Contactless Energy-Based Tire-Pavement Friction Models

Abstract: Tire-Pavement friction is a complex phenomenon affected by multiple factors including pavement surface texture, tire geometry and material characteristics, slip ratio, and skew angle. This paper discusses the findings from an ongoing study in Illinois implementing energy-based friction models (EBFM) to predict skid numbers (SN) and coefficient of friction (COF) on a wide range of pavement surfaces. The data collection was performed in two phases. In Phase I the data included high-density stationary laser texture scans and high-speed texture scans to estimate the SN values from the Locked Wheel Skid Trailer (LWST) and the COF measured using the dynamic friction tester (DFT) on 14 pavement sections. In Phase II, the data included high-density stationary laser texture scans and LWST measurements acquired from more than 15 highway segments to validate the findings form Phase I on operational highway segments. Surface texture was characterized using wavelets energy as an average representation of the power spectral density (PSD) within a given bandwidth. Then using a pareto mechanistic model, based on Persson’s friction model, the wavelets energy was used as the major explanatory variable to predict the SN and COF values at multiple speeds and using ribbed and smooth tires. Moreover, the study compared the models for peak and average SN values and compared the findings at multiple testing speeds and for dry DFT. The developed models had an R2 values ranging from 60% to 85% indicating a robust model given the vastly different surfaces and conditions included in the study. The presentation will discuss more detailed goodness of fit parameters. The models provided an estimation of the contribution of different frequency bands in the micro and macro texture ranges to the energy dissipation leading to frictional forces. In addition, the presentation will discuss the role of texture parameters in identifying hot spots for safety improvements and the use of texture parameters in safety evaluation for fiction management programs.