AW: Re: Position of the elbow FE joint

Hi Michael,
no problem as long as there will be an answer for me. I found the humerus fe
node and it´s location is ok, but the second joint patrner, the ulna has an fe
node too, but with other coordinates. In all the other joints both joint
partners have equal coordinates in the nodes which are used to define the joint.
They don´t differ much, but they are not equal. And there are some sRel defined
which i can´t understand and which doesn´t make the nodes fit better. I use the
free posture standing model. Because of this joint i´m now a little bit confused
about “r”, “sRel” and other funny positioning tools. If you can show me the part
of the file which makes the ulna.fe node and the humerus.fe node having the same
position this would be a great help.
With best regards, Maren

----- Ursprüngliche Mail ----
Von: AnyBody Support <support@anybodytech.com>
An: anyscript@yahoogroups.com
Gesendet: Freitag, den 7. Juli 2006, 11:28:36 Uhr
Betreff: [anyscript] Re: Position of the elbow FE joint

                        Hi Maren,

It’s Michael. Soeren is on vacation, so I shall try to cover for him.
I am not completely sure either; I have only little experience with
the arm model myself.

I found this node on the Humerus segment:

Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe

I believe ‘fe’ means Flexion-Extension. Inside it, there is a another
node, which is only rotate relative to the ‘fe’ node. It is called

Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe.RotNode

It is used to define the Flexion-Extension angular measure, as far I
can see.

I made an AnyDrawRefFrame
inside ‘Main.Model. HumanModel. Right.ShoulderAr m.Seg.Humerus. fe’ and it
looks like it is located in the elbow alright.

Please have a look and see if this is what you need, or write me again
and complain;-)

Best regards,

Michael
AnyBody Support

— In anyscript@yahoogrou ps.com, Maren Lühring <anystarter@ …> wrote:
>
> Hi Sören,
> the training in Aalborg helps me a lot when i´m now working with the
model. In this moment i try to find aout where the joints are, to
scale the model. Most of them are easy to find but with the elbow it
seems to be complicate: the RefNodes for the joint definition from
humerus and ulna are different while all the other joints have the
same RefNode coordinates for both joint partners. Which node defines
the position of the elbow fe joint?
> With best regards, Maren
>
>
> ------------ --------- --------- —
> Sie denken an Ihre Sicherheit? Das tun wir auch.
>
> [Non-text portions of this message have been removed]
>

[Non-text portions of this message have been removed]

Hi Michael,
did I understand right that before any analysis there is no need for the joint
coordinates to be the same for both joint partners? Till now I thougt that “r”
describes the absolute position of a reference system and “r0” is the position
from where on the analysis starts. But you say there are other constraints which
are violated. Where can I find them? So I recapitulate: If I like to now the
exact position of a joint I have to run a kinematic analysis. And only after
having done this I get the right positions for all joints. Is it the same for
all “RefNodes” ? Where they all influenced by the analysis or does for example
the “sRel” always stay the same? Now I tried out to click on the Nodes after a
kinematic analysis and they have got the same coordinates: my problem is solved,
but I´m still a bit confused about the definition of positions in AnyBody. How
does the joint decide where it will be after the kinematic analysis when there
are two different coordinates, or will
they meet in the middle? Is there any manual who makes the structure of this
human model clearer? With best regards, Maren

----- Ursprüngliche Mail ----
Von: AnyBody Support <support@anybodytech.com>
An: anyscript@yahoogroups.com
Gesendet: Freitag, den 7. Juli 2006, 13:39:16 Uhr
Betreff: [anyscript] Re: Position of the elbow FE joint

                        Hi Maren

I am not sure whether I understand your problem correctly

There is a Revolute joint called:

Main.Model.HumanMod el.Right. ShoulderArm. Jnt.FE

that uses the two nodes

Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe (the one I
mentioned)

and

Main.Model.HumanMod el.Right. ShoulderArm. Seg.Ulna. fe

The (global) position of these node, ‘r’, are indentical as far as I
can see, but only after you solve the kinematic analysis.

