foundation of burj dubai

Foundation design for the Emirates Twin Towers,Dubai

Harry George Poulos and Andrew J. Davids

Abstract: This paper describes the foundation design process adopted for two high-rise buildings in Dubai, theEmirates Twin Towers. The foundation system for each of the towers was a piled raft, founded on deep deposits of cal-careous soils and rocks. The paper outlines the geotechnical investigations undertaken, the field and laboratory testingprograms, and the design process and describes how potential issues of low skin friction and cyclic degradation of skinfriction due to wind loading were addressed. An advanced numerical computer analysis was used for the design pro-cess, which was carried out using a limit state approach. This necessitated analysis of a large number of load cases,and the paper describes how the information was processed to produce design information. A comprehensive programof pile load testing was undertaken, and class A predictions of both axial and lateral load–deflection behaviour were infair agreement with the load test results. Despite this agreement, the overall settlements of the towers observed duringconstruction were significantly less than predicted. The possible reasons for the discrepancy are discussed.

Key words: case history, footings and foundations, full-scale tests, piles, rafts, settlement.

Résumé : Cet article décrit le processus de conception des fondations adopté pour deux tours à Dubai, les « EmiratesTwin Towers ». Le système de fondation pour chacune des tours consistait en un radier sur pieux reposant sur des dé-pôts profonds de sols et roches de carbonate. Cet article donne les grandes lignes des études géotechniques réalisées,les programmes d’essais en laboratoire et sur le terrain, et le processus de conception, et décrit comment les problèmespotentiels de faible frottement de surface et de dégradation cyclique du frottement de surface due aux charges de ventpeuvent être traités. Une analyse numérique de pointe à l’ordinateur a été utilisée pour le processus de conception quia été réalisé au moyen de l’approche d’état limite. Ceci a nécessité l’analyse d’un grand nombre de cas de charge-ments, et l’article décrit comment les données ont été traitées pour produire les informations pour la conception. Unprogramme élaboré d’essais de chargement sur pieux a été entrepris et des prédictions de classe A du comportement endéflexion sous chargement tant axial que latéral ont montré une concordance assez bonne avec les résultats des essaisde chargement. En dépit de cette concordance, les tassements globaux des tours observés durant la construction étaientappréciablement inférieurs à ceux prédits. On discute des raisons possibles de cet inconsistance.

Mots clés : histoire de cas, semelles/fondations, essais à l’échelle naturelle, pieux, radiers, tassement.

[Traduit par la Rédaction] Poulos and Davids 730

Introduction

The Emirates project is a twin tower development inDubai, one of the United Arab Emirates (UAE). The towersare triangular in plan form, with a face dimension of approx-imately 50–54 m. The taller office tower has 52 floors andrises 355 m above ground level, whereas the shorter hoteltower is 305 m tall. These towers are more than double theheight of the nearby World Trade Centre, which was for-merly the tallest building in Dubai. The office tower is cur-rently the eighth tallest building in the world, and the hoteltower is the 17th tallest. The twin towers are on an approxi-

mately 200 000 m2 site, which also incorporates low-levelretail and parking podium areas.

Figure 1 shows a photograph of the towers just after thecompletion of construction.

The foundation system for both towers involved the use oflarge-diameter piles, in conjunction with a raft. This paperdescribes the geotechnical investigation undertaken for theproject and the process used for the foundation design. Italso presents the results of a major program of pile testingand compares predicted and observed test pile behaviour.Finally, some limited data on settlements during constructionof the towers are presented, together with the predicted values.

Can. Geotech. J. 42: 716–730 (2005) doi: 10.1139/T05-004 © 2005 NRC Canada

716

Received 21 April 2004. Accepted 24 November 2004. Published on the NRC Research Press Web site at http://cgj.nrc.ca on8 June 2005.

H.G. Poulos.1 Coffey Geosciences Pty Ltd., 8/12 Mars Road Lane Cove West, P.O. Box 125 North Ryde, NSW 1670, Australiaand The University of Sydney, Department of Civil Engineering, NSW 2006, Australia.A.J. Davids.2 Building Structures, Hyder Consulting Pty Ltd., 116 Miller St., North Sydney, NSW 2065, Australia.

1Corresponding author (e-mail: [email protected]).2Present address: Coffey Geosciences Pty Ltd., 8/12 Mars Road Lane Cove West, P.O. Box 125 North Ryde, NSW 1670, Australia.

Ground investigation and sitecharacterization

Preliminary geotechnical data for the site were availablefrom earlier investigations via a series of boreholes drilled toabout 15 m depth. These revealed layers of sand or siltysand overlying very weak to weak sandstone. For the twintower development, it was clear that this preliminary infor-mation was inadequate; hence, a comprehensive investiga-tion was carried out. This investigation involved the drillingof 23 boreholes to a maximum depth of about 80 m. Thedeepest boreholes were located below the tower footprints;boreholes below the low-rise areas tended to be considerablyshallower. Standard penetration tests (SPTs) were carried outat nominal 1 m depths in the upper 6 m of each borehole andthen at 1.5 m intervals, until an SPT value of 60 wasachieved. The SPT values generally ranged between 5 and20 in the upper 4 m, increasing to 60 at depths of 8–10 m.Rotary coring was carried out thereafter. Core recoverieswere typically 60%–100%, and rock quality designation val-ues were also between about 60% and 100%.

Figure 2 shows the borehole information along a sectionthat passes through the two towers. It was found that thestratigraphy was relatively uniform across the whole site, soit was considered adequate to characterize the site with asingle geotechnical model. The ground surface was typicallyat a level of +1 to +3 m Dubai Municipality datum (DMD),and the groundwater level was relatively close to the surface,typically between 0 and –0.6 m DMD. The investigation re-vealed seven main strata, which are summarized in Table 1by material descriptions commonly adopted in Dubai.

Foundation parameter assessment and thegeotechnical model

In situ and laboratory testingBecause of the relatively good ground conditions near the

surface, a piled raft system was deemed appropriate for thefoundation of each tower. The design of such a system re-quires information on both the strength and the stiffness ofthe ground. Consequently, a comprehensive series of in situtests was carried out. In addition to standard SPTs and per-meability tests, pressuremeter tests, vertical seismic shearwave testing, and site uniformity borehole seismic testingwere carried out.

