You are on page 1of 16

Proceedings of the Institution of

Civil Engineers
Geotechnical Engineering 160
April 2007 Issue GE2
Pages 105120

Paper 14428
Received 16/09/2005
Accepted 19/12/2005
Keywords: Luay Alrifai
foundations/geotechnical Principal Geotechnical
engineering/piles & piling Engineer, Hyder Consulting
(UK) Limited, Birmingham,
UK

Rock socket piles at Mall of the Emirates, Dubai


L. Alrifai, MSc, PhD, CEng, MICE

The first snow centre in the Middle East, Ski Dubai, was m number of rows perpendicular to lateral force direction
constructed as part of the Mall of the Emirates retail and n number of rows in lateral force direction
leisure complex in Dubai, United Arab Emirates. rock socket reduction factor
Opened in December 2005, the project comprises an 9 modified factor for E0
80 m high indoor ski slope and associated structures. The  rock socket correction factor
higher end of the slope spans over a three-level car park.  allowable displacement of pile in lateral direction
The site is underlain by a variable sequence of very
loose/loose silty sands, which in turn overlie solid geology 1. INTRODUCTION
comprising very weak/weak carbonate sandstone. It has Hyder Consulting Limited has been appointed as the Engineer
been considered that the most appropriate foundation of Record for the Snow Centre scheme, which is part of the
option to support the structure safely is a piled Mall of the Emirates in Dubai. The site is approximately 10 km
foundation. Bored and cast-in-place piles socketed into south-west of the centre of Dubai, close to interchange 4 along
the rock were proposed. Detailed analyses of single piles the Sheikh Zayed Road.
and pile groups under axial and lateral loading conditions
were undertaken. The seismicity and liquefaction A full foundation engineering service for the Snow Centre
potential were considered in the design in accordance scheme was provided. The geotechnical input included
with the requirements of Dubai Municipality. The planning and directing the ground investigation, detailed
preliminary test piles, which were carried out prior to foundation analysis, design verification and construction
the installation of the working piles, checked the pile supervision.
capacity and loadsettlement behaviour, confirmed the
effectiveness of the piling technique used, and provided Evaluation of foundation options was undertaken. Spread
greater assurance of the satisfactory performance of the foundations combined with ground improvement were
foundations. The results of the preliminary pile-testing considered. However, based on the proposed loading and
programme were compared with those obtained from prevailing ground conditions, it was concluded that rock socket
theoretical predictions based on the empirical piles were the most suitable foundation solution.
relationships between pile capacity and the unconfined
compressive strength. On the basis of this review, a The preliminary pile design was based on the design approach
guide for the design of rock socket piles in weak for bored piles in rock, given by Tomlinson. 1 With little
carbonate rocks suitable for use in Dubai is proposed. published information on pile performance in Dubai, it was
decided to carry out a preliminary pile testing ahead of the
NOTATION installation of working piles to verify the design. Based on the
D pile diameter observed pile behaviour, the design was re-evaluated.
DE reduction factor for SPT N value in seismic design
d ratio of spacing of piles to diameter The behaviour of pile groups was examined using a computer
Em deformation modulus of rock mass program based on a simplified continuum analysis using
Ep Youngs modulus of pile interaction factors. The deformation modulus adopted in the
E0 Youngs modulus of soil settlement analysis was back-analysed from the observed data
eg adjusting factor for kh to allow for pile group effect of the preliminary test piles.
FL liquefaction resistance factor
Gmax shear modulus at very small strain Comparisons between the predicted and observed pile
H lateral load capacity of pile behaviour have resulted in significant conclusions, which will
I moment of inertia of pile section contribute to pile design practice in Dubai.
k restriction factor of pile head
kh coefficient of lateral subgrade reaction 2. REGIONAL GEOLOGY
khg coefficient of lateral subgrade reaction of a group pile Figure 1 shows the geology of the United Arab Emirates and
L spacing of piles the Arabian Gulf, given in a KNGMG publication,2 which was

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 105
Gulf of Oman

Oman Musandam
Dibba
Iran Ras al Khor Fakkan
Khaimah Fujeirah
WD
N Umm al Quwain WJ
Oman Mountains
Ajman
Sharjah
Dubai
Al Ain
s
Jebel Ali

WD 5 Wadi Dhaid
WJ 5 Wadi Juweiza
Key
Aeolian sands, Quaternary
Abu Dhabi Gravel, sand, clay, Quaternary
Sabkha deposits, Quaternary
Aeolian sands with sabkha in depressions, Quaternary
Arabian Gulf
Gravel, alluvial dep. Quaternary, Tertiary
Tarif Aeolian sand with Pliocene outcrops in depressions
nk

s s
Ba

a Calcarenites, calcisiltites, marls and shales, Miocene, Pliocene


Liw
l
ar
Pe

Limestone, Cretaceous and Triassic


s s Metamorphic and igneous rock. Cretaceous to Permian
s s s Salt Dome Complex
s Ruweis
s

Saudi Arabia
Qatar

0 100 km

Fig. 1. Geology of the Arabian Gulf

reproduced from USGS & ARAMCO, 1963. The Arabian Gulf is northern extent of the Snow Centre, southwards to the
in an area of extensive carbonate sedimentation, and the beginning of mid station
nature and distribution of the sediments are governed by the (b) Section 2: mid station
recent geological history and structural setting of the Gulf, the (c) Section 3: upper end of the slope, which spans over the
orientation of the coastline, and the prevailing winds. 3 three-level West Car Park.
Conditions in the Dubai area consist essentially of a linear
coastline dissected by channels or creeks. Superficial deposits The eaves level at the northern end of Section 1 is
consist of beach and dune sands with development of sabkha approximately 23 m above ground. The elevations of the top of
deposits in certain areas, notably around the creeks. The Section 2 and Section 3 are 60 m and 80 m respectively above
superficial deposits are underlain by aeolian carbonate ground level.
sandstone. Close to the rockhead the sandstone often contains
a thin band of moderately to well cemented sand. The rockhead A site plan is presented in Fig. 2, and a sketch of the Snow
level varies appreciably over the area, being exposed at ground Centre model view (facing south) is presented in Fig. 3.
level in some localities in Dubai and occurring at depths of up
to 10 or 12 m elsewhere. 3 4. GROUND INVESTIGATION
A ground investigation was undertaken in January 2003 to
support the design and construction programme. At the time
3. PROPOSED DEVELOPMENT the investigation was planned, the design of the superstructure
The Snow Centre structure comprises an indoor ski slope was not fully developed and the ski model was not selected.
located above the retail areas, and a multi-storey car park. Therefore the ground investigation programme was designed to
Steel lattice trusses and columns provide clear spans, with provide an overall assessment of ground conditions and cover
reinforced concrete used to support the ski plate slab. The ski the main requirements of foundation design.
slope has an average gradient of 158 but is subdivided into
sections to provide interest and challenges for skiers. Fourteen boreholes up to 35 m deep and eight trial pits were
undertaken. The boreholes were advanced through soil deposits
The Snow Centre is divided into three sections using standard light cable percussion techniques. Rotary coring
techniques were used in the underlying rock strata. Standard
(a) Section 1: lower end of the slope, which extends from the penetration tests (SPTs) were carried out at 1.0 m intervals

