Discussion of “Examining Setup Mechanisms of Driven Piles in Sand Using Laboratory Model Pile Tests” by Z. Zhang and Y. H. Wang

Abstract

The authors conducted a pressurized chamber test, which was equipped with an advanced bender element system and tactile pressure sensors to examine the mechanisms of pile setup in sand. The findings of the paper are of significant interest and the discussers would like complement these findings with those from other experiments on this subject. The authors conclude that, in line with Axelsson (2000), the changes of stationary radial and hoop stresses in close proximity to the pile shaft during setup are minor and the soil stiffness and its corresponding rate of growth following pile installation are significantly higher. The same conclusion was drawn by Lim and Lehane (2015) from field experiments on instrumented jacked piles, which recorded a very small variation of stationary radial stress (σ 0 rs) over a period of 72 days after installation. A comparison of the measured variations of shear stress with radial effective stress during the last installation pushes and during subsequent load testing indicated a significant time dependence of the radial effective stress change during shearing (Δσ 0 rd). The authors suggest that additional loading imposed on the surrounding soil because of pile insertion triggers the subsequent aging (or creep) process. While soil displacement (and thus the additional loading induced to the sand) could broadly explain why setup is observed in driven piles but not bored piles, and why the rate of setup is higher on large-displacement piles than on small-displacement piles, field experiments reported by Lim and Lehane (2014a) show that cycles imposed during installation and the nature of the installation process also need to be considered. In the discussers’ experiments (summarized in Fig. 1), the closed-ended driven piles induced a large displacement to the sand but also required a very significant number of blows/shearing cycles. These piles exhibited a very low capacity shortly after driving but the capacity then increased substantially over the aging period. The short term capacity of the driven open-ended (pipe) piles was not as low as they were easier to drive and also imparted lower levels of displacement to the soil. In addition, the closedended jacked piles were installed relatively gently using a hydraulic system and these showed a higher initial capacity but very little gain in capacity with time. The discussers also conducted laboratory chamber tests to investigate the effects of pile setup in sand (Lim and Lehane 2014b). Despite a more basic experimental setup than the authors’ version, these tests offer some useful additional qualitative information concerning parameters influencing the aging phenomenon. Apart from significant setup on driven piles that was captured in the experiments, the absence of setup in buried piles (which impart no disturbance to the in situ sand) was also modelled successfully. The original aging during the compression period as described by the authors was considered by Lim and Lehane (2014b) as an undesired sample aging effect. This effect was minimised by extending the sample compression period to 7 days (characterized by the changes of miniature cone tip resistance measured at different time intervals) before piles were installed so that its interference with the setup effect would be negligible. One limitation of the authors’work is the very short setup period (80 h) employed in the experiment. To utilize the benefits of aging effects for foundation construction, the performance over a longer period of time (weeks to months) needs to be established. Experimental observations reported by Lim and Lehane (2014a), in conjunction with several recent full-scale pile tests (Jardine et al. 2006; Gavin et al. 2013; Karlsrud et al. 2014) presented in Fig. 2, show that the complete evolution of pile setup may involve a delay or a lower rate of setup during the initial period after installation and a plateau when equilibrium under the new environment following pile installation is close to being reached. Therefore, simple extrapolation using basic semi-logarithmic or power-law functions from a short-term observation is not appropriate. It would be