Issue with tendon compliance during CMC

[Ryan Byrne]

Hi all,

I am attempting to run CMC on a full body model with a lot of muscles (238 or so). I've noticed that in some example files from other users that at times the "ignore tendon compliance" box is checked for some pf the Millard muscles. When I check this for all of my muscles, CMC runs (although a bit slow) with moderate success. However, if I leave this unchecked for all muscles, CMC takes a long time to run and fails to find a solution to the initial conditions. It gives me the following suggestions, although I have checked and I'm pretty confident neither are the issue - especially since it works rather fine when ignoring tendon compliance.

CMC.computeControls: ERROR- Optimizer could not find a solution.
Unable to find a feasible solution at time = 0.1.
Model cannot generate the forces necessary to achieve the target acceleration.
Possible issues: 1. not all model degrees-of-freedom are actuated,
2. there are tracking tasks for locked coordinates, and/or
3. there are unnecessary control constraints on reserve/residual actuators.

Is anyone able to shed some light on why ignoring tendon compliance seems to be critical to running CMC here? Could it be that large passive muscle forces are making it difficult for the CMC tool to find a solution? The motion I am simulating is the lifting motion, where the starting position is the flexed position. Perhaps starting the simulation at the flexion pose could be an issue?

Thanks,
Ryan

[Dimitar Stanev]

The muscle forces are very sensitive to changes in the tendon parameters. The tendon parameters (e.g., tendon slack length) may need extra fine tuning in the range of motion that you are studying. More specifically, it is possible that some muscles don't operate in the region near the optimal fiber length, thus this may cause simulation instabilities.

https://simtk-confluence.stanford.edu/d ... cle+Models