Correlation Between Muscle Equations and Program Behavior
The computer program implements a Hill-type muscle model to simulate knee joint
torque generation by directly translating the analytical equations into numerical
procedures. Each block in the force generator diagram corresponds to a mathematical
relationship that governs how neural input, muscle state, and joint motion interact
dynamically.
The simulation begins with the neural excitation signal 𝑢(𝑡), which is processed by the
activation dynamics block to produce muscle activation 𝑎(𝑡). This step models the
physiological delay between neural input and muscle contraction. In the program, this is
implemented as a first-order di erential equation that gradually drives activation toward
the input value rather than changing instantaneously. As a result, when 𝑢(𝑡)increases
abruptly, the program produces a smooth rise in activation, reflecting realistic muscle
behavior.
Once activation is computed, muscle force is calculated using the multiplicative
relationship:
𝐹 (𝑡) = 𝑎(𝑡) 𝐹 𝑓(𝑙) 𝑓(𝑣)
This equation is the core of the program’s force generation logic. The program evaluates
each term sequentially, meaning that any change in muscle length or contraction velocity
directly scales the output force.
Force–Length Relationship and Program Response
The force–length relationship is implemented using the normalized quadratic equation:
𝑙(𝜃) − 𝑙
𝑓(𝑙) = 1 −
0.5 𝑙
In the program, joint angle 𝜃obtained from skeletal dynamics is converted into muscle
length 𝑙(𝜃). The software then evaluates 𝑓(𝑙)within the prescribed range 0.5 𝑙 ≤𝑙≤
1.5 𝑙 . This produces the bell-shaped curve observed in the force–length plot.
Programmatically, this means that:
When the knee angle corresponds to the optimal muscle length 𝑙 , the
computed force reaches its maximum.
When the joint moves toward excessive flexion or extension, the force term
decreases automatically.
If the joint moves outside the allowed length range, force generation drops sharply,
preventing non-physiological outputs.
Thus, the program’s reaction to joint angle changes is a direct numerical manifestation of
the force–length equation.
Force–Velocity Relationship and Program Response
The force–velocity behavior is governed by the piecewise function:
𝑣 −𝑣
, 𝑣≤0
𝑣 + 2.5𝑣
𝑓(𝑣) = 𝑣 − 𝑐𝑣
1.3 − 0.3 , 𝑣>0
1+𝑐 𝑣
Here, contraction velocity 𝑣is obtained by di erentiating joint angle 𝜃̇. In the program,
this value is continuously updated and fed into the force–velocity function.
This causes the following observable behavior:
During concentric contraction (𝑣 > 0), the program reduces force output as
velocity increases.
During eccentric contraction (𝑣 < 0), force output increases beyond the isometric
level.
The asymmetry in the plotted force–velocity curve emerges naturally from the
conditional structure of the equation implemented in code.
As a result, the program responds di erently to slow versus fast movements, even when
activation remains constant.
Torque Generation and Skeletal Motion
The muscle force computed by the program is converted into joint torque using:
𝜏 =𝐹 ⋅𝑟
This torque is summed with damping and elastic torques:
𝜏 =𝜏 +𝜏 +𝜏
Numerically, the program evaluates this sum at every simulation step and feeds it into
the skeletal dynamics equation:
𝑚𝑔𝑙
𝜏 − 2 sin 𝜃
𝜃̈ =
𝑚𝑙
4 +𝐼
The resulting angular acceleration is integrated to obtain angular velocity and joint angle.
This explains why changes in force–length or force–velocity parameters immediately alter
the knee motion trajectory in the simulation.
Closed-Loop Behavior in the Program
The most important correlation between the equations and program behavior is the
closed-loop interaction. Joint angle 𝜃a ects muscle length 𝑙, joint velocity 𝜃̇a ects
contraction velocity 𝑣, and both variables feed back into the muscle force equation. This
force then generates torque that drives joint motion again.
Consequently, the program does not merely evaluate equations independently; it
continuously couples them. Any modification to parameters such as 𝑙 , 𝑣 , or the
shape factor 𝑐immediately reshapes the force curves and alters joint kinematics. This
tight coupling is why the simulation exhibits realistic nonlinear responses rather than
simple linear motion.