Classical inverse dynamic analysis

Hi again,

Is it possible to perform a classical inverse dynamic analysis on the model?
I would like to calculate joint torques and reaction forces at the joints
required to perform a certain motion. In the tutorial this step is skipped,
saying: “Since we have no muscles so far, kinematic analysis is really all
that makes sense”. But in my case there is some relevant information I’d
like to extract from this.

I tried using an AnyForceMomentMeasure and including all muscles that act on
the joint, but it seems that these muscles are not considered to be applied
directly (am I right?), so I get a zero moment vector.

I also tried looking at the Constraints.Reaction, but those are just, as the
name says, the constrained reaction, and therefore do not provide the
torques in the non-constrained degrees of freedom.

How can I get the joint torque in the non-constrained degrees of freedom of
a joint?

Thanks you,

Jean-Olivier


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of AnyBody Support
Sent: 21 avril 2007 15:12
To: anyscript@yahoogroups.com
Subject: [anyscript] Re: Contraints.Reaction.Fout

Hello Jean-Olivier,

Contraints.Reaction.Fout is a menber of any constraint. It is the
force(s) in the constraint. If the constraint is a joint then
Contraints.Reaction.Fout contains the joint reactions just as you
anticipate.

But here is the point of the story: Joint reactions are very
different depending on whether you have muscles in the model or not.
Joint reaction forces in models relying only on joint torques, i.e.
classical inverse dynamics, are hugely under estimated. If you only
have joint torque providers in the model, then the joint reactions
are only what the external forces produce. When you include the
muscles, their contribution to the joint reaction is added and it is
often larger than what the xternal forces provide. A muscle can only
produce torque in a force couple with a joint reaction, so the real
joint reactions are much larger than what you get from a model with
pure joint torques.

In other words, it is correct that you get different joint reactions
depending on whether you include muscles in the model or not.

Best regards,
John

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
Jean-Olivier Racine <jean-
olivier.racine.1@…> wrote:
>
> Hello all,
>
>
>
> I am wondering what are the Contraints.Reaction.Fout values. As I
run my
> model with and without muscles, I get different values for those.
Therefore,
> I conclude that these are not the Reaction Forces, as defined when
doing
> classical inverse dynamics (reaction forces and moments passed from
one
> joint to the other), since they should not change with or without
muscles.
> So my questions:
>
>
>
> 1. What are these Constraint.Reaction.Fout value?
> 2. How can I get those reaction forces that come directly from
the
> classic inverse dynamics analysis?
>
>
>
> Thanks!
>
> Jean-Olivier
>
>
>
> [Non-text portions of this message have been removed]
>

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

I just realized that there’s a Reaction.Type not only in joints, but also in
drivers. So I suppose that if I set this to ON, I should get the required
information. I’ll try that and see what happens. Sorry about that.


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of Jean-Olivier Racine
Sent: 25 avril 2007 10:34
To: anyscript@yahoogroups.com
Subject: [anyscript] Classical inverse dynamic analysis

Hi again,

Is it possible to perform a classical inverse dynamic analysis on the model?
I would like to calculate joint torques and reaction forces at the joints
required to perform a certain motion. In the tutorial this step is skipped,
saying: “Since we have no muscles so far, kinematic analysis is really all
that makes sense”. But in my case there is some relevant information I’d
like to extract from this.

I tried using an AnyForceMomentMeasure and including all muscles that act on
the joint, but it seems that these muscles are not considered to be applied
directly (am I right?), so I get a zero moment vector.

I also tried looking at the Constraints.Reaction, but those are just, as the
name says, the constrained reaction, and therefore do not provide the
torques in the non-constrained degrees of freedom.

How can I get the joint torque in the non-constrained degrees of freedom of
a joint?

