Tight curve lofting improvment

[Anonymous]

I work a lot with irregular shapes that require lofted shapes with tight curves. I notice Fusion really does not like being asked to loft around tight radius's and find this to be my main reason for having to stop using Fusion. Not only does it not like it, it also adversely affects the rest of the lofted shape as a result. Other like modelers can handle these type of lofts with ease, and I would love to see Fusion 360 join the pack. I'd also like to keep using the other features included in Fusion, but without this simple upgrade I am forced to look to other software to fulfill this basic requirement.

Can you help?

[Anonymous]

 I too have suffered from this same issue although I feel as if the best solution to this problem is to use rails/guides when lofting although this fix doesn't always work. I feel the best way this could be improved is for users to be able to manually choose the path of the loft using lots of individual nodes to control the shape before the program automatically chooses the path of the loft.

[petrara]

@Anonymous could you post an example (screenshot)? From what I am reading in your message, it looks more like a bug than something suitable for IdeaStation... and if indeed is a feature that needs addressing (read: if is a bug), then you might get much faster attention from Autodesk if posting in Support section.
... just my 2p on the problem.

[Anonymous]

You will notice the lofted shape does not follow the centerline, and that the tight radius has negatively influenced the shape either side of the tight radius and it not perpendicular/normal to the profile as it should be. Guide curves are very difficult to insert for this shape as the profiles are all on separate planes. Very frustrating.

[Anonymous]

[Anonymous]

Just forget it.

A curved surface, the pairing of continuity, exact parametric curves from the snaps to other geometry and projection. These tasks are not for Fusion. Because Fusion works on a geometric kernel, which does not allow to calculate such geometry correctly and quickly. Even ordinary fillets are not built as they should be.

It is necessary to develop a geometric kernel from scratch, so that the geometry is correctly constructed and quickly considered. It is intellectually difficult and expensive.

This could turn Dassault Systemes, when he abandoned ACIS (CATIA V4) and created a CGM for CATIA V5, which are used 3DEXPERIENCE also. I don't believe Autodesk is capable of pulling off at such a deep and qualitative level. They develop in breadth, and work with the core requires a deep approach.

Before Fusion I have used only CATIA, then CATIA Inventor.
Now I have to use Rhino for basic forms then import into Fusion and finish. But there are also surprises: some patches will be long creation of fillets and all that.

You can try Alias and Inventor through AnyCAD, but I intend to test Creo to evaluate in a real workflow.

Yes, it's all uncomfortable, especially Creo or Alias, will have to get used to, but they have better mathematics on Board, which means that there will be no problems with the surfaces.

I would use Alias and Inventor via AnyCAD, but Alias is very expensive. I prefer CATIA (only for surfaces and Sudivision Surfaces), but for me CATIA is expensive, as is Alias. In Creo you get good math and create the whole design in one environment for one price, while the value for money is adequate.

If Creo will give too slow a workflow, then perhaps the best solution would be to make the shell Rhino, and then complete in Fusion.

In Rhino geometry is created dirty, but in Rhino it is not difficult to create shapes, and the result in surfaces is much better than in Fusion.

In General, this is all, of course, stupid and uncomfortable, but we have no other choice, and probably will not, given the depth of the problem.

This is my point of view and my experience, but you decide for yourself.

[Anonymous]

That's some great feedback, although not promising for Fusion (at all...) it does broaden my understanding of why Fusion is so weak. I can still hope for some recoding, but it sounds like putting chrome pipes on a Corolla won't quite have the desired effect. Shame on you Autodesk, for such a big player you should be much smarter than that.

Rhino has become my go-to software now for shapes like this, but as you are well aware it is lacking in the precision assembly that is required for such parts to be put into... I had my fingers crossed (like so many I'm sure) for a single platform to handle shapes like this and, even though they are not really that difficult to do, I can see it would require a level of coding where I can only guess the developers do not, or cannot get to in their efforts for a software package where each function actually DOES what each function SHOULD be doing. Grrrrr.