Trying to sweep&twist a line along closed spline but error occurs

[Anonymous]

Please watch the video and tell me what is causing the error.

https://autode.sk/2KNS705

thx in advance

[mavigogun]

A couple observations.   The swept profile in the sketch is constrained to the origin; from my experience, pinning to the origin may introduce undesirable consequences (ya, I recognize the sketch origin is also the intersected path projection).   In troubleshooting, consider removing the origin relationship- the same element could be constructed adjacent and moved into position.   I wonder if you have attempted the same sweep as an incomplete loop or omitting the rotation in attempt to isolate the nature of the problem?

[mavigogun]

I was able to perform the sweep without issue using the same elements- constraints, number of line segments, etc.   This may relate to some inperceivable issue with the splines- share the file and I'll report if the same error occurs on my end.

[ToddHarris7556 · Accepted solution]

With all due respect - this looks like a pretty simple case of a self-intersecting surface. If your path crosses itself, the feature will fail. 

Screencast attached. 

[ToddHarris7556]

Sorry - I should have also shown you how to do the pretzel-twist, which is possibly what you're after. 

[Anonymous]

I solved the problem: it was caused by self-crossing of sweep surface in
the middle of 8-like shape.

[ToddHarris7556]

Glad you figured it out 🙂

[Anonymous]

---

@mavigogun wrote:

A couple observations.   The swept profile in the sketch is constrained to the origin; from my experience, pinning to the origin may introduce undesirable consequences (ya, I recognize the sketch origin is also the intersected path projection).   In troubleshooting, consider removing the origin relationship- the same element could be constructed adjacent and moved into position.   I wonder if you have attempted the same sweep as an incomplete loop or omitting the rotation in attempt to isolate the nature of the problem?

---

I did it by constructing adjacent and constraining by dimensions, but the error remains.

[Anonymous]

---

@mavigogun wrote:

A couple observations.   The swept profile in the sketch is constrained to the origin; from my experience, pinning to the origin may introduce undesirable consequences (ya, I recognize the sketch origin is also the intersected path projection).   In troubleshooting, consider removing the origin relationship- the same element could be constructed adjacent and moved into position.   I wonder if you have attempted the same sweep as an incomplete loop or omitting the rotation in attempt to isolate the nature of the problem?

---

Yes I tried with omitting the rotation, but error remains. 

Incomplete loop solved the problem - if the sweep does not intersect itself.

Later I moved spline path defining points in 3D, to avoid intersecting, and this solved the problem even with complete loop.

[Anonymous]

Yes, self-intersecting was the problem.

(but why there is no "ACCEPT SOLUTION" button related to your answer?)

[Anonymous]

---

@ToddHarris7556 wrote:

Glad you figured it out 🙂

---

I don't care if there will be some self-intersecting or not because my model is not a machine assembly but just a jewelry model for 3D print.

Is there a possibility to turn off self-intersecting errors checking in Fusion 360?

[ToddHarris7556]

Re: turning off error-checking - in short, no.

The error isn't just giving you advice about the situation, it's actually saying that it can't solve it mathematically. 

In terms of a workaround, I'm guessing that you're probably trying to thicken this surface (as 3D printing a zero-thickness surface probably wouldn't do much for you in terms of jewelry) so I think that you're more likely to be successful working with 2D geometry, extruding up to get some 3D bas relief to work with, and then sculpting that as needed to get the texture you're after. 

That's only a guess, as I'm not really sure what your end goal is. 

[Anonymous]

---

@ToddHarris7556 wrote:

Re: turning off error-checking - in short, no.

The error isn't just giving you advice about the situation, it's actually saying that it can't solve it mathematically. 

In terms of a workaround, I'm guessing that you're probably trying to thicken this surface (as 3D printing a zero-thickness surface probably wouldn't do much for you in terms of jewelry) so I think that you're more likely to be successful working with 2D geometry, extruding up to get some 3D bas relief to work with, and then sculpting that as needed to get the texture you're after. 

That's only a guess, as I'm not really sure what your end goal is. 

---

My goal is to make 3D model of item shown on background jpg.

So, the edge of twisted strip will be the path for new sweep where two circles will be the profile.

Result will should a solid body. 

Self-intersecting will appear on the inner sides of curvatures when I increase twist variable enough (in this second twist feature).

The model would present a twisted rope, so intersecting should not be a problem, because the real rope will be squashed on the inner sides of the bends if we try to bend it this way.

So, if self-intersecting is not allowed - the profile should must not be the same on the inner and on the outer side of a curvature. It seems incredibly hard to make this model.

Have you an idea how to solve this?

Maybe to have two components which can intersect each other?

[mavigogun]

---

@ToddHarris7556 wrote:

With all due respect - this looks like a pretty simple case of a self-intersecting surface

---

If by "self-intersect" you're referring to the cross over point, I was able to perform such a Sweep without issue; if so, why then the discrepancy in outcome?

[ToddHarris7556]

Great question, and I don't know the answer. 

I tinkered with it quite a bit, and can reproduce 100% (so far) that as soon as either a surface or solid sweep body crosses itself, it fails. This has been my experience in Inventor, SWx and Fusion as far back as I can remember, so I've always just accepted it as a limitation. It was this question that prompted me to go check it out again, to make sure I wasn't overlooking something. 

If you've got a workflow that works, I'd love to see it. It's not something we fight everyday, but occasionally, it would come in handy.

[mavigogun]

---

@ToddHarris7556 wrote:

 If you've got a workflow that works, I'd love to see it. It's not something we fight everyday, but occasionally, it would come in handy.

---

I didn't save the build- and now my attempt includes the same error.   Curious.   I'll take another swing at it on the marrow.

[Anonymous]

---

@mavigogun wrote:

---

@ToddHarris7556 wrote:

With all due respect - this looks like a pretty simple case of a self-intersecting surface

---

If by "self-intersect" you're referring to the cross over point, I was able to perform such a Sweep without issue; if so, why then the discrepancy in outcome?

---

Yes, "self-intersecting" is cross over point (sorry, I'm not native English speaker).

mavigogun, please, record your workflow and publish it here, so we will maybe find the reason of discrepancy in outcome.

Also, publish the file, too.

Thx in Advance