Liu, Vinck, Ushev and Jardine (2024) In-Situ and Laboratory Characterisation of Stiff and Dense Geomaterials For Driven Pile Analysis and Design

Soils and Rocks
www.soilsandrocks.com
ISSN 1980-9743
An International Journal of Geotechnical and Geoenvironmental Engineering ISSN-e 2675-5475
In-situ and laboratory characterisation of stiff and dense

geomaterials for driven pile analysis and design
Tingfa Liu1# , Ken Vinck2 , Emil R. Ushev3 , Richard J. Jardine2  Article
Keywords Abstract
Laboratory testing Integrated field and laboratory characterisation of geomaterial behaviour is critical to
Glacial till foundation analysis and design for a wide range of offshore and onshore infrastructure.
Dense sand Challenges include the need for high-quality sampling, addressing natural and induced
Chalk micro-to-macro structures, and applying soil and stress states that represent both in-situ
Driven piles and in-service conditions. This paper draws on the Authors’ recent research with stiff
Offshore Engineering glacial till, dense marine sand and low-to-medium density chalk, and focuses particularly
on these geomaterials’ mechanical behaviour, from small strains to failure, their anisotropy
and response to cyclic loading. It considers a range of in-situ techniques as well as highly
instrumented monotonic and cyclic stress-path triaxial experiments and hollow cylinder
apparatus tests. The outcomes are shown to have important implications for the analysis
of large driven piles under monotonic-and-cyclic, axial-and-lateral loading, and the
development of practical design methods. Also highlighted are the needs for approaches
that integrate field observations, advanced sampling and laboratory testing, numerical and
theoretical modelling.
1. Introduction offshore windfarms around the UK and northern European

coastlines where such strata are often encountered. This
The potential capabilities and practical values of advanced requires: (i) high quality sampling (Hight & Jardine 1993);
laboratory testing have been demonstrated through ISSMGE (ii) addressing both natural geomaterials’ properties and
TC101’s International Symposia (IS) from Hokkaido 1994 to potential alterations induced by pile driving and soil-pile
Glasgow 2019. Jardine (2014, 2020, 2023a) highlighted their interaction; and (iii) applying appropriate effective stress
application with large piles driven to support key offshore states and perturbing stress paths (Jardine, 2014). This paper
and onshore infrastructure, noting that geomaterial behaviour draws on contributions made to the PISA, ALPACA and
under in-situ and in-service conditions can be affected ALPACA Plus Joint Industry Projects (JIPs) by Liu (2018),
markedly by pile installation effects as well as operational Ushev (2018) and Vinck (2021). The experiments conducted
and extreme loading. Representative numerical analyses rely on the three formations considered below have contributed
critically on experiments that match these conditions faithfully to making significant improvements in offshore wind turbine
to provide the fundamental physical basis for constitutive foundation design.
model development and calibration (Zdravković et al., 2021). • Cowden clay, stiff, high Yield Stress Ratio (YSR,
The worldwide drive for renewable offshore wind energy or apparent OCR) sandy glacial till sampled at the
has prompted large-scale research projects to tackle offshore PISA test site, NE England;
wind foundation design challenges; those involving advanced • Dunkirk sand, fine marine predominately silica sand
soil characterisation include the PISA (Byrne et al., 2017), from the PISA site in Northern France, consisting of
PICASO (Byrne et al., 2020a), WAS-XL (Page et al., 2021) dense to very dense normally consolidated hydraulic
ALPACA and ALPACA Plus (Jardine et al., 2023b) projects. fill (average relative density DR ≈ 100%) and Flandrian
High-quality pile test databases have also been collated by sand (average DR ≈ 75%);
Yang et al. (2017) and Lehane et al. (2020, 2022). • St Nicholas at Wade (SNW) chalk, low-to-medium
Establishing the behaviour of dense marine sands, stiff density B2/B3 grade Margate and Seaford chalk from
glacial tills, Eocene to Jurassic stiff clays and chalks has been the Upper Cretaceous sampled at the ALPACA site
particularly significant in developing large concentrations of in Kent, SE England.
#
Corresponding author. E-mail address: [email protected]
1
University of Bristol, School of Civil, Aerospace and Design Engineering, Bristol, United Kingdom.
2
Imperial College London, Department of Civil and Environmental Engineering, London, United Kingdom.
3
Wood Thilsted, London, United Kingdom.
Submitted on August 23, 2023; Final Acceptance on February 19, 2024; Discussion open until November 30, 2024.
https://doi.org/10.28927/SR.2024.009323
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.
Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 1

In-situ and laboratory characterisation of stiff and dense geomaterials for driven pile analysis and design
An overview of the ground conditions at these sites fine-to-medium sand; see Figure 1. The sub-angular particles
is presented, before considering three topics for each comprise ≈ 85% silica, plus calcium carbonate (CaCO3) shell
geomaterial: fragments and other minerals that leave the soil slightly alkaline
• Mechanical behaviour from very small to large (Chow, 1997). Table 1 summarises index properties, after
strains, emphasising non-linearity, anisotropy as Kuwano (1999) and Aghakouchak (2015). The site’s qt profiles
well as rate-and-time dependency, interpreted within are illustrated in Figure 2b. Groundwater was encountered at
suitable frameworks; 5.4 m in 2014, 1.4 m deeper than that found by Chow (1997)
• Synthesis of in-situ and laboratory measurements; with hydrostatic conditions below. While rotary coring was
• The geomaterials’ response to drained or undrained conducted for Brucy et al.’s (1991) investigation, coring for
cyclic loading. PISA was unsuccessful due to disturbance and drilling fluid
Mention is made briefly of how soil-steel interface contamination. The laboratory testing therefore had to rely
behaviour is affected by stress level, time (ageing), displacement, on reconstituted bulk samples.
surface roughness, pore water chemistry and other conditions. St Nicholas at Wade (SNW) The chalk site was
Attention is also drawn to how field and laboratory testing characterised by Diambra et al. (2014) and Ciavaglia et al.
can be applied in analyses of impact driven piles subjected (2017) prior to Buckley’s (2018) and Buckley et al. (2018)
to general loading. investigations for the Innovate UK project and the exhaustive
ALPACA and ALPACA Plus studies (Vinck, 2021; Vinck et al.,
2. Geotechnical overview 2022; Jardine et al., 2023c). Pure (98.6% CaCO3) white
Margate Chalk was found down to the Barrois’ Sponge
Bed at 5.2 m Above Ordnance Datum (AOD) above the
2.1 Site profiles Seaford white Chalk. The chalk classifies as CIRIA Grade
B3/B2 (structured, very weak to weak, low-to-medium
Cowden site encounters Bolders Bank formation density) over the depths of interest, with discontinuity
glacial tills, whose deposition involved cycles of ice apertures < 3 mm and fractures spaced at 60 to 600 mm.
loading and horizontal shearing and/or glaciotectonic Predominantly vertical linear micro-fissures were identified
passive pressures (Davies et al., 2011). The till comprises at all depths with ≈ 10 to 25 mm spacings. The ALPACA
principally stony, sandy and silty clay units that under-drain team located the (fresh) water table ≈ 0.9 m AOD, with ±
into two sand layers (Powell & Butcher, 2003; Ushev, 2018;
Zdravković et al., 2020a; Ushev & Jardine, 2022a). While
a water table is encountered at 1.0 m depth, pore-pressures
become under-drained at depth. The till’s index properties
are summarised in Table 1 and Figure 1. Figure 2a presents
typical profiles of corrected Cone Penetration Test (CPTu)
resistances (qt); the occasional spikes indicate hard stone
inclusions. Triple-barrel, wireline, Geobore-S rotary coring
and block sampling campaigns retrieved a sufficient stock
of high-quality intact specimens, although the till’s stony
nature and 8% gravel content limited the core recovery
rate to ≈ 42%.
Dunkirk conditions consist of mainly normally
consolidated marine Flandrian sands down to the Ypresian
Eocene marine clay found at 30 m (Brucy et al., 1991; Chow,
1997; Jardine et al., 2006; Zdravković et al., 2020a). Three
metres of hydraulic sand fill of the same origin were placed
over the Flandrian sands in the 1970s without compaction Figure 1. Particle size distributions of Cowden till, Dunkirk sand
or surcharging. Particle size analyses indicate uniform and SNW chalk over the depths of interest.
Table 1. Typical index properties of the three geomaterials.

Material D50 (mm) Gs wc (%) wP (%) wL (%) ρbulk (g/cm3)
Cowden glacial till 0.01-0.03 2.707-2.723 16.1-20.2 14.0-23.0 32.0-45.0 1.95-2.25
Dunkirk sand 0.21-0.28 2.65-2.66 5-7 (above WT) NA NA 1.74 (above WT)
22.7 (Below WT) 2.03 (below WT)
SNW chalk 0.003 2.71 27.9-33.0 22.3-23.9 30.2-32.0 1.85-1.98
Note: ‘NA’: Not Applicable.
2 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al.
Figure 2. Corrected cone resistances (qt) profiles for (a) Cowden, (b) Dunkirk and (c) St Nicholas at Wade (SNW) sites.
0.25 m variations. The chalk grains classify as fine silts (Vinck et al., 2019). Liu (2018) and Liu et al. (2020) provide
(D50 3-4 μm) with 1.43 Mg/m3 to 1.53 Mg/m3 Intact Dry full details of the Hollow Cylinder Apparatus (HCA) and
Densities (IDD) and a 0.91 average liquidity index; see Bishop ring shear apparatuses employed in parallel testing
Table 1. campaigns; supplementary experiments were undertaken at
Typical CPTu profiles are presented in Figure 2c, other research and commercial laboratories.
showing ‘spikes’ in thin flint bands. Significant chalk
de-structuration beneath and around CPT tips leads to 3. Cowden till
excess pore pressures up to 10 MPa at u1 (face) piezocone
positions (Buckley, 2018) and ≈ 4 MPa at u2 (shoulder) The PISA Cowden campaign first established profiles of
locations. Friction sleeve resistances indicate the chalk’s index properties and oedometer test outcomes. Comparisons
partial de-structuration towards the soft ‘putty’ found between the 1-D compression behaviours of natural and
around driven pile shafts. CPTu dissipation tests generally reconstituted specimens indicated a dense and insensitive
showed t50 times < 10 s from which Vinck et al. (2022) natural fabric. Initial triaxial tests also provided information
interpreted in-situ consolidation coefficients. Geobore-S on the till’s initial linear elastic response through to large-
rotary coring retrieved samples to ≈ 16 m depth, although strain critical states that fed into finite element modelling.
the chalk’s fractured and brittle nature limited the solid Liu (2018) and Ushev (2018) extended the scope to investigate
core recovery rate to 51% and a rock quality designation the till’s anisotropy, strain rate and pressure dependent
of 32%. Large block samples were taken from a 4 m deep monotonic shearing behaviour, as well its undrained cyclic
excavated pit. loading response, while Ushev & Jardine (2022a) explored
the till’s large-displacement soil-soil and soil-steel interface
2.2 Test equipment shearing characteristics.
The Imperial College London laboratory testing employed 3.1 Index and mechanical property profiles
fully instrumented automated Bishop-Wesley type triaxial
cells for 38 mm and 100 mm diameter specimens. Liu et al. The Cowden till comprises predominantly low plasticity
(2022a) detail the equipment and their stress and strain sensors’ (plastic limit ≈ 14-23%) clay over the depth of interest; the
resolutions and precisions. The LVDTs employed to measure 1.95 Mg/m3 – 2.25 Mg/m3 bulk densities reflect its gravel
local strains offered resolutions ≈ 0.1 μm. Mid-height sensors content. Figure 3a plots the 1-D compression yield stresses,
tracked local pore water pressures in the Cowden tests and σvy′, and Yield Stress Ratio (YSR, or apparent OCR = σvy′/σv0′)
lubricated top and base platens were deployed in most tests profiles from stage-loaded and constant rate of compression

