Sei sulla pagina 1di 11

Load Testing of a Closed-Ended Pipe Pile Driven

in Multilayered Soil
Daehyeon Kim1; Adriano Virgilio Damiani Bica2; Rodrigo Salgado3; Monica Prezzi4; and Wonje Lee5

Abstract: Piles are often driven in multilayered soil profiles. The accurate prediction of the ultimate bearing capacity of piles driven in
mixed soil is more challenging than that of piles driven in either clay or sand because the mechanical behavior of these soils is better
known. In order to study the behavior of closed-ended pipe piles driven into multilayered soil profiles, fully instrumented static and
dynamic axial load tests were performed on three piles. One of these piles was tested dynamically and statically. A second pile served as
reaction pile in the static load test and was tested dynamically. A third pile was tested dynamically. The base of each pile was embedded
slightly in a very dense nonplastic silt layer overlying a clay layer. In this paper, results of these pile load tests are presented, and the
lessons learned from the interpretation of the test data are discussed. A comparison is made of the ultimate base and limit shaft resistances
measured in the pile load tests with corresponding values predicted from in situ test-based and soil property-based design methods.
DOI: 10.1061/共ASCE兲1090-0241共2009兲135:4共463兲
CE Database subject headings: Bearing capacity; Pipe piles; Pile load tests; Layered soils.

Introduction pate very fast, and thus do not play the same role during instal-
lation or, afterwards, during loading of the pile as they do in clay.
While soil mechanics developed mostly by study of pure clay and When a closed-ended pipe pile is driven in mixed soil, excess
clean sand, real soil profiles do not need to conform to either soil pore pressure development and dissipation is considerably more
type. Piles are often driven into mixed soil profiles, i.e., profiles complex. The development of base resistance, which may be af-
consisting of multiple layers of soils that may consist of sand, fected by multiple soil layers if these are located sufficiently near
clay, silt, or a mixture of these three particle sizes. Closed-ended the pile base, is also more difficult to analyze. Finally, methods
and open-ended pipe piles have been widely used throughout the developed separately for sand and clay are by necessity used not
world. Generally, the ultimate bearing capacity of a closed-ended only for clean sand and pure clay, but also for soils that are
pipe pile is larger than that of an open-ended pile under the same classified as “sand” or “clay” but that in truth are neither. There-
load and soil conditions 共Salgado 2008兲, because of the larger fore, both theoretical and experimental efforts must be made to
displacement imposed on the soil surrounding the pile during develop a better understanding of the behavior of piles driven into
mixed soil profiles. This is the motivation for the work reported in
driving.
this paper.
The number of pipe pile load tests reported in the literature is
In the present study, two static axial load tests and a number of
small and limited to cases in which the piles were driven either in
dynamic axial load tests were carried out on a fully instrumented
clay or in sand. The load response of a closed-ended pipe pile
closed-ended pipe pile driven into a multilayered soil profile 共11
depends on the type of soil through which the pile is driven.
different soils兲 and embedded slightly in a very dense nonplastic
Driving a pile into a clayey soil will generate excess pore pres-
silt layer to a depth of 17.4 m. An extensive in situ and laboratory
sures that will dissipate slowly; upon pore pressure dissipation,
testing program was performed for characterization of the soils at
the pile bearing capacity increases. Driving a pile into a sandy
the test site. This included standard penetration tests 共SPTs兲, cone
soil will also generate excess pore pressures, but these will dissi-
penetration tests 共CPTs兲, triaxial compression tests, consolidation
tests, and index property tests.
1
Assistant Professor, Department of Civil Engineering, Chosun Univ., This paper presents the results of the pile load test program
Gwangju 501-759, Korea 共corresponding author兲. E-mail: dkimgeo@ and discusses some important lessons that were learned from the
chosun.ac.kr interpretation of the test data. The measured ultimate base resis-
2
Visiting Professor, School of Civil Engineering, Purdue Univ., West
tance and limit shaft resistance of the test piles were compared
Lafayette, IN 47907-2051. E-mail: abica@purdue.edu
3
Professor, School of Civil Engineering, Purdue Univ., West Lafay-
with the corresponding values predicted using several design
ette, IN 47907-2051. E-mail: rodrigo@purdue.edu methods available in the literature with the objective of evaluating
4
Associate Professor, School of Civil Engineering, Purdue Univ., their applicability to the case of piles driven in mixed soil.
West Lafayette, IN 47907-2051. E-mail: mprezzi@purdue.edu
5
Postdoctoral Fellow, School of Civil Engineering, Purdue Univ.,
West Lafayette, IN 47907-2051. E-mail: lee114@ecn.purdue.edu Pile Design Methods
Note. Discussion open until September 1, 2009. Separate discussions
must be submitted for individual papers. The manuscript for this paper In general, piles are designed for axial loads by first calculating
was submitted for review and possible publication on July 16, 2007; an ultimate capacity and reducing that by a suitable factor of
approved on June 26, 2008. This paper is part of the Journal of Geotech- safety. For typical design situations, the ultimate bearing capacity
nical and Geoenvironmental Engineering, Vol. 135, No. 4, April 1, Qult of a single pile consists of the ultimate base resistance Qb,ult
2009. ©ASCE, ISSN 1090-0241/2009/4-463–473/$25.00. and the limit shaft resistance QsL:

