Interpretation of failure load tests on micropiles in Apline Soils Bellato

7/25/2019 Interpretation of failure load tests on micropiles in Apline Soils Bellato

1/14

RIVISTA ITALIANA DI GEOTECNICA 1/2013

SummaryThis paper is concerned with the execution and interpretation of load tests on small diameter piles, commonly referred to as mi-

cropiles, drilled and grouted under gravity only into highly heterogeneous soils. These soils, forming the slopes of many areas in the

Italian Alpine Region, are composed of a chaotic and erratic mixture of gravel, sand with silt and clay including cobbles and boulders

and can be often poorly characterized, due to the difficulty in performing laboratory tests or in-situ tests, with the exception of the

classical dynamic penetration test. Nevertheless, these soils show significantly high particle interlocking and dilative mechanical

response under shear, thus providing both high base resistance and shaft friction at the relatively low overburden stress surrounding

the micropile, even if the latter is grouted without any additional grouting pressure. As a consequence of that, the micropile design

using customary approaches leads frequently to a much conservative estimate of the vertical limit load. To improve the design of mi-

cropiles in such soil deposits, an experimental test site, located in the Northeastern Italian Alps, has been selected, where pile tensile

and compressive load tests up to failure have been performed under controlled conditions. On the basis of the results of these tests,

the reliability of the most common micropile calculation methods is discussed.

Key words:Drilled piles, Micropiles, Heterogeneous soils, Pile Load Test

* Department ICEA, University of Padova** Venetian Transportation Authority, Sedico*** Department ICEA, University of Padova

Diego Bellato,* Sandro DAgostini,** Paolo Simonini***

Introduction

The behavior of any drilled micropile is strong-ly influenced by the construction technique adopt-ed to install the pile into the soil and especially bythe grouting procedure. According to the US Fed-eral Highway Administration [FWHA, 2005], drilledmicropiles are classified into types from A to D,primarily on the basis of grouting type and pressure[BRUCE et al., 1997]. Type A micropile is filled withgrout introduced into the drilled hole under grav-ity only through a tremie pipe lowered to the bot-tom of the borehole and gradually raised; type B isinstalled by grouting under low pressure the drilledhole using the temporary steel casing; type C is real-ized by injecting the grout via a sleeved grout pipe

without the use of a packer at a pressure of at least1 MPa; type D is created by forcing the grout underhigh pressure to compress the soil around the shaft.

A packer may be used inside the sleeved pipe to treatseveral times selected layers. Types C and D requirea primary low strength neat cement grout placed un-der gravity head.

Type A micropile, less expensive and simpler to

be constructed, is generally used in many civil en-gineering works in the Italian mountainous areas,where the subsoil condition is quite complex, dueto the relevant heterogeneity of soils usually formedby a chaotic and erratic mixture of gravel, sand withsilt and clay including cobbles and boulders. In thiscase, the geotechnical characterization is often verypoor and no in situ tests and/or laboratory tests arenormally carried out, with the exception of classicalSPT or DPSH [CESTARI, 2005; EN 1997-2:2007], thelatter being commonly used in the Alpine Regions.

Despite the recent improvement in pile design,the estimate of the limit load for gravity grouted mi-cropiles in heterogeneous soil conditions involves arelative high level of uncertainty.

To improve the design of such type of micropile,an experimental site located in the Northeastern Ital-ian Alps, namely the Spert Test Site (STS), was select-ed in order to install several micropiles on which loadtests, both in compression and tension, could be per-formed up to the failure of the micropile-soil system.

On the basis of the experimental results ob-tained from the load tests presented in this paper,the reliability of some calculation methods custom-arily used to design micropiles in highly heteroge-

neous soils such as the classical and method orthat proposed by BUSTAMANTE and DOIX [1985] arediscussed.

Interpretation of failure load tests on micropiles inheterogeneous Alpine soils


2/14

RIVISTA ITALIANA DI GEOTECNICA

4 BELLATO- DAGOSTINI- SIMONINI

Estimate of the limit load of micropiles

Interpretation of experimental load displacement curves

Several approaches exist to estimate the limitload of micropiles on the basis of the experimentalload-displacement curves obtained from in-situ axialload tests.

One of the most common and easy to implementis that proposed by Eurocode 7 [EN1997/1], whichrecommends to assume as ultimate failure load thatcorresponding to a displacement of the pile headw = 0.1D (D = pile diameter).

Many other methods have been developed in thepast. A brief introduction to the most used criteriafor the estimate of the ultimate load from the inter-pretation of experimental load-displacement curvesof axial compression load tests is reported hereafter.

Each interpretation procedure is identified by theresearchers who contributed to its formulation. DAVIDSSON[1972] proposes as limit load that lo-

ad corresponding to the displacement whichexceeds the elastic compression of the pile by a

value of 4 mm plus a factor equal to the diameterof the pile divided by 120.

DEBEER[1967] considers the experimental datain a double logarithmic load-displacement plotand defines the limit load of the pile as that de-termined by the intersection of the two approxi-mately straight lines giving the best fit of the da-

ta. BRINCH-HANSEN [1963] states that the limit load

of the pile is equal to the load that gives twicethe displacement of the pile head as obtainedfor 90% of that load (known as Brinch-Hansens90% criterion).

