Hi Prabhav
Sometimes the calibration runs faster when the modelview is switched of, I
think this is related to the graphical display of the muscles.
I think your problem could be related to that the modelview can not cope
with the analysis speed, but I am not sure.
Please try to check the output from the calibration and see if the position
variation of your joints is as expected. If these are ok, the issue is
related to the graphics only.
By the way why do you calibrate through a cycle of 50 steps, and not just
one position for example in the middle of your motion range so nstep=1?.
Best regards
Søren
From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of Prabhav Saraswat
Sent: 11 December 2007 21:25
To: anyscript@yahoogroups.com
Subject: Re: [AnyScript] Re: Muscle with multiple tendons
Hi,
I have a question about calibration study. I am using same drivers for
calibration and kinematic analysis. Problem is that the calibration analysis
runs only first few steps of drivers.
I have 50 time steps in the driver files and the graphics seems to work
fine in kinematic or inverse dynamic analysis. But when I run the
calibration analaysis, it runs for first few steps only in the graphics
window, but in the operation window, it seem to be going through all 50
steps.
It seems to me that driver move slower in calibratinon study. Is it just
the issue with graphics, its not able to catch up with calibration analysis?
OR I need to define calibration drivers differently?
I hope, I am able to explain my problem here, I will upload the model if it
doesnt make sense.
Prabhav.
On 9/10/07, Prabhav Saraswat <prabhav.saraswat@
<mailto:prabhav.saraswat%40gmail.com> gmail.com> wrote:
>
> Hi Soren,
> I am trying to follow what you suggested. I am driving the muscle
> intermediate segment with AnyKinMeasureLinComb with a ratio for
> muscle/tendon length. I defined the tendons as muscle model.
>
> It can solve the inverse dynamics but those tendons are providing
> force now. And these tendons forces dont add up to the total muscle force.
> Which muscle properties do you suggest me to change so that the
> tendons behave as elastic element?
>
> thanks
>
> prabhav.
>
>
> On 9/3/07, AnyBody Support <support@anybodytech
<mailto:support%40anybodytech.com> .com> wrote:
> >
> > Hi Prabhav
> >
> > Your absolute right in your description of the problem.
> >
> > You can create tendons in AnyBody and you can give it elastic
> > properties as you like. The fundamental problem is that this new
> > segment that the tendon is connected to, is driven to a certain
> > position so it will not automatically find it owns force
> > equilibrium. It would have been nice if this segment could find the
> > position that would give equilibrium between the four tendon units
> > and the muscle unit, but this is not directly possible.
> >
> > This was why i wrote that the force in the tendons will be
> > determined by the way this segment is driven, and then the muscle
> > will have to deliver a force that can match this.
> >
> > The solution is to have a strong muscle units as the tendons. In
> > principle this could also be an AnyReacForce. This setup will ensure
> > that when the muscle is activated forces can be transferred to all
> > four tendons. The load case will determine the load distribution
> > among the individual tendons.
> >
> > By the way you can use a AnyKinLinCombMeasure to drive the position
> > of the segment to alway keep the same realtive position wrt to the
> > length of the muscle.
> >
> > I hope this made it more clear
> >
> > Best regards
> > Søren, AnyBody Support
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
<anyscript%40yahoogroups.com>, “Prabhav
> > Saraswat”
> > <prabhav.saraswat@…> wrote:
> > >
> > > Thanks Soren,
> > > But I am a little confused with your suggestion. How can I
> > define
> > > tendons as muscles. What will be the properties of those
> > tendon “muscles”
> > > then. Tendons are elastic and should I model them as ligaments??
> > OR there is
> > > way to alter muscle properties to make them behave as elastic unit.
> > >
> > > Motivation of this whole procedure was to have one muscle and 4
> > tendons
> > > because physiological properties of muscle (say Extensor Digitorum
> > > Longus) is known as a single unit. When I defined them as 4
> > different
> > > muscles, I had 2 problems-
> > >
> > > (i) Physiological Properties are not known for each muscle
> > separately.
> > > (ii) Most of the force was taken up by one of those 4 muscles.
> > > But physiologically, all the tendons are connected to same muscle
> > and should
> > > be activated together.
> > >
> > > Idea of this muscle with bifurcating tendons, was that when
> > muscle is
> > > activated, force should be transfered to four tendons. I hope, I
> > am making
> > > sense here.
