Seismic Design Of Pile Foundations - Guru Nanak Dev .
Indian Geotechnical Conference – 2010, GEOtrendzDecember 16–18, 2010IGS Mumbai Chapter & IIT BombaySeismic Design of Pile FoundationsMadabhushi, S.P. GopalReadere-mail: mspg1@cam.ac.ukGeotechnical Engineering, University of Cambridge, CB2 1PZ, United KingdomABSTRACTPile foundations are widely used around the world to transfer loads from the superstructure into competentsoil strata below the ground. Earthquakes can create additional loading scenarios on piles that needspecial attention. This is particularly true if the pile foundations pass through shallow, liquefiable layerssuch as saturated, fine sands or silts. There are many examples of distress suffered by pile foundationsfollowing soil liquefaction in recent earthquakes. In this paper the experiences from the Haiti earthquakeof 2010 will be presented. Several failure mechanisms of piles in liquefiable soils such as pile bucklingand excessive settlement of piles in liquefied soil layers were developed at Cambridge University basedon extensive centrifuge experiments. It will be shown that some of these anticipated failure mechanismsof pile foundations are confirmed by the observations in Haiti. The value of dynamic centrifuge modellingin deciphering the pile behaviour through observation of failure mechanisms is emphasised.1. HAITI EARTHQUAKE OF 2010On 12th January 2010, a powerful earthquake with amoment magnitude Mw 7.2 hit Haiti that has causedextensive damage to civil engineering infrastructureand led to a death toll of nearly 250,000. Thisearthquake led to an unprecedented effort of usingsatellite and pictometric imaging to evaluate both thescale of damage and in damage assessments, Saito etal (2010). While the detailed analysis of such imageryis beyond the scope of this paper, specific use of highresolution pictometric and pre and post earthquakeimaging to assess the liquefaction damage at the portin Port-au-Prince will be considered in this paper.More details of the EEFIT mission to Haiti that wasundertaken with support from EPSRC, UK and supportfrom Institution of Structural Engineers, London canbe found at http://www.istructe.org/knowledge/EEFIT.Soon after the earthquake, it became clear thatextensive liquefaction has occurred in the vicinity ofthe main port in the capital city, Port-au-Prince.Excessive settlement of container crane structure asshown in Figure 1 made the use of port to receiveemergency relief from rest of the world impossible.This situation highlighted the importance of designingcritical facilities such as ports and harbours to withstandearthquake loading.Fig. 1: Settlement of Container Crane in Port-au-PrincePre and post earthquake images reveal the extentof damage to the port facilities as shown in Figures 2and 3.Fig. 2: Pre-earthquake Satellite Image of the Port
16S.P. Gopal MadabhushiFig. 3: Post-earthquake Satellite Image of the PortComparing these two satellite images it can beseen that several sections of the south wharf havecollapsed and the entire section of the north pier hascollapsed. In addition to the satellite images highresolution pictometric images were also available forthe port. Pictometric images are aerial photographstaken in-flight akin to the street-view images ofGoogleMaps. These offer an oblique view of the targetand give a perspective of damage suffered by thestructures. An example of these pictometric images isshown in Figure 4 that clearly depicts the damage tocontainer crane seen in Figure 1 above.Fig. 5: Settlement of Piers Cause Self-articulation of theBridge StructureSimilarly a wharf structure suffered excessivedamage as the pile foundations supporting the wharfsettled into liquefied soil. This is shown in Figure 6which depicts the formation of plastic hinges at thepile heads. This type of failure mechanism was predictedby Madabhushi et al (2009) based on a series ofextensive centrifuge tests that were conducted tounderstand settlement of piles into liquefied sandoverlying dense sand strata. More details of this studyare discussed later in this paper.Fig. 6: Hinging of Piles Supporting a Wharf StructureFig. 4: Pictometric Image Showing the Damage to theContainer Crane StructureIn addition to the above aerial imagery, the EEFITfield mission was also able to obtain valuable data onperformance of the port structures. The island seennear the top-left hand corner in Figure 4 above hassuffered excessive settlement following liquefaction. Asa result, connecting bridge and wharf structuressupported on pile foundations showed severe damage.In Figure 5 below the damage suffered by thebridge is shown. As the pier structure and the approachabutment on the island side settled, the bridge deckscracked to allow for these settlements. This caused the‘self-articulation’ of the bridge.Fig. 7: Failure Mechanism of Hinging Due to PileSettlements, Madabhushi et al (2009a)
