In this study, Traffic Conflict Technique was applied to the evaluation of safety at curve sections. The road observed was a two-lane highway in a mountainous area. Traffic flow volume was 14,500 vehicles per day. Eight curves were selected for the observation of traffic conflicts, and their radii ranged from 17 to 60 m. In total, 69 personal injury accidents had been recorded on the stretch of the road in the last five years. No conflicts were recorded during the observation but a relatively high correlation was found between the number of accidents in the last five years and the rate of vehicles partially or fully entering the opposite lane, which was measured in the two-hour observation. It is understood that the larger the number observation. It is understood that the larger the number of vehicles straying into the opposite lane, the greater the chances of head-on collisions. Therefore accident potential at curves can be assessed by observing the vehicles which enter the opposite lane, without having to observe conflicts. This idea may be considered as the extension of conflict definition, and this will be useful for application of Traffic Conflict Technique to where conflicts are seldom observed (A).
Abstract