Abstract

The Highway Capacity Manual (HCM) defines the level-of-service criteria for two-lane highways based on three performance measures: average travel speed (ATS), percent time spend following (PTSF) and percent of free-flow speed (PFFS). The performance measures are affected by the lack of passing opportunities, which is quantified through adjustment factors depending on the percentage of no passing zones. Passing zones are provided where available sight distance is greater than the minimum passing sight distance, and their length and frequency depend mainly on the physical constrains and cost limitations. HCM current evaluation of traffic performance excludes horizontal alignment variations, which could produce more or less sinuous alignments and thus affect potentially to both average travel speed and percentage time spent following. Moreover, HCM considers terrain type rather than the actual vertical profile, which could overestimate or underestimate the expected traffic performance depending on how accurately the roadway was adapted to the terrain. The objective of the research is to analyze the relationship between highway horizontal alignment, characterized by curvature change rate (CCR), highway vertical alignment, characterized by the gradient change rate, and traffic performance, based on ATS and PTSF. For the research, 25 Spanish two-lane rural highways were selected. Horizontal geometry was obtained from Google Earth using an azimuth-based restitution program. CCR was calculated and used to identify homogenous segments. Vertical alignment was recreated from GPS data. The sample included 112 homogeneous segments and 451 passing zones. After the geometry-based characterization, 20 homogeneous segments were introduced and simulated in the TWOPAS simulation model. TWOPAS was previously calibrated and validated with passing maneuvers data from Spanish highways. Traffic conditions (directional traffic volume, directional split and percentage of heavy vehicles) were varied. The analysis of the simulation results provided models of PTSF and ATS depending on CCR and gradient change rate, among other parameters. The results of this research will help highway designers and practitioners to predict operational performance based on CCR, type of gradient and expected hourly traffic volume, and could be used as management and design criteria.