106 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai
Hypermarket

Cinema

Hotel
Mall/retail
Mall/retail
Mall/retail

Section 1 Theatre

East Car Park

Section 3
Section 2

Key
Snow Centre

Fig. 2. Mall of the Emirates: site plan

Fig. 3. Sketch of the Snow Centre: model view

within the soil deposits. Due to the difficulty in obtaining full specification, and that the assumptions made when the
recovery in the weak rock, SPTs were specified between core scope of the investigation was established were still
runs in the underlying rock. Standpipes were installed in appropriate.
boreholes to monitor groundwater levels during and after the
fieldwork. A programme of laboratory testing was performed on samples
of soil, rock and groundwater recovered during the fieldwork.
The fieldwork was undertaken in accordance with BS 5930. 4 Test results were used to assist with sample description and
The ground investigation fieldwork was monitored on site to determination of engineering material properties. Chemical test
ensure that the work was carried out in accordance with the results were used to determine the corrosion potential of soil

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 107
and groundwater. The laboratory testing was carried out in Monitoring of the standpipes indicated that the groundwater
accordance with BS 1377. 5 level varied between 3.8 m and 4.5 m below ground level.
However, as dewatering was being undertaken in the area
The exploratory hole positions are shown in Fig. 4. adjacent to the site, the recorded water levels were not considered
to be representative of the natural water table. Because the
5. GROUND CONDITIONS groundwater was reported as 2 m below ground level in a
The stratigraphy beneath the site is summarised in Table 1. previous geotechnical survey, it was considered prudent to take
this level as the groundwater level for design purposes.
A geological long section through the eastern perimeter of the
site is shown in Fig. 5. The results obtained from the ground investigation have shown

BH10355

TP2024
BH20120 BH20320

TP2014 Zone 1 Zone Rockhead elevation


1 220 m DMD (5 mbgl)
2 247 m DMD (8 mbgl)
3 227 m DMD (6 mbgl)
BH1115 4 267 m DMD (10 mbgl)
TP2034
T1

TP2044 BH20525
C1

TP2064 TP2054 BH21320 Snow Centre pylon


C4
Zone 2 BH20625 BH20735
BH21120
BH20420 Zone 4
TP2074
C3
Key BH21220 BH21420
Preliminary test pile BH1720
Borehole Zone 3
TP2084 C2
Trial pit
BH20835
BH21025 Snow Centre pylon

BH20925

Fig. 4. Site zones

Stratum Brief description Depth of top of stratum: Thickness of stratum:


mbgl m

Made ground Sand fill 0 1.01.3


Dune/beach SAND Very loose to loose locally medium dense, light grey/light 1.01.3 1.43.0
brown, shelly, silty to very silty, fine to medium SAND
with silt bands
Transition zone Medium dense to very dense, light grey, silty SAND 2.54.0 1.24.5
between soil and containing some to frequent gravel-size fragments of very
rock weak sandstone
SANDSTONE Slightly to moderately weathered, light brown/brown, very 4.010 12.217.8
weak to weak SANDSTONE, with close to medium
fractures. Close to rockhead bands of moderately to well
cemented SAND and, at greater depth bands of weak
CONGLOMERATE
CALCISILTITE Slightly to moderately weathered, very pale orange/off- 19.130.5 Proved to 16 m
white, weak conglomeratic CALCISILTITE, containing
bands of weak conglomeratic SANDSTONE and
CONGLOMERATE

Table 1. Stratigraphy

108 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai
BH103 BH203 BH205 BH206 BH207
50
0
250
2100

Elevation: mDMD
2150
2200
2250
2300
2350
2400
2450
2500
2550
Key Solid
Made ground Sandstone Approximate ground level

Geological boundary
Superficial deposits Conglomerate
Groundwater level determined from
Very loose/loose silty sand Sequence of sandstone, standpipe monitoring in January/February 2003
conglomerate and calcisiltite
Borehole
Transition zone between
soil and rock Inferred rockhead profile
(SPT N > 50)

Fig. 5. Geological section

that the change from soil to rock is a transitional condition. would be classed as having significant contamination with
Therefore interpretation of the rockhead level was required. The both sulphate and chloride salts. Carbonate (as CaCO3 ) tests
criterion used to establish the interface between soil and rock undertaken on rock samples have shown that carbonate
was taken as SPT N value > 50. content in the sandstone and conglomerate varies between 25%
and 80%, with most of the results being . 50%.
For design purposes, the profile determined using this criterion
was taken as the rockhead. Fig. 4 shows that the site is divided 6. FOUNDATIONS OPTIONS
into four zones in relation to the rockhead depth.
6.1. Foundation loading
The unconfined compressive strength profile is shown in Fig. 6. Loading is applied on the ground through columns. The
The design line shown in the figure has been adopted. column loading at foundation level is shown in Table 2. The
loading is made up of dead loads and superimposed loads. The
Based on the chemical test results it is considered that the site latter result mainly from live, snow, wind and seismic loading.