Figure 3. Cowden site profiles of (a) 1-D yield stress σvy′ and YSR and (b) Su from triaxial compression (TXC) and extension (TXE) tests;
all tests performed on high-quality intact specimens (modified from Liu et al., 2020; and Ushev & Jardine, 2022a, b).
oedometer tests. The high YSRs noted over the upper 4 m 2023a) report HCA experiments that investigated the till’s
reflect complex glacial and post-glacial processes. Although anisotropy more rigorously.
Powell & Butcher (2003) interpreted K0 values exceeding
2.5 from various empirical procedures, Jardine (1985) argued 3.2 Triaxial monotonic shearing
that glacial lodgement till deposition leads to lower K0 values
that are more compatible with the networks of open vertical
fissures observed especially within the desiccated and more 3.2.1 Stiffness
stony upper sections. Therefore, the laboratory triaxial tests
and analyses limited K0 to 1.5 within the upper 7 m and unity The triaxial tests addressed behaviour from strains of
below this depth. 10-6 up to ultimate failure with deviatoric strains εs >25%.
The peak undrained triaxial compression and extension The natural Cowden till only exhibited elastic behaviour
strengths (Su) of natural specimens re-consolidated to in-situ under small stress and strain increments, no greater than a
stresses are reported in Figure 3b. The profiles exhibit similar few kPa and 0.002% respectively, that remained within the
shapes that, like the oedometer σvy′ profile in Figure 3a, till’s Y1 kinematic yield surface (Jardine, 1992). Undrained
differ markedly from the patterns expected in gravitationally vertical Young’s moduli established by regression analyses of
consolidated waterborne sediments after monotonic geostatic these initial portions led to the Eusec profile in Figure 4a along
loading and unloading. The compression profile indicates with secant stiffness profiles for axial (or εs) levels of 0.001%,
an average CPT Nkt ≈ 18, while the average SuTXC/SuTXE ratio 0.01%, 0.1% and 1%, which indicate nearly parabolic trends
is 1.25. Liu et al. (2020) argue that the latter ratio does not with depth. Analysis confirmed that Eu scales with p′ raised
reflect the till’s anisotropy because the extension strengths are to a partial power, as expressed in Equation 1, where pref is
affected by necking and different Lode angles (or b values) atmospheric pressure. The exponent β can be taken as 0.5 and
applied in compression and extension. Liu et al. (2020, the α values are largely constant with depth for each strain level;

Liu et al.
Figure 4. Cowden stiffness profiles of: (a) triaxial compression secant stiffness Eusec at axial strains of 0.001%, 0.01%, 0.1%, and 1%;
(b) normalised triaxial secant stiffness [Eusec/pref]/[p′/pref]0.5 at four axial strain levels, applying Equation 1 with β = 0.5; and (c) elastic
shear stiffnesses from laboratory and field geophysical measurements, with EuTXC/3 profile from triaxial compression tests (modified
from Ushev & Jardine, 2022a).
see Figure 4b. Similar traces were observed for the interpreted
triaxial tangent Young’s moduli (Eutan) (Ushev, 2018).
β
 p' 
E u = α pref   (1)
 pref 
 
Figure 5 shows the degradation of normalised secant

stiffness [Eusec/pref]/[p′/pref]0.5 with strain, with degradation
being steeper in extension than in compression.
As discussed in greater detail later, Cowden till shows
cross-anisotropic stiffnesses. Figure 4c plots laboratory
bender element Gvh, Ghv and Ghh profiles for intact specimens
along with in-situ seismic CPT (SCPT) Gvh data from
Zdravković et al. (2020a). The laboratory and SCPT Gvh
traces plot relatively close together, although the field values Figure 5. Typical degradation trends for normalised vertical secant
are (on average) slightly higher. Also shown in Figure 4c is undrained Young’s moduli from anisotropically consolidated triaxial
the equivalent isotropic octahedral shear moduli Goct taken compression and extension tests on intact Cowden till from upper
as EuTXC/3 which falls far below the bender element Gvh and layers (modified from Liu et al., 2020).
Ghh trends because of the till’s anisotropy.
3.2.2 Large-strain yielding and rate-dependent behaviour around 1.0, equivalent to ϕcs′ = 25.4°, c′ = 0 and applied in
constitutive modelling of the till. Figure 6 shows typical
Cowden till exhibits ductile, strain hardening, undrained effective stress paths from compression (TXC)
triaxial compression behaviour. Stable failure states were and extension (TXE) tests in q-p′ (q = (σv′ - σh′), p′ = (σv′ +
reached after imposing large strains, with the specimens’ 2σh′)/3) coordinates. Also indicated are large-scale ‘Y3 yield
barrelling without any evident shear banding. These points’ interpreted as the points (at q ≈ 100 kPa) where the
points were interpreted as critical states with MTXC = q/p′ effective stress paths rotated from their initial negative dp′/

dq ratios (that reflect the till’s elastic stiffness anisotropy) established the till’s stiffness anisotropy. A suite of undrained
to follow paths with positive, dilative, dp′/dq directions. HCA ‘αdσ-controlled’ experiments that achieved five final σ1 axis
While the TXC tests progressed to stable critical states, all orientations with respect to the vertical (αf values) spaced
TXE specimens developed necking failures which lead to equally between 0 and 90˚. HCA simple shear (HCASS) and
apparently low peak shear strengths and post-peak softening. HCA triaxial compression tests (HCA TXC) performed on
A notional Hvorslev surface is shown that runs from the block sample specimens from 2.9 m established in this way
final TXC critical state point back to the no-tension cut-off the till’s anisotropy up to shear failure. The αdσ-controlled
and encompasses the large strain undrained effective stress and HCASS tests were controlled to keep the intermediate
paths. Similar critical states and Hvorslev surfaces were principal stress parameter (b) close to 0.5.
assumed to apply in extension.
Noting that rapid lateral pile loading tests at Cowden 3.3.1 Cross-anisotropic elastic stiffnesses
delivered markedly higher peak loads and stiffness than
slower stage-loaded experiments (Byrne et al., 2020b), Ushev Treating the till’s linear elastic behaviour as cross-
(2018) investigated the till’s rate-dependency. Triaxial tests anisotropic and strain-rate independent allows dynamic and
run at axial strain rates of 5%/day, 50%/day and 500%/day static measurements to be considered as compatible and
showed Su and secant stiffness increasing by ≈ 6% and 10% inter-comparable. The cross-anisotropic stiffness matrix has
respectively per tenfold rate change. Compatible rate trends five independent parameters, Evʹ, Ehʹ, νhvʹ (or νvhʹ), Gvh, Ghh (or
and reversible isotach behaviour were observed in other tests νhhʹ) that can be determined experimentally (see for example
that alternated between 5% and 500%/day shearing rates. Kuwano & Jardine, 2002). Three independent parameters,
Evu, Ehu and Gvh apply under undrained conditions that can
3.3 Anisotropic stiffness and shear strength be established from the drained set (see Brosse et al., 2017)
characteristics or the combined drained and undrained probing approach
proposed by Nishimura (2014) that places emphasis on the
Differences observed in the till’s triaxial compression higher quality axial strain measurements.
and extension shear strengths and the triaxial and field stiffness The profiles with depth of Ghh/Gvh, Ehʹ/Evʹ and Ehu/Evu
profiles in Figures 3 and 4 prompted Liu et al. (2020) to derived from Liu et al.’s (2020) triaxial and HCA tests plotted
investigate the till’s stiffness anisotropy. Suites of triaxial in Figure 7 show a generally consistent trend for a stiffer
probing tests on samples from 4 depths applied drained vertical response under horizontal than vertical loading, which is
(dσvʹ), horizontal (dσhʹ) and undrained vertical (dq) stress
increments (of typically 2 kPa) at rates of 0.1-0.2 kPa/hour
Figure 6. Typical undrained triaxial compression and extension

effective stress paths for Cowden till samples from 3.0 to 3.5 m
depth (Axial shear strain rate: 5%/day) (modified from Ushev & Figure 7. Anisotropic stiffness ratio profiles for Cowden site
Jardine, 2022b). (modified from Liu et al., 2020).

Liu et al.
important to recognise when analysing lateral pile loading.

Also shown are Ghh/Gvh and Ghh/Ghv ratios from in-situ down-
hole and cross-hole shear velocity tests by Powell & Butcher
(2003); stiffness anisotropy clearly diminishes with depth.
3.3.2 Non-linear stiffness characteristics over full strain

range
Following Brosse et al. (2017), the αdσ-controlled

Cowden HCA tests allowed undrained vertical and horizontal
Young’s moduli and torsional shear stiffness to be tracked
over a wide strain range. The till’s marked elastic-plastic
stiffness degradation under HCA conditions is demonstrated
in Figure 8 by plotting Evu and Ehu against vertical strain
εz and average horizontal strain (εr + εθ)/2, respectively.
Despite showing initial scatter, the moduli exhibit consistent
degradation trends. Narrow spreads apply to each component,
suggesting that the cross-anisotropic stiffness components
are relatively insensitive to the shearing path followed.
The two HCA triaxial tests, which employed pre-shearing
b = 1.0 and ultimate bu = 0, showed overlapping Evu trends
that fall slightly below those of most b = 0.5 tests. Overall,
the Ehu curves plot above the Evu data trends, confirming that
stiffness anisotropy persists as strains grow.
When taken to failure, under nominally plane strain
conditions, in HCA tests, the highest shear strengths developed
when σ1 acted horizontally (b = 0.5, αf = 90˚) and the lowest
when σ1 acted vertically; the shear strengths specimens
applying at intermediate αf values spanned these limits,
confirming that any Su anisotropy inferred from comparing
triaxial compression and extension strengths (see Figure 3b)
Figure 8. Cowden stiffness degradation from HCA tests: (a) undrained
would be misleading.
vertical secant Young’s moduli Evu against vertical (axial) strain;
(b) undrained horizontal secant Young’s moduli Ehu against mean
3.4 Response to undrained cyclic loading horizontal strain (modified from Liu et al., 2020).
Cyclic loading, which can be critical for offshore

