simtk.org

Dear all,

I'm using the lowerlimb 2010 model and I'm trying to perform RRA in Opensim version 2.2.1
I've successfully completed IK and ID and the results of these look good. However when I perform RRA there are high peaks in the residual forces and moments (up to 500 N). When I look at the kinematics of the RRA results (as far as it runs) the left arm moves in a rotating way and backwards, while it doesn't do this in the IK and ID results. The peaks in the residuals occur at similar time points as when the arms moves strangely.

We tried running RRA in an older version of Opensim (v 1.9.1) and the results look fine, without peaks in the residual forces here and normal arm movement as well. Unfortunately I'm running windows 7 on my computer, so going back to version 1.9.1 is not an option. I did try running with the same settings (adjusted for version 2.2.1) in version 2.2.1, but no luck. The peaks and strange arm movement were still there.

Has anyone encountered similar problems? Or does anyone have any suggestions on how to solve this problem?

Thanks,
best wishes,

Paulien

I am having a similar problem as well. When I run RRA on my model in v2.2.1, my residual forces peak and RRA crashes for this particular model at the same time point each time I run RRA. I had downloaded the latest beta version of Opensim just to try it out and have been able to successfully run the same model through RRA. This has me confused as to what the differences are between v2.2.1 and the beta version. How RRA can run for the same model in one version of Opensim and not the other?

Is there any information available as to the differences between v2.2.1 and the latest beta version? Is the beta version designed to be more robust than 2.2.1?

Thanks,
Mike

Hi Michael & Paulien,

msamaan wrote: ...
This has me confused as to what the differences are between v2.2.1 and the beta version. How RRA can run for the same model in one version of Opensim and not the other?

Is there any information available as to the differences between v2.2.1 and the latest beta version? Is the beta version designed to be more robust than 2.2.1?

We do our best to make every release better than the previous one (by fixing bugs, cleaning up code, reducing unnecessary parameters and improving algorithms). The transition from 1.9.1 to 2.0+ stream involved a significant rewrite to make the computational layer more robust, use more solid math tools and state handling from SimTK, since then the changes are less broad but since the code has more mileage on it, and many improvements we expect it to be more robust. I'd recommend you sign up for the beta testers mailing list so you get notified when new release candidates are available and give us feedback in case we break something (we have a growing battery of tests that we run on regular basis and before making a release but the coverage is not as broad as we'd like it to be yet).

Every posted release has associated release notes that include the new functionality, bug-fixes and documentation updates, so please consult those or let me know if you can't find them and we'll try to make these more accessible.

Paulien, it would be great if you can try 2.4 beta and let us know so we can fix any issues before we publish the 2.4 release. If you do have an old model file, you can still run version 1.9.1 by modifying the file opensim.conf to use the (Window 7 compatible) JDK included with 2.2+.

Good luck and please let me know if you have any questions,
-Ayman