0% found this document useful (0 votes)
252 viewsICP Design Methods
Uploaded by
PietrachCopyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
252 viewsICP Design Methods
Uploaded by
PietrachCopyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF or read online on Scribd
/ 103
ICP DESIGN METHODS FOR
DRIVEN PILES IN SANDS AND CLAYS
Richard Jardine, Fiona Chow, Robert Overy and Jamie Standing ICP design methods
for driven piles
in sands and clays
Richard Jardine imperial College London
Fiona Chow WorleyParsons Australia
Robert Overy Shei! Uk Ltd
Jamie Standing imperial College London
—
‘L! ThomasTelford ICP design methods for driven piles in sands and clays
Page 3
CONTENTS
1 INTRODUCTION...
2 BACKGROUND... eee
2.1 Rationale for developing new design approaches
2.2 Imperial College research programmes. ose 7 40
2.2.1. Research aims........ : Seat pee
2.2.2 Research phases... 1"
2.2.3 Field tests with instrumented piles... Aa 13
2.2.4 Parallel experiments with field-scale driven piles... oa 13
3 DESIGN METHODS FOR PILES IN SILICA SAND 18
3.1 Introduction... so oe 45
3.20 Shaft friction annem Scecee ee 15
3.2.4 Basic mechanisms ee oi 15
3.22 Evaluating short-term shaft capacity of single cylindrical piles... 16
3.3. Base resistance. Aa eee 22
3.3.4 Introduction nn eee a 22
3.3.2 Closed-ended piles Selita elt pe.
3.3.3. Open-ended piles a ete 23
3.4 Axial capacity of piles with non-circular cross-sections te 26
3.4.1. Recommendations for rectangular piles Beer 26
3.4.2 Recommendations for H section piles. cer 26
4 DESIGN METHODS FOR PILES IN CLAY Set 28
441 Introduction .... ee 28
42. Shaft friction .. Eilts ser Set elect : 28
4.2.1 Basic mechanisms. pata : 28
42.2 Evaluating shaft capacity of single piles after pore pressure equalisation 29
4.3. Base resistance oe ee 36
5 RELIABILITY OF THE DESIGN METHODS... 38
5.1 Additional entries to the Chow pile load test database 38
5.2 Reliability of shaft capacity predictions in silica sand... : 42
5.2.1 Shaft capacity database for silica sand 42
5.2.2 _Reliablity of the ICP shaft method in sand.. Bae eee oA:
53. Shaft capacity in clay : Sea a : 46
5.3.1. Shaft capacity database for clay... 46
5.3.2 Reliability of ICP shaft method in clays 46
5.4 Base resistance in sand 49
5.4.1 End bearing database in sand pei 4g
5.4.2 Degree of fit for the ICP end bearing method in sand. 50 Page 4
10.
“1
12,
5.5. Base resistance in clay
5.5.1 End bearing database for clay
5.5.2 Degree of fitfor the ICP end bearing method in clay
5.6 _ Independent analyses of ICP methods’ predictive reliability
5.8.1 Reliability for square and H section piles...
5.6.2 Checks by other organisations on reliability for cylindrical driven piles
5.7 Selection of safety factors in design
5.7.1 Foundation COVs in mixed soil profiles.
5.7.2 Relabilty calibrated against well-established design methods
5.7.3 Reliability in terms of probability of failure...
5.7.4 Safety Factors for the ICP methods
TIME EFFECTS IN SAND AND CLAY
6.1. Time effects in sand
6.2. Time effects in clay
6.3 Implications
GROUP EFFECTS IN SAND AND CLAY
7.1. Group effects in sand...
7.2. Group effects in clay...
EXPERIENCE WITH OTHER SOIL PROFILES.
8.1. Micaceous sands
8.2. Calcareous sands
83 Sits and low plasticity clays a Be
8.3.1 Assessing whether to apply clay or sand design criteria.
83.2 Low plasticity, low YSR, sensitive clays and clay-
8.4 Diatomaceous clays and mudstones
8.5 Layered soil profiles 3
CYCLIC LOADING AND SEISMIC ACTION
9.1 General
Its
9.2 Recent cyclic pile testing in sand and clay...
9.3 Axial capacity of piles driven in clay under seismic loading,
CONCLUSIONS.
10.1 Main points.
10.2 Check list for sands.
10.3 Check list for clays.
ACKNOWLEDGEMENTS.