structures, was addressed in the PISA field pile testing. Ushev
(2018) and Ushev & Jardine (2022b) supported this work qmean/p′in-situ space, as shown in Figure 9. Five of the higher
by investigating Cowden till’s response to undrained triaxial amplitude cyclic tests failed after the indicated number of
cyclic loading, considering multiple nominally identical cycles (Nf). All other specimens remained un-failed when
samples from 3.4 mbgl re-consolidated to in-situ stresses q0 = loading halted after 1500 to 3500 cycles. Interpreted contours
-25 kPa, p0′ = 67 kPa. Mean (qmean) and cyclic (qcyc) deviatoric are shown of the conditions under which failure can be
stresses, in the -25 to 200 kPa and 10 to 150 kPa respective expected at Nf = 10, 100 or 1000. A stable zone is indicated
ranges, were applied to populate the qmean-qcyc interactive where cyclic failure is not expected within 1000 cycles, and
loading plane and investigate how the combinations affected regions where failures occurred by either (i) large cyclic
the evolving trends for effective stress, cyclic strain and and average strains that led to “abrupt failure” or (ii) mean
stiffness, damping ratios and any cyclic failures. strains accumulating and samples “creeping towards failure”.
3.4.1 Number of cycles to failure and cyclic interaction 3.4.2 Evolution of effective stress, cyclic strain and
diagram stiffness
The influences of the mean qmean and cyclic qcyc deviatoric Cowden till’s cyclic effective stress path and overall
stresses were considered in both (i) the conventional total cyclic loading behaviour varied systematically with the
stress qcyc/(2Su) and qmean/(2Su) interactive loading space and normalised loading parameters. Typical ‘stable’ and ‘unstable’
(ii) alternative ‘effective-stress’ normalised qcyc/p′in-situ and responses are illustrated in Figure 10 for specimens cycled

under qcyc = 25 kPa and a single test involving qcyc = 150 kPa and subsequently progressed to fail within 1000 cycles.
from ‘in-situ’ qmean = -25 kPa. Also shown are the no-tension One example is given in Figure 10b. The final stages of
lines, the Hvorslev surfaces and M lines identified from the ‘unstable’ experiments could involve axial strain rates up to
static tests. Figure 10a illustrates the stable outcomes of 200 times higher than the reference monotonic experiments,
experiments involving qcyc = 25 kPa (qcyc/(2Su) = 0.1). Shifts explaining why they were able to sustain transient conditions
can be seen from an initially “dilative” drift trend to one that could not be sustained under slow monotonic loading.
indicating slightly “contractive” behaviour. In contrast, the The specimens’ p′ values fell more significantly on unloading
higher amplitude qcyc = 100 and 150 kPa (qcyc/(2Su) = 0.4 and than they rose during loading, leading to an overall contractive
0.6) experiments conducted from qmean = -25, 50 and 125 kPa trend with the paths heading leftward towards the origin,
all initially violated the ‘slow shearing’ Hvorslev surface developing irregular loops and broad “butterfly-wing” shapes.
Figure 9. Contours of number of cycles to failure for intact Cowden till in normalised stress space (modified from Ushev & Jardine, 2022b).
Figure 10. Effective stress paths of Cowden till under undrained cyclic triaxial shearing: (a) typical stable response under low amplitude
cycling; (b) unstable cycling response under high qcyc level of 150 kPa (Ushev & Jardine, 2022b; Liu et al., 2023a).

Liu et al.
Typical patterns of permanent axial strain (εmean) effective stress point; Jardine (1992). The first elastic,
accumulation with N are plotted in Figure 11a on logarithmic Y1, region is approximately elliptical in shape and has a
scales. The strains developed under the lowest (10 kPa) cyclic maximum height of only about ± 2.5 kPa in-situ for the
amplitudes tended to stabilise after 100 to 1000 cycles to cyclic series’ specimens. The second Y2 kinematic yield
rates comparable to the slow residual creep rates (< 0.005%/ surface (KYS), defined here as the threshold limit to: (i)
day) applying before cycling started. Empirical power law stable effective stress conditions that involve less than 5%
relationships were found that relate the strain development of variation in p′ over 2000 ± 500 cycles and (ii) practically
stable and metastable tests to their normalised cyclic loading constant cyclic stiffness. Figure 12 marks the approximate
parameters. Figure 11b presents the cyclic peak-to-trough boundaries of the Y1 (red circles) and Y2 (black ellipses)
secant stiffnesses trends for tests performed at in-situ qmean. surfaces that surround each qmean point. The Y2 surfaces’
The tests that manifested stable effective stress loops showed vertical dimensions diminished as the pre-loading effective
stiffness either growing slightly with N or falling modestly conditions approached the Hvorslev surface, even though
before stabilising; unstable cycling response led to cyclic p′ increased as the samples moved along this undrained
stiffness falling sharply with N. loading path. While cyclic loading within the initial
Y2 kinematic yield surface does not appear deleterious,
3.4.3 Cyclic kinematic yield surfaces paths that engage Y2 on each loading and unloading stages
dissipate far more energy and provoke marked stiffness
The cyclic responses observed in the stable and degradation and strain accumulation.
metastable tests can be related to two kinematic yield
surfaces that, once engaged, move with the current 4. Dunkirk sand
The Dunkirk PISA triaxial testing focused initially on
calibrating the Taborda et al. (2014) state parameter-based
bounding surface plasticity model employed to design and
model lateral pile tests. The research scope extended later to
consider stiffness anisotropy and long-term drained cyclic
loading behaviour, again aiming to support cyclic pile tests.
Liu (2018) and Liu et al. (2019a, b) report on the sand’s
interface shearing behaviour.
Figure 11. Evolution trends for (a) mean (permanent) axial strain Figure 12. Y2 kinematic yield surfaces for Cowden till interpreted
and (b) cyclic secant stiffness against number of cycles for intact from cyclic undrained triaxial tests at four mean deviatoric stress
Cowden till specimen cycled from in-situ mean qmean = -25 kPa (qmean) points along a slow compression test path to failure (modified
(Ushev & Jardine, 2022b, Liu et al., 2023a). from Ushev & Jardine, 2022a, b).

4.1 Site stiffness profiles The following functions were fitted to correlate the cross-
anisotropic stiffnesses with the principal stresses σvʹ and σhʹ.
Extensive field SCPT and laboratory bender element
and small-stress probing tests led to the stiffness profiles av
plotted in Figure 13. The 2016 PISA SCPT tests indicated σ '
Ev ' = f (e)Cv  v  (2)
higher Gvh profiles than those reported by Chow (1997)  pref 
at a location 100 m distant, potentially reflecting local
ground variability and/or the effects of ageing over ≈ b
20 years under the hydraulic fill (Zdravković et al., 2020a). σ ' (3)
Eh ' = f (e)Ch  
Drained triaxial tests indicated consistently higher vertical  pref 
(Evʹ) than horizontal (Ehʹ) moduli, reflecting the applied
K0 = 0.4, although the Gvh and Ghh components showed avh bvh
σ '  σh ' 
less difference. The in-situ SCPT Gvh profiles plotted Gvh = f (e)Cvh  v    (4)
distinctly above laboratory bender element (BE) values,  pref   pref 
especially above the water table, which was ascribed to
long-term ageing, partial saturation in-situ and seasonal ahh bhh
σ '  σh '  (5)
water table fluctuations. Checking for, and addressing, such Ghh = f (e)Chh  v   
discrepancies is crucial to successful practical modelling  pref   pref 
of laterally loaded monopiles.
where Cv, Ch, Cvh, Chh, av, bh, avh, bvh, ahh and bhh are material
4.2 Full strain mechanical behaviour under monotonic coefficients and f(e) is taken as (2.17-e)2/(1+e).
loading Table 2 summarises the cross-anisotropic stiffness
coefficients determined from Liu’s (2018) study and Kuwano’s
(1999) earlier Dunkirk sand probing tests. Ch/Cv and Chh/
4.2.1 Linear elastic stiffness anisotropy
Cvh, were invariably greater than unity, indicating inherent
anisotropy induced by grain orientation during water (or air)
Liu (2018) undertook stress-path probing tests on fully
pluviation. Although the simple treatment generally works
instrumented 100 mm diameter specimens that were consolidated
well, the experiments indicated coefficients that varied
and swelled along five constant K (= σhʹ/σvʹ) stress ratio paths.
moderately with K with Cv values from K = 0.5 tests around
15% above those from K = 1 experiments; see Figure 14.
4.2.2 Non-linear stiffness and yielding
Vinck (2016) and Liu (2018) investigated Dunkirk

sand’s behaviour from its linear elastic range through to
critical state conditions in over sixty locally instrumented
tests, varying DR, effective stress level (p0′) and OCR as well
as platen configurations. Figure 15 illustrates the effects of
OCR on Ev,tan′/f(e) through samples tested at p0ʹ = 50 and
150 kPa after unloading to a range of OCRs. While over-
consolidation affected the initial Ev,tanʹ values by only 5%, it
extended the initial linear tangent stiffness plateau through
changes in the granular contacts and micro-structure that were
accompanied by minimal volume changes and so cannot be
captured by, for example, applying a simple f(e) function.
Rowe & Barden (1964), Bishop & Green (1965) and
Koseki (2021) identified how end friction affects large-strain
triaxial behaviour and shear strength of sands. However,
its impact on small-strain behaviour was only recently
investigated experimentally (with peat) by Muraro & Jommi
(2019) and numerically by Cui et al. (2021), considering
stiff, high OCR, London clay. Liu et al. (2022a) explored
Figure 13. Dunkirk stiffness profiles from in-situ SCPT, laboratory the impact of platen roughness on the pre-failure stiffness of
bender element (BE) and triaxial small-stress probing tests (Liu, siliceous sand and Figure 16 demonstrates how end constraint
2018). affects Dunkirk sand’s stiffness and degradation trends even

Liu et al.
Table 2. Cross-anisotropy stiffness coefficients of Dunkirk sand.

Sand batch DR (%) Method K Stiffness parameters
Batch 1 57-65 AP 1, 0.53, 0.35 Cv: 225 av: Ch: 228 bh: Cvh: 70 avh: Chh: 105 ahh: Ch/Cv: 1.01
(5.85-11.9 m bgl.) 0.55 0.50 0.36 bvh: 0.05 bhh: Chh/Cvh: 1.5
0.20 0.46
Cvh: 49 avh: Chh: 63 ahh: Ch/Cv: 1.08
Batch 3 Cv: 154 av: Ch: 166 bh:
73-92 WP 1, 0.5 0.30 bvh: 0.10 bhh: Chh/Cvh:
(0-0.8 m bgl.) 0.51 0.50
0.19 0.42 1.29
Notes: (1) Batch 1 data from Kuwano (1999); (2) Sample preparation method: AP = air pluviation; WP = water pluviation; (3) parameters Cv, Ch, Cvh and Chv in MPa; (4)
All tests performed with lubricated platens.
Figure 14. Correlations between Evʹ/f(e) and σvʹ/pref and the stiffness
Figure 15. Effects of effective stress level and history on the pre-
coefficients for the K = 1 and 0.5 small-stress probing tests on
failure small strain stiffnesses of Dunkirk sand (Liu, 2018).
Dunkirk sand (Liu, 2018).
when specimens had 2:1 initial height-to-diameter ratios and

strains were gauged locally over the central ≈ 50 mm height.
Figure 16 plots Ev,tan′/f(e) against σv′/pref for medium dense
specimens, considering axial strains of 0.001%, 0.01%,
0.1% and 1%. Medium density specimens tested at relatively
low stress levels were most heavily affected by sample end
constraints at low to modest strain levels.
4.2.3 Large-strain and critical state behaviour
End constraint also influences medium-to-large strain

behaviour. Full end constraint leads to misleadingly high peak
stress ratios and shear strengths, at apparently smaller strains,
and ultimate stress ratios that did not tend to converge to a
unique M with medium dense specimens. Deformation was
more uniform in tests employing enlarged lubricated platens,
Figure 16. Effects of platen conditions on the pre-failure small strain
although bulging about the vertical axis was observed with stiffnesses of medium dense (DR ≈ 53%) Dunkirk sand specimens
these specimens, but without visible bifurcation even with (Liu, 2018; Liu et al., 2022a).
very dense specimens.
Liu (2018) interpreted Dunkirk sand’s large strain and
ultimate shearing behaviour within the state parameter critical
state framework, providing the basis for bounding surface Fuentes et al. (2023) and Selby et al. (2023) demonstrated
plasticity (Taborda et al., 2014) and Nor-Sand (Jefferies these models’ application in the simulations of the PISA
& Been, 2015) modelling. Taborda et al. (2020), Castillo- Dunkirk pile tests.

