Multi-Stage Fatigue Characterization of Asphalt Mixtures Under Stress- and Displacement-Controlled Loading

Conventional laboratory fatigue tests are often limited to single-stage loading. As a result, they do not fully capture the progressive damage caused by repeated traffic loading. To address this limitation, this study proposes an integrated multi-stage fatigue testing (MSFT) framework. The framework combines Multi-Stage Stress-Controlled Dynamic (MSSD) testing and Multi-Stage Displacement-Controlled Dynamic (MSDD) testing. These two modes are used to characterize asphalt mixture fatigue behavior under complementary loading conditions. The proposed framework interprets stress-controlled and displacement-controlled responses together. This allows the two responses to be evaluated as part of one integrated system, rather than as separate tests. The experimental program included asphalt mixtures with different air-void contents, binder contents, and binder grades. Fatigue performance under MSDD loading was evaluated using dissipated-energy-based and force-response parameters. In contrast, MSSD response was characterized using fracture-mechanics-based Paris law fatigue indicators. The results showed that air-void content, binder content, and binder grade produced different fatigue responses under the two loading modes. Lower air-void content improved fatigue life and energy tolerance. High-air-void mixtures showed faster damage accumulation. Binder-rich mixtures performed better under displacement-controlled loading because they better accommodated the imposed displacement. However, they showed higher crack-growth sensitivity under stress-controlled loading. This confirms that asphalt mixtures can behave differently depending on the loading mode. The results show clear and systematic differences between the two loading modes. These differences reveal mode sensitivity that cannot be identified using a single fatigue test. Some mixtures performed well under one loading condition but showed lower resistance under the other. This presents the need for integrated interpretation when ranking mixtures. The proposed dual-mode framework can support tiered Balanced Mix Design decisions within performance-based specification systems.