Conventional laboratory tests, including classificationtests, chemical tests, unconfined compression tests, point loadindex tests, drained direct shear tests, and oedometer consol-idation tests, were carried out. In addition, a considerableamount of more advanced laboratory testing was undertaken.This included stress path triaxial tests for settlement analysisof the deeper layers, constant normal stiffness (CNS) directshear tests (Lam and Johnston 1982) for pile skin frictionunder both static and cyclic loading, resonant column testingfor small-strain shear modulus and damping of the founda-tion materials, and undrained static and cyclic triaxial sheartests to assess the possible influence of cyclic loading onstrength and to investigate the variation of soil stiffness anddamping with axial strain.

Test resultsFrom the viewpoint of the foundation design, some of the

relevant findings from the in situ and laboratory testing wereas follows:

(i) The site uniformity borehole seismic testing did not re-veal any significant variations in seismic velocity, thusindicating that it was unlikely that major fracturing orvoids would be present in the areas tested.

(ii) The cemented materials were generally very weak toweak, with uniaxial compressive strength (UCS) valuesranging between about 0.2 and 4 MPa, with most valueslying within the range of 0.5–1.5 MPa.

(iii) The average angle of internal friction of the near-surfacesoils, from direct shear box tests, was about 31°.

(iv) The oedometer data for compressibility were consideredunreliable, because of the compressibility of the appara-tus being of a similar order to that of some of the sam-ples.

(v) Cyclic triaxial tests were carried out with one-way cy-clic loading and a maximum stress of up to 930 kPa.These tests indicated that the unit 4 sand deposit had thepotential to generate significant excess pore pressuresunder cyclic loading and to accumulate permanent de-formations under repeated one-way loading. It couldtherefore be susceptible to earthquake-induced settle-ments.

© 2005 NRC Canada

Poulos and Davids 717

Fig. 1. The Emirates Twin Towers soon after completion of con-struction.

(vi) The CNS shear tests indicated that cyclic loading hadthe potential to significantly reduce or degrade the skinfriction after initial static failure and that a cyclic stressof 50% of the initial static resistance could cause failureduring cyclic loading, resulting in a very low postcyclicresidual strength.

Figure 3 summarizes the values of Young’s modulus ob-tained from the following tests: (i) seismic data (reduced bya factor of 0.2, to account for a strain level appropriate to theoverall behaviour of the pile foundation); (ii) resonant col-umn tests (at a strain level of 0.1%); (iii) laboratory stresspath tests, designed to simulate the initial and incrementalstress states along and below the foundation system; and(iv) unconfined compression tests (at 50% of ultimatestress).

Figure 4 shows the ultimate static shear resistance, de-rived from the CNS test data, as a function of depth belowthe surface. With the exception of one sample, all testsshowed a maximum shear resistance of at least 500 kPa. The

measured values from the CNS tests were within and beyondthe range of design values for static skin friction of piles incemented calcareous soils tentatively suggested by Poulos(1988); that range was between 100 and 500 kPa, dependingon the degree of cementation.

Geotechnical modelThe key design parameters for the foundation system were

the ultimate skin friction of the piles, the ultimate end-bearing resistance of the piles, the ultimate bearing capacityof the raft, and the Young’s modulus of the soils for both theraft and the pile behaviour under static loading. For the as-sessment of dynamic response under wind and seismic load-ing conditions, Young’s modulus values for rapid loadingconditions were also required, together with internal damp-ing values for the various strata.

The geotechnical model for foundation design under staticloading conditions was based on the relevant available insitu and laboratory test data and is shown in Fig. 5. The ulti-

© 2005 NRC Canada

718 Can. Geotech. J. Vol. 42, 2005

Fig. 2. Geotechnical conditions. DMD, Dubai Municipality datum.

Unit No. Designation Material descriptionAvg. elevation of baseof unit (m DMD)

1 Silty sand Uncemented calcareous silty sand, loose to moderately dense –3.32 Silty sand Variably and weakly cemented calcareous silty sand –8.13 Sandstone Calcareous sandstone, slightly to highly weathered, well cemented –26.84 Silty sand Calcareous silty sand, variably cemented, with localized well-cemented bands –33.15 Calcisiltite Variably weathered, very weakly to moderately well cemented –53.56 Calcisiltite As for unit 5 –68.57 Calcisiltite As for unit 5 –79.0

Note: DMD, Dubai Municipality datum.

Table 1. Summary of main strata at site.

mate skin friction values were based largely on the CNSdata, whereas the ultimate end-bearing values for the pileswere assessed on the basis of correlations with the UCS data(Reese and O’Neill 1988) and also on the basis of previousexperience with similar cemented deposits (Poulos 1988).The values of Young’s modulus were derived from the datasummarized in Fig. 3. Although inevitable scatter existsamong the different values, there is a reasonably consistentgeneral pattern of variation of modulus with depth. Consid-erable emphasis was placed on the laboratory stress pathtests, which should have reflected realistic stress and strainlevels within the various units. The values for the upper twounits were obtained from correlations with the SPT data.

The bearing capacity of the various layers for shallowfoundation loading, pu, was estimated from bearing capacitytheory for the inferred friction angles, the tangents of whichwere reduced by a factor of two thirds to allow for the ef-fects of soil compressibility, as suggested by Poulos andChua (1985).

Limit state design approach

Ultimate limit stateThe piled raft foundation was designed via a limit state

approach based on Australian Standard AS 2159–1995,“Piling—Design and installation” (AS 2159 1995). The de-sign criteria for the ultimate limit state were as follows:

[1] R s* ≥ S*

[2] Rg* ≥ S*

© 2005 NRC Canada


Fig. 3. Summary of Young’s modulus values.

Fig. 4. Ultimate skin friction values from CNS tests.

where R s* is design structural strength; Rg* is designgeotechnical strength; and S* is design action effect (fac-tored load combination). The above two criteria were ap-plied to the entire foundation system, and the structuralstrength criterion (eq. [1]) was also applied to each individ-ual pile. The values for R s* and Rg* were obtained from theestimated ultimate structural and geotechnical capacities,multiplied by appropriate reduction factors. In this case, thestructural reduction factor was taken as 0.6. For thegeotechnical ultimate limit state, the worst response arisingfrom the pile–soil–raft interaction may not occur when thepile and raft capacities are factored downwards. As a conse-quence, additional calculations were carried out forgeotechnical reduction factors of both less than 1 (0.6) andequal to 1.

