Bellato Et Al - Interpretation of Failure Load Tests On Micropiles PDF
Bellato Et Al - Interpretation of Failure Load Tests On Micropiles PDF
Bellato Et Al - Interpretation of Failure Load Tests On Micropiles PDF
Summary
This 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
Estimate of the limit load of micropiles ment. The correct value is that for which the plot
becomes approximately a straight line.
Interpretation of experimental load displacement curves FULLER AND HOY [1970] assume as limit load the
test load for where the load-displacement curve
Several approaches exist to estimate the limit is sloping 0,14 mm/kN.
load of micropiles on the basis of the experimental BUTLER AND HOY [1977] define Qu as the load at
load-displacement curves obtained from in-situ axial the intersection of the tangent sloping 0,14 mm/
load tests. kN and the tangent to the initial straight portion
One of the most common and easy to implement of the curve.
is that proposed by Eurocode 7 [EN1997/1], which CHIN [1970; 1971] proposes the well known in-
recommends to assume as ultimate failure load that verse slope method, the only between the pre-
corresponding to a displacement of the pile head viously described criteria that can be used for the
w = 0.1 D (D = pile diameter). prediction of the limit load even when the failu-
Many other methods have been developed in the re of the pile-soil system does not occur.
past. A brief introduction to the most used criteria The method assumes that the load-displacement
for the estimate of the ultimate load from the inter- curve when the load approaches the failure sta-
pretation of experimental load-displacement curves te is of hyperbolic shape of the type given by the
of axial compression load tests is reported hereafter. following equation:
Each interpretation procedure is identified by the
researchers who contributed to its formulation. (3)
DAVIDSSON [1972] proposes as limit load that lo-
ad corresponding to the displacement which After some initial variation, the experimental da-
exceeds the elastic compression of the pile by a ta, reported in a diagram w/P versus w, approach
value of 4 mm plus a factor equal to the diameter a straight line of equation y = ax + b. Qu is given
of the pile divided by 120. by the inverse of the slope of this line, i.e. Qu =
DE BEER [1967] considers the experimental data 1/a.
in a double logarithmic load-displacement plot However, the hyperbolic function fitting typical-
and defines the limit load of the pile as that de- ly provides an overestimate of the real ultimate
termined by the intersection of the two approxi- load of the pile [FELLENIUS, 1980]. For this rea-
mately straight lines giving the best fit of the da- son, Qu is customarily assumed 90% of the asym-
ta. ptotic value [MANDOLINI, 1995].
BRINCH-HANSEN [1963] states that the limit load The CEMSET method is a parametric formula-
of the pile is equal to the load that gives twice tion based on hyperbolic functions to describe
the displacement of the pile head as obtained both individual shaft and base performance of a
for 90% of that load (known as Brinch-Hansens single pile. The criterion accounts for the elastic
90% criterion). deformation of the pile and relies on the studies
BRINCH-HANSEN [1963] also proposes an 80% crite- of RANDOLPH and WROTH [1978; 1982] on the de-
rion defining the ultimate load as the load giving formation behaviour of vertically loaded piles.
four times the displacement of the pile head as The method allows to reconstruct the whole lo-
obtained for 80% of that load. The limit load can ad-displacement curve fitting the experimental
be estimated by plotting the experimental data data and, consequently, to determine the correct
in a diagram having the pile head displacements ultimate load.
along the x-axis and the ratio between the squa- The parameters used to represent the load-dis-
re root of the displacement and the applied load placement behavior of the pile are conventional
along the y-axis. In this way the data will lie ap- elastic soil properties, basic geometries of the sy-
proximately along a straight line of equation: stem, and the supposed base and shaft ultimate
loads. Among these parameters, the shaft flexibi-
(1) lity factor, Ms, may be regarded as the most diffi-
cult to be determined.
where P is the applied load, C 1 is the slope of the
line and C 2 the intercept with the y-axis. The ul-
timate load can finally be computed by: Estimate of the limit load of micropiles from static formu-
lae
(2)
In absence of experimental load-displacement
VANDER VEEN [1953] supposes different va- curves, the bearing capacity of micropiles can ge-
lues of limit loads and values computed from nerally be predicted by means of two analytical me-
ln (1 - P/Qu) are plotted against the displace- thods, namely:
the BUSTAMANTE and DOIX method [1985]; [VIGGIANI, 1993], v is the vertical effective stress,
the common method based on formulae used for and = Ktang , in which K is a factor representing
medium and large diameter bored piles, known the horizontal thrust coefficient (depending on the
as method or method, depending on the grouting technique) and is the interface friction
type of soil surrounding the pile. angle between the pile and the soil.
