International Journal of High-Rise Buildings

International Journal of
December 2015, Vol 4, No 4, 227-239
High-Rise Buildings
www.ctbuh-korea.org/ijhrb/index.php

Geotechnical Parameter Assessment

for Tall Building Foundation Design
Harry G. Poulos1,† and Frances Badelow2
1
Coffey Geotechnics, Australia 19F, Citadel Tower B, 799 Pacific Highway, Chatswood, NSW Australia 2067
2
Mott MacDonald, L10, 383 Kent Street, Sydney NSW 2000, Q1678, Australia

Abstract

This paper discusses the design parameters that are required for the design of high-rise building foundations, and suggests
that the method of assessment for these parameters should be consistent with the level of complexity involved for various stages
in the design process. Requirements for effective ground investigation are discussed, together with relevant in-situ and
laboratory test techniques for deriving the necessary strength and stiffness parameters. Some empirical correlations are also
presented to assist in the early stages of design, and to act as a check for parameters that are deduced from more detailed testing.
Pile load testing is then discussed and a method of interpreting bi-directional tests to obtain pile design parameters is outlined.
Examples of the application of the assessment process are described, including high-rise projects in Dubai and Saudi Arabia

Keywords: Case histories; foundations; geotechnical parameters; laboratory testing; in-situ testing; pile load tests

1. Introduction conditions, cost and project program requirements. Often

the subsurface conditions at high-rise building sites are
“Super-tall” buildings in excess of 300 m in height are far from ideal. Geotechnical uncertainty is the one of the
presenting new challenges to engineers, particularly in re- greatest risks in the foundation design and construction
lation to structural and geotechnical design. Many of the process. Establishing an accurate knowledge of the ground
traditional design methods cannot be applied with any conditions is essential in the development of economical
confidence since they require extrapolation well beyond foundation systems which perform to expectations.
the realms of prior experience, and accordingly, structural This paper summarises the process of characterising the
and geotechnical designers are utilizing more sophisticated ground conditions and quantifying the relevant geotechni-
methods of analysis and design. In particular, geotechnical cal parameters required for foundation design. The various
engineers involved in the design of foundations for super- stages of geotechnical assessment are discussed, together
tall buildings are leaving behind empirical methods and with the techniques available for estimating the geotech-
increasingly employing state-of-the art methods. nical parameters. Two case histories will be described
Foundation and building behaviour is highly interactive, briefly to illustrate the application of these principles to
with movements of the foundation resulting from the buil- practice.
ding loads, which in turn influence the behaviour of the
building. Foundation behaviour is mainly governed by the 2. Ground Information for Parameter
prevailing ground conditions, the foundation type, plus the Assessment
magnitude and distribution of the building loads. The foun-
dation design of high-rise buildings should therefore be In the assessment of geotechnical parameters for foun-
considered as a performance based soil-structure interac- dation design, it is first necessary to review the geology
tion (SSI) issue and not limited to traditional empirically of the site and identify any geological features that may
based design methods, such as a bearing capacity app- influence the design and performance of the foundations.
roach with an applied factor of safety. A desk study is usually the first step, followed by site
The type of foundation system for a high-rise building visits to observe the topography and any rock or soil
is dependent on the main design elements (building loads, exposures. Local experience, coupled with a detailed site
ground conditions and required building performance) as investigation program, is then required. The site investi-
well as other important factors like local construction gation is likely to include a comprehensive borehole drill-
ing and in-situ testing program together with a suite of
†
Corresponding author: Harry G. Poulos laboratory tests to characterize strength and stiffness pro-
Tel: +61-420-960-248 perties of the subsurface conditions. Based on the findings
E-mail: Harry.Poulos@coffey.com of the site investigation, the geotechnical model and asso-
228 Harry G. Poulos and Frances Badelow | International Journal of High-Rise Buildings

ciated design parameters are developed, which are then meters require assessment:
used in the foundation design process. 1. The ultimate skin friction for the various strata along
Increasingly, geophysical methods are being used to the pile.
supplement data from conventional borehole drilling. Such 2. The ultimate end bearing resistance for the founding
methods, which include downhole and cross-hole techni- stratum.
ques, have a number of major benefits, including: 3. The ultimate lateral pile-soil pressure for the various
1. They provide a means of identifying the stratigraphy strata along the pile.
between boreholes; 4. The ultimate bearing capacity of the raft.
2. The can identify localized anomalies in the ground 5. The stiffness of the soil strata supporting the piles,
profile (e.g., cavities, sinkholes or localised pockets in the vertical direction.
of softer or harder material); 6. The stiffness of the soil strata supporting the piles,
3. They can identify bedrock levels; in the horizontal direction.
4. They provide quantitative measurements for the 7. The stiffness of the soil strata supporting the raft.
shear wave and compression wave velocities. This It should be noted that the soil stiffness values are not
information can be used to estimate the in-situ values unique values but will vary, depending on whether long-
of soil stiffness at small strains, and hence to provide term values are required (for long-term settlement estima-
a basis for quantifying the deformation properties of tes) or short-term values are required (for dynamic res-
the soil strata. ponse to wind and seismic forces). For dynamic response
The site investigation works are desirably supplemented of the structure-foundation system, an estimate of the
with a program of instrumented vertical and lateral load internal damping of the soil is also required, as it may
testing of prototype piles (e.g., bi-directional load cell provide the main source of foundation damping.
(Osterberg Cell) tests) to allow calibration of the founda- Moreover, the soil stiffness values will generally vary
tion design parameters and hence to better predict the with applied stress or strain level, and will tend to dec-
foundation performance under loading. Completing the rease as the stress and strain levels increase.
load tests on prototype piles prior to final design can pro-
vide conformation of performance (i.e., pile construction, 5. Empirical Relationships for Preliminary
pile performance, ground behaviour and properties) or Design
else may provide data for modifying the design prior to
construction. For piles in soil, initial assessments for preliminary des-
ign are often based on the results of simple in-situ tests
3. Stages of Design such as the Standard Penetration Test (SPT) and the Static
Cone Penetration Test (CPT). For piles in rock, the correla-
There are commonly three broad stages in foundation ting factor is usually the unconfined compressive strength
design: (UCS). Some common correlations are summarized below.
1. A preliminary design, which provides an initial basis
for the development of foundation concepts and 5.1. Correlations with SPT
costing. Typical of the correlations that the authors have emplo-
2. A detailed design stage, in which the selected found- yed are those based on the work of Decourt (1982, 1995)
ation concept is analysed and progressive refine- using the SPT:
ments are made to the layout and details of the foun- Raft ultimate bearing capacity:
dation system. This stage is desirably undertaken
collaboratively with the structural designer, as the pur = K1. Nr kPa (1)
structure and the foundation are an interactive system.
3. A final design phase, in which both the analysis and Pile ultimate shaft resistance:
the parameters employed in the analysis are finalized.
It should be noted that the parameters used for each fs = a. [2.8 Ns + 10] kPa (2)
stage may change as knowledge of the ground conditions
increases, and the results of in-situ and laboratory testing Pile ultimate base resistance:
become available. The parameters for the final design stage
should desirably incorporate the results of foundation load fb = K2 . Nb kPa (3)
tests.
Soil Young’s modulus below raft:
4. Design Parameters
Esr = 2Nr MPa (4)
Many contemporary foundation systems incorporate
both piles and a raft, and in such cases the following para- Young’s modulus along and below pile (vertical loading):
 Geotechnical Parameter Assessment for Tall Building Foundation Design 229

