Inverse Dynamics Without Torso

[Josée Mallah]

Hello,
I have just started using OpenSim with data other than that of the tutorials. I have lower body marker data, so I modified the gait2392 model, removing the torso and associated muscles. I ran the scale tool successfully, and then inverse kinematics, all fine (max error ~4.2cm). When I run the inverse dynamics, results are completely off. Large residuals are expected, but joint moments should not be affected by the absence of the torso.
Is it an issue with my force plate data or what am I doing wrong? I am attaching some files here (P.S. force plate 2 for right foot and 1 for left).
Thank you

[Mohammadreza Rezaie]

Hi, could you share the scaled model as well as the IK and ID output files?

[Josée Mallah]

Hi and thank you for your reply. For some reason I am unable to save the scaled model, so I am attaching the scale setup file, IK output file, and ID results as .xlsx (couldn't attach .sto).

[Mohammadreza Rezaie]

The ID results for the right leg from time 1.5 to the end look good and the patterns are reasonable. This shows that the ground reaction forces were assigned correctly to the right leg. Looking at the external load file you shared elsewhere, the point identifier of the force plate beneath the left leg has not been defined correctly. It must be:
<point_identifier>ground_force_1_p</point_identifier>
(_v --> _p)

Hope this helps.

[Josée Mallah]

Ah I didn't even notice, thank you so much! I am attaching a sheet with a comparison between OpenSim results and the results published by C. Fukuchi et al. using Visual 3D. I am assuming the results are not way far off, are they?
And now what tools can I run on this torso-less model? You previously mentioned that CMC doesn't work and I can use static optimization instead. Do RRA and Forward Dynamics work?