Thanks you,

Jean-Olivier


From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com] On
Behalf
Of AnyBody Support
Sent: 21 avril 2007 15:12
To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
Subject: [anyscript] Re: Contraints.Reaction.Fout

Hello Jean-Olivier,

Contraints.Reaction.Fout is a menber of any constraint. It is the
force(s) in the constraint. If the constraint is a joint then
Contraints.Reaction.Fout contains the joint reactions just as you
anticipate.

But here is the point of the story: Joint reactions are very
different depending on whether you have muscles in the model or not.
Joint reaction forces in models relying only on joint torques, i.e.
classical inverse dynamics, are hugely under estimated. If you only
have joint torque providers in the model, then the joint reactions
are only what the external forces produce. When you include the
muscles, their contribution to the joint reaction is added and it is
often larger than what the xternal forces provide. A muscle can only
produce torque in a force couple with a joint reaction, so the real
joint reactions are much larger than what you get from a model with
pure joint torques.

In other words, it is correct that you get different joint reactions
depending on whether you include muscles in the model or not.

Best regards,
John

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
Jean-Olivier Racine <jean-
olivier.racine.1@…> wrote:
>
> Hello all,
>
>
>
> I am wondering what are the Contraints.Reaction.Fout values. As I
run my
> model with and without muscles, I get different values for those.
Therefore,
> I conclude that these are not the Reaction Forces, as defined when
doing
> classical inverse dynamics (reaction forces and moments passed from
one
> joint to the other), since they should not change with or without
muscles.
> So my questions:
>
>
>
> 1. What are these Constraint.Reaction.Fout value?
> 2. How can I get those reaction forces that come directly from
the
> classic inverse dynamics analysis?
>
>
>
> Thanks!
>
> Jean-Olivier
>
>
>
> [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 Jean-Oliver

Yes you are right, if you have the reaction.type in the drivers set
to “On” the driver will carry the load and you can read out its value.

Alternatively if you are not driving all the human joints directly
and want to have them carried you can also apply a AnyReacForce to
the joint, in this way the free dof in joint will also be carried
but no motion will be supplied.

If you are using a repository model and you are using a model with
nomuscles on it will also be possible to read out the joint moments.
By default these versions supply joint torque muscles to the joints
and by reading out the forces in these joint torque muscles you have
the joint moments. As i recall it these can be found named
JointMuscles. Please notice that there is a muscle for each dirction
of the moement.

You are right about the ForceMomentMeasure it will currently not
measure the moment generated by the muscles, this will however be
possible in future versions.

