Does SUBD have any use outside of freeform modeling? EVERY example I have seen has been free form.
Does SUBD have any use in a blueprint context to form shapes that are difficult with NURB? If so, are there any examples of this?
I use it to create fluidly curved objects in a process that I wouldn’t exactly describe as “free form”.
Taking tiling patterns from parakeet, turning them into SubD, and manipulating them, can yield interesting, orderly results that are curved but not really free form.
SubD is very good at quick Y-branch blends, even with large numbers of branches and 3D intersections. It will be very near perfect once McNeel adds a “circularize edge loop” command.
You can also use conversion SubD as a good way to create complicated edge blends on complicated planar polysurfaces. Things that will fail to fillet or blend in NURBS are relatively easy to get good edge blends on using SubD, even with the WIP.
1 Like
I do a lot of ship hulls. There are a lot of places that produce 5 sided hull sections that do not work well with NURBs. If you trim a 4-sided shape, the edges are off. Otherwise you have to chop the five-sized sections into oddly-shaped fours sided.
Given a closed edge and interior points within that closed edge, would subd provide any assistance in closing such gaps?
Can you post an example file? Doesn’t need to be a whole ship design. A simple mockup of what you’re describing?
Problem Hull.3dm (6.4 MB)
Here is a extract of what i have done in a problem area using NURBS. I was wondering if SUBD could help me do this better.
THX
Rhino~So would you conclude that SUBD is a practicable and better solution than doing the odd chopping of the area into 4-sided sections?
It is difficult to answer clearly until it is officially released ver7.
Clearly the answer is that it’s an advanced tool in a new format.
“Rhino SubD is a new object, new technology, and new user experience”
I hope this helps.
SubD is certainly more practicable and better than single surface hull modeling. It also serves well as a interactive modeling interface and preview geometry while generating fabrication geometry in the background. It is not a replacement for developable or semi developable plates, unrolling and generating cutting data - although it could be… but not yet.
I’m not sure what you mean by chopping? Are you saying that you re-model the fabrication pieces (sheets of metal, plates) as 4-sided untrimmed NURBS surfaces? It looks a bit like… which would explain the high density of control points in your model. That might be good for squishing, but unpractical from a designers and modelers view.
1 Like
"I’m not sure what you mean by chopping? "
You can see in the example I posted that there are areas at the ends of twin keels that do naturally not form 4-sided (or even 3-sided) surfaces that can be model. Those areas have to be chopped up in odd ways to create 4-sided (and as few 3-sided as possible) surfaces for Rhino. Plus there is the problem that the trailing arc above the shaft outlet creates a corner such that it intersect the hull at a right angle but the surface forward curves with the hull above and below.
This is how I broke it up:
Every video demonstrating subd has used freeform modeling. I have never seen an example of creating a surface to fit an odd shape like this.
1 Like
The perfect scenario for what I’ve described above. But quite some work…
Sorry I didn’t get back to you earlier.
A quick look at the original shows me that you’ve got some curves with poor continuity as edges for some of your surfaces. The panel at lower aft starboard is visibly rippled, even without turning on _zebra, and it’s lowermost edge curves have poor continuity when viewed with _curvaturegraph.
I’m not an expert on how hull design works, but I’d start by trying to create simpler more fluid edge curves by using _rebuild.
As for SubD: there’s definitely potential to get better surface continuity quickly by using it. A quick example would be to extract three or four vertical isocurves from the thruster surface and loft them using the SubD output option, and the rebuild with _ control points option.
Also, with patience, if you can recreate all these surfaces with SubD in ways that allow them to join each other with borders that have equal numbers of subdivisions, your surface continuity across the junctions will be perfect or near perfect.
You don’t need or want SubD for this. You’d be better off investing in Orca3D or ExpressMarine plugins. Rhino is the industry tool for ship hull design. You just have to use it correctly. Take a course on ship hull design.
1 Like
@Stratosfear, if you’re doing ship design professionally I agree 100% with you. But there are scenarios where SubD would be the answer in my opinion. There’s a growing number of people doing scale model building using 3D design software and 3D printing. While a SubD ship hull may not yield a 100% accurate result for professional use it would be both much easier to model and perfectly acceptable for a scale model.
Another case would be concept modeling. You could quickly model something that looks right for customer approval before moving on to building an accurate more time-consuming model.
Another thing that’s interesting is the Seanaptic image you posted, I’m not familiar with this plug-in at all, but it seems they rely on a whole bunch of NURBS surfaces in the bow of that ship hull, would be pretty cool if those were all connected as SubD limit surfaces instead of individual NURBS patches.
All quite exciting developments though.
Which tools in Orca3D would help with this?
I am not designing. I am working off of existing blueprints.
He expects the utmost accuracy to his curves. With that in mind he won’t be happy with SubD. Even then learning how to SubD properly (Quads only) is a long process to figure that out. I’m trying to be helpful.
I agree with everything you just said.
I’d say it depends what you want to do with the model. Do you need fabrication data, unrolled cutting contours for sheet metal, or a solid for 3D printing, or just something to calculate hydrostatics? Or the perfect shape for fluid dynamics simulations? Based on what input? Parametric? Manual modeling? So many options…