17Seismic Design of Pile FoundationsComparing Figures 6 and 7 it can be concludedthat the failure mechanism of piles in liquefiable soilsderived from dynamic centrifuge tests was validatedfrom the observations following Haiti earthquake.2. CURRENT DESIGN PRACTICESIn earthquake geotechnical engineering the liquefactionpotential of a given site is first established. Themaximum credible earthquake that is possible at thesite may be obtained from the seismic zonation mapsfor the region. The peak ground acceleration (amax)that can occur may be determined for the maximumcredible earthquake by using an appropriate attenuationrelationship, for example, Ambraseys et al., (2005).The cyclic shear stress τe generated by this earthquakecan be calculated following Seed and Idriss (1971) as;τ e 0.65amaxσ 0 rdgstiffness.It is both important and interesting to understandthe behaviour of structures on liquefiable soils. Withincreased pressure on land, the geotechnical engineershave to look at designing foundations of structures onsites which may be susceptible to liquefaction. Alsogeotechnical engineers may have to design retrofittingmeasures to existing historic structures that may belocated on liquefiable soils as is the case in southernEuropean countries like Italy and Greece. A large EUfunded project called NEMISREF that has concludedrecently, was aimed at developing novel liquefactionresistance measures for existing foundations. (for furtherdetails http://www.soletanche-bachy.com/nemisref andMitrani and Madabhushi (2008).0.6(1)where σo is the total stress and rd is a stress reductionfactor. Field test data from either SPT or morepreferably CPT tests can be used to establish theliquefaction potential of the site. The corrected SPTvalues (N160) can be used, for example. In Figure 8the liquefaction lines that demarcate ‘liquefaction’ and‘no liquefaction are shown for three types of sands,following Eurocode 8 Part 5 (2004) as re-plotted byMadabhushi et al (2009b). It is preferable to obtainCPT data compared to SPT testing. If CPT data isavailable for the given site then, similar relationshipscan be found between CPT data and the liquefactionpotential, for example as proposed by Robertson andWride (1998). While progress has been made in thearea of determining whether a given site is liquefiablethere are several uncertainties associated with soilliquefaction. For example, if the cyclic shear stressvalue from maximum credible earthquake and thecorrected SPT value for a layer give points very closeto liquefaction potential lines, then it is not clearwhether liquefaction should be expected or not. Forexample if the cyclic stress ratio (CSR) is 0.2 and theN160 is 20 for clean sand site, such a point will lieclose to but below the liquefaction potential line.However, good engineering judgement requires thatsignificant liquefaction be expected at such a site. Soeven when the N160 data suggests that the site willfall into ‘no liquefaction’ category, significant excesspore pressures can be generated during sufficientlystrong earthquake events. In such events, the structuresfounded on the site can suffer excessive settlementand/or rotations brought about by the lowering of theeffective stress and a consequent degradation in 1Silty s and (35% fines)Silty s and (15% fines)Clean s and ( 5% fines )0010203040N160Fig. 8: Liquefaction Potential Charts, Re-plotted FollowingEC8: Part 5 (2004), Madabhushi et al (2009b)In order to investigate earthquake geotechnicalengineering problems, it is necessary to conductdynamic centrifuge tests. These tests can be conductedon small scale physical models in the enhanced gravityfield of the centrifuge in which earthquake loadingcan be simulated. The technique is now well establishedand produces very valuable insights into failuremechanisms of foundations without the need to waitfor large earthquake events to happen. Similarly nonlinear FE analyses can be carried out to investigatethese problems, but it is essential that such analysesinclude realistic soil models and incorporate the soilplasticity accurately.
