Can't make a simple joint. What am I doing wrong?

[ipmcc]

Hello friends,

I've been fighting with this for most of two days, so please forgive me if I sound a little frustrated. 

I'm trying to model a glass-baffle stair railing. I've modeled a stanchion (a vertical pole) that is screwed into the floor on the bottom stair. I've also modeled the glass grippy clips (which I’ve chosen to call Grabbers) that grab the glass/wood/whatever baffles that prevent (human) children from falling off the side of the staircase. For simplicity's sake, I have only included the most basic situation here: One stanchion. One clip. Put 'em together.

The stanchion is in ‘Stanchion-Redo - 20260216-180203.f3d’, and the glass grabber that I would like to attach to it is in ‘Grabbers - 20260216-180203.f3d’ (There are two grabber components in that file. One is for mounting to the stanchion, and the other is for mounting to a flat wall on the other side, but that’s not relevant to this problem.) I've flailed quite a bit. I really thought that by creating explicit Joint Origins within each independent component, that I’d removed all ambiguity from this, but it seems there’s something else that I’m missing.

If you look at the screen cap video I've attached, here’s what I’m doing (or at least what I think I’m doing):

• In the attach video, I start with an empty document. I've already saved it, but it's a brand new, virgin document. 
• Next, I insert the stanchion from the second file. 
• Then I insert the ‘grabber’ from the third file. 
• After inserting the grabber, I move it away from the origin so that I can interact with the various UI on it. 
• I make the bolt that holds the grabber to the stanchion invisible so that I can more easily click the joint origin that it was obscuring. 
• Then I tried to build a joint between that joint origin and the joint origin on the stanchion. In other words, I'm trying to mount the grabber to the stanchion. 
• Nothing seems to happen. 
• OK, so I try another approach: I use the Align function to align the two components based on their joint origins. 
• Initially, I'm thinking, oh, lovely. At least it's in the right place now. I'll just make an as-built joint. 
• The motion preview for the as-built joint doesn't make any sense, but hey, I'll go on faith, and click OK. 
• Everything still seems to be correctly aligned, but let's try to move the stanchion. 
• …the grabber does not move with it, despite them supposedly being connected by a rigid joint.

I am completely baffled. Why doesn't this work like so many other joints I've made before? The ‘bigger problem’ is so much bigger than this, but if I can’t even get one Rigid Joint to work, then everything else is just a waste of effort. Anyone have any hot tips on a situation like this?

Thanks!

Ian

[davebYYPCU · Accepted solution]

Arriving cold to the file intent, my review - it appears that this Joint Origin is Jointed to a Sketch 1, point.

The Sketch point does not belong to anything but the Assembly Root position.  I figure you have successfully made a Rigid Joint between two different Joint Origins, the moving one is not related in any way to the component you wanted to move.  Duplicate Joint Origins after the Rigid Joint will be impossible to see.

The flat face was used to attach the Joint origin in the second grabber.

I hardly ever use Joint Origin, for this exact example of using duplicated items for the same job.  

Learn to use Joint offset values, and you don't need a Joint Origin.

Might help....

[TheCADWhisperer]

@ipmcc 

Whenever I am analyzing the work of someone else, I go through the entire design process step-by-step.

You should probably move this Fillet to before the Shell...

Now I will continue my analysis...

[TheCADWhisperer]

@ipmcc 

I forgot to mention that is it a good idea to Ground the first Component in an assembly so that everything is predictable behavior (even if you are going to move it later).

[TheCADWhisperer]

@ipmcc 

I would use Assembly Constraints for this.

Axis to Axis

Plane to Plane

Plane to Plane

[ipmcc]

@davebYYPCU I think I understand what you're saying in terms of what's wrong here. Thank you for the insight.

On the more philosophical issues around Joint Origins, I'm curious: I started using Joint Origins in this way because it feels like it captures the 'intent' associated with a part. I've also found them to be especially useful in cases where it's difficult/impossible to get the UI to find the natural/logical/implicit joint origin, like the cases seen here, although clearly this approach is not without its pitfalls. 😁

In terms of 'intent', if you consider the Stanchion part here, I added joint origins: one at the center of the 'cradle' to indicate that a railing tube can/should attach there, one at the base because that's where it should connect to the floor, and then one for where each of the two 'grabbers' attach, in a way that logically lines up with the direction of the cradle. These origins capture the way the part is intended to be used.

In terms of "difficult cases", it tuns out this is another decent example, because we have threaded holes being put into a round/convex tube -- the UI does not naturally show a snap point for this, as best I can tell. Similarly, on the grabbers that would mount to the stanchion, there's a (non-threaded) hole put into a concave surface. In both cases, in terms of signaling intent, I want the origin to be in the correct 'logical location', even though there is no material there: On the stanchion's grabber mount holes, I want the origin to sit 'proud' relative to any 'invisible flat surface' created by the act of drilling the hole -- in other words, where the center of the hole would've been on the un-disturbed surface of the tube/grabber body, prior to drilling and tapping. If I can get all the obvious/correct snap point from the UI, I typically won't create Joint Origins, I'll just use what the UI gives me, but I've found these cross-drilled-holes-in-round-material cases are particularly stymied by the Joint UI. This allows me to find a difficult-to-snap origin once, at creation time, and not every time I bring that part into another assembly 'up the chain'.

I have used also Joint Offsets before. My feeling has been that, just as creating explicit joint origins signals the 'intent', and joint offsets can hide intent. There's not a good example of it in these particular files, but what I've noticed working with Offsets in the past is that it can lead to cases where the little UI glyph representing the joint is no longer close by or logical with respect to where the two parts meet in 3D space. It's not the end of the world, but it is confusion, although I concede that the confusion I've had using Offsets has never been quite as cryptic as this particular Joint Origin problem has been. I've also created Joint Origins with Offsets on the Origin itself; I like this approach because it bakes in the offset, and seemingly unlike the joint glyph, the Origin glyph appears at the point after applying its offset. In other words, offsetting a joint origin feels more readable and self-evident that offsetting an actual joint (that uses implicit joint origins.) 

Does that make sense? Or am I being wrong-headed?

Regards,

Ian

[ipmcc]

Follow-up:

@davebYYPCU found the specific cause of the problem. I was able to fix it based on that info. I'll attach the files as they ended up. I'm still curious what davebYYPCU has to say about the philosophical debate on joint origins and designer intent, but I'm happy to report that I am unblocked for now, and extremely grateful!

I am also interested to try out Assembly Constraints, as @TheCADWhisperer suggested. But before I dive deep on those, what's the sentiment around Constraints vs Joints? Seeing the Assembly Constraints intro video, it raises the question: Why would you use joints any more? Assembly Constraints seem significantly more powerful (or maybe usable), at least for anything you could do with a Rigid joint, and probably more.

The video itself is titled: "Stop Wasting Time! The New Constraints in Autodesk Fusion Make Assemblies Easy", and the video appears to be an official Autodesk publication, so maybe I should get with the program!

Thanks again to all.

Ian

[jhackney1972]

One big thing, at least right now, is you can not animated (drive) Constraints.