Constraint not working on flexible component

[3D4Play]

I typed "constraint not working", because the constraint isn't failing and Inventor isn't throwing an error. However, when changing a LOD, the constraints that I have successfully applied to components made "flexible" now seem to have lost their effect. They are there, they are not suppressed, and they do not fail - but, they no longer work. My flexible components - previously snapped into the correct orientation when constrained - are now misplaced as though one of two applied constraints is ineffective. One constraint is an axis-axis constraint (labeled 1 in the screenshot), the other constraint (labeled 2) is a mate constraint, and relies on component flexibility in order to work. The highlighted assembly is made flexible. Does anyone have any immediate thoughts or suggestions?

[nfarringerHDYXZ]

I've ran into some weird situations where mates between flexible parts would act in this manner. I am not sure if it  would help in your situation, but with most every situation I ran into, I would remove flexibility and then re-add it to the parts. Not sure why it worked in my past situations, but it did.

Can you post the files in question?

Best Regards,

Nate

[johnsonshiue]

Hi! I believe you are on 2020 or earlier, right? Also, the component in question may have positional overrides, right? Please share the files that exhibit the behavior. This may have been resolved on 2021.

Many thanks!

[3D4Play]

Release 2021, build 183. The flexibility is in fact applied to a Positional Rep specifically for the purpose of applying flexibility - i.e. there are four positional reps: three with fixed position, and one with that (angular) positioning constraint suppressed so it can be made flexible to fit any angle in the master assembly.

[3D4Play]

I also tried that, it didn't work. See the attached screencast. The example component is a wall flange that is hinged to swivel. You're seeing a plan view, with a wall that I am dragging the component through, when it should be constraining flat to that wall surface, at the required angle. You can also see that something is happening when flexibility is removed, then reapplied (the flange angle changes, then resets). However, when that specific mate constraint takes effect, it should fully constrain the flange flat to the wall. It did, once, then it just stopped working.

[alex.haerens]

Postional reps and flexibility don't work very well together. I'm under the impression that is either one or the other. As recent as yesterday I came upon a similar problem. I had an assembly working perfectly, with several posreps and viewreps for documentation purposes. The last thing I did was changing the hardware in the assembly with a reference to the released files with the correct part numbers. The hardware itself is an assembly where I thought to be smart to have flexibility built-in, in order to allow customisation on a released, read-only file. I didn't check my posreps after that.

Coming back to the major assembly I noticed that posreps didn't work anymore. When I suppressed all hardware related constraints it was back functioning.

Note : I see you mentioning LOD. If not needed for performance I should stay away from that. Viewreps do the same in a much more reliable way.

Alex

[3D4Play]

The flexibility is being applied to a component that has a posrep explicitly for being used as a flexible assembly. In other words, that same posrep could be the master representation, with the constraint - normally suppressed for flexibility use - deleted altogether. I hate to take a developed component back to square one, when it is set up to work in several different assembly environments, but you're convincing me to swallow hard again, and go backwards.  

BTW, noted about the LOD's, thanks. I use them all of the time, and have had several discussions with AD about how and why they are important to my workflow, and about the issues they cause. They are unavoidable in my particular case, and if they worked correctly (or as intended, consistently, that is) they'd be a super-sweet treat. I'm anxiously following any new developments in this area.

[thomas.fitzgerald]

LOD's and View Reps are 2 entirely different tools to be used in different ways.  Never assume they are similar and one can be replaced by using the other.  LOD's suppress references, reducing memory consumption.  View Reps control visibility and view orientation.  Depending on what you are trying to accomplish will determine which method you use.  

Flexibility can be tricky, I'll admit.  And that is because of graphical dependencies.  Depending on which assembly constraints you use and how they are applied can give you different results and different behaviors.  Take a Mate constraint using the Flush solution with an offset as an example.  Depending on which component you select first verses second, will give you different results.  Personally, because of the nature of how Inventor works and how so many different people can make it work differently, I would recommend creating a video of the behavior, outlining what you expect or desire the outcome to be verses the observed behavior.  It's all about the Design Intent.

[3D4Play]

I deleted the pos reps in the component that was flexible. I save it, then updated the model, then changed the Pos Reps on the flexible components to "Master", then rebuild all. The constraints that had been applied in that model were still present, still not working and still not throwing an error. I had to delete them and reapply them, so I think this is a bug.

Behold the clarity of a screen shot movie, with the behavior in action. You can see I am trying to flex the component back into place, to no avail, until I delete the original mate constraint and apply a new, identical constraint.  Behold the power of an empty tin can sound.

[thomas.fitzgerald]

When looking at the video, Mate 40 was suppressed at the time you deleted it.  If a constraint is suppressed, obviously it won't work.  Also, I noticed that you were observing the LOD's and View Reps in lower level assemblies.  I think there may be a workflow issue with what you're expecting verses what is happening because of lower level assembly representations and how they interact with upper level representation settings.

[3D4Play]

I see what you mean and did not see the suppressed status just then. I may be able to undo back far enough to recheck this. I had about 15 of those components that had to be reconstrained. More to come...

[johnsonshiue]

Hi! Please share the files here or send them (zipped) to me directly (johnson.shiue@autodesk.com). I would like to take a look at the assembly closely.

Many thanks!

[3D4Play]

I've rechecked this, and am getting some very unfavorable and unpredictable constraint behavior. As first glance, it appears as though Inventor is suppressing constraints when a component is suppressed, but is not unsupressing the constraints when the component is unsuppressed. It may be due to using different LODs - I don't know yet. Now Inventor is throwing errors on suppressed constraints. I'm packing this dataset up and sending it to Johnson. By now, I have no more fudge time to casually screw with this - I need to get accurate deliverables out the door. Johnson, you'll get the dataset in its current state soon.