Thermally induced slab curvature, or curling, can cause 24-hour variability in the ride quality of Jointed Concrete Pavements (JCP). Locked-in warp curvature may be larger than typical curling curvature and can add to roughness. These curvature effects are often inseparable from persistent structural roughness, including faulting and cracking, in roughness calculations. This study presents a curvature-based framework to back-calculate the effective total slab temperature gradient from longitudinal wheel-path profiles and quantify the portion of the International Roughness Index (IRI) attributable to curl-and-warp slab deformation. Average slab curvature is extracted from field-measured wheel-path profiles after excluding joint-adjacent regions. The measured curvature is matched to predictions from finite-element method (FEM) simulations using an Artificial Neural Network (ANN). A brute-force ANN search identifies the effective total temperature gradient matching the measured curvature. The back-calculated curvature state generates synthetic profiles, from which a slab-curvature-only IRI component is computed for a 9,000-ft JCP test site.
Ahmed et al. (Fri,) studied this question.