Hi All,
I am a new comer to this group and AnyBody. One of the first
problems i have come across is determining is the number of degrees of
freedom of the joints that i utilize. Is there a list of these? Like if
i use the thorax and right arm model how many degrees of freedom are
available for each joint.
Shashank
Hi Shashank
Welcome to the group.
If you run the ModelInformation after the model is loaded, then a list
of constrains and DOF will come up. The DOF have names so you can see
where they belong.
Best regards
Christian, AnyBody Support
— In anyscript@yahoogroups.com, “shashu_23” <shashankraina@…> wrote:
>
> Hi All,
> I am a new comer to this group and AnyBody. One of the first
> problems i have come across is determining is the number of degrees of
> freedom of the joints that i utilize. Is there a list of these? Like if
> i use the thorax and right arm model how many degrees of freedom are
> available for each joint.
>
> Shashank
>
Hi Shashank
This is just some extra info to what Christain already wrote.
All applications have a file named Mannequin.any, this file has a
folder named “Posture” it list all the dof in the body with
descriptive names, so this file is a good place to get an overview
of the dof in the model. The FreePosture model is a good place
start looking in. It is a model where all dof in the human are
driven by the settings in the Mannequin.any file.
The file list the dof if you have a fullbody model, in case you are
using a subset like spine and right arm, you obviously need to
exclude the dof of these parts, the naming should explain this.
Best regards
Søren, AnyBody Support
— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Shashank
>
> Welcome to the group.
> If you run the ModelInformation after the model is loaded, then a
list
> of constrains and DOF will come up. The DOF have names so you can
see
> where they belong.
>
> Best regards
> Christian, AnyBody Support
>
> — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
wrote:
> >
> > Hi All,
> > I am a new comer to this group and AnyBody. One of the first
> > problems i have come across is determining is the number of
degrees of
> > freedom of the joints that i utilize. Is there a list of these?
Like if
> > i use the thorax and right arm model how many degrees of freedom
are
> > available for each joint.
> >
> > Shashank
> >
>
Hi,
Thank you for responses to the previous question.
I have a question regrading landing models. I was interested to know
if there are any basic models which are available to users?
Can we simulate flight phase in AnyBody or does is require a
segement to be always connected to the ground
Thanks
Shashank
Hi Shashank
I am a little unsure what kind of landing situation you are looking
for, but unfortunately we do not have a model which are in a
flyphase and then do a landing, that i aware of.
In principle then the segment do not have to be connected to the
ground BUT then the motion of the systems needs to be in perfect
balance and this is only possible in theory. So in practice there
need to be a connection. This can be done by creating “weak”
artificial muscles to the ground. If these muscles are weak they
will only be recruted if they are really needed to help balance the
model.
If the motion of the model is out of balance then these muscles will
act as a “hook” to the world, but obviously this should be avoided
if possible.
Please see the AnyGeneralMuscle in the manual.
Best regards
Søren, AnyBody Support
— In anyscript@yahoogroups.com, “shashu_23” <shashankraina@…>
wrote:
>
> Hi,
> Thank you for responses to the previous question.
> 1. I have a question regrading landing models. I was interested to
know
> if there are any basic models which are available to users?
>
> 2. Can we simulate flight phase in AnyBody or does is require a
> segement to be always connected to the ground
>
> Thanks
> Shashank
>
Hi Shashank and Soeren
We made the model with one single cow limb that was flying and then
landet. We just included all the data from the force plate, also when
the leg was not on it. I don’t recall we attached any weak muscles…
Christian
— In anyscript@yahoogroups.com, “AnyBody Support” <support@…> wrote:
>
> Hi Shashank
>
> I am a little unsure what kind of landing situation you are looking
> for, but unfortunately we do not have a model which are in a
> flyphase and then do a landing, that i aware of.
>
> In principle then the segment do not have to be connected to the
> ground BUT then the motion of the systems needs to be in perfect
> balance and this is only possible in theory. So in practice there
> need to be a connection. This can be done by creating “weak”
> artificial muscles to the ground. If these muscles are weak they
> will only be recruted if they are really needed to help balance the
> model.
>
> If the motion of the model is out of balance then these muscles will
> act as a “hook” to the world, but obviously this should be avoided
> if possible.
>
> Please see the AnyGeneralMuscle in the manual.
>
>
> Best regards
> S�ren, AnyBody Support
>
> — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
> wrote:
> >
> > Hi,
> > Thank you for responses to the previous question.