The BUSTAMANTE and DOIX method completely The tip bearing capacity in granular soils can be
neglects the tip bearing capacity of the micropile calculated as:
and proposes to calculate the skin friction mobilized
along the shaft as: (7)
(4) where v,p is the vertical effective stress at the pile tip
and Nq is a bearing capacity factor, which is related to
where Dsi is the effective diameter of the micropile the angle of friction of the soil surrounding the pile
after the grouting phase, Lsi is the bond length re- tip. For bored pile it is often suggested to use a Nq
lative to each i-th soil layer surrounding the shaft, according to the BEREZANTSEVs formulation [1961].
and qsi is the skin friction mobilized at the soil-pile It should be kept in mind that the end bearing
interface, function of NSPT and of the type of micro- capacity of micropiles is customarily neglected in
pile, i.e. IGU (Injection Globale et Unitaire) or IRS limit load calculations (e.g. ARMOUR et al., 1997), be-
(Injection Rptitive et Slective) [BUSTAMANTE and ing the tip settlement usually not sufficient to fully
DOIX, 1985]. mobilize the whole base resistance.
In the and method, the micropile lateral
resistance in cohesive and cohesionless soils is calcu-
lated using the following equations: Soil conditions at STS
From the ground level to the depth of approxi- Triaxial unconsolidated and undrained com-
mately 6 m, the soil is composed of a mixture of a pre- pression tests, carried out on previously saturated
dominant silt fraction mixed with sand and clay. Sub- specimens trimmed from undisturbed samples tak-
layers show the local presence of some gravel, which en between 1 m and 3 m below g.l., provided an un-
becomes particularly important between 4 m and 5 drained compressive strength, su, ranging approxi-
m below g.l. Atterberg limits of the fine grained frac- mately between 100 and 110 kPa. Since the fine
tion denote low plasticity silty clays in the upper part fraction in the upper layer exceeds 30% [OMINE et
of the layer and low plasticity silts in the lower one, al., 1994; COLA, 2002] and owing to the high rate of
respectively. stress application throughout the pile load tests, an
Between 6 and 8 m below g.l., gravel fraction be- undrained response for the silty-clayey mixture was
comes significant (beyond 50%), thus controlling tentatively assumed. Nevertheless, it should be em-
the mechanical response of the soil. To note that the phasized that a partial-saturation soil condition due
gravel particles are characterized by an angular or to presence of a deeper groundwater level would be
subangular shape, leading also to a high degree of a more reasonable assumption in this case, leading
particle interlocking. to a different strength from that reported above and
SPT value remains approximately constant measured in the laboratory in saturated conditions.
(NSPT = 10) up to 5 m below g.l. and rises to 50 and Dynamic Penetration Tests (both SPT and
over at higher depths. Static cone (standard CPT) DPSH) were used to estimate the value of friction
was pushed down only into the silty formation, angle of the deep sandy and silty gravel of approxi-
whereas DPSH allowed to investigate deeper the mately 42. In particular, NDPSH was first converted
ground up to 9 m. It is worth noticing the continu- into NSPT [LA CROIX and HORN, 1973 also reported
ous and regular increase of NDPSH with depth. in Fig. 2] and then introduced in empirical relation-
Fig. 3 Section of standard grouted micropiles. Fig. 4 Section of special micropile C3, showing the tech-
Fig. 3 Sezione di un micropalo colato standard. nical solution used to measure tip resistance only.
Fig. 4 Sezione del micropalo speciale C3, con riportata la
ships available from the literature and relating NSPT soluzione tecnica adottata per la determinazione della resistenza
with the angle of shearing resistance. di punta.
Tab. I Geometrical characteristics of the micropiles in- Each load increment was maintained until the
stalled at the STS. penetration/rise rate of the pile head reached 8
Tab. I Caratteristiche geometriche dei micropali installati nel m/min for a maximum time span of 30 min. Load-
sito sperimentale di Spert. ing-unloading cycles were carried out at different lo-
ad levels in all the tests, but mainly in the correspon-
MICROPILE L [cm] X [cm] Y [cm] Z [cm] dence of 50%, 75%, and 100% Qu.
