Insufficient understanding of the stress-strain behavior of pavements built over backfilled trenches, particularly with recycled aggregates, often leads to overdesign or overcompaction, raising costs and project delays. This research investigates how compaction levels during backfilling impact the pavement performance over these trenches. Various recycled material mixtures, both unbound and cement-treated, are compared with conventional crushed rock. Investigations included repeated load triaxial (RLT) tests, microstructural analysis with scanning electron microscopy, environmental assessments, and modeling with FlexPAVETM, a pavement response and performance analysis software. RLT test results were incorporated into the FlexPAVETM models by utilizing established constitutive resilient modulus models. Stress-strain responses of pavements over recycled aggregate backfill, compacted with standard and modified Proctor efforts, were compared with those over crushed rock and natural clay subgrades. Outcomes revealed that the standard compaction energy was sufficient for the desired performance. Fatigue and rutting strains with recycled mixtures closely resembled those with crushed rock, making them viable green alternatives. Pavements over backfilled trenches exhibited 1.5 and 1.8 times longer fatigue and rutting lives, respectively, than those over natural clay subgrades.
To satisfy the economic requirements and reduce the impact to the surrounding buildings and underground structures, the dynamic compaction (heavy tamping) and static compaction are combined used in the soil filling for airport subgrade. Despite compaction the subgrades in the same degree of compaction, the subgrades filled by dynamic and static compaction method show different increase potential in the permanent strain under cyclic loading, which then further result in the differential settlement and safety problems. This study firstly investigated the compaction characteristics under static compaction and different dynamic compaction scheme, during which the static and dynamic compaction strain and stress evolutions were monitored. The cyclic triaxial tests were then performed to investigate the sample preparation method derived difference in permanent strain under cyclic loading. Furthermore, to provide a microscopic interpretation to this difference, the pore size distributions of the silt samples based on mercury intrusion porosimetry (MIP) test and the internal particle contact stresses from discrete element method (DEM) simulation were respectively explored. The main conclusions are as follows: (1) The dynamic compaction processes can be divided into rapid and slow compaction strain stages determined by strain growth rate and compaction numbers, which further influences the homogeneity of soil samples; (2) The statically compacted samples have more significant permanent strain than the dynamic ones due to the localized stress concentration and different pore microstructures; the permanent strain increases with dynamic compaction energy until a stable stage is reached. (3) The MIP results show that the dynamic compaction transforms the macropores into mesopores; the higher compaction energy enhances this transforming effect but results in a decrease in the overall homogeneity.