REFERENCES...
APPENDIX A pelt
RING SHEAR TESTING METHODOLOGY
ICP design methods for driven piles in sands and clays
53
53
53
54
54
4
85
87
87
58
58
61
61
64
65
68
66
66
68
68
68
69
69
70
1
m1
72
72
74
a7
78
78
79
79
80
81
87
a7 ICP design methods for driven piles in sands and clays
Page §
A141 Principle of test 87
‘1.2 Specimen and interface preparation . 87
A1.3 Test procedure. 88
APPENDIX B. i a 7 91
CASE HISTORIES AND WORKED EXAMPLES FOR PILES IN SAND AND CLAY 9
B1 EURIPIDES 7 7 co 92
B1.1 Site conditions 7 7 92
B12 Test pile 92
B13. Pile capacity prediction... eo ah nee
B1.4 Comparison of calculated and measured capacity... 95
82 Pentre. Prat pelts 96
B2.1 Site conditions : 96
B22 Testpile 7 at Se 99
B23 Pile capacity prediction ied tenaeae 99
B2.4 Comparison of calculated and measured capacity. 101
APPENDIX C 102
LIST OF NOTATION 102 ICP design methods for driven piles in sands and clays Page 7
1 INTRODUCTION
In 1996 the UK Marine Technology Directorate (MTD) published a booklet by Jardine and Chow that
summarised a new integrated approach for calculating the axial capacity of tubular piles driven in
sands and clays. Axial capacity is often the governing criterion when designing driven piles and a
simplified treatment had been developed at Imperial College London through four consecutive PhD
Studies (Jardine 1985, Bond 1989, Lehane 1992 and Chow 1997). The work had been co-funded by
Industry, the UK's Health and Safety Executive and the Engineering and Physical Science Research
Council through MTD, with the main focus being on offshore pile behaviour. The recommendations
made by Jardine and Chow relied heavily on the earlier contributions made by Bond (1989) and
Lehane (1992), testing, extending, updating and applying their work in the light of the research by
Chow (1997)
‘The recommendations have been applied world-wide by the Authors and others in dozens of offshore,
marine and onshore projects. Applications have ranged from major new offshore platforms (including
13 such structures commissioned and operated by Shell Exploration and Production) to large bridges
and smaller scale foundations for light industrial facilities. Offshore Engineers often refer to the
procedures as the ‘MTD’ or Imperial College Pile (ICP) design method. However, the Marine
Technology Directorate no longer operates and we suggest that the acronym ‘ICP’ is now the most
‘appropriate.
This second edition, published by Thomas Telford Ltd, broadens and updates the original work. New
contributions are included and we emphasise the wide range of potential civil engineering apolications.
Reference is made to relevant research completed since 1996 and to lessons leamed through
Practice. Substantial new sections are included on choosing appropriate factors of safety, the selection
of geotechnical parameters, case histories, non-cylindrical pile shapes, ageing processes, a wider
range of soil types (including calcareous sands), group action, cyclic loading and seismic action. The
prediction of load-displacement behaviour, including the response to lateral loading, is not addressed
here. However, reference is made to other publications that describe how improved predictive
procedures have been developed from associated research at Imperial College
ur aim is to provide
‘+ Descriptions of the axial capacity calculation procedures that are sufficiently detailed and clear
to allow their application by suitably qualified geotechnical engineers,
+ Demonstrations of the theoretical and practical advantages offered in comparison with
conventional design methods
+ Evidence of the greater reliability and accuracy offered by the methods.