> > 1. I have a question regrading landing models. I was interested to
> know
> > if there are any basic models which are available to users?
> >
> > 2. Can we simulate flight phase in AnyBody or does is require a
> > segement to be always connected to the ground
> >
> > Thanks
> > Shashank
> >
>
Hi Shashank and Christian
Christain is right, if you have measured ground reactions forces and
model only a part of the human like the legs, you usually apply a
reaction force from pelvis to the ground. This reaction forces is in
principle the missing reaction force between the upper and lower
body, and it will make the weak muscles i mentioned redundant.
In other words, when the model is in the flyphase the reaction force
will carry the legs, so the weak muscles is not needed. These are
still needed though if you do a fullbodymodel, because then this
reaction force on pelvis should not be present, and something else
need to be able to ensure the balance.
Best regards
Søren, AnyBody Support
— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Shashank and Soeren
>
> We made the model with one single cow limb that was flying and then
> landet. We just included all the data from the force plate, also
when
> the leg was not on it. I don’t recall we attached any weak
muscles…
>
> Christian
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
wrote:
> >
> > Hi Shashank
> >
> > I am a little unsure what kind of landing situation you are
looking
> > for, but unfortunately we do not have a model which are in a
> > flyphase and then do a landing, that i aware of.
> >
> > In principle then the segment do not have to be connected to the
> > ground BUT then the motion of the systems needs to be in perfect
> > balance and this is only possible in theory. So in practice
there
> > need to be a connection. This can be done by creating “weak”
> > artificial muscles to the ground. If these muscles are weak they
> > will only be recruted if they are really needed to help balance
the
> > model.
> >
> > If the motion of the model is out of balance then these muscles
will
> > act as a “hook” to the world, but obviously this should be
avoided
> > if possible.
> >
> > Please see the AnyGeneralMuscle in the manual.
> >
> >
> > Best regards
> > S�ren, AnyBody Support
> >
> > — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
> > wrote:
> > >
> > > Hi,
> > > Thank you for responses to the previous question.
> > > 1. I have a question regrading landing models. I was
interested to
> > know
> > > if there are any basic models which are available to users?
> > >
> > > 2. Can we simulate flight phase in AnyBody or does is require
a
> > > segement to be always connected to the ground
> > >
> > > Thanks
> > > Shashank
> > >
> >
>
Hi,
I am Shashank Raina from University of Southern California and am
working on mechanics of wheelchair propulsion.
2.Is the total EMG activity of the muscle defined by the sum ( or some
other combination) of these individual fibers.
Thanks
Shashank
Hi Shashank
If you go to the file BRep/aalbrog/Arm3d/mus.any you will find the
definition of the muscle in the arms. All muscles have a line
something like this
AnyDrawMuscle DrwMus = {
#include “…/drawSettings/MusDrawSettings.any” };
This is the line which defines the graphical representation of the
muscle, if you change this line to for example
AnyDrawMuscle DrwMus = { RGB={0,0,1}; };
You will get a muscle which is blue, in this way you can find out by
which names the muscles are represented.
As i recall it, the deltoid_scapula muscle is numbered with the
muscle element no.1 closest to the spine and no. 6 is almost located
medial.
If you have measured EMG on the deltoid muscles i think you will
have to look at the results from several of the deltoid muscle
branches. Depending on the location of the electrode you will have
to selcet the closest of the branches. You can combine the results
from these in many ways, one way is to draw the activity envelope
form several of the deltoid branches.
Please ask again if you have further questions.
Best regards
Søren, AnyBody Support
— In anyscript@yahoogroups.com, “shashu_23” <shashankraina@…>
wrote:
>
> Hi,
> I am Shashank Raina from University of Southern California and
am
> working on mechanics of wheelchair propulsion.
>
> 1. I was looking at the muscles in the shoulder model.
> Of all the listed muscles for lets say the deltoid. Is there some
way i
> can find which are the anterior fibers and which one the middle.
> In the model they are listed as Deltoideus_clavicular_part1 and so
on.
>
> 2.Is the total EMG activity of the muscle defined by the sum ( or
some
> other combination) of these individual fibers.
>
> Thanks
> Shashank
>
Hi Søren,
Thanks for your response, it helped out.
I have a new question.