In addition to the normal design criteria, as expressed byeqs. [1] and [2], a criterion was imposed for the whole foun-dation to cope with the effects of repetitive loading fromwind action, as follows:

[3] ηRgs* ≥ S c*

where Rgs* is design geotechnical shaft capacity; S c* is maxi-mum amplitude of wind loading; and η is a factor assessedfrom geotechnical laboratory testing. The value selected forη, on the basis of laboratory data from CNS tests, was 0.5.The value for S c* was obtained from computer analyses,which gave the cyclic component of load on each pile forvarious wind loading cases.

Serviceability limit stateThe design criteria for the serviceability limit state were

as follows:

[4] ρmax ≤ ρall

[5] θmax ≤ θall

where ρmax is the maximum computed foundation settle-ment; ρall is the allowable foundation settlement, taken to be150 mm; θmax is the maximum computed local angular rota-tion; and θall is the allowable angular rotation, taken to be1/350 here.

Load combinationsFor each tower, 18 load combinations were analyzed: 1

loading set for the ultimate dead and live loading only; fourgroups of 4 loading sets for various combinations of dead,live, and wind loading for the ultimate limit state; and 1loading set for the long-term serviceability limit state (deadplus live loading).

AnalysesConventional pile capacity analyses were used to assess

the ultimate geotechnical capacity of the piles and raft. Forthe piles, this capacity was taken as the sum of the shaft andbase capacities. For the raft, account was taken of the layer-ing of the geotechnical profile and the large size of the foun-dation, and a value of 2.0 MPa was adopted for the ultimatebearing capacity. In these conventional analyses, it was as-sumed that the portion of the raft providing additional bear-ing capacity had a diameter of 3.6 m (three pile diameters)around each pile.

In additional to the conventional analyses, more completeanalyses of the foundation system were undertaken with thecomputer program geotechnical analysis of raft with piles(GARP) (Poulos 1994). This program uses a simplifiedboundary element analysis to compute the behaviour of arectangular piled raft when subjected to applied verticalloading, moment loading, and free-field vertical soil move-ments. The raft is represented by an elastic plate, the soil ismodelled as a layered elastic continuum, and the piles arerepresented by elastic–plastic or hyperbolic springs, whichcan interact with each other and with the raft. Pile–pile inter-actions are incorporated via interaction factors. Beneath theraft, limiting values of contact pressure in compression andtension can be specified so that some allowance can be madefor nonlinear raft behaviour. The output of GARP includesthe settlement at all nodes of the raft; the transverse, longitu-dinal, and torsional bending moments at each node in theraft; the contact pressures below the raft; and the verticalloads on each pile. In addition to GARP, the simplifiedboundary element program deformation analysis of pilegroups (DEFPIG) (Poulos and Davis 1980) was used to ob-tain the required input values of the pile stiffness and pile–pile interaction factors for GARP and for computing theoverall lateral response of the foundation system (ignoringthe effect of the raft in this case).

Both GARP and DEFPIG were used for the ultimate limitstate, with undrained soil parameters for the wind loadingcases and with drained soil parameters for the dead and liveloading only cases. The pile and raft capacities were fac-

© 2005 NRC Canada


Fig. 5. Geotechnical model adopted for design. Eu, undrainedYoung’s modulus; E ′, drained Young’s modulus, v ′, drainedPoisson’s ratio; fs, ultimate shaft friction; fb, ultimate pile endbearing; pu, ultimate lateral pile–soil pressure.

tored, as discussed in the “Ultimate limit state” sectionabove. They were also used for the serviceability limit state,but the pile and raft resistances were unfactored in this case.

Foundation design

Pile layoutThe number, depth, diameter, and locations of the founda-

tion piles were altered several times during the design pro-cess. The geotechnical and structural designers collaboratedclosely in an iterative process of computing structural loadsand foundation response. In the final design, the piles wereprimarily 1.2 m in diameter and extended 40 or 45 m belowthe base of the raft. In general, the piles were located di-rectly below 4.5 m deep walls, which spanned the raft andthe first-level floor slab. These walls acted as “webs”, whichforced the raft and the slab to act as the flanges of a deepbox structure. This deep box structure created a relativelystiff base for the tower superstructure, although the raft itselfwas only 1.5 m thick. Figure 6 shows the foundation layoutfor the hotel tower; the piles are generally located beneaththe load-bearing walls. Also shown in this figure are thecontours of predicted final settlement, which will be dis-cussed later.

Ultimate limit state: overall foundationTable 2 summarizes the maximum computed settlement

and angular rotation for each tower under the worst ultimatelimit state loadings as determined by the GARP analyses.For the cases here, the pile and raft capacities have been re-duced by a geotechnical reduction factor of 0.6. The calcu-lated values of settlement and angular rotation are notmeaningful, but the fact that the analysis does not predict

collapse (or very large settlements and angular rotations) in-dicates that the main geotechnical design criterion in eq. [2]is satisfied: that is, the reduced foundation resistancesclearly exceed the worst ultimate limit state loadings. Forboth foundation systems, the average ratio of cyclic compo-nent of load to design shaft resistance was found to be lessthan 0.5, thus satisfying the requirements of the criterion ineq. [3] for cyclic loading.

Serviceability limit state: overall foundationTable 3 summarizes the computed maximum settlement

and angular rotation under serviceability loading conditionsas determined by the GARP analyses. Although the com-puted values are relatively large, they nevertheless satisfy thedesign criteria set out in the “Serviceability limit state” sec-tion, above. Thus, the overall foundation systems were as-sessed to be satisfactory from the viewpoint ofserviceability.

Figure 6 shows the contours of settlement for the hoteltower after 24 months as computed by the GARP analyses.Similar settlement contours were developed for the officetower, which had a somewhat different pile layout. It can beobserved that for both towers, the predicted settlementsshowed a “dishing” pattern, with the settlements near thecentre being significantly greater than those near the edge ofthe foundation.

Raft designDuring the design process, the effects of raft thickness

were studied, but it was found that the performance of thefoundation was not greatly affected by raft thickness withinthe range of feasible thicknesses considered. It was thereforedecided to use a raft 1.5 m thick for the final design.

