Thank you for your suggestions. As I understand the whole point of motion skeletons is that they act like a both regular skeleton and a mechanism. But according to your explanation motion skeletons are no better than a regular mechanism. If this is the case then motion skeletons would be redundant, which doesn't mean what you say isn't true.
As to your other suggestions
- Snapshot constraints - These are difficult to maintain as they go out-of-date the moment anything in the model changes. I could setup the constraints each time but since I may have as many as 50 subassemblies that each contain the motion skeleton it would be very time consuming to setup the constraints each time I wanted to move a particular model.
- No sub-assemblies - Moving all the parts that are contained in the model of a combine into one assembly isn't practical. There are thousands of parts, hundreds of weldments, and hundreds of other assemblies. We definitely need subassemblies to have any hope of organization. Regardless this isn't as option as the company has specific standards on how the model tree is to be organized.
- Regeneration positions - Doing this would mean you could only ever see everything in one default position, you could never actually do any motion analysis. If I wanted to move something to other position then I would be back to square one - only one assembly will move while the surrounding assemblies with the same motion skeleton do not move.
Thanks again for the suggestions. I have also thought of these and tried them but they ended up being dead ends.