Abstract
ABSTRACT
One of the fundamental factors contributing to vehicle accidents is the lack of adequate friction at the tire-pavement interface. Pavement can prevent skidding-related accidents considerably when providing the pavement surface with an adequate friction. Thus, the researchers have tended recently to study the factors affecting pavement friction. Several models were developed based on aggregate characteristics and asphalt mixtures, whether conducting samples in the laboratories or collecting related data from the field, in order to estimate the existing skid resistance and provide the necessary friction to the pavement when needed.
This research aims at validating a system to predict the skid resistance for Hot Mix Asphalt (HMA) developed previously by Masad et al. (2009) an Kassem et al. (2013) as well as developing a new system that well be able to predict the skid resistance for Sealcoat. To achieve the mentioned objectives, measurements of various types of aggregates were conducted and pavement surface characteristics were collected in the state of Texas. These measurements include acquiring aggregate texture and angularity after polishing different time durations of polishing in the Micro-Deval, and sieve size analysis was conducted by TxDOT for all selected aggregates. With respect to data collected from the field, skid number was measured using the skid trailer for each selected sections over a number of years. In addition, The Dynamic Friction Tester (DFT) and Circular Texture Meter (CTMeter) were used to measure pavement friction and the mean profile depth (MPD). It was desirable also to collect the required traffic data to be considered in the analysis for the selected sections.
A system was modified for HMA in order to predict the skid number (SN), which is an indicator of skid resistance, based on a wide range of aggregates. This system includes models to estimate the initial international friction index (IFI), terminal IFI and also rate or change in the IFI. Finally, validating the skid prediction model by comparing the measured skid numbers from the field with the SN obtained from the developed model. In addition, a new system was developed for the sections that include Sealcoat based on the aforementioned tests as well. The results indicate good correlation between the measured and predicted SN values in spite of the variability in the pavement surfaces and existence of bleeding in some sections which affects the models adversely.