During a wheelchair propulsion task Supraspinatus activity is
generally seen during the recovery phase and some times even during
the push phase. When i look at the results of Fm given by the
Supraspinatus muscles in the Wheelchair Rancho data i do not see any
activity. As i have EMG data from the subject that was used to make
this model and there is Supraspin activity during this task. What
could be the reasons that the model does not show any activity in
that muscle?
Thanks
Shashank
— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Shashank
>
> If you go to the file BRep/aalbrog/Arm3d/mus.any you will find the
> definition of the muscle in the arms. All muscles have a line
> something like this
> AnyDrawMuscle DrwMus = {
> #include “…/drawSettings/MusDrawSettings.any” };
>
>
> This is the line which defines the graphical representation of the
> muscle, if you change this line to for example
>
> AnyDrawMuscle DrwMus = { RGB={0,0,1}; };
>
> You will get a muscle which is blue, in this way you can find out
by
> which names the muscles are represented.
>
> As i recall it, the deltoid_scapula muscle is numbered with the
> muscle element no.1 closest to the spine and no. 6 is almost
located
> medial.
>
> If you have measured EMG on the deltoid muscles i think you will
> have to look at the results from several of the deltoid muscle
> branches. Depending on the location of the electrode you will have
> to selcet the closest of the branches. You can combine the results
> from these in many ways, one way is to draw the activity envelope
> form several of the deltoid branches.
>
> Please ask again if you have further questions.
>
> Best regards
> Søren, AnyBody Support
>
> — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
> wrote:
> >
> > Hi,
> > I am Shashank Raina from University of Southern California and
> am
> > working on mechanics of wheelchair propulsion.
> >
> > 1. I was looking at the muscles in the shoulder model.
> > Of all the listed muscles for lets say the deltoid. Is there some
> way i
> > can find which are the anterior fibers and which one the middle.
> > In the model they are listed as Deltoideus_clavicular_part1 and
so
> on.
> >
> > 2.Is the total EMG activity of the muscle defined by the sum ( or
> some
> > other combination) of these individual fibers.
> >
> > Thanks
> > Shashank
> >
>
Hi Shashank
Unfortunately there can be multiple reasons for this difference.
I think the muscle activities have never been compared with measured
EMG before, in this model.
The main root for the error is most likely differences in the
kinematics. In the JointAndDriver.any file you will see that the SC
joint is driven to a fixed position this effectivly means that the
scapula has a fixed position wrt to the pelvis.
Without remembering the marker configuration in detail i think it
should be possible to free up the following dof of this joint
SternoClavicularProtraction
SternoClavicularElevation
and replace those by a marker driver of a marker close to the
acromion which should then drive the scapula up/down
anterior/posterior.
It can also be differences in the segment lenghts that can cause
differences…
Finally by reviewing the c3ddata folder i can see that there is a
measured torque “WHLCHRTQ” this torque has not been applied in the
environment.any file and i am not sure how to apply it since i am
unsure how it was measured, this will also give a difference.
So to improve the results i think these issues should be adressed.
Please ask again if you have further questions.
Best regards
Søren, AnyBody Support
— In anyscript@yahoogroups.com, “shashu_23” <shashankraina@…>
wrote:
>
> Hi Søren,
> Thanks for your response, it helped out.
> I have a new question.
> During a wheelchair propulsion task Supraspinatus activity is
> generally seen during the recovery phase and some times even
during
> the push phase. When i look at the results of Fm given by the
> Supraspinatus muscles in the Wheelchair Rancho data i do not see
any
> activity. As i have EMG data from the subject that was used to
make
> this model and there is Supraspin activity during this task. What
> could be the reasons that the model does not show any activity in
> that muscle?
>
> Thanks
> Shashank
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> wrote:
> >
> > Hi Shashank
> >
> > If you go to the file BRep/aalbrog/Arm3d/mus.any you will find
the
> > definition of the muscle in the arms. All muscles have a line
> > something like this
> > AnyDrawMuscle DrwMus = {
> > #include “…/drawSettings/MusDrawSettings.any” };
> >
> >
> > This is the line which defines the graphical representation of
the
> > muscle, if you change this line to for example
> >
> > AnyDrawMuscle DrwMus = { RGB={0,0,1}; };
> >
> > You will get a muscle which is blue, in this way you can find
out
> by
> > which names the muscles are represented.
> >
> > As i recall it, the deltoid_scapula muscle is numbered with the
> > muscle element no.1 closest to the spine and no. 6 is almost
> located
> > medial.