Figure 17. Normalised qcyc/p0ʹ-qmean/p0ʹ (CSR-η) stress states with reference to the peak and critical state envelopes of Dunkirk sand (Note
(q/pʹ)peakTXC = 1.5, McsTXC = 1.31, (q/pʹ)peakTXE = -1 and McsTXE = -0.9) (Liu, 2018).
4.3 Long-term drained cyclic loading behaviour
Aghakouchak (2015) and Aghakouchak et al. (2015)

conducted cyclic triaxial and hollow cylinder (HCA) tests to
model axial cyclic pile loading and showed how to cater for
the ‘near-shaft’ conditions, stress and loading histories that
govern field behaviour. Lateral cyclic pile loading engages
larger volumes in the passive, active and side shear reaction
regions located around the pile circumference. Predominantly
drained conditions applied in the PISA Dunkirk tests; conditions
are expected to be substantially drained under long-term
cyclic loading around even larger monopiles.
Liu’s (2018) cyclic triaxial testing programme for
Dunkirk therefore focussed on long-term drained constant
σr′ experiments, aiming to produce a benchmark dataset Figure 18. Deviatoric stress strain (q-εa) trends for a typical long-term
against which cyclic constitutive models could be validated, drained cyclic triaxial test involving two sets of stress-controlled
calibrated, and applied in boundary value problems. consolidation and creep stages, a cyclic shearing stage and post-
cyclic monotonic shearing to failure (Liu, 2018).
The tests applied the cyclic and mean deviatoric stress
combinations (qcyc, qmean) as defined in qcyc/p0ʹ-qmean/p0ʹ (or
CSR-η) space in Figure 17. Specimens were consolidated
to, and allowed to creep at, initial (qmean, p0′) states before interpret ratcheting, cyclic stiffness, damping and other
constant amplitude cycling to target conditions with N cycling characteristics from the high-resolution cycle-by-
> 104. All but two specimens survived cycling and were cycle stress-strain data.
subsequently sheared monotonically to failure to assess how Post-cyclic shear stiffness trends are considered in
cycling had modified their behaviour. Figure 18 illustrates Figure 20, plotting normalised secant Young’s moduli
the q–εa trends shown for a typical test on a very dense Ev,secʹ against axial strain for specimens cycled from the
(DR = 92%) specimen that started from (qmean, p0′) = (100, same (qmean, p0′) = (100, 133.3) (kPa) mean conditions but
133.3) (kPa) with qcyc = 40 kPa (CSR = 0.3). The local axial with different amplitudes (qcyc). Applying f(e) accounted
strains recorded with high resolution sensors manifested the for void ratio changes under cycling, showing how cycling
accumulation trends plotted logarithmically in Figure 19, led to higher initial stiffness, longer linear plateaux and
considering each of the 104 cycles. Also annotated are more gradual stiffness degradation. These changes grew
power law equations that fitted the experimental trends. markedly with the prior cyclic stress ratio, reflecting
Liu (2018) further related the fitting parameters to the micro-fabric optimisation under cycling. The significantly
qcyc/p0ʹ-qmean/p0ʹ loading conditions and showed how to better organised internal particle force chains and contact

Liu et al.
distributions also relocated and expanded the sand’s Y1 and non-uniform straining imposed by these tests. Triaxial BE
Y2 kinematic yield surfaces. measurements made on the same samples indicated Ghh/Gvh
≈ 0.5 in the shallow layers, which gradually rose to unity
5. St Nicholas at Wade (SNW) chalk at depth; the field seismic data show a similar, but more
muted, trend. Also plotted in Figure 21 is the Eh′/Ev′ profile
Extensive laboratory and in-situ characterisation developed from four suites of triaxial probing tests by Vinck
were conducted for the ALPACA piling JIP. Noting that (2021). Horizontal loading from in-situ stresses provokes
pile driving alters chalk properties significantly, the testing a far softer response than vertical compression, which is
considered the behaviour of both intact chalk and the de- important when analysing lateral pile loading. This aspect
structured ‘putty’ material that formed adjacent to the pile of the chalk’s marked anisotropy contributed to undrained
shafts during driving and controlled their axial behaviour. and drained triaxial compression paths following similarly
The programme encompassed monotonic and undrained cyclic inclined effective stress paths (Vinck et al., 2022).
triaxial loading, supported by geological logging, index,
unconfined compression strength (UCS), oedometer, direct 5.2 Chalk’s response to monotonic triaxial loading
simple shear (DSS), Brazilian tension (BT) and interface shear
tests (Vinck, 2021). A metallurgical corrosion study was also
undertaken by Vinck et al. (2023b) to provide fundamental 5.2.1 Pressure-dependent yielding and failure
insights into the piles’ highly age-dependent axial shaft
capacity trends (Buckley et al., 2018, 2020; Jardine et al., The SNW chalk’s cemented open micro-structure
2023b, c; Vinck et al., 2023a). (Alvarez-Borges et al., 2020) leads to significant sensitivity,
brittleness and pressure-dependent behaviour. Figure 22 shows
5.1 Elastic stiffness profiles example loading curves up to 0.5% axial strain from locally
instrumented triaxial and UCS tests conducted from in-situ
The chalk’s elastic stiffnesses were examined by p0′ and 300 kPa + p0′ stresses. Very stiff and nearly linear
laboratory and field testing and Figure 21 considers the shear pre-failure behaviour is evident, as well as a notably brittle
stiffness profiles measured at the smallest strains offered by post-peak response that involved tensile fracturing and led to
various techniques. As emphasised by Clayton et al. (2003), low strengths at large strains. High pressure triaxial tests that
the chalk’s meso-to-macro fabric plays a dominant role. explored conditions from in-situ p0′ (≈ 63 kPa) up to p0′ =
The mean SCPT Gvh and cross-hole Ghv trends amount to 12.8 MPa identified a gradual transition (shown in Figure 23)
around 2/3 of the shallow triaxial tests’ BE Gvh measurements from brittle to ductile behaviour under increasing p′ levels.
and converge more closely at depth, reflecting the reducing The chalk’s highly curved failure envelope (Figure 24)
occurrence of open fissures, which are naturally excluded reaches an ultimate (q/p′)ult = 1.25, equivalent to critical
from laboratory specimens. The direct simple shear (DSS) state ϕ′ ≈ 31o, that matches (as shown later) that of fully
Gvh maxima and the pressuremeter Ghh values fall far de-structured chalk.
below those interpreted from shear wave velocities, due to Figure 25 depicts how the initial Y1 and Y3 yield loci
their low resolution and the chalk’s brittle response to the of unfractured chalk may be represented with near elliptical
Figure 19. Accumulated cyclic axial strains for very dense (DR = 92%)
triaxial Dunkirk sand specimens cycled from (qmean, p0ʹ) = (100, Figure 20. Effects of cyclic loading history on the post-cyclic
133.3) (kPa) under fully drained conditions (Liu, 2018). monotonic stiffness degradation trends for Dunkirk sand (Liu, 2018).

Figure 21. SNW profiles and ratios of maximum stiffnesses: (a) laboratory and in-situ measurements of Gvh and Ghv; (b) Ghh profiles;
(c) ratios between laboratory and in-situ stiffnesses; and (d) profiles of stiffness anisotropy ratios (modified from Vinck et al., 2022).
surfaces in triaxial q-p′ space. Intact (cemented) chalk

displays linear behaviour over a far larger region of q-pꞌ
space than soils such as the unbonded Cowden till or Dunkirk
sand. Laboratory elastic vertical Young’s moduli Ev,max′ are
comparatively insensitive to applied p0′, with modest increases
up to the peak (7.7 GPa) found with p0′ = 650 kPa before
falling to a 4 MPa plateau as bonds break under increasing
p0′ (Liu et al., 2023b). However, the chalk’s systems of
micro-to-macro fissures, which gradually close under normal
stress, lead to the maximum mass chalk stiffnesses invoked
by pile loading being four times lower than the laboratory
values (Jardine et al., 2023c).
Chalk’s propensity to tensile fracture leads to BT
and DSS strengths 90% and 50% lower respectively than
those from UCS or triaxial compression tests, with many Figure 22. Example stress-strain curves up to 0.5% axial strain
implications for the modelling of geotechnical problems from locally instrumented triaxial and unconfined compression
that involve multi-directional loading. (UCS) tests on intact SNW chalk from in-situ p0′ and 300 kPa +
p0′ (Vinck et al., 2022).
5.2.2 Fully de-structured chalk’s response to undrained
monotonic shearing
Laboratory dynamic compaction, applied at in-situ considering the q-p′ effective stress paths. These are initially
water content, was employed to de-structure low-to-medium nearly vertical, suggesting that the re-consolidated and
density chalk in an analogous way to pile driving, producing (mildly aged) putty’s initial stiffness response was largely
uniform chalk putty with 9 ± 3 kPa fall-cone shear strengths isotropic. The effective stress paths rotated to follow
and liquid and plastic limits of 30.6% and 24.2% respectively. leftward (contractive) stages after mobilising modest ‘peak’
Instrumented, consolidated, triaxial specimens were formed resistances (at εa < 0.2%) and showed strain softening as
with (pre-shearing) void ratios and effective stress states shearing continued up to phase transformation (PT) points
matching those noted in the de-structured annuli identified at which their paths rotated abruptly and climbed towards
around driven pile shafts several months after driving ultimate (critical state) conditions. Continued straining led
(Buckley et al., 2018). to markedly higher ultimate strengths as the specimens
The de-structured chalk’s undrained triaxial compression attempted to continue dilating. Extension tests indicated
and extension behaviour is demonstrated first in Figure 26 by similar, yet not fully symmetric, paths and shear strengths.

