Is there any way I can pull this off?
Have two fixed hinges.
When the blue piece moves up.
The red and green pieces move with it but are still constrained by their connections. (Blue dots)
With look at constraints, the objects are not fixed to the blue section. So they float around and don't look like a legit pivot.
Then when I move the entire object as a whole the look at constraint starts to rotate the objects counteractive to to the up and down rotation of the pivot.
I had a quick try at this in maya, from what I understand, the green and red pieces rotate together at the same angle, so a aim constraint is not needed. The blue claw piece is only meant to face in a locked direction/axis and as the red and green pieces rotate the blue claw piece slides forward/up diagonally towards the center.
If this is the desired effect then all you need is a point constraint on the blue claw piece to a locator at the joint on the green piece. I had an additional locator at the same position of the first one that was orient constrained to the green piece's rotational joint, the locators pivot point was also located on the green pieces pivot point. The first locator was parented the the 2nd one.
This way when the green piece rotated it moved the 2nd locator via rotation, I'm not sure how to get translation in worldspace for that so the first locator I mentioned was parented to the rotating 2nd locator and that seems to work with the pivot constraint for the blue claw piece. For the red piece, you can either rotate it as you rotate the green or you could create a set driven key to control both rotational views or just throw in an aim constraint to a third locator at the red/blue piece joint.
This system could be easily built with IK. It would even work well with the non-parallel motion your second image would need to have.
Here's a quick example:
Create a bone for your control (red). Create a seperate bone chain for the rest of the hinge (green) then create an IK solver from the base of the green chain to the end. Parent the IK goal to the control bone. Skin/link geometry to the appropriate bones and you're done.
Replies
Then when I move the entire object as a whole the look at constraint starts to rotate the objects counteractive to to the up and down rotation of the pivot.
Can I over come these somehow?
If this is the desired effect then all you need is a point constraint on the blue claw piece to a locator at the joint on the green piece. I had an additional locator at the same position of the first one that was orient constrained to the green piece's rotational joint, the locators pivot point was also located on the green pieces pivot point. The first locator was parented the the 2nd one.
This way when the green piece rotated it moved the 2nd locator via rotation, I'm not sure how to get translation in worldspace for that so the first locator I mentioned was parented to the rotating 2nd locator and that seems to work with the pivot constraint for the blue claw piece. For the red piece, you can either rotate it as you rotate the green or you could create a set driven key to control both rotational views or just throw in an aim constraint to a third locator at the red/blue piece joint.
Here's a quick example:
Create a bone for your control (red). Create a seperate bone chain for the rest of the hinge (green) then create an IK solver from the base of the green chain to the end. Parent the IK goal to the control bone. Skin/link geometry to the appropriate bones and you're done.
Would that do what you want?
Thanks again everyone.