> >
> > If you have measured EMG on the deltoid muscles i think you will
> > have to look at the results from several of the deltoid muscle
> > branches. Depending on the location of the electrode you will
have
> > to selcet the closest of the branches. You can combine the
results
> > from these in many ways, one way is to draw the activity
envelope
> > form several of the deltoid branches.
> >
> > Please ask again if you have further questions.
> >
> > Best regards
> > Søren, AnyBody Support
> >
> > — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
> > wrote:
> > >
> > > Hi,
> > > I am Shashank Raina from University of Southern California
and
> > am
> > > working on mechanics of wheelchair propulsion.
> > >
> > > 1. I was looking at the muscles in the shoulder model.
> > > Of all the listed muscles for lets say the deltoid. Is there
some
> > way i
> > > can find which are the anterior fibers and which one the
middle.
> > > In the model they are listed as Deltoideus_clavicular_part1
and
> so
> > on.
> > >
> > > 2.Is the total EMG activity of the muscle defined by the sum (
or
> > some
> > > other combination) of these individual fibers.
> > >
> > > Thanks
> > > Shashank
> > >
> >
>
Hi Sylvain,
Thanks for the response, i was able to reduce the muscle strength
of the triceps( to a very low number) but when i did this an
interesting thing happens, the simulation is supposed to go for
20steps, but with the reduced triceps strength the simluation goes
till step 14 (where the hand is about just part the top dead center
of the wheel) and then it does not go ahead, it does not give an
error, the processing just keeps going without any results. Have you
seen something like this before?
I think it is taking the optimization routine a long time to find an
optimal solution for the problem, but how long i have let the
simulation run for 20mins with no steps forward or no errors. Would
like yo hear your thoughts on this.
Thanks
Shashank
— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
>
> Hi Shashank,
>
> 1. If i use AnyMan to change segment lengths with uniform
> scaling
> does the ligament lengths automatically get adjusted or do i have
to
> run a calibration study.
>
> Actually there are no ligaments in this model, but if there where
> some then yes you should run a calibration study each time you load
> the model (no matter if you scale it or not, you have to run it
> anyway). It is also the case for the MuscleModel_3E: if you use it
> you must run the calibration study each time you load the model.
But
> you are now using the simple muscle model so it is not necessary to
> run a calibration study.
>
> 2. What is the purpose of the tolerance error. How does it
> affect the
> function of the model.
>
> I suppose you are talking about the kinematics tolerance, it is the
> tolerance of matching of the joints. When you get an error message
> saying “kinematics error above tolerance…” that mean some joints
> cannot match with enough accuracy. It is possible to change this
> tolerance but it is definitely not a good idea because then your
> model loose accuracy and you will get noise in the results.
>
> 3. If i want to make specific muscles (eg triceps) have low
> strength
> or For ex a wheelchair users with C7 level injury does not have
> control of abdominal musculature what is the best way to represent
it
> in the model.
>
> Each muscle have is own muscle model located in BRep in the file
> Muscle_Parameters, so if you want a specific muscle to be weaker or
> stronger you can change the parameters in his own muscle model. In
> AnyBody the strength of the muscles is calculated with the PCSA,
this
> is the value you have to modify.
> For the case of the C7 level injury I am not sure of the exact
> physical consequences on the person you want to add in the model.
> Does that mean the person is not able to move the lumbar spine? In
> that case you should begin by constrain the movement of the spine
> with drivers. You should also think about whatever the person need
to
> be strapped to the seat or some other things. I guess that also
mean
> the person cannot contract his muscles and those muscles remain
> without any activity, you could represent that behaviour by setting
a
> very low strength to those particular muscles, you can’t totally
> remove them because the model won’t be balanced anymore.
>
> Best regards,
> Sylvain, AnyBody Support
>
>
>
> — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
> wrote:
> >
> > Hi Sylvain,
> > I have a few more generic questions.
> >
> > 1. If i use AnyMan to change segment lengths with uniform scaling
> > does the ligament lengths automatically get adjusted or do i have
> to
> > run a calibration study.
> >
> > 2. What is the purpose of the tolerance error. How does it affect
> the
> > function of the model.
> >
> > 3. If i want to make specific muscles (eg triceps) have low
> strength
> > or For ex a wheelchair users with C7 level injury does not have
> > control of abdominal musculature what is the best way to
represent
> it
> > in the model.
