| The PCE of heavy vehicles in work zones are developed based on delay caused by the heavy vehicles since delay captures the effects of heavy vehicles better than other parameters. A mathematical model was developed to determine the bounds on the delays in work zones. The results of the model for level terrain and upgrades were validated. This model can also be extended to model entry and exit ramps, which was not possible in earlier models. Relationships between the car-following parameters of VISSIM and capacity were established. A generalized calibration procedure, using these relationships was developed for VISSIM when the desired capacity and queue length are known for a site. In addition field data was used to validate VISSIM results for work zones under undersaturated conditions. Four factors that affect D_PCE, namely, work zone length, speed difference between cars and heavy vehicles, traffic volume and heavy vehicle percentage, were considered and the D_PCE values for various combinations and levels of these factors were determined. Nonlinear regression was performed to obtain the relationships between D_PCE values and the four factors mentioned above. The D_PCE values for 5 mph, 10 mph and 15 mph speed difference between cars and heavy vehicles for different volume levels and heavy vehicle percentages were computed. The effects of narrow lanes and lateral clearances on the speeds of cars and heavy vehicles in a work zone were studied. The speed reductions in work zones were found to be significantly greater than the speed reductions given in HCM 2000 for a basic freeway section of identical lane width. The recommended speed reductions for work zones are 10, 7, 4.4 and 2.1 mph for lane widths of 10, 10.5, 11 and 11.5 ft respectively. A new methodology is proposed to compute user delay and user costs in work zones. Unlike previous methodologies, the effect of differences in speeds within and between vehicle types is considered. The D_PCE values developed in this research are used to compute moving delay and input-output analysis is used to compute the queue delay. |
