It is pointed out that in attempting to evaluate devices or design alterations to minimize accident injury, there arise important questions of true injury hazard predicted by the test and of relative merit between designs, depending upon whether one employs a system test or a simplified laboratory impact procedure. These questions are illustrated first by describing some of the results of a series of accelerator tests of cadaver impact against a steering wheel and energy absorbing column assembly. A salient finding from this work is that, as a result of more favorable load distribution, the chest loading is in the range of one-half that which would be indicated by a simplified torso impact test. It is felt that in the future it will be particularly important to try to take into account in a simplified test the contribution of the shoulders to load distribution, as well as to alter the torso form to obtain more realistic dynamic deflection properties.

A second example draws a comparison between severity of impact of the belted cadaver’s head against a front pillar and adjacent windshield area with the results of free-fall laboratory drop tests using a 15 pound headform. An appreciably less head impact severity is shown by the system test, from which it can be inferred that drop or pendulous type of simplified test yields results comparable to field accidents of higher nominal velocity.

It can also be inferred that when important discrepancies exist between the system and simplified test result, an option of using the former should be allowed.