THE CORRELATION OF ACCIDENT RATES WITH GEOMETRIC DESIGN COMPONENT OF VARIOUS TYPES OF HIGHWAYS

THE RELATION OF DESIGN CHARACTERISTICS OF HIGHWAYS WITHOUT ANY ACCESS CONTROL TO ACCIDENTS IS INVESTIGATED. A DETAILED STUDY IS MADE OF HIGHWAY ACCIDENTS OCCURRING AT NON- INTERSECTIONAL MEDIAN-OPENINGS ON FOUR-LANE DIVIDED HIGHWAYS HAVING NO ACCESS CONTROL. A LIMITED INVESTIGATION IS ALSO MADE OF LANE TRANSITION AREAS AND FREQUENCY OF ACCIDENTS. FIFTY THREE RURAL SECTIONS OF DIVIDED HIGHWAYS IN NORTH CAROLINA HAVE NO CONTROL OF ACCESS. ELEVEN OF THESE WERE SELECTED FOR STUDY, THE PILOT INVESTIGATION OF A SITE ALONG US 301 IN WILSON COUNTY REVEALED LARGE CONCENTRATIONS OF ACCIDENTS AT MEDIAN OPENINGS. IT WAS DECIDED TO FOCUS THE PROJECT ON THIS SPECIFIC FEATURE. ACCIDENT DATA ON THESE 11 SITES WERE GATHERED FOR THE PERIOD 1959-1962. THE DATA SELECTION PHASE USED IBM PUNCH CARDS AND THE ORIGINAL ACCIDENT REPORT FORMS. ACCIDENTS PER MILLION VEHICLE-MILES WERE COMPUTED FOR ALL SITES TO ELIMINATE THE EFFECT OF THE VARIABLES OF VOLUME AND LENGTH WHEN MAKING COMPARISONS AMONG THE DIFFERENT SITES. COMMERCIAL VEHICLE ACCIDENTS AT MEDIAN- OPENINGS WERE INVESTIGATED SEPARATELY BECAUSE VEHICLE SIZE WAS THOUGHT TO BE RELATED TO ACCIDENT FREQUENCY. ALL ACCIDENT DATA WERE ANALYZED ACCORDING TO THE TYPE OF OCCURRENCE AND PATTERN OF MANEUVER OF THE VEHICLES. THESE TYPES WERE' 1/ ACCIDENTS IN WHICH VEHICLE WAS TURNING LEFT THROUGH THE OPENING FROM THE RIGHT LANE, AND WAS STRUCK BY ANOTHER VEHICLE PASSING IN THE LEFT LANE, 2/ ACCIDENTS IN WHICH ONE VEHICLE WAS STOPPING TO TURN LEFT AND WAS STRUCK IN THE REAR, 3/ ACCIDENTS IN WHICH ONE VEHICLE WAS COMPLETING A TURN ON THE OPPOSITE SIDE OF THE MEDIAN AND COLLIDED WITH ANOTHER VEHICLE TRAVELING IN THESE LANES, AND 4/ ALL MISCELLANEOUS TYPES OF MEDIAN-OPENING ACCIDENTS. ACCIDENT RATES AT ALL SITES WERE COMBINED INTO THREE MEDIAN WIDTH RANGES TO INVESTIGATE ANY POSSIBLE RELATIONSHIP. ACCIDENT RATES WERE ALSO COMPARED TO THE FREQUENCY OF OPENINGS AND TO THE PERCENTAGE OF RESTRICTED SIGHT DISTANCE ALONG THE SITES. IT WAS FOUND THAT ONE-FOURTH OF ALL ACCIDENTS OCCURRING BETWEEN INTERSECTIONS ON RURAL, NON CONTROL OF ACCESS, FOUR- LANE DIVIDED HIGHWAYS WERE HAPPENING AT MEDIAN-OPENINGS. MOST OF THE ACCIDENTS AT CROSSOVERS OCCURRED WHEN ONE VEH- ICLE WAS TURNING FROM THE RIGHT LANE ACROSS THE LEFT LANE TOWARD THE OPENING OR WAS STOPPING TO TURN AND WAS STRUCK IN THE REAR. COMMERCIAL VEHICLES WERE INVOLVED IN 65 OUT OF 230 MEDIAN-OPENING ACCIDENTS. SEVENTY-SEVEN PERCENT OF CROSS- OVER ACCIDENTS OCCURRED BEFORE A VEHICLE HAD MADE THE TURN THROUGH THE OPENING. THE INVESTIGATION OF THE TRANSITION AREAS SHOWED THAT AREAS WHERE THE FOUR-LANE HIGHWAY NARROWS TO A TWO-LANE ROAD WERE CRITICAL BECAUSE DRIVERS ATTEMPTING TO PASS WERE OFTEN FORCED INTO A SIDESWIPE COLLISION. IT IS RECOMMENDED THAT LEFT TURN SLOTS BE PROVIDED AT THOSE MEDIAN OPENINGS INTENDED TO SERVE COMMERCIAL ROADSIDE ESTABLISHMENTS AND RESIDENCES. THIS SHOULD REDUCE THE NUMBER OF ACCIDENTS AT THESE OPENINGS BY APPROXIMATELY THREE- FOURTHS.