Problems with calculation of hip forces with my ouwn hip model

Hi All,

I have a problem with the computation of the forces in the hip joint.
My hipmodel is assembled from the Repository pedal model. The pelvis
is fixed at the global reference frame. In my model there is the thigh
the only moveable body, the other bodys are removed. All muscles of
the thigh are contained, because the nodes, which connect the muscles
to the shank, are transfered to the segment of thigh. The ground
reaction force was set as outside force at the knee.
The calculated hip forces are twice as high as in the Gait3D-model.
Does someone have an idea, which the cause for it is?

The files are uploaded, you find this under “Pelvis-Thigh-G3”.zip
Thanks for your support

Best regards
Christian

Hi Christian,

There seems to be several causes for the high forces you experimented.

One is the muscle model. Because you modified the points of application of
the muscles, the 3 elements model may not be indicated anymore, even with
the calibration. The explanation is that the kinematic of some muscles is
changed as they have been moved from the shank to the thigh. In consequence
some of those muscles will stretch more than the normal and end up with zero
strength (this is what happens to the gluteus for example). Then the
remaining muscle try to compensate and this lead to an incorrect recruitment
with a lot of over activity in the muscle.

You should therefore use the simple muscle model for this particular
configuration.

Then I noticed that the forces you apply at the knee are quite high, more
than I would expect. I understand your model represent a gait situation. GRF
in gait are usually not much higher than the body weight but this force goes
up to 2000 N. Also applying the reaction force to the knee can be a little
tricky because it is not exactly the same reaction force as in the feet
which can be measured with force plates. May I ask how did you get the data
of that force applied at the knee joint?

Best regards,

Sylvain, AnyBody Support


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of kliewec
Sent: 23. januar 2009 10:12
To: anyscript@yahoogroups.com
Subject: [AnyScript] Problems with calculation of hip forces with my ouwn
hip model

Hi All,

I have a problem with the computation of the forces in the hip joint.
My hipmodel is assembled from the Repository pedal model. The pelvis
is fixed at the global reference frame. In my model there is the thigh
the only moveable body, the other bodys are removed. All muscles of
the thigh are contained, because the nodes, which connect the muscles
to the shank, are transfered to the segment of thigh. The ground
reaction force was set as outside force at the knee.
The calculated hip forces are twice as high as in the Gait3D-model.
Does someone have an idea, which the cause for it is?

The files are uploaded, you find this under “Pelvis-Thigh-G3”.zip
Thanks for your support

Best regards
Christian

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

Hi Silvian,

thanks for your support. I changed the force input file, now I think
it is correct. In addition I changed the muscle model to MUS1. But the
results are not really better- hipforce up to 5000N.Thus I set also
the kneeforce to zero and the mass of the thigh segment to Zero too.
So it should be calculated a Hipforce with Zero. Bur I get nearly 3000N.
That cannot be correct! I Have no idea, what the mistake is. I load up
the Folder with my model again (filename "Pelvis_Thigh_G3_09_02_05)

Best regards

Christian

— In anyscript@yahoogroups.com, AnyScript Support <anyscriptsup@…>
wrote:
>
> Hi Christian,
>
>
>
> There seems to be several causes for the high forces you experimented.
>
>
>
> One is the muscle model. Because you modified the points of
application of
> the muscles, the 3 elements model may not be indicated anymore, even
with
> the calibration. The explanation is that the kinematic of some
muscles is
> changed as they have been moved from the shank to the thigh. In
consequence
> some of those muscles will stretch more than the normal and end up
with zero
> strength (this is what happens to the gluteus for example). Then the
> remaining muscle try to compensate and this lead to an incorrect
recruitment
> with a lot of over activity in the muscle.
>
> You should therefore use the simple muscle model for this particular
> configuration.
>
>
>
> Then I noticed that the forces you apply at the knee are quite high,
more
> than I would expect. I understand your model represent a gait
situation. GRF
> in gait are usually not much higher than the body weight but this
force goes
> up to 2000 N. Also applying the reaction force to the knee can be a
little
> tricky because it is not exactly the same reaction force as in the feet
> which can be measured with force plates. May I ask how did you get
the data
> of that force applied at the knee joint?
>
>
>
> Best regards,
>
> Sylvain, AnyBody Support
>
>
>
>
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of kliewec
> Sent: 23. januar 2009 10:12
> To: anyscript@yahoogroups.com
> Subject: [AnyScript] Problems with calculation of hip forces with my
ouwn
> hip model
>
>
>
> Hi All,
>
> I have a problem with the computation of the forces in the hip joint.
> My hipmodel is assembled from the Repository pedal model. The pelvis
> is fixed at the global reference frame. In my model there is the thigh
> the only moveable body, the other bodys are removed. All muscles of
> the thigh are contained, because the nodes, which connect the muscles
> to the shank, are transfered to the segment of thigh. The ground
> reaction force was set as outside force at the knee.
> The calculated hip forces are twice as high as in the Gait3D-model.
> Does someone have an idea, which the cause for it is?
>
> The files are uploaded, you find this under “Pelvis-Thigh-G3”.zip
> Thanks for your support
>
> Best regards
> Christian
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

