The objective of this case study is to evaluate the development of soil arching in an embankment being widened. In two test sections, earth pressures at different locations were monitored to reveal the load redistribution mechanism due to soil arching. Tensile forces in uniaxial plastic geogrids were measured to evaluate the performance of geosynthetic reinforcement in the widened embankment. The soil arching effect was quantified in terms of stress concentration ratio, soil arching ratio and geomembrane effect. Several existing methods were selected to compare with the measured results. The results of the field tests indicate that a two-dimensional plane soil arch with some eccentricity affects the fill load distribution on pile caps and subsoil between piles, and a realistic critical arch height of less than 2.0 m with a ratio of 1.4 times the pile clear spacing was identified. Although few existing methods could generate close results comparable to the measured values, most of the methods overestimated the load carried by the geosynthetic layer. The results also suggest that the foundation soil reaction underneath the geosynthetic layer should be taken into account in the load transfer mechanism of geosynthetic-reinforced and pile-supported widened highway embankments.

A well-designed field experiment was carried out to enhance the understanding of a geogrid-reinforced and floating pile-supported (GRFPS) embankment constructed on alluvial soil with medium–low compressibility (ASMLC). Two test sections with identical reinforcement solutions and fill load were chosen for field monitoring which lasted approximately 32 months. The measurements, including pore water pressure, settlement, soil stress, strain in geogrid, and lateral displacement during and after the embankment filling, were reported and the performance of GRFPS embankment was examined. The results indicate that the downward movement of the floating pile allowed the fill load to transfer from the pile (cap) to the subsoil and mobilised more counter support from the subsoil. Less soil arching effect and tensioned membrane effect were observed, and so were the stress concentration ratio and strain in the geogrid. The floating pile with limited embedment depth is sufficient to control both the total settlement and lateral displacement. The existing design procedures are expected to be overly conservative for GRFPS embankment constructed on alluvial soil with medium–low compressibility, and some recommendations are proposed to help with the design.