Hi John
The AnyForce you have created will create a pulling force between the
two nodes when the AnyKinLinear measures a positive position and a
pushing force when it is negative, i belive. I am not sure if this
what you want. If you where using a AnyKinPline instead you would get
a pulling force for all nonzero positions of the two nodes.
If some of the support from the stilt is unilateral this is also
possible to model. If you look in the PedalDemoConditional example
you will find examples of such boundary conditions in the
JointAndDrivers.any file.
I am not sure i fully understand you last question, i willtry to
explain:
The AnyKinEqSimpleDriver have a property named Reaction.Type this is
by default switched On, so to get rid of the reactions one has to set
it to Off. The AnyReacForce can be used for creating forces between
objects without imposing any kinematics on them.
In your example with the two bodies free floating in space you have
created both a AnyKinEqSimpleDriver and a AnyReacForce between them,
you also write that the bodies are not linked to the ground. I do not
understand that this problem can be solved, if the bodies are
unsupport unless you have zero gravity. Anyway if the
AnyKinEqSimpleDriver has its reactions swithed on this could explain
why you do not see any effect of the AnyReacForce, in this case you
would have a double set of reations.
There is one possible drawback of using the AnyReacForce, if your
bodies moves wrt to each other you are adding energy into the system,
because then you have force and motion. This will also occur between
you feet and the stilts unless the connection is 100% rigid. I am not
sure if it will be a problem, but you can check the amount of motion
and force to get an impression on this.
I hope this helped clarify things
Best regards
Søren
— In anyscript@yahoogroups.com, “johnzengwu” <jwu@…> wrote:
>
> Søren:
>
> I have a further question, hopefully the last one, for this model.
The
> connection of the human knee to the stilt is not rigid in practical
> problem. I have linked the knee to the stilt using AnyForce and
> assumed a spring stiffness. I am not sure if it done correctly and
> whether the force is applied to the human leg or stilt in such a
> treatment.
>
> Another conceptual question about the interaction between two
bodies.
> Two bodies, A and B, are floating in sapce, none of them is linked
to
> the ground reference, the kinematic constrain between them is
> described using AnyKinEqSimpleDriver. The force interactions between
> these two bodies have to be described using AnyReactionForce. I have
> tested that, in this case, the Reaction.Type “On” or “Off” seems to
> have no effect. For example, the link between foot and stilt in my
> stilt model.
>
> “Environment.any” is uploaded for your reference. The remain files
in
> …v4 have not been changed.
>
> John
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
wrote:
> >
> > Hi John
> >
> > Thanks for the kind words.
> >
> > The footlength in the model is measured from the ankle joint to a
> > node named “ToeTip” on the big toe. If you need to display it you
can
> > find it in the file leg3d/seg.any and add a visible reference
frame
> > in it.
> >
> > The pelviswidth are the distance between the hipjoint centers.
> > When we find the time for it, we will add some graphics to
explain
> > these properties, sorry about this.
> >
> > I think this will make your model look more natural.
> >
> > Best regards
> > Søren
> >
> > — In anyscript@yahoogroups.com, “johnzengwu” <jwu@> wrote:
> > >
> > > Hi Søren:
> > >
> > > I always learn some new tricks by communicating with you. The
model
> > > seems almost working after your revision. Thanks for your help.
> > >
> > > Another trick I can play is to set the mass and mass moment of
> > enertia
> > > of the stils to zero, and add these to the human shanks. The
CoM of
> > > the shanks will also be modified. This would make the model
easier
> > to
> > > converge numerically.
> > >
> > > The mass of the stilt = 7.26 kg, is correct.
> > >
> > > Our subject: Foot length = 0.308 m; ASIS breadth = 0.262 m.
> > >
> > > I also wonder about the proportion of the human model by using
my
> > > parameters. Are these parameters the same as “FootLength” and
> > > “PelvisWidth” in your Gait3D model? It seems you have different
> > > definition for these parameters.
> > >
> > > John
> > >
> > >
> > >
> > >
> > > — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> > wrote:
> > > >
> > > > Hi John
> > > >
> > > > I have made some changes in the model and uploaded a new
version
> > > > named …v4.
