Pattern on Path, "Origin" of component messes it up

[kevin_trefflich2]

Hey,

I´ve included a screencast with spoken explanation where I describe my problem. It´s best you watch this, easier than writing it all.

https://screencast.autodesk.com/Embed/Timeline/8b04edcb-14ed-496b-9d2d-8ff4adc0c684

Shortform in text:

I need to create a pattern of a component along a curved line from a sketch. Problem is that the components do not follow the path perfectly. From what I can figure this is due to the "origin" of the compenent being vertically offset and not in line with the path. I suspect this is because the compenent is not perfectly symmetric along this axis.

To test this I have recreated the component in its most basic dimension but made it symmetrical, the "origin" of this compenent is now in line with the path. If I use this component to create the pattern everything lines up perfectly.

The question is how do I fix this, can I move the "origin" of the component somehow?

For clarification the actual origin of the part (the one that you can choose via the the list of the components on the left) actually is in line  with the path. Only the "origin" that is displayed when I choose to move the component is offset.

Thanks for your help

[aliobidi]

Hi , 

attach your file 

[kevin_trefflich2]

Here is the file.

Thanks for your help.

FYI, I´m not even sure if the "origin" is the problem, thats just what I´m suspecting.

[stiller.design]

Cable chains are difficult to have them parametric . If you really want to go down the rabbit hole you can try the rather new command TANGET RELATIONSHIP in the assemble section. Viel Erfolg 🙂

[kevin_trefflich2]

I know, I´ve tried with normal joints but there are just to many degrees of freedom with so many elements. Thats why I came up (rather copied from someone) this idea of having a parametric sketch along which the elements are placed.
If I want to move the drag chain I will always need to move back in the timeline before the sketch was created, move the endsegment of the chain, capture the position and jump back to the end of the timeline. The sketch will update accordingly and the elements will get placed along it.

This works perfectly with every single symmetric chain element I create, it´s just this unsymmetrical element that is provided by the manufacturer. It´s also not that the element is an imported STEP-File, before I had choosen a different cable chain from the same manufacturer that happened to be symmetrical and this approach worked fine.
I´ve also noticed that when you choose the identical orientation the ends of the cable chain will link up as intended but obviously this orientation will not work in my usecase.

Thanks for your help

[johan.rutgeerts]

Hi,

I faced the same problem. I consider this a bug.

I found following (limited) workaround. Mind you, it's ugly!

Consider a certain path:

Project it in a new sketch and manually draw a segment for each chain link. You need to two lines for each segment. E.g. if your pitch is 25mm (red arrow), then draw two parallel lines of 12,5mm each (green arrows). For the first chain link, only draw the 'second halve' of the segment (blue arrow):

Position your first chain link centered on the start of this newly sketched path:

Create the pattern, start point = 0, distance = the chain pitch:

There you go.

I told you it wasn't pretty. 😉

Above example file in attachment.

[johan.rutgeerts]

I made a cross post on the support forum to file this as a bug.

[jeff_strater]

in the future, please don't create duplicate posts.  There is nothing special about this forum vs the support one.  Bugs can be reported on either.  Thanks.

[kevin_trefflich2]

Thanks for the idea but sadly this wont work for me because I need to be able to move the drag chain and update accordingly (by altering the sketch).
In your way its not possible for a single element to "slide" from the radius to the straight horizontal part of the chain.

Thanks for all your help, maybe someone comes up with a fix in the future somehow.

[johan.rutgeerts]

---

@jeff_strater wrote:

in the future, please don't create duplicate posts.  There is nothing special about this forum vs the support one.  Bugs can be reported on either.  Thanks.

---

Ok I won't, yet I have to add that I find these kind of implicit rules confusing.

This blog post on "How to best report issues and get help" explicitly states: "Report bugs to the support forum"

This post also states: "This forum board [i.e. Fusion 360 Support] is the correct place to report bugs in Fusion 360."

Anyway, more important: given that 'Pattern on path' is not working as it should, did somebody register it as an open bug?

Regards,

Johan

[davebYYPCU]

for a single element to "slide" from the radius to the straight horizontal part of the chain.

I don't think it is a bug just a limitation of the Pattern on Path.  As others mention, you can get parametric result, with a combination of other patterns, with a change of geometry.  Some testing got me here, 

The path curves do not accord with the left end pivot on the component.

The axis between the two pivot points are not tangent to the radius at it's mid point, and therefore the pivot connections on the components wholly on the path radius overlap from one to another.  (top of pic)

You might see the the purple curve path end point above the red line, within the pin hole.

Align a component to the curved path with both holes on the curve one hole on the end point, use circular pattern to duplicate those, Path segment, multiples of the axis length.

Align the flat adjacent link, and rectangular pattern them for the duplicates, again, path length = multiple segments. 

Might help.....

[kevin_trefflich2]

Thanks for your help!
Im not 100% sure if I understand what you are doing, do you mind to the share the file?

[davebYYPCU]

Will be an hour or two.