Initially, GARP was used to obtain estimates of the largestbending moments and shears in the raft for any of the com-binations of ultimate limit state loadings. Subsequently, itwas realized that the moments thus computed were likely tobe greater than the actual moments, because no account wastaken of the effects of the stiffness of the structure itself inthese calculations. Therefore, for the final assessment of raftmoments and shears, the computed pile stiffnesses and raftcontact pressures were provided to the structural engineer,who put them into a program for the complete analysis ofthe structure and foundation. Although the settlements were

© 2005 NRC Canada


Fig. 6. Computed final settlement contours for the hotel tower.Contour interval: 5 mm.

TowerMax. settlement(mm)

Max. angularrotation

Office 185 1/273Hotel 181 1/256

Table 2. Computed maximum settlement and angu-lar rotation ultimate limit state.

TowerMax. settlement(mm)

Max. angularrotation

Office 134 1/384Hotel 138 1/378

Table 3. Computed maximum settlement and angu-lar rotation serviceability limit state.

generally similar to those computed by GARP, the resultingvalues of moment and shear from the structural analysiswere significantly smaller than those from the oversimplisticmodelling of the raft as a uniform flat plate in the GARPanalysis.

Pile designTo enable assessment of the piles from the standpoint of

structural design, the maximum axial force, lateral force, andbending moment in each pile were computed by the follow-ing process: (i) the maximum axial force was computedfrom the GARP analyses for the various loading combina-tions; and (ii) the maximum lateral shear force and bendingmoment were computed with the program DEFPIG, allow-ing for interaction effects among the piles but ignoring anycontribution of the raft to the lateral resistance. The overallgroup was analyzed under the action of the various windloadings.

It was found that the largest axial forces were developedin the piles near the corners and in two of the core piles. Anumber of the piles reached their full geotechnical design re-sistance, but the foundation as a whole could still supportthe imposed ultimate design loads and therefore satisfied thedesign criterion in eq. [2].

Combined with the moments developed by the lateralloading, the load on some of the piles fell outside the origi-nal design envelope for a 1.2 m diameter pile with 4% rein-forcement, as supplied by the structural engineer. To addressthe problem of overstressing of the piles, a number of op-tions were considered, including increasing the reinforce-ment in the 1.2 m diameter piles, increasing the number of1.2 m diameter piles in the problem areas, and increasing thediameter of the “problem” piles to 1.5 m. The second optionwas adopted, and for the office tower, the total number ofpiles was increased from the original 91 to 102, and for thehotel tower, the number of piles was increased from 68 to 92.

Dynamic responseThe structural design required information on the vertical

and lateral stiffness of the individual piles in the two towerblocks for a dynamic response analysis of the entire struc-ture–foundation system. For the pile stiffness calculations,the program DEFPIG was used, with the following simplify-ing assumptions:(i) Each pile carried an equal share of the vertical and lat-

eral loads (although this might not have been an entirelyrealistic assumption, it did enable the stiffness of eachpile within the group, allowing for interaction effects, tobe evaluated straightforwardly).

(ii) The loadings from the most severe case of wind loadingwere considered.

(iii) The loading was very rapid, so undrained conditionsprevailed in the soil profile.

(iv) The pile heads were fixed against rotation to simulatethe effect of the restraint provided by the raft.

(v) The dynamic stiffness of the piles in the group environ-ment was equal to the static stiffness.

To check the latter assumption, approximate dynamicanalyses were also undertaken, using the approach outlinedby Gazetas (1991), incorporating dynamic interaction factorsand dynamic pile stiffnesses. It was found that in the fre-

quency range of interest (up to about 0.2 Hz), dynamic ef-fects on stiffness were minor, and in general the static stiff-ness values provided an adequate approximation of thedynamic foundation stiffness.

The range of frequencies was assessed to be lower thanthe natural frequency of the soil profile, which was of the or-der of 0.7–0.8 Hz. As a consequence, little or no radiationdamping could be relied on from the piles, so all the damp-ing would be derived from internal damping of the soil.From the resonant column laboratory test data, the averagevalue of the internal damping ratio was found to be about0.05. Following the recommendations of Gazetas (1991), thefoundation damping ratio was taken to be 0.05 for verticaland rocking motions and 0.04 for lateral and torsional mo-tions.

Seismic hazard assessmentA seismic hazard assessment was carried out by a special-

ist consultant, and for a 500 year return period, the peakground acceleration was assessed to be 0.075g. Assessmentswere then made of the potential for ground motion amplifi-cation and for liquefaction at the site. Because of the lack ofdetailed information on likely earthquake time histories, thepotential for site amplification was estimated simply on thebasis of the site geology, related to the shear wave velocitywithin the upper 30 m of the geotechnical profile (Joynerand Fumal 1984). On this basis, the site was assessed tohave a relatively low potential for amplification.

The presence of uncemented sands near the ground sur-face and below the water table suggested that the possibilityof liquefaction during a strong seismic event. The gradingcurves for these soils indicated that they might fall into therange commonly considered to be easily liquefied. The pro-cedure described by Seed and de Alba (1986) was used as abasis for assessing liquefaction resistance with SPT data.Because of the greater propensity of the calcareous sand togenerate excess pore pressures under cyclic loading, a con-servative approach was adopted, in which only a smallamount of fines was considered, and the design earthquakemagnitude was assumed to be 7.5. The overall risk of lique-faction was assessed on the basis of the liquefaction poten-tial index defined by Iwasaki et al. (1984). This indexconsiders the factor of safety against liquefaction within theupper 20 m of the soil profile. On this basis, the risk of liq-uefaction was judged to be low to very low, depending onthe borehole considered. Consequently, there appeared to beno need to consider special measures to mitigate possible ef-fects of liquefaction within the upper uncemented soil lay-ers.

Site settlement study

An assessment was made of the settlement over the entiresite at various times after the commencement of construc-tion. This study was undertaken to facilitate the design ofstructural interfaces between various parts of the project andin particular the interface between the towers and the po-dium structure.

The methodology involved the integration of the settle-ment due to each of the towers (considered as distributedloadings) and due to the low-rise structures (considered as a

© 2005 NRC Canada


series of equivalent uniform loadings) at defined pointsacross the site. For each of these loadings, the relationshipbetween settlement and distance was obtained. The follow-ing procedure was developed:(i) For the towers themselves, the settlements were avail-

able from the GARP analyses for the serviceabilityloadings.

