simtk.org

I am currently trying to investigate how muscle forces are distributed throughout the lower limbs due to variations in ground reaction forces and also to changing muscle and joint parameters (such as those experienced after long periods of inactivity, say after achilles tendon problems).

I have tried to manipulate the muscle parameters in the Gait2354 and then run the static optimisation toolbox however the precessing takes around 20 hours and I am only observing changes in the muscles that I've edited rather than a redistribution of muscle forces in the leg which I would have expected to compensate for the change. Has anybody done anything similar before who could help out?

I was also thinking of manipulating the GRF CoP in the underlying C3D file for this example model and then observe the changes induced in muscle forces. Can anybody let me know the best way to go about this as I would have to alter many data points in the C3D file which I could not do one by one.

Many thanks.

Mark,

I am assuming that when you state that you are running the static optimization you are going all the way through the CMC. When you adjust the muscle and joint parameters are you only manipulating structural values (isometric force, lengths,...)? If so, your problem may lie in the adjustment of your actuator forces, with particular attention to the reserve forces. If you have the forces set too high, then the muscles in question may be compensating with higher activation levels, allowing the entire model to maintain the initial muscle values in the other limbs. Realize the other limbs may still require a greater activation to accomplish the compensation.

Adjusting these actuator levels to an optimal level may also help to speed up the processing time.

Thanks for your help Eric. I'm very new to this field and so was just wondering if you could clarify something.

So far I have indeed only been manipulating structural values, for instance reducing the max isometric force of gastrocnemius by 25%. In this scenario, how would I go about establishing what the other parameters should be in order for the model to run properly?