> > >
> > > Let me know of your thoughts about this.
> > >
> > > thanks
> > > prabhav.
> > >
> > >
> > >
> > >
> > > On 8/30/07, AnyBody Support <support@…> wrote:
> > > >
> > > > Hi Prabhav
> > > >
> > > > You can use the AnyDrawPLine for visualizing the line.
> > > >
> > > > Now that you have the kinematics running, the next step is to
> > look
> > > > into the force balance. As i recall the model the driver on the
> > > > PLine will carry the load in the tendon. To prevent this you
> > need to
> > > > swicth off the reactions in this driver. The tendon and the
> > muscles
> > > > should then carry the load, but if you have muscles on one side
> > and
> > > > tendon on the other side of this segment it will most likely not
> > > > work. This is because the force in the tendons are determined by
> > the
> > > > kinematics so the muscles will have to balance this force
> > exactly,
> > > > meaning that the muscles force will be effectively controlled by
> > the
> > > > way you drive this intermediate segment.
> > > >
> > > > The soltuion i to replace the tendon with real muscles, in this
> > way
> > > > they can balance any load.
> > > >
> > > > Best regards
> > > > Søren
> > > >
> > > > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com <anyscript%40yahoogroups.com><anyscript%
> > 40yahoogroups.com>, “Prabhav
> > > > Saraswat”
> > > > <prabhav.saraswat@> wrote:
> > > > >
> > > > > Thanks Soren,
> > > > > It works !!! One more question, I can’t make these
> > > > > AnyKinPLine to show in model view. Is there any AnyDraw…
> > function
> > > > specific
> > > > > for AnyKinPLine. If not, what other function can I use to make
> > it
> > > > appear on
> > > > > the model view.
> > > > >
> > > > > thanks again.
> > > > >
> > > > > prabhav.
> > > > >
> > > > >
> > > > > On 8/28/07, AnyBody Support <support@> wrote:
> > > > > >
> > > > > > Hi Prabhav
> > > > > >
> > > > > > I think the problem is that you try to drive the EDL2 which
> > is a
> > > > > > AnyKinPLine with multiple nodes on the midfoot segment to
> > have
> > > > zero
> > > > > > length. This is not possible because it has several nodes on
> > the
> > > > > > same segment which it would need to collaps to make the EDL2
> > > > length
> > > > > > zero.
> > > > > >
> > > > > > Otherwise i think the model looks ok. To obtain a good guess
> > on
> > > > the
> > > > > > what you should drive the lenght os EDL2 to be, you could do
> > the
> > > > > > folllowing
> > > > > > 1. create a driver for the primatic joint and drive the
> > > > intermediate
> > > > > > segment to the desired position
> > > > > > 2. then read out the pos value of the EDL2 measure
> > > > > > 3. remove the driver on the prismatic joint
> > > > > > 4 use the driver you have already on EDL2 and insert a
> > reaslictic
> > > > > > length
> > > > > >
> > > > > > Hope it solves the problem
> > > > > >
> > > > > > Best regards
> > > > > > Søren
> > > > > >
> > > > > > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com <anyscript%40yahoogroups.com><anyscript%
> > 40yahoogroups.com><anyscript%
> > > > 40yahoogroups.com>, “Prabhav
> > > > > > Saraswat”
> > > > > > <prabhav.saraswat@> wrote:
> > > > > > >
> > > > > > > Hi Soren,
> > > > > > >
> > > > > > > I have shown the steps below that I am following to
> > > > > > implement the
> > > > > > > multiple tendons on same muscle. I am getting this error-
> > > > > > >
> > > > > > > Failed to resolve kinematic constraints (final kin. error =
> > > > > > 1.052148E-001)
> > > > > > > Constraint no. 59 above error tolerance 0.000001, error =
> > > > > > 0.105215. -
> > > > > > > constr. #0 in ‘Main.GaitDrivers.EDLDriver’
> > > > > > > ERROR(OBJ.MCH.KIN3) : G:\Research\Foot Model\My
> > Folder\Anybody
> > > > > > Model\Foot
> > > > > > > Model\FootModel.any(49) : ‘MyStudy’ : Kinematic analysis
> > > > > > failed in time
> > > > > > > step 0
> > > > > > >
> > > > > > > I am probably doing something wrong with the driver. Please
> > > > let me
> > > > > > know,
> > > > > > > what am I doing wrong.