‘+ Worked examples with references to case
tories,
+ Commentaries on how pile shape, age, group action, cyclic loading and seismic action can
influence field performance
‘+ Guidance on applying the methods to a wider range of soil types. Page 8 ICP design methods for driven piles in sands and clays
Details of the experimental research, background theory and validation studies cannot be covered in
this deliberately short publication. Instead, a substantial list of references is provided and Appendices
are included covering (A) the methodology recommended for ring shear testing; (B) worked examples
of the ICP method's use in sands and clays; and (C) the notation and symbals employed, ICP design methods for driven piles in sands and clays Page 9
2 BACKGROUND
2.1 Rationale for developing new design approaches
The approaches applied by most practitioners to predict the axial capacity, Q, of displacement piles
are relatively unreliable. This may be demonstrated by comparing independent predictions. made by a
Statistically meaningful group of well qualified practitioners, with data from well conducted field tests in
‘predictions competitions’, or through database analyses in which a single individual (or team) applies
@ range of methods to a collection of load tests. The first approach relies on just one site, which might
ot be generally representative, while the second often has to rely on incomplete information drawn
from the literature, and may be biased by the subjective judgments of the single individual or team
concerned
Jardine et al. (2001) report a study of the first type that focused on piles driven in dense sand at the
Dunkirk research site in northem France. Full details of the site conditions and loading procedures
were published through a dedicated website. Axial capacity predictions were offered in confidence, to
an independent body, by a wide range of international consultants, researchers and specialists. Figure
1 shows the compression test capacity measured on site and the wide spread of predictions offered
which ranged from around one third of the measured capacity to about twice this value. The calculated
capacities Q. fell on average around 21% below the measured value Q,, and gave a Coefficient of
Variation (COV)' for (Q- /Qn) of around 0.53. Database studies by Briaud and Tucker (1988) and
Jardine and Chow (1996) show that even the best conventional approaches give COV of a similarly
high magnitude and may also be subject to substantial bias. Despite their limitations, predictions
competitions and database studies lead to similar conclusions: predictive reliability is generally tar
poorer than many practitioners recognise.
Pile load tests are specified in many projects to help mitigate the effects of predictive scatter.
However, this option is rarely available to offshore engineers and can be difficult to carry forward with
large piles in more general applications. Jardine and Chow (1996), and others since, considered how
well the procedures most commonly used by Offshore Engineers predict the capacities held in high
Quality databases. They found litle overall bias but report COV values as high as 0.5 to 0.7 that sit
uncomfortably with the relatively low safety margins (typically 1.5 to 2.0) that are commonly adopted
for offshore pile design. Jardine and Chow showed that the existing offshore methods are subject to
strong and systematic skewing of (Q. /Q,,) with respect to factors such as pile slenderness (L/D), sand
relative density (D,) or clay apparent over-consolidation ratio (YSR). Existing offshore methods may be
conservative in some cases, including low L/D piles in dense sands or high YSR clays, and non-
conservative in others, such as slender piles driven in loose sands or low YSR clays.
* The Coefficient of Variation (COV) is defined as the standard deviation, s, divided by the mean value
». In an ideal method y should tend to unity and the COV to zero. Page 10 ICP design methods for driven piles in sands and clays
6000
5000
4000
3000
z
1000
36035 32 11 17: «10: «18 «4 «13-307 «12 31 29534
Entry Number
Figure 1. Results from pile predictions competition based on tests in Dunkirk sand;
entries stacked in order of ascending total pile capacity estimate
Reports of piled structures experiencing difficulties due to axial capacity failures are rare. However, a
clear need exists to improve predictive methods to obtain economies, where possible, and enhance
Performance, and safety, in other cases. The implementation of improved methods needs to be co-
ordinated with any parallel developments in site characterisation techniques and the specification of
loading. The latter applies particularly in cases involving high levels of environmental load, and caution
's required when working with structural arrangements that impose unusual requirements on their
foundations.
2.2 Imperial College research Programmes
2.2.1 Research aims
The research carried out at Imperial College has sought to achieve: (i) a more fundamental and
thorough understanding of pile behaviour, and (i) practical design methods that capture the basic
mechanics of driven piles as simply as possible. The main tasks were to identity
+ How piles behave in different soils and layering sequences,
+ The scaling laws that relate the behaviour of models to that of full-scale piles.
* The effects on capacity of pile properties (dimensions, wall thickness, end conditions, surface
roughness, material hardness, etc.) and installation methods, ICP design methods for driven piles in sands and clays Page 11
+ Any changes in capacity and stiffness associated with time ater pile installation.
+ The response to different loading types, including group effects, cyclic loading and seismic
action.
* The controlling soil parameters that should be measured in site investigations.
The research reached a sufficiently developed state in 1996 for it to be applied practically. Ample
‘scope remained then, and now, for further discussion and research; several issues remain open to
academic discussion and potential improvement. However, as set out below, the ICP methods offer
‘substantially increased overall accuracy and hence tangible engineering benefits in improved reliability
‘and cost-effectiveness. They have also been widely applied and tested in practice since the mid
1990s.
