Hi,
I'm having a problem with Millard2012EquilibriumMuscle. When I compute the tendon force at activation 1 for e.g. the rectus femoris I get maximal tendon force of 120 N. When I change the muscles to Thelen2003Muscle, the maximal tendon force gets a maximal value of 920 N, which seems more reasonable.
I did the test with the gait2392_simbody model, with the normal.mot motion file of the tutorial. I did the computation both in the plotter of OpenSim and in the Matlab interface. Both gave the same result.
Does anyone knows what seems to be the problem here?
problem with Millard muscle
Re: problem with Millard muscle
Hi Kevin,
I did some digging and we figured out that its a known bug in the Millard model. Its been fixed on our development branch but it will still be seen in 3.2.
Essentially, the Millard muscle ignores the activation (from state) when you call equilibrate. Rather it sets the muscle activation to whatever the muscle models default activation is. If the default activation is 0.05, that's what get's used.
You can test this by changing the default activation of the rec fem to 1 and plotting the tendon force again.
If you wanted to to this in matlab, you would just update the default activation of the muscle once.
Hope that helps,
-james
I did some digging and we figured out that its a known bug in the Millard model. Its been fixed on our development branch but it will still be seen in 3.2.
Essentially, the Millard muscle ignores the activation (from state) when you call equilibrate. Rather it sets the muscle activation to whatever the muscle models default activation is. If the default activation is 0.05, that's what get's used.
You can test this by changing the default activation of the rec fem to 1 and plotting the tendon force again.
If you wanted to to this in matlab, you would just update the default activation of the muscle once.
Hope that helps,
-james