Liu et al.
Figure 23. Full-strain triaxial compression behaviour of intact

SNW chalk under a wide range of pressures (Liu et al., 2023b).
Figure 25. Yield stresses and interpreted pre-failure Y1 and large-

scale Y3 loci for intact SNW chalk (Liu et al., 2023b).
Figure 24. Peak and ultimate shear strength envelope for intact
SNW chalk (Liu et al., 2023b). Figure 26. Effective stress paths of de-structured chalk under undrained
triaxial compression and extension (modified from Liu et al., 2022b).
The ultimate critical states gave ϕcs′ ≈ 31°, matching the by high-level cycling with qcyc = 950 kPa and qmax/(2Su) =
high-pressure intact tests in Figure 24. 0.91. The axial strain and pore pressure records show little
or no sign of impending instability; little change was seen
5.3 Response to undrained cyclic triaxial loading in either the damping ratio (which remained at around 4%)
or the cyclic stiffness, until degradation set in over the last
30 cycles and brittle failure occurred abruptly after 181 cycles.
5.3.1 Intact chalk’s failure patterns In contrast, specimens that sustained large number (> 4,000) of
cycles developed minimal straining and excess pore pressure
Undrained cycling affects intact chalk differently to accumulation and manifested nearly visco-elastic behaviour.
saturated sedimentary sands or clays (Ahmadi-Naghadeh et al., The experimental outcomes allow cyclic interaction
2022). Figure 27 presents a typical unstable response observed diagrams to be established in qcyc/(2Su)–qmean/(2Su) space, as in
with a specimen loaded from p0′ = 42 kPa, qmean = 1225 kPa Figure 28. A tentative fan of linear Nf contours indicates the

Figure 27. Typical unstable cycling response of intact chalk:

Variation of axial strain and pore water pressure against number
of cycles (Test conditions: p0′ = 42 kPa, qmean = 1225 kPa, qcyc =
950 kPa and qmax/(2Su) = 0.91) (Ahmadi-Naghadeh et al., 2022).
main trends applying between 1 and 3000 cycles. The region Figure 28. Cyclic interaction diagram for intact SNW chalk tested
below the 3000-cycle contour in Figure 28 signifies the stable from p0′ = 42 kPa, showing the interpreted contours of number
of cycles to failure for specified qcyc-qmean conditions (Ahmadi-
region within which undrained cycling improved stiffness
Naghadeh et al., 2022).
without any loss of undrained shear strength.
Overall, the intact chalk’s behaviour appears more like
that of crystalline rocks or metals than sedimentary soils.
Repeated loading prompts progressive wear and shearing
between grains, forming microcracks that may coalesce into
macrocracks (Cerfontaine & Collin, 2018). The outcomes
can be interpreted as ‘S-N’ curves that plot the maximum
cyclic load (qmax = qmean + qcyc) against log N, with contours
interpolated to show how the qmax/(2Su) curves fall as qcyc/
qmean rises, expressed as:
qmax 1 + qcyc / qmean (6)

= f (N ) ⋅
2Su f ( N ) + qcyc / qmean
in which f(N) represents for each Nf line its projected intercept

on the vertical axis of the interactive qcyc/(2Su)-qmean/(2Su)
diagram and is expressed as:
1
f (=
N ) 0.35 + (7) Figure 29. Three-dimensional representation of the undrained
1.54 + 0.37 × [log10 ( N f )]2.75 cyclic failure characteristics of intact SNW chalk in qmax/(2Su)-qcyc/
qmean-log10(N) space (Ahmadi-Naghadeh et al., 2022).
A lower limit of f(N) = 0.35 is incorporated in this
expression which corresponds to a fatigue limit (or fatigue
strength) of qmax/(2Su) = 0.52, below which specimens 5.3.2 Cyclic failure characteristics of de-structured chalk
could sustain cycles indefinitely under the most critical
qcyc/qmean = 1 one-way loading condition. The above two Liu et al. (2022b) show that fully de-structured, re-
equations lead further to a three-dimensional representation consolidated, chalk putty responds to undrained cyclic
of intact chalk’s undrained cyclic failure characteristics in loading similarly to silts and silty sands, and markedly
a qmax/(2Su)-qcyc/qmean-log10(N) space, as shown in Figure 29. differently to the intact chalk. As illustrated in Figure 30,

Liu et al.
de-structured specimens respond to high-level cycling by The full suite of one- and two-way cyclic tests is set-
generating excess pore pressures that shift their effective out in an interactive loading diagram in Figure 31, which
stress paths leftwards. Both contractive and dilative phases adopts both qcyc/(2Su)-qmean/(2Su) and qcyc/(p0′)-qmean/(p0′) axes.
occur within individual cycles. Cyclic phase transformation A tentative family of contours is shown for number of cycles
lines with (q/p′) gradients of ≈ 0.54 and 0.38 can be identified to failure (Nf); the lower-levels, high Nf, contours show less
in compression and extension respectively, following Mao curvature and tighter spacings than those representing high-
& Fahey (2003), that fall well below the monotonic phase level cycling. As with all soils, chalk putty is more susceptible
transformation lines. The transient cyclic effective stress to high-level two-way (compression and extension, with qmean
paths can also cross the monotonic critical state gradients M. ≤ qcyc) loading than one-way compression cycling.
Figure 30. Effective stress paths for typical unstable tests on de-structured chalk and the identified cyclic phase transformation lines
(Liu et al., 2022b).
Figure 31. Cyclic interaction diagram for de-structured chalk expressed in normalised qcyc/(2Su) - qmean/(2Su) and qcyc/p0′ - qmean/p0′ stress
space for p0′ = 200 kPa and p0′ = 400 kPa test series (Liu et al., 2022b).

Significant axial strain development and stiffness losses major offshore wind projects. As noted below, advances
occur in unstable tests that accelerate markedly as cyclic have flowed in four main areas.
failure approaches. Liu et al. (2022b) report on these features
and the specimens’ damping ratios. Cycle-by-cycle effective 6.1 Pile behaviour under monotonic axial loading
stress changes were tracked in the experiments, enabling the
development of a laboratory-based global prediction method Jardine et al. (2005) and Jardine (2014) showed that
for axial shaft capacity under cyclic loading (Jardine, 2020; advanced finite element (FE) modelling based on the glacial
Buckley et al., 2023). All tests showed Δp′/p0′ decreasing tests by Jardine et al. (1984) and sand tests by Kuwano
continuously with N, although it is possible that cycling at (1999) offered good matches to full-scale field observations
lower levels than those applied would identify conditions made at North Sea clay till sites and tests to failure on 19 m
under which no reduction occurs. Liu et al. (2022b) show long, 457 mm diameter piles driven at Dunkirk. The most
that the mean effective stress drifts of tests run with (qcyc+ significant new findings to emerge regarding axial load-
qmean)/p0′ < 0.4 can be related to the number of cycles applied displacement behaviour, which is central to jacket supported
by the power-law Equation 8. structures, therefore concern the chalk, about which little
information was available previously. Wen et al. (2023a,
p' qcyc b) employed the laboratory testing to calibrate new 1D
=A × ( B + )× NC (8) load transfer and advanced 2D/3D FE models that provide
p0 ' p0 '
generally good fits to the ALPACA and ALPACA Plus field
Parameters A, B and C define the rates of p′ degradation tests on piles with diameters up to 1.8 m and embedded
and the maximum cyclic stress ratio that could lead to lengths up to 18 m.
beneficial, null, or deleterious cycling effects. Higher level
6.2 Pile behaviour under cyclic axial loading
cyclic tests that failed within just a few cycles show more
complex trends. Figure 32 reports the outcomes of cyclic
Buckley et al. (2023) and Liu et al. (2023c) report on
tests on samples consolidated to p0′ = 200 kPa and shows
how the laboratory chalk putty cyclic testing contributed to
how Equation 8, applied with A = -0.05 and B = -0.12 and
a simplified yet accurate global prediction method for axial
C = 3.48×qcyc/p0′ fits the experimental trends well.
cyclic pile design in chalk, which is now available for use
by industry. Jardine (2020) and Jardine et al. (2012) have
6. Applications of the research reported previously on the use of undrained cyclic tests on
sands and clays to predict pile response to axial loading,
The Cowden, Dunkirk and SNW characterisation
showing how the laboratory experiments by Aghakouchak et al.
programmes were vital to interpretating the high-value,
(2015) on Dunkirk sand enabled accurate predictions of the
large-scale, instrumented pile experiments run at the three
sites. They also led to new site investigation and practical cyclic axial pile tests reported by Jardine & Standing (2012).
pile design approaches that are being applied widely in The Cowden till cyclic testing data is also now available to
aid back-analysis of earlier cyclic axial pile tests (see Ove
Arup and Partners, 1988) as well as monotonic axial tests
reported by Karlsrud et al. (2014).
6.3 Pile behaviour under monotonic lateral loading
Modelling of field lateral pile loading response is

critical to the efficient design of monopile supported offshore
wind turbines. Taborda et al. (2020) and Zdravković et al.
(2020b) describe how the site characterisation described
above was central to advanced 3D FE analyses that aided
the design and interpretation of the PISA lateral loading tests
on instrumented piles with diameters up to 2 m driven at
Dunkirk and Cowden. The resulting PISA design approach
(Byrne et al., 2020c) has offered a step change in design
practice that has been applied internationally. Pedone et al.
(2023) describe how still more advanced 3D FE analyses
Figure 32. Mean effective stress trends during cycling and fitted developed from the chalk testing captured accurately the
trends (in dashed lines) for ‘putty’ tests with p0′ = 200 kPa and lateral loading behaviour shown by ALPACA tests in the
mostly (qcyc+ qmean)/p0′ < 0.4 (Liu et al., 2022b). fractured, highly brittle SNW chalk. These experiments and