Table 1. Correlation factors K1 and K2 (after Decourt, 1995) 5.2. Correlations with CPT
K2 K2 Two broad approaches have been adopted for utilising
K1
Soil Type Displacement Non-Displacement CPT data to predict pile capacity:
(Raft)
Piles Piles 1. Correlations between cone resistance values and both
Sand 90 325 165 ultimate shaft friction and ultimate base capacity
Sandy silt 80 205 115 (e.g., De Ruiter and Beringen 1979; Bustamante and
Clayey silt 80 165 100 Gianeselli 1982; Poulos 1989)
Clay 65 100 80 2. Correlations between sleeve friction and ultimate
pile shaft friction (Schmertmann 1975, 1978). This
Table 2. Typical Parameters for Small-strain Shear Modu- approach is often considered to be less reliable than
lus Correlations (after Hasancebi and Ulusay, 2007) the above method because of the difficulties in accu-
Soil Type X y rately measuring sleeve friction.
Sandy soils 90.8 0.32 A useful adaptation of the method of Bustamante and
Clayey Soils 97.9 0.27 Gianeselli (1982) is summarised by Frank and Magnan
All soils 90.0 0.31 (1995). The ultimate shaft friction fs and base capacity fb
are given by the following expressions:

Es = 3N MPa (5) fs = qc / ks ≤ fsl (7)

where Nr = average SPT (N60) value within the depth fb = kb · qc (8)

of one-half of the raft width
Ns = SPT value along pile shaft where qc = measured cone tip resistance
Nb = average SPT value close to pile tip ks = shaft factor
K1, K2 = factors shown in Table 1 fsl = limiting ultimate shaft friction
a = 1 for displacement piles in all soils & non- kb = base factor.
displacement piles in clays Table 3 gives recommended values of ks and fsl, which
a = 0.5-0.6 for non-displacement piles in depend on soil type and pile type. Values of kb are given
granular soils. in Table 4. Here, the value of qc used in Eq. (8) should be
Many correlations have been proposed to relate the the average value within a distance of 1.5 base diameters
small-strain shear modulus G0 to the SPT-N value. These above and below the base of the pile. Excessively large
generally take the following form: and low values are excluded from the average (Bustamante
and Gianeselli, 1982).
G0 ≈ X [N1(60)]y MPa (6) For square or circular shallow footings and rafts, MELT
(1993) suggests the following correlation for ultimate
where N1(60) = SPT value, corrected for overburden bearing capacity pur:
pressure and hammer energy
X and y are parameters that may depend on pur = a1[1 + a2.D/B] qc + q0 (9)
soil type.
Typical values of X and y are shown in Table 2. where a1, a2 are parameters depending on soil type
and condition

Table 3. Ultimate Shaft Friction Correlation Factors for CPT Tests (MELT, 1993)
Clay and Silt Sand and Gravel Chalk
Pile Type
Soft Stiff Hard Loose Med. Dense Soft Weathered
ks - - 751 - 200 200 200 125 80
Drilled
fsl (kPa) 15 40 80 40 80 - - 120 40 120
Drilled removed ks - 100 1002 - 1002 250 250 300 125 100
casing fsl (kPa) 15 40 60 40 80 - 40 120 40 80
Steel driven ks - 120 150 300 300 300
closed-ended (3)
fsl (kPa) 15 40 80 - - 120
ks - 75 - 150 150 150
Driven concrete (3)
fsl (kPa) 15 80 80 - - 120
(1) trimmed and grooved at the end of drilling
(2) dry excavation, no rotation of casing
(3) in chalk, fs can be very low for some types of piles; a specific study is needed.
230 Harry G. Poulos and Frances Badelow | International Journal of High-Rise Buildings

