Hi Yoann
Ok thanks, this gives a better view of what you want to do.
I will try to explain how the .stl files works. The stl files works
only as a visual representation of the bones surfaces. In order to
make muscles wrap around them, various geometrical primitives such
as spheres, cylinders etc. have been fitted to match their surface
in the regions where the muscle wrapping occurs. When you scale the
model the bones and the geometrical primitives will scale together.
The .stl files do not have any effect on the kinematics, the
kinematics is controlled by the various joints which can be defined
in AnyBody and non of these involes the bone surfaces.
So it is not possible in AnyBody to create a joint which takes
contact between two bony surfaces defined as .stl files into
account.
I have an idea on how to make a joint between two spheres of
different radius, but i am not sure if it will be of any use for
this purpose. If the difference between the glenohumeral raidus and
the humerus radius is small it will not have any effect, then the
kinematics will resemble the spherical joint. The joint can be
contructed as a mechanism using the standard elements in AnyBody. It
will constrain only one dof, so the two dof sliding between the
spheres will have to be driven manually. In other words this means
that the model will not by itself find the the correct contact point
between the two spheres this will have to be driven, in reality the
shoulder would find this contact point giving the equilibrium by
itself.
The correct contact point between the two spheres will be the point
where the contact force is radial to the two spheres, without need
of any tangential forces. In practise this will only work for static
situations since you will have to make a parameter study to
determine the correct location of the contact point between the two
spheres that will remove the need of any tangential force.
Here is a short description on the mechanism i have in mind:
1 a segment is attached to the scapula by a spherical joint in a
point close the gh joint we use today. It has the length equal to
the size of the radius of the scapula part (glenosphere)
2 a segment is attached to the humerus using a spherical in a
location lying close to the GH center we use now. This segment has a
length equal to the humerus cap size.
3 kinematic constraints between the two new segments is setup at
their end points. These kinematic constraints must ensure that long
axes of the new segments is alligned, meaning that their normals are
identical, and that the end points are in contact.
4 artificial strong contact muscles between the end points need to
be defined that will allow only for compression forces, sideways
reactions needs also to be allowed
5 by moving the humerus sligthly the contact point between the two
spheres will move around and the location where there is no
tangential force needed between the two new segments is the correct
contact point.
It will require a fair amount of work to implement this and make it
work as intended, but in principle i think it is possible. If you
decide to do so i would strongly recommend to build the joint from
scratch initially in a seperate model and play around with it to
make sure it works as intended before replacing the GH joint with it.
Best regards
Søren AnyBody support
— In anyscript@yahoogroups.com, “Yo” <yoann.collet.1@e…> wrote:
>
> Hi Soeren,
>
> In order to simplify, you have to know i am working on a reverse
> prosthesis (delta) of the soulder.
> What i wanna do is to see if we are able to mesure the impact of
the
> size of glenoid or humeral part.
> I want to test the effect of several diameters on the kinematics
and
> force in the muscles.
>
> So if you don’t have a real surface that could not work.
> You said to me the .stl files are only visual cause muscles are
> supported by the bonny surface defined before but how can we do to
> change it?
>
> What are the fondamental elements of Anybody, ie the virtual
> mechanical elements making the bones “acting” like surface? I know
> they are not FE but i didn’t understand their nature. What are the
> contact between two bones, can we modify those contacts?
> Maybe we will find a way to make something better.
>
> Tank you very much for your patience
>
> Yoann
>
>
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@a…>
wrote:
> >
> > Hi Yoann
> >
> > You are right about the sliding, the model will not predict any
> > slide, the motion is driven.
> >
> > It is also right that the sideways reactions will tell
> > which direction the slide would occur if possible.
> >
> > Best regards
> > Soeren AnyBody Support
> >
> >
> > — In anyscript@yahoogroups.com, “Yo” <yoann.collet.1@e…>
wrote:
> > >
> > > Hi Soeren and thank you for the precisions,
> > >
> > > Still some little things :
> > >
> > > > But sliding should be possible,
> > > > if you drive the sliding motion.
> > >
> > > Ok, but the aim is to see if there is sliding, when only
driving an
> > > abduction. It’s not possible if i well understood, is it?
> > > I mean, if you don’t drive a slide of the humeral head on the
> > glenoid,
> > > the centres of rotation will stay the same for the both. Am i
> > right?
> > >
> > > > Another option could be to switch off the side way reactions
in
> > the
> > > > GH joint. If you do so the muscles will have to carry the
> > sideways
> > > > forces. The joint will not dislocate but it will not be able
to
> > take
> > > > up sideways forces.
> > >
> > > If we can’t see a passive displacement of the humeral head, the
> > > sideways reactions would give directions where the head would
go if
> > > the model was able to let it. So we won’t switch off sideways
> > > reactions and maybe will keep this solution in mind.
> > >
> > > Thank you
> > >
> > > Yoann
> > >
> >
>