> > > > > > >
> > > > > > > **
> > > > > > > *1. Create an intermediate segment at the end of the muscle
> > > > unit
> > > > > > driving the
> > > > > > > four tendons. It should have zero mass and be small. *
> > > > > > > **
> > > > > > > // Intermediate Segments
> > > > > > > //for EDL
> > > > > > > AnySeg LateralSuperiorSurface2Cuneiforms = {
> > > > > > > r0= { 0.1156,0.0815,-0.4957};
> > > > > > > //Axes0= {{-0.0540,0.9964,0.0656},{-0.9739,-0.0380,-
> > 0.2237},{-
> > > > > > 0.2204,-0.0760
> > > > > > > ,0.9724}}; Mass = 0;
> > > > > > > Jii = {0,0,0};
> > > > > > > AnyRefNode node = {
> > > > > > > sRel ={0,0,0};
> > > > > > > AnyDrawNode drw={ScaleXYZ={0.02,.02,.02}; RGB={0,1,0};};
> > > > > > > };
> > > > > > > };
> > > > > > > //End of EDB
> > > > > > >
> > > > > > >
> > > > > > > 2. Create a prismatic joint between some close element on
> > the
> > > > > > foot and the
> > > > > > > intermediate segment that only allows the segment to model
> > in
> > > > line
> > > > > > with the
> > > > > > > muscle.
> > > > > > >
> > > > > > >
> > > > > > > // Intermediate Joints
> > > > > > >
> > > > > > > AnyPrismaticJoint EDLJoint = {
> > > > > > >
> > > > > > > Axis = x;
> > > > > > >
> > > > > > > AnyRefNode &HindFootNode =
> > > > > >
> > > …FootSegmentsRight.HindFoot.LateralSuperiorSurface2Cuneiforms;
> > > > > > >
> > > > > > > AnyRefNode &IntSegNode =
> > > > > > > …FootSegmentsRight.LateralSuperiorSurface2Cuneiforms.node;
> > > > > > >
> > > > > > > };
> > > > > > > *3. Attach the muscle to the intermediate segment. *
> > > > > > >
> > > > > > > AnyViaPointMuscle EDL = {
> > > > > > >
> > > > > > > AnyMuscleModel &MusMdl = …FootMuscleParameters3E.EDL2Par;
> > > > > > >
> > > > > > > AnyRefNode &Org
> > > > > > = …FootSegmentsRight.Shank.AnteriorProximalInterosseus;
> > > > > > >
> > > > > > > AnyRefNode &Via1
> > > > = …FootSegmentsRight.Shank.LateralRetinaculum;
> > > > > > >
> > > > > > > AnyRefNode &Ins =
> > > > > > > …FootSegmentsRight.LateralSuperiorSurface2Cuneiforms.node;
> > > > > > >
> > > > > > > AnyDrawMuscle DrwMus = {
> > > > > > >
> > > > > > > Bulging = …Bulge;
> > > > > > >
> > > > > > > RGB = …RGB;
> > > > > > >
> > > > > > > Transparency = …Trans;
> > > > > > >
> > > > > > > MaxStress = …MaxS;
> > > > > > >
> > > > > > > };
> > > > > > >
> > > > > > > };
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > 4. Replace the four tendons with AnyKinPLine’s and attach
> > the
> > > > > > origin of
> > > > > > > them to the intermediate segment.
> > > > > > >
> > > > > > > //Tendons
> > > > > > >
> > > > > > > AnyKinPLine EDL2 = {
> > > > > > >
> > > > > > > AnyRefNode &Org =
> > > > > > > …FootSegmentsRight.LateralSuperiorSurface2Cuneiforms.node;
> > > > > > >
> > > > > > > AnyRefNode &Via3 =
> > > > > >
> > > >
> > > …FootSegmentsRight.MidFoot.PosteriorMedialSuperiorTuberosityMT2;
> > > > > > >
> > > > > > > AnyRefNode &Via4 =
> > > > > > > …FootSegmentsRight.MidFoot.AnteriorSuperiorTuberosityMT2;
> > > > > > >
> > > > > > > AnyRefNode &Via5 =
> > > > > >
> > > >
> > > …FootSegmentsRight.Toes.ProximalPosteriorSuperiorTuberosityToe2;
> > > > > > >
> > > > > > > AnyRefNode &Ins =
> > > > > >
> > > …FootSegmentsRight.Toes.DistalPosteriorSuperiorTuberosityToe2;
> > > > > > >
> > > > > > > };
> > > > > > > 5. Create drivers that keep the length of each AnyKinPLine
> > > > > > constant (An
> > > > > > > AnyKinPLine is a kinematic measure, and hence it can be
> > > > driven).