Unconfined compressive strength, UCS: MN/m2


01 1 5 10 100
0
Very weak Weak Mod. weak Mod. strong Strong

2500

21000
Elevation: mDMD

21500
Design line

22000

22500

23000

Fig. 6. Unconfined compressive strength against elevation

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 109
Dome section Loading range: kN Vibroflotation involves the
installation of compacted
Axial Lateral granular columns in granular
soil by means of radial
Section 1: lower end 10005000 10400 compaction. In granular soils
Section 2: mid station 500012 000 100500 containing significant fines
Section 3: higher end 18 00056 000 3003000
or cohesive layers radial
compaction has little effect,
Table 2. Foundation loading
so the granular columns act
as vertical reinforcement. The
use of this option to support
6.2. Spread foundations the foundations in the lower end of the structure raised the
Spread foundations were considered for the lower end of the following concerns.
structure, where the proposed loads are the lowest. The upper
stratum is very loose/loose sand. In granular deposits, the (a) Because of the need to limit the bearing pressures below
tolerable settlement is generally the predominant factor 200 kN/m2 , large footing sizes may be required for the
governing allowable bearing pressures. Spread foundations upper end of the loading range. The use of the stone
founded within the upper sand will be subject to significant column foundation to support loading in excess of 200
total and differential settlements. Therefore this option was kN/m2 is considered to be a risk. If the bearing pressure
discounted. exceeds 200 kN/m2 , the column response departs from the
elastic behaviour. As column yield occurs, the effectiveness
Replacement of the upper stratum was not considered to be of the column diminishes with increasing pressure. 6
cost-effective as it requires excavation of up to 5 m depth, (b) Vibro replacement techniques suit saturated soils.
support to the sides of the excavation, material disposal, and Experience has shown that it may not be possible to
placement of large amounts of imported material. Furthermore, construct stone columns in dry sand, and it will be
the possibility of dewatering the excavation could not be ruled necessary to excavate up to the depth of the groundwater
out. This is due to the uncertainty of groundwater levels arising and start operations at this level. As the groundwater levels
from dewatering being undertaken in the vicinity of the site. at the time of construction cannot be predicted, and
depend on the dewatering conditions within the adjacent
6.3. Ground improvement areas, the feasibility of this option could not be confirmed.
The concept of ground improvement involves the application
of a geotechnical process to improve the strength and/or
7. PILED FOUNDATIONS
compressibility of the soil so that the stability and settlement
criteria under structural loading are satisfied. This option was
7.1. Introduction
considered for improving the upper very loose/loose sand
Piled foundations were considered to be the most appropriate
stratum for the lower end of the structure (Section 1) only.
solution. Rock socket piles were proposed. The piles were
Improving the ground to support the high column loads of the
designed to support axial and lateral loading, and limit ground
mid station and higher end are outside the capability of ground
movements. They act primarily as friction piles, as they derive
improvement. Two methods of ground improvement were
most of their carrying capacity from the skin friction between
considered: pre-compression and vibroflotation/vibro
the embedded surface of the pile and the surrounding rock.
replacement stone columns. The use of dynamic compaction,
which involves dropping a heavy weight onto the surface of
A single pile foundation beneath columns was used within the
the ground to compact the soil to great depth, was discounted
lower end of the structure.
because of the effect of the vibrations produced on the
adjacent areas under construction.
In the mid station and upper end, where loading is high, piles
were placed in groups linked together with a pile cap. The
Pre-compression involves compressing the soil under
proposed underside of the pile cap is at approximately 2 m
temporary applied load such as earth fill prior to placing or
below existing ground level.
completing the permanent structure in order to develop
settlements prior to construction. This method is most
attractive in granular soil as the settlement occurs rapidly so 7.2. Design criteria
that the time required for pre-loading is short. The The design criteria for the piled foundation were as follows.
consolidation of the silt bands within the upper sand stratum
may require the placement of a pre-load for some considerable (a) The working loads should meet the specification of BS
time. The process can be speeded up by the installation of 8004,7 which requires that average compressive stress
vertical drains. However, this will significantly increase the under working load should not exceed 25% of the specified
cost of the process. It has therefore been concluded that the concrete cube strength at 28 days.
presence of the silt bands within the sand stratum makes this (b) There must be an adequate factor of safety for the working
option undesirable. loads.
(c) The vertical settlement of a single pile and pile group must
The vibro replacement stone columns technique was developed be within an acceptable limit.
in the 1960s as an extension to the vibroflotation method. (d) The ultimate lateral pile capacity should not be exceeded.

110 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai
(e) The lateral displacements of a single pile and pile group that field tests showed that grouted piles offered superior load-
must be within an acceptable limit. carrying capacity to driven piles in calcareous sediments.
( f ) Seismicity and liquefaction potential should be considered
in design.
(g) Durability of piles in the long term must be ensured. 7.3.2. Piles in carbonate rocks Limited information on the
behaviour of piles in weak carbonate rocks has been published.
7.3. Behaviour of piles in carbonate sands and rocks The main reference found in the literature on pile design in
carbonate rocks is a method proposed for the design of grouted
7.3.1. Piles in carbonate sands The main information in the piles in weak carbonate rocks presented at the 17th Offshore
literature on the behaviour of piles in carbonate sands relates Technology Conference, Texas. 10 The method was developed
to offshore engineering. A review of the occurrence and during a study of weak carbonate rocks prevalent in offshore
composition of carbonate sediments and their behaviour in areas of the Middle East, where offshore platforms are
relation to foundation of offshore structures was presented at supported by grouted steel pipe piles.
the International Symposium on Offshore Engineering in
1983. 8 A summary of the main conclusions is given below. The rocks in question are typically carbonate sandstones and
siltstones with unconfined compressive strengths in the range
Carbonate sediments are abundant in warm marine waters. 0.55.0 MN/m2 . Typical piles utilised are 610914 mm
Many deposits of carbonate sands coincide with oil exploration diameter pipe piles grouted into predrilled holes 150 mm larger
and production areas, such as the Arabian Gulf, North Rankin in diameter than the pipes. The method proposes the
and the Bass Strait in Australia. Particles may cover the full relationship shown in Table 3 between the skin friction
range of sizes, from gravel-sized coral and shell to sand-sized resistance developed from the rock to grout bond and the
foraminifera, shells, fine silt and clay-sized fossils. The unconfined compressive strength (UCS)
proportion of carbonate to total minerals present may vary
from a trace to essentially 100% in any case. When almost all Examination of this proposed relationship indicates that the
particles are carbonates and within the sand-size range, the ultimate skin friction resistance predicted by this method is in
resulting sediment is commonly referred to as a carbonate the same range as that provided by the conventional pile
sand. design methods in non-carbonate rocks.

The review indicates that the presence of carbonate sands has The conclusions of the study indicate that the proposed method
important consequences for design and planning of offshore is based on limited observations and little theoretical basis.
foundations, primarily because piles driven through this type However, back-analyses of existing offshore platforms that
of material have values for both driving resistance and load have been in place for many years have given confidence in
capacity which are low when compared with the range of the approach.
values usually found for the better known silica sands.
However, whereas the results of load capacity tests are
remarkably different, the same does not occur in the laboratory 7.4. Pile ultimate capacity
in the usual strength tests for the two kinds of sand. Laboratory There is a lack of published information on observed pile
test results have shown that the angle of internal friction of behaviour and specific pile design methods in carbonate rocks
carbonate sands is in the same range as that of silica sands. in Dubai. Therefore, in order to predict the pile ultimate
capacity, the commonly used relationship proposed by
The review identifies four properties that seem to be Williams and Pells 11 as given in reference 1 was used. This
responsible for most of the differences between carbonate relates the ultimate rock socket skin friction to the unconfined
sands and the more commonly encountered silica sands. These compressive strength of intact rock. The relationship applies a
are hardness of grains, fabric, intragrain porosity, and skin friction reduction factor , which should be corrected by a
cementation. Grain crushing attributed to grain morphology factor () that depends on the mass factor of the rock
and breakdown of cementation have been cited as the formation. The mass factor is related to the RQD (rock quality
determining factors for the low capacities of piles driven into designation) or fracture index (fracture spacing of the rock
carbonate sands. It was concluded that the low skin friction mass).
capacities of these piles are primarily from volume reduction
due to grain crushing, which is responsible for low lateral The mechanism of load transfer of a socket pile and the
pressure adjacent to the pile, and not from any inherent low distribution of support between the shaft and end bearing have
frictional resistance of carbonate minerals relative to silica been discussed by various authors. 1,12
sands.
UCS: MN/m2 Ultimate unit skin friction resistance: MN/
A paper published in the proceedings of the Australia New m2
Zealand Conference on Geomechanics, 1992, 9 reports that piles
driven into calcareous sediments show low-capacity behaviour ,1.0 0.375 3 UCS
1.03.0 0.375 + [(UCS  1) 3 0.1875]
similar to that observed in carbonate sands. It indicates that
.3.0 0.75
conventional driven steel pipe piles in calcareous sediments
may show very low unit skin friction values, and as a result
Table 3. Skin friction related to unconfined compressive
drilled and grouted insert piles have generally been used for strength (UCS) of weak carbonate rocks
support of offshore platforms in such soils. The paper reports

