Professional Documents
Culture Documents
Abstract
Pile jacking (also known as press-in piling) is an innovative technique for the installation of preformed piles, in which hydraulic rams are used to push the piles into the ground. Machines have been developed to allow this technique to be used quickly and efficiently. Advantages over traditional techniques such as dynamic pile driving include greatly reduced noise and vibration during installation. The use of pile jacking presents new challenges for the geotechnical engineer. Designers of jacked piles for any loading must predict the short-term axial force that will resist pile installation, so that a piling machine with a sufficiently high capacity to overcome this resistance can be selected. If unexpectedly high resistance is encountered, the piler may be unable to advance the piles to the required embedment. In many cases this installation resistance may be different from the long-term ultimate capacity which the pile can mobilise to resist applied axial loads. This paper describes a field investigation into the force required to jack piles into sand and silt. A series of field tests involving installing a 15m-long closed-ended tubular steel pile at a site in Japan is reported. During each test the total installation resistance was recorded and sensors in the pile measured the forces on the base and shaft. These resistances were found to be consistently less than standard predictions of the medium-term ultimate axial capacity from CPT (cone penetration test) data. This difference appears to be caused by rate-dependent partially-drained installation behaviour. Excess pore water pressure generated during installation reduces the installation resistance, making it lower than the drained medium-term capacity. The significance of this behaviour for designers of jacked piles is discussed, and techniques for predicting it are investigated.
Jackson, A. Pile Jacking in Sand and Silt. 2007. Previous research (3) has identified many advantages of jacked piling over traditional techniques, including: very low levels of noise and vibration; enhanced capacity due to the low number of stress cycles experienced by the soil around the pile during installation; enhanced stiffness due to the full pre-loading experienced by the soil around the pile during installation; the potential for extraction and reuse of temporary piles, bringing cost savings and environmental benefits; and the ability to use the measured installation resistance as an indicator of the pile capacity.
However, the use of pile jacking presents new challenges for the geotechnical engineer. Designers of axially-loaded piles of any type require a prediction of the long-term ultimate axial strength that can be mobilised to support the applied load. Designers of jacked piles must also predict the shortterm axial force that will resist pile installation, so that a piling machine with a sufficiently high capacity to overcome this resistance can be selected. Much previous research has considered the ultimate axial strength of piles, but this is often not the primary factor controlling pile design; serviceability stiffness criteria or the strength required to resist other loading conditions may be more critical. Less research has considered the installation resistance, but this dictates piler selection irrespective of any stiffness criteria or later loading conditions. In many cases, the short-term axial force resisting pile installation is different from the longer-term pile capacity. If unexpectedly high installation resistance is encountered, the piler may be unable to advance the piles to the required embedment. As discussed later, the resistance and strength also vary with time after installation, in processes known as set-up (increasing resistance) or set-down (reducing resistance). If set-up occurs after initial installation it may be impossible to advance the piles after a delay or to extract temporary piles after use (Figure 2).
a) Pile Installation
b) Increased Pile Capacity Topples Piler During Extraction Figure 2: Problems with Jacked Piling
2/8
tan
Equation 2
Embedded length L
Shaft Perimeter resistance
is the angle of friction between the soil and the pile shaft. , and are empirical parameters which require calibration. The UWA-05 method provides values of the parameters to predict the medium-term capacity of driven tubular piles in sand, calibrated against a large database of load tests. This is currently the only case for which the method is validated. However, the equations and parameters can be attributed to physical phenomena occurring during installation and loading, so it is expected that other installation, soil and loading conditions can be considered by appropriate variation of these parameters (5).
Qs
Diameter
Base resistance
Qb
One phenomenon which may cause the parameters to vary in soil conditions other than sand is partial drainage. In low-permeability soils, installation of a pile induces excess pore water pressure around the pile tip. The magnitude of the induced pressure is dependent on the installation velocity, and it affects the base and shaft resistance. For example, positive excess pore water pressures generated by partially drained compression of the soil under the pile base will reduce the effective stress in the soil around the pile, and thus reduce the shear stress which can be mobilised. Figure 4 shows how the pile installation resistance (point A) may therefore be different from the long-term drained capacity (point B), and may depend on the installation velocity , the soil property (the coefficient of horizontal consolidation), and the pile diameter . Whilst other phenomena such as viscous behaviour may cause the installation resistance to differ from the medium-term resistance, partial drainage is expected to be the primary cause of such behaviour in the soils considered in this project.