Table 4. Base Capacity Factors for CPT (after MELT, 1993) In employing such correlations, it should be recognised
Soil Type qc (MPa) kb kb that, in the field, they may be influenced by geological
A soft <3 features and structure that cannot be captured by a small
Clay and generally intact rock sample. Nevertheless, in the ab-
B stiff 3-6 0.40 0.55
Silt sence of other information, such correlations provide at
C hard (clay) >6
A loose <5 least an indication of the order of magnitude.
Sand More detailed correlations for rock mass modulus are
B medium 8-15 0.15 0.50
Gravel provided by Hoek and Diederichs (2006), who relate the
C dense >20
A soft <5 0.20 0.30
rock mass modulus to the Geological Strength Index, GSI,
Chalk and a disturbance factor that reflects the geological struc-
B weathered >5 0.30 0.45
ture.
ND=non displacement pile; D=displacement pile
5.4. Parameters for lateral pile response
Table 5. Parameters A and B for Ultimate Bearing Capacity The above correlations are for vertical loading on piles
of Square Shallow Footings and Rafts (after MELT, 1993)
and rafts. For lateral response analyses of piles, the above
Soil Type Condition a1 a2 correlations need to be modified, and as a first approxi-
Clay, silt All 0.32 0.35 mation, the following adjustments are suggested:
Loose 0.14 0.4 1. Young’s modulus values for vertical loading should
Sand, gravel Medium 0.11 0.4 be reduced by multiplying by a factor of 0.7, to allow
Dense 0.08 0.4 for the greater soil strain levels arising from lateral
Chalk - 0.17 0.27 loading.
2. The ultimate lateral pile-soil pressure, py, can be
approximately related to the ultimate end bearing fb,
q0 = overburden pressure at level of base as follows:
D = depth of embedment below surface
B = width of footing or raft. py = η · fb (11)
Table 5 gives recommended values of a1 and a2.
The small strain shear modulus G0 has also been corre- where η = 0.22 (1+z/d) ≤ 1.0
lated to the measured CPT value, for example, as sugge- z = depth below ground surface
sted by Mayne et al. (2009) and Mayne and Rix (1993). d = pile diameter or width.
The latter source gives the following relationship:
6. Laboratory Testing
G0 = 406(qc)0.695 / e01.130 kPa (10)
6.1. Triaxial and stress path testing
where qc = cone resistance, in kPa Conventional triaxial testing is of limited value for
e0 = initial void ratio. assessing design parameters for pile foundations, as the
method of stress application does not reflect the way in
5.3. Correlations with unconfined compressive which load transfer occurs from the piles to the surroun-
strength ding soil. However, cyclic triaxial testing may be useful
For piles in rock, it is common to correlate design in providing an indication of the potential degradation
parameters with the unconfined compressive strength, qu, effects on the stiffness/strength properties of the founda-
at least for preliminary purposes. Some of the available tion ground material due to cyclic loading. For the Burj
correlations are summarised in Table 6. Khalifa project, cyclic triaxial test results indicated that a

Table 6. Correlations of design parameters for piles in rock

Parameter Correlation Remarks
a0 can vary from about 0.1 for extremely poor quality rock to 24 for intact
Ultimate bearing capacity (raft) pur = a0 qu high-strength rock (Merifield et al, 2006). A value of 2 is likely to be rea-
sonable and conservative in many cases.
a generally varies between 0.20 and 0.45
Ultimate shaft friction, fs fs = a (qu) b
b in most correlations is 0.5
a1 generally varies between 3 and 5
Ultimate end bearing, fb fb = a1 (qu) b1 b1 in most correlations is 1.0, although Zhang and Einstein (1998) adopt b1
= 0.5
Young’s modulus for a2 varies between about 100 and 500 for a wide range of rocks.
Esv = a2 (qu) b2
vertical loading, Esv b2 is generally taken as 1.0
 Geotechnical Parameter Assessment for Tall Building Foundation Design 231

degree of degradation was possible in the mass ground

strength/stiffness properties. However, under the anticipa-
ted applied loading, the foundations would be loaded to
small strain levels and hence the potential degradation of
strength and stiffness would be limited.
More sophisticated stress path testing can be useful in
providing stiffness parameters over a range of stress app-
ropriate to the foundation system. Such testing can be used
to compare with values from other means of assessment.

6.2. Resonant column testing

The resonant column test is commonly used for labora-
tory measurement of the low-strain properties of soils. It
subjects solid or hollow cylindrical specimens to torsional
or axial loading by an electromagnetic loading system,
usually harmonic loads for which frequency and ampli- Figure 1. Constant Normal Stiffness Test Setup.
tude can be controlled. It can be used to measure the small
strain shear modulus and damping ratio of a soil or rock
sample, and the variation in modulus and damping ratio parameters via correlations such as those mentioned above.
with increasing shear strain level. Such data are valuable
for carrying out dynamic response analyses of the foun- 7.2. Pressuremeter testing
dation system. Pressuremeter testing can be used to estimate both
strength and stiffness properties of the ground. The inter-
6.3. Constant Normal Stiffness (CNS) testing pretation of test data is discussed by Briaud (1992) and
It is generally accepted by practitioners that there is no Mair and Wood (1987). The stiffness values relevant to
suitable laboratory test that can be reliably used to meas- foundation design are generally the values derived from
ure the ultimate shaft friction fs. However, there has been an unload/reload loop.
a significant development over the past 20-25 years in
direct shear testing of interfaces, with the development of 7.3. Geophysical testing
the constant normal stiffness (CNS) test (Ooi and Carter, Geophysical testing is becoming more widely used in
1987; Lam and Johnston, 1982). The basic concept of this geotechnical investigations, offering at least three major
test is illustrated in Fig. 1, and involves the presence of a advantages:
spring of appropriate stiffness against which the normal 1. Ground conditions between boreholes can be
stress on the interface acts. This test provides a closer inferred.
simulation of the conditions at a pile-soil interface than 2. Depths to bedrock or a firm bearing stratum can be
the conventional constant normal stress direct shear test. estimated.
The normal stiffness Kn can be “tuned” to represent the 3. Shear wave velocities in the various layers within
restraint of the soil surrounding the pile, and is given by: the ground profile can be measured, and tomogra-
phic images developed to portray both vertical and
Kn = 4Gs / d (12) lateral inhomogeneity.
4. From the measured shear wave velocity (vs), the
where Gs = shear modulus of surrounding soil small-strain shear modulus, G0, can be obtained as
d = pile diameter. follows:
The effects of interface volume changes and dilatancy
can be tracked in a CNS test, and the results are parti- G0 = ρ vs2 (13)
cularly enlightening when cyclic loading is applied, as
they demonstrate that the cyclic degradation is due to the where ρ = mass density of soil.
reduction in normal stress arising from the cyclic displace- From G0, allowance must be made for the effects of
ments applied to the interface. shear strain within the soil which will lead to a reduction
in the secant modulus value that may be useful for routine
7. In-situ Testing design, as discussed below.

7.1. Penetration testing 8. Derivation of Secant Values of Soil

Conventional SPTs and CPTs are usually undertaken as a Modulus for Foundation Analysis
means of classifying and approximately quantifying the soil
strata, and of facilitating estimation of geotechnical design For application to routine design, allowance must be
232 Harry G. Poulos and Frances Badelow | International Journal of High-Rise Buildings

ance is made for strain level effects, modulus values deri-

ved from geophysical tests can correlate well with those
from pressuremeter tests. Fig. 3 reproduces such an exam-
ple in which a reduction factor of 0.2 has been applied to
the small-strain modulus values derived from cross-hole
seismic test results. The modulus values so derived were
found to be consistent with values obtained from subse-
quent pile load tests.

