Evaluation of alternative superelevation distribution methods

Abstract

Understanding vehicle performance on horizontal curves ics. is needed to improve current design standards. Superelevation distribution methods determine the amount of supererogation that is used on each horizontal curve. The supererogation distribution methodologies currently used in the United States do not provide a clear understanding of the operational characteristics of these methods. These methodologies were developed on the 1930's, 1940's, and 1950's. Since then, roadway, pavement, and vehicle characteristics have greatly evolved. This thesis documents an evaluation of the two supererogation distribution methods recommended in, 4 Policy on Geometric Design of Highways and streets (i.e., Methods 2 and 5, J). Specifically, the focus of this evaluation was on predicting the portion of the driver population experiencing adequate side friction demand when traveling on a horizontal curve. Because of driver expectancy, adequate side friction demand should be lower than the comfortable limit and higher than zero. Good design guidelines should provide adequate side friction demand for the majority of drivers. The results indicated that Method 2 provides adequate side friction demand for the majority of drivers. Additionally, Method 5 provides side friction demand below the maximals comfortable side friction limit for a majority of the drivers. However, it was found that Method 5 fails to provide the majority of drivers with positive side friction. The results of this thesis indicate that the portion of drivers experiencing negative side friction increases with increasing curve radius. Also, the portion of drivers experiencing negative side friction decreases with increasing 95th percentile speed of the facility. Recommendations include investigating crashes related to negative side friction demand, and considering both fast and slow driver in curve design. Also, Method 5 should be modified to consider the 5th percentile driver.