Millard muscle passive torque computation

[Francesca Dell''Eva]

Dear everyone,

I am working on the gait10dof18 model which uses Millard muscles. I defined a gait trajectory and my aim is to compute the torque needed at a given coordinate (knee_angle_r, for example) in order to track this movement.

To obtain this, I compute the inverse dynamics using the InverseDynamicsTool with Matlab and I consider only the column of the resulting matrix relative to the right knee.
Then, to compute the passive torques applied by the muscles at the right knee joint, I use a Matlab function which simply multiplies the passiveFiberForceAlongTendon (function computePassiveFiberForceAlongTendon()) and MomentArm (function computeMomentArm()) of the muscle with respect to the knee joint.

The problem is that I get reasonable torques for ID while the passive torques I get are really small (order of 10^-7 or even smaller).
If I try the same process using the gait2354 model which uses Thelen muscles, I don't have this problem.

Does someone have any idea of which could be the problem? Do you think it is related to the fact that my model uses Millard muscles? Because I checked and it seems that the same functions are available for both muscle types so I do not understand which could be the problem...

Thank you!

[Thomas Uchida]

It sounds like you have performed analogous calculations on two models and have obtained different answers. This will be true in general. To investigate further, you could compare the lengths of the musculotendon actuator paths (which will not depend on the muscle model). Please also be aware of the limitations of each model, some of which are listed on the "Musculoskeletal Models" page in the documentation: https://simtk-confluence.stanford.edu:8 ... tal+Models.