Flat assembly destroyed, constrains flipped - FIX IT challenge
Hi Cris,
Today, Played with your data and deleted all constraints and then re-create every constraints from scratch in Inventor 2019, so far, everything goes well and the constrains flipped is not happen any more and the result is much more stable and robust.
Please check out the data and switch the pos-rep to view the result.
Notes, the Max position you've provide is not valid since the arm length is not enough.
Min
Max
Drag-able video -
Note, the entire constraint chain contains >6 DOFs, so it's better to disable some of them when drag.
By the way, please try with flexible for big/small props which should work correctly as well in Inventor 2019.
Hope it helps!
Xun
Hmm...
I don't fill the limits with the range in. I don't see any problem in the video, do you mean the conflitions? the reason for the conflitions happen exactly because of the picture you show me is not match the model.
Limits is an addition condition for sure, but it should not affect the result. Hope you can understand that limits is not the major factor which cause the constraint chain problem.
Flexible stability is more robust since the possible solution with axial direction involved and the result is become unique for entire constrain chain.
You can image that if the system constrain chain is unique, the result is more robust. Flexible is a way to compute sub-constraint chain in the top, so you have to make sure each level constraint chain is unique before leverage flexible approching.
Thanks!
Xun
Your challenge is rig and can never work because of your unreasonable request.
@Cris-IdeasWrote:
Design intent is simple - This is to be constrained in the way so it's DOFs are equivalent to real life construction. (This was actually build and used on construction site).
Criteria for the challenge:
- it must be stable,
- it must be draggable within its proper envelope. So it was clear that constrains positioning upper beams can be applied separately from structure's internal constrain set,
1) In real life construction you cannot drag the beams to work, this is turn buckle adjustment in real life and that is what you would have to do in Inventor to see it work.
2) If the location of the "Motion Envelope" is correct with these numbers your design has failed already and cannot fit into the window you are asking.
If that is the correct location the "rama 1" is not design correctly from the start as you can see in the photo below.
I can make this work like it was real life by turning the buckles but it would never fit your Motion Envelope.
As you can see in @Xun.Zhang photos the "Motion" Envelope has moved, I believe this would not meet your criteria because you had a location for the "Motion Envelope".
This would be easy to do.
Just a suggestion if you want to animate anything in Inventor I have found that offset constraints makes it unstable, and mating to assembly plains not grounding the 1st part and constraining from it also makes animation unstable. Not say you can't just more stable if you don't.
Hello Cris,
Thank you so much for the comments and you are simply awesome!!!
Good to know the flip problem is gone.
Is that possible to study the case further? Please help modify the data on top of it and add range limits, make or modify any constant based on the data to correct component position and send it back to me.
I will talk with the Math guy who major working on the solver and see if there is something we can improve for sure.
By the way, please try Drive the constrain instead of drag, as far as I know, drag may lead the value beyond limit boundaries and the result is not that obvious.
Please let me know if you have any more findings.
Thank you so much!
Xun
Drag would not lead to solution outside limits if props are properly limited.
Hi @Cris-Ideas,
I don't have the time to go through the whole assembly, but I had a quick look, and one of the things that stood out to me was the use of sketches and work points for use in the constraints. I have not found that method to be a reliable way to work, and so I wonder if another approach would help.
Here's one example of an alternative approach to the "props" that came to mind for me, but I'm sure there are others. I'm not suggesting that this is the way you need to do it, but I would suggest thinking about some other approach to see if getting away from the use of workpoints and sketches helps.
As for dragging, keep in mind that you can often defeat the solver by dragging out side of the limits before your computer's processor has a chance to catch up and do the calculations... so dragging a complex mechanism can be unreliable sometimes.
I hope this helps.
Best of luck to you in all of your Inventor pursuits,
Curtis
http://inventortrenches.blogspot.com
Hi Cris,
This is a case you and I took a look a while back. I think the key issue here is the concept of symmetry in the actuators. The both ends of the actuators have to move inward or outward in the same amount. Although there is degree of freedom, you cannot use one constraint to drive it. It is like you have multiple drivers in the system fighting each other. I personally think this is a limitation in the solver. I am not saying it is ideal. Nor do I think it should lock up like this. But, I don't see a solution to the problem at the moment.
Many thanks!
Johnson Shiue (johnson.shiue@autodesk.com)
Software Test Engineer