9. Pile Load Testing

9.1. Introduction
From the designer’s viewpoint, pile load testing should
ideally be able to satisfy the following requirements:
• provide information on all the above design issues
Figure 2. Typical example of ratio of secant shear modulus • be able to be undertaken on pre-production, prototype
to small-strain value (Poulos et al., 2001).
piles
• be able to be undertaken on any of the production
piles without special preparation
made for the reduction in the shear modulus because of • be relatively inexpensive
the relatively large strain levels that are relevant to • provide reliable and unequivocal information which
foundations under normal serviceability conditions. As an can be applied directly to the design process.
example, Poulos et al. (2001) have suggested the reduc- Some of the common methods of pile testing are sum-
tion factors for clay soils shown in Fig. 2 for the case where marised below, and then suggestions for the interpretation
G0/su=500 (su=undrained shear strength). This figure indi- of the tests are offered.
cates that:
1. The secant modulus for axial loading is about 30- 9.2. Types of test
40% of the small-strain value for normal factors of 9.2.1. Static vertical load test
safety. This test type is the most fundamental and involves the
2. The secant modulus for lateral loading is smaller application of vertical load directly to the pile head, usu-
than that for axial loading, typically 30% for compa- ally via a series of increments. Test procedures have been
rable factors of safety. developed and specified by various codes, for example,
Haberfield (2013) has demonstrated that, when allow- ASTM D1143. The static load test is generally regarded

Figure 3. Comparison of modulus values from pressuremeter and cross-hole seismic tests (Haberfield, 2013).
 Geotechnical Parameter Assessment for Tall Building Foundation Design 233

as the definitive test and the one against which other types stiffness at the design load. However, it is expected to be
of test are compared. The test may take a variety of forms, increasingly inaccurate as the load level approaches the
depending on how the reaction for the applied pile load- ultimate value.
ing is supplied. This is the type of test that the designer For heavily loaded foundations such as those supporting
would like to carry out, as it best simulates the way in tall buildings, dynamic load testing is generally not fea-
which a structural load is applied to the pile. Unfortuna- sible as insufficient energy can be imparted to the pile to
tely, the ideal test cannot usually be achieved in practice, fully mobilise its capacity. The test may however provide
and the reaction system interacts with the test pile, thus a convenient means of obtaining the head stiffness of a
creating some problems with the interpretation of the test single pile.
data.
The usual basic information from such a test is the load- 9.2.4. Bi-directional (Osterberg cell) test
settlement relationship, from which the load capacity and This test was developed by Osterberg (1989) while a
pile head stiffness can be interpreted. However, such inter- similar test has been developed in Japan (Fujioka and
pretation should be carried out with caution, as the meas- Yamada, 1994), and has been used increasingly over the
ured pile settlement may be influenced by interaction bet- past decade or so. A special cell is cast in or near the pile
ween the test pile and the reaction system. Such interac- base, and pressure is applied. The base is jacked down-
tion tends to lead to over-estimates of both capacity and wards while the shaft provides reaction and is jacked
stiffness, and therefore can lead to unconservative results, upwards. The test can continue until the element with the
unless appropriate allowances are made for the effects of smaller capacity reaches its ultimate resistance. Using the
the interaction between the test pile and the reaction and/ Osterberg cell, load tests of up to 150 MN have been
or settlement measuring system. carried out. Despite its ability to provide “self-reaction”,
the Osterberg cell test (like all tests) has its limitations
9.2.2. Static lateral load test and shortcomings, including the following:
There are several forms of lateral load test, but the most 1. it is applicable primarily to bored piles
common and convenient is that which involves the jack- 2. the cell must be pre-installed prior to the test
ing of one pile against one or more other piles; for exam- 3. there is interaction between the base and the shaft,
ple, ASTM Standard D3966 outlines a procedure for lat- and each will tend to move less than the “real”
eral load testing and test interpretation. movement so that the apparent shaft and base
As with the static vertical load test, there are “side stiffness will tend to be larger than the real values.
effects” if two piles are jacked against other piles. In par-
ticular, because the direction of loading of each pile is 9.2.5. Statnamic test
different, the interaction between the piles will tend to Statnamic testing was jointly developed in Canada and
cause a reduced pile head deflection, and as a conse- the Netherlands (Middendorp et al., 1992; Bermingham et
quence, the measured lateral stiffness of the pile will be al., 1989), and has also found considerable use and devel-
greater than the true value. opment in Japan. Comparative tests on piles subjected to
conventional static testing and Statnamic testing have
9.2.3. Dynamic load test shown good agreement in load-settlement performance
The principles of the dynamic load test are very well- (Bermingham et al., 1994).
established (Rausche et al., 1985; Goble, 1994). The test Statnamic testing appears to offer some advantages over
procedure is now accepted as routine, especially for qua- other test types, including:
lity control and design confirmation purposes. Despite its 1. the test is quick and easily mobilized
widespread use, the dynamic pile load test has a number 2. high loading capacity is available
of potential limitations, including the fact that the load- 3. the loading is accurately centred and can be applied
settlement behaviour estimated from the test is not unique, to both single piles and pile groups
but is a best-fit estimate. Two measurements (strain and 4. the test does not require any pre-installation of the
acceleration versus time) are taken, and from these, the loading equipment
complete distribution of resistance along the pile, as well 5. the test is quasi-static, and does not involve the dev-
as the load-settlement behaviour, are interpreted. Also, the elopment of potentially damaging compressive and
load is applied far more rapidly than in most situations in tensile stresses in the test pile.
practice, and hence time-dependent settlements are not Inevitably, there are also some potential shortcomings,
developed during the test. Fortunately, under normal des- including:
ign load levels, the amount of time-dependency (from both 1. certain assumptions need to be made in the interpre-
consolidation and creep) is relatively small as most of the tation of the test, especially in relation to the unload-
settlement arises from shear deformation at or near the ing of the pile
pile-soil interface. Hence, the dynamic test may give a rea- 2. it cannot provide information on time-dependent
sonable (if over-estimated) assessment of the pile head settlements or movements. While this may not be of
234 Harry G. Poulos and Frances Badelow | International Journal of High-Rise Buildings

