topic Re: Problem in loft in Fusion Design, Validate & Document Forum

Problem in loft

cdotrodrigues — Sun, 28 Jan 2024 01:06:02 GMT

Hi all,

I´m stuck here in what I think is a simple loft.

Trying to create a loft surface from dark gray nose body edges to profile in skt_3 using the top and bottom extruded surfaces edge and the middle (skt_top3) as rails. An erros is thrown but I can´t really figure out why ... any help would be appreciated. File uploaded as "fuse_test.f3d"

Thanks

Carlos

Re: Problem in loft

jhackney1972 — Sun, 28 Jan 2024 01:29:34 GMT

See if this works for you. I created a projected sketch on the existing surface body because it kept saying that end has multiple paths. Model is attached.

Re: Problem in loft

cdotrodrigues — Sun, 28 Jan 2024 01:44:49 GMT

Thaks for your answer. I did make the body using the sketch from the nose body. Anyway I was trying to figure out why does fusion throws that error when I use the 2 edges from the node body. In fact I´m not searching for a workaround but trying to find what´s wrong with my model. Thanks

Re: Problem in loft

cdotrodrigues — Sun, 28 Jan 2024 01:50:35 GMT

And if you need to use tangent or curvature to match both surfaces you cannot do that using the sketch as profile.

Re: Problem in loft

davebYYPCU — Sun, 28 Jan 2024 02:13:10 GMT

Your first Loft is corrupt. Why?

You used a Point profile, that belongs to a rail sketch.

All the inputs must be from individual bodies or sketches. (Big section first is 2 body edges and 3 sketches - works)

Second Loft is 3 body edges and 2 sketches.

You have the correct preparation, but you should not be segmenting lofts as you move along the fuselage. Loft is finicky and even with a tangent connection, you will see the transitional segments from segment lofting. (Like a starving horse it will give your ribs. See highlights in this pair of lofts.

I prefer to make separate rail curves first, then Project > Intersect those to profile sketches.

Might help...

Re: Problem in loft

TrippyLighting — Sun, 28 Jan 2024 14:53:58 GMT

@cdotrodrigues this loft should absolutely work and I don't see any fault on you side. Period. end of story!

@davebYYPCU @jhackney1972 Please report bugs or undesirable behavior so we can get these tools improved. Waving this off as "loft is finicky", without a true explanation is not OK.

@Phil.E can you please report this as an improvement request?

Edit:

I tried this in another CAD software (2020 release) and it worked on the first try. Even with the G2 (curvature continuity) enabled across this surface transition. The sketches and surfaces in the Fusion 360 model support this, but the loft tool fails already in the preview if that transition is changed from tangent to curvature continuity.

Re: Problem in loft

davebYYPCU — Mon, 29 Jan 2024 02:46:04 GMT

What part of

Your first Loft is corrupt. Why?

You used a Point profile, that belongs to a rail sketch.

(My opinion and experience) All the inputs must be from individual bodies or sketches. (Big section first is 2 body edges and 3 sketches - works)

Second Loft is 3 body edges and 2 sketches.

(Successful Lofts provided in the example file)

Is without explanation?

Re: Problem in loft

laughingcreek — Mon, 29 Jan 2024 04:54:12 GMT

@cdotrodrigues -a few thoughts and then I'm off to bed.

I do agree with @davebYYPCU that lofts are finicky. so you have to be very careful with your input geometry. I also agree that it might be better to loft this all in one go instead of segmenting. (but it kind of depends on what your desired outcome is)

and while I agree that lofting to a point should be avoided when possible, I disagree that the loft is corrupt because a "point profile that belongs to a rail sketch was used". The end results of Dave's still isn't particularly great.

one thing that draws my attention in your model is the surface edge running down the middle of the nose cone-

this is b/c you used 2 splines joined at a point instead of a single spline in sketch skt_top2. despite your efforts with the constraints, this is what the curvature looks like across those 2 splines-

