Why does Rhino have lots of issues in filleting complex surfaces while Solidworks usually fillets easily?

The purpose of my example was not to show the flaws in filletEdge. Everybody knows about that.

There is only one reason Rhino has a reputation for having lots of issues in filleting complex surfaces and that reason is that McNeel has wasted resources working on a tool that doesn’t work and likely never will work and has ignored development of a tool that does work well but is extremely time consuming and difficult to use. If just 10% of the development time was devoted to automating the filletSrf command, Rhino’s reputation for filleting would improve enormously.

I posted that file as one example of how easy it would be to improve the workflow for the filletSrf command. The purpose was to show how ridiculous it is that the user has to keep showing Rhino where to make the next fillet over and over and over again in situations that it should be obvious where the user is going to request a series of connecting fillets. .

2 Likes

This has been requested long ago.
Its not very difficult to extract the bare surface without the extensions. The problem is that there is no way to accurately get the radius that is needed or wanted at each end of the unextended (Variable)filletSrf surface.

1 Like

sounds like the pong to the ping of this thread.

ApplyEdgeSoftening command does not help because it is crude - no matter what options I choose, it always chamfers the edges.

1 Like

See the replies in this thread:

I tried Weld and WeldEdge on the extracted render mesh. It did not work.

1 Like

Free OnShape program makes much better fillets than Rhino: https://www.youtube.com/watch?v=xTZpGM2P9u4

1 Like

Not sure if it would help to make it easier to create fillet on that model but it looks like he uses variable fillets in solidworks and onshape but not in Rhino?

1 Like

These discussions help, you say? Then please allow me to bump it and let it never fall off the main page. :slight_smile:

At work, we’re lucky to have seats for Catia, Alias, Solidworks, NX and Rhino, so we can use the best tool for the job at hand, but I work at the largest industrial design firm in our country, so I realize not everyone has that luxury.

But if I had to design a product from start to finish in Rhino, I’d probably tear my hair out, if I may be blunt.

1 Like

If you upload that 3dm model here, I will send you an updated version with working fillets. :slight_smile:

P.S. What’s the font that you used in your scren-shot? Really beautiful one.

1 Like

“But if I had to design a product from start to finish in Rhino, I’d probably tear my hair out, if I may be blunt.”

Yes, it’s excruciating. Even with simple products as the one below, when you need maybe three quality design proposals for a quality client, all G2 surfaces and fillets, draft angle, and you can’t bang this out comfortably with all the interactive control and analysis tools you need for NURBS, it’s a pain, and the engineers will have to model it all over again - and that wastes much time and, more importantly, cost.

2 Likes

Every Rhino user has his or her own favourite set of tools and modeling techniques, but in my opinion the following tools and improvements should be the most urgent priority for future Rhino releases:

  • Match surface. Free-form surfacing is crucial part of Rhino and robust match surface capabilities are a must.

1-1. Make it possible to combine “Position”, “Tangency” and “Curvature” when using “Match up to four surface edges”. Currently, only one of these could be used for all edges, which is a huge limitation.

1-2. Allow partial match of surface edge based on user-selected region (a percentage length of the surface edge) or by selecting certain surface control points (pictured below). In many occasions the modeler needs to match only a portion of the surface edge while keeping the rest of it intact. Product design will benefit from this functionality. I already posted that request two months ago in another thread…

1-3. Surface match influence. With other words, allow user-defined mix between “Tangent” and “Curvature” so that the surface match is being something in-between both types. Could be also useful for creating crease lines, if the start of the surface edge has “G1 and G2 influence = 100%” and the end of the surface edge has “G1 and G2 influence = 0%”. This command will work best if it allows the user to divide (not actual edge splitting) the surface edge into sections (such like 2, 3, 4, 5 etc) and set custom influence parameters to each of them. By default, it could have only two sections (start and end), and extra ones could be added by the user at any time during the command. Using 4 regions and making regions 1 and 2 same (lets say G2 influence 100%), and 3 and 4 same (G2 influence 0), will ensure that both the start and the end of the surface edge have equal matching parameters. Example of that with a surface edge divided into 5 adjustable regions (start, end and 3 inner regions). Note that the middle region is optional for greater control over the shape, and would not exist if there were just 4 regions:

1-4. Match with draft angle. This is very important for cast or vacuum formed products where the recommended split line of the mould must be about 2 degrees or more.

1-5. Show the maximum deviation or a warning for critical areas where the matched surface edge(s) go too far from the target edge(s) (similar to what the “Maximum deviation” in “Rebuild surface” does).

  • Blend surface.

2-1. Add “Match target isocurve direction”, similar to what of the option of the same name does in the “Match surface” tool.

2-2. Ability to pick individual Blend surface handle and match it to the same direction as the edge of the target surface. Example:

2-3. Visible “Planar sections” handle that could be adjusted at any given moment via manual manipulation (with the mouse pointer or the Gumball) or by user-defined angle of rotation. The current implementation requires turning off the “Planar sections” option upon every time of changing its direction. This will not allow for a clear and direct visual comparison between different directions for the “Planar sections”.

2-4. Sometimes “Blend surface” behaves in a strange way and adds considerably more control points for the same resulting shape compared to its predecessor in Rhino 5. Turning off the “Interior shapes” reduces the amount of control points, but the tool fails to properly blend the surface, unlike “Blend surface” in Rhino 5.

15 Likes

As someone learning Rhino, I too have had simple fillets go awry. (Fortunately, my work does not require the precision that others here need.) Most times, I have to change my fillet radius to get it to work.

At any rate, I have on occasion wanted to have varying radius fillets and have no idea how to achieve them. I also must be doing something really wrong with FilletSrf as seen in my attached files. (Pic is object, fillet with FilletEdge, fillet with FilletSrf.)

This discussion is probably beyond my skills at this point, but if someone could point me a good tutorial on getting good fillets in Rhino, especially varying radii, it would be appreciated.

FilletTest.3dm (412.9 KB)

1 Like

In this case, FilletEdge.

image

-Pascal

1 Like

Can I get a fillet to taper along an edge?

1 Like

Thanks, I know how to fillet it manually, it’s just tedious work that in this day and age I just don’t think I should have to do (and that I don’t have to do if I switch to basically any other CAD software). :stuck_out_tongue:

As for the font… well, that’s just my sloppy handwriting using a tablet. :smiley:

3 Likes

I have to say that at least with your particular example there, that’s something you’d be able to knock out super quick in SubD with V7…

However, when you then use toNurbs, I bet Rhino is going to have even more trouble filleting those surfaces, because they will most likely have rather weird control point topology.

Wow, your handwriting is so cool. :slight_smile: Would be nice if you post an image consisting the whole English alphabet. :smiley: The green letters seem to be bold version and look especially good.

2 Likes

In this particular example, the upper 3-sided fillet surface fails to build tangency to the long fillets at either side.

I have yet to come across a top tier or even second tier client who accepts subdivision surface modelled data.

1 Like