18S.P. Gopal MadabhushiIn this paper the emphasis will be on dynamiccentrifuge modelling. After a brief introduction to thetechnique itself, the paper will present the case of pilebehaviour in liquefiable soils and how centrifugemodelling helped to clarify the possible failuremechanisms of pile foundations.3. DYNAMIC CENTRIFUGE MODELLINGPhysical modelling in earthquake engineering withreduced scale experiments in the centrifuge is now widelyconsidered as ‘the’ established experimental technique ofobtaining data in controlled conditions to help engineersand researchers to understand the mechanisms involvedin the response of soil – structure systems to seismicloading. This experimental approach recreates the stressstate in soils which is a fundamental requirement toobserve realistic soil behaviour. In Figure 9 a view of the10 m diameter Turner beam centrifuge at Cambridge ispresented.Fig. 10: A View of the SAM Earthquake ActuatorFig. 11: Plan View of a Centrifuge Model of SlopingGround with Square and Circular Piles Loaded on theCentrifugeTable 1: Specifications of SAM ActuatorFig. 9: The 10m Diameter Turner Beam CentrifugeThe success of earthquake geotechnical engineeringat Cambridge depended to a large extent on thesimple mechanical actuators that have been used formore than 30 years. First attempts of centrifuge shakingtables were mechanical 1D harmonic devices based onleaf spring device (Morris, 1979), bumpy road tracks(Kutter, 1982). The current earthquake actuator atCambridge relies on Stored Angular Momentum (SAM)to deliver powerful earthquakes at high gravities wasdeveloped and is in operation for 14 years, Madabhushiet al (1998). In Figure 10 the front view of the SAMactuator is shown while in Figure 11 a view of theSAM actuator loaded onto the end of the centrifuge ispresented. In Table 1 the specifications of the SAMactuator are presented. The model seen in Figure 11was from an investigation carried out by Haigh andMadabhushi (2005), on lateral spreading of liquefiedground past square and circular piles.ParameterMaximum g-level ofoperationValue100 gEarthquake duration of choice56 m (L) 25 m (B) 22 m(H)80 m (L) 25 m (B) 40 m(H)Up to 0.4g of bed rockaccelerationFrom 0 s to 150 sEarthquake frequency ofchoiceFrom 0.5 Hz to 5 HzSwept sine wave capabilityDimension of the soil modelsEarthquake strength of choiceNote: All parameters above are in prototype scaleThe SAM earthquake actuator is a mechanicaldevice which stores the large amount of energy requiredfor the model earthquake event in a set of flywheels.At the desired moment this energy is transferred tothe soil model via a reciprocating rod and a fast
19Seismic Design of Pile Foundationsacting clutch. When the clutch is closed through ahigh pressure system to start the earthquake, theclutch grabs the reciprocating rod and shakes with anamplitude of 2.5 mm. This is transferred to the soilmodel via a bell crank mechanism. The leveringdistance can be adjusted to vary the strength of theearthquake. The duration of the earthquake can bechanged by determining the duration for which theclutch stays on. Earthquakes at different frequencytone bursts can be obtained by selecting the angularfrequency of the flywheels.With the capabilities of the SAM actuator and thegeotechnical centrifuge facilities described above, it ispossible to investigate a wide range of earthquakegeotechnical engineering problems including soilliquefaction. In this paper, the particular example ofpile behaviour in liquefiable soils will be considered.It will be demonstrated how dynamic centrifugemodelling can help understand the failure mechanismsinvolved. Further it will be shown how the thinkingabout pile behaviour in liquefiable soils has evolved asmore information is deciphered from the valuablecentrifuge test data, culminating in development ofsimplified procedures for estimating settlements of pilegroups.More recently a new servo-hydraulic earthquakeactuator was developed to enhance earthquake testingcapabilities at Cambridge. This device is able tocomplement the capabilities of SAM actuator with theability to simulate desired realistic earthquake motionssuch as a Kobe earthquake motion or Northridgeearthquake motion. This device was custom-built toadapt to the existing services and configuration of the10m Turner beam centrifuge shown in Figure 9. Itsmaximum operational g level is limited to 80g. Thespecifications of this actuator are presented in Table 2.An early example of the performance of the new servohydraulic earthquake actuator is presented in Figure 13.The shaker was commanded to perform a Kobe motion andit responded satisfactorily as seen in Figure 13. The higherfrequency components still need further amplification ascan be seen in the Fourier transform of the demand andachieved traces in Figure 14.Table 2: Specifications of the new Servo-Hydraulic ActuatorFig. 