multilane highways. The headway acceptance characteristics of U-turning vehicles are analyzed at unsignalized intersections on four-lane divided roadways.
Headway Acceptance Characteristics of U-Turning Vehicles at Unsignalized Intersections Pan Liu, Xu Wang, Jian Lu, and Gary Sokolow of the Highway Capacity Manual (HCM) contains a procedure for estimating the capacity and level of service of minor traffic streams at two-way stop-controlled intersections (1). Headway acceptance theory forms the foundation of the HCM capacity analysis method, and critical headway plays an important role in the headway acceptance theory. The HCM provides critical headways for some minor movements at unsignalized intersections on two- and four-lane divided roadways. These minor movements include left turns from a minor street, left turns from a major street, right turns from a minor street, and crossing movements. Unfortunately, HCM does not provide the critical headway for U-turn movements. A vehicle that makes a U-turn at an unsignalized intersection conflicts with the major-street traffic stream and the right-turning traffic stream from the driveway (Figure 1). A U-turning vehicle must yield the right-of-way to the major-street traffic stream; a driver must wait at the left-turn bay until the headway in the major-street traffic stream is suitable to make a U-turn. The headway acceptance characteristics of U-turning vehicles on four-lane divided roadways are more complex than those of other turning movements, partly because they are likely to be affected by some external factors (e.g., widths of median and roadway). For a U-turn design to be satisfactory, the width of the roadway plus the width of the median should be wide enough to permit a vehicle to turn from an exclusive left-turn lane in the median into the lane next to the outside shoulder (Figure 1). Vehicles at such locations should be able to perform continuous U-turn maneuvers without any impedance. On six- to eight-lane divided roadways that have at least three lanes in each direction, vehicles usually can make U-turns without considering the width of the median because the roadway widths already provide sufficient turning radii to U-turning vehicles. On four-lane divided roadways with only two lanes in each direction, however, roadway width usually is limited. The turning radius of a U-turning vehicle is partly provided by the median width. If the median is narrow, then a vehicle might have limited physical space to perform a U-turn. A driver in such a situation may make a tight U-turn at a slower turning speed and may require a larger headway to enter the major-street traffic stream. Although several studies have evaluated the operational effects of U-turn movement on multilane highways (2–6), little documentation is available concerning the headway acceptance characteristics of U-turning vehicles at unsignalized intersections. The only previous research concerning the headway acceptance characteristics of U-turning vehicles was undertaken by the University of South Florida (USF) in 2001 (7 ). The study’s research team collected headway acceptance data at 10 selected median openings. Of these 10 openings, 7 were located on six-lane divided roadways with three lanes in each
During the past two decades, more and more state departments of transportation have installed nontraversable medians and directional median openings on multilane highways. These installations have resulted in an increased number of U-turning vehicles at unsignalized intersections on multilane highways. The headway acceptance characteristics of U-turning vehicles are analyzed at unsignalized intersections on four-lane divided roadways. The critical headway for U-turn movement is estimated by using the maximum likelihood method. A binary logit model is developed to identify the influence of median width on the headway acceptance characteristics of U-turning vehicles. Results of this binary logit model indicate that the median width at a median opening significantly affects the decisions of U-turning drivers about accepting or rejecting a particular headway in the major-street traffic stream. Results of the maximum likelihood estimation indicate that a vehicle making a U-turn at an unsignalized intersection with a wide (≥21 ft) median has a smaller critical headway than one making a U-turn at an unsignalized intersection with a narrow (60 vehicles/h).
(6)
and the linearization of the logit model is ln
P ( xi ) = xi′β 1 − P ( xi )
(i = 1, 2, . . . , n )
(7)
where P(xi) denotes the probability that a headway i is accepted and x i′β is the multiple linear combination, also known as the utility function. The utility function can be expressed as xi′β = β 0 + β1 x1i + . . . + β k x ki
Some geometric characteristics of the selected sites are listed in Table 1. The range of median widths is 3 to 45 ft. At Sites 1, 2, and 3, many passenger cars must encroach onto the shoulder to perform U-turns (e.g., 73.2% at Site 1). This high rate of encroachments implies that median widths at these sites are insufficient for vehicles to make continuous U-turns. At Sites 4, 5, and 6, most passenger cars make U-turns from the leftturn bay into the lane next to the outside shoulder, and few encroachments onto the shoulder were observed (e.g., >77.8% at Site 4;