Best regards
Søren, AnyBody Support

— In anyscript@yahoogroups.com, Jean-Olivier Racine <jean-
olivier.racine.1@…> wrote:
>
> I just realized that there’s a Reaction.Type not only in joints,
but also in
> drivers. So I suppose that if I set this to ON, I should get the
required
> information. I’ll try that and see what happens. Sorry about that.
>
>
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of Jean-Olivier Racine
> Sent: 25 avril 2007 10:34
> To: anyscript@yahoogroups.com
> Subject: [anyscript] Classical inverse dynamic analysis
>
>
>
> Hi again,
>
> Is it possible to perform a classical inverse dynamic analysis on
the model?
> I would like to calculate joint torques and reaction forces at the
joints
> required to perform a certain motion. In the tutorial this step is
skipped,
> saying: “Since we have no muscles so far, kinematic analysis is
really all
> that makes sense”. But in my case there is some relevant
information I’d
> like to extract from this.
>
> I tried using an AnyForceMomentMeasure and including all muscles
that act on
> the joint, but it seems that these muscles are not considered to be
applied
> directly (am I right?), so I get a zero moment vector.
>
> I also tried looking at the Constraints.Reaction, but those are
just, as the
> name says, the constrained reaction, and therefore do not provide
the
> torques in the non-constrained degrees of freedom.
>
> How can I get the joint torque in the non-constrained degrees of
freedom of
> a joint?
>
> Thanks you,
>
> Jean-Olivier
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com] On
> Behalf
> Of AnyBody Support
> Sent: 21 avril 2007 15:12
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [anyscript] Re: Contraints.Reaction.Fout
>
> Hello Jean-Olivier,
>
> Contraints.Reaction.Fout is a menber of any constraint. It is the
> force(s) in the constraint. If the constraint is a joint then
> Contraints.Reaction.Fout contains the joint reactions just as you
> anticipate.
>
> But here is the point of the story: Joint reactions are very
> different depending on whether you have muscles in the model or
not.
> Joint reaction forces in models relying only on joint torques, i.e.
> classical inverse dynamics, are hugely under estimated. If you only
> have joint torque providers in the model, then the joint reactions
> are only what the external forces produce. When you include the
> muscles, their contribution to the joint reaction is added and it
is
> often larger than what the xternal forces provide. A muscle can
only
> produce torque in a force couple with a joint reaction, so the real
> joint reactions are much larger than what you get from a model with
> pure joint torques.
>
> In other words, it is correct that you get different joint
reactions
> depending on whether you include muscles in the model or not.
>
> Best regards,
> John
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> Jean-Olivier Racine <jean-
> olivier.racine.1@> wrote:
> >
> > Hello all,
> >
> >
> >
> > I am wondering what are the Contraints.Reaction.Fout values. As I
> run my
> > model with and without muscles, I get different values for those.
> Therefore,
> > I conclude that these are not the Reaction Forces, as defined
when
> doing
> > classical inverse dynamics (reaction forces and moments passed
from
> one
> > joint to the other), since they should not change with or without
> muscles.
> > So my questions:
> >
> >
> >
> > 1. What are these Constraint.Reaction.Fout value?
> > 2. How can I get those reaction forces that come directly from
> the
> > classic inverse dynamics analysis?
> >
> >
> >
> > Thanks!
> >
> > Jean-Olivier
> >
> >
> >
> > [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]
>

Thanks Søren,

I tried to make the drivers supply the movement generating moment. To do so,
I commented out the diagnostic muscle part (#include “JointMuscles.any”,
RightLegOnly/ LegNoMuscles.root.any:31) and set reaction.type={on} (#include
Drivers_NoMuscles.any, STS.main.any:70) for my knee and ankle flexion
drivers. It gave me the following warning:

  1. Inverse dynamic analysis…

WARNING(OBJ.MCH.DYN2) : C:\Documents and Settings\Jean-Olivier
Racine\Desktop\LIO\Anybody\Repository.6\ARep\LIO\STS Test
3\InverseDynamicsStudy.any(1) : ‘InverseDynamicsStudy’ : Model contains
redundant constraint reactions :

List of redundant reactions:

0: Main.Model.HumanModel.Right.Leg.Jnt.Knee.Constraints.Reaction[0]

1: Main.Model.HumanModel.Right.Leg.Jnt.Knee.Constraints.Reaction[1]

Everything ran well besides that warning. But when I looked at the moments,
the knee flexion moment looks more like noise than any realistic value. Not
to mention that it has a very low amplitude (moments of 10Nm?). So I figure
that the redundant knee reaction is competing for the moments, therefore
falsifying my values. Where is that other reaction come from? Too bad it’s
not specified in the warning message… I looked everywhere it seemed logical
to have it but couldn’t find it…

Since my solution didn’t work, I figured I’d try the other one, with the
diagnostic muscles. So I’ve set my reaction.type to off (#include
Drivers_Muscles.any, STS.main.any:69) again and re-added the joint muscles
(#include “JointMuscles.any”, RightLegOnly/ LegNoMuscles.root.any:31). But
both the negative and positive muscles give me zero strength! I only get
fibber length values (lm, lmt, lmdot, lmtdot)…

Clearly, I must be doing something wrong, but what?

Thanks!

Jean-Olivier

P.-S. Model uploaded to the group in case you need it (stsjoints.rar)


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of AnyBody Support
Sent: 26 avril 2007 03:17
To: anyscript@yahoogroups.com
Subject: [anyscript] Re: Classical inverse dynamic analysis

Hi Jean-Oliver

Yes you are right, if you have the reaction.type in the drivers set
to “On” the driver will carry the load and you can read out its value.

