BIOCODYN

Other physical models have been evaluated or developed over the last 15 years to study injury risks in this context (mainly for the case of the rear effects of body armor), but all have crippling drawbacks. The dummies developed for automotive accidentology (Hybrid III type) are not valid in the deformation speed regime under consideration. Systems based on deformable membranes (DSTL, Biokinetics) are valid only for the prediction of pulmonary contusions or other intra-thoracic lesions in the very high energy and speed regime (sniper fire on rigid-plate vests); moreover, membrane technology is not yet perfected, and the issues of wear, repeatability and reproducibility have not been mastered. Artificial torsos" made up of artificial rib cages and more or less complex assemblies of synthetic organs (Kneubuehl, AUSMAN, JAYCOR, John Hopkins HSTM...) are representations whose anthropomorphic appearance is misleading because their biofidelity is not established and necessarily remains limited due to the materials used; on the other hand these models pose complex instrumentation problems, and consequently the validity of the measurements made is generally questionable and their interpretation is very delicate due to the complexity of the systems.

Numerical simulation is a complementary approach to testing on physical (or biological) models. This avenue has been explored for some twenty years, during which time various numerical models of the human body have been developed. While some models have now reached an acceptable level of maturity for low-velocity impacts (notably the head), the case of ballistic impacts still faces major difficulties in determining the laws of behavior of organic materials in large deformations at high speeds, and in managing the complexity of the phenomena to be described in numerical computation; moreover, these developments are too rarely validated by comparison with sufficiently characterized experimental models.

In the absence of a physiologically-based model for the interpretation of the "dynamic cone" test, specifications are currently based on first-level observables (derived from lesion criteria used notably in automobile accidentology) and on calibrations obtained by means of tests on known equipment (bullet-proof vests or ALR projectiles), and as a result the results remain essentially comparative.

Improving test interpretation would require an evolution towards criteria based on biomechanical analysis, and specific to the different types of possible injury.