Predicting Pavement Skid Resistance via Multi-Source Contact Mechanics and Texture Morphology at the Beijing RIOHTRACK

Pavement texture contributes to friction only where it engages the tire, limiting the interpretability of whole-surface texture indices. This study develops a contact-informed framework for predicting skid resistance at RIOHTrack, Beijing Highway Traffic Test Field. Pressure-sensitive film and three-dimensional laser scanning characterized contact stress and surface morphology under a 25 kN tire load and 830 kPa inflation pressure. British Pendulum Number, peak contact stress, and effective texture density were related to the side-force coefficient at 60 km/h using seven asphalt-section means. Asphalt surfaces exhibited partial, or semi-interlocking, contact, with effective depths mainly of 2–3 mm and effective-to-maximum depth ratios of 0.3–0.4; concrete surfaces showed different ranges. The reported regression achieved adjusted R2 = 0.92 and an overall F-test p-value of 0.013, but two individual predictors were not significant at the 0.05 level. Reanalysis of the tabulated data gave a leave-one-section-out root-mean-square error of 4.56 SFC units, exceeding the in-sample error of 0.97. Two-year lane comparisons associated the PG82-22 binder with lower attenuation of excess effective texture density than PG76-22. The framework links texture selection to partial tire engagement, while the small calibration sample requires independent validation before engineering deployment.