Some factors effecting left-turn capacity and delay at unsignalized intersections.

This paper considers some approaches to improving the mathematical modelling of traffic capacity and delay at an intersection with a left turn into a lane which is crossed by pedestrians. It is no longer assumed that each driver has the same gap acceptance; instead, the sequence of gap acceptance times is modelled as a stochastic process in which the psychophysical characteristics of the drivers are parameters. The use of the cumulative distribution function of gap acceptance, instead of a constant value for critical gap, in the acceptance model increases the precision of the estimates of traffic capacity and delay. The analysis of the relation, between the vehicle stop-time in front of a crosswalk and the pedestrian flow volume, has led the authors to the conclusion that the percentage of vehicles crossing the crosswalk without delay decreases sharply when the volume of pedestrian flow increases; in a typical case, this percentage fell from 78% at 200 persons/ hour to 21% at 1500 persons/hour. It is therefore necessary to include pedestrian flow in accurate methods for calculating left-turn capacity.