great importance for single piles, it can be a major then the parameters for the curves derived via a process
limitation when testing pile groups, especially if of trial and error.
compressible layers underlie the pile tips. If an elastic-plastic soil model is assumed, then a distri-
bution of Young’s modulus and ultimate shaft friction
9.3. Test interpretation with depth must be assumed and again, a trial and error
9.3.1. Ultimate axial capacity process will generally be required to obtain a fit between
For conventional static load testing, it is common for the load-settlement behaviour from the theoretical model
the test to be stopped prior to complete plunging failure and the measured load settlement behaviour. More often
being achieved. A vast number of suggestions have been than not, there will be no instrumentation along the pile
made on how the ultimate axial load capacity can be esti- so that there is no detailed load transfer information along
mated from such tests, some of which have been reviewed the pile shaft. Thus, an assumption has to be made regar-
and assessed by Hwang et al. (2003). They can be classi- ding the distribution of soil stiffness and strength with
fied into the following categories: depth. This needs to be done in relation to the geotech-
1. “Conspicuous turning point of the load-settlement nical profile in order to obtain reliable results.
curve”. This is often a subjective assessment. If instrumentation has been installed in the pile, and if
2. Settlement S of the pile head, including: proper account is taken of residual stresses in the inter-
a. S =10% of diameter typically (Terzaghi, 1943). pretation of the results, then the value of Young’s modulus
b. Tangent flexibility of pile head, for example, of the ground, Es, between each adjacent set of instrum-
Fuller & Hoy (1971). entation can be interpreted by use of the following rela-
3. Residual settlement (Sp) of pile head. Examples tionship developed by Randolph and Wroth (1978).
include Davisson (1972), who suggests that the ulti-
mate capacity is the load at which the pile head settle- Es = (τ/ws)d(1+υ)ln(2rm/d) (14)
ment = 0.15 + 0.1d (inches), where d = pile dia-
meter, in inches, and DIN4026 (Germany) in which where τ = local shear stress
the residual settlement upon unloading from the ulti- ws = local settlement
mate load is 2.5% of the diameter. d = pile diameter
4. Creep rate of head settlement, where the ultimate υ = ground Poisson’s ratio
capacity is taken as the load at which a sudden rm = radius at which displacements become
increase in the slope of the settlement-time curve very small
occurs τ/ws = the slope of the derived load transfer (t-
5. Coordinate transformation of the load-settlement z) curve.
curve, with the procedure of Chin (1970) being typi- Randolph and Wroth (1978) give an expression for rm
cal. This involves plotting the ratio of settlement to and indicate that it is in the order of the length of the pile.
load as a function of settlement, and defining the
ultimate capacity from the slope of the straight line 10. Example 1 - The Emirates Towers,
portion of this plot. Dubai
6. Employing a specified shape of load-settlement cur-
ve, such as that employed by Hirany and Kulhawy 10.1. Ground investigation and site characterization
(1989). The detailed investigation involved the drilling of 23
Hwang et al. (2003) concluded that the approach attri- boreholes, to a maximum depth of about 80 m. The deepest
buted to Terzaghi (1943) was preferable to many of the boreholes were located below the tower footprints, while
other approaches. boreholes below the low-rise areas tended to be consi-
The emergence of the bi-directional cell test has derably shallower. SPTs were carried out at nominal 1 m
facilitated the interpretation of the ultimate load capacity, depths in the upper 6 m of each borehole, and then at 1.5
since a well-designed test will permit full (or almost full) m intervals until an SPT value of 60 was achieved. SPT
mobilization of both the shaft and base resistances. values generally ranged between 5 and 20 in the upper 4
m, increasing to 60 at depths of 8 to 10 m. Rotary coring
9.3.2. Ground modulus values was carried out thereafter. Core recoveries were typically
Interpretation of the pile load test to assess the pile and 60-100% and rock quality designation (RQD) values were
ground stiffness characteristics requires that account be also between about 60 and 100%.
taken of the site stratigraphy. For the model of ground be- It was found that the stratigraphy was relatively uniform
haviour assumed in the pile analysis, the relevant ground across the whole site, so that it was considered adequate
parameters need to be interpreted from the measured to characterize the site with a single geotechnical model.
load-settlement behaviour. For example, if a load transfer The ground surface was typically at a level of +1 to +3
(t-z) approach is adopted, the initial slope and subsequent m DMD, while the groundwater level was relatively close
shape of the load transfer curves must be assumed and to the surface, typically between 0 m DMD and -0.6 m
 Geotechnical Parameter Assessment for Tall Building Foundation Design 235

DMD (DMD = Dubai Municipality datum).

10.2. Parameter assessment and geotechnical model

10.2.1. In-situ and laboratory testing
Because of the relatively good ground conditions near
the surface, it was assessed that a piled raft foundation
system would be appropriate for each of the towers. The
design of such a foundation system requires information
on both the strength and stiffness of the ground. As a con-
sequence, a comprehensive series of in-situ tests was
carried out. In addition to SPTs and permeability tests,
pressuremeter tests, vertical seismic shear wave testing,
and site uniformity borehole seismic testing were carried
out.
Conventional laboratory testing was undertaken, consi-
sting of conventional testing, including classification tests,
chemical tests, unconfined compressive tests, point load
index tests, drained direct shear tests, and oedometer con-
solidation tests. In addition, a considerable amount of more Figure 4. Summary of Young’s modulus values.
advanced laboratory testing was undertaken, including
stress path triaxial tests for settlement analysis of the
deeper layers, constant normal stiffness direct shear tests the potential to significantly reduce or degrade the
for pile skin friction under both static and cyclic loading, skin friction after initial static failure, and that a cy-
resonant column testing for small-strain shear modulus clic stress of 50% of the initial static resistance could
and damping of the foundation materials, and undrained cause failure during cyclic loading, resulting in a
static and cyclic triaxial shear tests to assess the possible very low post-cyclic residual strength.
influence of cyclic loading on strength, and to investigate Fig. 4 summarizes the values of Young’s modulus ob-
the variation of soil stiffness and damping with axial tained from the following tests:
strain. • seismic data (reduced by a factor of 0.2 to account
for a strain level appropriate to the overall behaviour
10.2.2. Test results of the pile foundation)
From the viewpoint of the foundation design, some of • resonant column tests (at a strain level of 0.1%)
the relevant findings from the in-situ and laboratory test- • laboratory stress path tests, designed to simulate the
ing were as follows: initial and incremental stress state along and below
• The site uniformity borehole seismic testing did not the foundation system
reveal any significant variations in seismic velocity, • unconfined compression tests (at 50% of ultimate
thus indicating that it was unlikely that major frac- stress).
turing or voids were present in the areas tested. Fig. 5 shows the ultimate static shear resistance derived
• The cemented materials were generally very weak to
weak, with UCS values ranging between about 0.2
MPa and 4 MPa, with most values lying within the
range of 0.5 to 1.5 MPa.
• The average angle of internal friction of the near-
surface soils was about 31 degrees.
• The oedometer data for compressibility were consi-
dered to be unreliable because of the compressibility
of the apparatus being of a similar order to that of
some of the samples.
• The cyclic triaxial tests indicated that the Unit 4 sand
deposit had the potential to generate significant ex-
cess pore pressures under cyclic loading, and to ac-
cumulate permanent deformations under repeated one-
way loading. It could therefore be susceptible to
earthquake-induced settlements.
• The constant normal stiffness (CNS) shear tests (Lam
and Johnston, 1982) indicated that cyclic loading had Figure 5. Ultimate skin friction values from CNS tests.
236 Harry G. Poulos and Frances Badelow | International Journal of High-Rise Buildings

