Hi Raul

When you load the model the model will calculate the orientations

(Axes0) of the segments from the values in the mannequin file.

Prevoiusly we had to set the Axes0 matrix for each of the segments

in the body to adjust its position at load time.

In the file InitialPositions.any we have made a “replicate” of the

real kinematics of the model so it will set the Axes0 using the

values from the mannequinfile. When doing this we had to make some

assumptions such as that if the angle between thorax and pelvis is

for example 30 deg. this angle will be distributed equally between

the lumbar joints, this is however not exactly like the “real”

kinematics of the model works since there is a spine rhythm, but it

is good enough as the first guess on the segments position.

This difference means that after you have loaded the model and solve

the first of the kinematics steps the model will move slightly,

primarily due to the assumptions made on the intial positions of the

spine segments. Please keep in mind that the initial positions is

only the first guess on the segments positions once the kinematics

is solve they have no influence at all.

(sorry about this long explanation)

If you look in Mannequin.any file of the FreePostureModel you can

find values that will give you a neutral position of the SC joint.

You are right that in the model you have uploaded even though you

have constant values for gh flexion and abduction these

will “change” when you apply non constant rotation. The reason for

this comes from the way the rotational measure are defined as a

AnykinRotational with the type RotAxesAngles.

Here is the explanation from the manual:

Cardan angles measure a 3-D rotation by three angles

of planar rotation about the reference system’s axes. One should be

aware that the sequence of these rotations is important, and that

any rotation subsequent to the first is a rotation about a local

axis arising from the previous rotation(s).

The GH rotational measure that is used in the model can be found in

the brep/aalborg/arm3d/interface.any. I think it will be possible to

change the sequence of the rotations so that the last rotation is

around the humerus long axis, this will make it behave in more

realistic way, and avoid that the abduction and flexion is “changed”

for humerus rotation I will to do this, but i can not promise when

it will be finished. You can also define your own rotational measure

that behaves in the desired way, you can either modify the one in

the interface file or create a new one.

Best regards

SÃ¸ren, AnyBody Support

— In anyscript@yahoogroups.com, RaÃºl Borraz Lamiel

<raulborrazlamiel@…> wrote:

>

> Hi Soren,

>

> First of all sorry for my explication but my english isn’t very

good.

>

> Why in the initial position in kinematic analysis the

> SternoClavicular goes ahead when the SternoClavicularProtraction

> files are zero?

>

> What angles should we introduce to have a neutral position?

>

> Another question

>

> When we introduce any GlenohumeralRotation, the other angles

> (flexion and abduction) are not respected.

>

> I send you the files of anybody and some pictures about this

> movement whit anybody model and other biomechanical model.

>

> Best regards.

>