Unable to move slider joint

[R4SMEs]

I am trying to have a drawer sliding in a table.

The slider joint looks fine in the process of defining it, i.e., moves as expected for "animate".
However, it is not possible to drag the drawer to open it using the mouse.

For simplicity, I have the drawer implemented as a simple rectangular solid block which can be seen in orange.

The structure of the model is as follows:

Component: post
Body: post [grounded]

Component: foot
...

Component: framework, etc.
...

At this stage, a rigid group was defined to include all of the above, i.e., all components of the table with the exception of the subsequently added component testDrawer (orange block)

So, the timeline is as follows:

Post component/ post body/ ground post/ Foot component/ .../ Framework components/ ... / Rigid Group for all components up to this stage/ testDrawer (orange block) component/ Slider joint for testDrawer to slide relative to the side panel of the framework.

It is not possible to drag the drawer (orange block) to open it in the direction in which it is to slide.
The orange block behaves as if it is part of the rigid group.

However, it was defined (and its slider joint defined) after creation of the rigid group.
The fact that it is not part of the rigid group is seen in the graphic which results from clicking on the rigid group in the browser.
Everything changes to blue, whilst the testDrawer remains in orange:

By clicking on the slider joint in the browser, its location is illustrated:

The grounding of the post and rigid group provide the functionality expected - nothing can be moved.

The problem is that the slider is also locked.

A screen print taken during animation of the slider demonstrates that the slider seems correct, i.e., animation shows the expected motion:

[jeff_strater]

@R4SMEs - we'd have to see the design to be certain.  Can you share it here?  One thing to note:  The "Animate" in the Joint command dialog is not a full system solve (that would be too slow), it only shows what the joint itself will try to do.  Most likely, what is going on is that there are other joints on that drawer that are blocking movement in the entire system.  But, that's just a guess, without the design to look at.  Thanks!

[R4SMEs]

Model is attached - thanks for your assistance!

[jeff_strater]

thanks for the model.  Hmmm...  Stumped so far.  It all looks right.  Animate Model (which does do a full solve, so should be accurate), shows the right behavior, but I can't get that thing to drag, either.  I even re-did the rigid group, with a smaller subset of components (because some are already related with other joints), and re-did the slider, still drag would not work...  Still looking, though.

[R4SMEs]

Thinking about other joints in the model which could possibly be culprits ...

There are various screws - probably they are not culprits.

There is a dowel that was created as a cylinder.
It was patterned x4 and each dowel then joined to locate between holes in the end panels and the rear panel.

There are also sliding wedges in the wedged mortise and tenon joint (post to foot).

For the left hand post/ foot, these behave as expected, i.e., do indeed slide.

I did some mirroring to implement for the right-hand post/ foot.
Although the wedges appear there, for some reason they don't slide.
This is a somewhat similar issue to the non-sliding drawer - so may possibly provide a clue to where problem(s) may be?

I have no real need to be able to slide the wedges in that right-hand post/ foot wedged mortise and tenon - so I didn't investigate why sliding could not be performed.

[jeff_strater]

thanks.  I investigated the other sliding joints, and even suppressed those features, and it didn't seem to make a difference in the drawer behavior.  There is something very strange here.  I did crash Fusion at one point, which I'll have to investigate further...

[jeff_strater]

just to let you know - I have not forgotten this design, just gotten busy the past few days and been unable to look at it.  The design seems pretty straightforward, I don't understand why it doesn't work.  Appreciate the patience...

[R4SMEs]

Thanks for your update and support.

I myself haven't further investigated it but I had a couple of thoughts/ queries:

1) Consider the sequence:
Create a component
Anchor that component
Mirror that component

Something like that may apply to the leg (or post as it is called for the trestle table).

Is it okay to anchor a component that is mirrored.
Does the anchoring need to be before or after the mirroring - or does the sequence matter?

2) If one creates a rigid group by selecting a component which includes subcomponents,
what happens if another subcomponent is subsequently added within that component?

My question here is whether the rigid group relates only to things preceding it in the timeline?
Or, does it encompass what may possibly be added later within the component which is the subject of the rigid group?

In other words, assume there was a component "engine" which had the subcomponents "piston" and "gasket",
we define a rigid group for "engine" and its subcomponents.
Subsequently, we add "connecting rod" as a subcomponent to "engine".

The question is whether or not "connecting rod" would be a member of the previously defined rigid group?

I don't think this is an issue with the table since the rigid group was defined as one of the last steps in the timeline.

[laughingcreek]

interestingly, just grounding the component "front frame strip" without making any other changes allows the slider to work. 

[jeff_strater]

@R4SMEs - I finally got a chance to play with this design.  I suppose you've moved on, but I wanted to try to see what might be going on with this.

To be honest, it is a bit concerning.  I crashed 5-6 times just doing Drive Joints on various joints.  But, I have the design in a state that I think is stable.  Here's what I did:

1. I edited the the slider joint for the bolt/nut joint:  "Slider Nut with washer to bolt".  It is very early in the design.  I noticed that shortly after that joint was added, when I tried to drive it, Fusion crashed.  So, I converted it to a rigid joint.  I was able to leave the wedge slider joints as sliders.
2. after that, the design was pretty well behaved as I stepped through the timeline, until I got to the rigid group near the end of the timeline.  After that joint, the entire design is locked up.  I noticed that rigid group had 66 selections, which is way more than needed.  So, I deleted it, and re-created a rigid group with a minimal number of component selected (13).  This created a group that included all of the right side leg, foot, and top, plus all of the supporting components.  I added another as-built rigid joint between the top and the right side leg, foot, etc.  This created a rigid structure for the whole desk.
3. Then, I stepped to the end, where the drawer is created, and the final slider joint is created.  At this point, the design was in a weird state.  I could make the drawer slide using Drive Joint or Animate Model, but not by component drag.
4. Finally, as @laughingcreek noted, if you then ground the "Front frame strip" component, drag suddenly worked.  I have no idea why...

