Balancing Fracture and Fatigue Resistance of Marshall-Designed Asphalt Mixtures with High Contents of Multi-Source Fractionated RAP

This study examines how reclaimed asphalt pavement (RAP) dosage, virgin-binder grade, and virgin-binder content govern the cracking response of Marshall-designed mixtures. The experimental matrix comprised four RAP levels (30%, 40%, 50%, and 60%), two modified virgin binders (PG 76-22 and PG 88-34) evaluated at their respective optimum asphalt contents (OACs), and two binder-rich PG 76-22 variants (OAC + 0.3 and OAC + 0.5 percentage points); virgin mixtures served as controls. Cracking behavior was characterized by low-temperature semi-circular bending (SCB), the Illinois Flexibility Index Test (I-FIT), and direct-tension cyclic fatigue (DTCF). Fracture and fatigue indicators were subsequently integrated into cracking balance design diagrams, and k-means clustering was used to derive provisional, dataset-specific performance boundaries for long-term oven-aged mixtures. The results show that the higher-performance binder improved long-term cracking resistance at moderate RAP contents. By contrast, mixtures with 50% or 60% RAP and PG 76-22 displayed a distinctly brittle response. Increasing the PG 76-22 content above OAC produced only limited gains. These findings demonstrate that binder quality, rather than a small increase in binder dosage, is the more effective lever for balancing fracture and fatigue resistance in mixtures with substantial RAP contents.