Evaluating Pavement Response and Performance with Different Simulative Tests

Simulative tests refer to the Full-scale accelerated pavement testing (APT) and laboratory wheel tracking testing, which are widely used for evaluation of pavement responses and performance under a controlled and accelerated damage conditions in a compressed limited time. This dissertation focuses on comparative evaluations under ALF, MMLS 3 and APA tests, in terms of rut depth, strain response, seismic stiffness, and contact stress using both experimental and numerical simulation results. Test slabs extracted from the ALF test lanes, are trafficked with the MMLS3 under comparable environmental conditions at laboratory in Virginia Tech. Some specimens were cut from the slabs for APA tests at VTRC. It is found that the monitored parameters yielded by the MMLS 3 test were comparable to the related full-scale ALF test results in terms of intrinsic material characteristics and pavement performance. The wireless sensor network based on Internet of things technology is implemented in laboratory for the MMLS 3 test, which provides a convenient solution for researchers on long-term observation and monitoring without being physically presented.

The numerical simulations of ALF, MMLS 3 and APA in ABAQUS are used to supplement the investigation on the pavement response and performance under repeated moving loading. The viscoelastic-viscoplastic model is adopted to characterize rate and temperature dependent properties of asphalt mixtures. The 3D finite element models are capable of predicting the pavement response at critical locations while underestimates the rut depth because the permanent deformation induced by volumetric change cannot be represented in simulation.

According to the test results, a power law function fits well for the accumulated rut depth versus number of load repetitions before the material reaches tertiary stage in MMLS 3 test. The rut depth development of APA tests exhibits a close-to-liner regression with number of load cycles after the initial 500 load repetitions. A regression model for predicting rut depth after 500 loads has a satisfying agreement with the experimental measurement. The calibrated MEPDG fatigue model can be used to estimate the allowable load repetitions in MMLS 3 trafficking. Besides, the effects of tire configuration, tire pressure, axle load amplitude, wheel load speed and temperature on pavement responses are investigated in this dissertation using the finite element model.

It is concluded that MMLS 3 is an effective, economic and reliable trafficking tool to characterize rutting and fatigue performance of pavement materials with due regard to the relative structures. MMLS 3 test can be employed as the screen testing for establishing full-scale testing protocols as desired or required, which will significantly enhance economics of APT testing.