BRINCH-HANSEN[1963] also proposes an 80%crite-riondefining the ultimate load as the load givingfour times the displacement of the pile head asobtained for 80% of that load. The limit load canbe estimated by plotting the experimental datain a diagram having the pile head displacementsalong the x-axis and the ratio between the squa-re root of the displacement and the applied loadalong the y-axis. In this way the data will lie ap-proximately along a straight line of equation:

(1)

where Pis the applied load, C1is the slope of theline and C2the intercept with the y-axis. The ul-timate load can finally be computed by:

(2)

VANDER VEEN [1953] supposes different va-lues of limit loads and values computed fromln (1 - P/Qu) are plotted against the displace-

ment. The correct value is that for which the plotbecomes approximately a straight line.

FULLERANDHOY[1970] assume as limit load thetest load for where the load-displacement curveis sloping 0,14 mm/kN.

BUTLERANDHOY[1977] define Quas the load atthe intersection of the tangent sloping 0,14 mm/kN and the tangent to the initial straight portionof the curve.

CHIN[1970; 1971] proposes the well known in-verse slope method, the only between the pre-viously described criteria that can be used for theprediction of the limit load even when the failu-re of the pile-soil system does not occur.

The method assumes that the load-displacementcurve when the load approaches the failure sta-te is of hyperbolic shape of the type given by thefollowing equation:

(3)

After some initial variation, the experimental da-ta, reported in a diagram w/Pversus w,approacha straight line of equationy = ax + b. Quis givenby the inverse of the slope of this line, i.e. Qu=1/a.However, the hyperbolic function fitting typical-ly provides an overestimate of the real ultimateload of the pile [FELLENIUS, 1980]. For this rea-son, Quis customarily assumed 90% of the asym-

ptotic value [MANDOLINI, 1995]. The CEMSET method is a parametric formula-tion based on hyperbolic functions to describeboth individual shaft and base performance of asingle pile. The criterion accounts for the elasticdeformation of the pile and relies on the studiesof RANDOLPHand WROTH[1978; 1982] on the de-formation behaviour of vertically loaded piles.The method allows to reconstruct the whole lo-ad-displacement curve fitting the experimentaldata and, consequently, to determine the correctultimate load.

The parameters used to represent the load-dis-placement behavior of the pile are conventionalelastic soil properties, basic geometries of the sy-stem, and the supposed base and shaft ultimateloads. Among these parameters, the shaft flexibi-lity factor, Ms, may be regarded as the most diffi-cult to be determined.

Estimate of the limit load of micropiles from static formu-lae

In absence of experimental load-displacement

curves, the bearing capacity of micropiles can ge-nerally be predicted by means of two analytical me-thods, namely:


3/14

5

GENNAIO - MARZO 2013

INTERPRETATIONOFFAILURELOADTESTSONMICROPILESINHETEROGENEOUSALPINESOILS

the BUSTAMANTEand DOIXmethod [1985]; the common method based on formulae used for

medium and large diameter bored piles, knownas method or method, depending on thetype of soil surrounding the pile.The BUSTAMANTE and DOIX method completely

neglects the tip bearing capacity of the micropileand proposes to calculate the skin friction mobilizedalong the shaft as:

(4)

whereDsiis the effective diameter of the micropileafter the grouting phase, Lsi is the bond length re-lative to each i-th soil layer surrounding the shaft,and qsiis the skin friction mobilized at the soil-pileinterface, function of NSPTand of the type of micro-pile, i.e. IGU (Injection Globale et Unitaire) or IRS

(Injection Rptitive et Slective) [BUSTAMANTEandDOIX, 1985].

In the and method, the micropile lateralresistance in cohesive and cohesionless soils is calcu-lated using the following equations:

(5)

(6)

where D and Lare, respectively, the diameter and

the length of the pile, zis the depth, limis the mo-bilized unit shaft friction resistance, suis the undrai-ned shear strength of the cohesive soil surroundingthe shaft, is an empirical factor depending on su

[VIGGIANI, 1993], v is the vertical effective stress,and = Ktang , in which Kis a factor representingthe horizontal thrust coefficient (depending on thegrouting technique) and is the interface frictionangle between the pile and the soil.

The tip bearing capacity in granular soils can becalculated as:

(7)

wherev,pis the vertical effective stress at the pile tipand Nqis a bearing capacity factor, which is related tothe angle of friction of the soil surrounding the piletip. For bored pile it is often suggested to use a Nqaccording to the BEREZANTSEVs formulation [1961].

It should be kept in mind that the end bearingcapacity of micropiles is customarily neglected inlimit load calculations (e.g.ARMOURet al.,1997), be-

ing the tip settlement usually not sufficient to fullymobilize the whole base resistance.

Soil conditions at STS

Figure 1 shows the location of the geotechni-cal investigations at the STS, consisting of boreholesand undisturbed sampling, Standard PenetrationTests, Dynamic Penetration Tests, and Cone Penetra-tion Tests. On the same figure, the positions of themicropiles subjected to load test are also depicted.

