INVESTIGATION OF SOIL ARCHING UNDER DIFFERENT MODES OF SOIL MOVEMENT AND SURFACE LOADING

Abstract

Soil arching exists in many geotechnical applications, including tunnels, buried pipes and culverts, and Geosynthetic-Reinforced Pile-Supported (GRPS) embankments. The existence of these buried structures or structural elements within soil masses causes redistribution of stresses, which is referred to as soil arching. The relative stiffness and differential settlement between these buried structures and their surrounding soils affect the magnitude and distribution of vertical stresses. Soil arching has been mostly investigated using trapdoor tests under soil self-weight and/or uniform surcharge. In real applications, localized surface loading, such as traffic loading, may be applied onto soil and affect or degrade soil arching. Also, additional stresses caused by traffic loading on a buried structure may cause excessive deformations and even failure of the buried structure. Geosynthetics have been used in GRPS embankments or over buried pipes and may have effects on soil arching mobilization and degradation under localized surface loading. Expanded Polystyrene (EPS) geofoam, a lightweight material, has been increasingly used above buried structures as a compressible inclusion to reduce vertical stresses acting on the buried structures. The effects of surface traffic loading and geosynthetics on soil arching have not yet been well investigated. Therefore, the main objective of this study was to investigate soil arching under different modes of soil movement and surface loading. To fulfill the above research objective, a comprehensive experimental study and numerical analysis were conducted. The experimental study included two experimental series. The first experimental series consisted of reduced-scale models of a buried box culvert that were constructed in a test box under a plane-strain condition. This study adopted the Induced Trench Installation (ITI) method to place the concrete culvert overlaid with an EPS geofoam and investigated the effects of EPS geofoam, including geofoam stiffness and thickness, on the distribution of vertical stresses above a rectangular concrete culvert under surface footing loading. The second experimental series utilized the trapdoor test setup to investigate the effects of localized surface loading on soil arching mobilization and degradation in geosynthetic-reinforced and unreinforced embankments under a plane-strain condition. The trapdoor was supported by compressible springs of a known stiffness and could move under fill self-weight and surface loading to simulate soil subsidence and/or consolidation of foundation (soft) soil between rigid supports. In both experimental series, the backfill material was a dry, poorly-graded Kansas River sand. The footing load was applied parallelly to the culvert or the trapdoor axes. Earth pressure cells were used to monitor the vertical stress distributions above the culvert, the trapdoor, and the surrounding soil. To comprehensively assess the effects of localized surface loading with different configurations, numerical models simulating trapdoor tests were built and validated against the results of the experimental tests. A series of parametric studies were conducted to investigate: the effects of fill height, the most critical condition of the surface loading (as for the footing width and location), and the effects of non-uniform trapdoor displacements by multi-segment trapdoors on soil arching mobilization. The experimental results of the buried box culvert show that the EPS geofoam reduced the vertical stresses on the buried structure due to the mobilization of soil arching. However, soil arching was found to be partially mobilized based on the measured soil arching ratio due to the low modulus ratio of soil to geofoam that caused limited compression of the geofoam. The lower stiffness and thin geofoam had more effect on the vertical stress reduction. Cyclic loading minimized the soil arching effect induced by the compressible geofoam. This study also examines the test results with available analytical solutions. The effects of soil arching and the induced vertical stresses above the rigid structure under static footing loading were considered separately. The analytical solutions were found to match well with the experimental results. The trapdoor test results show that the displacement of the trapdoor during the fill placement induced progressive mobilization of soil arching and geosynthetic reinforcement minimized soil arching mobilization due to the change of the soil deformation. Localized surface loading increased the degree of soil arching at low applied pressure (approximately 50 kPa); however, under higher footing loading, soil arching degraded or stress recovered due to larger trapdoor displacement. Single and double layers of geosynthetic reinforcement helped maintain soil arching under localized surface loading. Geosynthetic reinforcement increased the applied surface load required to fully degrade soil arching and eliminate the benefit of the geosynthetic. Soil arching exhibited arching degradation and even collapse under static loading; however, arching degradation was less pronounced under cyclic loading as the applied pressure increased beyond 80 kPa due to larger differential settlement within the fill. The results of the numerical simulations show that the degree of soil arching increased as the fill height (H) increased due to the additional shear forces mobilized throughout the fill material. Consequently, less pressure was applied on the trapdoor and more pressure transferred to the supports as the fill height increased from H/B of 1 to 3 (B is the trapdoor width). The model with a footing width of 0.5B was the most critical width and had the highest vertical pressure on the trapdoor for H/B of 2; however, the model with a footing width of 1B had the highest pressure on the trapdoor for both H/B of 1 and 3. The model with a footing offset of 0.0B from the centerline of the trapdoor had the highest vertical pressure on the centerline of the trapdoor. Also, as the footing offset increased to 1B, less pressure reached the trapdoor and more pressure transferred onto the support. In this study, an analytical solution was proposed based on Terzaghi’s theory but for localized footing loading along the centerline of the trapdoor. This solution well predicted the measured vertical pressures on the trapdoor under localized footing loading as compared with the trapdoor test results obtained in this study. In addition to the experimental tests, eight numerical models with different fill height to trapdoor width ratios (H/B = 1, 2, and 3) and different footing widths (0.25B, 0.5B, 1B, 1.5B, 2B, and 5B (uniform)) were selected and their numerical results were compared well with the proposed solution. The numerical results further validated the proposed solution for soil arching over a trapdoor or a yielding soil zone under localized footing loading.