I’ve designed A LOT of products in the past years, using Catia. These products always have - let’s say - a high ergonomic-factor: saddles, ski-boots, helmets, etc.
The typical design process was: scan of the human body part of interest, design a good base surface that approximates the scan, pull the aesthetic on that surface and build the product (for certain products the outer / aesthetic surface is an offset of the base surface, so it has to be clean enough).
This works fine with Catia, since if any adjustment of the base surface is needed it is easy to modify the base surface and re-parametrize all the stuff.
A few months ago I joined the Rhinoceros world and I’ve started to build my personal workflow, trying to mimic the Catia workflow: design of the base surface (usually a Network Srf or a SubD) and the curves to project or pull on that surface; then all this stuff as inputs in Grasshopper to have easy adjustments and then… struggling with GH to design all the surfaces there (usually Network Srfs).
This works with the easy stuff, for the more “mechanical” parts and for the meshes (great!) but for the complex shapes I’ve found it really intricate.
So I came back to Rhino NURBS modelling basics and I started with a “cleaner” approach, following the “Alias golden rules” and the precious tutorials from Sky Greenawaldt.
Question:
4 Sided edge surfaces and single-spans are really the way to go, but what about the base surfaces that I’ve mentioned? Those super-organic surfaces that approximate the body scan? Is the Network Srfs to avoid ALWAYS in favour of a “patch layout base surface” or not? Please consider that sometimes the outer / aesthetic surface is an offset of the base surf, so several patches on the base can be hard to handle.
Thanks for your reply.
Yes, I guess it is a good option. What I’m thinking of is: "does it make sense to struggle in designing something ‘organic’ following the ‘class-a’ rules?
depends if your product needs to be class A. If so, then yes, if not then no.
I’m all for modeling appropriately for the intended product.
modeling class A for a toy or product that will be machined or pantographed in asia, in a mold that is finished by hand, is a waste of your modeling time.
Modeling an audi r8 doorskin that will be cnc machined then used for production directly with no human intervention? then yep, class a all the way.
So glad you enjoy my series! I would second everything Kyle has said - I think these’s a proper tool and a proper approach for each project, and that’s not always going to be classic patch modeling/class-A etc. Especially for the use cases you describe, where the thing you’re trying to conform to is something not easily described by that approach - subd’s are often better in this case. Heck, I’m working on an aircraft cowl right now, and speed is of the essence, as is the ability to iterate quickly. For this - subd’s all the way. Could I make a final product that has even better reflections if I took that subd concept and did a full re-surface of it in NURBS - sure. Will this make a difference for THIS project - and more importantly is the client going to see the value in that rework, given the large degree of hand finishing this piece will have? Nope. Take a survey of the tools at your disposal for each project, weigh the pros and cons, and pick accordingly I say.
I admit that sometimes I’m too much of a perfectionist and I totally agree on the machining process topic.
Switching from Catia to Rhino I’ve noticed that I’m focusing more on the patch cleanness, since that means “more control in editing” as well. For me, working with a ‘full parametric surface modeler’ sometimes gave the chance to add too many patches than needed.
@sgreenawalt : I hope to see other series in the future.
