So I’ve been working in Solidworks for a number of years. My background is in Mechanical Engineering but I’ve been branching off and exploring some generative design concepts. I’ve been working with Processing with the intent to transition to Grasshopper and Rhino.
So my next step is to start learning Rhino and I was wondering if any of you had experience making the transition from Solidworks to Rhino. I know that they’re not interchangeable, I’m just looking to add some new tools to my kit.
If anyone has any advice or resources for learning Rhino, making the switch from Parametric to NURBS, that would be greatly appreciated. Thanks.
How much experience do you have with complex surface modeling in Solidworks? The process is actually very similar to Rhino in so much as creating surfaces from curves, surfaces or a combination of surfaces and curves. The only real differences are that you use different tools in the same situations (e.g. Solidworks’ loft tool is very powerful while the swept surfaces kind of suck. Rhino is the opposite.) and that Rhino lacks the powerful history editing capabilities of a parametric modeler. Find a couple quick tutorials on creating NURBS curves in Rhino (which is a thousand times simpler than complex splines and 3d sketches in solidworks), and you’ll be off and running.
To be clear, SW is also NURBS, it just happens to retain the history of those functions (macros). So the way things are built will feel very familiar. In other words, it’s not like moving to poly modeling, which can be an entirely different workflow.
The stuff that takes getting used to, for me:
No dialogs or wizards; all messages and prompts appear in the text box up top
It is very easy to accidentally be imprecise. Add this to the fact that the onscreen visuals are always approximated, so something that looks to be in the right spot might actually be a little off, unless you make sure that you’re snapping correctly.
Whereas SW will yell at you for every little invalid thing you do, Rhino will just let you live with your shitty geometry. This is really good for quick modeling, but terrible for diagnosing failures.
My mind is in more of a Photoshop mentality when I’m using Rhino. Meaning: liberal use of layers, duplicate+hide any object that may need its changes reverted. To me it’s a much more sculptural process; less step-by-step than SW.
No, “solid modeling” is only a theme for orienting the user interface and feature set. All CAD systems today use NURBS to build geometry, but a “solid modeler” never lets you do anything that leaves the model in a non-closed state. So the model is always at least theoretically manufacturable, but the workflow for making curvy shapes is unnecessarily convoluted.
Of course that definition is decades old and increasingly not entirely true, today “solid modeling” means history-based everything-plus-the-kitchen-sink mechanical engineering, and “surface modeling” means Rhino.
History(Rhino has history now too)doesn’t actually have anything to do with some attempt at a basic definition “solid” vs “surface” modeling, though of course (and there certainly were such things,) a rigidly “solid” modeler without some sort of history or parametrics would be kind of painful to use. I was saying that today “solid modeling” means whatever Dassault wants to call it.
The confusion is totally reasonable; “solid modeling” is sort of a disorienting term.
“Solid modeling” is, in a way, a fictional concept; it’s just a workflow method where the software continuously checks that your NURBS object is a watertight volume. If the object becomes non-manifold it either scolds you or it switches to traditional surface modeling. A watertight NURBS cube is identical to a “solid” cube.
Parasolid is NURBS too. The basic geometry in all CAD software is NURBS. That’s why the distinction is increasingly becoming difficult, at the core they’re the same thing. “Surface modeling” just means having the option to work with the geometry a little more directly.
Another way to think about it. All commercial CAD systems use definitions of the boundaries, the surfaces.
If you took the cross section of a CAD cube, and took any point inside the cube, there is no gemetrical information written for that specific point. The system only assesses if that point lies inside or outside the calculated solid.
If it was a real world solid, that point inside a wooden cube, would be wood. In a true “solid modeler” then each unit of space would have a definition. That file would be huge depending on resolution. Computations would not be fast.
Voxel modeling is the closest analogy. Scan and Solve uses this approach for FEA analysis.
I meant the concept of “solids” as some novel way of generating geometry is misleading, because it is. “Solid modeling” is a workflow method, controlled at the software level, not at the geometry generation level. New versions of SW aren’t just now incorporating NURBS features; they’re implementing NURBS ideas that have been around for ages (see: Alias), but SW is just now figuring out a way to implement them without wrecking the solid-modeling style.
So really the idea of Parasolid is not relevant to the concept of geometry calculation. Generically, Parasolid is a kernel which means it’s basically an interpreter. It takes the math of spline boundaries and says “Hey software, here’s what you’re dealing with” and then the software handles the display and manipulation of that data.
All of this put more succinctly by JimC5…
Good description by nxakt too
Sorry to the OP for taking this on a curvature-continuous tangency.
To answer your last question, I’m sure there are tons of tutorials out there due to Rhino being a more accessible program. My troubles going from Pro/E and ISDX to Rhino were that it felt ‘flimsy’, like I wanted to lock more things down. But certain surfaces and features were much easier to do in Rhino vs SolidWorks so sometimes we’d patch the two together.
Anyone remember CDRS? A great surfacing program (splines were hair-thin and renderings looked like photos), unfortunately Parametric Technologies (makers of Pro/E) bought Evans & Sutherland and ran CDRS to the ground.
That was the first CAD program for ID that I used, circa 1997. It was a stand-alone (at the time) partner to Pro/E. It ran on a crazy expensive Digital machine with top-of-the-line performance. I changed jobs from a pretty fun agency to a dysfunctional corporate gig just to use that software, and learned it quite well.
Isn’t CDRS essentially what became ISDX and then Wildfire?