Work flow for sketching symmetric splines in 2D

[Anonymous]

Hello, 

I'm struggling for quite some time with this as I'm trying to make organic interesting shapes with splines but can never get a line curve to be the same in relation to a line. 

Some people would suggest making only half a spline but then when you mirror it, it creates this not too smooth transaction along the line mostly a pointy edge like this one: 

Any solution for that or this is just the nature on splines in fusion 360?

Thanks for your time.

[robduarte]

I don't quite understand what you're asking, but I can say for sure that it's not a limitation of Fusion 360. I can almost guarantee you can do what you're aiming for. Can you post a public URL to your design so we can see the sketch that you're struggling with? 

[g-andresen]

Hi Mike,

try it this way.

good luck

günther

[Anonymous]

What I'm trying to achieve is for these spline have a symmetric curve across the middle line and it won't do even after setting up symmetry constrain with the two ends this picture describes what's wrong: 

here is the link for that particular example: 

http://a360.co/2FitczD

[Anonymous]

I tried this but it either still makes that non G2/G3 complaint and not continues curve.  

[g-andresen]

Hi Mike,

The symmetry of the sketch conditions only determine the distance of the points to an axis, but not the behavior of the tangents in length and radius.

But if you make only half of it and mirror it, then you have identical tangent behavior on both sides of the mirror axis.

(screencast)

günther

[TrippyLighting]

---

@Anonymous wrote:

But if you make only half of it and mirror it, then you have identical tangent behavior on both sides of the mirror axis.

---

That does not agree my experience! I reference this thread as it discusses the same thing.

[Anonymous]

So I understand there is no simple way to apply symmetry to multiple tangent lines in the dots of a single spline.

[TrippyLighting]

That was a misunderstanding on my side, based on another thread that discussed mirroring a spline.

In that case the scale of the tangent handles are calculated independently on a per spline basis, regardless of applied symmetry.

If you have a single spline and want to make it symmetric, applying equal dimensions, or applying the equal constraints will keep the spline symmetric.

[chrisplyler]

---

@TrippyLighting wrote:

...based on another thread that discussed mirroring a spline.

In that case the scale of the tangent handles are calculated independently on a per spline basis, regardless of applied symmetry.

---

That's true if you only mirror the spline line. If you mirror the spline AND the tangent handles of all its points within a single mirror event, it seems to work as you would expect.

For some reason this approach creates extraneous sketch lines where the tangent handles are on the mirrored side, which you will only notice once you move any of the points or reconfigure any of the tangent handles, but it seems they can be deleted with no ill effect.

[Anonymous]

But when you apply the equal constraint to the tangent lines the relative angle is not affected making it not that symmetric and it is not ideal once you have many points also still when you mirror it it created two separate faces that are not smooth like shown in the first screenshot.

[TrippyLighting]

I personally prefer a different method for modifying symmetry in splines isn that also allows me to modify spines without actively editing the sketch:

[TrippyLighting]

@chrisplyleras already explained in the other thread, mirroring a spline seems to work now, but certainly did not in Aug 2017 and it was confirmed by the Fusion 360 and in offline conversations with 2 other Expert Elites.

But the good this is tat in the meantime work on debugging the sketch engine is showing results!

I wonder how you have been able to mirror a spline without the tangent handles ? Ive not been able to replicate that. Whenever I mirror a spline , the mirrored spline also has tangent handles with symmetry constraints applied.

[Anonymous]

Yes this is kind of the best solution for that but again it's very labor intensive and once you have to deal with more than one point that becomes messy.

[TrippyLighting]

If you can share your design maybe we can find a solution that works.

[g-andresen]

Hi,

In such cases I define the control point over length and curve.

günther