C1 705 40 0 665
C2 705 45 0 660
Test results
C3 705 40 627 38
T1 689 35 0 654 Figures 5a and 5b present the load-displacement
T2 596 34 0 562 curves from tests under compressive and tensile lo-
T3 705 44 0 661 ad, respectively. From a general point of view, the re-
sponse is characterized by a progressive hardening
behaviour, with a gradual accumulation of irrecove-
to a couple of surrounding micropiles realized speci- rable displacements since the very early stage of the
fically to this purpose. tests. Additional features to note are:
The compression or tension load was applied by the response of micropile C1 does not show cle-
means of a hydraulic jack to the steel tubular rein- arly the failure condition, but it is characterized
forcement or to the tendon, respectively, where- by a continuous hardening behaviour. This has
as the vertical displacement of the tested micropile been associated with a probable significant in-
head was measured by three dial gauges according fluence on the axial response caused by impor-
to ASTM D-1143. tant volumes of cementitious slurry loss in the
The loading procedure consisted in the appli- subsoil during the grouting of both the reaction
cation of load increments equal to about 1/8 of the piles necessary to anchor the frame structure
pile ultimate load Qu, estimated by rough prelimi- used to test micropile C1;
nary computations. Once a load equal to 100% Qu the response of micropile C2 shows a more pro-
was applied without the collapse occurrence, the nounced yielding coupled with the occurrence
load increments were reduced to 5% Qu up to the of large displacements even under small load in-
failure, which is generally associated with the de- crements, thus confirming the mobilization of
velopment of relevant displacements, here assu- full shaft friction and base resistance;
med greater than 25% D. Furthermore, load incre- the response of special micropile C3 (realized to
ments of 5% Qu were also applied during the lo- measure base resistance only) is characterized by
ading procedure whenever unexpected behaviors a sudden break in the load-displacement curve
were observed. due to the overcoming of the limit bond stress
Fig. 5a Load-displacement curves from compression load Fig. 5b Load-displacement curves from tensile load
tests. tests.
Fig. 5a Curve carico - spostamento ottenute dalle prove di Fig. 5b Curve carico - spostamento ottenute dalle prove di
carico in compressione. carico a trazione.
Fig. 6 Hyperbolic fitting of load tests in compression. Fig. 7 Hyperbolic fitting of load tests in tension.
Fig. 6 Interpolazione iperbolica delle curve di carico in Fig. 7 Interpolazione iperbolica delle curve di carico a trazione.
compressione.
the pile shaft and the surrounding heteroge-
between the grout surrounding the pile tip and neous material.
the tubular steel reinforcement; From the load-displacement curves depicted in
the load-displacement behaviour of tensile piles figures 5a and 5b, it was tentatively possible to evalua-
T2 and T3 is characterized by a gradual pull-out te the ultimate load relative to each tested micropile.
at increasing load; Assuming a limit load corresponding to a displace-
micropile T1 shows a slightly stiffer response ment of the pile head equal to 10% of the micropile
followed by a sudden pull-out of the pile from diameter (for bored pile the load at w/D = 0.1 is typi-
the soil. This could be probably due to an unex- cally accepted as limit load according to EN1997/1),
pected failure of some local blockages between the micropile reference ultimate loads in both com-
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.
Fig. 9 CEMSET fitting procedure of the experimental Fig. 10 CEMSET fitting procedure of the experimental
load-settlement curve of micropile C2. load-settlement curve of micropile T3.
Fig. 9 Procedura CEMSET per linterpretazione della curva Fig. 10 Procedura CEMSET per linterpretazione della curva
carico - spostamento relativa al micropalo C2. carico - spostamento relativa al micropalo T3.
pression and tension were calculated and compared with other methods, namely those of Davisson, De
with the theoretical calculations described above. Beer, Brinch Hansen, Vander Veen, Fuller and Hoy,
Butler and Hoy, and CEMSET [DAVISSON, 1972; DE
BEER, 1967; DE BEER and WALLAYS, 1972; BRINCH HAN-
Interpretation of load-displacement curves SEN, 1963; VANDER VEEN, 1953; FULLER and HOY, 1970;
BUTLER and HOY, 1977; FLEMING, 1992]. The limit lo-
Ultimate failure load in compression and tension ads obtained using the above methods were compa-
red with the ultimate failure loads estimated from
To interpret the experimental load-displace- the in situ load test results. The outcome of the cal-
ment curves of both groups of piles tested in com- culations is reported in figure 8.
pression and tension, a hyperbolic function was se- For micropiles tested in tension, all the above
lected as suggested by CHIN [1970]. The result of the methods provide values in agreement with the field
best fitting is shown in figures 6 and 7 for piles under
compression and tension, respectively.