oratory stress path tests, which should have reflected rea-

listic stress and strain levels within the various units. The
values for the upper two units were obtained from cor-
relations with SPT data.
The bearing capacity of the various layers for shallow
foundation loading, pu, was estimated from bearing capa-
city theory for the inferred friction angles, the tangent of
which was reduced by a factor of 2/3 to allow for the
effects of soil compressibility, as suggested by Poulos and
Chua (1985).

10.3. Foundation analyses and performance

The behaviour of the test piles at the site was predicted
on the basis of the assessed parameters, and found to agree
reasonably well with the measured load-settlement beha-
viour. However, the analyses for the piled raft foundation
systems were found to over-predict the settlements consi-
derably, and this over-prediction was found to be attribu-
table to an over-estimation of the pile interaction factors.
The analysis of the test piles and the foundation systems
for the Emirates Towers is described in detail by Poulos
Figure 6. Geotechnical model adopted for design. and Davids (2005).

11. Example 2 - Tower in Jeddah, KSA

from the CNS test data, as a function of depth below the
surface. With the exception of one sample, all tests showed 11.1. Introduction
a maximum shear resistance of at least 500 kPa. The mea- For a high-rise project in Jeddah Saudi Arabia, involving
sured values from the CNS tests were within and beyond a tower over 390 m high, potentially karstic conditions
the range of design values of static skin friction of piles were identified in some parts of the site. A piled raft
in cemented soils suggested tentatively by Poulos (1988) foundation system was developed for this tower, as it was
of between 100 and 500 kPa, depending on the degree of considered that such a system would allow the raft to
cementation. redistribute load to other piles in the group if cavities
caused a reduction in capacity or stiffness for some piles
10.2.3. Geotechnical model within the group. The means by which the geotechnical
The key design parameters for the foundation system model and parameters were assessed is described below.
were the ultimate skin friction of the piles, the ultimate
end bearing resistance of the piles, the ultimate bearing 11.2. Geotechnical model
capacity of the raft, and the Young’s modulus of the soils The quantitative data from which engineering properties
for both the raft and the pile behaviour under static loa- could be estimated was relatively limited, and included
ding. For the assessment of dynamic response under wind the following:
and seismic loading conditions, Young’s modulus values • Unconfined compression tests (UCS)
for rapid loading conditions were also required, together • Shear wave velocity data
with internal damping values for the various strata. • Pressuremeter testing
The geotechnical model for foundation design under sta- • SPT data in the weaker strata.
tic loading conditions was based on the relevant available Use was made of these data to assess the following engi-
in-situ and laboratory test data, and is shown in Fig. 6. neering properties which were required for the settlement
The ultimate skin friction values were based largely on analysis, primarily, the Young’s modulus of the ground
the CNS data, while the ultimate end bearing values for deposits (long-term drained values), the ultimate distribu-
the piles were assessed on the basis of correlations with tion of pile shaft friction with depth and the ultimate pile
UCS data (Reese and O’Neill, 1988) and also previous end bearing capacity. The values adopted for the analyses
experience with similar cemented carbonate deposits are summarized in Table 7, and the procedures adopted to
(Poulos, 1988). The values of Young’s modulus were deri- assess each of these parameters are described briefly
ved from the data summarized in Fig. 3. While inevitable below.
scatter exists among the different values, there is a reason-
ably consistent general pattern of variation of modulus 11.2.1. Long-term young’s modulus
with depth. Considerable emphasis was placed on the lab- The assessment of this parameter is critical as it greatly
 Geotechnical Parameter Assessment for Tall Building Foundation Design 237

Table 7. Soil properties used for tower analysis

Depth at bottom of geo-unit (m) Description of Geo-Unit Ev (MPa) fs (MPa) fb (MPa)
20 Coralline Limestone (1) 450 0.2 2
50 Coralline Limestone (2) 600 0.2 9.8
70 Coralline Limestone (3) 1200 0.35 9.8
100 Coralline Limestone (4) 3000 0.4 9.8
*Ev = modulus of soil for vertical pile response; fs = ultimate pile shaft skin friction; fb = ultimate pile base load.

