DEM analysis of the plugging effect of open-ended pile during the installation process

Highlights

• A novel DEM sample generation method was proposed.

• Plug formation is more sensitive to pile diameter and installation method.

• Plug resistance is highly related to IFR development.

• Plug is divided into dense zone, transition zone and shear zone.

Abstract

Mechanisms governing the sand plug behavior inside an open-ended pile are examined using the discrete element method. A series of numerical pile penetration tests have been conducted by considering the influence of soil density, pile geometry and installation method. A novel sample generation method, based on the replication of unit cell, is applied to produce a large and homogeneous sample efficiently. According to the soil deformation pattern, a “nose cone”, with the length about one pile diameter, has been observed beneath the small diameter jacked pile at the end of penetration. Plugging effect is shown to be more prevalent for jacked piles than dynamic installed piles. Also, larger penetration, smaller diameter and higher soil density all seem to promote plug formation, while the influence of wall thickness is not that obvious. This conclusion is later verified by the development of Incremental Filling Ratio (IFR) and porosity distribution. Furthermore, remarkable stress concentration has been observed at the lower part of the soil plug. The development of installation resistance indicates that jacking produces the largest resistance while dynamic installation methods ease pile penetration. Further analysis based on particle movements, contact force chain distribution and stress orientation provides a micromechanical perspective of the plug behavior. Finally, the plug resistance mobilization process at different plugging conditions and the formation process for soil plug are illustrated.

Introduction

Open-ended piles are frequently used as foundations for both onshore and offshore structures. During the installation process, part of the soil mass underneath will be pushed into the pile, forming a soil column called “soil plug” (Paikowsky, 1989; Randolph et al., 1991). The interplay of soil plug and pile, which is commonly referred to as plugging effect, poses a great influence on the pile installation resistance and bearing capacity, thus leading to the distinctive characteristics of open-ended piles compared to that of equivalent close-ended piles (Gavin and Lehane, 2003; Paik et al., 2003).

The plugging effect of open-ended pile has been widely discussed in literatures. Many researchers have reported that the soil plug state is highly dependent on the installation method, pile geometry and soil condition (Henke et al., 2008; Henke and Bienen, 2013; Lehane and Gavin, 2001). Hight (1996) proposed that a pile would plug when the soil plug length exceeded a critical height, the value of which was influenced by the pile diameter and soil density. It has also been reported that the maximum pile diameter for plug formation was 1.5m (Lüking and Kempfert, 2013; Jardine et al., 2005). De Nicola and Randolph (1997) found that the plugging effect was more remarkable in jacked piles than in driven piles, and increasing sand density may lead to plug growth in driven piles, but not in jacked piles. Furthermore, Henke et al. (2008, 2012, 2014) performed systematic analysis of the influence of pile installation methods via field tests, centrifuge experiments and numerical modelling; the research pointed out that for dynamic installation methods, the inertia of the soil mass inside the pile prevented the formation of soil plug, while for jacked piles, high horizontal stress could accumulate, thus facilitating plug formation. Ko et al. (2016) simulated the driving process of open-ended piles in sand using the Coupled Eulerian-Lagrangian (CEL) approach; parametric analysis revealed that the plugging effect was most influenced by driving energy, followed by pile diameter and pile embedment depth. Based on centrifuge tests, Wang et al. (2018) found that during the loading process of open-ended pile, the resistance components would take effect in the order of shaft resistance, annular resistance and plug resistance; also, the ultimate bearing capacity was reached synchronously with the mobilization of the maximum plug resistance. Li et al. (2019) proposed an innovative pile foundation design by setting up restriction plates inside the pipe pile; centrifuge tests showed that the restriction plates could promote plug formation, thus leading to higher bearing capacity compared to that of the conventional open-ended pile; however, the installation resistance (hammer blow counts) was not provided.

To better predict the bearing capacity of open-ended pile, some of the acknowledged design methods, such as HKU-12 (Yu and Yang, 2012) and UWA-5 (Lehane et al., 2005), have incorporated the plugging effect into the design guidelines by introducing correction factors related to plug length change (e.g. plug length ratio PLR, incremental filling ratio IFR, final filling ratio FFR). However, the complexity of plug behavior has made the above parameters hard to obtain and interpret in field tests. Meanwhile, the publication of more reliable database (Lehane et al., 2017; Han et al., 2019) and new design methods (Labenski et al., 2016b; Lehane et al., 2020) demonstrates that our knowledge of plugging effect is still advancing.

The above and related researches mainly focus on plug behavior at macro level, while microscopic investigation on plugging effect has been less favored. Regarding this, the discrete element method (DEM) provides a convenient and comprehensive alternative to investigate both the macroscopic and microscopic behavior of granular soil (Jiang et al., 2020; O'Sullivan, 2011). Zhou et al. (2010) conducted model tests and DEM simulations of open-ended piles jacking into sands and proposed to describe the soil plug formation in three stages: initial stage, formation of “active arch” and formation of “passive arch”. Thongmunee et al. (2011) carried out experimental study and DEM simulation of the push-up load tests, where sand plugs inside steel pipe piles were pushed upwards using a rigid platen; test results showed that the push-up force increased significantly with increasing aspect ratio and sand relative density. Li et al. (2019) simulated the sand plug behavior during the pile jacking process; remarkable stress concentration was observed at the lower part of the pile, where the soil plug was highly compacted. This phenomenon has been explained by Lehane and Randolph (2002) that the high vertical stress acting at the base of soil plug would lead to large radial stresses since the sand was confined. Also, Liu et al. (2019) found that most of the plug resistance was mobilized at the lower 2–3 diameters of the pile and the ratio of base resistance, including annular resistance and plug resistance, to total resistance increased with increasing pile diameter. Wang and Yin (2020) conducted 2D DEM simulation of the installation, operation and failure process of caisson foundation; it showed that a caisson foundation with a higher M/(DH) (homogenous moment to horizontal load ratio) would fail at much lower horizontal displacement or rotational angle.

In this paper, a systematic analysis of the plugging effect of open-ended pile is conducted using DEM. Several parameters, like pile installation method, pile geometry and soil density have been considered during the numerical modelling of pile penetration process. The main work of this paper can be summarized as follows: first, the sample generation and parameter calibration process are provided; then the model setup and pile installation process are introduced; afterwards, numerical results are presented to characterize the influence of different macroscopic parameters on the plugging effect and reveal the corresponding microscopic response.