Investigation of flexible bridge deck overlays in Kansas

Abstract

In Kansas, the bridges continuously deteriorate due to frequent freeze-thaw cycles and repeated traffic loads. Therefore, the bridge decks need to be overlaid intermittently. A hot-mix asphalt (HMA) overlay for bridge decks requires to be impermeable, friction-improved, and rut-resistant. This research focuses on determining whether modified Reflective Cracking Interlayer (RCI) mixtures, used on highways, can function effectively as a thin-bridge-deck overlay while also protecting the deck from moisture infiltration. This study used primary aggregates from the RCI mixtures, a gradation band based on the New Jersey Bridge Deck Wearing Surface Course, and a PG 70-28 RCI binder. Friction data from a freshly paved RCI overlay project showed an average Mean Profile Depth (MPD) of 0.021 in. (0.54 mm) and a mean British Pendulum Number of 82. While the RCI mixture meets the minimum MPD required by FHWA for highways, higher friction would be needed for an RCI-based HMA bridge deck overlay mix. In this study, the target design air voids at 50 gyrations for the HMA mix were 2.5±0.5% and an MPD of 0.039 in. (1 mm) or higher. Thirty-five trial gradations were assessed using the Laser Texture Scanner for friction, the falling-head permeability test for permeability, and the IDEAL-RT test for rutting. Concurrent Hamburg Wheel Tracking Device (HWTD) rutting tests were also conducted. The results showed that the initial fine-graded primary aggregate mixtures were impermeable but lacked sufficient friction. Adding calcined bauxite slightly increased the MPD. Subsequently, the study introduced secondary coarse aggregates in the mix design. Based on their volumetrics and MPD, five mix designs were selected. These mix designs achieved an MPD of 0.039 in. (1.0 mm) or higher, while all mixtures except one showed the nearly impermeable coefficient of permeability (k) of 0 to 1.97 ×10⁻⁵ in./s (0 to 5×10⁻⁵ cm/s). Although none of the mixtures met the target of 15,000 wheel passes at a 0.5 in. (12.5 mm) rut depth in the HWTD test, two mixtures exhibited higher rutting resistance. Statistical analysis using Response Surface Methodology and constrained multi-response optimization quantified the relationships between mix design variables (air voids, dust content, and coarse aggregate and coarse sand contents) and the three performance responses. Only one mixture satisfied all three performance thresholds (MPD ≥ 0.039 in. (1.0 mm), k ≤ 1.97 ×10⁻⁵ in./s (5 × 10⁻⁵ cm/s), RT Index ≥ 65), confirming it as the best-balanced mixture for bridge deck overlay applications. The mathematical optimum identified within the design space indicated that achieving all three performance targets simultaneously requires a coarse, near-gap-graded mixture with low fines and low air voids.