2.22 Research phases
The research at Imperial College has taken place in five main phases. principally involving the sites
and profiles identified in Table 1 and Figure 2, but also supplemented by data gathered at other
locations ranging from Belfast to Mexico City.
The first phase of work involved developing the ICP instrumented piles and experimental procedures.
Multiple ICP tests and other experiments were then performed at the Building Research
Establishment’s (BRE) Canons Park test site. The research was summarised by Bond (1989) and
Bond and Jardine (1990, 1981).
Bothkernar
Pentre. —
eunkirk
Cartons
f
Loberne,
Figure 2. Locations of ICP test sites in UK and France Page 12, ICP design methods for driven piles in sands and clays
‘The scope was broadened in Phase 2 to cover tests in sand at the French Ponts et Chaussées’ test
site at Labenne, the BRE's stif til site at Cowden, and the Engineering and Physical Sciences
Research Council's (EPSRC) former national soft clay test site at Bothkennar. At each location an
advanced site investigation was performed, a field pile testing facil established, and a programme of
multiple (closed-ended) ICP tests carried out. The Phase 2 work was reported by Lehane (1992) and
Jardine and Lehane (1894); clear and striking results emerged from the experiments that allowed new
design approaches to be proposed for closed-ended piles (Lehane et al., 1993: Lehane and Jardine,
1994a, b). The tentative proposals of Lehane and Jardine (1994c) and Lehane et al. (1984) provided
‘many of the key elements of the later ‘MTD’ procedures.
The third phase reported by Chow (1997) and Chow and Jar
}e (1996) involved:
* Establishing facilities and performing advanced site investigations and multiple ICP tests at
the Pentre (clay-silts/laminated clays) LDP research site and at the Dunkirk ‘CLAROM’ dense
sand research site
* Interpreting and performing tests on full-scale driven open-ended piles (with diameters up to
760 mm) at the ICP sites to assess the effects of scale, installation methods and pile-end
conditions,
* Field experiments to assess pile group and ageing effects in dense sand.
* Using the above to refine the new approaches for closed-ended piles and extend the design
methods to cover open-ended driven piles.
+ Collating an up-to-date and critically approved database of full-scale pile tests that met
rigorous quality criteria.
* Using the above to calibrate and validate the new methods for a wide range of practical
applications
Phase 4 (Jardine & Standing, 2000) comprised research on full-sized piles performed at Dunkirk
between 1998 and 1999. Eight 456mm OD open-ended piles (six of which were 19m long, two 10m
long) were driven to examine the effects of cyclic loading, pile age and base grouting
‘The fi phase of relevant work at Imperial College work involved a number of smaller projects
conducted (often in collaboration with other groups) between 1997 and 2003:
* Studies by Thompson (1997) into driven pile capacity in calcareous sands and by Cowley
(1998) into the effects of pile shape on axial capacity,
+ Field tests with Trinity College Dublin (TCD) on square section concrete piles driven in soft
Belfast clay to examine group action, cyclic loading and pile ageing (see Lehane and Jardine
2003 and Lehane et al., 2004),
* Field tests in collaboration with the Building Research Establishment at Canons Park to
investigate long-term pile ageing effects in clay (Pellew, 2002),
* Research into the effects of pile shape, clay type and seismiclcyciic action for piles driven in
Mexico City (Saldivar-Moguel, 2002).
‘+ Further research into the interface shearing properties of clays and sands.
