Strength and Fatigue: Experiments and Modeling - Human Motion Simulation: Predictive Dynamics - page 139

Figure 6.10 demonstrates the concept of comparing torque values calculated
from predictive dynamics with those measured experimentally using a 2D torque-
velocity curve for simplicity. The white curve depicts torque values over time for
a single joint relative to expected torque-velocity percentile strength curves. Note
that for this example, the task requires from

25% of maximum to 75% of

, .

maximum peak torque, which varies across time.
Once strength models have been created, there are essentially two methods for
applying torque models in DHM:

a. Pre-processing approach: Embedding the torque/strength models as limits or
constraints in the optimization algorithm which yields a motion that is
constrained to the available peak strength, i.e., a motion that can be
accomplished by the digital human if a solution exists. The benefit of this
method is that it simulates human motion more realistically as strength limits
(e.g., 50 th %ile) cannot be violated to accomplish the motion, if the task can
be accomplished. However, the disadvantages include greater computation
cost (i.e., a nonlinear constraint that varies as a function of the parameters
being optimized) and it does not allow the DHM to model where a task could
exceed the strength surface for any given strength percentile, which may be
useful for predicting risks for musculoskeletal injury.
b. Post-processing approach: Comparing the results of the predicted DHM
torques resulting from the optimization algorithm using only simplified
strength limits (e.g., a constant average value) to the 3D strength surfaces
without using the surface models to limit or constrain the DHM. This method
is computationally simpler and provides insight into what joint has exerted a
torque that may exceed expected normative percentile levels. However, it can
also yield motions that would otherwise not be possible, particularly at faster
velocities or end-ranges of motion where the simple strength limit is most
likely to be in error relative to the 3D strength models.

6.7 Fatigue
Localized muscle fatigue can be briefly defined as the loss of force-producing
capability following muscle activity ( Bigland-Ritchie and Woods, 1984 ). Thus,
fatigue is temporary, recovers following rest, and is distinct from weakness,
pathology or traumatic injury. Practically, the development of fatigue is important
for dynamic DHM because the temporal decay in muscle strength is a very com-
mon and real phenomenon. Clearly, as we perform tasks either for greater lengths
of time or of higher intensity, we typically develop greater fatigue.
A curvilinear relationship between fatigue and intensity has been well documented
for over 50 years, often referred to as Rhomert's curve or the intensity-endurance time
(ET) curve ( Rohmert, 1960 ). Numerous authors have proposed various versions of
intensity-ET curves, including several joint-specific models ( El ahrache et al., 2006;