(ii) The settlements due to tower loadings that occurred atpoints outside the towers were computed with the pilegroup settlement (PIGS) computer program (developedin-house by the first author). This program uses a sim-plified approach to compute the settlements both withinand outside pile groups subjected to vertical loading.The PIGS program uses the equations of Randolph andWroth (1978) to compute the single-pile stiffness val-ues, and the approximate approach described in Fleminget al. (1992) is used to compute pile interaction factors.The Mindlin equations are then used to compute groundsettlements outside the loaded area.

(iii) The loads acting on the low-rise areas were modelled asa series of uniformly loaded circular areas. The com-puter program finite layer elastic analysis (FLEA)(Small 1984) was used to compute the variation of sur-face settlement with distance from each loaded area.

(iv) The rate of settlement over time for both the tower foun-dations and the distributed loads was calculated on thebasis of two- and three-dimensional consolidation the-ory, allowing for the gradual increase of load with time(Taylor 1948). For these calculations, a coefficient ofconsolidation of 800 m2/year was assumed on the basisof field permeability tests and the assessed Young’smodulus values for the various layers. The rate of settle-ment of the towers was based on the solution for anequivalent isolated surface circular load, 40 m in diame-ter, located above a compressible material 60 m deep,with an impermeable base layer and a free-draining sur-face layer (Davis and Poulos 1972). For the low-rise ar-eas, the solutions for the rate of settlement of a stripfoundation were used, to allow for the continuity ofloading.

For the PIGS and FLEA analyses, linear soil behaviourwas assumed. A large Excel spreadsheet was developed toallow the summation of the effects of all 188 circular loadsand two towers assumed in the model. The settlement at 289points over the site was computed for 6-month intervals afterthe commencement of construction. A typical contour plotfor 24 months is shown in Fig. 7.

Pile load test program

As part of the foundation design process, a program ofpile load testing was undertaken, the main purpose being toassess the validity of the design assumptions and parameters.The test program involved the installation of three test pilesat or near the location of each of the two towers. Table 4summarizes the tests carried out. All piles were drilled underbentonite slurry support, with steel casing being provided inthe upper 3–4 m of each shaft. Because of the very large de-sign loads on the piles, it was not considered feasible to testfull-size piles in compression, and as a consequence, themaximum pile diameter for the pile load tests was 0.9 m.Nevertheless, it will be observed from Table 4 that the twocompression tests on the 0.9 m diameter piles involved avery high maximum test load of 30 MN.

Test detailsFigure 8 shows the test setup for the 0.9 m diameter test

piles. For the compression tests, the loading was supplied bya series of jacks, and the reaction was provided by 22 an-chors drilled into the underlying unit 5 calcisiltite. Each an-chor had a diameter of 100 mm and a total length of 40–45 m. The anchors were connected to the test pile via twocrowns (a larger one above a smaller unit) located above thejacks and load cells. For the tension tests, the reaction wassupplied by a pair of reaction piles 12 m long, with a cross-beam connecting the heads of the test and reaction piles. Inthe lateral load tests, the test pile was jacked against the ad-jacent 0.9 m diameter compression test pile, the centre-to-centre spacing between the piles being 4.5 m.

For piles P2(O) and P(2)H, the cyclic tension tests werecarried out prior to the lateral loading test. In each of the cy-clic tension tests, four parcels of uniform one-way cyclicload were applied.

Four main types of instrumentation were used in the testpiles: (i) strain gauges (concrete embedment vibrating wiretype), to allow measurement of strains along the pile shafts

© 2005 NRC Canada


Fig. 7. Computed settlement contours around the site after24 months. Contour interval: 10 mm.

TowerTestpile No.

Diameter(m)

Length(m) Test type

Max. testload (MN)

Hotel P3(H) 0.9 40 Compression 30.00P1(H) 0.6 25 Static tension 6.50a

P2(H) 0.6 25 Cyclic tension 3.25b

P2(H) 0.6 25 Lateral 0.20Office P3(O) 0.9 40 Compression 30.00

P1(O) 0.7 25 Static tension 6.50a

P2(O) 0.7 25 Cyclic tension 3.25b

P2(O) 0.7 25 Lateral 0.20aInitial estimated value; actual value was different.bMaximum load in cyclic test = 0.5 × maximum load in static tension

test.

Table 4. Summary of pile load tests.

and hence estimation of the axial load distribution: (ii) rodextensometers, to provide additional information on axialload distribution with depth; (iii) inclinometers (a pair, at180°, for each pile for the lateral load tests), to enable mea-surement of rotation with depth and hence assessment of lat-eral displacement with depth; and (iv) displacementtransducers, to measure vertical and lateral displacements.

For the two P3 piles, 44 strain gauges were used, 4 ateach of 11 levels, and extensometers were installed at 8 lev-els; for the P1 and P2 piles, there were 32 strain gauges, 4 ateach of 8 levels, and extensometers at 5 levels. In general,the strain gauges performed reasonably reliably. For the of-fice piles, only 3 of the strain gauges, all on P3(O), did notfunction properly, and for the hotel piles, 13 strain gauges(out of a total of 108) did not function properly: 1 on P3(H),4 on P2(H), and 8 on P1(H). The strain gauge readings weregenerally consistent with the extensometer readings.

Class A predictionsTo provide some guidance on the expected behaviour of

the piles during the test pile program, class A predictions ofthe load–deflection response of the test piles were calculatedand communicated to the main consultant, prior to the com-mencement of testing. The geotechnical model was similarto that used for design (see Fig. 5), with some minor modifi-cations to allow for the specific conditions revealed during

installation of the test piles. The following programs wereused to make the predictions: (i) PIES, for static compres-sion and tension tests (Poulos 1989); (ii) Static and CyclicAxial Response of Piles (SCARP), for cyclic tension tests(Poulos 1990); and (iii) ERCAP, for lateral load tests (CPI1992). All three programs were based on simplified bound-ary element analyses that represented the soil as a layeredcontinuum and were capable of incorporating nonlinearpile–soil responses and considering the effects of the reac-tion piles. Young’s modulus for the piles was assumed to be30 000 MPa. The input geotechnical parameters for the pre-dictions were those used for the design, as shown in Fig. 5.The program SCARP, however, required additional data oncyclic degradation characteristics for skin friction and endbearing. Some indication of skin friction degradation wasavailable from the CNS test data, but some of the parametersrelating to displacement accumulation had to be assessed onthe basis of judgement and previous experience with similardeposits (Poulos 1988). It was therefore expected that thepredictions for the cyclic tension test would be less accuratethan those for the static tests.

