Performance Evaluation of Design Methods for Geosynthetic-Reinforced Pile-Supported Embankments

Geosynthetic-reinforced pile-supported (GRPS) embankments are widely used in soft soil regions to support roadways. Various design methods for GRPS embankments have been developed; however, engineering experience and recent research have shown that the performance of these design methods varies case by case. This paper systematically evaluates the performance of several empirical GRPS embankment design methods, which include the BS8006 method, the Nordic method, the EBGEO method, the FHWA method, and the CUR226 method. A preliminary assessment of these methods using three field case studies confirms that their performances differ from each other and from field measurements, the differences sometimes being quite significant. To enable a more systematic evaluation of the design methods, a validated finite element model (FEM) was first used to conduct a comprehensive parametric study considering typical soft soil conditions and various geometrical parameters, the results of which served as a baseline for the evaluation of the design methods. The efficacy, stress concentration ratio, maximum differential settlement, and maximum reinforcement tension were used as indicators for the method evaluation. The results show that, overall, the CUR226 method outperforms other design methods. The percentage difference between the CUR226 method and FEM can be as small as 5%, and the maximum value is 130%. It was also found that the Nordic method does not apply to small embankment heights, and that the BS8006 method provides a large overestimation of reinforcement tension when pile spacing is large (>4 ft) for all embankment heights considered.