Does not allow for actual snaking movement, without way more work.  I did see a YouTube video, with animated chain on cogs, or tank track style of rotary motion, but for cable chain that may not work in your case.  Static solution with multiple patterns, so far.

Might help....

[davebYYPCU]

I would abandon the Pattern on Path.  Using the current path as a guide.

Rectangular Pattern 25 components horizontally.  I have used as built Revolve joints, Normal Joint for a few places,

grounded this one, to demo, will work correctly with the correct Joint Limits, 0 to 10 deg max, or will not form a nice curve.

Might help....

[johan.rutgeerts]

---

@davebYYPCU wrote:

I don't think it is a bug just a limitation of the Pattern on Path. 

---

Imo it's an unfinished implementation, that someone deemed "good enough to be released", as it is to some extent functional (e.g. if your pitch is small enough wrt the curvature, the position error becomes neglectable). Unfortunately, it is also buggy.

Consider this:

Set 0.5 as start point, the pattern disappears:

Revert start point to zero, the pattern appears inverted !?

While making these screenshots, I noticed that the reason is that the trajectory is flipped when the start point becomes larger than the pitch (??). So when reverting to zero, the start point is at the other side of the trajectory (compare to the initial screenshot):

Call it a bug, call it a limitation, unexpected behavior, whatever. In any case if I were responsible for this implementation, I would not be proud of it. Even the documentation seems unfinished: I did not find any official documentation as to the exact function of the parameter 'Start point'.

[davebYYPCU]

Yes, Start point other than zero, usually pushes the parts off any curved path.

The limitation here is attempting to align the holes in the part to the path, whereas the function will likely be to align an arbitrary point within the object being patterned.  Start point may well be that parameter for the object's offset.

However, the file is for animating the cable chain in this question, - does not need the Pattern on Path.

Might help...

[johan.rutgeerts]

---

@davebYYPCU wrote:

Yes, Start point other than zero, usually pushes the parts off any curved path.

Ok but that's a bug. It is not normal behavior. Or at the very least the way it works should be clearly documented.

The limitation here is attempting to align the holes in the part to the path, whereas the function will likely be to align an arbitrary point within the object being patterned.  Start point may well be that parameter for the object's offset.

The way it aligns is as follows:

From the start point of the trajectory (at least when parameter 'start point' is set to zero): locate the point along the path at a distance as set by parameter 'Distance'. E.g. in this case, Distance is 25mm, which equals to 17,50 mm linear + 7,5mm arc length:

At that point, place the next instance, oriented according to the tangent of the path at that point.

This is relative to the pose of the initial component wrt the trajectory, e.g. compare with this (first link moved 12,5mm to the right):

Or this (first link moved 12,5mm to the right and rotated 90deg CCW):

For the record:  this is not buggy behavior. It is limited, in that you cannot adequately model chains with large pitch wrt the curve radius. But it has its relevance.

What is buggy is:

1. That it is not documented, so you have to find out this info on your own,
2. The behavior of the 'start point' parameter.

However, the file is for animating the cable chain in this question, - does not need the Pattern on Path.

---

Yes, unfortunately my use case is not a cable chain...

Furthermore, using a rotational joint per link is rather computationally intensive, as well as results in a highly underconstrained kinematic chain, which has its own issues...

Regards,

Johan

[kevin_trefflich2]

So, to sum up, nobody got a real clue why "Pattern on Path" shows this weird behaviour or how to solve it?

If that´s the case, can I get a confirmation from someone of the fusion team that this is officially considered a bug and looked into?

I´m still hoping someone can figure out a workable solution but thanks for all you guys` help!

[davebYYPCU]

Seems you have a misconception that Pattern on a Path is the right tool for the job, from someone else’s example.

Pattern on a Path can align parts to the curve, and did so, according to the data existing, Your alignment, 1, was off, and expectation, 2, is unrealistic.  Your Component geometry does not allow three points of a line, to convert into an arc.

If I read it correctly, you want variable number of links, driven by a changed sketched path length, the Rectangular Pattern does that too, but Joints can not be patterned.

I suggested a workable solution for the file, you are under no obligation to use it.

[johan.rutgeerts]

---

@davebYYPCU wrote:

Pattern on a Path can align parts to the curve, and did so, according to the data existing, Your alignment, 1, was off, and expectation, 2, is unrealistic.  Your Component geometry does not allow three points of a line, to convert into an arc.

It is perfectly feasible to implement functionality that, given a path and a pitch length:

• Calculates the position of chord lines of length 'pitch' along that path, and
• Positions components accordingly.

There's no need for any other data or geometric info etc. for this to work. So I disagree that this expectation is unrealistic. 

We do agree though that this functionality does not currently exist, and that it probably will not be implemented anytime soon. 🙂

That being said, the weird behavior of the 'start point' parameter should be addressed, or at least its behavior should be explained in the documentation. Unless there is some clear logic behind it, I don't think one can consider this other than a bug. So I agree with @kevin_trefflich2 that I would like to see this confirmed (for what it's worth...) as a bug.

Regards,

Johan