Figure 7. Young’s modulus values derived from various sources.

influences the predicted settlement. Three different me- prediction.

thods of assessment have been used: This is an example of a case where there was no detailed
• Modulus values from the pressuremeter (PMT) tests laboratory testing, so that the parameters derived from
• Values correlated to UCS via the correlation Es = empirical correlations were adjusted on the basis of the
100UCS, where Es is long-term Young’s modulus pile load tests.
• Values derived from the small-strain Young’s modu-
lus values obtained from shear wave velocity meas- 11.2.2. Ultimate pile shaft friction and end bearing
urements, but scaled by a factor of 0.2 to allow for Use was made of correlations between the ultimate shaft
the effects of practical strain levels. friction, fs, and end bearing, fb, with UCS. For the reefoi-
Fig. 7 compares the values obtained from each of these dal coral deposits, the following conservative relationship
three approaches. was used for our assessment.
On the basis of these data, the following assumptions
were originally made: fs = 0.1(UCS)0.5 MPa (15)
• From the surface to a depth of 20 m, an average long-
term Young’s modulus (for vertical loading), Es, is where UCS = unconfined compressive strength (MPa)
150 MPa The average ultimate shaft friction for the upper 50 m
• From 20 m to 50 m, Es = 200 MPa was thus taken to be 0.2 MPa (200 kPa). The subsequent
• From 50 m to 70 m, Es = 400 MPa pile load test revealed that this was a conservative esti-
• Below 70 m, Es = 1000 MPa, which reflects the grea- mate of shaft friction, as values of about 500 kPa were
ter stiffness expected because of the smaller levels of mobilized along some portions of the test pile, with an
strain within the ground at greater depths. average value of about 310 kPa.
Subsequent to these initial assessments, a pile load test The following correlation suggested by Zhang & Ein-
was undertaken using the Osterberg Cell technique. The stein (1998) was employed for base ultimate capacity:
pile head stiffness derived from this test was considerably
larger than that implied by the initially-selected values of fb = 4.8 (UCS)0.5 MPa (16)
Young’s modulus. Accordingly, the initially-selected values
were multiplied by a factor of 3 for the final settlement On this basis, for an average UCS of 4 MPa, fb was 9.6
238 Harry G. Poulos and Frances Badelow | International Journal of High-Rise Buildings

MPa. This value assumes that there are no cavities in the Such parameters can be used for preliminary design
area of influence of the base of a pile. and checks on parameters derived by other means.
2. Detailed assessment via in-situ and laboratory tests.
11.2.3. Analysis of foundation system These will usually form the basis for detailed found-
The foundation analysis has been described by Poulos ation design.
et al. (2013). The initial foundation analyses assumed that 3. Parameters derived from pile load testing. These can
no significant cavities exist below the pile toes. If cavities be used to confirm the design assumptions and the
were to be found during construction, then it would be adequacy of the construction techniques, and can also
necessary to re-assess the performance of the foundation provide information for adjusting the foundation de-
system and make provision for grouting of the cavities if sign before, or during, the construction process.
this was deemed to be necessary. Thus, subsequent to the Two examples have been provided of how the combin-
foundation design, a further series of analyses was under- ation of the three stages of parameter assessment can be
taken to investigate the possible effects of cavities on the used to characterise the ground conditions and provide
settlements and also on the raft bending moments and pile the necessary information for the foundation design.
loads. For these analyses, the commercially-available finite
element program PLAXIS 3D was used. References
Firstly, the effect of a single cavity at different locations
along the centre line of the raft at different depths was Bermingham, P. and Janes, M. (1989). “An Innovative App-
examined. The cavity was introduced into the finite ele- roach to Load Testing of High Capacity Piles”. Proc. Int.
ment mesh at various depths. Generally the locations of Conf. Piling & Deep Founds, 1, pp. 409~413.
cavities beneath the foundation are not known, and only Bermingham, P., Ealy, C. D., and White, J. K. (1994). “A
cavities found in specific boreholes can be precisely loc- Comparison of Statnamic and Static Field Tests at Seven
FHWA Sites”. Proc. Int. Conf. Des. and Constrn. of Deep
ated. The effect of cavities at random locations and of
Founds., FHWA, Orlando, 2, pp. 616~630.
random sizes was therefore gauged. To this end, a random
Briaud, J-.L. (1992). “The Pressuremeter”. Balkema, Rotter-
number generator was used to select the location and size dam.
of the cavity. The centre of the cavity was constrained to Bustamante, M. and Gianeselli, L. (1982). “Pile bearing cap-
lie within the footprint of the raft, and the depth was con- acity prediction by means of static penetrometer CPT”.
strained to be within a 70 m depth. The number of ran- Proc. ESOPT II, Amsterdam, 2, pp. 492~500
domly placed cavities was limited to 5 for each of the Chen, Y.-J. and Fang, Y.-C. (2009). “Critical Evaluation of
cases analysed. Compression Interpretation Criteria for Drilled Shafts”. J.
It was found that the consequences of cavities, while Geot. Geoenv. Eng., ASCE, 135, pp. 1056~1069.
not insignificant, may not be as serious as might be feared, Chin, F. K. (1970). “Estimation of the Ultimate Load of Pile
because of the inherent redundancy of the piled raft foun- from Tests not Carried to Failure”. Proc. 2nd S.E. Asian
dation system. The maximum settlement increased from Conf. Soil Mech. Found. Eng., Singapore, 81-92.
about 56 mm to 74 mm for the range of cases examined. Davisson, M. T. (1972). “High capacity piles”. Proc. Lecture
Series on Innovations in Foundation Construction. ASCE
The maximum bending moment in the raft was increased
Illinois Section, Chicago.
by only about 13%.
Decourt, L. (1982). “Prediction of the Bearing Capacity of
While the analyses undertaken were insufficient to en- Piles Based Exclusively on N values of the SPT”. Proc.
able a quantitative assessment of risk to be assessed, they ESOPT II, Amsterdam, 1, pp. 29~34.
did give a good appreciation of the sensitivity of the com- Decourt, L. (1995). “Prediction of Load-Settlement Relation-
puted foundation response to the presence of random cav- ships for Foundations on the Basis of the SPT-T”. Ciclo
ities. Clearly, using redundant foundation systems may de Conferencias Intern. “Leonardo Zeevaert”, UNAM,
not only reduce the risks associated with building towers Mexico, 85-104.
on karstic limestone but also provide a much more eco- De Ruiter, J. and Beringen, F. L. (1979). “Pile foundations for
nomical foundation than using deep foundation piles in large North Sea structures”. Marine Geotechs. 3(3), pp.
an attempt to carry foundation loads through the karstic 267~314.
zones. Frank, R. and Magnan, J.-P. (1995). “Cone Penetration Test-
ing in France: National Report”. Proc. CPT '95, Linkoping,
Swedish Geot. Society, 3, pp. 147~156.
12. Conclusions Fujioka, T. and Yamada, K. (1994). “The Development of a
New Pile Load Testing System”. Proc. Int. Conf. Des.
This paper has set out a process for characterising the and Constrn. of Deep Founds., FHWA, Orlando, 2: 670-
ground conditions and quantifying the relevant geotech- 684.
nical parameters required for tall building foundation Fuller, F. M. and Hoy, H. E. (1971). “Pile Load Tests Includ-
design. Three stages of parameter assessment have been ing Quick Load Test Method, Conventional Methods and
described: Interpretations”. High. Res. Rec. 333, pp. 74~86.
1. Preliminary assessment via empirical correlations. Goble, G. G. (1994). “Pile Driving - An International State-
 Geotechnical Parameter Assessment for Tall Building Foundation Design 239