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 111
A review of the information indicates that the development of where H is the lateral capacity of the pile (kg); kh is the
end bearing resistance depends on the socket geometry, moduli coefficient of lateral subgrade reaction (kg/cm3 ); D is the pile
of the pile and rock, cleaning of the base during pile diameter (cm);  is the allowable displacement of the pile in
installation, and rock conditions along the sides and base of the lateral direction (cm) (2% of pile diameter in sands was
the rock socket. Based on the review it was considered that it is adopted 16 ); E p is the Youngs modulus of the pile (kg/cm2 );
uncertain that any end bearing support will be developed. and I is the moment of inertia of the pile section (cm4 ).
Therefore for design purposes it was assumed that skin friction
resistance along the pile shaft will provide the capacity of the
2 kh 0:29E0 D3=4
pile, with negligible load being transferred to the base.

7.5. Working loads where 9 is a modified factor for E0 (usually 2), and E0 is the
In order to calculate the pile working load, the ultimate Youngs modulus of the surrounding soils (kg/cm2 ).
capacity was divided by a factor of safety of 2.5 in
compression and 3.0 in tension. The ultimate capacity in 3 E0 25N
tension is the same as for compression, excluding the self-
weight of the pile. Table 4 shows the compression working
loads calculated for a range of pile diameters and rock socket where N is the SPT N value.
lengths.
7.7.2. Lateral load capacity of a single pile The lateral load
capacities were determined using the above approach. For
7.6. Settlement at working load
comparison, the ultimate lateral loads were calculated using
The settlement of a single pile at working load was estimated
the method proposed by Broms, 17 which relates the ultimate
using the procedures proposed by Rowe and Armitage. 13 The
lateral load capacity to ultimate resistance moment (yield
deformation modulus (Em ) of the rock mass was determined
moment). The results are presented in Table 5. The lateral
using the relationship between the unconfined compressive
working load was determined by dividing the ultimate capacity
strength and deformation modulus. 1 A value of Em 200
by 2.5.
MN/m2 for the upper rock layers was adopted for estimates of
settlement for a single pile. The maximum settlement under the
7.7.3. Pile group effect In a pile group, the coefficient of
anticipated working loads was estimated to be in the order of
lateral subgrade reaction (kh ) should be adjusted depending on
5 mm.
the arrangement of piles 18 as follows.

7.7. Lateral resistance of pile khg kh 3 eg


4

7.7.1. Design approach The design of piled foundations to


resist lateral loading has been presented by many authors. 1418 where khg is the coefficient of lateral subgrade reaction of a
group pile, kh is the coefficient of lateral subgrade reaction
The performance of a vertical pile when subjected to lateral (equation (2)), and eg is an adjusting factor for kh to allow for
load is controlled mainly by the properties of the soil near the the pile group effect.
surface, for example, the upper 35 m. 14
The adjusting factors of the coefficient of lateral subgrade
For evaluating the lateral resistance of a single pile, the design reaction are a function of the spacing of the piles and the
method outlined below 15 was adopted. configuration of the piles within the group. Furthermore, the
direction in which the lateral forces are applied in relation to
The following equation (Changs), derived from elastic theory, the axes of the group has a bearing on the pile group effect.
was used to calculate the lateral load capacity of the single pile
related to the allowable lateral deflection of a pile shaft with a : :
eg f1  5[1  (0:6  0:25k) 3 d0 300 2 k ]
rigid type of connection.
5 : :
3 (1  m0 22 3 n0 09 )g4=3
kh D
1 H
kh D=4Ep I1=4
where k is the restriction of the pile head (usually k 0.6); m

Pile diameter: Lateral load capacity Lateral load capacity


Pile Pile diameter: Rock socket length mm related to allowable related to ultimate
type mm lateral displacement: resistance moment:
7m 10 m 13 m kN kN

A 600 2350 3700 5100 650 370 450


B 750 2900 4600 6350 750 500 650
C 1000 3900 6350 8700 1000 950 1000

Table 4. Working loads in kN Table 5. Lateral load capacity of single pile

112 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai
is the number of rows perpendicular to the lateral force group can be determined. Poulos 21 has reported four programs
direction; n is the number of rows in the lateral force direction; based on continuum analysis methods as follows.
d is the ratio of spacing of piles to diameter (L/D); L is the
spacing of the piles; and D is the diameter of the piles. (a) DEFPIG: non-linear continuum analysis using interaction
factors
7.8. Seismicity and liquefaction potential (b) GAPFIX: non-linear continuum analysis, complete solution
(c) M-PILE: simplified continuum analysis using interaction
7.8.1. Seismicity The UBC (Uniform Building Code) 1997 19 is factors
the standard used in Dubai to assess seismic zones and ground (d) PGROUP: complete linear continuum analysis.
accelerations. The UBC classifies seismic zones from 0, with no
seismic risk, to zone 4 with maximum seismic risk. Pile group analysis was undertaken using the computer
program PC-MPILE 22 (derivative of M-PILE).
The UBC Zone assigned is dependent upon the nature of the
structure and the ground conditions. More onerous seismic The program determines the response of a group of piles to a
design criteria apply to tall buildings and other high-risk set of loads applied to a common pile cap. It calculates the load
structures, whereas temporary works are typically exempt from distribution among the piles in the group, and the vertical and
consideration of seismic effects. lateral deformation of the group, for any given pile layout and
applied loads.
In the UBC, Dubai is considered to be within seismic zone 0.
However, Dubai Municipality requires that buildings of 10 The program assumes that the ground behaves elastically, with
storeys or more and other high-risk structures be designed to a shear modulus of elasticity that increases linearly with depth.
meet the requirements of seismic zone 2A, where a peak
horizontal acceleration of 15% (0.15g) should be applied. In Pile group analyses under static and seismic loading conditions
order to comply with the Municipalitys building regulations were undertaken. The preliminary settlement estimates of pile
for seismic qualification, seismic loads were included in groups under static loading conditions were calculated to be
foundation loading for Sections 2 and 3 of the Snow Centre in within 10 mm. The seismic analysis was undertaken using the
accordance with the requirements of seismic zone 2A. shear modulus at very small strain (Gmax ). Gmax was determined
from the correlation between SPT N and Gmax . 23,24 Values of
The determination of the shear modulus at very small strain Gmax for the upper sand deposit were estimated to be between
(Gmax ), which represents the ground stiffness under earthquake 25 and 50 MN/m2 .
loading conditions, is presented in section 7.9.
7.10. Durability and protection of foundations
As indicated before, chemical testing on soil, rock and
7.8.2. Liquefaction potential Two methods can be used to groundwater samples indicates elevated levels of water-soluble
evaluate the effect of liquefaction. 15,20 The adopted method 16 salts in the form of sulphates (as SO3 ) and chlorides (as Cl). The
uses the SPT N value, depth of test, anticipated ground sulphate attack causes deterioration of concrete in foundations.
acceleration, groundwater level and fines content of the soil in The chloride content indicates the risk of corrosion of the
order to determine the soil liquefaction potential. Soils that reinforcement. The sulphate and chloride content were assessed
have a liquefaction resistance factor (FL ) of less than 1.0 are using BRE Special Digest 1 25 and CIRIA Special Publication
vulnerable to liquefaction. SPT N values applicable to these 31. 26
soils should therefore be reduced in the seismic design by the
appropriate reduction factor (DE ). 7.11. Preliminary pile-testing programme