Figure 2 sketches the soil profile determinedon the basis of the geotechnical investigation per-formed at the site. No water table was detected atthe STS.

Fig. 1 Location of site investigations and pile load tests.

Fig. 1 Posizione dei punti di investigazione e dei pali di prova.


4/14



From the ground level to the depth of approxi-mately 6 m, the soil is composed of a mixture of a pre-dominant silt fraction mixed with sand and clay. Sub-layers show the local presence of some gravel, whichbecomes particularly important between 4 m and 5m below g.l. Atterberg limits of the fine grained frac-tion denote low plasticity silty clays in the upper partof the layer and low plasticity silts in the lower one,respectively.

Between 6 and 8 m below g.l., gravel fraction be-comes significant (beyond 50%), thus controllingthe mechanical response of the soil. To note that thegravel particles are characterized by an angular orsubangular shape, leading also to a high degree ofparticle interlocking.

SPT value remains approximately constant(NSPT = 10) up to 5 m below g.l. and rises to 50 andover at higher depths. Static cone (standard CPT)

was pushed down only into the silty formation,

whereas DPSH allowed to investigate deeper theground up to 9 m. It is worth noticing the continu-ous and regular increase of NDPSHwith depth.

Triaxial unconsolidated and undrained com-pression tests, carried out on previously saturatedspecimens trimmed from undisturbed samples tak-en between 1 m and 3 m below g.l., provided an un-drained compressive strength, su, ranging approxi-mately between 100 and 110 kPa. Since the finefraction in the upper layer exceeds 30% [OMINE etal.,1994; COLA,2002] and owing to the high rate ofstress application throughout the pile load tests, anundrained response for the silty-clayey mixture wastentatively assumed. Nevertheless, it should be em-phasized that a partial-saturation soil condition dueto presence of a deeper groundwater level would bea more reasonable assumption in this case, leadingto a different strength from that reported above andmeasured in the laboratory in saturated conditions.

Dynamic Penetration Tests (both SPT andDPSH) were used to estimate the value of frictionangle of the deep sandy and silty gravel of approxi-

mately 42. In particular, NDPSHwas first convertedinto NSPT[LACROIXand HORN, 1973 also reportedin Fig. 2] and then introduced in empirical relation-

Fig. 2 Soil profile and soil properties at the STS.

Fig. 2 Profilo del sottosuolo e principali propriet del terreno presso il sito sperimentale di Spert.


5/14

7



ships available from the literature and relating NSPT

with the angle of shearing resistance.

Micropile installation

The drill rig used at the STS was a hydraulic rota-ry unit with a drilling diameter of 200 mm (198 mmrotary cutter). The typical drilling technique adoptedin the Alpine Region is the so called rotary concen-tric percussive duplex [FHWA, 2005], whereby thedrill rods inside the casing and the casing itself are si-multaneously percussed, rotated, and advanced. Thismethod guarantees against the instability of the dril-led hole. Compressed air was used for cleaning andremoving of the spoil material during drilling.

After reaching the maximum depth, a tubularsteel reinforcement (diameter 127 mm, thickness 10mm) was inserted. Grouting was then performed bygravity only to fill the whole cavity from the bottomof the hole. An additional 32 mm diameter Dywidagbar was placed inside the previous steel tubular rein-forcement to carry out the subsequent tensile loadtests.

Figure 1 shows the locations of the 200 mm dia-meter micropiles, whose geometrical characteristics

are reported in table I. Three micropiles were tested,respectively, under compressive loads and three un-der tensile loads.

In order to measure the base resistance only, aspecial micropile - labeled C3 - was designed andconstructed. To this purpose, the tubular reinforce-ment was enveloped with a PVC pipe for a length Y =627 cm to avoid any contact between the reinforce-ment itself and the surrounding soil. Then a prede-fined amount of grout was introduced in the tubularreinforcement previously pierced at the toe with ablowtorch to create a tip of the micropile of length Zof about 38 cm, reasonably assuming no loss of groutoccurring in the surrounding layers due to the stiffnature of the deposits. Once the setting of the pri-mary grout was completed, the space inside the ste-el tubular reinforcement was filled with additionalgrout. Figures 3 and 4 sketch, respectively, a three-dimensional longitudinal section of the standard pi-les and of the special micropile C3 installed into theground at the STS.

Testing procedure

Load tests were carried out in accordance with

the specifications provided by ASTM D-1143 [ASTM,1994] and ASTM D-3689 [ASTM, 1995]. The reac-tion frame in compression load tests was anchored

Fig. 3 Section of standard grouted micropiles.

Fig. 3 Sezione di un micropalo colato standard.

Fig. 4 Section of special micropile C3, showing the tech-

nical solution used to measure tip resistance only.

Fig. 4 Sezione del micropalo speciale C3, con riportata la

soluzione tecnica adottata per la determinazione della resistenza

di punta.


6/14



to a couple of surrounding micropiles realized speci-fically to this purpose.