It is worth noting that at large displacements,
the experimental behavior can be well described by
the hyperbolic curve, thus allowing the estimate of
the limit load represented by the inverse slope of the
corresponding straight line. Notwithstanding, the
hyperbolic function fitting produced an overestima-
te of the actual failure load and, therefore, a 90% of
the asymptotic limit value was assumed according to
MANDOLINI [1995].
In the case of the tensile load tests, the ultimate
load corresponding to the pull-out of the piles from
the soil is clearly appreciable from the experimental
curves and can be easily compared with that provi-
ded by the hyperbolic interpretation. Unfortunately,
for piles in compression, the unique curve leading to
a proper estimate of the limit load is that referred to
pile C2 (Fig. 5a).
In addition to the modified Chin method (90% Fig. 11 Double logarithmic plot of the uplift displace-
of the failure load determined with the Chin proce- ments during time of micropile T3.
dure), the load-displacement curves obtained from Fig. 11 Grafico doppio-logaritmico del sollevamento misurato
the compression and tension tests were interpreted nel tempo per il micropalo T3.
Fig. 12 Accumulation of displacement during tensile Fig. 13 Displacement rate from load-displacement curve
load test carried out on micropile T3. of micropile T3.
Fig. 12 Accumulo di deformazione plastica durante il test di Fig. 13 Velocit di deformazione ottenuta dalla curva carico -
carico a trazione eseguito sul micropalo T3. spostamento relativa al micropalo T3.
results, with the exception of the De Beers and the pidly increasing trend of w/D with time approaching
Butler and Hoys ones, that are too conservative [FEL- the pull-out of the pile.
LENIUS, 1980]. Figures 13 shows the displacement rate vs. time
For micropile C2 it is possible to observe an in- at different load levels close to the maximum pull-
creased variability in limit load values, being the out load. A progressive reduction of the displace-
CEMSET, the modified Chins, and the Brinch- ment rate with time and a stabilization of the dela-
Hansens procedures the most reliable methods. It yed deformation over a long period, namely a sort
is worthwhile to note that the best interpretation of primary creep phase [FABRE and PELLET, 2006],
method seems to be the CEMSET one, that provi- can be observed. Under the maximum load of 350
des both the limit load and the entire load-displa- kN, an initial primary creep phase characterized by
cement curve, the latter based on best fitting opti- a strain rate reduction is followed by a sudden in-
mization. In particular, the results of this procedu- crease in the strain rate up to the failure state (ter-
re are presented in figures 9 and 10 for micropiles tiary creep phase) without a clearly detectable in-
C2 and T3. termediate secondary creep phase. This is probably
due to the diffuse and progressive damage of the
soil matrix occurred during the previous load in-
Time-dependent effect on micropile pull-out crements.
Base resistance
Figure 16 shows that the end bearing resistan- interaction system, typically denoted by a ductile re-
ce of the micropile C3 increases rapidly providing a sponse, but rather by the interface between the steel
contribution of around 45% of the total bearing ca- reinforcement and the grout.
pacity at a head displacement of about 10% of the Laboratory tests were carried out on the grout
pile diameter. This can be attributed to the excel- used to form the micropiles at the STS to investiga-
lent mechanical characteristics of the underlying lay- te the shearing strength (unconfined compressive
er (composed of very dense gravel) in which the pile strength) of the cementitious mortar. In accordan-
base was embedded. ce with CEB-FIP Model Code [1990], the limit bond
The contribution of the shaft resistance has been stress between the grout surrounding the pile tip
finally obtained by subtracting the end bearing resi- and the smooth reinforcing steel tube can be com-
stance curve from the reconstructed load-displace- puted as follows:
ment curve shown in figure 16. The resulting hyper-
bolic trend provided higher loads than those observed (8)
at comparable displacements in the curves relative to
the micropiles tested under tension, ranging between Considering a characteristic compression
1.20 times those measured from pile T2 and 1.05 times strength value, fck, of 32,3 MPa, the axial force ne-
those collected from pile T3, thus giving a ratio of the cessary to cause the reaching of the limit bond stress
shaft resistance under tension over that under com- along Z (see Fig. 4) is approximately equal to 260
pression lower than unity. Despite some difference in kN, that is comparable with the ultimate load measu-
the range of variation, this ratio seems to be in reaso- red at the end of the load test. Therefore, it is presu-
nable accordance with the observations of DE NICOLA mable that this limit load was due to the overcoming
and RANDOLPH [1999] and HAN and YE [2006]. of the allowable shear stress between the grout and
the pile steel reinforcement.