13: Simulation of the Kobe Earthquake MotionParameterMaximum g-level ofoperationDimension of the soilmodelsEarthquake strength ofchoiceEarthquake duration ofchoiceEarthquake frequency ofchoiceFig. 12: A View of the New Servo-Hydraulic EarthquakeActuator Developed at Cambridge UniversityValue80 g56 m (L) 25 m (B) 22 m(H)80 m (L) 25 m (B) 40 m(H)Up to 0.6g of bed rockaccelerationFrom 0 s to 80 sFrom 0.5 Hz to 5 HzRealistic earthquake motioncapabilityNote: All parameters above are in prototype scaleFig. 14: Fourier Spectra of the Demand and Achieved Traces
20S.P. Gopal Madabhushi4. MODELLING OF SINGLE PILES INLIQUEFIABLE SOILSPiles are long, slender members that carry large axial loadsas they transfer super structure loads into competent soilstrata. In normal soils the slenderness is not an issue, aslateral soil pressures support the pile in the radial directions.However in soft soils such as marine clays, piles can sufferbuckling failure. This problem of pile buckling in soft clayshas been investigated earlier by many researchers, forexample, Siva Reddy and Valsangkar (1970).In sands the problem of pile buckling does not arise aslarge lateral pressures that offer lateral support to the pilecan be easily generated. However when the sandy soils areloose and saturated, they can suffer liquefaction. Underthose circumstances piles can lose all the lateral supportand can suffer buckling failure as illustrated in Figure 15by forming a plastic hinge close to the base of the pile.Madabhushi et al (2009a) describe several possible failuremechanisms for single piles. It must also be pointed outthat the piles need to be ‘rock socketed’ at the base so thatno settlement of the pile itself is possible. This type of basecondition can occur when piles are transferring the loadfrom the ground surface onto the bed rock.is 0.5.The centrifuge model was tested at 50g’s and subjectedto a base acceleration of 0.2g at the bedrock level. Thiscaused the loose sand to liquefy fully which was confirmedby excess pore pressure measurements as shown in Figure16. The piles have suffered buckling failure as seen in Figure17. After the test the model piles were extracted from thesoil model to examine the location of plastic hingeformation. As seen in Figure 18, the plastic hingingoccurred at approximately ¼ of the length below the pilehead.Fig. 16: Excess Pore Pressure Time Histories at Base, Middleand Top of the Sand Layer (Dashed Line Indicates FullLiquefaction)Fig. 15: Buckling of a Pile in Liquefied SoilThis mechanism of failure for single piles wasevaluated using dynamic centrifuge modelling byBhattacharya et al (2004). Tubular model piles made fromaluminium alloy were placed in loose, saturated sand. Thepile tips were fixed to base of the model container allowingno displacement or rotation. Axial load on these single pileswas modelled using brass weights. These loads werecalculated using Euler’s critical load formula for slendercolumns given by;Pcr π2EI(βL )2(2)where EI is the flexural rigidity of the pile, L is the lengthof the pile and β is a effective length factor that dependson the end fixity conditions. For the case of pile that isfixed at the base and free at the top, β is 2. For a pile group,where both ends of the pile may be considered as fixed, βFig. 17: Centrifuge Model ShowingBuckling of three Piles inLiquefied SoilsFig. 18: A View of theBuckled PileBased on these series of tests, it can be concluded thatpile buckling occurs in liquefied soils for the case of singlepiles carrying axial loads close to the Euler critical loads.Any imperfections in the pile will of course bring downthe Euler critical loads significantly. However single pilesare very rarely employed in geotechnical design. One suchcase is the Showa Bridge foundations in which the bridgepiers were extended below the water level and supportedon rows of single piles. Based on the evidence of the
