simtk.org

In opensim 4.4, simulations were performed on the "arm26.osim" model using python.
The elbow joint angles were controlled by inputting excitation to the "BIClong" and "BICshort" muscles.
As shown in the figure, when a step input was applied to each muscle, the elbow joint angle suddenly increased sharply at an excitation value of 0.326 and showed a very strange behavior.
The joint angle at the time of the steep rise was approximately 145.4[deg].
What could be the cause of such a result?
In the attached figure, the first axis on the left is the excitation, the second axis on the right is the elbow joint angle value, and the horizontal axis is time.
The same value of excitation is given to both "BIClong" and "BICshort" as step inputs, and the other muscles (TRIlong, TRIlat, TRImed, and BRA) are given the minimum value of 0.1, which is always the minimum value of excitation as input, resulting in the value of activation to be constant at 0.
The python code used in the simulation is also attached.

Hi Yokoya,

By applying excitations with step inputs in the 'arm26.osim' muscles, you are essentially changing the stiffness of the system incrementally. To me, it looks like that you are transitioning between an overdamped system at the lower activation levels to an underdamped system (i.e., the oscillations) at the higher stiffness levels.

This is less "strange" behavior and more the dynamics of a spring-damper system, which simple models with few muscles can sometimes approximate.

Best,
Nick