1 Normalised resistance: qb / qb,drained qt / qt,drained s / s,drained 0 C B Drained load test after dissipation A Set-up A Partially drained installation Pile installation resistance Pile mediumB term capacity CPT installation C resistance
Partial drainage also provides a mechanism for the variation of capacity with time discussed above. After installation the induced pore water pressures dissipate away, reducing their affect on the pile capacity. This is represented by the transition from point A to point B in Figure 4. Again, other phenomena such as creep, aging and corrosion can cause the pile capacity to change with time, but 3/8
Jackson, A. Pile Jacking in Sand and Silt. 2007. partial drainage is expected to be the primary cause of such behaviour in the soils and relatively short timescales considered in this project. Finally, the cone penetration test itself may also be partially drained, affecting its resistance (point C on Figure 4). In general, a jacked pile designer therefore needs to be able to predict point A, point B and the behaviour in between, using information such as that provided by point C. A modified form of the UWA-05 method, adapted to account for the partially-drained behaviour shown in Figure 4, is suggested. This paper concentrates on predicting the partially-drained installation resistance (point A) from partially-drained CPT data (point C).
Test Methodology
Field tests were conducted at the Takasu test site in Kochi, Japan during July and December 2006. 15m-long closed-ended cylindrical piles with an external diameter of 318.5mm were used. During each test, a pile was installed and extracted using an AT150 piler specially modified for these tests by Giken Seisakusho (Figure 5). The axial ground resistance was calculated using measurements from calibrated pressure cells in the hydraulic circuit driving the hydraulic cylinders of the piler, with a correction applied to account for the measured pile weight and hydraulic losses. Some of the tests used an instrumented pile in which a load cell in the pile base plate measured the base resistance separately from the shaft resistance, and pore pressure transducers measured the pore water pressures near the pile base. Each test took place at an undisturbed part of the site. Figure 6 shows the results of ground investigations conducted at the site. All the data suggests a transition at approximately 10.5m below ground; above this level the soil is primarily sand, and below it is primarily silt.
0m BG Made ground Silt Sandy gravel Silty sand 0 qt [MPa] 10 0 u2 [MPa] 0.6
Pressure cells in hydraulic circuit measure force exerted by piler Data to datalogger
Piler chuck grips pile Coiled wire extensometer measures penetration of pile
Figure 5: Test Arrangement
Silt
15.8
Further details of the test methodology and site conditions are given in (7) and (8). 4/8
Results
Figure 7 shows the results of the field tests on the instrumented pile. These suggest the following observations: The measured base resistance is very consistent between tests, suggesting that the site is laterally homogeneous. However, there is a much greater variation in the shaft resistance. Other research (7) has shown that the shaft resistance is sensitive to small discrepancies in the installation process. For example, the variation may be caused by horizontal movements of the piler chuck imposed by the piler operator during installation to correct any minor misalignment of the pile (8). Both the base and shaft resistance show a significant difference between the results in the sand and silt.
0 50 100 150 200
0 2 4
Qs [kN]
0 2 4 6 8 0 100 200 300 400 500
6 8 10 12 14
10 12 14
a) Base Resistance b) Shaft Resistance Figure 7: Results of Tests on Instrumented Pile
Sand Silt
Figure 8 shows the results of tests on piles installed at different velocities. As suggested by Figure 4, faster installations (in which the induced pore water pressures are expected to be higher) appear to mobilise a lower resistance. The long-term capacity after the excess pressures have dissipated is therefore expected to be higher than the installation resistance (set-up). Other tests not documented here did indeed show set-up occurring at the site (7). Pore pressure transducers on the instrumented pile also showed positive Qb+Qs [kN] excess pressures; as described previously, 0 200 400 600 this is consistent with a reduced resistance 0 Average installation resistance during installation. 2 In the following section, the influence of soil condition is considered in more detail for the base and shaft resistance.
Tip Depth [m]
4 6 8 10 12 of eight monotonic tests installed at v=20mm/s. Average installation resistance of eleven monotonic tests installed at v=30mm/s.