Alternatively if you are not driving all the human joints directly
and want to have them carried you can also apply a AnyReacForce to
the joint, in this way the free dof in joint will also be carried
but no motion will be supplied.

If you are using a repository model and you are using a model with
nomuscles on it will also be possible to read out the joint moments.
By default these versions supply joint torque muscles to the joints
and by reading out the forces in these joint torque muscles you have
the joint moments. As i recall it these can be found named
JointMuscles. Please notice that there is a muscle for each dirction
of the moement.

You are right about the ForceMomentMeasure it will currently not
measure the moment generated by the muscles, this will however be
possible in future versions.

Best regards
Søren, AnyBody Support

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
Jean-Olivier Racine <jean-
olivier.racine.1@…> wrote:
>
> I just realized that there’s a Reaction.Type not only in joints,
but also in
> drivers. So I suppose that if I set this to ON, I should get the
required
> information. I’ll try that and see what happens. Sorry about that.
>
>
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com]
On Behalf
> Of Jean-Olivier Racine
> Sent: 25 avril 2007 10:34
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [anyscript] Classical inverse dynamic analysis
>
>
>
> Hi again,
>
> Is it possible to perform a classical inverse dynamic analysis on
the model?
> I would like to calculate joint torques and reaction forces at the
joints
> required to perform a certain motion. In the tutorial this step is
skipped,
> saying: “Since we have no muscles so far, kinematic analysis is
really all
> that makes sense”. But in my case there is some relevant
information I’d
> like to extract from this.
>
> I tried using an AnyForceMomentMeasure and including all muscles
that act on
> the joint, but it seems that these muscles are not considered to be
applied
> directly (am I right?), so I get a zero moment vector.
>
> I also tried looking at the Constraints.Reaction, but those are
just, as the
> name says, the constrained reaction, and therefore do not provide
the
> torques in the non-constrained degrees of freedom.
>
> How can I get the joint torque in the non-constrained degrees of
freedom of
> a joint?
>
> Thanks you,
>
> Jean-Olivier
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com] On
> Behalf
> Of AnyBody Support
> Sent: 21 avril 2007 15:12
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [anyscript] Re: Contraints.Reaction.Fout
>
> Hello Jean-Olivier,
>
> Contraints.Reaction.Fout is a menber of any constraint. It is the
> force(s) in the constraint. If the constraint is a joint then
> Contraints.Reaction.Fout contains the joint reactions just as you
> anticipate.
>
> But here is the point of the story: Joint reactions are very
> different depending on whether you have muscles in the model or
not.
> Joint reaction forces in models relying only on joint torques, i.e.
> classical inverse dynamics, are hugely under estimated. If you only
> have joint torque providers in the model, then the joint reactions
> are only what the external forces produce. When you include the
> muscles, their contribution to the joint reaction is added and it
is
> often larger than what the xternal forces provide. A muscle can
only
> produce torque in a force couple with a joint reaction, so the real
> joint reactions are much larger than what you get from a model with
> pure joint torques.
>
> In other words, it is correct that you get different joint
reactions
> depending on whether you include muscles in the model or not.
>
> Best regards,
> John
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> Jean-Olivier Racine <jean-
> olivier.racine.1@> wrote:
> >
> > Hello all,
> >
> >
> >
> > I am wondering what are the Contraints.Reaction.Fout values. As I
> run my
> > model with and without muscles, I get different values for those.
> Therefore,
> > I conclude that these are not the Reaction Forces, as defined
when
> doing
> > classical inverse dynamics (reaction forces and moments passed
from
> one
> > joint to the other), since they should not change with or without
> muscles.
> > So my questions:
> >
> >
> >
> > 1. What are these Constraint.Reaction.Fout value?
> > 2. How can I get those reaction forces that come directly from
> the
> > classic inverse dynamics analysis?
> >
> >
> >
> > Thanks!
> >
> > Jean-Olivier
> >
> >
> >
> > [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]
>

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

Hi Jean-Oliver

I have looked through your model and found a few things.