I imagine you did this so you could get separate surface bodies to use as rails? it is better to use a continuous spline like this. (you'll also need to use a fitpoint spline to control where the point of the nose lands)-

in order to get your 2 surface bodies to use as rails, create profiles-

extrude them one at a time-

now when you do your loft you'll end up with a single surface for the nose, and it will have much better curvature quality-

more thoughts-

you got luck with the sketch skt_top3. the curvature of it is pretty good, but you'll note that sketch skt_side3 isn't curvature continuous between the splines despite having a curvature continuous constraint on them🤔-

also already mentioned is that it's bad practice to chain your lofts together like this. but if you rgoing to at least use continuous spines.

when you loft an edge free, with nothing to control the curvature, you don't really no what your going to get. using that edge in another loft is a crap shoot. preferably all edges of a loft (profile and rails) will be from other edge surfaces that have been created with things like extrude, revolve or sweep.

one possibility to avoid lofting to a point (if the above doesn't work good enough) would be to revolve a disk at the very tip of the nose cone and loft to the edges of that instead of a point. just a thought.

are you creating an idea from scratch, or are you trying to recreate an existing object? do yo have design sketches or pics you can post?

Re: Problem in loft

Phil.E — Mon, 29 Jan 2024 17:43:21 GMT

I'm happy to log a feature request. But I'm having trouble figuring out what the request is by reading this post. Can you elaborate?

Re: Problem in loft

laughingcreek — Mon, 29 Jan 2024 23:21:26 GMT

@Phil.E wrote:

I'm happy to log a feature request. But I'm having trouble figuring out what the request is by reading this post. Can you elaborate?

@Phil.E -

I'll let @TrippyLighting elaborate on his findings, but as long as we have your ear-

clearly the curvature continuous constraint isn't working right on the splines in the sketch skt_side3:

also, the loft error message "one or more profiles have multiple edge loops" probably doesn't mean much to someone who isn't involved with programming cad software to do lofts. And "cannot create tool body for loft" is super vague. for both errors it would be nice to have more relavant verbiage that can be understood by your average non-programmer, and also some visual feedback on what/where the issue is.

Re: Problem in loft

Phil.E — Tue, 30 Jan 2024 15:44:49 GMT

@laughingcreek Thanks. I've asked the sketch team to look at this and sort out the issues. I agree that the error messages could be more "human". Do you have any suggestions for the text? If you know what the error means, it's interesting what you think it should say.

Thanks!

Re: Problem in loft

TrippyLighting — Mon, 05 Feb 2024 23:38:37 GMT

@laughingcreek wrote:

@Phil.E wrote:

I'm happy to log a feature request. But I'm having trouble figuring out what the request is by reading this post. Can you elaborate?

@Phil.E -

I'll let @TrippyLighting elaborate on his findings, but as long as we have your ear-

clearly the curvature continuous constraint isn't working right on the splines in the sketch skt_side3:

also, the loft error message "one or more profiles have multiple edge loops" probably doesn't mean much to someone who isn't involved with programming cad software to do lofts. And "cannot create tool body for loft" is super vague. for both errors it would be nice to have more relavant verbiage that can be understood by your average non-programmer, and also some visual feedback on what/where the issue is.

Those transitions are G2 curvature continuous. That just isn't easily observable as the transition happens exactly at an inflection point.

(view in My Videos)

Re: Problem in loft

laughingcreek — Tue, 06 Feb 2024 17:42:31 GMT

interesting. I didn't realize it was only the magnitude of curvature, not the direction, that mattered. I thought that was was just tangency. So by definition then, two arcs that are tangent with the same radius are curvature continuous?

I'm not sure I've ever read anything that directly covers the curvature at an inflection point. from observation I had just assumed curvature had to go to zero at the inflection point to be able to get to G2-

(control point spline with 2 control handles each side co-linear-)

the zebra stripes across the point in question on the OP models sure LOOKS like it's just G1-