Results of IK depends on the joint position

[Michala Cadova]

Hi,
I have probably a silly question, but I want to be sure, that I understand the relationship right.
I am right now working with the model of the mandible, so I will use this to describe my problem, but it should hold for every model.

First of all, some relationship: origin of the coordinate system (CS) of the skull is [0,0,0] as well as the mandible [0,0,0].

The temporomandibular joint (TMJ) is located in [-0.027; 0.063; 0.0414] (right joint; coordinates are the same for parent body as well as for the jaw itself).

Now, I run the inverse kinematics (IK) to get the generalized coordinates in time based on the applied movement of the markers which are connected to the lower jaw (mandible). I defined the TMJ with 6 degree of freedom. Then, from the IK I have three rotations and three translation. And here goes my first question:

The generalized coordinates describe the relationship between the two coordinate system, right? Between the CS of the skull and that of the mandible. The translation goes from the jaw’s origin [0,0,0] and NOT from the position of the joint? The generalized coordinates are connected to the movement of the whole body, is that right?

So why do I get different results (from IK), if I, for the same model, define the joint in the left TMJ, or somewhere in between, let’s say in [0,0,0]? The movement should follow the displacement of the three LEDs I use, so the results should be the same.

Then, I run the inverse dynamics (ID) with some external forces and with the movement I got from IK. Now the results for 6DOF are torques and forces in the joint? That means the force relationship in the position of the joint ([-0.027; 0.063; 0.0414]), right?

I hope, that the questions are not too confusing.

Thanks for any hints.

Michala