> >
> > Thanks
> >
> > Shashank
> >
> > — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> > wrote:
> > >
> > > Hi Shashank,
> > >
> > > One of the causes of the difference of activity of the scapular
> > > muscles can be the shoulder external rotation driver, precisely
> the
> > > one I couldn’t replace by a marker driver. That mean the
shoulder
> > > external rotation is fixed instead of following the marker
> > movement.
> > > The difference in the motion is not big and as you saw the
model
> is
> > > running, but still that small difference will induce somehow a
> > change
> > > in some muscles activity (most probably in the shoulder/scapula
> > > area). That’s why it would be good to have that DOF driven by
the
> > > marker.
> > >
> > > It seems your understanding of the simple driver is not
correct,
> > > things are like this: when you drive a joint with a simple
driver
> > the
> > > muscles are involved in the inverse dynamics analysis as well
as
> if
> > > you drive it with a marker. There is a parameter in the Driver
> > class
> > > that controls if the muscles will be active or not, it is the
> > > ReactionType. If ReactionType is “Off” then muscles are
normally
> > > caring the load for that joint, if switched “On” the load in
the
> > join
> > > won’t be carried by the muscles but by artificial forces
> generated
> > in
> > > the joint. Default value of ReactionType is “Off”.
> > > So clearly the difference between those two types of driver is
> that
> > > you cannot drive such complex motion with a simple driver than
> with
> > a
> > > marker driver. And it affects your model in the way I explained
> > > before.
> > >
> > > Best regards,
> > > Sylvain, AnyBody Support
> > >
> > >
> > >
> > >
> > >
> > > — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
> > > wrote:
> > > >
> > > > Sylvain,
> > > > I ran the model and it works fine.
> > > > I also noticed that the scapular muscles are also active
(this
> > was
> > > > not the case with Rancho model). I will look in and see how
> drove
> > > the
> > > > scapula motion.
> > > >
> > > > My question is when we drive something with a simple driver.
> That
> > > > means the muscles for that joint won’t get involved in the
> > inverse
> > > > dynamics (am i correct). Could you tell me if it affects any
> > other
> > > > results of the model.
> > > >
> > > > Thanks a lot,
> > > >
> > > > Shashank
> > > >
> > > >
> > > > — In anyscript@yahoogroups.com, “AnyBody Support”
<support@>
> > > > wrote:
> > > > >
> > > > > Hi,
> > > > >
> > > > > I think one problem comes from the initial position, some
> > > segments
> > > > > was not close enough to the marker. So I modified some
values
> > in
> > > > the
> > > > > mannequin file. I also changed some of the drivers for the
> > trunk
> > > > that
> > > > > was giving me troubles: I fixed the pelvis to the ground
and
> I
> > > > drive
> > > > > 2 rotations of the thoracic spine with marker while the
third
> > one
> > > > > keeps giving troubles and I had to drive it with a simple
> > driver.
> > > > > I have the model running now but there is a second driver
(in
> > > plus
> > > > of
> > > > > the trunk rotation) that I could not replace by a marker.
> It’s
> > > the
> > > > > one for the shoulder external rotation. So far I haven’t
find
> > > what
> > > > > cause the error for this driver, it might be a problem of
> > initial
> > > > > position again.
> > > > > I will keep digging the problem. I uploaded the running
model
> > > > > (wheel_shank2).
> > > > >
> > > > > Best regards,
> > > > > Sylvain, AnyBody Support
> > > > >
> > > > >
> > > > > — In anyscript@yahoogroups.com, “shashu_23”
> <shashankraina@>
> > > > > wrote:
> > > > > >
> > > > > > Hi,
> > > > > > I was trying to run a wheelchair model using the
> template
> > > > given
> > > > > in
> > > > > > WheelChairRancho. I was trying to keep it simple so i
just
> > > tried
> > > > to
> > > > > > replace the marker data and kept the forces the same.
> > > > > > Here is what is did.
> > > > > >
> > > > > > 1. The motion markers i have are the same as the ones in
> the
> > > > model
> > > > > so
> > > > > > that step is fine.
> > > > > >
> > > > > > 2. I aligned the segments so they are close to the
markers
> in
> > > the
> > > > > > starting position.
> > > > > >
> > > > > > 3. The model loads up fine. The problem is when i start
> > > removing
> > > > > the
> > > > > > drivers from the joint and drivers file. I get errors in
> > right
> > > > away
> > > > > > and i have tried different combinations but it still
keeps
> > > > failing.
