Proposal of Michael P. McGuire of Virginia Tech University


Surface Deformation in Geosynthetic-Reinforced Column-Supported Embankments


Geosynthetic-Reinforced, Column-Supported Embankments (GRCSEs) have been used in soft ground conditions when there is a need to accelerate construction and/or protect adjacent facilities from the settlement that would otherwise be induced by the new embankment load. For instances when the embankment height is short relative to the clear spacing between adjacent columns, there is the risk of excessive levels of strain developing within the geosynthetic reinforcement. This condition can also lead to surface expression of the differential settlement that occurs between the columns and the soft soil at the foundation level. Currently, there is no consensus regarding procedures to design embankments to be safe against such surface deformation. The purpose of this research is to explore the mechanisms of deformation within a GRCSE using a combination of physical and numerical modeling. The ultimate goal of this research is to develop a set of design guidelines that mitigate the risk of this type of surface deformation.


A bench-scale test apparatus has been developed to study the relationship between surface deformation and differential settlement of the foundation for a variety of testing conditions. The apparatus consists of a circular open tank with an inside diameter of just over 22 inches. Three column arrangements can be used to study the interaction of the embankment material and geosynthetic reinforcement to differential movement between columns and the foundation soil: 1) a single 3 inch diameter column, 2) a 5x5 square array of 1.25 inch columns on a 3.5 inch spacing, and 3) a 5x5 square array of 2 inch columns on a 3.5 inch spacing. The apparatus is instrumented to capture the magnitude of penetration of the column(s) and as well as the load for the single column setup. Column displacement is controlled by a motorized jack that advances the column(s) into the base of the soil sample at a rate of about 1 inch in 20 minutes. At various increments of penetration, the column motion is stopped and the surface of the sample is scanned using a profiling device that consists of a non-contact laser distance transducer and a string-pot sensor. The laser device measures the distance from the instrument to the sample surface and the string-pot measures the horizontal position of the laser. The profiling system is capable of capturing the shape of the sample surface with a high degree of accuracy and resolution at four orientations around the circular sample tank.


A poorly-graded manufactured sand is used as the sample material and samples ranging in height from about 2 to 10 inches can be prepared dry at selected relative densities using the technique of air pluviation. The relative density of the sample can be selected with a reasonably good degree of confidence. In addition to testing samples of different height and relative density, the apparatus includes a vacuum system that provides the ability to test samples at sub-atmospheric conditions. The system is capable of applying up to 425 psf of equivalent overburden pressure on the surface of the sample.


Three types of biaxial polypropylene netting with different tensile stiffnesses were selected for use as the geosynthetic reinforcement. The netting was selected based on stiffness and aperture size criteria determined to be appropriate by approximate application of scaling laws for a 1:10 to 1:20 scale model. The influence of reinforcement stiffness on surface deformation is investigated by changing either the stiffness or number of the reinforcement layers. The bottom layer of reinforcement is placed inch above the base of the sample with inch separating any overlying layers.


The results from the bench-scale testing will be used to verify and calibrate axisymmetric and 3D numerical models of a GRCSE. The final 3D numerical model should allow for an unprecedented look at the stress distribution within the geosynthetic reinforcement in the vicinity of the pile cap, where stress concentrations are thought to occur but are generally ignored in current design procedures. The model should also provide the opportunity for better understanding of how geosynthetic reinforcement can be used to mitigate surface deformation in GRCSEs.