Optimization of rubber powder-composite fiber modified asphalt mixture and its pavement performance application in ultra-thin wearing course

This research presents a rubber powder - composite fiber (RP-CF) modified asphalt technology for the ultra-thin wearing course (UTWC). Rubber powder is one of the materials employed to enhance the properties of asphalt. Utilizing this recycled rubber powder as a common and cost - effective additive not only contributes to improving asphalt performance but also aids in environmental protection. The combined incorporation of basalt fiber (BF) and lignin fiber (LF) can address the drawbacks associated with BF, such as its high density and cost, as well as the negative impact of excessive LF on the fatigue resistance of asphalt mixtures. By applying the response surface methodology (RSM), the interactions among the rubber powder (RP) content, asphalt - aggregate ratio (AAR), and composite fiber (CF) proportion, along with Marshall stability (MS), flow value (FV), and Cantabro abrasion loss (CAL), were optimized. When compared with the common BF- or LF- modified control specimens, the optimized RP-CF mixture demonstrates superior pavement performance characteristics. Specifically, the dynamic stability (DS) increases by 18.23% - 71.08%, the residual stability (RS) improves by 2.11% - 5.75%. The permeability coefficient decreases by 4.52% - 29.29% (down to 16.9 mL/min), and the low-temperature performance increases by 3.74%-32.14%. This indicates enhanced rutting resistance, moisture resistance, and compactness. Moreover, the amount of fiber used is reduced, effectively cutting down on production costs. The RSM framework effectively resolves the multi - factor interactions, offering a reliable approach for the design of sustainable UTWC mixtures. • Response surface methodology optimized the rubber powder-composite fiber asphalt mixture. • The optimized mix achieved Marshall stability of 10.23 kN, and flow value of 22.7. • Dynamic stability and Marshall residual stability increased by over 59.20% and 2.11%. • Permeability reduced by 27.16% - 29.29%, improving mixture densification and durability. • All predictive errors were below 5%, validating the model’s reliability.