A three-year study of 722 freeway ramps carrying over 2- billion vehicles on which 1643 accidents occurred was undertaken to investigate the geometric features which play important roles in ramp safety and to classify these geometric features by ramp type and relative safety measures the study included classification of ramp type and presented the results of correlations between accident rates and various geometric and non-geometric considerations. A basic assumption made was that ramp experience is essentially related to the number of vehicles using the ramp without regard to the distance travelled. The 722 study ramps, of which 32 percent were accident- free, experienced 18 percent of the accidents occurring on the freeways which they served. Generally, it was concluded that design differences of the ramps studied did not play a significant role when comparisons were made of accident- free ramps and those ramps with accidents. It was found, however, that the off-ramp accident rates were consistently higher than the on-ramp accident rates, the differences being attributed primarily to the difference in number of single-vehicle accidents occurring on the ramps. It was also considered that acceleration lanes of 800 feet or greater and deceleration lanes of 900 feet or greater result in lower accident rates than shorter speed-change lanes.
Abstract