Researchers all over the world have ducted over 100 simulated car crashes in which eadavers have been restrained by ous types of safety belt systems. The large investment of time, effort, and funds associ~ ated with these research investigations warrants an analysis of the resulting data from a gìobal viewpoint to identify injury trends and possible design criteria for near-term application. Such an attempt has been made and is discussed.

Large variations in test instrumentation, test conditions, and belt system configura tions utilized by the various researchers limited the analysis to those measured engineering parameters and observed injury evaluations that were common to all the experiments.

This very minimal set of event descriptors, when subjected to an analysis, has shown injury (defined in this context to be the number of observed thoracic fractures) to be a statistically significant function of the maximum upper torso belt force, cadaver weight, and cadaver age at death. Implications and consequences of such results are discussed.