Fig. 17 Comparison between ultimate load bearing capacities calculated from static formulae using the factor proposed
by VIGGIANI [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.
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.
the BUSTAMANTE and DOIX method [1985]; along with the excellent mechanical properties of
the common method based on static formulae the gravelly layer in which the pile tip was embed-
known as method or method, depend- ded.
ing on the type of soil surrounding the pile. The and method underestimated both the
The calculation of the ultimate bearing capacity tip and the shaft resistance. This may be due mainly
of the six micropiles at the STS with both the above to two reasons.
methods (assuming an effective diameter equal to Firstly, the end bearing capacity factor, Nq, as-
200 mm) did not provide results in agreement with sumed in the calculation according to BEREZANTSEV
the failure loads measured during the compressive [1961], theoretically corresponds to a critical tip
and tensile load tests and with the estimated values displacement of 5% of the pile diameter (usually
computed by the 90% Chin criterion and the CEM- indicated as Nq*). However, in this case, the above
SET method. The outcomes are presented in figure assumption yielded to an ultimate base load lower
17. The differences obtained may be tentatively ex- than that obtainable at the same w/D from figure 16.
plained by the following considerations. In order to compute an ultimate end bearing resis-
BUSTAMANTE and DOIX [1985] proposed different tance referred to a w/D = 0.1 [EN1997/1], a bearing
unitary skin friction curves depending on the type capacity factor Nq = 1,4 Nq* has been proposed.
of soil in which the micropile is installed and on two Secondly, the shaft resistance was calculated
grouting techniques, i.e. IGU and IRS. A relation- with a rather small factor, in accordance with usu-
ship for the grouting procedure by gravity head on- al geotechnical design recommendations for bored
ly was not provided by the authors. Therefore, the piles [VIGGIANI, 1993]. Nevertheless, the experimen-
assumption of the same skin friction suggested for tal findings from the pull-out tests showed that a
IGU technique caused an overestimation of the uni- factor equals to 0.9-1.0, estimated assuming an und-
tary lateral resistance along the shaft. rained soil behavior, seems to be give more reliable
In addition, the Bustamante and Doix method results with respect to those proposed by VIGGIANI
completely neglects the base resistance providing a [1993]. On the other hand, it should be emphasized
total bearing capacity lower than that measured in that the hypothesis of undrained conditions could
the micropiles tested under compression load. be misleading in this case, as the in situ material is
From the experimental findings, in fact, it was more correctly characterized by partial saturation
possible to notice that the typical assumption of neg- conditions, that in turn involve a variation in soil
ligible tip resistance cannot be properly considered strength from the assumed values.
for the micropiles installed at the STS, because of In order to determine the shaft resistance of mi-
the very stiff response of the soil underneath the pile cropiles grouted under gravity head only in granu-
lar soils, the factor was chosen equal to K0tan, ASTM (1995) D3689-90 Standard Test Method for in-
according to common design recommendations for dividual piles under static axial tensile load.
bored piles [e.g. FHWA, 2010]. BELLATO D., SIMONINI P., DAGOSTINI S. (2011) Com-
Thus, from the above considerations, the ulti- portamento di micropali in terreni eterogenei. Proc. XX-
mate bearing capacities of the micropiles were recal- IV Convegno Nazionale di Geotecnica, Napoli, 22-
culated. These new values fit relatively well the fail- 24 giugno 2011, pp. 61-68.
ure loads obtained from the experimental load-dis- BEREZANTSEV V.G., KHRISTOFOROV V., GOLUBKOV V.
placements curves, as presented in figure 18, which (1961) Load-bearing capacity and deformation of piled
shows, however, still some difference for micropile foundation. Proc. 5th ICSMGE, Paris, 1961, vol. II,
C3, whose tip limit resistance was not achieved dur- pp. 11-15.
ing the static load test, but only predicted on the ba- BRINCH HANSEN, J. (1963) Discussion: Hyperbolic stress-
sis of the measured data. strain response. Cohesive soils. ASCE, J. SMFD, 89, n.