> > > >
> > > > The changes are the following:
> > > > All reactions between the markers and the human has been
swithed
> > off,
> > > > if these are on the model “hooks” up to the global ref. (this
was
> > in
> > > > the environment.any file)
> > > >
> > > > The stilts needed to be carried by the humanmodel, so
assuming
> > that
> > > > there is a rigid connection between the foot and the stilt i
have
> > > > created a reaction force taking care of that. It locks 6 dof
> > between
> > > > the foot and the stilt. Furthermore i have created reactions
> > between
> > > > the stilt and the knee, i am not sure if these are coorect it
> > depends
> > > > on how the connection is made in practice. (these changes are
> > also in
> > > > enviroment.any)
> > > >
> > > > The mass of the stilts are set to 7.26 each, are they really
that
> > > > heavy?
> > > >
> > > > The pelvis and foot looked very large in my view, so i have
> > scaled
> > > > them down a bit (this was in AnyMan.any)
> > > >
> > > > The model now runs on the default kinematictol settings.
> > > >
> > > > The muscle activities are very large, i think you should
check
> > the
> > > > following:
> > > > Weight of the stilts?
> > > > Ground reactions especially the moment is this correctly
applied
> > > > Check it the accellerations of the segments are ok this is
the
> > rrdot
> > > > property
> > > >
> > > > You can try add AnyReacForces to the individual human joints
in
> > order
> > > > to detect which dof the model has difficulties in carrying?
> > > >
> > > >
> > > > I hope this helps you move on, it is a nice model!
> > > >
> > > > Best regards
> > > > Søren, AnyBody support
> > > >
> > > >
> > > > — In anyscript@yahoogroups.com, “johnzengwu” <jwu@> wrote:
> > > > >
> > > > > Hi Søren:
> > > > >
> > > > > I got the stilt-walking model running kinematically. Not
> > perfect, I
> > > > > have to set the KinematicTol to 0.05.
> > > > >
> > > > > I have tried to run inverse dynamic analysis. The model
works
> > only
> > > > by
> > > > > setting all Reaction.Type in drivers on the stils to “On”.
The
> > > > > analysis fails once the Reaction.Type’s in the stilt
drivers
> > are set
> > > > > back to “Off”. I tried many tricks and this state is the
best I
> > can
> > > > > get. Do you have any idea or suggestions to resolve the
> > problem? I
> > > > > have up-loaded my model as “Gait3D_w_Stilt_v3.zip” to let
you
> > check
> > > > it.
> > > > >
> > > > > John
> > > > >
> > > > > — In anyscript@yahoogroups.com, “johnzengwu” <jwu@> wrote:
> > > > > >
> > > > > > S�ren,
> > > > > >
> > > > > > Thanks for the great job. I can now work to adjust the
model.
> > In
> > > > fact,
> > > > > > the attachment locations of the stilts to the legs have
not
> > been
> > > > > > carefully adjusted. I will import the actual stilt scans
to
> > > > precisely
> > > > > > define the attachment locations. Once this is done, the
> > numerical
> > > > > > problems should be resolved.
> > > > > >
> > > > > > In reviewing your corrections, I found you mainly removed
the
> > > > driving
> > > > > > markers in the foot, using the driving markers of the
stilt
> > > > instead.
> > > > > > The stilts have been treated virtually as extended shanks
in
> > this
> > > > > > way. I think it’s a clever solution for the problem.
> > > > > >
> > > > > > John Wu
> > > > > >
> > > > > > — In anyscript@yahoogroups.com, “AnyBody Support”
> > <support@>
> > > > wrote:
> > > > > > >
> > > > > > > Hi John
> > > > > > >
> > > > > > > Interesting model !
> > > > > > >
> > > > > > > I have reviewed the joint and drivers and the
> > environment.any
> > > > files
> > > > > > > and made various changes. A new model named
> > > > Gait3D_w_StiltV2.zip has
> > > > > > > been uploaded. This model runs 34 steps before failing,
so
> > it
> > > > is not
> > > > > > > perfect yet, but you have something which is running to
> > start
> > > > out
> > > > > > > with.