Hi Christian,

I could not reproduced the 3000 N when setting the knee force to zero
and the femur mass to zero. I get zero reaction force on the hip
joint as expected. So for me the model behaves correctly on this
point.

What I am concerned about is once again the force applied to the
knee. As there is no shank segment you cannot just apply some
measured knee reaction force. I think this was what you did on the
last posting? I see you modified it but I cannot tell if it is the
correct one without more information. I need to know what is this
force you apply to help you (where did you get it?).
Actually this force should reflect the effect of the GRF at the knee
point. I am thinking of using AnyForceMomentMeasure2 to read an
equivalent force and moment of the GRF at the knee joint. It is just
a rough idea but maybe it could work.

Also did you consider using a model of the entire leg? That would
make things easier. I am also a bit worried about moving the
insertion point of the muscles on the thigh segment, that can have
some side effect I am not quite sure about yet.

Best regards,
Sylvain, AnyBody Support

— In anyscript@yahoogroups.com, “kliewec” <christian.kliewe@…>
wrote:
>
> Hi Silvian,
>
> thanks for your support. I changed the force input file, now I think
> it is correct. In addition I changed the muscle model to MUS1. But
the
> results are not really better- hipforce up to 5000N.Thus I set also
> the kneeforce to zero and the mass of the thigh segment to Zero too.
> So it should be calculated a Hipforce with Zero. Bur I get nearly
3000N.
> That cannot be correct! I Have no idea, what the mistake is. I load
up
> the Folder with my model again (filename "Pelvis_Thigh_G3_09_02_05)
>
> Best regards
>
> Christian
>
>
>
> — In anyscript@yahoogroups.com, AnyScript Support <anyscriptsup@>
> wrote:
> >
> > Hi Christian,
> >
> >
> >
> > There seems to be several causes for the high forces you
experimented.
> >
> >
> >
> > One is the muscle model. Because you modified the points of
> application of
> > the muscles, the 3 elements model may not be indicated anymore,
even
> with
> > the calibration. The explanation is that the kinematic of some
> muscles is
> > changed as they have been moved from the shank to the thigh. In
> consequence
> > some of those muscles will stretch more than the normal and end up
> with zero
> > strength (this is what happens to the gluteus for example). Then
the
> > remaining muscle try to compensate and this lead to an incorrect
> recruitment
> > with a lot of over activity in the muscle.
> >
> > You should therefore use the simple muscle model for this
particular
> > configuration.
> >
> >
> >
> > Then I noticed that the forces you apply at the knee are quite
high,
> more
> > than I would expect. I understand your model represent a gait
> situation. GRF
> > in gait are usually not much higher than the body weight but this
> force goes
> > up to 2000 N. Also applying the reaction force to the knee can be
a
> little
> > tricky because it is not exactly the same reaction force as in
the feet
> > which can be measured with force plates. May I ask how did you get
> the data
> > of that force applied at the knee joint?
> >
> >
> >
> > Best regards,
> >
> > Sylvain, AnyBody Support
> >
> >
> >
> >
> >
> > _____
> >
> > From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
> On Behalf
> > Of kliewec
> > Sent: 23. januar 2009 10:12
> > To: anyscript@yahoogroups.com
> > Subject: [AnyScript] Problems with calculation of hip forces with
my
> ouwn
> > hip model
> >
> >
> >
> > Hi All,
> >
> > I have a problem with the computation of the forces in the hip
joint.
> > My hipmodel is assembled from the Repository pedal model. The
pelvis
> > is fixed at the global reference frame. In my model there is the
thigh
> > the only moveable body, the other bodys are removed. All muscles
of
> > the thigh are contained, because the nodes, which connect the
muscles
> > to the shank, are transfered to the segment of thigh. The ground
> > reaction force was set as outside force at the knee.
> > The calculated hip forces are twice as high as in the Gait3D-
model.
> > Does someone have an idea, which the cause for it is?
> >
> > The files are uploaded, you find this under “Pelvis-Thigh-G3”.zip
> > Thanks for your support
> >
> > Best regards
> > Christian
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>

Hi Silvian,

sorry about may late answer.

