| description abstract | Geosynthetic-reinforced and pile-supported embankments (GRPSEs) encompass two core components of the aforementioned reinforced embankment and the underlying composite foundation. Previous centrifuge modeling of GRPSEs generally simplified or omitted the key aspects related to the core components, which can weaken the similarity of working performance between models and prototypes. In the current study, to accurately investigate the effects of the change of pile arrangement on the overall working mechanism of GRPSEs, two groups of centrifuge model tests on GRPSEs (Groups A and B with different pile cap sizes and pile spacings) were conducted by taking into account the similarity in design of materials, geometry, and pile‒soil friction. Without simplifications and omissions of the core aspects of the composite foundation, the test models involve the embankment slope, geotechnical reinforcement, large-diameter cast-in-situ concrete pipe (PCC) piles, and soft soil foundation. It has been verified that increasing the area replacement ratio by expanding the pile-cap area and enlarging the pile spacing is an economical way to reduce the final settlement and postsettlement, allowing for more loading to be concentrated on the PCC piles with a lower pile-soil stress ratio. Additionally, the influences of the 1g model preparation and g-ring loading method were analyzed and discussed. The subsurface settlements during the embankment loading stage, as well as the total rebound deformation, were recorded, and it was found that, based on the four-stage global settlement characteristics investigation, the settlement in the embankment construction stage and the final settlement are both overestimated. This study provides essential insights into the influence of modeling stress history on the deformation in centrifuge testing and serves as a reference for the pile arrangement design of GRPSEs. | |