Liu et al.
analyses are now being applied in the design of major new the development of suitable predictive approaches.
windfarms at offshore chalk sites. While the intact chalk’s response was distinctly
different to that of the till and sand, resembling that
6.4 Pile behaviour under cyclic lateral loading of crystalline rocks or metals, the chalk putty’s cyclic
behaviour was closer to that shown by the glacial
It is essential to consider lateral cyclic loading cases till and marine sand.
when designing monopile foundations. The PISA cyclic
pile tests are applied widely as benchmarks against which Acknowledgements
alternative modelling strategies are routinely calibrated and
tested. The Authors’ recently published cyclic laboratory The first author is deeply grateful to Lidija Zdravković,
test programmes offer a basis for re-analyses of the Cowden Erdin Ibraim and Andrea Diambra who provided essential
and Dunkirk cyclic PISA pile tests, as well as the recent guidance for writing up this paper and lecture. Parts of the
PICASO tests (Byrne et al., 2020a) at Cowden. The cyclic laboratory testing campaigns on Cowden till and Dunkirk sand
laboratory programmes on chalk will also pave the way for were undertaken to support the PISA JIP project managed by
well-calibrated lateral cyclic analyses at chalk sites. Ørsted. The PISA sponsors’ and Academic Working Group’s
support is acknowledged thankfully, as is the financial support
7. Summary and conclusions provided by Ørsted for the later Post-PISA Experimental
Project. The studies on St Nicholas at Wade (SNW) chalk
This paper summarises recent research into the stiff were undertaken as part of the ALPACA project funded by
glacial till, dense marine sand, and low-to-medium density the Engineering and Physical Science Research Council
chalk encountered at the PISA, ALPACA and ALPACA (EPSRC) grant EP/P033091/1 and Royal Society Newton
Plus JIPs instrumented pile test sites. Coordinated field and Advanced Fellowship NA160438. The authors acknowledge
laboratory studies investigated the materials’ stiffness patterns,
gratefully additional financial and technical support from
full strain mechanical behaviour, anisotropy and response to
Atkins, Cathie Associates, Equinor, Fugro, Geotechnical
cyclic loading. Their outcomes are being employed widely to
Consulting Group (GCG), Iberdrola, Innogy, LEMS, Ørsted,
aid the modelling of field behaviour under axial-and-lateral,
Parkwind, Siemens, TATA Steel and Vattenfall. Ken Vinck
monotonic-and-cyclic loading. By enabling more efficient
acknowledges Imperial College’s EPSRC Centre for Doctoral
driven pile foundation design, the research is contributing
Training (CDT) in Sustainable Civil Engineering and the
significantly to more efficient offshore wind energy generation
DEME Group (Belgium) for supporting his doctoral research.
and CO2 emission reductions. The large volume of new
The kind support and contributions of many collaborators
findings supports four general conclusions.
are acknowledged gratefully including those of Amin
1. Monotonic laboratory testing to failure showed
Aghakouchak, Reza Ahmadi-Naghadeh, Amandine Brosse,
how the geomaterials behave over their full strain
Róisín Buckley, Byron Byrne, Matteo Ciantia, Matthew
range, including the distinct properties of the fully
de-structured chalk putty. Combining these data Coop, Wenjie Cui, Pedro Ferreira, Stavroula Kontoe, James
with in-situ measurements enables representative Lawrence, Ross McAdam, Satoshi Nishimura, Catherine
modelling of the pile test outcomes and shows the O’Sullivan, Giuseppe Pedone, David Potts, David Taborda,
way forward for their practical application; Kai Wen, Hongjie Zhou and others. Invaluable technical
2. Laboratory and field measured elastic stiffnesses support by Steve Ackerley, Graham Keefe, Prash Hirani,
may show significant divergence. While broadly Stef Karapanagiotidis, Graham Nash and Gary Jones is
similar profiles applied at Cowden, triaxial tests on acknowledged gratefully. Finally, the authors thank their
reconstituted Dunkirk sand underpredicted the site’s families for their patient and essential support, which was
field stiffnesses, which had grown after 20 years of crucial to enabling the presented research.
ageing under hydraulic fill. In contrast, the chalk’s
mass field stiffness fell far below the measurements Declaration of interest
made from laboratory tests on intact samples because
they did not capture the important effects of partially The authors have no conflicts of interest to declare. All
open meso-to-macro fractures; co-authors have observed and affirmed the contents of the
3. All three geomaterials displayed elastic stiffness paper and there is no financial interest to report.
anisotropy. Vertical stiffnesses exceeded horizonal
moduli in the normally consolidated dense Dunkirk Authors’ contributions
sand and in the high YSR intact SNW chalk, while
the opposite applied in high YSR Cowden glacial till; Tingfa Liu: conceptualization, investigation, data
4. The cyclic programmes showed a wide range of curation, methodology, formal analysis, writing – original
behaviours and provide vital new evidence to aid draft. Ken Vinck: conceptualization, investigation, data

curation, methodology, formal analysis, writing – review DSS Direct simple shear
& editing. Emil R. Ushev: conceptualization, investigation, D50 Mean particle diameter
data curation, methodology, formal analysis, writing – review Ev′, Eh′ Drained vertical and horizontal Young’s moduli
& editing. Richard J. Jardine: supervision, methodology, respectively
project administration, funding acquisition, writing – review E vu, E hu Undrained vertical and horizontal Young’s moduli
& editing. respectively
EuTXC Undrained Young’s modulus determined from
Data availability triaxial compression tests
Esecu Undrained secant Young’s modulus
Some data that support the findings of this study are Ghh Shear modulus in horizontal plane
available from the corresponding author upon reasonable Ghv,Gvh Shear modulus in vertical plane
request. Gzθ (= Gvh) Shear modulus in vertical plane
G0 Elastic shear stiffness
List of symbols and abbreviations Goct Octahedral shear modulus
Gs Specific gravity
av, avh, ahh Cross-anisotropy stiffness parameters; see HCA Hollow Cylinder Apparatus
Equations 2-5 IDD Intact dry density
b Intermediate principal stress factor (= (σ2 - σ3)/ JIP Joint industry project
(σ1 - σ3)) KUC Ko-consolidated undrained compression
bh, bvh, bhh Cross-anisotropy stiffness parameters; see KUE Ko-consolidated undrained extension
Equations 2-5 K0 Coefficient of earth pressure at rest
c′ Soil cohesion LVDT Linear variable differential transformer
e Void ratio M Critical state q/p′ stress ratio
e0 Specimen initial void ratio McsTXC, McsTXE Critical state stress ratio under triaxial
p′ Mean effective stress compression and extension
p′in-situ In-situ mean effective stress N Number of cycles
p0′ Initial mean effective stress Nf Number of cycles to failure
pref Reference atmospheric pressure (= 101.3 kPa) Nkt Cone factor defined as Nkt = (qt – σv0)/Su
q Deviatoric stress OCR Over-consolidation ratio
q c, q t Measured cone resistance and cone resistance PI Plasticity index
corrected for pore pressure respectively PISA PIle Soil Analysis
qcyc Cyclic deviatoric stress amplitude (= (qpeak – PT Phase transformation
qtrough)/2) SCPT Seismic cone penetration test
qf Deviatoric stress at failure Sr Saturation degree
qmean Mean q applied in stress cycle Su Undrained shear strength
qmax Maximum q applied in stress cycle (= qmean SuTXC, SuTXE Undrained shear strength under triaxial
+ qcyc) compression and extension
qtrough Minimum q applied in stress cycle (= qmean – TXC, TXE Triaxial compression and extension
qcyc) respectively
qPT Deviatoric stress at the phase transformation YSR Yield stress ratio
state α Angle between the vertical and the direction of
wc Water content σ1 axis
wL Liquid limit α dσ Angle between the vertical and the direction of Δσ1
wp Plasticity limit axis
ALPACA Axial-Lateral Pile Analysis for Chalk αf Angle between the vertical and the direction of σ1
Applying multi-scale field and laboratory testing axis at ultimate failure
AOD Above ordnance datum γzθ Torsional shear strain
BE Bender element test Δu Excess pore water pressure
CPTu Cone penetration test with pore pressure δult′ Ultimate soil-structure interface shear resistance
measurement angle
CSR Cyclic stress ratio (= qcyc/p0′) εa, εv Axial (vertical) strain
Cv, Ch, Cvh, Chh Cross-anisotropy stiffness parameters, see εmean Permanent axial strain accumulated in cyclic
Equations 2-5 tests
DR Relative density εr Radial (horizontal) strain

Liu et al.
εs Shear strain (= εa for undrained triaxial Buckley, R.M. (2018). The axial behaviour of displacement
condition) piles in chalk [Doctoral thesis]. Imperial College London,
ε z, ε r, ε θ Axial, radial and circumferential strains respectively London, UK. Retrieved in August 23, 2023, from https://
in HCA space spiral.imperial.ac.uk/handle/10044/1/78617
ε v, ε h Vertical (axial) and horizontal (radial) strains Buckley, R.M., Jardine, R.J., Kontoe, S., Parker, D., &
respectively in triaxial space Schroeder, F.C. (2018). Ageing and cyclic behaviour
ε 1, ε 3 Major and minor principal strain respectively of axially loaded piles driven in chalk. Géotechnique,
η Mean stress ratio (= qmean/p0′) 68(2), 146-161. http://doi.org/10.1680/jgeot.17.P.012.
ρbulk Bulk density Buckley, R.M., Jardine, R.J., Kontoe, S., Barbosa, P., &
σ 1, σ 2, σ 3 Major, intermediate and minor principal stresses Schroeder, F.C. (2020). Full-scale observations of dynamic
respectively and static axial responses of offshore piles driven in
σvʹ, σhʹ Vertical and horizontal effective stress chalk and tills. Géotechnique, 70(8), 657-681. http://doi.
respectively org/10.1680/jgeot.19.TI.001.
σvyʹ One-dimensional (1-D) compression yield Buckley, R.M., Jardine, R.J., Kontoe, S., Liu, T., Byrne, B.W.,
stress McAdam, R.A., Schranz, F., & Vinck, K. (2023). Axial
σv0ʹ In-situ vertical effective stress cyclic loading of piles in low to medium density chalk.
τzθ Torsional shear stress Géotechnique, 1-14. In press. http://doi.org/10.1680/
ϕʹ Effective angle of shearing resistance jgeot.22.00044.
ϕʹcs Critical state shear resistance angle Byrne, B.W., McAdam, R.A., Burd, H.J., Houlsby, G.T.,
ϕ′peak Shear resistance angle at peak Martin, C.M., Beuckelaers, W.J.A.P., Zdravkovic, L.,
Taborda, D.M.G., Potts, D.M., Jardine, R.J., Ushev, E.,
Liu, T., Abadias, D., Gavin, K., Igoe, D., Doherty, P.,
References
Gretlund, J., Andrade, M.P., Wood, A.M., Schroeder,
F.C., Turner, S., & Plummer, M.A.L. (2017). PISA: new
Aghakouchak, A. (2015). Advanced laboratory studies
design methods for offshore wind turbine monopiles.
to explore the axial cyclic behaviour of driven piles
In Proceedings of the 8th International Conference on
[Doctoral thesis]. Department of Civil and Environmental
Offshore Site Investigation and Geotechnics (pp. 142-
Engineering, Imperial College London, London, UK.
161), London. http://doi.org/10.3723/OSIG17.142
Retrieved in August 23, 2023, from https://spiral.imperial.
Byrne, B.W., Aghakouchak, A., Buckley, R.M., Burd, H.J.,
ac.uk/handle/10044/1/28900
Gengenbach, J., Houlsby, G.T., Mc Adam, R.A., Martin,
Aghakouchak, A., Sim, W.W., & Jardine, R.J. (2015). Stress-
C.M., Schranz, F., Sheil, B.B., & Suryasentana, S.K.
path laboratory tests to characterise the cyclic behaviour (2020a). PICASO: cyclic lateral loading of offshore wind
of piles driven in sands. Soil and Foundation, 55(5), turbine monopiles. In Proceedings of the 4th International
917-928. http://doi.org/10.1016/j.sandf.2015.08.001. Symposium Frontiers in Offshore Geotechnics, Austin,
Ahmadi-Naghadeh, R., Liu, T., Vinck, K., Jardine, R.J., Texas.
Kontoe, S., Byrne, B.W., & McAdam, R.A. (2022). Byrne, B.W., McAdam, R.A., Burd, H.J., Beuckelaers,
A laboratory characterization of the response of intact W.J., Gavin, K.G., Houlsby, G.T., Igoe, D.J., Jardine,
chalk to cyclic loading. Géotechnique, 1-13. http://doi. R.J., Martin, C.M., Muir Wood, A., Potts, D.M., Skov
org/10.1680/jgeot.21.00198. Gretlund, J., Taborda, D.M.G., & Zdravković, L. (2020b).
Alvarez-Borges, F., Madhusudhan, B.N., & Richards, D. Monotonic laterally loaded pile testing in a stiff glacial
(2020). Mechanical behaviour of low-medium density clay till at Cowden. Géotechnique, 70(11), 970-985.
destructured White Chalk. Géotechnique Letters, 10(2), http://doi.org/10.1680/jgeot.18.PISA.003.
360-366. http://doi.org/10.1680/jgele.20.00009. Byrne, B.W., Houlsby, G.T., Burd, H.J., Gavin, K.G., Igoe,
Bishop, A.W., & Green, G.E. (1965). The influence of end D.J., Jardine, R.J., Martin, C.M., McAdam, R.A., Potts,
restraint on the compression strength of a cohesionless D.M., Taborda, D.M.G., & Zdravković, L. (2020c). PISA
soil. Géotechnique, 15(3), 243-266. http://doi.org/10.1680/ design model for monopiles for offshore wind turbines:
geot.1965.15.3.243. application to a stiff glacial clay till. Géotechnique,
Brosse, A.M., Hosseini Kamal, R., Jardine, R.J., & Coop, 70(11), 1030-1047. http://doi.org/10.1680/jgeot.18.P.255.
M.R. (2017). The shear stiffness characteristics of four Castillo-Fuentes, E., Zdravković, L., & Potts, D.M. (2023).
Eocene-to-Jurassic UK stiff clays. Géotechnique, 67(3), On the calibration and application of the NorSand model.
242-259. http://doi.org/10.1680/jgeot.15.P.236. In L. Zdravković, S. Kontoe, D.M.G. Taborda, & A.
Brucy, F., Meunier, J., & Nauroy, J.F. (1991). Behaviour of pile Tsiampousi (Eds.), Proceedings of the 10th European
plug in sandy soils during and after driving. In Proceedings Conference on Numerical Methods in Geotechnical
of the 23rd Offshore Technology Conference (pp. 145-154), Engineering, London, UK. http://doi.org/10.53243/
Houston, TX, USA. http://doi.org/10.4043/6514-MS. NUMGE2023-242.