* Application of the pile design procedures to circumstances and combinations of soil conditions
that were not covered by the range of tests collated by Chow (1997), including the reuse of
pile foundations in construction projects. ICP design methods for driven piles in sands and clays Page 13
2.2.3. Field tests with instrumented piles
A central feature of the first three phases of research has been the development of the accurate and
reliable on-pile Imperial College Pile (ICP) instrumentation to study the pore pressures, radial total
stresses, local shear stresses and temperatures developed on pile shafts. Until reliable instruments
became available, the stress conditions surrounding driven piles had been open to conjecture. The
ICP gauges were mounted on 6 to 20 m long, 102 mm diameter, closed-ended” steal pipe piles for
intensive testing programmes in the geotechnical profiles summarised in Table 1. The ICPs were
installed by fast jacking, allowing comprehensive measurements of the effective stress conditions
developed close to the shafts to be made at multiple levels during installation, long-term equalisation
and load testing to failure. In particular, it was possible to make direct measurements of the residual
loads set up by installation, a feature that is often hard to assess in conventional pile tests
Detailed site investigations were performed at each site, involving In-situ tests and advanced
laboratory experiments. ‘Strain Path Method’ numerical simulations of the ICP tests performed at
Canons Park and Bothkennar were also carried out in conjunction with Professor Whittle from MIT as
described by Bond (1989) and Lehane (1992)
Table 1. Summary of Imperial College pile research sites
Site
Soil conditions
4. Canons Park
London Clay: stiff to very stif, high plasticity, Eocene marine clay; high YSR
2. Cowden
Cowden tl: stiff to very stif, lean, glacial lodgement til; high YSR
3. Bothkennar
Carse clay: soft, high plasticity, moderately organic. Holocene shallow-
marine/estuarine clay-sit ighty cemented: moderate YSR
/4, Labenne Dune sand: loose to medium dense, medium-sized, Holocene; low YSR
5. Pentre Glacio-lacustrine clay-sitt and laminated clays: very soft to fir, low plasticity, ow
YsR
6. Dunkirk Marine sand: dense to very dense, shelly medium-sized sand, Flandrian: low
Ysr
Note: Yield Stress Ratio (YSR) is the apparent OCR
2.2.4 Parallel experiments with field-scale driven piles
‘As mentioned above, less heavily instrumented open-ended driven tubular piles were driven and
tested at four of the ICP sites, providing data with which to check the potential effects on axial capacity
Of ple tip detail and installation method. Offshore scale open-ended piles had been tested at Pentre as
Part of the Large Diameter Pile (LDP) project described by Clarke (1993), while those at Dunkirk and
Cowden were installed for earlier projects run by the CLAROM group and BRE respectively.
* The use of closed-ended piles allowed more accurate and robust instrumentation to be deployed. Page 14 ICP design methods for driven piles in sands and clays
The Canons Park driven piles were installed by the Imperial College research team and the same
group worked with full-scale driven tubular plles at Dunkirk between 1998 and 1999 to study cyclic
loading, pile age and base grouting. Recent research on driven square section solid piles, performed
with Trinity College Dublin (TCD), has examined the effects of group action, cyclic loading and pile
ageing at the Kinnegar soft clay site in Belfast (see for example Lehane et al., 2004). Other projects
have investigated the effects of ageing in London Clay and pile shape, seismic action and cyclic
loading in Mexico City clay. ICP design methods for driven piles in sands and clays Page 15
3 DESIGN METHODS FOR PILES IN SILICA SAND
3.1 Introduction
The following paragraphs describe the methods proposed for evaluating the short-term static axial
capacity of single displacement piles in silica sands. Shaft friction and end bearing are covered under
separate headings. Attention is concentrated first on cylindrical piles, with either closed or open ends,
before discussing how capacity may be assessed for rectangular or H section piles. Subsequent
sections discuss the predictive methods required for calcareous and micaceous sands.
Recent research and experience shows that pile capacity varies with age and testing style. The
methods set out below are intended to predict the capacities that may be mobilised in slow maintained
load tests, conducted around ten days after driving’, on previously un-failed piles. The calculation
Procedures are intended to be compatible with modem pile testing practice, which includes pause
Periods that allow creep straining to stabilise between loading increments.
Section 4 sets out the equivalent procedures for piles driven in clay, while Section 5 offers an
assessment of how well both sets of the predictive equations compare with an updated assembly of
pile load tests. The later Sections (6, 7 and 9) consider the effects of age, group action and cyclic
loading. Section 8 comments on the methods’ use in micaceous and calcareous sands; silts and low
plasticity clays; diatomaceous clays and mudstones; layered soil profiles. Worked examples covering
both a sand and a clay site are included in Appendix B.
3.2 Shaft friction
3.2.1 Basic mechanisms
‘The ICP experiments at Labenne and Dunkirk showed that at failure the local shear stresses acting on
the pile shaft, r, follow the simple Coulomb failure criterion:
= oy tan &
‘The radial effective stress acting on the shaft at failure, depends on o'. the equalised value acting a
few days after installation (when pore pressures and radial stresses are relatively stable) combined
with any changes developed during pile loading. The 8
[Shear stresses in tension reduced by a further 10%
Note: Above recommendations apply to ()) silica sands, (I) piles with ocular cross-sections (ii)
capacities available in ‘first-time’ slow maintained loading tests conducted around ten days after
driving
Page 21 Page 22 ICP design methods for driven piles in sands and clays
3.3.1. Introduction
‘Theoretical investigations of deep penetration problems show that the surface shallow bearing
capacity factors and arbitrary upper limits employed in some design codes are unlikely to be
representative of field behaviour. Spherical cavity expansion solutions provide more useful analogues
for closed-ended piles, indicating how pressure-dependent dilation characteristics, non-linear stiffness
behaviour and anisotropic shear strength can all affect ultimate Q,; particle crushing also plays an
important role at the elevated stresses developed by piles in sand. These complex factors make
theoretical calculations both difficult to perform and relatively unreliable.
