Contributions to the tire-pavement friction characterization from the traffic safety perspective

Abstract

The WHO recognizes traffic-related crashes as a public health problem, caused by a complex set of factors, many of which are related to transportation engineering. When it comes to infrastructure, the Traffic Safety Performance (TSP) is a function of the tire-pavement friction. Even so, roadway operators and urban authorities generally ignore this property in the pavement design and construction practices, taking safety-related strategies and decisions only after the occurrence of crashes. The two first stages of this work aimed to employ Digital Image Processing (DIP) techniques such as the Aggregate Imaging Measurement System (AIMS) and the Close-Range Photogrammetry (CRP) to predict the tire-pavement friction coefficient (μ) in the laboratory and to measure it in the field. Firstly, this research applied a predictive model of μ from aggregate shape properties, mixture gradations, and traffic. In the second stage of this research, a typical airport pavement mixture was designed and compacted in full-scale sections using 8 different protocols. For the texture analysis, traditional tests (sand patch – SP, and British pendulum - BP) and DIP techniques (AIMS and CRP) were applied. As a result, models for estimating the post-construction μ were proposed from parameters obtained from the AIMS texture analysis (MPD), and from the analysis of three-dimensional CRP models (MPD and Vmp). Finally, the assessment of vehicular conflicts in different urban scenarios was performed using microscopic traffic simulation (VISSIM software) and Surrogate Measures of Safety (SMoS) to support the proposition of minimum and ideal requirements of μ based on TSP parameters. TSP indicators are used to compare two main conditions: the necessary deceleration rates for vehicles to avoid a crash (DRAC), and the maximum available deceleration rates (MADR). DRAC was obtained using microscopic traffic simulation and vehicle trajectory analysis tools (SSAM3 software). MADR is a function of μ, vehicles’ speeds, as well as road characteristics. Typical corridors of Fortaleza were represented in this simulation effort, resulting in the definition of the μ limits of 0.20 (minimum) and 0.65 (ideal) for the Fortaleza urban scenarios. Based on this analysis, material selection can be assessed in terms of the suitability for use in the urban environment. The main framework of this thesis can be summarized as follows: once a traffic simulation is performed, the TSP analysis using the SSAM tool and applying SMoS allows the estimation of required μ limits. The DIP tools presented (AIMS and CRP) can support the assessment of the estimated μ limits through contactless protocols. Furthermore, μ prediction based on material characteristics integrated into a database can indicate the suitability of construction or maintenance strategies for use in the urban environment.