In the file Marker.any the reactions on the markers drivershad been
turned off, leaving them with nothing to carry them.

In the Forces.anyt file there where too many constraints on the force
plate, the moments had been turned off but the linear reaction
between the plate and the environment was still there. I think this
was the reason for the redundant constraints. Please look at
the “LinDriver” object.

With these changes the model runs with no warnings and it produces
moment by the joint muscles, which i have switched in again in the
LegNoMuscles.any file.

A new version JeanOlivier.zip will be uploaded shortly

Best regards
Søren, AnyBody Support

— In anyscript@yahoogroups.com, Jean-Olivier Racine <jean-
olivier.racine.1@…> wrote:
>
> Thanks Søren,
>
>
>
> I tried to make the drivers supply the movement generating moment.
To do so,
> I commented out the diagnostic muscle part
(#include “JointMuscles.any”,
> RightLegOnly/ LegNoMuscles.root.any:31) and set reaction.type={on}
(#include
> Drivers_NoMuscles.any, STS.main.any:70) for my knee and ankle
flexion
> drivers. It gave me the following warning:
>
>
>
> 4) Inverse dynamic analysis…
>
> WARNING(OBJ.MCH.DYN2) : C:\Documents and Settings\Jean-Olivier
> Racine\Desktop\LIO\Anybody\Repository.6\ARep\LIO\STS Test
> 3\InverseDynamicsStudy.any(1) : ‘InverseDynamicsStudy’ : Model
contains
> redundant constraint reactions :
>
> List of redundant reactions:
>
> 0: Main.Model.HumanModel.Right.Leg.Jnt.Knee.Constraints.Reaction
[0]
>
> 1: Main.Model.HumanModel.Right.Leg.Jnt.Knee.Constraints.Reaction
[1]
>
>
>
> Everything ran well besides that warning. But when I looked at the
moments,
> the knee flexion moment looks more like noise than any realistic
value. Not
> to mention that it has a very low amplitude (moments of 10Nm?). So
I figure
> that the redundant knee reaction is competing for the moments,
therefore
> falsifying my values. Where is that other reaction come from? Too
bad it’s
> not specified in the warning message… I looked everywhere it seemed
logical
> to have it but couldn’t find it…
>
>
>
>
>
> Since my solution didn’t work, I figured I’d try the other one,
with the
> diagnostic muscles. So I’ve set my reaction.type to off (#include
> Drivers_Muscles.any, STS.main.any:69) again and re-added the joint
muscles
> (#include “JointMuscles.any”, RightLegOnly/
LegNoMuscles.root.any:31). But
> both the negative and positive muscles give me zero strength! I
only get
> fibber length values (lm, lmt, lmdot, lmtdot)…
>
>
>
> Clearly, I must be doing something wrong, but what?
>
> Thanks!
>
> Jean-Olivier
>
>
>
> P.-S. Model uploaded to the group in case you need it
(stsjoints.rar)
>
>
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of AnyBody Support
> Sent: 26 avril 2007 03:17
> To: anyscript@yahoogroups.com
> Subject: [anyscript] Re: Classical inverse dynamic analysis
>
>
>
> Hi Jean-Oliver
>
> Yes you are right, if you have the reaction.type in the drivers set
> to “On” the driver will carry the load and you can read out its
value.
>
> Alternatively if you are not driving all the human joints directly
> and want to have them carried you can also apply a AnyReacForce to
> the joint, in this way the free dof in joint will also be carried
> but no motion will be supplied.
>
> If you are using a repository model and you are using a model with
> nomuscles on it will also be possible to read out the joint
moments.
> By default these versions supply joint torque muscles to the joints
> and by reading out the forces in these joint torque muscles you
have
> the joint moments. As i recall it these can be found named
> JointMuscles. Please notice that there is a muscle for each
dirction
> of the moement.
>
> You are right about the ForceMomentMeasure it will currently not
> measure the moment generated by the muscles, this will however be
> possible in future versions.
>
> Best regards
> Søren, AnyBody Support
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> Jean-Olivier Racine <jean-
> olivier.racine.1@> wrote:
> >
