This research was initiated to determine if post-mounted reflective delineators placed along interstate highways were effective and valuable enough as a traffic control device to warrant their installation and maintenance. The primary investigation consisted of the measurement of vehicle speed and placement at a single point within each of 4 test sections, 2 tangents and 2 curves, located on a single segment of interstate 10 in southwest louisiana. These measurements were made with a radar speed meter and a bureau of public roads placement tape for a minimum sample of 100 independent passenger-car vehicles for each combination of experimental conditions' daylight vs nighttime, nondelineated va post-delineated vs edge-striped delineation, eastbound vs westbound direction of travel, and tangent vs curve geometrics, a total of 24 combinations. In addition, supplementary speed and placement, nighttime driver interview and test-vehicle distance-lapse film studies were conducted to help verify primary investigations. Analysis of the principal study data showed mean speeds to be some 2 mph lower under delineated conditions than under nondelineated conditions but this difference was not considered to be significant from a practical standpoint. Also there was no significant effect of delineation on vehicle placement. /in general it was shown that vehicles travel closer to the centerline at nighttime as compared to daytime and at speeds over 64 mph as compared with traveling at speeds less than 55 mph./ driver interviews provided origin-and-destination data on vehicles using the test sections at night and yielded an almost unanimous approval of delineation by drivers. Analysis of test-vehicle distance-lapse films provided placement profiles for individual vehicles or drivers and determined a new traffic characteristic, the placement profile smoothness index; however, the limited number of such studies obtained did not provide a positive basis for distinguishing the effectiveness of roadside delineation. /author/.
Samenvatting