The CPT test produces direct measurements of in-situ resistance under conditions that resemble
closely those at a pile tip and this advantage is recognised in the Dutch and French national CPT
based design approaches (de Ruiter & Beringen, 1979 and Bustamante & Gianeselli, 1982). The
similar contained failure system and boundary conditions allow the CPT data to be used directly,
without having to decouple the full set of complex soil parameters that determine both problems. The
dp values (where Q, = q,xD‘/4) observed during penetration in the ICP tests at Labenne and Dunkirk
correlated directly with the g. traces, as did the values developed in static load tests. However, the ICP
tests could not show whether gi/a. varies with pile scale, or in-situ stress level. Nor could they offer
insights into the base resistances of open-ended piles
3.3.2 Closed-ended piles
Chow's assessment of work completed by others, and interpretation of the limited database of full-
scale measurements, led to the provisional recommendations given in Table 3 where qp is calculated
‘as set out in Step C1. The approach relies on CPT measurements, incorporates a scale effect and
distinguishes crucially between open and closed-ended piles. The aim is to predict the tip resistance
Q, available at settlements of up to DI10, allowing for its use in simplified calculation procedures that
do not account for the different rates of development of shaft and base resistance, or the possibility of
progressive failure. Higher base resistances will often be available in load tests that continue until an
Ultimate ‘plunging’ failure develops at much larger settlements. With closed-ended piles qu/q. is
considered to be less than unity, with the ratio falling with pile scale, expressed here by diameter.
Continuum based approaches do not predict any such scale effect for ultimate resistance and
potential explanations for scale effects involve (i) reference to particle scale processes such as
localised crushing zones or shear bands forming within the zone of contained failure beneath the pile
tip (Which could reduce the global influence of peak strength, stiness and dilation rates), (ii) the
definition of capacity as the load mobilised at a pile head displacement of D/10 and (ii) the effects of
variations in sand state over short intervals of depth,
The approach taken to select appropriate q. values from often variable CPT traces can have a
considerable influence on the pile capacity calculations. Following from Bustamante & Gianeseli
(1982), averaging a. over 1.5 pile diameters above and below the pile toe is recommended for Steps
C1 and D2, provided that (i) the variations in q. are not extreme and (i) the depth intervals between
the peak and trough qe. values are no greater than D/2. A q. value below the mean should be adopted
for design if these conditions are not met as the base capacity may be controlled by a localised failure
within any significant weaker layer. Equally, if the pile tip could possibly terminate within pile ICP design methods for driven piles in sands and clays Page 23
Giameters of a consistently less competent stratum, then account should be taken of any potential
reduction the lower layer might cause to end bearing capacity
Alternative interpretations nave been made of closed-ended driven pile databases assembled by
Chow (1997) and others. Randolph, Jamiokkowski and Zdravkovie (2004) propose that qu/a. may be
{aken as 0.4 when designing closed-ended driven piles, the value found from Table 3 (Step C1) when
D = 570 mm. Recognising their work, the lower limit given for qu/ae in Table 3 has been set at 0.3,
corresponding to a ‘cut-off diameter of 804 mm. However, few reliable data points exist for driven piles
with diameters greater than 600 mm and we are not aware of any test measurements on closed-
ended piles larger than 1.2 m diameter, There is an urgent need for well designed field tests to clarify
whether end bearing is affected by pile scale.
3.3.3 Open-ended piles
Open-ended tubular piles develop their base capacities through a combination of
‘+ The ‘internal skin-friction' transferred through the internal soil column.
+ Resistance beneath the annular area of the pipe.