4, pp. 241-242.
BROWN D.A., TURNER J.P., CASTELLI R.J. (2010) Drilled
Conclusions shafts: construction procedures and LRFD Design Meth-
ods. Federal Highway Administration, FHWA-
The results and the interpretation of axial load NHI-10-016.
tests in tension and in compression performed on BRUCE D.A., DIMILLIO A.F., JURAN I. (1997) Micro-
six micropiles installed in heterogeneous soils typi- piles: the state of practice. Part 1: characteristics, defini-
cal of the Alpine mountainous area were presented tions and classifications. Ground Improvement, 1,
and discussed in this paper. n.1, pp. 25-35.
After the preliminary site characterization carri- BUSTAMANTE M., DOIX B. (1985) Une mthode pour
ed out by means of in situ and laboratory tests, three le calcul des tirants et des micropieux injects. Bulletin
compressive and three tensile load tests were execut- Liaison Laboratoire des Ponts et Chausses, n. 140,
ed on simply gravity filled micropiles of 200 mm in pp. 75-95.
diameter. BUTLER H.D., HOY H.E. (1977) Users manual for the
Through the construction of a special micropile Texas quick-load method for foundation load testing.
with minimized contribution of skin friction, it was FHWA, Washington, 59 pp.
possible to interpret the load distribution between CEB-FIP COMITE EURO-INTERNATIONAL DU BETON
shaft and base resistance of the micropiles. The re- (1993) CEB-FIP Model Code 1990. Thomas Telford
sults emphasized that the usual assumption of ne- Services Ltd.
glecting tip resistance in micropile design can be CEN EUROPEAN COMMITTEE FOR STANDARDIZATION
considered sometimes too conservative in these geo- (2004) EN 1997/1 Eurocodice 7. Part 1 - Geotechnical
technical contexts. Design, General rules. Bruxelles.
The reliability of a number of methods for the CESTARI F. (2005) Prove geotecniche in sito. Ed. Geo-
interpretation of load-displacement curves obtai- graph, Segrate, Milano, 416 pp.
ned from pile load tests was investigated, and the CHIN F.K. (1970) Estimation of the ultimate load of
CEMSET procedure proved to be the most efficient piles not carried to failure. Proc. 2nd Southeast Asian
for the estimate of the ultimate bearing capacity. Conf. on Soil Engng., Singapore 1970, pp. 81- 90.
Finally, the outcomes of the micropile design COLA S. (2002) On modeling the behaviour of mlang-
approach introduced by Bustamante and Doix we- es. Proc. 5th European Conference on Numerical
re compared with the classic method, customa- Methods in Geotechnical Engineering, Paris 2002,
rily adopted in bored pile design, and the load test pp. 143-148.
results. From the comparison, it was found that in DAVISSON M.T. (1972) High capacity piles. Proceed-
some particular conditions, such as that discussed in ings, Lecture Series, Innovations in Foundation
the paper, significant underestimation of the effecti- Construction, ASCE, Illinois Section, 52 pp.
ve limit load may lead to an excessively conservative DE BEER E.E. (1967) Proefondervindelijke bijdrage tot
micropile design. de 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,
Bibliografia pp. 41-88; n. 4, pp. 321-360; n. 5, pp- 395-442; n. 6,
pp. 495-522.
ARMOUR T., GRONECK P., KEELEY J., SHARMA S. (1997) DE NICOLA A., RANDOLPH M.F. (1999) Centrifuge mod-
Micropile Design and Construction Guidelines. Imple- eling of pipe piles in sand under axial loads. Gotech-
mentation Manual. Federal Highway Administra- nique, 49, n.3, pp. 295-318.
tion, FHWA-SA-97-070. FELLENIUS B.H. (1980) The analysis of results from rou-
ASTM (1994) D1143-81 Standard Test Method for piles tine pile load tests. Ground Engineering, 13, n. 6, pp.
under static axial compressive load. 19-31.
