Description of the Millard2012AccelerationMuscle

[Yannik Schröder]

Hello everybody,
i am currently in the progress of writing my masters thesis and have been using the Millard2012Acceleration model for some CMC- and forward dynamics calculations. So i'm currently looking for some information on how exactly the muslce model works and what the differences to other models (say the Thelen2003 model) are, so i can give a quick overview of the muscle models in my thesis.

While i found some informations of the Thelen model (and the other two Millard models), information of the MillardAccelerationModel seems sparse. All i found was the description in the doxygen (https://simtk.org/api_docs/opensim/api_ ... ml#details), but i didn't exactly understand how the model works and calculates the resulting force in dependence on the activation, fiber length and fiber velocity.

Further i don't understand, which mass exactly is mentioned in the doxygen-description. Is it the mass of the bodies the muscle needs to move? Or am i using a specific mass of a muscle?

Lastly i don't understand why the Thelen2003 model is a two state model, since a force-velocity function is also taken into account? Or am i misinterpreting the term "two state model"?

I hope someone can shed some light on the muscle model theory.

Thanks in advance!

Best Regards,
Yannik

[Brian Umberger]

Hi Yannik,

While I can't help with a detailed description of how the Millard2012Acceleration muscle model works, I am pretty sure that the mass referred to is the muscle mass, not the mass of the segment(s) that the muscle acts upon. I'm sure other people of the forum can confirm whether that is correct.

Regarding your other question, the Thelen2003 muscle model is referred to as a "two-state" model because its behavior is governed by two, first-order differential equations. The two state variables are activation and fiber length. The time derivatives of the state variables are: 1) the rate of change of activation, which comes from the activation dynamics equation, and 2) fiber velocity, which comes from the force-velocity equation.

I hope this helps.

Cheers,
Brian

[Jennifer Hicks]

Yannik -

Our group recently published a paper documenting the new muscle models. You can find it along with some other muscle model information here:
http://simtk-confluence.stanford.edu:80 ... blications

Jen

[Yannik Schröder]

Thanks for the replies so far!

But as far as i can see only the Millard2012EquilibriumMuscle model and some of its variations are described in http://simtk-confluence.stanford.edu:80 ... cle+Models

The Millard acceleration model seems to use a different approach than the classic Hill type muscle models which i don't quite understand. Is there a name for the kind of muscle model the Millard2012Acceleration model is based on?

Regards
Yannik

[Fabian Bayerlein]

Hey Yannik,

Matt Millard also held a Webinar about his new Muscle models:
http://opensim.stanford.edu/support/eve ... html?id=47

The Information is very detailed and structured and I'm sure it would further your understanding.

Best Regards,
Fabian