Valle-Javier, Dafne

Profile Picture

Filters

20151
Reset filters

Settings

Citations

Publication Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    Publication
    Crash modification factors for rumble strips treatment in freeway
    (2015-12) Valle-Javier, Dafne; Colucci-Ríos, Benjamín; College of Engineering; Maldonado Fortunet, Francisco; Figueroa Medina, Alberto M.; Ortiz Rodríguez, Rosario A.; Department of Civil Engineering; Lorenzo, Edgardo
    Longitudinal shoulder rumble strips is a safety feature or treatment on a paved shoulder, capable of alerting drivers that their motor vehicle is departing the traveled lane. This treatment has proven to be effective specially preventing roadway departure crashes. In 2009, Puerto Rico Highway and Transportation Authority (PRHTA) started the implementation of non-continuous longitudinal rumble strips on the right hand shoulder on freeways on the island. In 2010, the Highway Safety Manual (HSM) was published by AASHTO to provide tools for decision making to estimate how effective a countermeasure or set of countermeasures will be in reducing crashes at a specific location. The Crash Modification Factors (CMF) are used to quantify the effect of a particular treatment on expected crash frequency. On the HSM, CMF’s for freeway applications using non-continuous longitudinal rumble strips on shoulders were not included. This investigation documents the process of the development of CMF for non-continuous longitudinal shoulder rumble strips on freeway segments, in rolling to mountainous topography. The study area is highway PR-52, a 108.3 kilometers toll freeway facility that is part of the National Highway System (NHS) that originates in the north at San Juan, Capital of Puerto Rico, crossing the central mountain range and ends in the city of Ponce. Its AADT ranges from 165,800 vpd in its origin in the urban area with level to rolling terrain to a minimum of 18,600 vpd in a rural mountainous region. The maximum speed limit is 105 km/hr (65 mph) with approximately 8% of heavy trucks. The specific segment evaluated in this research starts on the South Caguas Toll Plaza (km 23.1) and ends at the exit ramp towards the town of Salinas (km 66.3). An observational study known as Empirical Bayes Method was performed to calculate the effectiveness of this treatment. This method can predict the number of crashes that would occur in the “after” period if the treatment or countermeasure had not been implemented and then compares these values with the actual count of crashes in the “before” period of the treated site. The first step on this method is the development of a multivariate crash prediction model or “Safety Performance Function” (SPF) specifically adapted to Puerto Rico’s jurisdiction for total crashes and run off the road (ROR) crashes for the reference group. This investigation revealed that, other than length and AADT, the presence of freeway ramps and climate category (dry areas versus moist/wet areas) were statistically significant in the prediction of total crashes. Furthermore, the SPF associated with ROR crashes for the reference group, had statistically significant variables such as segment length, AADT and presence of freeway ramps. No major outliers were detected for both SPF models on the final verification process. In addition, a jurisdictional SPF was calibrated for average crashes on Puerto Rico’s freeway network. The EB Method showed that for longitudinal intermittent shoulder rumble strips in the NHS PR-52 toll freeway there was a decrease on ROR and total crashes by 5% and 2%, respectively. The estimation of CMF for longitudinal intermittent shoulder rumble strips associated with ROR crashes was 0.95 and for total crashes was 0.98. The estimated CMF’s had standard errors of less than 0.1 indicating reliable results based on AASHTO guidelines.