Rob, I think you can measure the gap by using BRepClosestPoint. Project the offending corner onto the other quad, and it will tell you how far it had to travel. If you get the new version you can then put those deviations into a BarGraph component and see what the distribution is like :)
-- David Rutten Robert McNeel & Associates On Oct 10, 5:21 pm, robhenderson <[EMAIL PROTECTED]> wrote: > Hey Patrick, David > > Thanks, guys. > > Patrick, thanks for your post. And as beautiful as your's is, I would > need a simpler solution due to budget constraints (this one is > supposed to be built...) > > David, what I've got here is two edges and one point of each panel > which are not coincident, and the normal of each panel is based on a > three point plane from the 3 remaining points. This gives a better > result than just using a plane normal to a point on the surface. I'm > still intersted in knowing what the gap at the open corner of my > current model is, because there would be tolerances in a glazing > system that we could work with. > > Anyhow, what's the basic logic for creating planar, non-rectangular, 4- > sided panels. I'm guessing they are almost rectangles, because my > tolerances are pretty low. > > Thanks again. > rob > > On Oct 9, 9:37 pm, Patrick <[EMAIL PROTECTED]> wrote: > > > Rob, > > > I have posted an example of converting a doubly curved surface into a > > set of flat panels with triangles filling the gaps. I am not sure if > > this relates to what you are trying to do, but just in case, here are > > the links. > > >http://grasshopper3d.googlegroups.com/web/flat-panels.pnghttp://grass... > > > Patrick > > > On Oct 9, 6:18 am, robhenderson <[EMAIL PROTECTED]> wrote: > > > > Hi All > > > > I have panelized a fairly simple, yet doubly curved surface with > > > planes based on a variable 2d interval. Nothing extreme, but I end up > > > with at tolerance (read gap) at the bottom left corner in my case. I > > > would like to be able to measure that gap, but have been unable. > > > > The gap is always between pt2 (series 'cull 2') and pt4 (series 'cull > > > 4') of diagonally adjacent panels (obvious in the file). I've tried > > > re-culling the cull 2 and cull 4 series again; tried sorting from the > > > original applied 2d interval; and tried 'nearest surface pt' to create > > > another series of pts to measure to. . > > > > The re-culling gave the closest results but did not always connect the > > > two closest points, which is simply what I need it to do. Unless I'm > > > missing a 'nearest point' button or something else obvious, what's the > > > best way to solve this. I hope y'all see the other possibilities of > > > precision and testing this would open up. > > > > I posted the file "planar panels tolerance measurement needed.wrm" > > > with the group > > > files:http://groups.google.com/group/grasshopper3d/files?hl=en&&sort=date > > > > rob > > > > p.s. David, I've gotten the 'rhino resource was...' crash several more > > > times. I don't know much, but I'd go with the buffers... good luck.
