Jump to content

mmyoung

Member
  • Posts

    155
  • Joined

  • Last visited

Posts posted by mmyoung

  1. We've discovered that the Unfold Surfaces Tool fails at a crucial point in a sculpture we are fabricating for an artist here.

    I've included two .vwx files. "3 turn spiral A.vwx" is the one we fabricated from. The other file, "spiral mmy 01b.vwx" is another attempt with the same geometric result (unbuildable).

    The red unfolded surface in "3 turn spiral A.vwx" is the one that pulls away from the structure about half-way up, though the other two surfaces seem to be fine.

    Attached you'll see two photographs ("unfold error 1.JPG" and "unfold error 2.JPG") from a complete set of documents showing the entire process, in which we attempt to fabricate a small-scale mock-up for 30' tall sculpture. Our complete documentation set shows exactly where the part failed, and we can send it if it will help.

    The last photo ("spiral fabbed.jpg") shows a completed spiral in which the piece that failed was re-fabricated by hand (tedious and difficult even for our crew). The final, full-scale piece cannot tolerate even minute irregularities, however, and building it at full scale would be too risky, so we have not yet taken the project all the way to completion.

    Any assistance or insight would be very welcome!

    Has anyone else had any experience with the Unfold Surfaces tool like this?

  2. Hi Michael K ?

    modo is a very competent subD modeler with a gorgeous rendering engine and a good animator (no rigging though), that was started by some refugees from Lightwave. It's got some of the best tutorials in the business, made by their on-board genius, Andy Brown, which apply to any subD modeler.

    The modeling environment isn't as fast or intuitive as Silo, but it has excellent spline patch modeling and a suite of functions that make it a better hard-surface modeler in many respects, though the snapping functions in modo are a little strange. Actually they are real strange. Silo's snapping is straightforward and very good, but it lacks rotational snapping.

  3. I suppose it's worth adding, having mentioned SuperNURBS, that whoever developed that is a *&^$#! genius, and I wish it had worked. You could pull and push on control points, just like you do on an interpolated surface, only it was *easy* to do: You didn't have to engage in a prickly struggle with NURBS control points, and neither did you have to deal with oddball sub-D cage geometry; you just built something and started tugging on it. And the result was parameterized.

    This strikes me as substantially the fusion so many developers are seeking between direct poly modeling and parametrics.

  4. tallboy, subD has to do with modeling a 'cage' or polygon mesh, which can be 'subdivided' (hence the subD moniker); as the mesh is subdivided, it collapses towards a 'limit surface,' which I believe is an interpolated NURBS surface, though I'm not sure of that. I haven't done the math. Electric Image's modeler, before they started bundling Silo with their app, had a feature called 'SuperNurbs,' which was mind-blowing but sufficiently flawed that they abandoned it (I have heard they might revive it). It allowed you to do subD-style 'character modeling' and end up with a full-parameterized NURBS surface you could hand off to a CNC miller, so I understand. I worked with it a little bit and loved it and hated it because it blew up so much, but this is where I got the idea that "Catmull-Clark subdivision surface modeling" (read about it in Wikipedia for a quick upload) produces surfaces that are inherently parameterized, i.e., every point on them is defined mathematically and resolution-independent.

    Yes, you could model something in Silo (excellent, free/cheap [depending]), Wings (free, a little balky), Blender (free and very sophisticated, multi-platform), or of course 3DS or Maya (fancy, very good, and expensive), then port it to STL (which VW handles very nicely), or you could get it 3D printed. This "maquette" or small-ish 3D model could be carved into, drawn on, added to, or otherwise manipulated, re-digitized with a 3D scanner and re-manipulated digitally, and so forth, until you have what you want.

    NextEngine:

    Look at:

    http://www.youtube.com/watch?v=ggvzcGdZsTc

    and

    and

    http://www.popularmechanics.com/automotive/jay_leno_garage/4320759.html

    (reprise with text)

    A nice aspect of working this way, is that you can port your model back into VW and leverage its strengths: Work out your armature and supporting hardware, get full dimensioning and shop drawings, make a bomber presentation to your client, fit up the rebar, picking eyes, mesh attachments, and all manner of other details. This would give your sculptor something to work from and take a lot of the dog work out of building the thing, and it gives you dimensioned points to work off of, which is probably going to be important if you want to install it somewhere tricky.

    Now you have the best of parameterized control (VW) and free-form creation, and you can predict the outcome reasonably well, giving the model and accurate dimensions to a diligent sculptor or someone like Digital Stone or Talix (ha ha the range of money here really covers the waterfront).

    This is one of the bleeding edge frontiers of 3D, in my opinion. Keep at it!

    *Ideally*, you'd model the thing up and pass it to someone with a large-scale, 5-axis gantry mill.