Cerfontaine, B., & Collin, F. (2018). Cyclic and fatigue 2, pp. 201-245), Lyon. Lisse: Balkema. http://doi.
behaviour of rock materials: review, interpretation org/10.1201/9780203970812.
and research perspectives. Rock Mechanics and Rock Jardine, R.J., Standing, J.R., & Chow, F.C. (2006). Some
Engineering, 51(2), 391-414. http://doi.org/10.1007/ observations of the effects of time on the capacity of piles
s00603-017-1337-5. driven in sand. Géotechnique, 56(4), 227-244. http://doi.
Chow, F. (1997). Investigations into the behaviour of org/10.1680/geot.2006.56.4.227.
displacement piles for offshore foundations [Doctoral Jardine, R.J., & Standing, J.R. (2012). Field axial cyclic
thesis]. Imperial College, University of London, London, loading experiments on piles driven in sand. Soil and
UK. Retrieved in August 23, 2023, from https://spiral. Foundation, 52(4), 723-736. http://doi.org/10.1016/j.
imperial.ac.uk/handle/10044/1/7894 sandf.2012.07.012.
Ciavaglia, F., Carey, J., & Diambra, A. (2017). Time-dependent Jardine, R.J., Andersen, K., & Puech, A. (2012). Cyclic
uplift capacity of driven piles in low to medium density loading of offshore piles: potential effects and practical
chalk. Géotechnique Letters, 7(1), 90-96. http://doi. design. In Proceedings of the 7th International Conference
org/10.1680/jgele.16.00162. on Offshore Site Investigations and Geotechnics (pp.
Clayton, C.R.I., Matthews, M.C., & Heymann, G. (2003). 59-100). London: Society for Underwater Technology.
The chalk. In T. Tan, S. Phoon, D.W. Hight, & S. Leroueil Jardine, R.J. (2014). Advanced laboratory testing in research
(Eds.), Proceedings of the International Workshop on and practice: the 2nd Bishop Lecture. Geotechnical Research,
Characterisation and Engineering Properties of Natural 1(1), 2-31. http://doi.org/10.1680/geores.14.00003.
Soils (pp. 983-1020), Singapore. CRC Press. Jardine, R.J. (2020). Geotechnics, energy and climate change:
Cui, W., Potts, D.M., Pedro, A.M.G., & Zdravković, L. (2021). the 56th rankine lecture. Géotechnique, 70(1), 3-59. http://
Numerical assessment of the effects of end-restraints and doi.org/10.1680/jgeot.18.RL.001.
a pre-existing fissure on the interpretation of triaxial tests Jardine, R.J. (2023a). Time-dependent vertical bearing
on stiff clays. Géotechnique, 71(9), 765-780. http://doi. behaviour of shallow foundations and driven piles:
org/10.1680/jgeot.19.P.318. the 6th McClelland Lecture. In Proceedings of the 9th
Davies, B.J., Roberts, D.H., Bridgland, D.R., Cofaigh, C.O., International offshore Site Investigation and Geotechnics
& Riding, J.B. (2011). Provenance and depositional Conference, London.
environments of Quaternary sediments from the western Jardine, R.J., Buckley, R.M., Liu, T., Byrne, B.W., Kontoe,
North Sea Basin. Journal of Quaternary Science, 26(1), S., McAdam, R.A., Schranz, F., & Vinck, K. (2023b).
59-75. http://doi.org/10.1002/jqs.1426. The ALPACA and ALPACA Plus Joint Industry studies
Diambra, A., Ciavaglia, F., Harman, A., Dimelow, C., Carey, of driven pile behaviour in low-to-medium density
J., & Nash, D.F. (2014). Performance of cyclic cone chalk. In Proceedings of the 9th International offshore
penetration tests in chalk. Géotechnique Letters, 4(3), Site Investigation and Geotechnics Conference, London.
230-237. http://doi.org/10.1680/geolett.14.00050. Jardine, R.J., Buckley, R.M., Liu, T., Andolfsson, T., Byrne,
Hight, D.W., & Jardine, R.J. (1993). Small strain stiffness B.W., Kontoe, S., McAdam, R.A., Schranz, F., & Vinck,
and strength characteristics of hard London Tertiary clays. K. (2023c). The axial behaviour of piles driven in chalk.
In A. Anagnostopoulos, F. Schlosser, N. Kalteziotis, & Géotechnique, 1-17. In press. http://doi.org/10.1680/
R. Frank (Eds.), Geotechnical engineering of hard soils- jgeot.22.00041.
soft rocks (Vol. 1, pp. 533-552). Rotterdam: Balkema. Jefferies, M., & Been, K. (2015). Soil liquefaction: a critical
Jardine, R.J., Symes, M.J., & Burland, J.B. (1984). The state approach. Boca Raton: CRC Press. http://doi.
measurement of soil stiffness in the triaxial apparatus. org/10.1201/b19114.
Géotechnique, 34(3), 323-340. http://doi.org/10.1680/ Karlsrud, K., Jensen, T.G., Wensaas Lied, E.K., Nowacki,
geot.1984.34.3.323. F., & Simonsen, A.S. (2014). Significant ageing effects
Jardine, R.J. (1985). Investigations of pile-soil behaviour, for axially loaded piles in sand and clay verified by new
with special reference to the foundations of offshore field load tests. In Proceedings of the Offshore Technology
structures [Doctoral thesis]. Imperial College, University Conference (Paper OTC-25197-MS), Houston, TX, USA.
of London, London, UK. Retrieved in August 23, 2023, http://doi.org/10.4043/25197-MS.
from https://spiral.imperial.ac.uk/handle/10044/1/8519 Koseki, J. (2021). Several challenges in advanced laboratory
Jardine, R.J. (1992). Some observations on the kinematic testing of geomaterials with emphasis on unconventional
nature of soil stiffness. Soil and Foundation, 32(2), 111- types of liquefaction tests. Geomechanics for Energy and
124. http://doi.org/10.3208/sandf1972.32.2_111. the Environment, 27, 100157. http://doi.org/10.1016/j.
Jardine, R.J., Standing, J.R., & Kovacevic, N. (2005). gete.2019.100157.
Lessons learned from full scale observations and the Kuwano, R. (1999). The stiffness and yielding anisotropy of
practical application of advanced testing and modeling. sand [Doctoral thesis]. Imperial College London, London,
In Proceedings of the International Symposium on UK. Retrieved in August 23, 2023, from https://spiral.
Deformation Characteristics of Geomaterials (Vol. imperial.ac.uk/handle/10044/1/8519

