topic Re: loft surface quality issues and failed computations in Fusion Design, Validate & Document Forum

loft surface quality issues and failed computations

cekuhnen — Tue, 19 Jul 2016 15:35:27 GMT

here is a file where I experienced many loft command issues.

the video explains it well.

I am curious what the reasons are and why the loft works that way.

https://drive.google.com/file/d/0Byzv_NlyKp_2ckJuLUtaTUJBNjA/view

Re: loft surface quality issues and failed computations

TrippyLighting — Tue, 19 Jul 2016 16:47:57 GMT

That wing tip is really proving to be a challenge!

Re: loft surface quality issues and failed computations

cekuhnen — Tue, 19 Jul 2016 17:24:05 GMT

@TrippyLighting

So easy to model with sub-d hahahaha

but with NURBS you really need top notch surfacing tools

Both approaches we provided can somewhat create the wing tip shape fine.

I am quite sure that however making this flow nicely into a wing will be an issue

with the approach we selected.

This task I just tried in Alias and it works but it is also there incredible work intensive to get it smooth.

I am curious if the quality of the Fusion model however will be fine for the project?

This is not Siemens NX or Catia!

Re: loft surface quality issues and failed computations

michallach81 — Wed, 20 Jul 2016 12:16:19 GMT

@cekuhnen

I know it might cause distraction and you could never get an answer to your question, but while I was playing with your model (from first post),I've noticed an improvement in Loft tool (maybe it's even more then just Loft improvement, it looks like ADSK did improve recognition of self intersection). Take a look:

Not so long ago we would receive an error message or loft would create self-intersecting solid. Now it can make a hole, nice. Nothing did change to Boundary Fill, so we still can get in to trouble, trying to do this: http://forums.autodesk.com/t5/design-validate-document/uroborus-bug/td-p/6252189

Re: loft surface quality issues and failed computations

cekuhnen — Wed, 20 Jul 2016 13:58:10 GMT

@michallach81 Would be great for AD to join here. I already have few posts with critical technical questions and not getting any response at all.

Re: loft surface quality issues and failed computations

michallach81 — Wed, 20 Jul 2016 14:01:34 GMT

would be

Re: loft surface quality issues and failed computations

cekuhnen — Wed, 20 Jul 2016 14:19:17 GMT

@michallach81 I remember also with a different problem that they stated that solid loft and patch loft should work the same but oddly somehow do not work the same resulting in different outcomes. the solid command seems to work better.

Re: loft surface quality issues and failed computations

wilkhui — Thu, 21 Jul 2016 16:42:48 GMT

Hi everyone, sorry for the delayed response, we're still looking into this one.

Indy

Re: loft surface quality issues and failed computations

wilkhui — Fri, 22 Jul 2016 15:02:50 GMT

Hi again - sometimes splitting the loft into different features gives a more desirable result with this type of geometry but in this case I still saw the surface intersection:

I think there's something we can improve here but I want to suggest a change in the design (I know that's not ideal) to allow you to continue with your work...

Adding a tiny amount of convexity to the 'End' profile avoids the self-intersection, can you try it out and see what you think? I can't find the file I made but this is an extremely exaggerated view of what I did:

Let me know if that doesn't make sense and I'll upload a file.

Thanks,

Indy

Re: loft surface quality issues and failed computations

cekuhnen — Fri, 22 Jul 2016 16:38:44 GMT

@wilkhui

If the end profile is a triangle the loft works

If the end profile is as previous it fails

this seems not to be very logical to me nor intuitive.

also with a triangle there is no self intersection.

I could split the end profile and the skinned mid section surface into two parts

but then loft cannot offer G1 or G2 along Rails - I long requested and criticized lack in functionality!

So this approach is not even possible in Fusion.

I would say that the loft code in this area really really needs some improvement to make this work better.

Re: loft surface quality issues and failed computations

cekuhnen — Fri, 22 Jul 2016 17:38:36 GMT

I rebuild the design with clean Degree 3 curves

and encounter no self intersections in Alias with the surface tools (sweep rail)

This is what I would expect Fusion to be able to do.

Re: loft surface quality issues and failed computations

michallach81 — Fri, 22 Jul 2016 18:16:59 GMT

With this exact set of curves Fusion also will succeed.

Re: loft surface quality issues and failed computations

cekuhnen — Fri, 22 Jul 2016 18:43:16 GMT

@michallach81 no Fusion cannot because it fails since the beginning to offer surface continuity beyond G0 along rails!

So cross sections cannot be lofted has halfs but only as fulls- so you can split the rails but not split the profile curves.

That's why you cannot model half models in Fusion well but only full models. The loft code even works worse with surfaces

than solids in certain situations.

But besides the lofting short commings, also the sketch engines fails to work well here because the spline tool bends curves

in a crazy way unlike in AI or how I could work with a CV curve.

The screenshot shows how and where the expected linear segment deforms. So you cannot even comfortably sculpt half a profile

with one curve when a certain part has to remain linear meaning you need to make blend curve and split it. but then you have

4 segments not two.

All this should have been a focus for the advanced modeling and sketching tool set but that project got put onto the shelve.

Loft in Fusion vs Alias

Fusion: No G1 G2 along rails but able to loft to a point + point G1 ! pretty Awesome but not so good mesh structure later

Alias: G0 G1 G2 along all profiles and rails

Patch in Fusion vs Alias:

Fusion: G0 G1 G2 for all edges the same way and now eight control - fusion can dome end tubes via draping and trimming but it is a messy surface.

Alias: G0 G1 G2 and implied G1 G2 for each individual edge and weight control

To be fair - Alias square and Fusion patch are not the same tool. I just used it to illustrate what the tools should have in terms of functions.

Re: loft surface quality issues and failed computations

michallach81 — Fri, 22 Jul 2016 18:35:27 GMT

Can you pass these curves?

Re: loft surface quality issues and failed computations

cekuhnen — Fri, 22 Jul 2016 18:47:09 GMT

here it is

the network is not perfected perfectly smoothed out and I used G1.

To get this into a bezier surface quality one would need even more patches.

but I was only interested in the end of the wing tip

In MOI I also did not get any self-intersection.

Re: loft surface quality issues and failed computations

michallach81 — Fri, 22 Jul 2016 18:55:52 GMT

thx

Re: loft surface quality issues and failed computations

cekuhnen — Fri, 22 Jul 2016 19:16:33 GMT

@michallach81 I totally understand Alias is not Fusion or vs and many things in Fusion internal are solved and work nicer.

But the fact that loft and such commands do not have the requested features (some are around 2 years old and common in all other parametric CAD apps) is really

a frustrating disappointment.

It would make Fusion really a killer app!

Re: loft surface quality issues and failed computations

cekuhnen — Fri, 22 Jul 2016 21:25:33 GMT

did you find a way to make this work in Fusion?