I have dumbed the following values from a model by double-clicking
the ‘r’-objects:
------------ --------- --------- -Objects dump-------- --------- ---------
Main.Model.HumanMod el.Right. ShoulderArm. Seg.Ulna. fe.r = {0.1215727,
1.315634, 0.2169811};
------------ --------- --------- -Objects dump-------- --------- ---------
Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe.r =
{0.1215727, 1.315634, 0.2169811};

Before the kinematical analysis has done at least one step (either
by the SetInitialPosition or the real kinematic analysis), the
values look like this:
------------ --------- --------- -Objects dump-------- --------- ---------
Main.Model.HumanMod el.Right. ShoulderArm. Seg.Ulna. fe.r = {-
0.007244633, 1.249246, 0.2856014};
------------ --------- --------- -Objects dump-------- --------- ---------
Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe.r = {-
0.009015867, 1.25663, 0.2830825};

Here the points do not coincide, but this is because the kinematic
constaints are violated. The violation is small because the guy that
made the model did a good job on making the initial positions, but
he didn’t do it perfectly (he has no reason to). Good initial
conditions are just needed
On the other joints he might have done a perfect job of matching the
joint translations (if ‘r’ match), but he didn’t have to. Also, it
is not all joint type that actually do match the translations in all
directions.

‘sRel’ is the local (segment-fixed) coordinates of the node,
whereas ‘r’ is the global position - just to make that clear. The
global position is given by

r_node = r_owner + A_owner*s_rel

where A_owner is the segment frame rotational transformation matrix
(i.e. ‘Axes’ in AnyScript terms).

Just a hint for locating stuff in the files:
If you want to find ulna.fe or something else in the code and you
have the objects in the Model Tree. Then right-click the object in
question and select “Dumb locations”. In the Operations/Report View
it will dumb some text with some file-links. Here you get
Construction and sometimes also Initialization locations. You can
double-click the file location and you will jump to the place in the
file.
I know this is well hidden, but it is a nice feature for getting to
specific places in the files.

I hope some of this helps you,

Michael
AnyBody Support

— In anyscript@yahoogrou ps.com, Maren Lühring <anystarter@ …>
wrote:
>
> Hi Michael,
> no problem as long as there will be an answer for me. I found the
humerus fe node and it´s location is ok, but the second joint
patrner, the ulna has an fe node too, but with other coordinates. In
all the other joints both joint partners have equal coordinates in
the nodes which are used to define the joint. They don´t differ
much, but they are not equal. And there are some sRel defined which
i can´t understand and which doesn´t make the nodes fit better. I
use the free posture standing model. Because of this joint i´m now a
little bit confused about “r”, “sRel” and other funny positioning
tools. If you can show me the part of the file which makes the
ulna.fe node and the humerus.fe node having the same position this
would be a great help.
> With best regards, Maren
>
> ----- Ursprüngliche Mail ----
> Von: AnyBody Support <support@… >
> An: anyscript@yahoogrou ps.com
> Gesendet: Freitag, den 7. Juli 2006, 11:28:36 Uhr
> Betreff: [anyscript] Re: Position of the elbow FE joint
>
> Hi Maren,
>
> It’s Michael. Soeren is on vacation, so I shall try to cover for
him.
> I am not completely sure either; I have only little experience
with
> the arm model myself.
>
> I found this node on the Humerus segment:
>
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe
>
> I believe ‘fe’ means Flexion-Extension. Inside it, there is a
another
> node, which is only rotate relative to the ‘fe’ node. It is called
>
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe.RotNode
>
> It is used to define the Flexion-Extension angular measure, as
far I
> can see.
>
> I made an AnyDrawRefFrame
> inside ‘Main.Model. HumanModel. Right.ShoulderAr m.Seg.Humerus.
fe’ and it
> looks like it is located in the elbow alright.
>
> Please have a look and see if this is what you need, or write me
again
> and complain;-)
>
> Best regards,
>
> Michael
> AnyBody Support
>
> — In anyscript@yahoogrou ps.com, Maren Lühring
<anystarter@ …> wrote:
> >
> > Hi Sören,
> > the training in Aalborg helps me a lot when i´m now working
with the
> model. In this moment i try to find aout where the joints are, to
> scale the model. Most of them are easy to find but with the elbow
it
> seems to be complicate: the RefNodes for the joint definition
from
> humerus and ulna are different while all the other joints have
the
> same RefNode coordinates for both joint partners. Which node
defines
> the position of the elbow fe joint?
> > With best regards, Maren
> >
> >
> > ------------ --------- --------- —
> > Sie denken an Ihre Sicherheit? Das tun wir auch.
> >
> > [Non-text portions of this message have been removed]
> >
>
>
>
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