21Seismic Design of Pile Foundationscentrifuge test data, Bhattacharya et al (2005) argued thatthe failure of the Showa bridge during the 1964 Niigataearthquake could be attributed to buckling of the piles.Interesting as the above results and their implicationsare, a few queries arise. Firstly, compared to Figure 15 inwhich the plastic hinging in the pile occurs close to thebase of the pile, the location of the plastic hinge in thecentrifuge test occurs quite a way up as seen in Figure 18.As plastic hinges form when the maximum bendingmoments exceed the plastic moment capacity of the pilesection, clearly the liquefied soil below this level must beoffering some support to the pile. This may be due to themonotonic shearing of the liquefied soil demanded by thedeforming pile causes dilation and therefore a temporaryreduction of the excess pore pressure. This can cause theliquefied soil to gain some strength temporarily and offerresistance to pile buckling. As a result the pile cannot buckleat the base but can only do so higher up, where the soildilation is compromised due to inflow of water fromsurrounding soil. Secondly, would pile buckling be an issuefor pile groups as the effective length of a pile in a groupcan be significantly smaller due to the end fixity conditions(fixed at base and top and therefore β 0.5 in Eq.2). Thisaspect is considered next.5. PILE GROUPS IN LIQUEFIED SOILSPiles are most often used in groups with a pile cap connectingall the pile heads. As a result there is considerable fixity ofpiles heads that provides resistance to rotation. In additionthe pile heads are all forced to translate together. Madabhushiet al (2009) describe several possible failure mechanisms forpile groups by forming plastic hinges. Typical examples forliquefiable level ground are shown in Figure 19 with fourhinge and three-hinge mechanisms.groups made from aluminium alloy were tested in loosesaturated sands. A typical cross-section of the centrifugemodels is shown in Figure 20. It must be noted that thebrass weights on the pile caps are restrained in the directionof earthquake shaking to remove the inertial effects i.e. thepile groups
Seismic Design of Pile Foundations Madabhushi, S.P. Gopal Reader e-mail: mspg1@cam.ac.uk Geotechnical Engineering, University of Cambridge, CB2 1PZ, United Kingdom ABSTRACT Pile foundations are widely used around the world to transfer loads from the superstructure into competent soil strata below the ground. Earthquakes can create additional loading scenarios on piles that need special .
Under seismic loads, deep foundations with fixed pile/pile-cap connection may be subjected to a large curvature demand at the pile head. Damage induced by local inelastic deformation depends on the . A common approach for seismic design of pile foundations assumes that a laterally loaded soil-pile system can be analyzed as a flexural member .
To observe the design capacity, a test pile is constructed and estimated load is given upon the designed pile. There are three kinds of static pile load testing. 1. Compression pile load test. 2. Tension pile load test. 3. Lateral pile load test. A lobal Journal of Researches in Engineering V olume XVI Issue IV Ve rsion I 41 Year 201 E
In contrast, pile-supported foundations transmit design loads into the adjacent soil mass through pile friction, end bearing, or both. This chapter addresses footing foundations. Pile foundations are covered in Chapter 5, Pile Foundations-General. Each individual footing foundation must be sized so that the maximum soil-bearing pressure does not exceed the allowable soil bearing capacity of .
The pile driving analyzer (PDA) was developed in the 1970's as a method to directly measure dynamic pile response during driving. As the name implies, it was developed to analyze pile driving and evaluate pile driveability, including the range of stresses imparted to the pile, hammer efficiency, etc.
4. The next step is the design of pile-groups, which settlements are larger than that obtained for a single pile. However this research evalu-ates a single pile, which can be considered as a first step in design of pile foundation. Settle-ment of a single pile is a prerequisite for esti-mation of pile-group settlements (either from
PHC pile, with a diameter of 1000mm and a wall thickness of 130mm , is adopted for the wall pile frame structure of this project. The pipe pile prefabrication is carried out by dalian prefabrication factory. The parameters of the PHC prestressed concrete pipe pile are as follows: Table 1 parameters of PHC pile of wall pile frame structure .
2.2 High-Strain Dynamic Pile Testing. 2.2.1 The contractor shall perform dynamic pile testing at the locations and frequency required in accordance with section 4.0 of this special provision. 2.2.2 Dynamic pile testing involves monitoring the response of a pile subjected to heavy impact applied by the pile hammer at the pile head.
Schiavo ex rel. Schiavo, _ F.3d _, 2005 WL 648897 (11th Cir. Mar. 23, 2005) (Schiavo I), stay denied, _ S. Ct. _, 2005 WL 672685 (Mar. 24, 2005). After that appeal was taken, the plaintiffs filed an amended complaint on March 22, 2005, adding four more counts, and a second amended complaint on March 24, 2005, adding a fifth count. On the basis of the claims contained in those new .