The compression or tension load was applied by

means of a hydraulic jack to the steel tubular rein-forcement or to the tendon, respectively, where-as the vertical displacement of the tested micropilehead was measured by three dial gauges accordingto ASTM D-1143.

The loading procedure consisted in the appli-cation of load increments equal to about 1/8 of thepile ultimate load Qu, estimated by rough prelimi-nary computations. Once a load equal to 100% Qu

was applied without the collapse occurrence, theload increments were reduced to 5% Quup to thefailure, which is generally associated with the de-

velopment of relevant displacements, here assu-med greater than 25%D. Furthermore, load incre-ments of 5% Qu were also applied during the lo-ading procedure whenever unexpected behaviors

were observed.

Each load increment was maintained until thepenetration/rise rate of the pile head reached 8m/min for a maximum time span of 30 min. Load-ing-unloading cycles were carried out at different lo-ad levels in all the tests, but mainly in the correspon-dence of 50%, 75%, and 100% Qu.

Test results

Figures 5a and 5b present the load-displacementcurves from tests under compressive and tensile lo-ad, respectively. From a general point of view, the re-sponse is characterized by a progressive hardeningbehaviour, with a gradual accumulation of irrecove-rable displacements since the very early stage of thetests. Additional features to note are: the response of micropile C1 does not show cle-

arly the failure condition, but it is characterizedby a continuous hardening behaviour. This hasbeen associated with a probable significant in-fluence on the axial response caused by impor-tant volumes of cementitious slurry loss in thesubsoil during the grouting of both the reactionpiles necessary to anchor the frame structureused to test micropile C1;

the response of micropile C2 shows a more pro-nounced yielding coupled with the occurrenceof large displacements even under small load in-crements, thus confirming the mobilization of

full shaft friction and base resistance; the response of special micropile C3 (realized tomeasure base resistance only) is characterized bya sudden break in the load-displacement curvedue to the overcoming of the limit bond stress

Fig. 5a Load-displacement curves from compression load

tests.Fig. 5a Curve carico - spostamento ottenute dalle prove di

carico in compressione.

Fig. 5b Load-displacement curves from tensile load

tests.Fig. 5b Curve carico - spostamento ottenute dalle prove di

carico a trazione.

MICROPILE

L [cm] X [cm] Y [cm] Z [cm]C1 705 40 0 665

C2 705 45 0 660

C3 705 40 627 38

T1 689 35 0 654

T2 596 34 0 562

T3 705 44 0 661

Tab. I Geometrical characteristics of the micropiles in-

stalled at the STS.

Tab. I Caratteristiche geometriche dei micropali installati nel

sito sperimentale di Spert.


7/14

9



between the grout surrounding the pile tip andthe tubular steel reinforcement;

the load-displacement behaviour of tensile pilesT2 and T3 is characterized by a gradual pull-outat increasing load;

micropile T1 shows a slightly stiffer response

followed by a sudden pull-out of the pile fromthe soil. This could be probably due to an unex-pected failure of some local blockages between

the pile shaft and the surrounding heteroge-neous material.From the load-displacement curves depicted in

figures 5a and 5b, it was tentatively possible to evalua-te the ultimate load relative to each tested micropile.

Assuming a limit load corresponding to a displace-ment of the pile head equal to 10% of the micropile

diameter (for bored pile the load at w/D= 0.1 is typi-cally accepted as limit load according to EN1997/1),the micropile reference ultimate loads in both com-

Fig. 6 Hyperbolic fitting of load tests in compression.

Fig. 6 Interpolazione iperbolica delle curve di carico in

compressione.

Fig. 7 Hyperbolic fitting of load tests in tension.

Fig. 7 Interpolazione iperbolica delle curve di carico a trazione.

Fig. 8 Comparison between limit loads estimated from load-displacement curve interpretation and load tests.Fig. 8 Confronto tra i carichi limite stimati a partire dallinterpretazione delle curve carico - spostamento e misurati al termine delle prove

carico.


8/14



pression and tension were calculated and comparedwith the theoretical calculations described above.

Interpretation of load-displacement curves

Ultimate failure load in compression and tension

To interpret the experimental load-displace-ment curves of both groups of piles tested in com-pression and tension, a hyperbolic function was se-lected as suggested by CHIN[1970]. The result of thebest fitting is shown in figures 6 and 7 for piles undercompression and tension, respectively.

It is worth noting that at large displacements,the experimental behavior can be well described bythe hyperbolic curve, thus allowing the estimate ofthe limit load represented by the inverse slope of thecorresponding straight line. Notwithstanding, thehyperbolic function fitting produced an overestima-te of the actual failure load and, therefore, a 90% ofthe asymptotic limit value was assumed according toMANDOLINI[1995].

In the case of the tensile load tests, the ultimateload corresponding to the pull-out of the piles fromthe soil is clearly appreciable from the experimentalcurves and can be easily compared with that provi-ded by the hyperbolic interpretation. Unfortunately,for piles in compression, the unique curve leading toa proper estimate of the limit load is that referred topile C2 (Fig. 5a).