    Or you could work out something like the holy grail, which would be large-scale deposition forming, i.e., 3D 'printing' out of structural or at least permanent material (the structural bits might be the armature inside it).

  5. If you even considered modeling an angel in VW, you'd probably be willing to try it in subD. Silo is an inexpensive and excellent subD modeler, free trial, and it talks to VW through the .3ds format. It would be tedious work, but it might at least point you at something you'd like to try.

    subD modeling takes getting used to; there's a wicked 'knack' thing in the learning curve.

  6. Ah ha... either everyone is being very kind by not pointing out what a dufus I am for not being able to figure this out... or maybe no one else is trying to do this. hmm. It's still really important, that we can fillet NURBS etc. Any comments appreciated.

  7. The Help file says, with regard to the Fillet tool:

    "Apply fillets to lines, rectangles, polygons, polylines, circles, circular arcs, NURBS curves, 3D polygons, and walls. With rectangles, polygons, or polylines..."

    I can fillet lines, arcs, circles, and polygons.

    But I can't fillet NURBS, polylines, 3D polygons, or composed objects.

    Is there a trick to this, or is the Help file mistaken? This seems like pretty important stuff, be be able to fillet a NURBS or polyline to a line or arc, for instance.

  8. One you've rotated your model into an axo view (non-perspective) that you like,

    - Go to the "Visualization" tool set, then:

    - Get into perspective mode quickly by choosing the Walkthrough tool. (But don't move the cursor around.)

    - Now choose the Translate View tool. You can slide the model around in the perspective window with it.

    - While you are in the Translate View tool (in perspective), hold down the Alt/Option key. Now click and drag left and right, up and down.

    - Alt/Option Drag LEFT moves the model into the distance.

    - Alt/Option Drag RIGHT move the model closer.

    - Alt/Option Drag DOWN increases the perpective effect.

    - Alt/Option Drag UP decreases the perspective effect

    - Dragging UP and DOWN changes the angle of the lens, controlling how dramatic your perspective is. Drag DOWN to widen the lens, making the perspective more dramatic.

    - Draging DOWN to increase the perspective effect also makes the model appear closer. Drag LEFT to decrese its size.

    - Dragging UP decreases the perspective effect, but makes the model seem smaller. Drag RIGHT to increase its size.

    To get out of perspective, go to View > Perspective > Orthogonal. If you do a lot of perspective, you'll want to use a shortcut for that.

  9. Oh yeah, the three tools in this group that consistently have trouble in our shop are:

    ? the Trim Tool, which frequently removes the whole line instead the segment I clicked on,

    ? the Fillet Tool, which consistently trims off the wrong part of the object (in a very nice, thorough response from tech support, after I submitted a reproducible bug, thank gosh, I was advised that a workaround is to Compose the object I'm filleting).

    ? the Connect/Combine Tool, which often fails by removing the wrong portion of a line or connecting the wrong ends of spline

    & You're right ? these are fundamental drafting tools that malfunction. I'm kind of jumping up and down about this, because it costs us a lot of time. I resort to the Split Tool a lot to sort this out, and that tool never fails, but it's tedious to have use that when the other tools ought to do it. I've sent in some bug reports, and I'm pretty sure they're being worked on. I hope they get to the top of the queue. It's important.

    These issues come up a lot more if you are doing complex drafting with splines and so on.

  10. Tom,

    When I compose an object using arcs and straight line segments, so far offsetting them seems to be fine, they offset like any other object. I may not be pushing it too hard, but it seems all right. In 2008, I didn't have any trouble with this either.

    Invoking the 2D Reshape Tool doesn't reveal a welter of tiny segments, just a few polylines, and they are easily editable... so far.

    Decomposing the offset polylines seems to cause the addition of perhaps one or two new arc segments, but nothing prohibitive.

    Is that what you were asking?

  11. I am hoping the Parasolids kernel will affect the functioning of these 3 crucial tools when it gets fully implemented. At present, I consistently have failures with 2D filleting, trimming, and combine/connect. I've sent in a few bug reports, some of which were well acknowledged ? but 2009 fails in the same ways 2008 and v12.5 did in this regard. I'm sure NNA is aware of it and working on it, but I'd be glad to send everything again, if that would help. It's a really important fix, I think.

    I realize these are not as important to architecture as they are to machine design and ID, but now with the fully capable modeling engine, the drafting has got to come up to speed. Thanks for giving us a hand on this ??

    Many thanks, as usual.

  12. Thanks Pat! That's working out really nicely for us.

    The way we control the tilting and banking of an extrude, in order to extrude a fancy profile into a railing or trim, is to make a loft of many carefully placed profiles. Wouldn't it be good if you could specify that the profile maintained its orientation, so for example the "top" of the profile would remain normal to the Z axis, or something like that. Would save hours.

×
×
  • Create New...