The development of crash research techniques at the general motors proving ground is discussed. A ground-level roll-over test has been developed which gives more reproducible results and is more realistic than earlier roll -over tests using ramps. barrier impact tests and impacts with a tree were studied. Deceleration levels in these collisions are discussed. The safety of various roadside designs was investigated in simulated run-off roadway accidents. The distance by which a car left the road was investigated for 56 proving ground accidents. Elimination of roadside obstacles appears to offer impressive potential for reducing serious accidents. Guardrail accidents were also studied. Major improvements are possible in guardrail design and installation, and inexpensive modifications would eliminate the most serious guardrail deficiencies. In automotive engineering development work, the interests are generally in one component or in a few components at a time. Tools for such development work have been developed. A heavy duty snubber and an indoor acceleration device are used to produce the decelerations necessary to simulate collision decelerations economically and precisely. Other tests were developed for evaluating steering wheels, instrument panel pads and seat backs. An appendix contains data from illinois on median encroachments on a divided highway. Angle of departure, lateral distance of encroachment and the length of travel in the median are presented. The data suggest that a roadside free of obstacles for 33 feet would provide safety for at least 80 percent of the drivers leaving the road. In more than 10 percent of the cases studied the vehicle crossed a 40 foot median into the opposing traffic lane.
Abstract