Predicted and measured test pile behaviour

Compression testsComparisons between predicted and measured test pile

behaviour were made after the results of the tests were made

© 2005 NRC Canada


Fig. 8. Setup for axial pile load tests.

available. Figure 9 compares the measured and predictedload–settlement curves for test P3(H) and reveals a fair mea-sure of agreement in the early stages. The predicted settle-ments, however, exceed the measured values, and themaximum applied load of 30 MN exceeded the estimated ul-timate load capacity of about 23 MN. The correspondingcomparison for office tower test pile P3(O) also revealedgood agreement in the early stages, but again, the predictedultimate load capacity of 23 MN was exceeded. Indeed, it isclear from Fig. 9 that the actual ultimate load capacity islikely to be well in excess of the maximum applied load of30 MN.

The fact that the actual capacity exceeded the predictedvalue was significant, because the values of ultimate skinfriction used for the predictions were well in excess of val-ues commonly used for bored pile design at that time inDubai.

Figure 10 shows the measured and predicted distributionsof axial load with depth for two applied load levels. Theagreement at 15 MN load is reasonable, but at 23 MN themeasured loads at depth are less than those predicted, indi-cating that the actual load transfer to the soil (i.e., the ulti-mate shaft friction) was greater than predicted.

Static tension testsFigure 11 compares the measured and predicted load–dis-

placement curves for the static tension test on pile P1(H) andindicates good agreement up to about 2 MN. At higherloads, the actual displacement exceeded the predicted value,but the maximum applied load of 5.5 MN exceeded the pre-dicted ultimate value of about 4.7 MN. For the office towertest pile, a similar measure of agreement was obtained, al-though the maximum load in that case was about 7.5 MN,because the test pile had a larger diameter (700 mm) thanthe originally planned 600 mm on which the predictionswere based.

Figure 12 shows the values of ultimate skin friction in-ferred from the axial load distribution measurements forboth the compression and the tension tests. These values arederived for the maximum applied test loads and are likely tobe less than the actual ultimate values. Also shown are theultimate values adopted for the design process, and these arein reasonable agreement with the measured values; indeed,the values used for design appear to be comfortably conser-

© 2005 NRC Canada


Fig. 9. Predicted and measured load–settlement behaviour forpile P3(H).

Fig. 10. Predicted and measured axial load distribution for pileP3(H). DMD, Dubai Municipality datum.

Fig. 11. Predicted and measured load–uplift behaviour for ten-sion test on pile P1(H).

vative. It is interesting to note that the design values weresubstantially larger (by about a factor of two) than the de-sign values commonly used in the UAE prior to the project.It appears that the CNS tests, which were used as the pri-mary basis for selecting the design values of skin friction,hold significant promise as a means of measuring relevantpile skin friction characteristics in the laboratory.

Cyclic tension testsFigure 13 shows the results of the cyclic tension test for

hotel tower pile P2(H). Four parcels of one-way cyclic loadwere applied, and for each parcel there was an accumulationof displacement with increasing number of cycles; this accu-mulation was more pronounced at higher load levels. Thepredictions from the SCARP analysis are also shown inFig. 13. Although the predictions at loads of <1 MN are rea-sonable, the theory significantly underestimates the accumu-lation of displacement at higher load levels. A similar (andlimited) level of agreement was obtained for the test on of-fice tower test pile P2(O). It had been expected that predic-tions of cyclic response might not be accurate, and thisexpectation was borne out by the comparisons. Clearly, theextent of displacement accumulation under cyclic loadingwas underestimated in the SCARP analysis. This analysis,based on an empirical approach originally developed byDiyaljee and Raymond (1982), assumes that the accumu-lated settlement of the soil at each element of the pile is re-lated to the number of cycles of loading and the cyclic stresslevel at that element. The computed settlements are imposedon the pile as external soil movements, and the analysiscomputes the resulting accumulated pile head settlement and

the distribution of load along the pile. The approach is veryapproximate and clearly has been unable to reproduce thereal behaviour of the pile under cyclic tension loading.

Nevertheless, from a practical viewpoint, the importantfeature of the cyclic tension tests was that a load of about50% of the static ultimate load could be applied without thepile failing (i.e., reaching an upward displacement of the or-der of 1%–2% of diameter). However, the tests indicatedthat the foundation rotations under repeated wind loadingcould be larger than predicted if the piles were subjected to a

© 2005 NRC Canada


Fig. 12. Ultimate skin friction values: design values and valuesderived from load tests.

Fig. 13. Measured and predicted load–uplift behaviour for cyclicuplift test on pile P2(H).

Fig. 14. Measured and predicted lateral load versus displacementfor pile P2(H).

© 2005 NRC Canada


cyclic tension in excess of about 25% of the ultimate staticuplift load capacity.

Lateral load testsFigure 14 shows the predicted and measured load–

displacement curves for the hotel tower test pile. Both thetest pile and the reaction pile responses are plotted. Theagreement in both cases is reasonably good, although thereis a tendency for the predicted deflections to be smaller thanthe measured values as the load level increases. A similarmeasure of agreement was found for the office tower pile,although the initial prediction had to be modified to allowfor the larger as-constructed diameter of the test pile. Itshould be noted that the predictions took account of the in-teraction between the test pile and the reaction pile. Had thisinteraction not been taken into account, the predicted deflec-tions would have been considerably larger than those mea-sured.

Figure 15 shows the predicted and measured deflectionprofiles along the hotel tower test pile at an applied load of150 kN. The agreement is generally good, although the mea-surements indicate a reversal of direction of deflection atabout 3.5 m depth, a characteristic that was not predicted.This characteristic has been observed from other analyses(for example, Matlock and Reese 1960) and may also reflectthe fact that the stiffness of the ground beyond about re-duced level (RL) –4 m was greater than assumed in the anal-ysis. The sharp kink in the measured deflection profile mayalso be due to the change in stiffness caused by the transi-tion from a cased to an uncased pile.

Measured and predicted buildingsettlements

Comparisons during constructionThe generally good agreement between measured and pre-

dicted performance of the test piles gave rise to expectationsof similar levels of agreement for the entire tower structurefoundations. Unfortunately, this was not the case. Measure-ments were available only for a limited period during theconstruction process, and these are compared with the pre-dicted time–settlement relationships in Fig. 16 for two typi-cal points within the hotel tower. The time–settlementpredictions were based on the predicted distribution of finalsettlement (Fig. 6), an assumed rate of construction, and arate of settlement computed from three-dimensional consoli-dation theory. At the time of the last available measure-ments, the tower had reached about 70% of its final height(i.e., a height of about 215 m). Figure 16 shows that the ac-tual measured settlements were significantly smaller thanthose predicted, being only about 25% of the predicted val-ues after 10–12 months. A similar level of disagreement wasfound for the office tower.