The results of the evaluation indicate that near-surface very 7.11.1. Test programme The preliminary pile-testing
loose/loose sands could be susceptible to liquefaction under programme was implemented ahead of the installation of
seismic loading. For piled foundations, a loss of lateral support working piles for the following reasons
and negative skin friction would occur in the upper section of
the pile as a result of liquefaction. However, because of the (a) to check the ultimate pile capacity derived using theoretical
limited thickness of these soils, and the shallow depth at which methods
they occur, it is considered that, if liquefaction occurs, the (b) to determine the loadsettlement behaviour, especially in
magnitude of negative skin friction generated and loss of the region of the working loads
support will not adversely affect the piles. (c) to check the effect of pile installation method on pile
capacity.
7.9. Pile group analysis
Settlement of the pile group is greater than that of a single pile The main details of the preliminary test piles are summarised in
carrying the same working load as that on each pile within the Table 6.
group. 1 This is because the zone of soil or rock that is stressed
by the entire group extends to a much greater width and depth The test piles were installed at locations considered to represent
than the zone beneath the single pile. the general conditions across the site. The locations are shown
in Fig. 4.
Computer programs have been established to model pilesoil
interaction behaviour, from which the settlement of a pile 7.11.2. Piling contract procedure The piling works were

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 113
Test Type Location (section/zone) Pile diameter: Length measured Working load: Maximum test
pile ref. mm from ground level: m kN load: kN

C1 Compression Section 1 600 15.0 3000 7 500


C2 Compression Section 2 750 12.8 3000 7 500
C3 Compression Section 3 1000 17.8 7400 18 500
C4 Compression Zone 4 (ref. Fig 4) 750 17.8 5600 14 000
T1 Tension (Section 1. Area of 750 11.8 2000 5 000
maximum tension loading)

Table 6. Preliminary test pile details

carried out as a separate contract ahead of the main strain gauges embedded in the concrete at discrete levels to
construction programme with the engineer responsible for evaluate load distribution and skin friction resistance along the
design and supervision. The engineer was responsible for pile shaft during the pile load test. The strain gauges were
deciding the working loads, settlement criteria, working installed as follows.
stresses, fabrication methods, pile diameters and penetration
depths. Tenders were invited on the basis of a detailed (a) Level 1 at 300 mm below pile head.
specification and drawings. The piling contractor was (b) Level 2 slightly higher than the rockhead level.
responsible for installing the piles and carrying out the pile (c) Level 3 generally midway between level 2 and level 4 or,
load tests in accordance with the specification. where applicable, at the boundary between the upper
sandstone and conglomerate.
(d) Level 4 at 300 mm above the toe of the pile.
7.11.3. Pile installation A crawler-mounted hydraulic piling
rig was used to install the bored cast-in-place piles. The main
installation procedure comprised drilling the pile borehole 7.11.4. Test procedure Before the pile load test was initiated,
under a water support system, lowering the reinforcement cage caliper logging and integrity testing were undertaken.
into the drilled borehole, and concreting the pile. The specified Furthermore, results of the concrete cube strength tests were
concrete strength was C-60. Fig. 7 shows the reinforcement checked to confirm that the concrete had reached the minimum
cage being lowered into the borehole. strength as per the specification.

The instrumentation programme comprised vibrating-wire The test type was the maintained load test, in which the load
was increased in stages to 2.5 times the working load, with
time, load and movements recorded at each stage of loading
and unloading. Each successive load increment was maintained
constant and held until the creep criteria were met. The test
procedure was based on the ICE specification. 27

The load was applied to the compression test pile by a


hydraulic jack, using as the reaction either kentledge or cable
anchors tied back to a crown system, as shown in Figs 8 and 9.
Dywidag bars were used to transfer the load from the cable
anchors to the crown by bearing plates. For the tension test
pile the conventional cross-beam and jack method was used.

Fig. 7. Preliminary test pile installation: reinforcement cage Fig. 8. Test pile reaction arrangement: kentledge used as
being lowered into the drilled borehole reaction system

114 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai
Load: kN
0 1850 3700 5550 7400 9250 11 100 12 950 14 800 16 650 18 500
0

210000

Settlement: mm 3100
220000

230000

240000

250000

260000

270000

Fig. 10. Preliminary test pile C3: loadsettlement graph

Load: kN
0 1400 2800 4200 5600 7000 8400 9800 11 200 12 600 14 000
0

250000

Settlement: mm 3100
2150000

2300000

2200000

2250000

2300000

Fig. 9. Test pile reaction arrangement: cable anchors tied back


to a crown used as reaction system Fig. 11. Preliminary test pile C4: loadsettlement graph

7.11.5. Results of the preliminary test piles The results of the assumption that the skin friction resistance along the pile shaft
preliminary pile-testing programme are summarised in Table 7. carries most of the load, with negligible load being transferred
to the base. The significant load transmitted to the pile base in
The loadsettlement relationships for the preliminary test piles test pile C4 appears to be due to the skin friction being fully
C3 and C4 are presented in Figs 10 and 11 respectively. mobilised and at its ultimate value.

The skin friction values along the pile shaft, derived from the Table 9 shows the skin friction values adopted for the design of
strain gauge measurements at the maximum test load (250% of compression piles, which were based on the data presented in
working load), are presented in Table 8. Table 8.

7.11.6. Evaluation of test results The results of the In addition, a maximum value of ultimate unit skin friction
preliminary test piles C1, C2 and C3 have confirmed the design of 100 kN/m2 was considered for the very dense/dense

Pile Pile Ground Rock Pile Pile toe Test load Observed
ref. dia.: elevation: socket length: elevation: residual
mm mDMD* length: mbgl mDMD* 100% of working load 250% of working load movement:
m mm
Load: Observed Load: Observed
kN pile head kN pile head
movement: movement:
mm mm

C1 600 +3.2 10.0 15.0 11.8 3000 1.67 7 500 4.62 +0.16
C2 750 +3.2 6.8 12.8 9.6 3000 1.36 7 500 3.36 +0.19
C3 1000 +3.3 11.8 17.8 14.5 7400 1.45 18 500 6.36 2.14
C4 750 +3.3 7.8 17.8 14.5 5600 3.49 14 000 25.24 15.74
T1 750 +3.1 6.8 11.8 8.7 2000 +0.71 5 000 +4.12 +1.04

*DMD: Dubai Municipality Datum.


