NOTES.txt

Rajagopal et al. 2016. Tendons were modeled as rigid when the tendon slack length was less than the muscle optimal fiber length.

In rigid tendons, there is no fiber_length variable in state.
In compliant tendons, force-velocity relationship, i.e., muscle dynamics would be ignored.

In Millard2012EquilibriumMuscle:
ignoreTendonCompliance     computeEquilibrium     ForceVelocityMultiplier
        False                     False                     0
        False                     True                      1*
        True                      False                     values
        True                      True                      values
* current implementation

PCSA = max isometric force / 60 (specific tension used by Rajagopal et al. (2016) (N/cm^2))
muscle volume = muscle length * PCSA
muscle stress = muscle force / PCSA
contractile element = activity * max isometric force * active force-length multiplier * force-velocity multiplier
parallel elasic element = max isometric force * passive force-length multiplier
fiber force along tendon = (contractile element + parallel elasic element) * cos pennation angle

total muscle tendon length (optimal fiber length plus tendon slack length) van der Krogt 2016

Equality constraint means that the constraint function result is to be zero 
Inequality constraint means that it is to be non-negative.

Maximum muscle recruitment could be achieved by minmax criteria (Rasmussen2001). However, discontinuities may occur which is not appealing physiologically.
Currently, I'm working on this issue and the newest idea is muscle moment-weighted cost function. But, it has its own downsides.