> > I just realized that there’s a Reaction.Type not only in joints,
> but also in
> > drivers. So I suppose that if I set this to ON, I should get the
> required
> > information. I’ll try that and see what happens. Sorry about that.
> >
> >
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com]
> On Behalf
> > Of Jean-Olivier Racine
> > Sent: 25 avril 2007 10:34
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > Subject: [anyscript] Classical inverse dynamic analysis
> >
> >
> >
> > Hi again,
> >
> > Is it possible to perform a classical inverse dynamic analysis on
> the model?
> > I would like to calculate joint torques and reaction forces at
the
> joints
> > required to perform a certain motion. In the tutorial this step
is
> skipped,
> > saying: “Since we have no muscles so far, kinematic analysis is
> really all
> > that makes sense”. But in my case there is some relevant
> information I’d
> > like to extract from this.
> >
> > I tried using an AnyForceMomentMeasure and including all muscles
> that act on
> > the joint, but it seems that these muscles are not considered to
be
> applied
> > directly (am I right?), so I get a zero moment vector.
> >
> > I also tried looking at the Constraints.Reaction, but those are
> just, as the
> > name says, the constrained reaction, and therefore do not provide
> the
> > torques in the non-constrained degrees of freedom.
> >
> > How can I get the joint torque in the non-constrained degrees of
> freedom of
> > a joint?
> >
> > Thanks you,
> >
> > Jean-Olivier
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com] On
> > Behalf
> > Of AnyBody Support
> > Sent: 21 avril 2007 15:12
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > Subject: [anyscript] Re: Contraints.Reaction.Fout
> >
> > Hello Jean-Olivier,
> >
> > Contraints.Reaction.Fout is a menber of any constraint. It is the
> > force(s) in the constraint. If the constraint is a joint then
> > Contraints.Reaction.Fout contains the joint reactions just as you
> > anticipate.
> >
> > But here is the point of the story: Joint reactions are very
> > different depending on whether you have muscles in the model or
> not.
> > Joint reaction forces in models relying only on joint torques,
i.e.
> > classical inverse dynamics, are hugely under estimated. If you
only
> > have joint torque providers in the model, then the joint
reactions
> > are only what the external forces produce. When you include the
> > muscles, their contribution to the joint reaction is added and it
> is
> > often larger than what the xternal forces provide. A muscle can
> only
> > produce torque in a force couple with a joint reaction, so the
real
> > joint reactions are much larger than what you get from a model
with
> > pure joint torques.
> >
> > In other words, it is correct that you get different joint
> reactions
> > depending on whether you include muscles in the model or not.
> >
> > Best regards,
> > John
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > Jean-Olivier Racine <jean-
> > olivier.racine.1@> wrote:
> > >
> > > Hello all,
> > >
> > >
> > >
> > > I am wondering what are the Contraints.Reaction.Fout values. As
I
> > run my
> > > model with and without muscles, I get different values for
those.
> > Therefore,
> > > I conclude that these are not the Reaction Forces, as defined
> when
> > doing
> > > classical inverse dynamics (reaction forces and moments passed
> from
> > one
> > > joint to the other), since they should not change with or
without
> > muscles.
> > > So my questions:
> > >
> > >
> > >
> > > 1. What are these Constraint.Reaction.Fout value?
> > > 2. How can I get those reaction forces that come directly from
> > the
> > > classic inverse dynamics analysis?
> > >
> > >
> > >
> > > Thanks!
> > >
> > > Jean-Olivier
> > >
> > >
> > >
> > > [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]
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

Good day,

Sorry for the delay before I answered.

Thanks for taking a look at my model. Took me a while to understand why
markers needed to be carried. But now it all makes sense.