Negative values represent downward movement.

Table 7. Results of preliminary test piles

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 115
Preliminary Unit skin friction: kN/m2 Load carried The ultimate pile capacity
test pile by end derived from the pile test
Transition Upper Conglomerate bearing: % results was also compared
zone sandstone
with predictions using other
design methods. A summary
C1 100 280 N/A 3 of the more common
C2 169 317 N/A 1
C3 160 347 440 2 methods 10,13,2837 is
C4 75 295 495 16 presented in Table 10, and
T1 139 204 N/A N/A the ratios of the predicted to
the observed capacity are
N/A: not applicable. presented in Table 11. It can
Table 8. Preliminary test pile results: observed skin friction and end bearing resistance at be seen that widely differing
maximum test load predictions result from the
various methods.

Table 11 shows that the


prediction provided by Horvath and Kenney 28 for the ultimate
Stratum Elevation: mDMD Ultimate unit skin pile capacity is the closest to the observed value. It is therefore
friction: kN/m2
considered that this design method provides a reasonable
estimate of the ultimate capacity of rock sockets in the
Upper Rockhead to 10 280
carbonate sandstone in Dubai.
sandstone
Conglomerate 10 to 18 440
The method is used by Hyder Consulting for preliminary pile
Table 9. Compression piles: ultimate unit skin friction designs in Dubai.

7.11.8. Comparison between predicted and observed pile


sand horizons within the transition zone above the settlement The settlement observed during the preliminary pile
rockhead. testing indicates that the test piles performed better than
expected. A value of deformation modulus Em 300 MN/m2
The measured skin friction value in tension shown in Table 8 was back-analysed from the test data and adopted in the final
was used for the design of tension piles. The ratio of pile pile group settlement analysis.
capacity in tension to that in compression was calculated to
be 0.73.
7.12. Final pile design
The rock socket lengths were the main uncertainty in the
7.11.7. Comparison between predicted and observed ultimate preliminary design. These were re-calculated using the design
pile capacity Figure 12 shows a comparison between the skin friction values derived from the preliminary test pile
preliminary assessment of the ultimate pile capacities estimated programme.
using the Williams and Pells approach 11 and those derived
from the preliminary test pile programme. The comparison The use of the observed parameters in the final design made it
indicates that the Williams and Pells method significantly possible to reduce the factor of safety to 2.0 on the ultimate
overestimated the ultimate pile capacity. skin friction capacity.

The adopted working loads were multiplied by the factor of


Ultimate axial pile capacity: kN
0 5000 10000 30000
safety to obtain the ultimate design capacities. The ultimate
15000 20000 25000
0 capacities were used to calculate the required rock socket
lengths, which determined the founding levels. Consequently,
Pile elevation: mDMD

25 schedules giving pile toe elevations for all zones were issued
for construction.
210

7.13. Proof loading tests on working piles


215
A programme of proof pile load tests using the maintained load
was performed during the pile construction stage.
220

The total number of load tests carried out was seven (0.5% of
Pile diameter D 5 600 mm Predicted using Williams the number of piles). They covered a selection of pile
Pile diameter D 5 750 mm and Pells11 diameters, with a range of working loads, located within all site
Pile diameter D 5 1000 mm Derived from preliminary test
zones.
results

The piles were loaded to 150% of the specified working loads.


Fig. 12. Rock socket design (Zone 3) showing two sets of
results, predicted and derived from preliminary pile testing
The results obtained from the proof tests showed behaviour

116 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai
Pile design method Formula for ultimate unit skin friction based on Comments
unconfined compressive strength (UCS): MN/m2

:
Horvath and Kenney 28 0.20.25 (UCS)0 5 Based on results of load tests on rock socket piles
with diameter between 410 and 1220 mm. The
majority of the piles were socketed into
sedimentary rocks. Mostly shale or mudstone.
Williams and Pells 11 * (UCS) Based on rock sockets in sandstone, shale and
mudstone
:
Horvath et al. 29 0.20.3 (UCS)0 5 Revised the Horvath and Kenney 28 correlation
between the UCS and the ultimate skin friction,
and proposed a method for estimating the effects
of socket wall roughness on the skin friction
capacity.
Abbs and Needham 10 0.375 (UCS) (UCS , 1 MN/m2 ) Recommended for cast-in-place concrete piles in
0.375 + [(UCS  1) 3 0.1875] (UCS weak carbonate rocks.
13 MN/m2 )
:
Rowe and Armitage 13 0.45 (UCS)0 5 Recommended for regular clean rock sockets.
Reese and ONeill 30 0 15 (UCS) (UCS , 1.9 MN /m2 )
. The first equation was taken from Carter and
0.2 (UCS)0-5 (for higher strengths) Kulhawy 31 and the second equation from Horvath
and Kenney. 28 The method is adopted by the US
Department of Transportation. 32
:
Zhang and Einstein 33 0.4 (UCS)0 5 (smooth sockets) Based on a review of theory of Rosenberg and
. :
0 85 (UCS)0 5 (rough sockets) Journeaux, 34 Meigh and Wolski, 35 Williams and
Pells, 11 , Horvath, 36 and Kulhawy and Phoon. 37

*Rock socket reduction factor given by Williams and Pells. 11


Rock socket correction factor allowing for discontinuities given by Williams and Pells. 11
Application of a correction factor to the UCS to allow for discontinuities is recommended.

Table 10. Common design methods for rock socket piles in weak rocks

Pile design method Predicted ultimate unit skin Predicted/Observed


friction: kN/m2

Upper Conglomerate Upper Conglomerate


sandstone sandstone

Horvath and Kenney 28 200250 335418 0.710.89 0.760.95


Williams and Pells 11 445 612 1.59 1.39
Horvath et al. 29 200300 335502 0.711.07 0.761.14
Abbs and Needham 10 305 612 1.09 1.39
Rowe and Armitage 13 450 752 1.61 1.71
Reese and ONeill 30 150 335 0.54 0.76
Zhang and Einstein 33 400850 6691422 1.433.0 1.523.23

Table 11. Comparison between predicted and observed skin friction capacities

similar to that observed in the preliminary pile testing, with the 7.15. Feedback from piling contractors in Dubai
settlements at 150% of the working load to be within 5 mm. Discussions have been conducted with four main piling
contractors in Dubai with local knowledge about the pile
7.14. Construction supervision design approach adopted in the carbonate rocks in Dubai.
The installation of the working piles was monitored by a
geotechnical engineer. A key role of the supervising engineer It has been pointed out that, because of the limited published
was to identify any variation of ground conditions from that information on observed pile behaviour and pile design
established from the ground investigation. Recording of the practice in Dubai, the design approach is generally based on
ground strata in the pile boreholes is an essential check on the the available results of pile load tests undertaken as part of
ground investigation data. previous piling schemes.

