Drive constraint

[roberts4S7QE]

Hi All,

So besides the HIGHLY annoying "Cannot solve at (or near) this point" error when trying to drive an assembly with more than one sub-assembly set to flexible, and Autodesk whom have stated they are investigating it since 2021 with no change, I have to suppress all but one of any of the flexible sub-assemblies in order to be able to animate the constraint drive. But then I found this new problem I can't seem to fix...

Running the drive tool back and forth, everything works perfectly, until I hit "OK", then everything goes to hell.

See attached video.

This still happens even after removing all flexible sub-assemblies entirely, and I have deleted all constraints and reassembled from scratch only to end up with  the exact same glitch again.

Has anyone experienced something like this before and come up with a fix?

[JDMather]

---

@roberts4S7QE wrote:

…and come up with a fix?

---

Attach your assembly here.

[roberts4S7QE]

Hi JD,

I am unable to create a zip folder from pack and go to attach the assembly. Even forcing pack and go to create a zip folder on file creation, the zip folder becomes corrupted

[CStilesCARE]

First thing I'd look for after seeing that is to check that any and all angle constraints are using the "explicit reference vector" option (choice on the right side of 3 options) and change them to use that if not.

Angle constraints are notorious for flipping 180 degrees at random times if they're not limited by a reference, so I will use that option 99.9% of the time for angle constraints, unless I know for sure that direction of rotation won't matter (like when constraining component origin planes to face same way as assembly origin planes).

For the 3rd selection option, just pick something that can act as the "axis" for rotation. The constraint will use that to determine which way 0 degrees is, and which direction is positive/negative rotation.

[johnsonshiue]

Hi! I guess I was the one offering to help take a look. If I recall correctly, we don't have a good solution yet. But, it is possible my memory is corrupted. Please use Pack&Go to copy all the files to a separate folder, zip up the entire folder, and share it here. Or, send it to me directly johnson.shiue@autodesk.com. I would like to take a look to see if this is a known issue.

Many thanks!

[Frederick_Law]

Hate those flipping constrains.

Sometime I cheat and use linear to drive it.

Don't allow the linkage to got straight and able to flip to the other side.

That apply to CAD and real.

[roberts4S7QE]

Hi CStiles,

There are no angle constrains in this assembly. I used mates, flush mates, and insert constrains for rotation. The idea is to drive the doors open as they would in reality. I never figured out how angle constrains work anyway 😂

[CStilesCARE · Accepted solution]

Ah, ok. In this case, maybe adding an angle constraint between the little braces (circled in snip below) and the lower/upper frame would help prevent the one door from flipping over.

You could use its bolting location at the frame as the 3rd "axis" selection, then (after establishing which rotation direction is positive) just give it a min and max range that lay a little outside its actual range of movement, so it's not limited, but should stop it from being able to flip.

An alternative option (if the doors always move symmetrically to each other) could be to use a symmetry constraint between the vertical lengthwise origin plane on the brace of both doors, with the center vertical plane of the overall assembly (awkward sketch below). This assumes the parts and assembly were created centered on their respective origins, though you can create work planes tha

[roberts4S7QE]

Thanks CStiles.

Creating a symmetric constraint between the two arms works beautifully!