FLEMING W.G.K. (1992) A new method for single pile Benevento, 568 pp.
settlement prediction and analysis. Gotecnique, 42, n. VIGGIANI C. (1993) Further experience with auger piles
3, pp. 411- 425. in Naples area. Proc. of the Second Int. Seminar on
FULLER R.M., HOY H.E. (1970) Pile load tests includ- Deep Foundations on Bored and Auger Piles, BAP
ing quick-load test method, conventional methods and in- II, Van Impe (Ed.), Balkema, Rotterdam, pp. 445-
terpretations. HRB 333, pp. 78-86. 458.
HAN J. YE S.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.
JEON S.S. (2004) Interpretation of load tests on mini- Interpretazione di prove di carico a
piles. Proc. ICE, Geotechnical Engineering, 157, n.
2, pp. 85-90.
rottura su micropali installati in terreni
JURAN I., BRUCE D.A., DIMILLIO A., BENSLIMANE A. alpini eterogenei
(1999) Micropiles: the state of practice. Part 2: design
of single micropiles and groups and networks of micro-
piles. Ground Improvement, 3, n. 3, pp. 89-110. Sommario
KORECK W. (1978) Small diameter bored injection piles. Larticolo tratta di importanti aspetti legati alla costruzione
Ground Engineering, 11, n. 4, pp. 14-29. e allinterpretazione di prove di carico condotte su pali di piccolo
LACROIX Y., HORN H. (1973) Direct Determination and diametro, comunemente detti micropali, installati mediante
Indirect Evaluation of Relative Density and Its Use on trivellazione e getto a gravit in terreni fortemente eterogenei.
Earthwork Construction Projects. Evaluation of Rela- Queste formazioni, in genere incontrate nelle zone di pendio
tive Density and Its Role in Geotechnical Projects delle regioni alpine italiane, sono prevalentemente costituite da
Involving Cohesionless Soils, ASTM Special Techni- una miscela caotica e granulometricamente irregolare di ghiaia
cal Publication 523, pp. 251-280. e sabbia con limo e argilla spesso contenente ciottoli e massi.
MANDOLINI A., 1995 Prove di carico su pali di fondazio- Sono pertanto difficilmente caratterizzabili dal punto di vista
ne. Ed. Hevelius, Benevento, 64 pp. geotecnico, date le problematiche relative allesecuzione di prove di
RANDOLPH M.F., WROTH C.P. (1982) Recent develop- laboratorio o in sito, ad esclusione delle classiche prove dinamiche.
ments in understanding the axial capacity of piles in clay. Ciononostante, questi terreni mostrano un significativo grado
Ground Engineering, 15, n.7. di mutuo incastro e una risposta meccanica a taglio fortemente
RANDOLPH M.F., WROTH C.P. (1978) Analysis of the de- dilatante, dalla quale dipende lo sviluppo di unelevata resistenza
formation of vertically loaded piles. J. Geotech. Engng. sia di punta sia per attrito laterale anche in micropali di tipo
Div., ASCE, 104, n.12, p. 1465. colato, vale a dire in configurazioni caratterizzate da un limitato
RUSSO G. (2004) Full-scale load tests on instrumented confinamento laterale. Un simile comportamento non tipicamente
micropiles. Proc. ICE, Geotechnical Engineering, descritto dagli usuali approcci di progettazione dei micropali
157, n. 3, pp. 127-135. che, spesso, forniscono una stima troppo conservativa del carico
SABATINI P.J., TANYU B., ARMOUR T., GRONECK P., KEELEY limite verticale. Al fine di rendere tali metodi di progettazione pi
J. (2005) Micropile Design and Construction Refer- adatti a rappresentare la realt fisica del sito, stato realizzato un
ence Manual. FHWA-NHI-05-039. apposito campo sperimentale nelle Alpi nord-orientali nel quale
SABINI G., SAPIO G. (1981) Behaviour of small diam- condurre diverse prove a rottura in compressione e a trazione in
eter bored piles under axial load. Proc. 10th ICSMFE, condizioni controllate. Sulla base dei risultati di queste prove
Stockholm 1981, pp. 823-828. stata valutata laffidabilit dei pi comuni metodi di calcolo dei
VANDER VEEN C. (1953) The bearing capacity of a pile. micropali.
Proc. 3rd ICSMFE, Zurich, vol. II, pp. 84-90. Parole chiave: pali trivellati, micropali, terreni eterogenei, test
VIGGIANI C. (1999) Fondazioni. Ed. Hevelius, di carico su pali