simtk.org

Hi,

I'm having some issues with CMC running at footstrike times. I have a feeling that this could be due to heelstrike transient effects and have tried several GRF filtering freqencies. The CMC is failing at exactly the same time step (a few m/s before foot strike) despite me checking the following:

1) Model is strong enough (this was checked by examining a simulation that ran to just before foot strike) - reserve actuators were made "sufficiently high" and there are no signs of actuator saturation.

2) Kinematics are tracking well. No divergance there.

I've tried changing start times and also the cmc_time_window and have checked for NaNs in the inputs but this is all OK.

The only constant pattern is that the failure occurs just before the foot strike periods (there are multiple of them and they all fail about that local region). CMC will work fine between these footstrike periods but not near them.

Does anyone have thoughts that they could share?

Thanks!

Hi,

I got same problem when I drive CMC to simulate unilateral landing task and never solved yet. I doubled all of isometric force of gait2392 as previous study and RRA and SO worked correctly . CMC doesn't work around the foot strike, while the other part correctly simulated. I think I should modify control constraints file or actuator file, but I couldn't find good way. Input kinematics and GRF data ware filtered at 12Hz.

I hope someone have solved the problem already and please let me know.

Thank you,

Ryo