Driving a model with constant angular velocity

I have uploaded two files to the files section of the group -
Validation I - cst arm angles, and Validation II - driven by Vicon. I
am having problems with Validation I - cst arm angles. Like I said
earlier, when I drive the model by the elbow AND shoulder
simultaneously, the model stops running at step 81. If I drive the
model with the elbow OR shoulder angle, then the model compiles and
runs completely. I can however, get the model running with the
MinMaxOOSolQP, however for a simple model I’m not sure why I would
need that solver. Ultimately, I am hoping that once compiled and run,
both models of mine will produce the same results (or nearly the same
results). As of now however, that is not the case.

Any help would be appreciated.

Thanks!
Sarah

Hello Sarah,

I apologize for the belated reply.

> I have uploaded two files to the files section of the group -
> Validation I - cst arm angles, and Validation II -
> driven by Vicon. I
> am having problems with Validation I - cst arm angles. Like I said
> earlier, when I drive the model by the elbow AND shoulder
> simultaneously, the model stops running at step 81.

You are quite right - I also get that error. However, when I use your
model together with the next version of the repository, which we are
working intensively on at the moment, then I do not get the error. I
wish I could say precisely what is the difference, but we have made
so many changes that it is quite impossible to determine which one
makes the difference.

In any case, the easy fix is to update you to the new release of the
BRep part of the repository. I’ll clean a temporary version and uload
it to the group.

> If I drive the
> model with the elbow OR shoulder angle, then the model compiles and
> runs completely. I can however, get the model running with the
> MinMaxOOSolQP, however for a simple model I’m not sure why I would
> need that solver.

The MinMaxOOSolQP should give you exactly the same result as any of
the other solvers if you do not use any penalization parameters.
However, I understand that this model is for validation, and in that
case it could be most useful to compare the results of the alorithms.
Please have a look at the muscle recruitment tutorial:
http://www.anybodytech.com/514.0.html
for an n-depth explanation of how to switch between the different
recruitment formulations.

> Ultimately, I am hoping that once compiled and run,
> both models of mine will produce the same results
> (or nearly the same

If he movement and load is the same, then the results should also be
the same. I was not able to load the Validation II model, however,
because of a syntax error in one of the joint definitions.

Best regards,
John

Hi Sarah,

I have now uploaded a very temporary version of the new repository,
Repository.6.5.zip.

The ARep part now only contains your two models. Please check that
Validation I can analyse. Validation II cannot load due to an error in
a joint definition, but you are probably better at fixing that than I
am.

Let me know if you want to discuss the validation issue. Notice that
Bergmann’s new shoulder reaction force data are now published as an in-
press paper in Journal of Biomechanics homepage. It would be great to
validate the model against these data.

Best regards,
John