الفهرس 
General IntroductionOnly 14 pages are availabe for public view
 |  | from | 178 |  |  |
| | | from | 178 | | |
Abstract
Existing geometric design guides provide deterministic design criteria for highway elements that ignore the uncertainty associated with many geometric design parameters. Several studies have advocated probabilistic geometric design where reliability analysis can be used to account for the uncertainty in the design parameters and to provide a risk measure of the degree of deviation from design standards.
This study used for the first time reliability analysis for highway geometric design in Egypt to evaluate twelve horizontal curves in Menoufia, Egypt, to check if the available sight distance (ASD) and the available horizontal curve radius satisfy the required sight distance and the required curve radius by road users or not. For evaluating the ASD, non-compliance occurs whenever the ASD falls short of the stopping sight distance (SSD). The inputs of the SSD were the operating speed, driver’s perception reaction time (PRT) and deceleration of vehicles (a). All of them were random variables. The mean value and standard deviation were assumed for ”PRT” and ”a” based on previous studies. However the operating speeds were obtained based on spot speed data collection from sites, instead of using the previous models, as the case in the previous works. These operating speed models vary significantly from one country to another because of changing in vehicle and driver characteristics, climate condition and design criteria between different countries, so calculation of speed from models as in previous studies was uncertain and could be a mistake. On another hand the inputs of the ASD were deterministic. The results showed that about 85% of studied curves have probability of non-compliance (𝑃𝑛𝑐) less than 10%.
For evaluating the available curve radius, non-compliance occurs whenever the required curve radius exceeds the available curve radius. The inputs of the required radius were the superelevation, the operating speed and the lateral friction. The superelevation was deterministic and its value was determined from data collection and both of the operating speed and the lateral friction were random variables. The mean of the lateral friction was obtained based on the previous model and the standard deviation was assumed. However the mean and the standard deviation of the operating speed were obtained based on spot speed data collected at sites. On another hand the available radius was deterministic. The results showed that more than 90% of studied curves satisfy the required radius by curve users and have probability of non-compliance (𝑃𝑛𝑐) less than 1%.
Finally, this study provided calibrated design charts for the middle ordinate (M), defined as the lateral distance between side obstruction and centerline of the inner traffic lane, at different probability of non-compliance (𝑃𝑛𝑐) levels. The calibrated values were compared with AASHTO values. The results showed that the calibrated values of (M) were generally lower than those derived from the AASHTO design guide.
The results of our study can help the highway designer in increasing the safety of geometric design standards in case of horizontal curve design and quantifying the safety level built in design values that are deemed acceptable. Also results indicating that as the frequency of accidents increases when Pnc values increase, due to difference between the available variable and the required variable by road users. In other words, higher values of probability of non-compliance interrupt the homogeneity of drivers’ control of the vehicle and may lead to driver errors and accidents.