The first component can make @ substantial contribution (at pile head settlements of 0/10) if strong
arching develops at the base of the internal pile soil column. The arching action is enhanced by high
values of 9, and strong dilation in the sand and at the internal pile interface, both factors correlate
with high relative density, However, arching is likely to be less effective as pile diameter increases’
and a simplified empirical plugging criterion is offered in Step D1, showing the minimum relative
density required to achieve a full arch in a pile of given diameter; Figure 6 presents the field data from
which this tentative criterion was inferred. The plugging action is likely to be highly sensitive to the soil
Conditions close to the pile tip and more research is required into the stress regime inside and around
the potential pile plug,
The base response developed by a fully arching plugged pile is softer than that of a similar closed-
‘ended pile because: (i) some local settlement is required to establish the arch, and (i) the soll beneath
the soll column has not experienced the same degree of pre-stressing and pre-stiffening during
driving. Lehane and Gavin (2001) report that the base response of jacked piles that plugged during
Griving (giving final Incremental Filling Ratios”, IFR = 0) was far stiffer than that of similar piles that
were driven with final IFRs ~ 1.0. They argue that a coring penetration mode leads to far smaller
residual loads, and that the latter reduce the resistance available at a displacement of 0/10. Recording
reliable IFR data is recommended: it can prove valuable to any subsequent re-assessmont of
capacity
While fully plugged and closed-ended driven piles might develop similar ‘plunging’ ultimate base
capacities after large settlements, Chow's interpretation of the available data shows that the ‘fully
Plugged’ end resistance mobilised at a settlement of D/10 is typically around half of that available for a
closed-ended pile: Step D2 is therefore recommended when assessing the utiisable open-ended
plugged capacity. The lower limit to qo/q. has been revised up to 0.18 (applying when D>0.90 m)
reflecting the findings of Lehane and Randolph (2002) and Randolph, Jamiolkowski and Zdravkovic
* This conclusion is reinforced by the CLAROM research at Dunkirk (Brucy et al, 1991) and the
GCG/Shell study described by Hight etal. (1996).
"IFR = 4H,/AL where Hy is the change in plug height and Al is the change in embedded pile length. Page 24 ICP design methods for driven piles in sands and clays
(2004) who propose qy/a_ = 0.2 as a minimum value for open-ended piles. Step D2 also specifies that
the fully plugged design capacity should be no less than the unplugged capacity calculated from Step
D3,
Itis recommended that Step D3 should be adopted if the pile is likely to core during static failure. Note
that the full CPT resistance q. is multiplied by the pile’s ‘annular area’, despite the different shapes of
the cone and pile annulus. Numerical analyses indicate that gx/g is less than unity below the annulus
and suggest a typical value around 0.7. However, the additional capacity computed by taking qs = de
offsets the internal skin friction components that may not be accounted for by Step B3, Table 2. The
additional ‘end bearing’ component is approximately equivalent to applying over the internal pile wall
the (maximum) external shaft shear stresses (that apply at the tip) up to a level 30 to 40 wall
thicknesses above the toe. Step D3 is considered to be marginally conservative, giving reasonable
‘agreement with both the very sparse field static field test database and driving data obtained with
large open-ended piles at several sand sites.
As before, the selection of appropriate q. values should account for the form of the CPT traces,
Because the postulated annular end bearing mechanism can develop over a relatively short depth
range of perhaps three pile wall thicknesses, the design value should reflect the weakest sufficiently
thick sub-layer within the soil unit in which the pile tip might credibly be terminated. Equally,
consideration should be given to the possibility of a more critical fully plugged failure mode developing
if @ generally weaker layer exists within 8 pile diameters of the expected final tip depth
20
© Plugaea °
18 | © Unslugaed
ost
Internal diameter (m)
os
Proved coring
oak \ :
o2zF ‘
Probable plugging
pp eee Jeo eee eee eee
o 0 2 % 4 50 60 70 80 9 100
Dr at ple toe (%)
Figure 6. Field evidence for the adopted rigid-plugging criterion in sand ICP design methods for driven piles in sands and clays
Table 3. Procedures for base capacity calculations in sand
c
|BASE CAPACITY OF CLOSED-ENDED PILES
oT
1Q, = qu D4
Ide = ge [1 = 0.5 log (D/Deer)}
Pile base resistance is related to the average CPT
Jond resistance at the founding depth, and the
relative pile and CPT diameters.
|A lower limit of q, = 0.30q. is suggested for piles
lwith D > 0.90 m.