The problem possibly lies in my model, because the kneeforce produces a
moment , against the muscles must work. May be, the orientation and the
lever arm of the kneeforce leads to other results.

The force, wich I used at the knee is the knee reaction force, measured at
the Gait-3D-Modell without muscles
(to find in
Study/Output/Model/HumanModel/Right/Leg/JntDOF/Hip/Constrains/Reaction/Fout).
Therefore this force is a little bit smaler then the ForcePlate1 (reduced by
the segment mass of shank and foot).This force should reflect the GRF in the
knee, as you already determined.

Your proposal to use an AnyForceMomentMeasure2 to get the force and the
moment at the knee should be the same??? Otherwise it could be possible to
read out the torque in the hip, what the muscle have to realize for a
special movement.

How the function “AnyForceMomentMeasure2” is to define and where?

The reason, why we want to make the model as simple as possible, is the
computing time. We want to combine the simulation model with an robot, so we
have restrictions with data exchange. We need for our investigations only
the resultant hip force.

Best regards

Christian

----- Original Message -----
From: “AnyBody Support” <anyscriptsup@anybodytech.com>
To: <anyscript@yahoogroups.com>
Sent: Monday, February 09, 2009 9:53 AM
Subject: [AnyScript] Re: Problems with calculation of hip forces with my
ouwn hip model

> Hi Christian,
>
> I could not reproduced the 3000 N when setting the knee force to zero
> and the femur mass to zero. I get zero reaction force on the hip
> joint as expected. So for me the model behaves correctly on this
> point.
>
> What I am concerned about is once again the force applied to the
> knee. As there is no shank segment you cannot just apply some
> measured knee reaction force. I think this was what you did on the
> last posting? I see you modified it but I cannot tell if it is the
> correct one without more information. I need to know what is this
> force you apply to help you (where did you get it?).
> Actually this force should reflect the effect of the GRF at the knee
> point. I am thinking of using AnyForceMomentMeasure2 to read an
> equivalent force and moment of the GRF at the knee joint. It is just
> a rough idea but maybe it could work.
>
> Also did you consider using a model of the entire leg? That would
> make things easier. I am also a bit worried about moving the
> insertion point of the muscles on the thigh segment, that can have
> some side effect I am not quite sure about yet.
>
> Best regards,
> Sylvain, AnyBody Support
>
>
>
> — In anyscript@yahoogroups.com, “kliewec” <christian.kliewe@…>
> wrote:
>>
>> Hi Silvian,
>>
>> thanks for your support. I changed the force input file, now I think
>> it is correct. In addition I changed the muscle model to MUS1. But
> the
>> results are not really better- hipforce up to 5000N.Thus I set also
>> the kneeforce to zero and the mass of the thigh segment to Zero too.
>> So it should be calculated a Hipforce with Zero. Bur I get nearly
> 3000N.
>> That cannot be correct! I Have no idea, what the mistake is. I load
> up
>> the Folder with my model again (filename "Pelvis_Thigh_G3_09_02_05)
>>
>> Best regards
>>
>> Christian
>>
>>
>>
>> — In anyscript@yahoogroups.com, AnyScript Support <anyscriptsup@>
>> wrote:
>> >
>> > Hi Christian,
>> >
>> >
>> >
>> > There seems to be several causes for the high forces you
> experimented.
>> >
>> >
>> >
>> > One is the muscle model. Because you modified the points of
>> application of
>> > the muscles, the 3 elements model may not be indicated anymore,
> even
>> with
>> > the calibration. The explanation is that the kinematic of some
>> muscles is
>> > changed as they have been moved from the shank to the thigh. In
>> consequence
>> > some of those muscles will stretch more than the normal and end up
>> with zero
>> > strength (this is what happens to the gluteus for example). Then
> the
>> > remaining muscle try to compensate and this lead to an incorrect
>> recruitment
>> > with a lot of over activity in the muscle.
>> >
>> > You should therefore use the simple muscle model for this
> particular
>> > configuration.
>> >
>> >
>> >
>> > Then I noticed that the forces you apply at the knee are quite
> high,
>> more
>> > than I would expect. I understand your model represent a gait
>> situation. GRF
>> > in gait are usually not much higher than the body weight but this
>> force goes
>> > up to 2000 N. Also applying the reaction force to the knee can be
> a
>> little
>> > tricky because it is not exactly the same reaction force as in
> the feet
>> > which can be measured with force plates. May I ask how did you get
>> the data
>> > of that force applied at the knee joint?
>> >
>> >
>> >
>> > Best regards,
>> >
>> > Sylvain, AnyBody Support
>> >
>> >
>> >
>> >
>> >
>> > _____
>> >
>> > From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
>> On Behalf
>> > Of kliewec
>> > Sent: 23. januar 2009 10:12
>> > To: anyscript@yahoogroups.com
>> > Subject: [AnyScript] Problems with calculation of hip forces with
> my
>> ouwn
>> > hip model
>> >
>> >
>> >
>> > Hi All,
>> >
>> > I have a problem with the computation of the forces in the hip
> joint.
>> > My hipmodel is assembled from the Repository pedal model. The
> pelvis
>> > is fixed at the global reference frame. In my model there is the
> thigh
>> > the only moveable body, the other bodys are removed. All muscles
> of
>> > the thigh are contained, because the nodes, which connect the
> muscles
>> > to the shank, are transfered to the segment of thigh. The ground
>> > reaction force was set as outside force at the knee.
>> > The calculated hip forces are twice as high as in the Gait3D-
> model.
>> > Does someone have an idea, which the cause for it is?
>> >
>> > The files are uploaded, you find this under “Pelvis-Thigh-G3”.zip
>> > Thanks for your support
>> >
>> > Best regards
>> > Christian
>> >
>> >
>> >
>> >
>> >
>> > [Non-text portions of this message have been removed]
>> >
>>
>
>
>
>
> ------------------------------------
>
> Users of the AnyBody Modeling System help each other create biomechanical
> models in the AnyScript language.Yahoo! Groups Links
>
>
>

