Understanding the Impacts of Ramp Configuration on Commercial Motor Vehicle Rollover Crashes
Commercial Motor Vehicle (CMV) rollover crashes are a major public safety concern in the United States. 52 percent of large truck occupant deaths in 2011 occurred in crashes in which their vehicles rolled over. This thesis aims to provide an improved understanding of how ramp configurations and driver characteristics impact the likelihood of a rollover crash. Specifically, this thesis includes 1.) an analysis of CMV crash data at on- and off-ramps from 2005 to 2013; 2.) an equivalent property damage only (EPDO) clustering analysis of all CMVs from 2013 to 2014 using geographic information system (GIS) software; 3.) a linear clustering crash analysis of both CMV and non-CMV rollover crashes from 2007 to 2012. All of the data and analyses were for crashes which took place in the Commonwealth of Massachusetts. The analysis of all on- and off-ramp crashes had a sample size of 2,149 crashes, resulting in an even split of crashes across all age groups. 79.3 percent of ramp crashes occurred in daylight conditions, and 78.9 percent of crashes took place on dry pavement. 35.6 percent of crashes did not have a reported driver contributing code, which leaves a huge gap in determining recurring human factors. A different analysis of the same data looked at only rollover crashes, consisting of a sample of 111 crashes. Rollovers occurred evenly across all age groups. 63.2 ii percent of vehicles which rolled over had a gross vehicle weight rating over 26,000 pounds, the highest classification for all vehicles. 80.2 percent of crashes occurred in daylight conditions, and the roadway was dry in 91.0 percent of instances. 47.7 percent of drivers were observed to be speeding, compared to the overall average of 4.7 percent. 49.5 percent of rollovers took place on free-flow loops and 3.6 percent occurred on diamond ramps. The large percentage of rollovers on free-flow loops are a result of the continuous curvature changes that are present at these ramps. Diamond ramps, on the other hand, are mostly linear with minimal curvature, leading to a lower crash rate. The EPDO clustering using GIS used a sample of 2013 and 2014 data involving all commercial motor vehicle crashes. The geo-statistical analysis shows that urbanized locations with dense populations and high traffic volumes are the most likely to encounter CMV crashes. This data sample had 49 rollovers, resulting in 2 fatal injuries and 12 injury crashes. 9 towns had two rollover crashes, including the town or Revere, which showed up as a high z-score location on the GIS map. Further analysis revealed that two rollovers occurred at the same rotary at the same time. Overall, the mapping of these rollover crashes does not give any obvious answers as to which locations should be improved because the sample size is too small and the sample area is too large. The linear clustering analysis resulted in 27 rollover data points; 17 of these had useful driver data and 5 had additional citation data. 16 of these drivers were male, 14 were between the ages of 30 and 59, and 3 of the operators were unlicensed. All but one of the trucks weighed over 26,000 pounds, and the majority of crashes occurred during daylight hours on dry pavement. 10 of the crashes were a direct result of driver error, while 4 crashes did not have a record of this information. iii The completion of research tasks within the framework of this thesis achieved the overall objective of providing insight on ramp types and crash data statistics of vehicle and driver demographics. Sex, age, roadway lighting, unfavorable road conditions and inclement weather were non-influencing factors in the vast majority of these crashes. Driver error seen in vehicles over 26,000 pounds is by far the most noticeable cause. Moving forward, the insights established within this thesis may prove useful in the establishment of target locations and specific countermeasures.
This document is currently not available here.