> > > > > >
> > > > > > 4. I was hoping if someone could look at the model and
help
> > me
> > > > > figure
> > > > > > out the problem with the drivers. I have uploaded the
> folder
> > > > called
> > > > > > wheel_shank with the model in it. I have also put in the
> c3d
> > > > data
> > > > > > file if you want it. (as i mentioned before i am still
> using
> > > the
> > > > > > forces from the rancho file as there is some vector
length
> > > issues
> > > > > > with the forces)
> > > > > >
> > > > > > It would be of great help is someone could help me figure
> out
> > > how
> > > > > to
> > > > > > solve this problem
> > > > > >
> > > > > > Thanks
> > > > > > Shashank
> > > > > >
> > > > >
> > > >
> > >
> >
>
Hi Sashank,
This definitely looks like an infinite loop, that mean the solver
gets stuck in the calculation and will never go to the next step. You
can fix that by changing the solver in the study folder. If you are
now using the MinMaxNRSimplex try to switch to MinMaxSimplex. This
solver is stronger in term of calculation and that usually avoid the
infinite loop problem.
I don’t really know what you mean by “cuff” markers, but I think that
does not really matter. The important thing is to know to which bones
the markers are related to, then you can create and modify all the
markers you want in AnyBody corresponding to the ones of your motion
capture. The Kit-Vaughn set is the one we usually use, but if they
don’t correspond to what you need you can adjust the markers
coordinates and create new markers if you miss some.
Best regards,
Sylvain, AnyBody Support.
— In anyscript@yahoogroups.com, “shashu_23” <shashankraina@…>
wrote:
>
> Hi Sylvain,
> Thanks for the response, i was able to reduce the muscle
strength
> of the triceps( to a very low number) but when i did this an
> interesting thing happens, the simulation is supposed to go for
> 20steps, but with the reduced triceps strength the simluation goes
> till step 14 (where the hand is about just part the top dead center
> of the wheel) and then it does not go ahead, it does not give an
> error, the processing just keeps going without any results. Have
you
> seen something like this before?
>
> I think it is taking the optimization routine a long time to find
an
> optimal solution for the problem, but how long i have let the
> simulation run for 20mins with no steps forward or no errors. Would
> like yo hear your thoughts on this.
>
>
> 2. This is a different question regarding a different setup. I did
a
> study where i used cuff markers instead of joint landmarks. Should
i
> estimate the markers defined in the Kit-Vaughn marker sets before
> importing the model into AnyBody or is there another way of using
> cuff marker data (like a set of four markers on the forearm).
>
>
> Thanks
> Shashank
>
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> wrote:
> >
> >
> > Hi Shashank,
> >
> > 1. If i use AnyMan to change segment lengths with uniform
> > scaling
> > does the ligament lengths automatically get adjusted or do i have
> to
> > run a calibration study.
> >
> > Actually there are no ligaments in this model, but if there where
> > some then yes you should run a calibration study each time you
load
> > the model (no matter if you scale it or not, you have to run it
> > anyway). It is also the case for the MuscleModel_3E: if you use
it
> > you must run the calibration study each time you load the model.
> But
> > you are now using the simple muscle model so it is not necessary
to
> > run a calibration study.
> >
> > 2. What is the purpose of the tolerance error. How does it
> > affect the
> > function of the model.
> >
> > I suppose you are talking about the kinematics tolerance, it is
the
> > tolerance of matching of the joints. When you get an error
message
> > saying “kinematics error above tolerance…” that mean some joints
> > cannot match with enough accuracy. It is possible to change this
> > tolerance but it is definitely not a good idea because then your
> > model loose accuracy and you will get noise in the results.
> >
> > 3. If i want to make specific muscles (eg triceps) have low
> > strength
> > or For ex a wheelchair users with C7 level injury does not have
> > control of abdominal musculature what is the best way to
represent
> it
> > in the model.
> >
> > Each muscle have is own muscle model located in BRep in the file
> > Muscle_Parameters, so if you want a specific muscle to be weaker
or
> > stronger you can change the parameters in his own muscle model.
In
> > AnyBody the strength of the muscles is calculated with the PCSA,
> this
> > is the value you have to modify.
> > For the case of the C7 level injury I am not sure of the exact
> > physical consequences on the person you want to add in the model.
