Laboratory and computational mechanics-based framework for the analysis and design of cold recycled pavement layers

Cold recycling technologies are increasingly popular rehabilitation techniques for asphalt pavements. Those technologies allow to construct new pavement layers with minimal addition of heat and minimal need for transporting the material. In fact, the mixtures are usually prepared in mobile plants or directly in place. The products obtained through the use of cold recycling technologies are called Bituminous Stabilized Materials (BSM). BSM are considered partially-bonded materials since they have mechanical characteristics which are in between fully bonded materials, such as Hot-Mix-Asphalt (HMA) or cemented materials, and unbounded materials, such as crushed aggregates. Their mechanical response is simultaneously dependent on testing temperature and applied confining pressure and the main concern related to those materials in the field is the accumulation of permanent deformation under traffic loading application. Nonetheless, BSM are currently assumed to give a linear elastic response in pavement design and analysis. In this research study, two different mechanical characterization methods for BSM are compared. Indirect Tensile Strength (ITS) test and shear strength and resilient modulus tests in triaxial configuration were performed in the laboratory in order to identify the best suited approach to calculate elastic and plastic global properties for BSM. Tests in triaxial configuration showed the ability to fully capture the plastic behavior of the tested mixtures and for this reason were selected as characterization method for the study. Triaxial shear strength (TSS) and resilient modulus (TMR) tests were performed subjecting the material to different lateral confining pressures and to different temperature conditions. This allowed to collect information on material response under a tridimensional stress state and under a wide range of realistic temperature scenarios. On the basis of those results, a constitutive model for BSM was designed. Laboratory reaction force-displacement curves from TSS tests were matched with three-dimensional finite element elastoplastic model simulations in order to extract local elastic and plastic constitutive properties for the material. The local properties calibrated and validated in this stage of research were subsequently used as input parameters in multilayer elastoplastic pavement models for pavement structure evaluation. Different structural solutions with and without BSM as base layer were initially simulated in order to verify the ability of BSM to guarantee a comparable service life in terms of rutting accumulation with respect to traditional base layers. Afterwards, different scenarios with different constant temperature distribution throughout the pavement structures were simulated and elastoplastic analysis was compared to the traditional linear elastic analysis. Number of allowable load repetitions before fatigue and rutting failure were calculated respectively on the basis of horizontal strains at the bottom of HMA and vertical strains on top of subgrade taken from the model simulation results. The model was then implemented with functions able to consider a realistic temperature distribution with depth and consequently adjust material properties at the different locations with respect to the pavement surface. As last step, considerations on the impact of BSM curing stage on overall plastic response of the pavement structure were made. BSM mechanical properties were estimated at different curing days basing on laboratory results and equations found in the literature. Different structural solutions were simulated before the HMA overlay placement and trends of plastic deformation accumulation on the pavement surface were calculated for the first fourteen days after construction. The research study indicated the importance of considering elastoplastic models for partiallybonded and unbounded materials in the design and analysis of pavement structures. In addition, it was shown that the effect of temperature on BSM mechanical response cannot be neglected for an accurate pavement evaluation. In terms of curing, the simulations predicted a similar behavior of the pavement structure in between seven and fourteen days after construction. This information needs to be furtherly investigated with field and laboratory testing. The possibility of overlaying the BSM and re-opening the roadway to traffic at an earlier curing stage could guarantee significant savings on cost. Overall, this dissertation presents a framework for the analysis and design of BSM based on laboratory tests and computational mechanics analysis which could be adopted for future studies. In addition, this work gives a contribution for the improvement of current methods for pavement design and analysis including considerations on plasticity, indirect confining pressure effects and realistic temperature distribution with depth.