Interesting Phenomenon in Assemblies

Jul 21, 2006 4 Replies

I saw a very interesting phenomenon in assemblies today. Someone had built a model of a linkage with a hydraulic cylinder driving it. When the hydraulic cylinder was added in, the model started to disassemble itself when it was simply rotated graphically. The problem turned out to be that the joints that the cylinder were mated to were ever so slightly out of parallel. The measure command gave angles on the order of .000000082 degrees. The revolute joints on either end of the cylinder were connected through several levels of subassembly to each other.



When this assembly was rotated the cylinder would sometimes end up a kilometer or so away from the main assembly. Parts of the assembly not even connected to the hydraulic cylinder would end up "exploded" from the main assembly and they would stay that way even through a forced rebuild.



I just wonder if anyone else has seen this kind of thing.



I think so, but with much more easily identified over-constrained conditions. Nice work tracking that one down.

Was the cylinder an imported body?

No the cylinder was made in house in great detail. Flexible or fully defined, the problem still existed.

I can understand the very small angles causing the constraint solver fits, but how this got to mess with the graphics is another thing.

Dale Dunn wrote:

Oh, I managed to miss or forget that part of your post. I've never seen anything move during a view rotation. That's just weird. Do you use a Spaceball? That's the only thing I can think of that can normally affect both view orientatioon and component position. It just shouldn't happen simultaneously.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required