> > Does that mean the person is not able to move the lumbar spine?
In
> > that case you should begin by constrain the movement of the spine
> > with drivers. You should also think about whatever the person
need
> to
> > be strapped to the seat or some other things. I guess that also
> mean
> > the person cannot contract his muscles and those muscles remain
> > without any activity, you could represent that behaviour by
setting
> a
> > very low strength to those particular muscles, you can’t totally
> > remove them because the model won’t be balanced anymore.
> >
> > Best regards,
> > Sylvain, AnyBody Support
> >
> >
> >
> > — In anyscript@yahoogroups.com, “shashu_23” <shashankraina@>
> > wrote:
> > >
> > > Hi Sylvain,
> > > I have a few more generic questions.
> > >
> > > 1. If i use AnyMan to change segment lengths with uniform
scaling
> > > does the ligament lengths automatically get adjusted or do i
have
> > to
> > > run a calibration study.
> > >
> > > 2. What is the purpose of the tolerance error. How does it
affect
> > the
> > > function of the model.
> > >
> > > 3. If i want to make specific muscles (eg triceps) have low
> > strength
> > > or For ex a wheelchair users with C7 level injury does not have
> > > control of abdominal musculature what is the best way to
> represent
> > it
> > > in the model.
> > >
> > > Thanks
> > >
> > > Shashank
> > >
> > > — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> > > wrote:
> > > >
> > > > Hi Shashank,
> > > >
> > > > One of the causes of the difference of activity of the
scapular
> > > > muscles can be the shoulder external rotation driver,
precisely
> > the
> > > > one I couldn’t replace by a marker driver. That mean the
> shoulder
> > > > external rotation is fixed instead of following the marker
> > > movement.
> > > > The difference in the motion is not big and as you saw the
> model
> > is
> > > > running, but still that small difference will induce somehow
a
> > > change
> > > > in some muscles activity (most probably in the
shoulder/scapula
> > > > area). That’s why it would be good to have that DOF driven by
> the
> > > > marker.
> > > >
> > > > It seems your understanding of the simple driver is not
> correct,
> > > > things are like this: when you drive a joint with a simple
> driver
> > > the
> > > > muscles are involved in the inverse dynamics analysis as well
> as
> > if
> > > > you drive it with a marker. There is a parameter in the
Driver
> > > class
> > > > that controls if the muscles will be active or not, it is the
> > > > ReactionType. If ReactionType is “Off” then muscles are
> normally
> > > > caring the load for that joint, if switched “On” the load in
> the
> > > join
> > > > won’t be carried by the muscles but by artificial forces
> > generated
> > > in
> > > > the joint. Default value of ReactionType is “Off”.
> > > > So clearly the difference between those two types of driver
is
> > that
> > > > you cannot drive such complex motion with a simple driver
than
> > with
> > > a
> > > > marker driver. And it affects your model in the way I
explained
> > > > before.
> > > >
> > > > Best regards,
> > > > Sylvain, AnyBody Support
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > — In anyscript@yahoogroups.com, “shashu_23”
<shashankraina@>
> > > > wrote:
> > > > >
> > > > > Sylvain,
> > > > > I ran the model and it works fine.
> > > > > I also noticed that the scapular muscles are also active
> (this
> > > was
> > > > > not the case with Rancho model). I will look in and see how
> > drove
> > > > the
> > > > > scapula motion.
> > > > >
> > > > > My question is when we drive something with a simple
driver.
> > That
> > > > > means the muscles for that joint won’t get involved in the
> > > inverse
> > > > > dynamics (am i correct). Could you tell me if it affects
any
> > > other
> > > > > results of the model.
> > > > >
> > > > > Thanks a lot,
> > > > >
> > > > > Shashank
> > > > >
> > > > >
> > > > > — In anyscript@yahoogroups.com, “AnyBody Support”
> <support@>
> > > > > wrote:
> > > > > >
> > > > > > Hi,
> > > > > >
> > > > > > I think one problem comes from the initial position, some
> > > > segments
> > > > > > was not close enough to the marker. So I modified some
> values
> > > in
> > > > > the
> > > > > > mannequin file. I also changed some of the drivers for
the
> > > trunk
> > > > > that
> > > > > > was giving me troubles: I fixed the pelvis to the ground
> and
> > I
> > > > > drive
> > > > > > 2 rotations of the thoracic spine with marker while the
> third
> > > one
> > > > > > keeps giving troubles and I had to drive it with a simple
> > > driver.