Thanks for all your help!

Jean-Olivier


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of AnyBody Support
Sent: 3 mai 2007 04:01
To: anyscript@yahoogroups.com
Subject: [anyscript] Re: Classical inverse dynamic analysis

Hi Jean-Oliver

I have looked through your model and found a few things.

In the file Marker.any the reactions on the markers drivershad been
turned off, leaving them with nothing to carry them.

In the Forces.anyt file there where too many constraints on the force
plate, the moments had been turned off but the linear reaction
between the plate and the environment was still there. I think this
was the reason for the redundant constraints. Please look at
the “LinDriver” object.

With these changes the model runs with no warnings and it produces
moment by the joint muscles, which i have switched in again in the
LegNoMuscles.any file.

A new version JeanOlivier.zip will be uploaded shortly

Best regards
Søren, AnyBody Support

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
Jean-Olivier Racine <jean-
olivier.racine.1@…> wrote:
>
> Thanks Søren,
>
>
>
> I tried to make the drivers supply the movement generating moment.
To do so,
> I commented out the diagnostic muscle part
(#include “JointMuscles.any”,
> RightLegOnly/ LegNoMuscles.root.any:31) and set reaction.type={on}
(#include
> Drivers_NoMuscles.any, STS.main.any:70) for my knee and ankle
flexion
> drivers. It gave me the following warning:
>
>
>
> 4) Inverse dynamic analysis…
>
> WARNING(OBJ.MCH.DYN2) : C:\Documents and Settings\Jean-Olivier
> Racine\Desktop\LIO\Anybody\Repository.6\ARep\LIO\STS Test
> 3\InverseDynamicsStudy.any(1) : ‘InverseDynamicsStudy’ : Model
contains
> redundant constraint reactions :
>
> List of redundant reactions:
>
> 0: Main.Model.HumanModel.Right.Leg.Jnt.Knee.Constraints.Reaction
[0]
>
> 1: Main.Model.HumanModel.Right.Leg.Jnt.Knee.Constraints.Reaction
[1]
>
>
>
> Everything ran well besides that warning. But when I looked at the
moments,
> the knee flexion moment looks more like noise than any realistic
value. Not
> to mention that it has a very low amplitude (moments of 10Nm?). So
I figure
> that the redundant knee reaction is competing for the moments,
therefore
> falsifying my values. Where is that other reaction come from? Too
bad it’s
> not specified in the warning message… I looked everywhere it seemed
logical
> to have it but couldn’t find it…
>
>
>
>
>
> Since my solution didn’t work, I figured I’d try the other one,
with the
> diagnostic muscles. So I’ve set my reaction.type to off (#include
> Drivers_Muscles.any, STS.main.any:69) again and re-added the joint
muscles
> (#include “JointMuscles.any”, RightLegOnly/
LegNoMuscles.root.any:31). But
> both the negative and positive muscles give me zero strength! I
only get
> fibber length values (lm, lmt, lmdot, lmtdot)…
>
>
>
> Clearly, I must be doing something wrong, but what?
>
> Thanks!
>
> Jean-Olivier
>
>
>
> P.-S. Model uploaded to the group in case you need it
(stsjoints.rar)
>
>
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com]
On Behalf
> Of AnyBody Support
> Sent: 26 avril 2007 03:17
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [anyscript] Re: Classical inverse dynamic analysis
>
>
>
> Hi Jean-Oliver
>
> Yes you are right, if you have the reaction.type in the drivers set
> to “On” the driver will carry the load and you can read out its
value.
>
> Alternatively if you are not driving all the human joints directly
> and want to have them carried you can also apply a AnyReacForce to
> the joint, in this way the free dof in joint will also be carried
> but no motion will be supplied.
>
> If you are using a repository model and you are using a model with
> nomuscles on it will also be possible to read out the joint
moments.
> By default these versions supply joint torque muscles to the joints
> and by reading out the forces in these joint torque muscles you
have
> the joint moments. As i recall it these can be found named
> JointMuscles. Please notice that there is a muscle for each
dirction
> of the moement.
>
> You are right about the ForceMomentMeasure it will currently not
> measure the moment generated by the muscles, this will however be
