I've attached a picture:
(consider opening on another tab to make it easier to see - middle button on many browsers)
1 - An axle/arbor (Pinned coincident to a datum axis - free on other directions)
2 - A gear (Well attached to the Axle, where it IS to be)
3 - Washer/Separator (attached to the gear and the arbor) [there's another one behind the gear]
4 - Inner runner of a ball bearing (Attached to the Washer/Separators (3))
5 - Outer runner of a ball bearing (pinned and aligned with the inner runner) - This part is important, because when I attach these parts to another assembly, the outer runner is the interface
More clarifications:
If I restrain the movement of the outer runners, I can still move (spin) the gear (which also moves the axle and the inner runners)
Which is the behavior I want
Why Creo disallows converting the gear onto a gear (when I attach these parts to another assembly containing another gear) ?
(Unfortunately, the orange circled arrow belongs to the bearing, not the gear)
I am not sure I understand the desired results. However, you should be able to use the pin constraint at higher levels but you just have to be careful of making sure your ground elements remain... or worse, that new ground elements are not being created. You can turn this on in the model/mechanism tree.
Quite often I need to create, or create as a matter of convenience a dummy part only to manage the assembly and motion. I find that mechanism and the logic of your device do not follow the same logic. There is also something going on when axes are not oriented as the software expects. I have trouble with simulated rack and pinion constraints where the axes are parallel for instance. I am only trying to achieve a rotation to linear displacement, but the software wants specific orientations as well. I only mention this in case a spur gear setup does not recognize bevels, for instance.
In general, I get everything I want out of mechanism but rarely the 1st time. I create extra servo motors and dummy parts if the system won't cooperate. Many times the motion I get is not even close to logical to what I thought I set up. Sometimes, elements get stuck where I have to delete the entire constraint set to "unstick" the problem. Point is, what seems to be a limitation isn't necessarily the real problem.
Also, don't forget you can add multiple constraint sets. I think you may have done this, but this helps by not over constraining a primary constraint.
A lot to cover here !
If I'm interpretating correctly your description, you've stumbled on bugs !
I think me too.
I did put many constraints, following the real world logic.
And it works like a charm. Using the Drag component.
There IS a structure I could call dummy.
I did "delete entire constraints" many times, actually.
Create more than one set when convenient. Probably the case here.
If you find yourself free time, would you (or anyone) mind opening my assembly (turret.asm).
(And don't forget to right click and activate the XSEC0001 section - the last feature of the model tree)
I can open it later this week. I don't have the right computer with me.
I did learn that pistons and pins act completely different. So yes, a lot of this is about knowing what the software is really thinking rather than the description.
No need to hurry
I'm only asking because it's been some long I've been trying to make this work.
Usually I'd do it myself. Meshing mechanisms is a game to me. And I DO like to play with it (I bet many of the peers around this community also enjoy).
But really, this one is making me give up !
Don't mind the mess the file is. I haven't yet made acleanup on all those datums.