The uplift capacity of helical anchors normally increases with the number of helical plates. The rate of capacity gain is variable, considering that the disturbance caused by the anchor installation is generally more pronounced in the soil mass above the upper plates than above the lower plates, because the upper soil layers are penetrated more times. The present investigation examines the effect of the number of helices on the performance of helical anchors in sand, based on the results of centrifuge model tests. Uplift loading tests were performed on 12 different types of piles installed in two containers of dry sand prepared with different densities. The measured fractions of the uplift capacity related to each individual helical plate of multi-helix anchors were compared with the fractions predicted by the individual bearing method. The results of this investigation indicate that in double- and triple-helix anchors, the contributions of the second and third plate to the total anchor uplift capacity decreased with the increase of sand relative density and plate diameter. In addition, these experiments demonstrated that the variation of the anchor load–displacement behavior with the number of helices also depends on these parameters.

The use of helical anchors has been increasing in recent years; however, the current understanding of its behaviour does not allow for reliable estimates of uplift capacity and load–displacement response. The need for improvement of the design methods has motivated the present study, which investigates the effect of installation on the helical anchor response in very dense sand. This work is divided into three parts. The first part describes the results of centrifuge model tests and microtomographic analysis performed to examine the uplift response of a single-helix anchor and the soil disturbance caused by the installation process. The aim of these tests was to define appropriate hypotheses for the numerical prediction of helical anchor performance. In the second part, the load–displacement curve of a tested helical anchor was numerically simulated with the finite difference software FLAC^3D, considering the installation effect observed experimentally. In the third part of the paper, the adjusted numerical model was used in a parametric analysis to provide some guidelines for the design of helical anchors in dense sand. Additionally, from the results of the numerical investigation and experimental observations, two different approaches are presented for improving the accuracy of analytical predictions of helical anchor capacity. Both suggested methods showed good agreement with centrifuge and field testing data.

Rigid piles are used to reinforce soft soil foundations and thus increase embankment stability. This technique is improved by placing one or more geosynthetic reinforcement (GR) layers inside or at the base of the embankment. A series of 33 small scale models were tested using a geotechnical centrifuge. Soft soil settlement was imposed by the downward displacement of a tray. First, a series of models were prepared to examine how the load transmitted to the pile network increased with the embankment thickness. Using the same configuration, two identical models were prepared to successively test two different types of GR (geosynthetic reinforcement). Another approach was used to study how the external surcharges applied on the embankment affect load transfer. The results showed that, compared to the piled embankment, the load transfer increased for the case of the geosynthetic-reinforced pile-supported embankment (GRPSE) due to the membrane effect. The membrane effect is higher when the GR is stiff and its vertical distance from the pile is reduced. Numerical modelling reveals that, when another GR layer is added, the second GR has an effect only if punching is sufficient. However, the second layer did not reduce embankment settlement.