[Non-text portions of this message have been removed]

Hi Maren

I believe you understand things correct. Before you run the
SetInitialConditions the joint coordinates doesn’t have to be the same.
You should see the joints as a way of removing DoF. When you connect two
AnyRefNode in a joint, you connect their two coordinate systems
together. One segment has 6 DoF, because it can move around in the
coordinate system in 3 directions and it can rotate around 3 axes. Two
segments has 12 DoF. When you tie two segments together with a joint you
remove at least three DoF, because they cannot move relatively to each
other. Depending on which joint you are using, you remove the rotational
DoF. You are right that r0 is the initial position and r is the absolute
position. i.e. if the model is driven by some kinematic data the r will
change during the movement.
The sRel is the position in the local coordinate system, and it stays
the same… I believe…
Well the joint position is decided from somewhere else, i.e. the model
could be attached to the global coordinate system in the most proximal
part or it could be just driven by kinematic markers on bony landmarks.
I don’t think there is a manual to any of the models, but I am not sure?
Hope I answered some of your questions, else I can try again:D

Regards Christian

Maren Lühring wrote:
>
> Hi Michael,
> did I understand right that before any analysis there is no need for
> the joint coordinates to be the same for both joint partners? Till now
> I thougt that “r” describes the absolute position of a reference
> system and “r0” is the position from where on the analysis starts. But
> you say there are other constraints which are violated. Where can I
> find them? So I recapitulate: If I like to now the exact position of a
> joint I have to run a kinematic analysis. And only after having done
> this I get the right positions for all joints. Is it the same for all
> “RefNodes” ? Where they all influenced by the analysis or does for
> example the “sRel” always stay the same? Now I tried out to click on
> the Nodes after a kinematic analysis and they have got the same
> coordinates: my problem is solved, but I´m still a bit confused about
> the definition of positions in AnyBody. How does the joint decide
> where it will be after the kinematic analysis when there are two
> different coordinates, or will
> they meet in the middle? Is there any manual who makes the structure
> of this human model clearer? With best regards, Maren
>
> ----- Ursprüngliche Mail ----
> Von: AnyBody Support <support@anybodytech.com
> <mailto:support%40anybodytech.com>>
> An: anyscript@yahoogroups.com <mailto:anyscript%40yahoogroups.com>
> Gesendet: Freitag, den 7. Juli 2006, 13:39:16 Uhr
> Betreff: [anyscript] Re: Position of the elbow FE joint
>
> Hi Maren
>
> I am not sure whether I understand your problem correctly
>
> There is a Revolute joint called:
>
> Main.Model.HumanMod el.Right. ShoulderArm. Jnt.FE
>
> that uses the two nodes
>
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe (the one I
> mentioned)
>
> and
>
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Ulna. fe
>
> The (global) position of these node, ‘r’, are indentical as far as I
> can see, but only after you solve the kinematic analysis.
>
> I have dumbed the following values from a model by double-clicking
> the ‘r’-objects:
> ------------ --------- --------- -Objects dump-------- --------- ---------
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Ulna. fe.r = {0.1215727,
> 1.315634, 0.2169811};
> ------------ --------- --------- -Objects dump-------- --------- ---------
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe.r =
> {0.1215727, 1.315634, 0.2169811};
>
> Before the kinematical analysis has done at least one step (either
> by the SetInitialPosition or the real kinematic analysis), the
> values look like this:
> ------------ --------- --------- -Objects dump-------- --------- ---------
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Ulna. fe.r = {-
> 0.007244633, 1.249246, 0.2856014};
> ------------ --------- --------- -Objects dump-------- --------- ---------
> Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe.r = {-
> 0.009015867, 1.25663, 0.2830825};
>
> Here the points do not coincide, but this is because the kinematic
