to this. Here is how i did it. I know it is not using the filled surface but it gets something close to what you are wanting I think.
Timothy, that's quite close but there is a tangency problem at the mirror joins. However, this seems to be cured if you choose "Normal to Profile" for the start and end constraints. I still think filled surface should work.
Stuart, I'm not sure why that doesn't work, but ellipses can sometimes exhibit odd behavior like that. I've attached a version that takes advantage of the symmetry of the part--though you could also accomplish this slightly different with a part that didn't have such symmetry by breaking the fills into pieces if necessary.
+1 for Jeff's method.
I try and not let surface fills curve beyond 90°. Because this curves back 180° in two directions, I am not surprised surface fill doesn't work. Also, surface fill often gives bad results when you have intersecting constraint curves (although this is not a hard-and-fast rule).
yes, Boundary has problems with two or more crossing constrain curves. My best advice is as others have recommended would be to do one quarter. An even better solution is to use boundary or sweep solution instead. If it is a perfect ellipse that you want with elliptical cross-sections then consider Boundary or Sweep. See attached file and observe three different configurations. I usually would not recommend the boundary method because you could be setting yourself up for a de-generate point issue (turn on Display>Curvature to see it) at the top of the dome. Fortunately our Shell algorithm is getting better every release and when you have degeneracy in low curvature areas like these, the sheller has no problem with it.
EDIT - BTW - I'm not sure why Fill has this limitation especially when it is ellipse constrain curves. I'm traveling this week but when I get back in the office I'll ask the developer who created the Fill feature and get his feedback.
Thank you all for your comments and examples. The reason I had assumed that Surface fill would work is that some months ago, I took the surface modelling training course that our VAR was running and part of that course included modelling a remote control. One of the aspects of this modelling excercise was removing a planar face and replacing it with a face which is curved in two directions using two guide curves in much the same way as I have tried with the ellipses. I can only assume, as has been suggested here, that surface fill just isn't happy with ellipses. I see what Charles is saying about angles but it does fill if you select only one of the guide curves.
Timothy's method seems to work well if you choose 'normal to profile' for start and end conditions but I don't know how precise the elliptical shape is. I like Jeff's method also, using the surface extrudes to get the tangency.
Anyhow, as I said, this was only an excercise, trying not to forget too much of what I had learnt on the course.
..this is after the fact and I don't have the latest release (2011) but I can view the model..
btw, this was something we did way back when on some old forum,.. it's using a fill on 1/4 of the elip...
I'll attach how we did it back then for anyone who does not have the latest release and wants to have a looksee.. (btw, it's a SW2006 version) ..
Retrieving data ...