Workflow? Wire-frame first?


Although a long time Rhino user, I am not a seasoned modeler. I use Rhino only occasionally. Currently I am modeling a tablet case, with the goal of actually producing it, at least in theory. So I take care about dimensions, availability of materials, and price. All in all, it is an interesting exercise.

My workflow is normally like this:

  1. Create wire-frame, i.e. only using curves (in extreme cases this can involve doing some surface operations followed by DupEdge).

  2. Create surfaces from wire-frame.

  3. Merge surfaces into solids.

  4. If something needs to be modified, go back to step 1, editing the wire-frame.

In the current file, I have pairs of layers. For each wire-frame layer, there is a matching surfaces layer. Also, I have some layers with help lines.

How do seasoned users approach modelling? Like that, or do you work directly with surfaces as much as possible?

I can imagine that this question has been asked before on the newsgroup, but it isn’t yet on the Discourse. :wink:

(Pascal Golay) #2

Hello - I guess I would be wary of always working from the wireframe directly- it may not always make sense to do this if you can get an understanding of the underlying surface shapes. For example, if one of your surfaces is let’s say for the sake of illustration, spherical- maybe it is only a small portion of a large sphere- nearly flat, with some arbitrarily shaped outline where it meets other surfaces… making this from the surrounding wireframe might be a messy process, but making the sphere and trimming it may be simple and exact. I usually try to determine what the basic, larger underlying shapes are and and then decide on how these should be made. Another consideration is that curves derived from trimmed surface edges are frequently not the best input for making new surfaces. They can be quite dense in points and not as smooth as you might want or expect.


Thanks for your input, @pascal! Don’t get hung up too much on the DupEdge. I have only used it on some rare occasions (extreme cases) in the past, and I am well aware of the limitations. That was more more to make the wire-frame look better. I wouldn’t necessarily use that part to construct a surface, but instead proceed like in the example that you mentioned.

My question is more if the pros keep a wire-frame like I do: For each surface layer, one wire-frame layer. Or something like that. Perhaps I’m wasting my time?


I’m not this fastidious (read: I’m much more lazy). I keep one layer which I call Dead which is always off and locked that has the input curves or base surfaces of things that might be difficult to reproduce down the line, but it often isn’t that much. More often I just save many iterations of the file. Any major change I’ll do a save first, then a save as before I start hacking into something, just be consistent with your file naming so you can find it later, for me it is usually something like [projectNumber]_[projectName]_[quickNote]_YYMMDD.3dm

I don’t do everything as curves first, certainly not the whole project. I may lay in some overall guide lines which describe the project as a whole, but for the most part I am working in curves just to get the next surface I’m building, not to mock up the whole thing in curves first. Once that surface is built, I’ll get rid of the curves I no longer need, or throw them on the dead layer if I think I’ll need them later, so they don’t get lost. I prefer using untrimmed surface edges more than the input curves because they can be blended to, matched, or used as rails which have curvature information. A lot of it you just learn from enough times of saying to yourself “I shouldn’t have done that.”


Thanks for the insight into your workflow @SamPage - very helpful!

By the way, for software development, which I do also professionally, there are plenty of guidelines and opinions concerning how to structure workflow, how to document, how to do versioning, etc. Developers can get quite worked up on that topic.


It is rare that a 3D model needs to be explained in much detail to a new modeller on the project. So documentation is usually lacking, unlike for software development. “A picture says a thousand words.”

My layers are similar to @SamPage, but I keep a “construction” layer for building bits in then distribute to finished layers as the parts are completed. Under “construction” there is a “used” layer for lines and other stuff that is used to build the parts. If the parts are complicated I’ll add a number of used layers to keep things in some sort of order. The “used” layers are locked and off.


Thanks, @Ncik! Note that I keep my curves not for documentation, but more as a history, to have a clean starting point in case I need to make a change. Thinking about it, that often seems superfluous: For example, if there is a planar surface, then I believe I can reconstruct the exact original curves by doing DupEdge.

(Willem Derks) #8

Also Many times I can retrieve original surfaces by using Untrim* (in it’s various flavours)



Just used it - very handy. Thanks for the reminder!