Scaling subject mass on non-full body models

[Will Flanagan]

I just wanted to get some clarification on scaling a subject's mass when we have a non-full body model (like gait_2392 where there are no arms).

The model documentation page (https://opensimconfluence.atlassian.net ... 354+Models) says that this model represents a subject with 75kg mass, so I'm wondering if the mass of the arms was lumped into the torso? This is because in terms of scaling, the scaling documentation (https://opensimconfluence.atlassian.net ... Properties) says that if input target mass > 0, the scaled model will have a mass equal to the target mass. So if I have a subject with actual mass of 65kg, my new scaled model (without arms) will be 65kg, which if the arms weren't lumped into the torso would result in the model segments having higher mass than the subject's actual segments (since all the segments will be "sharing" the total mass of the arms). This would affect my inverse dynamics results, so it would be great to better understand how to account for mass of segments that the subject has but the model doesn't.

Thanks,
Will Flanagan, UCLA

[Nicholas Bianco]

Hi Will,

I don't know for sure gait2392 lumped the arm masses into the torso, but I would assume that to be the case.

So if I have a subject with actual mass of 65kg, my new scaled model (without arms) will be 65kg, which if the arms weren't lumped into the torso would result in the model segments having higher mass than the subject's actual segments (since all the segments will be "sharing" the total mass of the arms).

This is correct, if the arms weren't lumped into the torso. The scaling tool will distribute mass throughout the skeleton based on the ratio of original body masses to the total body mass. So if the original model torso mass seems correct to you, then everything is fine, but you can also change it to a new value if you think it is too heavy/light.

Hope that helps,
Nick