In addition to the modified Chin method (90%

of the failure load determined with the Chin proce-dure), the load-displacement curves obtained fromthe compression and tension tests were interpreted

with other methods, namely those of Davisson, DeBeer, Brinch Hansen, Vander Veen, Fuller and Hoy,Butler and Hoy, and CEMSET [DAVISSON,1972; DEBEER,1967; DEBEERandWALLAYS,1972; BRINCHHAN-SEN, 1963;VANDERVEEN, 1953;FULLERand HOY, 1970;BUTLERand HOY, 1977; FLEMING, 1992]. The limit lo-ads obtained using the above methods were compa-

red with the ultimate failure loads estimated fromthe in situ load test results. The outcome of the cal-culations is reported in figure 8.

For micropiles tested in tension, all the abovemethods provide values in agreement with the field

Fig. 9 CEMSET fitting procedure of the experimentalload-settlement curve of micropile C2.

Fig. 9 Procedura CEMSET per linterpretazione della curva

carico - spostamento relativa al micropalo C2.

Fig. 10 CEMSET fitting procedure of the experimentalload-settlement curve of micropile T3.

Fig. 10 Procedura CEMSET per linterpretazione della curva

carico - spostamento relativa al micropalo T3.

Fig. 11 Double logarithmic plot of the uplift displace-

ments during time of micropile T3.Fig. 11 Grafico doppio-logaritmico del sollevamento misurato

nel tempo per il micropalo T3.


9/14

11



results, with the exception of the De Beers and theButler and Hoys ones, that are too conservative [FEL-LENIUS,1980].

For micropile C2 it is possible to observe an in-creased variability in limit load values, being theCEMSET, the modified Chins, and the Brinch-Hansens procedures the most reliable methods. Itis worthwhile to note that the best interpretation

method seems to be the CEMSET one, that provi-des both the limit load and the entire load-displa-cement curve, the latter based on best fitting opti-mization. In particular, the results of this procedu-re are presented in figures 9 and 10 for micropilesC2 and T3.

Time-dependent effect on micropile pull-out

It is interesting to analyze the time-dependentbehavior of micropiles during the pull-out tests. Fi-gure 11 shows, for instance, the log-log plot of themicropile T3 head displacement versus time, wherethe typical time-dependent deformation of the mi-cropile-soil system at constant load (creep) can benoticed. For the last three load increments, the tren-ds are approximately parallel showing a linear de-formation behavior characterized by almost the sa-me slope translated towards higher displacement va-lues with the increase of the applied load. The defor-mation proceeds under the constant maximum load(350 kN) up to a yielding point (approximately 9-10mm) beyond which failure suddenly occurs. The di-mensionless displacement w/Dis plotted versus time

in figure 12, from which it can be seen that failurein tension took place at a relative displacement ofaround 3-4% of the pile diameter. To note the ra-

pidly increasing trend of w/Dwith time approachingthe pull-out of the pile.

Figures 13 shows the displacement rate vs. timeat different load levels close to the maximum pull-out load. A progressive reduction of the displace-ment rate with time and a stabilization of the dela-

yed deformation over a long period, namely a sortof primary creep phase [FABRE and PELLET, 2006],

can be observed. Under the maximum load of 350kN, an initial primary creep phase characterized bya strain rate reduction is followed by a sudden in-crease in the strain rate up to the failure state (ter-tiary creep phase) without a clearly detectable in-termediate secondary creep phase. This is probablydue to the diffuse and progressive damage of thesoil matrix occurred during the previous load in-crements.

Shaft resistance

Based on the micropile ultimate loads, the mo-bilized unit shaft resistance at failure was estima-ted. Assuming tentatively no normal stress variationagainst the shaft durin the pull-out tests and shearstress uniformity along the whole lenght of the pile

with an effective diameter of 200 mm, the mobilizedunit shaft resistance, lim, was found to range betwe-en 84 and 93 kPa.

The assumption about the effective diameter ofthe micropiles was confirmed at the end of the loadtests, when the soil surrounding piles was removed.In all the cases, the measured diameter was appro-

ximately equal to the nominal one. This is of coursedue to the grouting technique not forcing the groutagainst the wall of the hole. Figures 14 and 15 show,

Fig. 12 Accumulation of displacement during tensile

load test carried out on micropile T3.Fig. 12 Accumulo di deformazione plastica durante il test di

carico a trazione eseguito sul micropalo T3.

Fig. 13 Displacement rate from load-displacement curve

of micropile T3.Fig. 13 Velocit di deformazione ottenuta dalla curva carico -

spostamento relativa al micropalo T3.


10/14


11/14

13



Figure 16 shows that the end bearing resistan-ce of the micropile C3 increases rapidly providing acontribution of around 45% of the total bearing ca-pacity at a head displacement of about 10% of thepile diameter. This can be attributed to the excel-lent mechanical characteristics of the underlying lay-er (composed of very dense gravel) in which the pilebase was embedded.