> > > > > > >
> > > > > > > I have made numerous changes in the jointanddrivers.any
> > file
> > > > and in
> > > > > > > the environment.any so i can not list them here one by
one,
> > > > please
> > > > > > > compare with your old files if needed. One of the most
> > > > signifigant
> > > > > > > changes is that i have locked the ankle motion and
changed
> > the
> > > > point
> > > > > > > on the foot which is connected to the stilt from being
the
> > > > metatarsal
> > > > > > > marker to be the foot origo.
> > > > > > >
> > > > > > > I hope this helps you move on…
> > > > > > >
> > > > > > > Best regards
> > > > > > > S�ren, anyBody support
> > > > > > >
> > > > > > >
> > > > > > > — In anyscript@yahoogroups.com, “johnzengwu” <jwu@>
wrote:
> > > > > > > >
> > > > > > > > S�ren:
> > > > > > > >
> > > > > > > > Following your suggestions, I have adjusted the model
> > > > posture, and
> > > > > > > got
> > > > > > > > the model working user my own data.
> > > > > > > >
> > > > > > > > I am working further with my model, trying using
Gait3D
> > to
> > > > simulate
> > > > > > > > a problem “walking on stilt”. The stilt is attached
to
> > the
> > > > foot via
> > > > > > > > the constrains
> > > > > > > > in 3 displacemental directions and at knee in 2
> > horizontal
> > > > > > > directions.
> > > > > > > > The model seems logic, and posture is well adjusted.
But,
> > I
> > > > have
> > > > > > > > difficulty to set the model to the initial
conditions. Do
> > you
> > > > have
> > > > > > > any
> > > > > > > > suggestions to get the model working? I understand
this
> > might
> > > > not
> > > > > > > > be a basic question, would take you some time.
> > > > > > > >
> > > > > > > > I have up-loaded my model (Gait3D_w_Stilt.zip).
> > > > > > > >
> > > > > > > > John
> > > > > > > >
> > > > > > > > — In anyscript@yahoogroups.com, “johnzengwu” <jwu@>
> > wrote:
> > > > > > > > >
> > > > > > > > > — In anyscript@yahoogroups.com, “AnyBody Support”
> > > > <support@>
> > > > > > > wrote:
> > > > > > > > > S�ren:
> > > > > > > > >
> > > > > > > > > Thanks for the fast response and detailed
instruction.
> > I
> > > > have
> > > > > > > > > downloaded your files and it runs. I can now refine
the
> > > > model
> > > > > > > from here.
> > > > > > > > >
> > > > > > > > > John
> > > > > > > > >
> > > > > > > > > > Hi John
> > > > > > > > > >
> > > > > > > > > > I will try to guide you through some steps, these
are
> > the
> > > > steps
> > > > > > > i
> > > > > > > > > > followed trying to make the model run
kinematically.
> > > > > > > > > >
> > > > > > > > > > 1: Make sure the model has the correct segment
> > lengths,
> > > > this
> > > > > > > can be
> > > > > > > > > > changed in the file "AnyMan.any "
> > > > > > > > > >
> > > > > > > > > > 2:Check the locations of the markers, the blue
> > markers on
> > > > the
> > > > > > > model
> > > > > > > > > > should be located at the same location as the
real
> > ones
> > > > in the
> > > > > > > > > > experiment. You can change these locations in the
> > > > > > > > > >
> > > > file “BRep\Aalborg\ToolBox\Markersets\KitVaughanMarkers.any”
by
> > > > > > > > > > making a copy of it and put in inside your
gaitmodel
> > > > directory,
> > > > > > > in
> > > > > > > > > > this way the original is left unchanged.
> > > > > > > > > >
> > > > > > > > > > 3: In the Mannequin.any file try to alter the
joint
> > > > angles
> > > > > > > until the
> > > > > > > > > > load position is close to the location of the
grey
> > > > markers.
> > > > > > > This
> > > > > > > > > > makes it easier for the model to get a good
starting
> > > > quess.
> > > > > > > > > >
> > > > > > > > > > 4: Then next file to look into is
> > > > the “JointAndDriver.any”,
> > > > > > > this is
> > > > > > > > > > the file where the connection between the markers
are
> > > > made.
> > > > > > > Here i
> > > > > > > > > > started by adding joint drivers to the ankles
and
> > > > removing the
> > > > > > > > > > drivers for the metatarsals. This i did because
it
> > looked
> > > > like
> > > > > > > the
> > > > > > > > > > model had problems fullfilling the constraints in
the
> > > > ankle
> > > > > > > joint.
