Assessing Safety Performance Function by Integrating Macrotexture and Friction on Urban/Rural Roads

Abstract: In 2021, the National Highway Traffic Safety Administration (NHTSA) reported 42,939 motor vehicle traffic fatalities, with 40% occurring in rural areas and 60% in urban areas. While the overall fatality rate dropped to 1.13 fatalities per 100 million vehicle miles traveled (MVMT) in early 2024 (down from 1.18 in the same period of 2023), rural roads remain notably affected by crashes, with a fatality rate 2.4 times higher than that of urban roads. Therefore, ensuring safety on rural and urban roadways is significant as it accounts for most highway fatalities. Driving speed significantly affects the severity and frequency of crashes through its impact on pavement frictional demand. As vehicle speed increases, tire-pavement contact decreases, leading to reduced friction and increasing crash risk, particularly on rural roads. Research shows that low speed roadways experience friction challenges and loss of macrotexture due to frequent braking, high traffic volumes, and insufficient maintenance, further increasing crash risk potential, especially in wet weather conditions. Thus, implementing appropriate frictional properties according to the roadway classifications and speed limits is significant in reducing the fatality rate and serious injury crashes. This study focuses on exploring the interaction between pavement friction and macrotexture in both rural and urban road facility types, using Safety Performance Functions (SPFs). The dataset includes road functional classification, friction at lower slip speeds (SFN(40)), macrotexture (MPD), traffic volumes (AADT), surface mix types, road geometry, and historical crash data. Specifically, it investigates the combined friction index (integrating microtexture and macrotexture) and its effects on crash risks on rural multilane roads, urban arterials, and two-lane rural roads. Friction indices (FRS) are calculated at various speeds (40, 50, 60, and 80 km/h) using the ASTM E1960-07 speed correction. The analysis shows that models incorporating the FRS perform consistently better than models that include friction and macrotexture as individual predictors. This is established using performance metrics such as Akaike Information Criterion (AIC) and negative log-likelihood (NLL). Further, the findings reveal that the estimated available friction at 80 and 40 km/h, have considerably greater impact on crash risk potential on rural and urban roads, respectively. SPFs enabled the development of crash modification factors (CMFs) for computing the potential reduction in crash risk due to changes in both friction and macrotexture. Last, the CMF calculations resulted in significant crash reduction when enhancing friction and macrotexture properties using FRS indices.