simtk.org

Hello,

I was hoping to find some information about the effect of locking coordinates on the outcome of static optimization. Specifically, does locking a coordinate exclude it from the optimization solution entirely?

I have observed that the activation of actuators crossing locked coordinates go to a minimum, and there are less constraint violations reported. However, the activation of the muscles of interest (for unlocked coordinates) is similar to when all the coordinates are unlocked.

I just want to try to understand this process better, and to know if I am right to lock the coordinates as I have.

Any help would be greatly appreciated.

Thanks,

Alex

Hi Alex,
Static optimization works by calculating the muscle forces required to generate the torque about a joint. When you lock a joint, you are adding a lock constraint that provides the generalized force (torque) for that joint. The muscle activation behavior you are seeing would be then expected as muscles that cross that joint no-longer have a torque to account for.

Torques across 'non-locked' joints haven't changed, nor has its need to be satisfied. So the muscles crossing them would see almost no change.

Hope that helps,
-james

Hi James,

Thanks for your help, that explains it nicely.

Alex