Hi Christian,

I understand your need for a simple model. However cutting the leg is
creating complications as you see it. If the whole leg model is still
acceptable for you in term of computation time, then I keep thinking it will
be better.

But let see what we can do about it.

First point is that you can’t use the knee reaction force. Because this
reaction force is mainly the result of the muscles spanning and pulling
around the joint, and you have no more joint here and neither muscles
around. The knee reaction force doesn’t reflect the GRF, it is way too high.
The solution I see is to use AnyForceMomentMeasure2 in the full gait model
to read an equivalent of the GRF at the knee joint. This object calculate
the equivalent moment+force of any kind of forces at a certain point of
interest. Please see in the reference manual for the full explanation. So
you can read the equivalent of the GRF at the level of the knee (it is
different than the knee reaction force).

You can define it wherever in the code, it just has to pint at the good
objects. It will look like this:

AnyForceMomentMeasure2 <ObjectName> =

{

AnyForceBase &<Insert name0> = < GRF, the force you want the equivalent of
>;

AnyRefFrame &<Insert name0> = < Knee joint, the point of interest >;

AnySeg &<Insert name0> = < Femur, the mechanical system >;

};

The force and moment (in global coo) returned by this object can then serve
as input for the simple pelvis-femur model and applied to the knee. You can
make sure that the results are correct by comparing the hip reaction force
of the two models for the same motion. Notice that this method will only be
valid if the motion of the simple model is the same as the gait model. If
you want to model another load case then this particular GRF is not valid
anymore and you can’t use it to load the simple model.

Best regards,

Sylvain, AnyBody Support


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of Ch. Kliewe
Sent: 17. februar 2009 15:37
To: anyscript@yahoogroups.com
Subject: Re: [AnyScript] Re: Problems with calculation of hip forces with my
ouwn hip model

Hi Silvian,

sorry about may late answer.

The problem possibly lies in my model, because the kneeforce produces a
moment , against the muscles must work. May be, the orientation and the
lever arm of the kneeforce leads to other results.

The force, wich I used at the knee is the knee reaction force, measured at
the Gait-3D-Modell without muscles
(to find in
Study/Output/Model/HumanModel/Right/Leg/JntDOF/Hip/Constrains/Reaction/Fout)
.
Therefore this force is a little bit smaler then the ForcePlate1 (reduced by

the segment mass of shank and foot).This force should reflect the GRF in the

knee, as you already determined.

Your proposal to use an AnyForceMomentMeasure2 to get the force and the
moment at the knee should be the same??? Otherwise it could be possible to
read out the torque in the hip, what the muscle have to realize for a
special movement.

How the function “AnyForceMomentMeasure2” is to define and where?

The reason, why we want to make the model as simple as possible, is the
computing time. We want to combine the simulation model with an robot, so we

have restrictions with data exchange. We need for our investigations only
the resultant hip force.

