ID: double support phase and RRA: large mass adjustments

[Stefan van Drongelen]

Hello,

I'm starting to use OpenSim for my research. I started to simulate normal gait of a healthy person. My scaling and IK are fine, although it can be improved.
My question however concerns the ID as I wanted to compare the results to PlugIn Gait. The output shows noise/interference during the double support phase. Is this always the case or how can I change my setup/ work around it to minimize this effect?

My second question concerns the RRA: The recommended mass adjustment is pretty large: 14 kg for an 88 kg man.
I suspect that this is just a setting issue, so I hope somebody knows a solution.

Thanks
Stefan

[jimmy d]

Hi Stefan,

Unfortunately it sounds as if the scaling and/or IK solution is not very good. ID is not expected to have any noise/interference during double support, if all is correct. The examples provided with OpenSim have double support periods and show expected results.

This noise is being seen in RRA. RRA is a forward dynamics simulation that resolves the F=ma equations so that the mass and accelerations match the forces. If you are having to make large adjustments its an indicator of large errors in your model and/or IK.

Let us know how it goes,
-james

[Brian Umberger]

Hi Stefan,

In addition to checking the quality of the scaling and IK results as James suggested, how do your center of pressure data look? The center of pressure can sometimes be a bit noisy around heel strike and toe off (i.e., at the beginning and end of double support). We have seen noisy joint moments during double support that were resolved by cleaning up the center of pressure data. Hope this helps.

Brian

[Stefan van Drongelen]

Dear James, dear Brian,

I have looked into my scaling and my IK with someone using OpenSim for a longer time. I also checked the GRF and the COP. Nothing supsicious was found. With cleaning and filtering I can resolve part of the peaks during double support but not all of it.
I happen to notice that also in the *.mot files which come with the OpenSim models in the installation this phenomenon is present.
Can this be real then? I mean the legs are not completely independent from one other as is in Vicon PluginGait. So the GRF of the one leg during double support will have an influence on the other.

Stefan