5/8
Analysis
The suitability of the modified UWA-05 method for predicting the installation resistance can be assessed by back-analysing the test results. Firstly, consider the base resistance. Figure 9 shows =Qb/Aqt 0 0.2 0.4 0.6 0.8 1 1.2 a plot of a rearrangement of Equation 1 using the 0 measured installation base resistance. A value of 2 the parameter = 0.2 is assumed. Ignoring the data for very shallow installation, where the ground 4 is expected to be variable, the results are reasonably consistent with an assumption of a single value of 6 the parameter for the sand stratum above 8 10.5mBG and another for the silt stratum below 10.5mBG. In both strata, the ratio of pile base to Sand 10 CPT resistance is consistently less than predictions Silt of the medium-term base resistance in sand from 12 the UWA-05 method and by other researchers (9). 14 This agrees with the general behaviour shown in Figure 9: Back-Analysis of the Base Resistance Figure 4; partial drainage reduces the pile resistance in comparison with the CPT resistance. The difference in the ratios between the sand and the silt is caused by the different relative positions of their CPT and pile installations on Figure 4.
Tip Depth L [m] Sand average Silt average
2 4
UWA-05
Back-analysis and predictions of the shaft resistance are more complicated than that of the base resistance, because the total shaft resistance is influenced by more than one stratum of soil. Each stratum has a different strength and drainage behaviour, and has had a different time for the dissipation of excess pore water pressures since their generation as the pile tip passed. However, an initial simplifying assumption is that the base and shaft resistances contributed by a given stratum of soil are influenced to the same extent by the local excess pore water pressure. This is logical, since it is the high stresses generated around the pile tip which cause both the base resistance and, to a large degree, the shaft resistance. Based on this approach, Equations 1 and 2 can then be combined to give Equation 3, which analyses the shaft resistance a tan / in terms of the measured base resistance in an 0 0.01 0.02 0.03 0.04 0 attempt to eliminate the effects of differences between pile and CPT installation: tan =
= =0
max , 2
UWA-05
6 8 Sand Silt
10 12
c=-0.2
14
Figure 10: Back-Analysis of Shaft Resistance
Figure 10 shows values of the parameter tan derived from Equation 3 using the measured installation resistances. Scatter in the shaft resistance makes accurate analysis difficult. The ratio is close to that for medium-term resistances derived from the UWA-05 method, because the effect of partial drainage has been eliminated, leaving only the soil strength parameters. The small variation with depth is primarily due to a variation in soil strength rather than 6/8
Equation 3
Jackson, A. Pile Jacking in Sand and Silt. 2007. in drainage condition. The assumption that the shaft resistance is affected by partial drainage to the same degree as the base resistance seems to give reasonable results.
7/8
Jackson, A. Pile Jacking in Sand and Silt. 2007. Piler operators should be aware of the possibility for set-up during delays to cause refusal or trapped piles, and partial drainage curves could be used to predict the significance of this behaviour for a new site. Piler operators should be aware that, somewhat counter-intuitively, faster installation causes a lower installation resistance.
Acknowledgments
The author would like to thank the following people for their contribution to the research: Dr D.J. White, Prof M.D. Bolton, Mr A.D. Deeks, Mr M. Gillard, Miss H. Dingle, Mr T. Nagayama, and all the staff and engineers at Giken who provided assistance, advice and support. This research was funded by Giken Seisakusho Ltd. as part of a long-term research collaboration with Cambridge University Engineering Department, which forms part of the activities of the International Press-in Association (www.press-in.org).
References
1. [Online] http://www.giken.com. 2. [Online] http://www.grandarcade.co.uk/Scheme/Construction/Vibration.htm. 3. White, D.J. and Deeks, A.D. Recent research into the behaviour of jacked foundation piles. Yokosuka . Proceedings of the International Workshop on Recent Advances of Deep Foundations. 2007. pp. 3-26. 4. Lehane, B.M., Schneider, J.A. and Xu, X. The UWA-05 method for prediction of axial capacity of driven piles in sand. Perth. Proceedings of the International Symposium on Frontiers in Offshore Geotechnics. A A Balkema, 2005. pp. 683-689. 5. White, D.J. A general framework for shaft resistance on displacement piles in sand. London : Proceedings of the International Symposium on Frontiers in Offshore Geotechnics (ISFOG). Taylor & Francis, 2005. pp. 697-703. 6. Silva, M.F. Numerical and physical models of rate effect in soil penetration. Cambridge University Engineering Department. 2005. PhD Thesis. 7. Jackson, A. The Setup of Jacked Piles. Cambridge University Engineering Department. 2007. MEng Thesis. 8. Dingle, H. The testing and analysis of jacked foundation piles. Cambridge University Engineering Department. 2006. MEng Thesis. 9. White, D.J. and Bolton, M.D. Comparing CPT and base resistance in sand. Proceedings of the Institution of Civil Engineers: Geotechnical Engineering, Vol. 158(GEI). 2005. pp. 3-14.
8/8