> possible in future versions.
>
> Best regards
> Søren, AnyBody Support
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> Jean-Olivier Racine <jean-
> olivier.racine.1@> wrote:
> >
> > I just realized that there’s a Reaction.Type not only in joints,
> but also in
> > drivers. So I suppose that if I set this to ON, I should get the
> required
> > information. I’ll try that and see what happens. Sorry about that.
> >
> >
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com]
> On Behalf
> > Of Jean-Olivier Racine
> > Sent: 25 avril 2007 10:34
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > Subject: [anyscript] Classical inverse dynamic analysis
> >
> >
> >
> > Hi again,
> >
> > Is it possible to perform a classical inverse dynamic analysis on
> the model?
> > I would like to calculate joint torques and reaction forces at
the
> joints
> > required to perform a certain motion. In the tutorial this step
is
> skipped,
> > saying: “Since we have no muscles so far, kinematic analysis is
> really all
> > that makes sense”. But in my case there is some relevant
> information I’d
> > like to extract from this.
> >
> > I tried using an AnyForceMomentMeasure and including all muscles
> that act on
> > the joint, but it seems that these muscles are not considered to
be
> applied
> > directly (am I right?), so I get a zero moment vector.
> >
> > I also tried looking at the Constraints.Reaction, but those are
> just, as the
> > name says, the constrained reaction, and therefore do not provide
> the
> > torques in the non-constrained degrees of freedom.
> >
> > How can I get the joint torque in the non-constrained degrees of
> freedom of
> > a joint?
> >
> > Thanks you,
> >
> > Jean-Olivier
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com] On
> > Behalf
> > Of AnyBody Support
> > Sent: 21 avril 2007 15:12
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > Subject: [anyscript] Re: Contraints.Reaction.Fout
> >
> > Hello Jean-Olivier,
> >
> > Contraints.Reaction.Fout is a menber of any constraint. It is the
> > force(s) in the constraint. If the constraint is a joint then
> > Contraints.Reaction.Fout contains the joint reactions just as you
> > anticipate.
> >
> > But here is the point of the story: Joint reactions are very
> > different depending on whether you have muscles in the model or
> not.
> > Joint reaction forces in models relying only on joint torques,
i.e.
> > classical inverse dynamics, are hugely under estimated. If you
only
> > have joint torque providers in the model, then the joint
reactions
> > are only what the external forces produce. When you include the
> > muscles, their contribution to the joint reaction is added and it
> is
> > often larger than what the xternal forces provide. A muscle can
> only
> > produce torque in a force couple with a joint reaction, so the
real
> > joint reactions are much larger than what you get from a model
with
> > pure joint torques.
> >
> > In other words, it is correct that you get different joint
> reactions
> > depending on whether you include muscles in the model or not.
> >
> > Best regards,
> > John
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > Jean-Olivier Racine <jean-
> > olivier.racine.1@> wrote:
> > >
> > > Hello all,
> > >
> > >
> > >
> > > I am wondering what are the Contraints.Reaction.Fout values. As
I
> > run my
> > > model with and without muscles, I get different values for
those.
> > Therefore,
> > > I conclude that these are not the Reaction Forces, as defined
> when
> > doing
> > > classical inverse dynamics (reaction forces and moments passed
> from
> > one
> > > joint to the other), since they should not change with or
without
> > muscles.
> > > So my questions:
> > >
> > >
> > >
> > > 1. What are these Constraint.Reaction.Fout value?
> > > 2. How can I get those reaction forces that come directly from
> > the
> > > classic inverse dynamics analysis?
> > >
> > >
> > >
> > > Thanks!
> > >
> > > Jean-Olivier
> > >
> > >
> > >
> > > [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]
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

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