The contribution of the shaft resistance has beenfinally obtained by subtracting the end bearing resi-stance curve from the reconstructed load-displace-ment curve shown in figure 16. The resulting hyper-bolic trend provided higher loads than those observedat comparable displacements in the curves relative tothe micropiles tested under tension, ranging between1.20 times those measured from pile T2 and 1.05 timesthose collected from pile T3, thus giving a ratio of theshaft resistance under tension over that under com-

pression lower than unity. Despite some difference inthe range of variation, this ratio seems to be in reaso-nable accordance with the observations of DENICOLAand RANDOLPH[1999] and HANandYE[2006].

Failure of the end bearing micropile C3

The sudden brittle failure observed in micropi-le C3, characterized by a heavy dull sound occurredduring the static load test, suggested that the beha-

vior of the pile at collapse was presumably due to a

slippage mechanism controlled not by the soil-pile

interaction system, typically denoted by a ductile re-sponse, but rather by the interface between the steelreinforcement and the grout.

Laboratory tests were carried out on the groutused to form the micropiles at the STS to investiga-te the shearing strength (unconfined compressivestrength) of the cementitious mortar. In accordan-ce with CEB-FIP Model Code [1990], the limit bondstress between the grout surrounding the pile tipand the smooth reinforcing steel tube can be com-puted as follows:

(8)

Considering a characteristic compressionstrength value, fck, of 32,3 MPa, the axial force ne-cessary to cause the reaching of the limit bond stressalong Z (see Fig. 4) is approximately equal to 260

kN, that is comparable with the ultimate load measu-red at the end of the load test. Therefore, it is presu-mable that this limit load was due to the overcomingof the allowable shear stress between the grout andthe pile steel reinforcement.

Evaluation of applicability of analytical methodsto predict bearing capacity of micropiles

To evaluate the applicability of analytical me-thods to predict the bearing capacity of micropiles,

two methods were here considered, namely:

Fig. 17 Comparison between ultimate load bearing capacities calculated from static formulae using the factor proposed

byVIGGIANI[1993] and those derived from load-displacement curve interpretation.

Fig. 17 Confronto tra le capacit portanti ultime calcolate da formule statiche utilizzando i valori del coefficiente a proposto da VIGGIANI[1993] e quelle derivate dallinterpretazione delle curve carico - spostamento.


12/14



Fig. 18 Comparison between ultimate load bearing capacities calculated from static formulae using the experimentally de-

termined factor and those derived from load-displacement curve interpretation.

Fig. 18 Confronto tra le capacit portanti ultime calcolate da formule statiche utilizzando i valori del coefficiente determinati

sperimentalmente e quelle derivate dallinterpretazione delle curve carico - spostamento.

theBUSTAMANTEand DOIXmethod [1985]; the common method based on static formulae

known as method or method, depend-ing on the type of soil surrounding the pile.

The calculation of the ultimate bearing capacityof the six micropiles at the STS with both the abovemethods (assuming an effective diameter equal to200 mm) did not provide results in agreement withthe failure loads measured during the compressiveand tensile load tests and with the estimated valuescomputed by the 90% Chin criterion and the CEM-SET method. The outcomes are presented in figure17. The differences obtained may be tentatively ex-plained by the following considerations.

BUSTAMANTEand DOIX[1985] proposed differentunitary skin friction curves depending on the typeof soil in which the micropile is installed and on twogrouting techniques, i.e. IGU and IRS. A relation-ship for the grouting procedure by gravity head on-ly was not provided by the authors. Therefore, theassumption of the same skin friction suggested forIGU technique caused an overestimation of the uni-tary lateral resistance along the shaft.

In addition, the Bustamante and Doix methodcompletely neglects the base resistance providing atotal bearing capacity lower than that measured inthe micropiles tested under compression load.

From the experimental findings, in fact, it waspossible to notice that the typical assumption of neg-

ligible tip resistance cannot be properly consideredfor the micropiles installed at the STS, because ofthe very stiff response of the soil underneath the pile

along with the excellent mechanical properties ofthe gravelly layer in which the pile tip was embed-ded.

The and method underestimated both the

tip and the shaft resistance. This may be due mainlyto two reasons.Firstly, the end bearing capacity factor, Nq, as-

sumed in the calculation according to BEREZANTSEV[1961], theoretically corresponds to a critical tipdisplacement of 5% of the pile diameter (usuallyindicated as Nq*). However, in this case, the aboveassumption yielded to an ultimate base load lowerthan that obtainable at the same w/Dfrom figure 16.In order to compute an ultimate end bearing resis-tance referred to a w/D= 0.1 [EN1997/1], a bearingcapacity factor Nq= 1,4 Nq*has been proposed.

Secondly, the shaft resistance was calculatedwith a rather small factor, in accordance with usu-al geotechnical design recommendations for boredpiles [VIGGIANI, 1993]. Nevertheless, the experimen-tal findings from the pull-out tests showed that a factor equals to 0.9-1.0, estimated assuming an und-rained soil behavior, seems to be give more reliableresults with respect to those proposed by VIGGIANI[1993]. On the other hand, it should be emphasizedthat the hypothesis of undrained conditions couldbe misleading in this case, as the in situ material ismore correctly characterized by partial saturationconditions, that in turn involve a variation in soil

strength from the assumed values.In order to determine the shaft resistance of mi-

cropiles grouted under gravity head only in granu-


13/14

15



lar soils, the factor was chosen equal to K0tan,according to common design recommendations forbored piles [e.g. FHWA, 2010].