Figure 17 shows the contours of measured settlement at aparticular time during construction for the hotel tower. Al-though the magnitude of the measured settlements is farsmaller than predicted, the distribution bears some similarityto that predicted. The predicted ratio of final settlement atT4 to that at T15 is about 0.7, which is a similar order tothat measured. Thus, despite the considerable thickness ofthe raft and the apparent stiffness of the structure, the foun-dation experienced a dishing distribution of settlement,which is similar to that measured on some other high-risestructures on piled raft foundations, particularly theMesseturm in Frankfurt, Germany (Sommer 1993; Franke etal. 1994).

Possible reasons for discrepanciesThe disappointing lack of agreement between measured

Fig. 15. Measured and predicted deflection distributions for pileP2(H). DMD, Dubai Municipality datum.

Fig. 16. Measured and predicted time–settlement behaviour forthe hotel tower.

and predicted settlement of the towers prompted a “postmor-tem” investigation of possible reasons for the poor predic-tions. At least five reasons were suggested:(i) Some settlements may have occurred prior to the com-

mencement of measurements.(ii) The time–load pattern may have differed from that as-

sumed.(iii) The rate of consolidation may have been much slower

than predicted.(iv) The effects of pile interaction within the piled raft foun-

dation may have been overestimated.(v) The stiffness of the ground below RL –53 m may have

been underestimated.On the basis of the information available during construc-

tion, the first two of these suggested reasons did not seem tobe likely, and the last two were considered to be the mostlikely. Calculations were therefore carried out to assess thesensitivity of the predicted settlements to the assumptionsmade in deriving interaction factors for the piled raft analy-sis with GARP. When the program DEFPIG was used to de-rive the interaction factors originally used, it had beenassumed that the soil or rock between the piles had the samestiffness as that around the pile and that the rock below thepile tips had a constant stiffness for a considerable depth. In

reality, the ground between the piles is likely to be stifferthan that near the piles, because of the lower levels of strainwithin the rock below the pile tips is also likely to increasesignificantly with depth, because of both the increasing levelof overburden stress and the decreasing level of strain. TheDEFPIG program was therefore used to compute the interac-tion factors for a series of alternative (but credible) assump-tions regarding the distribution of stiffness both radially andwith depth. The ratio of the modulus of the soil between thepiles to that near the piles was increased to 5, and the modu-lus of the material below the pile tips was increased fromthe original 80 MPa to 600 MPa (the value assessed for therock at depth). The various cases are summarized in Table 5.

Figure 18 shows the computed relationships between in-teraction factor and spacing for a variety of parameter as-sumptions. It can be seen that the original interaction curve(No. 1) used for the predictions lies considerably above thosefor what are considered (in retrospect) more realistic as-sumptions. Since the foundations analyzed contained many

© 2005 NRC Canada


Fig. 17. Measured settlement contours for the hotel tower. Con-tour interval: 1 mm.

Fig. 18. Sensitivity of computed interaction factors to analysisassumptions. s, pile spacing; d, pile diameter.

CaseModulus below53 m (MPa)

Ratio of max.modulus to near-pile modulus

Max.settlement(mm)

Min.settlement(mm)

% loadon piles

Original calculations 80 1 138 91 93Case 2 80 5 122 85 93Case 3 200 5 74 50 92Case 4 600 5 40 23 92Case 5 600 1 58 32 92

Table 5. Summary of revised calculations for the hotel tower.

piles, the potential for overprediction of settlements wasconsiderable, as small inaccuracies in the interaction factorscan translate into large errors in the predicted group settle-ment (for example, Poulos 1993). In addition, Al-Douri andPoulos (1994) indicated that the interaction between piles incalcareous deposits may be much lower than that betweenpiles in a laterally and vertically homogeneous soil. Unfortu-nately, this experience was not incorporated into the class Apile group settlement predictions for the towers.

Revised settlement calculations, on the basis of these in-teraction factors, gave the results shown in Table 5. The in-teraction factors used clearly have a great influence on thepredicted foundation settlements, although they have almostno effect on load sharing between the raft and the piles. Themaximum settlement for case 4 is reduced to 29% of thevalue originally predicted, and the minimum settlement isabout 25% of the original value. If this case had been usedfor the calculation of the settlements during construction, thesettlement at point T15 would have been about 12 mm after11 months, which is in much closer agreement with the mea-sured value of about 10 mm than the original predictions.

The importance of proper assessment of the geotechnicalmodel in computing the effects of group interaction hasagain been emphasized by this case history.

Conclusions

The comprehensive investigation and testing program forthe Emirates project enabled the site to be characterizedmore completely than is usually possible with many pro-jects. Modern methods of in situ and laboratory testing wereused in conjunction with advanced methods of foundationanalysis to design the piled raft foundations. The limit stateapproach used for the foundation design involved a greatdeal of analysis (particularly because of the large number ofload combinations to be considered). One of the major chal-lenges was to process and portray the results from the analy-ses as useful design information.

The substantial design effort was complemented by acomprehensive program of pile load testing, including staticcompression tests, static and cyclic tension tests, and lateralload tests. The compression tests were among the largestcarried out in the Middle East and involved loads up to 30MN. Class A predictions of the performance of the test pileswere found to be in fair agreement with the measurements,although generally more conservative. In particular, the val-ues of ultimate skin friction along the pile inferred from theload tests were in good agreement with the values used fordesign, which were derived from CNS laboratory tests.There appears to be potential for this type of test to providea rational means of measuring pile skin friction characteris-tics in the laboratory.

The expectation that the tower settlements would be aswell predicted as the settlements in the load tests was not re-alized. The measured values during construction were onlyabout 25% of the predicted values. Possible reasons for thesignificant discrepancy were investigated, and it was foundthat at least some of the differences could be attributed tothe conservative assumptions made in deriving the pile set-tlement interaction factors that were used in the piled raftanalysis. When more realistic (in retrospect) assumptions

were made about the modulus values for soil between andbelow the piles, a much closer match to the measured settle-ments was possible. The importance of taking proper ac-count of interaction effects in pile group analyses and ofallowing for a more realistic distribution of ground stiffnessat depth was therefore reemphasized.

