Abstract
IntroductionVehicles in transport sometimes leave the travel lane and encroach onto natural or artificial objects on the roadsides. These types of crashes are called run-off the road crashes, which account for a large proportion of fatalities and severe crashes to vehicle occupants. In the United States, there are about one million such crashes, with roadside features leading to one third of all road fatalities. Traffic barriers could be installed to keep vehicles on the roadways and to prevent vehicles from colliding with obstacles such as trees, boulder, and walls. The installation of traffic barriers would be warranted if the severity of colliding with the barrier would be less severe than colliding with other fix objects on the sides of the roadway. However, injuries and fatalities do occur when vehicle collide with traffic barriers. A comprehensive analysis of traffic barrier features is lacking due to the absence of traffic barrier features data. Previous research has focused on simulation studies or only a general evaluation of traffic barriers, without accounting for different traffic barrier features. MethodThis study is conducted using an extensive traffic barrier features database for the purpose of investigating the impact of different environmental and traffic barrier geometry on this type of crash severity. This study only included data related to two-lane undivided roadway systems, which did not involve median barrier crashes. Crash severity is modeled using a mixed binary logistic regression model in which some parameters are fixed and some are random. ResultsThe results indicated that the effects of traffic barrier height, traffic barrier offset, and shoulder width should not be separated, but rather considered as interactions that impact crash severity. Rollover, side slope height, alcohol involvement, road surface conditions, and posted speed limit are some factors that also impact the severity of these crashes. The effects of gender, truck traffic count, and time of a day were found to be best modeled with random parameters in this study. The effects of these risk factors are discussed in this paper. Practical applicationsResults from this study could provide new guidelines for the design of traffic barriers based upon the identified roadway and traffic barrier characteristics.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have