> constaints are violated. The violation is small because the guy that
> made the model did a good job on making the initial positions, but
> he didn’t do it perfectly (he has no reason to). Good initial
> conditions are just needed
> On the other joints he might have done a perfect job of matching the
> joint translations (if ‘r’ match), but he didn’t have to. Also, it
> is not all joint type that actually do match the translations in all
> directions.
>
> ‘sRel’ is the local (segment-fixed) coordinates of the node,
> whereas ‘r’ is the global position - just to make that clear. The
> global position is given by
>
> r_node = r_owner + A_owner*s_rel
>
> where A_owner is the segment frame rotational transformation matrix
> (i.e. ‘Axes’ in AnyScript terms).
>
> Just a hint for locating stuff in the files:
> If you want to find ulna.fe or something else in the code and you
> have the objects in the Model Tree. Then right-click the object in
> question and select “Dumb locations”. In the Operations/Report View
> it will dumb some text with some file-links. Here you get
> Construction and sometimes also Initialization locations. You can
> double-click the file location and you will jump to the place in the
> file.
> I know this is well hidden, but it is a nice feature for getting to
> specific places in the files.
>
> I hope some of this helps you,
>
> Michael
> AnyBody Support
>
> — In anyscript@yahoogrou ps.com, Maren Lühring <anystarter@ …>
> wrote:
> >
> > Hi Michael,
> > no problem as long as there will be an answer for me. I found the
> humerus fe node and it´s location is ok, but the second joint
> patrner, the ulna has an fe node too, but with other coordinates. In
> all the other joints both joint partners have equal coordinates in
> the nodes which are used to define the joint. They don´t differ
> much, but they are not equal. And there are some sRel defined which
> i can´t understand and which doesn´t make the nodes fit better. I
> use the free posture standing model. Because of this joint i´m now a
> little bit confused about “r”, “sRel” and other funny positioning
> tools. If you can show me the part of the file which makes the
> ulna.fe node and the humerus.fe node having the same position this
> would be a great help.
> > With best regards, Maren
> >
> > ----- Ursprüngliche Mail ----
> > Von: AnyBody Support <support@… >
> > An: anyscript@yahoogrou ps.com
> > Gesendet: Freitag, den 7. Juli 2006, 11:28:36 Uhr
> > Betreff: [anyscript] Re: Position of the elbow FE joint
> >
> > Hi Maren,
> >
> > It’s Michael. Soeren is on vacation, so I shall try to cover for
> him.
> > I am not completely sure either; I have only little experience
> with
> > the arm model myself.
> >
> > I found this node on the Humerus segment:
> >
> > Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe
> >
> > I believe ‘fe’ means Flexion-Extension. Inside it, there is a
> another
> > node, which is only rotate relative to the ‘fe’ node. It is called
> >
> > Main.Model.HumanMod el.Right. ShoulderArm. Seg.Humerus. fe.RotNode
> >
> > It is used to define the Flexion-Extension angular measure, as
> far I
> > can see.
> >
> > I made an AnyDrawRefFrame
> > inside ‘Main.Model. HumanModel. Right.ShoulderAr m.Seg.Humerus.
> fe’ and it
> > looks like it is located in the elbow alright.
> >
> > Please have a look and see if this is what you need, or write me
> again
> > and complain;-)
> >
> > Best regards,
> >
> > Michael
> > AnyBody Support
> >
> > — In anyscript@yahoogrou ps.com, Maren Lühring
> <anystarter@ …> wrote:
> > >
> > > Hi Sören,
> > > the training in Aalborg helps me a lot when i´m now working
> with the
> > model. In this moment i try to find aout where the joints are, to
> > scale the model. Most of them are easy to find but with the elbow
> it
> > seems to be complicate: the RefNodes for the joint definition
> from
> > humerus and ulna are different while all the other joints have
> the
> > same RefNode coordinates for both joint partners. Which node
> defines
> > the position of the elbow fe joint?
> > > With best regards, Maren
> > >
> > >
> > > ------------ --------- --------- —
> > > Sie denken an Ihre Sicherheit? Das tun wir auch.
> > >
> > > [Non-text portions of this message have been removed]
> > >
> >
> >
> >
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>
>