Best regards

Christian

----- Original Message -----
From: “AnyBody Support” <anyscriptsup@
<mailto:anyscriptsup%40anybodytech.com> anybodytech.com>
To: <anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com>
Sent: Monday, February 09, 2009 9:53 AM
Subject: [AnyScript] Re: Problems with calculation of hip forces with my
ouwn hip model

> Hi Christian,
>
> I could not reproduced the 3000 N when setting the knee force to zero
> and the femur mass to zero. I get zero reaction force on the hip
> joint as expected. So for me the model behaves correctly on this
> point.
>
> What I am concerned about is once again the force applied to the
> knee. As there is no shank segment you cannot just apply some
> measured knee reaction force. I think this was what you did on the
> last posting? I see you modified it but I cannot tell if it is the
> correct one without more information. I need to know what is this
> force you apply to help you (where did you get it?).
> Actually this force should reflect the effect of the GRF at the knee
> point. I am thinking of using AnyForceMomentMeasure2 to read an
> equivalent force and moment of the GRF at the knee joint. It is just
> a rough idea but maybe it could work.
>
> Also did you consider using a model of the entire leg? That would
> make things easier. I am also a bit worried about moving the
> insertion point of the muscles on the thigh segment, that can have
> some side effect I am not quite sure about yet.
>
> Best regards,
> Sylvain, AnyBody Support
>
>
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“kliewec” <christian.kliewe@…>
> wrote:
>>
>> Hi Silvian,
>>
>> thanks for your support. I changed the force input file, now I think
>> it is correct. In addition I changed the muscle model to MUS1. But
> the
>> results are not really better- hipforce up to 5000N.Thus I set also
>> the kneeforce to zero and the mass of the thigh segment to Zero too.
>> So it should be calculated a Hipforce with Zero. Bur I get nearly
> 3000N.
>> That cannot be correct! I Have no idea, what the mistake is. I load
> up
>> the Folder with my model again (filename "Pelvis_Thigh_G3_09_02_05)
>>
>> Best regards
>>
>> Christian
>>
>>
>>
>> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
AnyScript Support <anyscriptsup@>
>> wrote:
>> >
>> > Hi Christian,
>> >
>> >
>> >
>> > There seems to be several causes for the high forces you
> experimented.
>> >
>> >
>> >
>> > One is the muscle model. Because you modified the points of
>> application of
>> > the muscles, the 3 elements model may not be indicated anymore,
> even
>> with
>> > the calibration. The explanation is that the kinematic of some
>> muscles is
>> > changed as they have been moved from the shank to the thigh. In
>> consequence
>> > some of those muscles will stretch more than the normal and end up
>> with zero
>> > strength (this is what happens to the gluteus for example). Then
> the
>> > remaining muscle try to compensate and this lead to an incorrect
>> recruitment
>> > with a lot of over activity in the muscle.
>> >
>> > You should therefore use the simple muscle model for this
> particular
>> > configuration.
>> >
>> >
>> >
>> > Then I noticed that the forces you apply at the knee are quite
> high,
>> more
>> > than I would expect. I understand your model represent a gait
>> situation. GRF
>> > in gait are usually not much higher than the body weight but this
>> force goes
>> > up to 2000 N. Also applying the reaction force to the knee can be
> a
>> little
>> > tricky because it is not exactly the same reaction force as in
> the feet
>> > which can be measured with force plates. May I ask how did you get
>> the data
>> > of that force applied at the knee joint?
>> >
>> >
>> >
>> > Best regards,
>> >
>> > Sylvain, AnyBody Support
>> >
>> >
>> >
>> >
>> >
>> > _____
>> >
>> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com]
>> On Behalf
>> > Of kliewec
>> > Sent: 23. januar 2009 10:12
>> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
>> > Subject: [AnyScript] Problems with calculation of hip forces with
> my
>> ouwn
>> > hip model
>> >
>> >
>> >
>> > Hi All,
>> >
>> > I have a problem with the computation of the forces in the hip
> joint.
>> > My hipmodel is assembled from the Repository pedal model. The
> pelvis
>> > is fixed at the global reference frame. In my model there is the
> thigh
>> > the only moveable body, the other bodys are removed. All muscles
> of
>> > the thigh are contained, because the nodes, which connect the
> muscles
>> > to the shank, are transfered to the segment of thigh. The ground
>> > reaction force was set as outside force at the knee.
>> > The calculated hip forces are twice as high as in the Gait3D-
> model.
>> > Does someone have an idea, which the cause for it is?
>> >
>> > The files are uploaded, you find this under “Pelvis-Thigh-G3”.zip
>> > Thanks for your support
>> >
>> > Best regards
>> > Christian
>> >
>> >
>> >
>> >
>> >
>> > [Non-text portions of this message have been removed]
>> >
>>
>
>
>
>
> ------------------------------------
>
> Users of the AnyBody Modeling System help each other create biomechanical
> models in the AnyScript language.Yahoo! Groups Links
>
>
>

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