[See text regarding g. selection
IDcor= 0.036 m,
BASE CAPACITY OF OPEN-ENDED PI
LES.
0}
ner < 0.02 (D,- 30)
Dinoe/Dcor € 0.083q./P,
[A rigid basal plug can develop during static
lloading if these criteria are satisfied
In the first equation Din is in metres and D, in %.
[Absolute atmospheric pressure, P,= 100 kPa.
D2
1s = qo IT Reser
19s =e [0.5 - 0.25 log (D/Dce)}
[Fully plugged piles develop 50% of the end
resistance of closed-ended piles of the same
|diameter (C1 above) after a pile head displacement
lof D/10. Two lower limits apply: () the fully plugged
lcapacity should be no less than the unplugged
|capacity (D3 below) and (ii) q should not fall below
(0.15g. (as predicted for D > 0.90 m). See text
regarding q, selection.
D3
}Q, = 0077 (Router - Pra)
se =9-
Unplugged piles are assumed to sustain end
bearing on the annular pile base area only with qua
|=q.; see text regarding parameter selection,
[Contributions from internal shear stresses are not
lconsidered explicity,
Page 25 Page 26 ICP design methods for driven piles in sands and clays
3.4 Axial capacity of piles with non-circular cross-sections
The ICP methods were originally developed to model tubular steel offshore piles. However, they are
fully applicable to driven pre-stressed concrete pipe piles, and closed-ended driven cylinders made
from steel or concrete. The rules are likely to be conservative for downwardly tapering, or stepped,
driven piles that may develop less marked decays of racial effective stress with f, and hence higher
capacities, than prismatic piles with constant cross-sections. However, modifications are required to
allow the ICP approach to be applied to other common driven pile geomeiries such as solid
rectangular section pre-stressed piles or H section steel piles.
Cowley (1998) considered how the ICP procedures might be applied to square and H section piles,
and also to sheet piles. His approach was to assemble databases of reliable load tests from the
literature and assess how well calculations predicted on alternative hypotheses corresponded to the
available field data, His database for square piles amounted to a total of 16 tests in sand and clay,
with a further 16 tests on H piles. These data populations are too low to define reliable statistics for
particular cases and Cowley considered it appropriate to seek rules that were:
+ compatible with the thrust of the ICP work;
+ as simple as possible;
+ equally applicable to sands and clays.
3.4.1. Recommendations for rectangular piles
Cowley's proposals for rectangular piles are sot out in Table 4 in which equivalent effective shatt
perimeters P and base areas A, are defined that may be used in combination with the ICP rules for
solid cylindrical piles. He found that R*, the term that determines the rate of decay ino’, with h in the
shaft ealoulations could be calculated (as with open-ended tubular piles) as R* = (Ay, Note also
that his database included no pile wider than 500 mm and that his base capacity formula includes no
scale effect, but takes an average q./q, = 0.7, which is equivalent to the ratio found from Table 3 Step
Ct for D = 143 mm
The proposals set out in Table 4 consider the general case of rectangular section piles (where breadth
b may not be equal to width @), involving a slight extension of Cowley’s results. As with cylindrical
piles, the calculations are intended to predict the capacities mobilised in slow maintained load tests
performed around ten days after installation. As detailed later in Section 5.6.1, the design rules
provide a generally good fit to the load test database.
3.4.2. Recommendations for H section piles
Table 4 also summarises Cowiey's proposals for H piles, defining their equivalent shaft perimeter P
and base area A, in terms of the geometry shown in Figure 7, following from De Beer et al. (1979). As
before, R" is calculated as [Ap/m|°*. None of the H section piles was wider than 535 mm and the base
capacity formula assumes q./q. = 1 with no scale effect. The expressions apply equally to sands and
clays. As discussed later in Section 5.6.1, the recommendations give a satisfactory degree of fit to the
database of 16 tests covering all material types ICP design methods for driven piles in sands and clays
Page 27
Table 4. Recommendations for ICP method application to rectangular and H section
piles: applies to both sands and clays; after Cowley (1998)
Parameter Square or H piles
Rectangular
piles
Dimensions width | o Geometry shown in Figure 7
Breadth | 6
Pile perimeter | P = 2(+b) P= 2(0+8)
Base Area | A, =d Ae = As + 2Xq(D-2T)
Where: A, = Area of H section steel
(A, definition for H piles after De
X= BB if 82 <(D.2N