> > > > > > I have the model running now but there is a second driver
> (in
> > > > plus
> > > > > of
> > > > > > the trunk rotation) that I could not replace by a marker.
> > It’s
> > > > the
> > > > > > one for the shoulder external rotation. So far I haven’t
> find
> > > > what
> > > > > > cause the error for this driver, it might be a problem of
> > > initial
> > > > > > position again.
> > > > > > I will keep digging the problem. I uploaded the running
> model
> > > > > > (wheel_shank2).
> > > > > >
> > > > > > Best regards,
> > > > > > Sylvain, AnyBody Support
> > > > > >
> > > > > >
> > > > > > — In anyscript@yahoogroups.com, “shashu_23”
> > <shashankraina@>
> > > > > > wrote:
> > > > > > >
> > > > > > > Hi,
> > > > > > > I was trying to run a wheelchair model using the
> > template
> > > > > given
> > > > > > in
> > > > > > > WheelChairRancho. I was trying to keep it simple so i
> just
> > > > tried
> > > > > to
> > > > > > > replace the marker data and kept the forces the same.
> > > > > > > Here is what is did.
> > > > > > >
> > > > > > > 1. The motion markers i have are the same as the ones
in
> > the
> > > > > model
> > > > > > so
> > > > > > > that step is fine.
> > > > > > >
> > > > > > > 2. I aligned the segments so they are close to the
> markers
> > in
> > > > the
> > > > > > > starting position.
> > > > > > >
> > > > > > > 3. The model loads up fine. The problem is when i
start
> > > > removing
> > > > > > the
> > > > > > > drivers from the joint and drivers file. I get errors
in
> > > right
> > > > > away
> > > > > > > and i have tried different combinations but it still
> keeps
> > > > > failing.
> > > > > > >
> > > > > > > 4. I was hoping if someone could look at the model and
> help
> > > me
> > > > > > figure
> > > > > > > out the problem with the drivers. I have uploaded the
> > folder
> > > > > called
> > > > > > > wheel_shank with the model in it. I have also put in
the
> > c3d
> > > > > data
> > > > > > > file if you want it. (as i mentioned before i am still
> > using
> > > > the
> > > > > > > forces from the rancho file as there is some vector
> length
> > > > issues
> > > > > > > with the forces)
> > > > > > >
> > > > > > > It would be of great help is someone could help me
figure
> > out
> > > > how
> > > > > > to
> > > > > > > solve this problem
> > > > > > >
> > > > > > > Thanks
> > > > > > > Shashank
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
Hi Sylvain,
I had a question on the optimization routine for solving muscle
forces. As i understand the load is distributed according to the
strength of each muscle spanning the joint and in this way the fatigue
in each muscle is delayed.
My question was if the goal of the task was to generate very high force
ignoring fatigue and stabilization of joints(seen in some atlhletes)
can that be used as a criteria for calculating muscle force. So in
other words are there other routines to solve for muscle forces.
I hope my question is clear.
Thanks
Shashank
Hi Shashank,
Currently there are no other routines to calculate the muscles recruitment.
I mean the criterion is always to minimize the muscle activity. But it is
possible to influence the way the muscle activity is minimized (maximum
activity, sum of activities .) by using the linear or quadratic solver and
by playing around with the linear penalty and quadratic penalty. For example
using a high linear penalty will have the effect of recruiting only the
muscles that are best situated around a joint and have a maximum effect on
it. This might be close to what you want.
There is a tutorial dealing with the different parameters of the recruitment
( http://www.anybodytech.com/691.0.html ), you are of course very welcome to
look at it.
Best regards,
Sylvain, AnyBody Support
From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of shashu_23
Sent: 4. november 2008 19:19
To: anyscript@yahoogroups.com
Subject: [AnyScript] Question
Hi Sylvain,
I had a question on the optimization routine for solving muscle
forces. As i understand the load is distributed according to the
strength of each muscle spanning the joint and in this way the fatigue
in each muscle is delayed.
My question was if the goal of the task was to generate very high force
ignoring fatigue and stabilization of joints(seen in some atlhletes)
can that be used as a criteria for calculating muscle force. So in
other words are there other routines to solve for muscle forces.
I hope my question is clear.
Thanks
Shashank
[Non-text portions of this message have been removed]