INVESTIGATION OF SOIL ARCHING STABILITY UNDER STATIC AND CYCLIC SURFACE LOADING USING TRAPDOOR MODEL TESTS

Abstract

Soil arching is a phenomenon describing pressure redistribution due to relative movement between adjoining portions. It commonly exists when soil interacts with structure elements, for example, tunnels, retaining walls, buried structures, and piles in pile-supported embankments. Therefore, soil arching is a key mechanism of load transfer in these geotechnical applications. The performance of these applications, where differential settlement, complete loss of support, or differential stiffness occurs, highly depends on the stability of the soil arching. Trapdoor tests have been widely used by researchers to demonstrate and investigate the soil arching phenomenon. However, most trapdoor tests have been conducted under soil self-weight or soil self-weight plus uniform static surface load. In other words, the soil arching was investigated focusing on particle-particle interaction instead of stress transfer due to localized external loading. In addition, earth structures are often subjected to cyclic surface loading (due to moving vehicles and railroad crossings) and dynamic-in-depth loading (due to pile driving, blast waves, and earthquakes). Unfortunately, limited research of cyclic or dynamic loading on soil arching stability was conducted. Moreover, current design methods for geosynthetic-reinforced earth structures involving soil arching, such as geosynthetic over voids and geosynthetic-reinforced pile-supported embankments, were mostly based on the findings from trapdoor studies without any geosynthetic. This extrapolation lacks appropriate theoretical and experimental justifications. This study is to address the aforementioned points by conducting a series of physical model tests under a plane strain condition. Fourteen model tests were conducted including two baseline tests and twelve other tests. The two baseline tests were carried out under only footing loading, one with static loading and another with cyclic loading. The remaining twelve tests consisted of both trapdoor and loading tests to evaluate the stability of the soil arching. Kansas River sand was used as a granular fill material. Both unreinforced and geosynthetic-reinforced embankments were investigated. Fully mobilized soil arching was first reached by lowering the trapdoor, and then a footing load was applied on the surface. Both static and cyclic loads were applied to simulate traffic loading. Pressure distribution, footing and trapdoor displacements, geosynthetic strains, and embankment soil movement were monitored during each test. The trapdoor test results show that the progressive displacement of the trapdoor affected the mobilization of the soil arching. Soil arching started to mobilize as the pressure on the trapdoor decreased and then deteriorated as the pressure on the trapdoor increased under soil self-weight after the trapdoor displacement increased to more than 2.5% of its width. However, the use of geosynthetic reinforcement prevented the deterioration of the soil arching and lowered the equal settlement plane height, although the trapdoor was lowered more than 4% of its width. The loading test results show that soil arching was not stable under surface loading without a geosynthetic, and the geosynthetic stabilized soil arching. To evaluate the progressive change of soil arching, soil arching ratio is defined as the ratio of the measured pressure on the trapdoor at a trapdoor displacement to the measured pressure on the trapdoor at no displacement. Soil Arching Degradation Pressure (SADP) is defined as an applied footing pressure required to eliminate soil arching (i.e., the soil arching ratio equal to 1.0). In the unreinforced embankment tests under static and cyclic loading, the SADPs were the same and equal to 54.0 kPa. Also, mobilizing soil arching under static and cyclic footing load (i.e., lowering the trapdoor under footing load) further decreased the SADPs to 45.0 kPa. The SADPs under static footing loading were increased from the unreinforced embankment to the reinforced embankment by 38.2% and 99.6% with the use of uniaxial and biaxial geogrids, respectively. Geosynthetic reinforcement further increased the SADPs under cyclic footing loading as compared to those under static footing loading by 17.5% and 9.13 % with the use of uniaxial and biaxial geogrids, respectively. Finally, the SADPs in the double layer of geosynthetic reinforcement tests were lower than those in the single layer of geosynthetic reinforcement tests.