> > > > > > >
> > > > > > > AnyKinEqSimpleDriver EDL2Driver = {
> > > > > > >
> > > > > > > AnyKinMeasure &TendonLoength =
> > > > > > Main.FootModel.FootMusclesRight.EDL2;
> > > > > > >
> > > > > > > DriverPos = {0};
> > > > > > >
> > > > > > > DriverVel = {0.0};
> > > > > > >
> > > > > > > Reaction.Type = {1};
> > > > > > >
> > > > > > > };
> > > > > > >
> > > > > > >
> > > > > > > Thanks for you help.
> > > > > > > **
> > > > > > > Prabhav.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > On 7/11/07, AnyBody Support <support@> wrote:
> > > > > > > >
> > > > > > > > Hi Prabhav
> > > > > > > >
> > > > > > > > The solution you describe is currently the only way to do
> > > > this.
> > > > > > > >
> > > > > > > > You need to define an intermediate segment for
> > connecting the
> > > > > > > > muscles.
> > > > > > > >
> > > > > > > > This poses the question on how to drive this segment, in
> > > > reality
> > > > > > > > this point of the muscle will find an equilibrium
> > positions
> > > > by
> > > > > > > > itslef, but in the model we need to drive thsi manually.
> > > > Here i
> > > > > > > > would recommend using linear combination measures, in
> > this
> > > > way
> > > > > > you
> > > > > > > > may drive the segment to stay for example exactly in the
> > > > middle
> > > > > > of
> > > > > > > > the muscle, or at some certain fraction.
> > > > > > > >
> > > > > > > > Depending on the muscle configuration the muscles may be
> > > > able to
> > > > > > > > balance the forces on this segment, but additional
> > reactions
> > > > to
> > > > > > the
> > > > > > > > underlying segment may be needed at least in the normal
> > > > > > direction of
> > > > > > > > the bone and also the moment reactions (which should be
> > zero
> > > > but
> > > > > > > > needs to be carried anyway).
> > > > > > > >
> > > > > > > > Plase ask again if you have further questions
> > > > > > > >
> > > > > > > > Best regards
> > > > > > > > Søren, AnyBody Support
> > > > > > > >
> > > > > > > > — In anyscript@yahoogrou
<mailto:anyscript%40yahoogroups.com>
ps.com<anyscript%40yahoogroups.com><anyscript%
> > 40yahoogroups.com><anyscript%
> > > > 40yahoogroups.com><anyscript%
> > > > > > 40yahoogroups.com>, “Prabhav
> > > > > > > > Saraswat”
> > > > > > > > <prabhav.saraswat@> wrote:
> > > > > > > > >
> > > > > > > > > Hello everyone,
> > > > > > > > > I am working on a foot model and I have been modeling
> > > > > > > > digotorum
> > > > > > > > > extensor and flexor muscle as 4 different muscles. I am
> > > > > > curious,
> > > > > > > > if there is
> > > > > > > > > a way to define a muscle with multiple tendons. I am
> > > > getting
> > > > > > > > results with
> > > > > > > > > only one of those 4 muscles taking up all the force. It
> > > > will be
> > > > > > > > better to
> > > > > > > > > have a single muscle with bifurcating tendons.
> > > > > > > > >
> > > > > > > > > I am aware that I can do it by defining an intermediate
> > > > segment
> > > > > > > > and
> > > > > > > > > using AnyKinPLine
> > > > > > > > > function, but a detailed explanation will be helpful.
> > And,
> > > > is
> > > > > > > > there any
> > > > > > > > > other better/easier method?
> > > > > > > > >
> > > > > > > > > thanks
> > > > > > > > >
> > > > > > > > > -prabhav.
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > [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]
[Non-text portions of this message have been removed]