The Emirates project involved close interaction betweenthe structural and geotechnical designers in designing piledraft foundations for two complex and significant high-risestructures. Such interaction has some major benefits inavoiding oversimplification of geotechnical matters by thestructural engineer and oversimplification of structural mat-ters by the geotechnical engineer. Such interaction thereforepromotes the development of effective and economical foun-dation and structural designs.

Acknowledgements

The permission of His Highness General Shaikh Moham-med bin Rashid al-Maktoum to publish this paper is grate-fully acknowledged. Dr. James Apted provided expert advicein relation to pile testing. Patrick Wong, Robert Lumsdaine,Leanne Petersen, Paul Gildea, Jeff Forse, and Strath Clarkewere involved in various aspects of the field and designwork and the subsequent supervision of pile constructionand testing. Dr. Julian Seidel carried out the CNS testing atMonash University, Australia.

References

Al-Douri, R., and Poulos, H.G. 1994. Interaction between jackedpiles in calcareous sediments. In Proceedings of the 13th Inter-national Conference on Soil Mechanics and Foundation Engi-neering, New Delhi, January 1994. Oxford Press, UK. Vol. 3,pp. 1669–1672.

AS 2159. 1995. Piling—Design and installation (AS 2159–1995).Standards Australia, Sydney, Australia.

CPI. 1992. ERCAP users’ manual. Coffey Partners International,Sydney, Australia.

Davis, E.H., and Poulos, H.G. 1972. Rate of settlement under threedimensional conditions. Géotechnique, 22(1): 95–114.

Diyaljee, V.A., and Raymond, G.P. 1982. Repetitive load deforma-tion of cohesionless soil. Journal of the Geotechnical Engi-neering Division, ASCE, 108(GT10): 1215–1229.

Fleming, W.G.K., Weltman, A.J., Randolph, M.F., and Elson, W.K.1992. Piling engineering. 2nd ed. Halsted Press, New York.

Franke, E., Lutz, B., and El-Mossallamy, Y. 1994. Measurementsand numerical modeling of high-rise buildings on Frankfurtclay. In Vertical and Horizontal Deformations of Foundationsand Embankments: Proceedings of Settlement ’94, College Sta-tion, Tex., 16–18 June 1994. Edited by A.T. Yeung and G.Y.Felio. American Society of Civil Engineers, Geotechnical Spe-cial Publication 40, pp. 1325–1336.

Gazetas, G. 1991. Foundation vibrations. In Foundation engineer-ing handbook. 2nd ed. Edited by H.V. Fang. Van NostrandReinhold, New York. pp. 553–593.

Iwasaki, T., Arakawa, T., and Tokida, K.-I. 1984. Simplified proce-dure for assessing liquefaction during earthquakes. Soil Dynam-ics and Earthquake Engineering, 3(1): 49–58.

Joyner, W.R., and Fumal, T. 1984. Use of measured shear-wave ve-locity for predicting geological site effects on strong motion. InProceedings of the 8th World Conference on Earthquake Engi-

© 2005 NRC Canada


© 2005 NRC Canada


neering, San Francisco, Calif. Prentice-Hall, Englewood Cliffs,N.J. Vol. 2, pp. 777–783.

Lam, T.S.K., and Johnston, I.W. 1982. A constant normal stiffnessdirect shear machine. In Proceedings of the 7th South EastAsian Conference on Soil Engineering, Hong Kong. pp. 805–820.

Matlock, H., and Reese, L.C. 1960. Generalized solution for later-ally loaded piles. Journal of the Soil Mechanics and FoundationsDivision, ASCE, 86(SM5): 63–91.

Poulos, H.G. 1988. The mechanics of calcareous sediments. JohnJaeger Memorial Lecture, Australian Geomechanics, SpecialEdition. pp. 8–41.

Poulos, H.G. 1989. PIES user’s manual. Centre for GeotechnicalResearch, University of Sydney, Sydney, NSW, Australia.

Poulos, H.G. 1990. SCARP user’s manual. Centre for GeotechnicalResearch, University of Sydney, Sydney, NSW, Australia.

Poulos, H.G. 1993. Settlement of bored pile groups. In Deep foun-dations on bored and auger piles. Edited by W.F. van Impe. A.A.Balkema, Rotterdam, The Netherlands. pp.103–117.

Poulos, H.G. 1994. An approximate analysis of pile–raft interac-tion. International Journal for Numerical and AnalyticalMethods in Geomechanics, 18: 73–92.

Poulos, H.G., and Chua, E.W. 1985. Bearing capacity of founda-tions on calcareous sand. In Proceedings, 11th InternationalConference on Soil Mechanics and Foundation Engineering, SanFrancisco, Calif. Vol. 3, pp. 1619–1622.

Poulos, H.G., and Davis, E.H. 1980. Pile foundation analysis anddesign. John Wiley, New York.

Randolph, M.F., and Wroth, C.P. 1978. Analysis of deformation ofvertically loaded piles. Journal of Geotechnical Engineering,ASCE, 104(GT12): 1465–1488.

Reese, L.C., and O’Neill, M.W. 1988. Drilled shafts: constructionprocedures and design methods. US Department of Transporta-tion, Washington, D.C. Publication No. FHWA-H1-88-042.

Seed, H.B., and de Alba, P. 1986. Use of SPT and CPT tests forevaluating the liquefaction resistance of sands. In Use of in SituTests in Geotechnical Engineering: Proceedings of In Situ ’86, aSpecialty Conference, Blacksburg, Va. Edited by S.P. Clemence.American Society of Civil Engineers, New York. GeotechnicalPublication No. 6, pp. 281–302.

Small, J.C. 1984. User’s manual: program FLEA—Finite LayerElastic Analysis. School of Civil and Mining Engineering, Uni-versity of Sydney, Sydney, NSW, Australia.

Sommer, H. 1993. Development of locked stresses and negativeshaft resistance at the piled raft foundation—Messeturm Frank-furt Main. In Deep foundations on bored and auger piles. Editedby W.F. van Impe. A.A. Balkema, Rotterdam, The Netherlands.pp. 347–349.

Taylor, D.W. 1948. Fundamentals of soil mechanics. John Wiley,New York.

foundation of burj dubai

Documents