> > > > > > > > > > This made the model run, then i added back again
the
> > > > drivers
> > > > > > > for the
> > > > > > > > > > metatarsals in the local x dircection of the foot
> > > > (dorsal) and
> > > > > > > remove
> > > > > > > > > > the ankle joint driver for the flexion and the
model
> > > > would
> > > > > > > still run.
> > > > > > > > > >
> > > > > > > > > > 5: I have uploaded a model named Gait3DJohnWu.zip
to
> > the
> > > > file
> > > > > > > > > > section, please add in your own data. The model
runs
> > > > > > > kinematically
> > > > > > > > > > but there are for sure room for improvements. I
think
> > you
> > > > > > > should look
> > > > > > > > > > more into the location of the markers, and the
> > segment
> > > > lenghts.
> > > > > > > Once
> > > > > > > > > > they are ok then try remove the joint drivers in
the
> > > > ankle. It
> > > > > > > is an
> > > > > > > > > > iterative process that make take some iterations
> > before
> > > > it
> > > > > > > works out.
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Best regards
> > > > > > > > > > S�ren, AnyBody Support
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > — In anyscript@yahoogroups.com, “johnzengwu”
<jwu@>
> > > > wrote:
> > > > > > > > > > >
> > > > > > > > > > > Hi,
> > > > > > > > > > >
> > > > > > > > > > > I try to run Gait3D using our own motion data.
I
> > copied
> > > > the
> > > > > > > AnyBody
> > > > > > > > > > > Gait3D model, run it using your data, it works
> > fine. I
> > > > then
> > > > > > > replaced
> > > > > > > > > > > all your data using our own data, I can load
the
> > model,
> > > > but
> > > > > > > cannot
> > > > > > > > > > > even initiate it. The problem is the conflict
> > kinematic
> > > > > > > constrains,
> > > > > > > > > > > numerous error messages, e.g.:
> > > > > > > > > > >
> > > > > > > > > > > Set Initial Conditions: (Operation:
> > > > > > > > > > Main.Study.SetInitialConditions):
> > > > > > > > > > > 0) Design variables have been updated.
> > > > > > > > > > > 1) Load time positions have been re-established.
> > > > > > > > > > > Failed to resolve kinematic constraints (final
kin.
> > > > error =
> > > > > > > > > > 2.282212E-001)
> > > > > > > > > > > Constraint no. 90 above error tolerance
0.000001,
> > error
> > > > =
> > > > > > > > > > 0.000607. -
> > > > > > > > > > > constr. #0
> > > > > > > > > >
> > in ‘Main.Model.HumanModel.Right.Leg.Jnt.Ankle.Constraints’
> > > > > > > > > > > Constraint no. 91 above error tolerance
0.000001,
> > error
> > > > =
> > > > > > > > > > 0.027297. -
> > > > > > > > > > > constr. #1
> > > > > > > > > >
> > in ‘Main.Model.HumanModel.Right.Leg.Jnt.Ankle.Constraints’
> > > > > > > > > > > Constraint no. 92 above error tolerance
0.000001,
> > error
> > > > =
> > > > > > > > > > 0.006170. -
> > > > > > > > > > > constr. #2
> > > > > > > > > >
> > in ‘Main.Model.HumanModel.Right.Leg.Jnt.Ankle.Constraints’
> > > > > > > > > > > Constraint no. 93 above error tolerance
0.000001,
> > error
> > > > =
> > > > > > > > > > 0.054458. -
> > > > > > > > > > > constr. #3
> > > > > > > > > >
> > in ‘Main.Model.HumanModel.Right.Leg.Jnt.Ankle.Constraints’
> > > > > > > > > > > Constraint no. 94 above error tolerance
0.000001,
> > error
> > > > =
> > > > > > > > > > 0.000200. -
> > > > > > > > > > > constr. #0
> > > > > > >
in ‘Main.Model.HumanModel.Right.Leg.Jnt.Knee.Constraints’
> > > > > > > > > > > …
> > > > > > > > > > >
> > > > > > > > > > > I can send you my motion data let you try it.
> > > > > > > > > > >
> > > > > > > > > > > John Wu
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>