This version of the design is attached, FYI

I'm going to take all this info (the crashes, the design you sent and create a bug for the assembly modeling team to look at., to see if we can figure out what is going on.  It's a good case, because it is relatively simple, and does not have that many joints.  Thanks for sharing this design, and your patience.

[edit]  The Fusion bug is FUS-64693

[R4SMEs]

Many thanks for your ongoing efforts.

I shall need to look again at the design to better understand the changes you mentioned.

An aspect of Rigid Group which I find rather irritating is that it does not respect the basis of the original definition of the group.

If, for example, there is:

A. Component
A.1 Subcomponent within component A
A.2 Another subcomponent within component A

I define Rigid Group as being "A Component" with the "Include child components" box ticked.

I later create another subcomponent "A.3 Yet another subcomponent within component A" with the timeline positioned before the Rigid Group was created.

However, the Rigid Group does not consider that this child component is within the RG.
Logically, it would seem that it should since I had requested that the RG was the "A Component" with the "Include child components"!
I feel that the RG should properly remember my design intent!

You write:
> ... rigid group had 66 selections, which is way more than needed.

As I recall, I simply clicked on one or more of the top level components, being careful to include everything with the exception of the "Test block" (orange simple "drawer") component.
I considered this to be a reasonable approach to join all components apart from the drawer that needed to slide.
Are you suggesting that it is "bad practice" to have a RG of 66?
[I certainly didn't myself click on 66 components - as I explain above, I had simply selected high level components with the "Include child components" box ticked.]

[jeff_strater]

@R4SMEs - these are both very good questions that reflect a pretty deep understanding of what is going on...  I can discuss them a bit, but don't really have great answers for either of them.

1. Rigid Group and "include child components".  Your observation is correct, and your concern is valid.  Today, "include child components" is a "command time" option.  Meaning that it is just a way to quickly select all the current child components of a component.  As you noticed, it is not a compute time option.  It probably should be, I agree.  There is an analog in the Fillet command:  the "Tangent Chain" option is similarly a command time option today, but should be a compute time option.  We are in the process of fixing that for Fillet.  Hopefully we will get to Rigid Group at some point
2. Is it bad to select lots of components for a Rigid Group?  In theory, no.  The concern here is not so much the number of components as it is the fact that such a wide net includes components that are already joined.  Again, in theory, this should be OK.  In your design, though, you have things like those wedges, which were joined (jointed?  what is the right verb here?) using a slider joint.  If those components are included in the Rigid Group, the RG effectively overrides the sliders.  Again, should be OK.  But, there is a chance that these kinds of joint conflicts can result in the kind of freezing you see, or even "solve failed" errors, because both the slider and the rigid relationship are still present in the solver.  This may be the problem I observed with the fasteners - once I converted the nut/bolt joint to rigid, the design seemed to behave better.  The simplest case should be where components are already rigidly joined, and you put them into a larger RG.  Should be no conflict there.  But, I have seen cases where this kind of double-constraint can lead to errors, so I tend to avoid it.  What I did in my version of the RG is a bit brute-force, but it's how I tend to do things.  I dragged around bits of the desk to see what was free to move.  Then I reverted the move, and created the RG and just picked components (in the graphics area) that were clearly free to move (like the desk top and some of the under-pieces).  Then, I drag again to see what else was left behind, edit the RG, add more stuff, etc.  Tedious, but minimalist.

The other thing that I tend to do is "joints as you go".  Some people prefer to put all the joints at the end - build all the components, and stick them together at the end.  I prefer to build up the sub-assemblies, with joints, as I go.  That way, I usually only need one joint to fasten the sub-assembly to some other component.  But, that's a personal preference.

[Anonymous]

FYI

Hi, I had also a similar problem.

A Frame with 4 sliding doors joint within in sliding profiles. The sliding profiles are different per pair of doors. (part of a bigger frame). The complete frame is grounded but I only can move the doors when the bottom sliding profiles are ground to !! 

Frame is build up with Bosch strut profiles.

greetz

[jhackney1972]

Please attach your model.  If you do not know how to attach your Fusion 360 model follow these easy steps. Open the model in Fusion 360, select the File menu, then Export and save as a F3D or F3Z file to your hard drive. Then use the Attachments section of a forum post to attach it.

[Anonymous]

Thanks for your reaction. 

I attached the design. (still working on it)

[davebYYPCU]

I do not see any problem with your doors and joints.  3 doors have joints 1 door has not yet.

In Fusion all components are free to move until you prevent it. 

The slider rail can be rigid joint or rigid group, but this is required.

Yellow Icons can be fixed with Undo command when they turn up.  Should be fixed as soon as possible.

Might help....

[Anonymous]

Hi, I know that I can move the sliding doors. I was just wondering why I have to 'ground' the sliding rails before I can move the sliding doors.

thx

[davebYYPCU]

All components are free to be dragged in Fusion., when created.

To make the door slide on another component - one component has to be locked from free move.  Just as if you lock the door, the rail slides.  If both are not locked, Fusion will not know what to do, most likely both will move but not slide.

Might help....