Liu et al.
Kuwano, R., & Jardine, R.J. (2002). On the applicability of medium density chalk under a wide range of pressure
cross anisotropic elasticity to granular materials at very conditions. Soil and Foundation, 63(1), 101268. http://
small strains. Géotechnique, 52(10), 727-749. http://doi. doi.org/10.1016/j.sandf.2022.101268.
org/10.1680/geot.2002.52.10.727. Liu, T., Jardine, R.J., Vinck, K., Ahmadi-Naghadeh, R.,
Lehane, B.M., Liu, Z., Bittar, E., Nadim, F., Lacasse, S., Kontoe, S., Buckley, R.M., Byrne, B.W., & McAdam,
Jardine, R., Carotenuto, P., Rattley, M., & Gavin, K. R.A. (2023c). Cyclic characterisation of low-to-medium
(2020). A new ‘unified’ CPT-based axial pile capacity density chalk for offshore driven pile design. In Proceedings
design method for driven piles in sand. In Proceedings of of the 9th International offshore Site Investigation and
the 4th International Symposium on Frontiers in Offshore Geotechnics Conference, London.
Geotechnics (pp. 462-477), Austin, TX, USA. Mao, X., & Fahey, M. (2003). Behaviour of calcareous soils
Lehane, B.M., Liu, Z., Bittar, E.J., Nadim, F., Lacasse, S., in undrained cyclic simple shear. Géotechnique, 53(8),
Bozorgzadeh, N., Jardine, R.J., Ballard, J.C., Carotenuto, 715-727. http://doi.org/10.1680/geot.2003.53.8.715.
P., Gavin, K., Gilbert, R.B., Bergan-Haavik, J., Jeanjean, Muraro, S., & Jommi, C. (2019). Implication of end restraint
P., & Morgan, N. (2022). CPT-based axial capacity design in triaxial tests on the derivation of stress–dilatancy
method for driven piles in clay. Journal of Geotechnical rule for soils having high compressibility. Canadian
and Geoenvironmental Engineering, 148(9), 04022069. Geotechnical Journal, 56(6), 840-851. http://doi.
http://doi.org/10.1061/(ASCE)GT.1943-5606.0002847. org/10.1139/cgj-2018-0343.
Liu, T. (2018). Advanced laboratory testing for offshore Nishimura, S. (2014). Assessment of anisotropic elastic
pile foundations under monotonic and cyclic loading parameters of saturated clay measured in triaxial apparatus:
[Doctoral thesis]. Imperial College London, London, appraisal of techniques and derivation procedures. Soil
UK. http://doi.org/10.25560/101146. and Foundation, 54(3), 364-376. http://doi.org/10.1016/j.
Liu, T., Chen, H., Buckley, R., Quinteros, S., & Jardine, sandf.2014.04.006.
R.J. (2019a). Characterisation of large-displacement Ove Arup and Partners. (1988). Research on the behaviour
interface shearing behaviour for application in design of
of piles as anchors for buoyant structures (Offshore
steel piles driven in sands. In A. Tarantino, & E. Ibraim
Technology Report, No. OTH 86 215). London: Department
(Eds.), Proceedings of the 7th International Symposium on
of Energy, 80 p.
Deformation Characteristics of Geomaterials, Glasgow,
Page, A.M., Klinkvort, R.T., Bayton, S., Zhang, Y., & Jostad,
UK. http://doi.org/10.1051/e3sconf/20199213001.
H.P. (2021). A procedure for predicting the permanent
Liu, T., Quinteros, S., Jardine, R.J., Carraro, J.A.H., &
rotation of monopiles in sand supporting offshore wind
Robinson, J. (2019b). A unified database of ring-shear
turbines. Marine Structures, 75, 102813. http://doi.
interface tests on sandy-silty soils. In Proceedings of
org/10.1016/j.marstruc.2020.102813.
the XVII European Conference on Soil Mechanics and
Geotechnical Engineering (ECSMGE-2019), Reykjavik, Pedone, G., Kontoe, S., Zdravković, L., Jardine, R.J., Vinck,
Iceland. http://doi.org/10.32075/17ECSMGE-2019-0268. K., & Liu, T. (2023). Numerical modelling of laterally
Liu, T., Ushev, E.R., & Jardine, R.J. (2020). Anisotropic loaded piles driven in low-to-medium density fractured
stiffness and shear strength characteristics of a stiff glacial chalk. Computers and Geotechnics, 156, 105252. http://
till. Journal of Geotechnical and Geoenvironmental doi.org/10.1016/j.compgeo.2023.105252.
Engineering, 146(12), 04020137. http://doi.org/10.1061/ Powell, J.J.M., & Butcher, A.P. (2003). Characterisation of
(ASCE)GT.1943-5606.0002387. a glacial till at Cowden, Humberside. In T.S. Tan, K.K.
Liu, T., Jardine, R.J., Vinck, K., & Ackerley, S.K. (2022a). Phoon, D.W. Hight, & S. Leroueil (Eds.), Proceedings
Optimization of advanced laboratory monotonic and of the International Workshop on Characterisation and
cyclic triaxial testing on fine sands. Geotechnical Testing Engineering Properties of Natural Soils (Vol. 2, pp.
Journal, 45(6), 1087. http://doi.org/10.1520/GTJ20210190. 983-1020), Singapore.
Liu, T., Ahmadi-Naghadeh, R., Vinck, K., Jardine, R.J., Rowe, P., & Barden, L. (1964). Importance of free ends
Kontoe, S., Buckley, R.M., & Byrne, B.W. (2022b). in triaxial testing. Journal of the Soil Mechanics and
An experimental investigation into the behaviour of Foundations Division, 90(1), 1-27. http://doi.org/10.1061/
de-structured chalk under cyclic loading. Géotechnique, JSFEAQ.0000586.
1-13. In press. http://doi.org/10.1680/jgeot.21.00199. Selby, M., Zhou, H., Liu, T., & Boylan, N. (2023). Back
Liu, T., Ushev, E.R., & Jardine, R.J. (2023a). Anisotropy analysis of PISA pile tests at Dunkirk using a practical
and cyclic loading characteristics of a stiff Bolders Nor-Sand model. In Proceedings of the 9th International
Bank glacial till at Cowden. In Proceedings of the 8th Conference on Offshore site Investigations and Geotechnics.
International Symposium on Deformation Characteristics London: Society for Underwater Technology.
of Geomechanics (IS-Porto), Porto. Taborda, D.M.G., Zdravkovic´, L., Kontoe, S., & Potts, D.M.
Liu, T., Ferreira, P.M.V., Vinck, K., Coop, M.R., Jardine, (2014). Computational study on the modification of a
R.J., & Kontoe, S. (2023b). The behaviour of a low-to- bounding surface plasticity model for sands. Computers

and Geotechnics, 59, 145-160. http://doi.org/10.1016/j. Geotechnical Journal, 60(10), 1475-1490. http://doi.
compgeo.2014.03.005. org/10.1139/cgj-2022-0441.
Taborda, D.M.G., Zdravković, L., Potts, D.M., Burd, H.J., Vinck, K., Liu, T., Jardine, R.J., Byrne, B.W., & Willow, A.
Byrne, B.W., Gavin, K.G., Houlsby, G.T., Jardine, R.J., (2023b). The effects of steel corrosion on the interface
Liu, T., Martin, C.M., & McAdam, R.A. (2020). Finite- shearing behaviour of chalk. In Proceedings of the 9th
element modelling of laterally loaded piles in a dense Int. Conf on Offshore site Investigations and Geotechnics.
marine sand at Dunkirk. Géotechnique, 70(11), 1014-1029. London: Society for Underwater Technology.
http://doi.org/10.1680/jgeot.18.PISA.006. Wen, K., Kontoe, S., Jardine, R.J., & Liu, T. (2023a). An
Ushev, E.R. (2018). Laboratory investigation of the mechanical axial load transfer model for piles driven in chalk. Journal
properties of Cowden till under static and cyclic conditions of Geotechnical and Geoenvironmental Engineering,
[Doctoral thesis]. Imperial College London, London, UK. 149(11), 04023107. http://doi.org/10.1061/JGGEFK.
http://doi.org/10.25560/78220. GTENG-11368.
Ushev, E.R., & Jardine, R.J. (2022a). The mechanical behaviour Wen, K., Kontoe, S., Jardine, R.J., Liu, T., & Pan, L. (2023b).
of Bolders Bank till. Canadian Geotechnical Journal, Non-linear finite-element analysis of axially loaded piles
59(12), 2163-2183. http://doi.org/10.1139/cgj-2021-0436. driven in chalk. In L. Zdravković, S. Kontoe, D.M.G.
Ushev, E.R., & Jardine, R.J. (2022b). The behaviour of Taborda, & A. Tsiampousi (Eds.), Proceedings of the
Bolders Bank glacial till under undrained cyclic loading. 10th European Conference on Numerical Methods in
Géotechnique, 72(1), 1-19. http://doi.org/10.1680/ Geotechnical Engineering, London. http://doi.org/10.53243/
jgeot.18.P.236. NUMGE2023-337.
Vinck, K. (2016). Laboratory testing of Dunkerque sand Yang, Z.X., Guo, W.B., Jardine, R.J., & Chow, F. (2017).
[Master’s dissertation]. Imperial College, London, UK.. Design method reliability assessment from an extended
Vinck, K., Liu, T., Ushev, E., & Jardine, R.J. (2019). An database of axial load tests on piles driven in sand.
appraisal of end constraint effects on triaxial monotonic Canadian Geotechnical Journal, 54(1), 59-74. http://
and cyclic tests on a range of geomaterials. In A. doi.org/10.1139/cgj-2015-0518.
Tarantino, & E. Ibraim (Eds.), Proceedings of the 7th Zdravković, L., Jardine, R.J., Taborda, D.M.G., Abadias,
International Symposium on Deformation Characteristics D., Burd, H.J., Byrne, B.W., Gavin, K., Houlsby, G.T.,
of Geomaterials, Glasgow, UK. http://doi.org/10.1051/ Igoe, D., Liu, T., Martin, C.M., McAdam, R.A., Muir-
e3sconf/20199202007. Wood, A., Potts, D.M., Skov Gretlund, J., & Ushev, E.
Vinck, K. (2021). Advanced geotechnical characterisation (2020a). Ground characterisation for PISA pile testing
to support driven pile design at chalk sites [Doctoral and analysis. Géotechnique, 70(11), 945-960. http://doi.
thesis]. Imperial College, London, UK. Retrieved in org/10.1680/jgeot.18.PISA.001.
August 23, 2023, from https://spiral.imperial.ac.uk/ Zdravković, L., Taborda, D.M.G., Potts, D.M., Abadias, D.,
handle/10044/1/107416 Burd, H.J., Byrne, B.W., Gavin, K., Houlsby, G.T., Jardine,
Vinck, K., Liu, T., Jardine, R.J., Kontoe, S., Ahmadi- R.J., Martin, C.M., McAdam, R.A., & Ushev, E. (2020b).
Naghadeh, R., Buckley, R.M., Byrne, B.W., Lawrence, Finite element modelling of laterally loaded piles in a
J., Mcadam, R.A., & Schranz, F. (2022). Advanced in-situ stiff glacial clay till at Cowden. Géotechnique, 70(11),
and laboratory characterisation of the ALPACA chalk 999-1013. http://doi.org/10.1680/jgeot.18.PISA.005.
research site. Géotechnique, 1-15. In press. http://doi. Zdravković, L., Potts, D.M., & Taborda, D.M.G. (2021).
org/10.1680/jgeot.21.00197. Integrating laboratory and field testing into advanced
Vinck, K., Liu, T., Mawet, J., Kontoe, S., & Jardine, R.J. geotechnical design. Geomechanics for Energy and
(2023a). Field tests on large scale instrumented piles the Environment, 27, 100216. http://doi.org/10.1016/j.
driven in chalk: results and interpretation. Canadian gete.2020.100216.

Liu, Vinck, Ushev and Jardine (2024) In-Situ and Laboratory Characterisation of Stiff and Dense Geomaterials For Driven Pile Analysis and Design

Uploaded by

Copyright:

Available Formats

Liu, Vinck, Ushev and Jardine (2024) In-Situ and Laboratory Characterisation of Stiff and Dense Geomaterials For Driven Pile Analysis and Design

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Liu, Vinck, Ushev and Jardine (2024) In-Situ and Laboratory Characterisation of Stiff and Dense Geomaterials For Driven Pile Analysis and Design

Uploaded by

Copyright:

Available Formats

Soils and Rocks

In-situ and laboratory characterisation of stiff and dense

1. Introduction offshore windfarms around the UK and northern European

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 1

Table 1. Typical index properties of the three geomaterials.

2 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 3

4 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Figure 5 shows the degradation of normalised secant

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 5

Figure 6. Typical undrained triaxial compression and extension

6 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

important to recognise when analysing lateral pile loading.

3.3.2 Non-linear stiffness characteristics over full strain

Following Brosse et al. (2017), the αdσ-controlled

Cyclic loading, which can be critical for offshore

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 7

8 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 9

4.2.2 Non-linear stiffness and yielding

Vinck (2016) and Liu (2018) investigated Dunkirk

10 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Table 2. Cross-anisotropy stiffness coefficients of Dunkirk sand.

when specimens had 2:1 initial height-to-diameter ratios and

4.2.3 Large-strain and critical state behaviour

End constraint also influences medium-to-large strain

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 11

4.3 Long-term drained cyclic loading behaviour

Aghakouchak (2015) and Aghakouchak et al. (2015)

12 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 13

surfaces in triaxial q-p′ space. Intact (cemented) chalk

14 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Figure 23. Full-strain triaxial compression behaviour of intact

Figure 25. Yield stresses and interpreted pre-failure Y1 and large-

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 15

Figure 27. Typical unstable cycling response of intact chalk:

qmax 1 + qcyc / qmean (6)

in which f(N) represents for each Nf line its projected intercept

16 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 17

6.3 Pile behaviour under monotonic lateral loading

Modelling of field lateral pile loading response is

18 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 19

20 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 21

22 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323 23

24 Liu et al., Soil. Rocks, São Paulo, 2024 47(3):e2024009323

You might also like