Integrity testing was also carried out to test the structural The available pile test results do not appear to be
integrity of the bored and cast-in-place piles. In addition, tests comprehensive, owing to the difficulties experienced over the
to control the quality of materials and mixing of concrete were years in pile testing, as summarised below.
undertaken. The piling works were completed in February
2004. (a) It is not often feasible to obtain pile load test results to

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 117
failure, owing to the available capacity of the reaction
system to the load test set-up: that is, the reaction system
fails before pile failure is reached.
(b) It is not always possible, generally because of budget
constraints, to carry out well-instrumented load tests to
measure load distribution and skin friction resistance along
the pile shaft as well as the load transferred to the pile
base.
(c) The ultimate pile capacity is based on developing an
intimate bond between the concrete and the rock. The
formation of a smeared zone, or bentonite slurry, on the
bore of the shaft during pile installation may reduce this
bond, consequently reducing the pile skin friction capacity.

The piling contractors reported ratios of ultimate unit skin


Fig. 14. Mid station and upper end of the Snow Centre during
friction to UCS used in the preliminary pile design within the construction
range 0.150.3. It follows that a wide range of ultimate pile
capacities is predicted.

It is understood that end bearing resistance was allowed for in


9. CONCLUSIONS
the earlier pile designs in Dubai (a value of unit end bearing
The conclusions of this paper can be summarised as follows.
resistance in the order of UCS for the rock was reported to have
been used). However, as the instrumented pile test results have
shown a very small contribution of end bearing to the rock (a) An integrated geotechnical engineering service ranging
socket capacity, it is currently unusual to include it in the from management of ground investigation through to
design. detailed foundation analysis and design verification was
provided for a snow centre project in Dubai. The scheme
comprises an 80 m high slope and associated structures
8. CONSTRUCTION OF THE SUPERSTRUCTURE subject to axial, lateral and seismic loading.
Two photographs of the Snow Centre during construction (b) The available published information on the geology of the
showing the three main sectionsthe lower end, mid station Arabian Gulf in general and the United Arab Emirates/
and upper endare shown in Figs 13 and 14. Dubai in particular was studied. The Arabian Gulf is in an

Fig. 13. Lower end of the Snow Centre during construction

118 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai
area of extensive carbonate sedimentation. The geology of (iv) Confirm the effectiveness of the pile installation
Dubai is dominated by beach/dune sands with development technique.
of sabkha plains at the head of the creeks. The recent ( j) Literature review has shown that information on pile
sediments overlie aeolian carbonate sandstone. The surface behaviour in carbonate sands and rocks specific to the
level of the sandstone varies appreciably over the area, Arabian Gulf/Dubai is limited. Furthermore, although
being exposed at ground level in some areas in Dubai, and extensive piling works have been undertaken in Dubai in
occurring at depths of up to 12 m elsewhere. the last two decades, no published data on pile load tests
(c) Appropriate ground investigation is considered to be key to are available.
economic design. Inadequate geotechnical data require (k) Comparisons were made between the observed behaviour
conservative design to manage the risks of insufficient of the preliminary test piles and predictions obtained using
information. the more common design methods for rock-socketed piles
(d) An evaluation of foundation options was undertaken. As a with a view to identifying the most suitable method for pile
result, a rock-socketed piled foundation was selected. The design in Dubai. The results of the comparisons have
rock socket lengths ranged between 5 m and 15 m. shown that most of the design methods overestimate the
(e) It is considered that the main geotechnical risk in the rock ultimate capacity of the piles. However, the method
socket design was the variation of the rockhead level proposed by Horvath and Kenney 28 is considered to be the
across the site. To control the risk, the site was divided into most suitable for initial estimates of the ultimate capacity
zones in relation to the depth to rockhead. The variation of of piles in the carbonate sandstone in Dubai.
the rockhead level was considered in the ground model (l) Feedback from piling contractors in Dubai indicates that
used for foundation analysis. The design of the rock the results of instrumented pile load tests show that skin
sockets was based on the relevant rockhead level at each friction resistance along the rock socket provides the
zone. carrying capacity of the piles, with insignificant load being
(f) Pile working loads were calculated for a range of pile transferred to the base. Furthermore, piling contractors
diameters and rock socket lengths. This information was reported a wide range of predictions of ultimate pile
used to provide the most economic pile groups for all capacity. It is considered that the proposed guide for pile
sections of the proposed structure. design will assist with refinement of the current design
(g) A holistic approach for designing the piled foundation was practice.
adopted, taking into account the single pile behaviour,
effects of group action, and soilpile interaction behaviour.
10. ACKNOWLEDGEMENTS
Furthermore, supervision of the installation of the working
The author wishes to acknowledge the contribution of the
piles was provided to ensure that conditions encountered
project manager, Lewis OBrien of Hyder Consulting Middle
were those that had been anticipated.
East, who acted as a focal point for liaison with the design and
(h) The responsibilities of the engineer and contractor were
site teams as well as the client representative. The contribution
clearly defined in the piling contract and specification. A
of the senior structural engineer, Mark Grantham, is
single-point responsibility was undertaken by the engineer
particularly appreciated. Further acknowledgements are due to
for all aspects of the design of the piled foundation. It is
Fugro Middle East, who carried out the ground investigation,
recommended that, where the contractor is invited to
and Bauer International, who undertook the pile testing for the
design the piles on the basis of specified working loads,
data provided. Finally, thanks are due to the project managers,
and acceptability criteria, his responsibility should be
Mace International Ltd, who represented the client, Majid Al
precisely defined. It is essential to state in the piling
Futtaim Investments, for granting permission to publish this
contract whether the contractors responsibility covers the
paper.
design of individual piles or the whole piled foundation
that is, pile group behaviour etc.
(i) The preliminary pile testing undertaken in advance of the REFERENCES
installation of working piles achieved the following 1. TOMLINSON M. J. Pile Design and Construction Practice, 4th
objectives. edn. E & FN Spon, London, 1994.
(i) Check the ultimate pile capacity predicted using the 2. MAURENBRECHER P. M. and VAN DER HARST M. The
relationship between the ultimate capacity and geotechnics of the coastal lowlands of the United Arab
unconfined compressive strength proposed by Williams Emirates. Proceedings of the KNGMG Symposium on
and Pells. 11 A comparison between the predicted and Coastal Lowlands: Geology and Geotechnology, The Hague,
observed results indicates that the relationship used 1987, pp. 321335.
overestimated the ultimate pile capacity. The observed 3. EPPS R.J. Geotechnical practice and ground conditions in
data were therefore used to derive the ultimate capacity coastal regions of the United Arab Emirates. Ground
of the compression and tension piles in the final Engineering, 1980, 13, No. 5. 1922 and 2425.
design. 4. BRITISH STANDARDS INSTITUTION. Code of Practice for Site
(ii) Determine the loadsettlement behaviour of piles and Investigations. BSI, Milton Keynes, 1999, BS 5930.
the stiffness of the soil/rockpile system. A back- 5. BRITISH STANDARDS INSTITUTION. Methods of Test for Soils for
analysis of the test data enabled the deformation Civil Engineering Purposes. BSI, Milton Keynes, 1990, BS
modulus to be re-evaluated and subsequently used to 1377.
refine the pile group settlement analysis. 6. ALRIFAI L. Settlement and Consolidation of Clay Reinforced
(iii) Confirm that the piles behaved as friction piles, with only with Stone Columns. PhD thesis, University of Birmingham,
13% of the load transmitted to the base of the pile. UK, 1991.

Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai 119
7. BRITISH STANDARDS INSTITUTION. Code of Practice for 24. CRESPELLANI T. and VANNUCCHI G. Dynamic properties of
Foundations. BSI, Milton Keynes, 1986, BS 8004. soils. In Seismic Hazard and Site Effects in the Florence
8. CELESTINO T. B. and MITCHELL J. K. Behaviour of carbonate Area (VANNUCCHI G. (ed.)). Associazione Geotecnica
sands for foundation of offshore structures. In Proceedings Italiana, Rome, 1991, pp. 7180.
of the International Symposium on Offshore Engineering, 25. BUILDING RESEARCH ESTABLISHMENT. Concrete in
Rio de Janeiro, 1983, pp. 85102. Aggressive Ground. Part 1: Assessing the Aggressive
9. ABBS A. F. Design of grouted offshore piles in calcareous Chemical Environment. BRE, Garston, 2001, BRE Special
soils. In Proceedings of the Australia New Zealand Digest 1.
Conference on Geomechanics, Christchurch, 1992, pp. 26. CIRIA. The CIRIA Guide to Concrete Construction in the
128132. Gulf Region. Construction Industry Research and
10. ABBS A. F. and NEEDHAM A. D. Grouted piles in weak Information Association, London, 1984. CIRIA Special
carbonate rocks. In Proceedings of the 17th Offshore Publication 31.
Technology Conference, Houston, 1985, pp. 105112. 27. INSTITUTION OF CIVIL ENGINEERS. Specification for Piling and
11. WILLIAMS A. F. and PELLS P. J. N. Side resistance rock Embedded Retaining Walls. ICE, London, 1996.
sockets in sandstone, mudstone and shale, Canadian 28. HORVATH R. G. and KENNEY T. C. Shaft resistance of rock-
Geotechnical Journal, 1981, 18, No. 4, 502513. socketed drilled piers. In Proceedings of a Symposium on
12. WYLLIE D. C. Foundations on Rock, 2nd edn. E & FN Spon, Deep Foundations. ASCE, New York, USA, 1980, pp.
London, 1999. 182214.
13. ROWE R. K. and ARMITAGE H. H. Theoretical solutions 29. HORVATH R. G., KENNEY T. C. and KOZICKI P. Drilled piers
for the axial deformation of drilled shafts in rock. socketed into weak shale: methods of improving the
Canadian Geotechnical Journal, 1987, 24, No. 1, 114125, performance of drilled piers in weak rock. Canadian
126142. Geotechnical Journal, 1983, 20, No. 4, 758772.
14. SIMONS N. and MENZIES B. A Short Course in Foundation 30. REESE L. C. and ONEILL M. W. Drilled Shafts: Construction
Engineering, 2nd edn. Thomas Telford, London, 2000. Procedures and Design Methods. Federal Highway
15. JAPAN ROAD ASSOCIATION. Specification for Highway Administration, Washington, DC, USA, 1988. Publication
Bridges. Japan Road Association, Tokyo, 1996. No. FHWA-HI-88-042.
16. CIRIA. Design of Laterally loaded Piles. Construction 31. CARTER J. P. and KULHAWY E. H. Analysis and Design of
Industry Research and Information Association, London, Foundations Socketed into Rock. Geotechnical Engineering
1984, CIRIA Report 103. Group, Cornell University, 1987, Research Report 1493-4.
17. BROMS B. The lateral resistance of piles in cohesionless 32. CHARLES, W. W. Ng, SIMONS N. and MENZIES B. A Short
soils, Journal of the Soil Mechanics Division, ASCE, 1964, Course in Soil-Structure Engineering of Deep Foundations,
90, No. SM3, 123156. Excavations and Tunnels. Thomas Telford, London, 2004.
18. Japanese Railway Structure Design Standards and 33. ZHANG L. and EINSTEIN H. H. End bearing capacity of drilled
Guidelines for Foundations and Retaining Structures, 2000. shafts in rock. Journal of Geotechnical and
19. UNIFORM BUILDING CODE. International Conference of Geoenvironmental Engineering, ASCE, 1998, 124, No. 7,
Building Officials; Whittier, CA, USA, 1997, vol. 2 574584.
Structural Engineering Design Provisions. 34. ROSENBERG P. and JOURNEAUX N. L. Friction and end bearing
20. SEED H. B. and IDRISS I. M. Simplified procedure for tests on bedrock for high capacity socket design. Canadian
evaluating soil liquefaction potential. Journal of the Soil Geotechnical Journal, 1976, 13, August, 324333.
Mechanics and Foundation Division, ASCE, 1971, 97, 35. MEIGH A. C. and WOLSKI W. Design resistance rock sockets
No. 1, 12491273. in sandstone, mudstone and shale. Canadian Geotechnical
21. POULOS H. G. Pile behaviour: theory and application. Journal, 1979, 18, No. 4, 502512.
Geotechnique, 1989, 39, No. 3, 365415. 36. HORVATH R. G. Drilled Piers Socketed into Weak Shale:
22. PC-MPILE. Computer program for the analysis of pile Methods of Improving Performance. PhD thesis, University
groups under general three-dimensional loading of Toronto, Canada, 1982.
conditions. Mott MacDonald software in association with 37. KULHAWY E. H. and PHOON K. K. Drilled shaft side resistance
the Road Research Laboratory, 1990. in clay soil to rock. Proceedings of the Conference on Design
23. CIRIA. The Standard Penetration Test (SPT): Method and and Performance of Deep Foundations: Piles and Piers in
Use. Construction Industry Research and Information Soil and Soft Rock. ASCE, New York, 1993, Geotechnical
Association, London, 1995, CIRIA Report 143. Specification Publication No. 38, pp. 172183.

What do you think?


To comment on this paper, please email up to 500 words to the editor at journals@ice.org.uk
Proceedings journals rely entirely on contributions sent in by civil engineers and related professionals, academics and students. Papers
should be 20005000 words long, with adequate illustrations and references. Please visit www.thomastelford.com/journals for author
guidelines and further details.

120 Geotechnical Engineering 160 Issue GE2 Rock socket piles at Mall of the Emirates, Dubai Alrifai

You might also like