of-the-Art”. Proc. Int. Conf. Des. and Constrn. of Deep Balkema, Rotterdam, 581-588.
Founds., FHWA, Orlando, 1, pp. 1~26. Ooi, L. H. and Carter, J. P. (1987). “A Constant Normal Stiff-
Haberfield, C. M. (2013). “Tall Tower Foundations - from ness Direct Shear Device for Static and Cyclic Loading”.
concept to construction”. Advances in Foundation Engin- Geot. Testing Jnl., ASTM, 10, pp. 3~12.
eering, Ed. K. K. Phoon et al., Research Publishing Ser- Osterberg, J. (1989). “New Device for Load Testing Driven
vices, 33-65. and Drilled Shafts Separates Friction and End Bearing”.
Hasancebi, N. and Ulusay, R. (2007). “Empirical Correlations Proc. Int. Conf. Piling and Deep Founds., London, 421-
Between Shear Wave Velocity and Penetration Resistance 427.
for Ground Shaking Assessments,” Bull. Eng. Geol. En- Poulos, H. G. (1988). “The Mechanics of Calcareous Sedim-
viron. 66, pp. 203~213. ents”. John Jaeger Memorial Lecture, Aust. Geomechanics,
Hirany, A. and Kulhawy, F. H. (1989). “Interpretation of Load Spec. Edition, 8-41.
Tests on Drilled shafts I: Axial Compression”. Foundation Poulos, H. G. (1989). “Pile Behaviour-Theory and Applica-
Engineering: Current Principles and Practices, GSP 22, tion”. Geotechnique, 39(3), pp. 365~415.
Ed. F. H. Kulhawy, ASCE, New York: 1132-1149. Poulos, H. G. and Chua, E. W. (1985). “Bearing Capacity of
Hoek, E. and Diederichs, M. S. (2006). “Empirical estimation Foundations oin Calcareous Sand”. Proc. 11th Int. Conf.
of rock mass modulus”. International Journal of Rock Soil Mechs. Found. Eng., San Francisco, 3, pp. 1619~
Mechanics and Mining Sciences, 43(2006), pp. 203~215, 1622.
Elsevier. Poulos, H. G., Carter, J. P., and Small, J. C. (2001). “Founda-
Hwang J. H., Li J. C. C., and Liang N. (2003). “On Methods tions and Retaining Structures-Research and Practice”.
for Interpreting Bearing Capacity from a Pile Load Test”. Proc. 15th Int. Conf. Soil Mechs. Geot. Eng., Istanbul, 4,
Geotechnical Engineering, 34(1), pp. 27~39. pp. 2527~2606.
Lam, T. S. K. and Johnston, I. W. (1982). “A Constant Nor- Poulos, H. G., Small, J. C., and Chow, H. S. W. (2013). “Foun-
mal Stiffness Direct Shear Machine”. Proc. 7th S.E. Asian dation Design for High-Rise Towers in Karstic Ground”.
Conf. on Soil Eng., Hong Kong, 805-820. ASCE GSP229, Foundation Engineering in the Face of
Mair, R. J. and Wood, D. M. (1987). “Pressuremeter Testing: Uncertainty, Ed. J. L. Withiam, K.-K. Phoon and M. H.
methods and interpretation”. CIRIA, UK. Hussein, 720-731.
Mayne, P. W., Coop, M. R., Springman, S. M., Huang, A.- Poulos, H. G. and Davids, A. J. (2005). “Foundation Design
B., and Zornberg, J. G. (2009). “Geomaterial Behavior for the Emirates Twin Towers, Dubai”. Can. Geot. Jnl. 42,
and Testing”. Proc. 17th Int. Conf. Soil Mechs. Geot. Eng., pp. 716~730.
Alexandria, Egypt, 4, pp. 2777~2872. Randolph, M. F. and Wroth, C. P. (1978). “Analysis of Defor-
Mayne, P. W. and Rix, G. J. (1993). “Gmax - qc relationships mation of Vertically Loaded Piles”. Jnl. Geot. Eng., ASCE,
for clays”. Geot. Testing Jnl., 16(1), pp. 54~60. 104(GT12), pp. 1465~1488.
MELT (1993). “Regles techniques de conception et de calcul Rausche, F., Goble, G. G., and Likins, G. (1985). “Dynamic
des fondations des ouvrages de genie civil”. CCTG, Fasci- Determination of Pile Capacity”. Jnl. Geot. Eng., ASCE,
cule No. 62, Titre V, Min. de L'Equipement du Lodgement 111(3), pp. 367~383.
et des Transport, Paris. Reese, L. C. and O’Neill, M. W. (1988). “Drilled Shafts:
Merifield, R. S., Lyamin, A. V., and Sloan, S. W. (2006). “Li- Construction Procedures and Design Methods”. Pub. No.
mit analysis solutions for the bearing capacity of rock FHWA-H1-88-042, US Dept. Transportation.
masses using the generalised Hoek-Brown criterion.” Int. Terzaghi, K. (1943). “Theoretical Soil Mechanics”. John
J. Rock Mechs. Min. Sciences, 43, pp. 920~937. Wiley, New York.
Middendorp, P., Bermingham, P., and Kuiper, B. (1992). Zhang, L. and Einstein, H. (1998). “End bearing capacity of
“Statnamic Load Testing of Foundation Piles”. Proc. 4th drilled shafts in rock”. Jnl. Geot. Eng., ASCE, 124(7), pp.
Int. Conf. App. of Stress Wave Theory to Piles, the Hague, 574~584.