Page 297 - Biomedical Engineering and Design Handbook Volume 1, Fundamentals

274 BIOMECHANICS OF THE HUMAN BODY

0 ms 500 ms 600 ms
FIGURE 11.9 Articulated total body (ATB) model prediction for the response of an unrestrained
standing child to panic braking in an automobile at 500 and 600 ms after commencing braking.
(von Gierke, 1997.)

dashboard), seat belts, and air bags. The models may include advanced harness systems and wind
forces to model seat ejection from aircraft and can also describe aircraft crash. An example of
the use of these models to predict human response in crashlike situations is shown in Fig. 11.9
(von Gierke, 1997). In this diagram, the motion of an unrestrained child standing on the front seat
of an automobile is shown in response to panic braking. The output of the model has been calculated
at the onset of braking and after 500 and 600 ms. It can be seen from the diagram that the model pre-
dicts the child will slide down and forward along the seat until its head impacts the dashboard of the
vehicle.
Finite-Element (FE) Models. The detailed response of selected parts of the body and the
environment being simulated (e.g., head and neck, spine, and vehicle seating) have been modeled by
using finite elements (FEs). In the MADYMO model, the FEs can interact with the multibody model
elements. Examples of human body subsystems that have been modeled with FEs include the spine
(Seidel, 2001) to predict the injury potential of vertebral compression and torsional loads (Seidel, 2005),
and the head and neck to predict rotation of the head and neck loads during rapid horizontal
deceleration (see Fig. 11.12) (RTO-MP-22, 1999).
Artificial Neural Network Models. The response of the spine to vertical accelerations has also
been modeled by an artificial neural network (Nicol et al., 1997), which permits the magnitude-
dependent (i.e., nonlinear) characteristics of the human body to be included (see, for example, panels
2, 4, 6, and 8 of Fig. 11.4). The parameters of the network were established by “training” the model
using repeated shocks, which were applied to a seated person with back unsupported in the head-
2
ward direction, with peak amplitudes from 10 to 40 m/s . Since a neural network model has no
predictive power beyond its “training,” the applicability of the model is restricted to shocks and
impacts in this direction and with this range of accelerations.

11.3.2 Anthropometric Manikins
Mechanically constructed manikins, or dummies, are used extensively in motor vehicle crash testing
and for evaluating aircraft escape systems and seating. Several have been developed for these
purposes (Mertz, 2002a; AGARD AR-330, 1997), some of which are commercially available.

Hybid III Manikin. The Hybrid III manikin, shown in Fig. 11.10, was originally developed by
General Motors for motor vehicle crash testing, and has since become the de facto standard for sim-
ulating the response of motor vehicle occupants to frontal collisions and for tests of occupant safety
restraint systems. The manikin approximates the size, shape, and mass of the 50th percentile North
American adult male, and consists of metal parts to provide structural strength and define the over-
all geometry. This “skeleton” is covered with foam and an external vinyl skin to produce the desired
shape. The manikin possesses a rubber lumbar spine, curved to mimic a sitting posture. The head,
neck, chest, and leg responses are designed to simulate the following human responses during rapid