Thus, from the above considerations, the ulti-mate bearing capacities of the micropiles were recal-culated. These new values fit relatively well the fail-ure loads obtained from the experimental load-dis-placements curves, as presented in figure 18, whichshows, however, still some difference for micropileC3, whose tip limit resistance was not achieved dur-ing the static load test, but only predicted on the ba-sis of the measured data.

Conclusions

The results and the interpretation of axial loadtests in tension and in compression performed on

six micropiles installed in heterogeneous soils typi-cal of the Alpine mountainous area were presentedand discussed in this paper.

After the preliminary site characterization carri-ed out by means of in situ and laboratory tests, threecompressive and three tensile load tests were execut-ed on simply gravity filled micropiles of 200 mm indiameter.

Through the construction of a special micropilewith minimized contribution of skin friction, it waspossible to interpret the load distribution betweenshaft and base resistance of the micropiles. The re-

sults emphasized that the usual assumption of ne-glecting tip resistance in micropile design can beconsidered sometimes too conservative in these geo-technical contexts.

The reliability of a number of methods for theinterpretation of load-displacement curves obtai-ned from pile load tests was investigated, and theCEMSET procedure proved to be the most efficientfor the estimate of the ultimate bearing capacity.

Finally, the outcomes of the micropile designapproach introduced by Bustamante and Doix we-re compared with the classic method, customa-rily adopted in bored pile design, and the load testresults. From the comparison, it was found that insome particular conditions, such as that discussed inthe paper, significant underestimation of the effecti-

ve limit load may lead to an excessively conservativemicropile design.

Bibliografia

ARMOUR T., GRONECKP., KEELEY J., SHARMA S.(1997) Micropile Design and Construction Guidelines. Imple-mentation Manual. Federal Highway Administra-

tion, FHWA-SA-97-070.ASTM (1994) D1143-81Standard Test Method for piles

under static axial compressive load.

ASTM (1995) D3689-90 Standard Test Method for in-dividual piles under static axial tensile load.

BELLATOD., SIMONINIP., DAGOSTINIS. (2011) Com-portamento di micropali in terreni eterogenei. Proc. XX-IV Convegno Nazionale di Geotecnica, Napoli, 22-24 giugno 2011, pp. 61-68.

BEREZANTSEV V.G., KHRISTOFOROV V., GOLUBKOV V.(1961) Load-bearing capacity and deformation of piled

foundation.Proc. 5th ICSMGE, Paris, 1961, vol. II,pp. 11-15.

BRINCHHANSEN, J.(1963) Discussion: Hyperbolic stress-strain response. Cohesive soils. ASCE, J. SMFD, 89, n.4, pp. 241-242.

BROWND.A., TURNERJ.P., CASTELLIR.J. (2010) Drilledshafts: construction procedures and LRFD Design Meth-ods. Federal Highway Administration, FHWA-NHI-10-016.

BRUCE D.A., DIMILLIO A.F., JURAN I. (1997) Micro-

piles:the state of practice. Part 1: characteristics, defini-tions and classifications. Ground Improvement, 1,n.1, pp. 25-35.

BUSTAMANTE M., DOIX B. (1985) Une mthode pourle calcul des tirants et des micropieux injects. BulletinLiaison Laboratoire des Ponts et Chausses, n. 140,pp. 75-95.

BUTLERH.D., HOYH.E. (1977) Users manual for theTexas quick-load method for foundation load testing.FHWA, Washington, 59 pp.

CEB-FIP COMITE EURO-INTERNATIONAL DU BETON(1993) CEB-FIP Model Code 1990. Thomas Telford

Services Ltd.CEN EUROPEAN COMMITTEE FOR STANDARDIZATION(2004) EN 1997/1 Eurocodice 7. Part 1 - Geotechnical

Design, General rules. Bruxelles.CESTARIF. (2005) Prove geotecniche in sito. Ed. Geo-

graph, Segrate, Milano, 416 pp.CHIN F.K. (1970) Estimation of the ultimate load ofpiles not carried to failure. Proc. 2nd Southeast AsianConf. on Soil Engng., Singapore 1970, pp. 81- 90.

COLAS.(2002) On modeling the behaviour of mlang-es. Proc. 5th European Conference on NumericalMethods in Geotechnical Engineering, Paris 2002,pp. 143-148.

DAVISSONM.T. (1972) High capacity piles. Proceed-ings, Lecture Series, Innovations in FoundationConstruction, ASCE, Illinois Section, 52 pp.

DEBEERE.E.(1967) Proefondervindelijke bijdrage totde studie van het grensdraag vermogen van zand onder

funderingen op staal. Tijdschrift der Openbar Wer-ken van Belgie, 1967, n. 6, pp. 481-506; 1968: n. 1,pp. 41-88; n. 4, pp. 321-360; n. 5, pp- 395-442; n. 6,pp. 495-522.

DENICOLAA., RANDOLPHM.F. (1999) Centrifuge mod-eling of pipe piles in sand under axial loads. Gotech-nique, 49, n.3, pp. 295-318.

FELLENIUSB.H.(1980) The analysis of results from rou-tine pile load tests. Ground Engineering, 13, n. 6, pp.19-31.


14/14



FLEMINGW.G.K.(1992) A new method for single pilesettlement prediction and analysis. Gotecnique, 42, n.3, pp. 411- 425.

FULLERR.M., HOYH.E.(1970) Pile load tests includ-ing quick-load test method, conventional methods and in-

terpretations. HRB 333, pp. 78-86.HANJ. YES.L. (2006) A field study on the behaviour

of micropiles in clay under compression or tension. Ca-nadian Geotechnical Journal, 43, n. 1, pp. 19-29.

JEONS.S. (2004) Interpretation of load tests on mini-piles. Proc. ICE, Geotechnical Engineering, 157, n.2, pp. 85-90.

JURAN I., BRUCE D.A., DIMILLIO A., BENSLIMANE A.(1999) Micropiles: the state of practice. Part 2: designof single micropiles and groups and networks of micro-

piles. Ground Improvement, 3, n. 3, pp. 89-110.KORECKW. (1978) Small diameter bored injection piles.

Ground Engineering, 11, n. 4, pp. 14-29.

LACROIXY., HORNH. (1973) Direct Determination andIndirect Evaluation of Relative Density and Its Use on

Earthwork Construction Projects. Evaluation of Rela-tive Density and Its Role in Geotechnical ProjectsInvolving Cohesionless Soils, ASTM Special Techni-cal Publication 523, pp. 251-280.

MANDOLINIA., 1995 Prove di carico su pali di fondazio-ne. Ed. Hevelius, Benevento, 64 pp.

RANDOLPH M.F., WROTHC.P. (1982) Recent develop-ments in understanding the axial capacity of piles in clay.Ground Engineering, 15, n.7.

RANDOLPHM.F., WROTHC.P. (1978) Analysis of the de-

formation of vertically loaded piles. J. Geotech. Engng.Div., ASCE, 104, n.12, p. 1465.RUSSOG. (2004)Full-scale load tests on instrumented

micropiles. Proc. ICE, Geotechnical Engineering,157, n. 3, pp. 127-135.

SABATINIP.J., TANYUB., ARMOURT., GRONECKP., KEELEY

J.(2005) Micropile Design and Construction Refer-ence Manual. FHWA-NHI-05-039.

SABINIG., SAPIOG. (1981) Behaviour of small diam-eter bored piles under axial load. Proc. 10th ICSMFE,Stockholm 1981, pp. 823-828.

VANDERVEENC.(1953) The bearing capacity of a pile.Proc. 3rdICSMFE, Zurich, vol. II, pp. 84-90.

VIGGIANI C. (1999) Fondazioni. Ed. Hevelius,

Benevento, 568 pp.VIGGIANIC.(1993) Further experience with auger piles

in Naples area. Proc. of the Second Int. Seminar onDeep Foundations on Bored and Auger Piles, BAPII, Van Impe (Ed.), Balkema, Rotterdam, pp. 445-458.

Interpretazione di prove di carico a

rottura su micropali installati in terreni

alpini eterogenei

Sommario

Larticolo tratta di importanti aspetti legati alla costruzione

e allinterpretazione di prove di carico condotte su pali di piccolo

diametro, comunemente detti micropali, installati mediantetrivellazione e getto a gravit in terreni fortemente eterogenei.

Queste formazioni, in genere incontrate nelle zone di pendio

delle regioni alpine italiane, sono prevalentemente costituite da

una miscela caotica e granulometricamente irregolare di ghiaia

e sabbia con limo e argilla spesso contenente ciottoli e massi.

Sono pertanto difficilmente caratterizzabili dal punto di vista

geotecnico, date le problematiche relative allesecuzione di prove di

laboratorio o in sito, ad esclusione delle classiche prove dinamiche.

Ciononostante, questi terreni mostrano un significativo grado

di mutuo incastro e una risposta meccanica a taglio fortemente

dilatante, dalla quale dipende lo sviluppo di unelevata resistenza

sia di punta sia per attrito laterale anche in micropali di tipocolato, vale a dire in configurazioni caratterizzate da un limitato

confinamento laterale. Un simile comportamento non tipicamente

descritto dagli usuali approcci di progettazione dei micropali

che, spesso, forniscono una stima troppo conservativa del carico

limite verticale. Al fine di rendere tali metodi di progettazione pi

adatti a rappresentare la realt fisica del sito, stato realizzato un

apposito campo sperimentale nelle Alpi nord-orientali nel quale

condurre diverse prove a rottura in compressione e a trazione in

condizioni controllate. Sulla base dei risultati di queste prove

stata valutata laffidabilit dei pi comuni metodi di calcolo dei

micropali.

Parole chiave:pali trivellati, micropali, terreni eterogenei, test

di carico su pali

Documents

Interpretation of failure load tests on micropiles in Apline Soils Bellato