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009 / 463

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
n

Qult = Qb,ult + QsL = 共qb,ultAb兲 + 兺


i=1
qsLiAsi 共1兲

where qb,ult = ultimate unit base resistance; qsLi = limit unit shaft
resistance for soil layer i; Ab = area of pile base; Asi = pile shaft
area interfacing with layer i; and n = number of soil layers.
There are two largely empirical approaches that are used to
design single piles subjected to axial loads: 共1兲 the indirect 共or
soil property-based兲 design methods; and 共2兲 the direct 共or in situ
test-based兲 design methods. The soil property-based design meth-
ods are either based on total stresses 共e.g., the ␣ method, typically
used for clayey soils兲 or effective stresses 共e.g., the ␤ method,
typically used for sandy soils兲. The indirect design methods re-
quire that the soil properties of interest be estimated first from in Fig. 1. Layout of the piles and soundings of the load test program
situ or laboratory tests. For sand, these properties include state
variables, such as relative density DR and initial effective stress
state, and intrinsic variables 共mainly the critical-state friction
angle ␾c兲. For clay, required properties include any two of void layers are sandy soils, whereas all the other soil layers are either
ratio, overconsolidation ratio, or current effective stress state, and clayey or silty soils, except for the first soil layer 共which consists
basic index properties. From these variables, performance-based of organic soil兲. Isotropically consolidated undrained triaxial
variables 共i.e., indicators of soil stiffness or strength兲—such as G0 compression tests 共CIUC兲 were performed on Shelby tube
共small-strain shear modulus兲, ␾⬘p 共peak friction angle兲, and su samples collected from the third, seventh, eighth, ninth, and tenth
共undrained shear strength兲—may be determined as needed. soil layers, as well as consolidation tests 共for the first, third, sev-
In the direct design methods, in situ test results 共in most cases, enth, ninth, and tenth soil layers兲.
from either CPTs or SPTs兲 are used directly as input variables. As shown in Fig. 1, one CPT 共C2兲 at the main test pile location
For instance, the CPT cone resistance qc or the SPT blow count and one CPT 共C1兲 at a reaction pile location 共RCEP-1兲 were
NSPT, multiplied by different factors depending on soil type, pile performed. Two SPT logs were also carried out. Undisturbed tube
type, and pile installation method, are directly linked with qb,ult samples were obtained from several depths. The groundwater
and qsL. In order to evaluate the accuracy of both indirect and table was relatively shallow, at a depth of about 1 m from the
direct methods, we calculated Qult, Qb,ult, and QsL of the closed- ground surface. Figs. 2共a and b兲 show the CPT profiles and SPT
ended test pipe pile using several indirect and direct methods that results down to a depth of 20 m. The subsoil profile at the test site
are widely used in pile design practice and discuss the results in consists of multiple layers of various types of soils. The mixed
subsequent sections. soil layers extending down to a depth of 17 m are underlain by a
1.1 m thick very dense silt layer, by an 8 m thick soft to stiff clay
layer and finally by the bedrock. Several preliminary analyses
Pile Load Test Program using the GRLWEAP program to evaluate pile drivability with the
given pile materials, pile sections, and hammer revealed that the
A pile load test program was undertaken to study the behavior of piles 共especially, the main CE pile兲 would have difficulty pen-
closed-ended pipe piles. The main test pile and the reaction piles etrating the very dense silt layer, which showed values of cone
were driven into a mixed soil profile 共alternating layers of clayey, resistance of about 50 MPa. The pile base was embedded slightly
silty, and sandy soils兲. Two static axial load tests were performed in the silt layer, at the depth of 17.4 m.
on the main test pile. In addition, large-strain dynamic tests using
the GRL PDA 共pile driving analyzer兲 were performed in accor-
dance with ASTM D4945-89 on all the piles 共test and reaction Instrumented Test Pile and Reaction Piles
piles兲 during initial driving and also during restrikes at various A total of eight piles were installed at the test site, as shown in
times during a 5 month period so that time-dependent changes in Fig. 1. The main test pile, a closed-ended steel pipe pile 共MCEP兲,
pile capacity could be studied. Both steel closed-ended pipe and was driven to a depth of about 17.4 m. The main test pile had an
H piles were used as reaction piles for the performance of the outside diameter of 356 mm and a wall thickness of 12.7 mm.
static axial-load tests. The pile base was closed by a 25.4 mm thick steel plate welded at
the base. This pile was not filled with concrete after driving. The
reaction piles were five HP310⫻ 110 H piles and one closed-
Subsurface Investigation Using In Situ and Laboratory
ended pipe pile identical to MCEP. In addition to the main pile
Tests
and the six reaction piles, an extra pile 共designated as ECEP兲 was
The site for the pile load tests is located on SR 49, over Oliver driven solely for performing dynamic restrike tests over time.
ditch, in Jasper County, Indiana. In addition to the preliminary In order to obtain qsL for each soil layer and qb,ult, 34
boring data obtained for the bridge project at this location, addi- vibrating-wire 共Geokon Model 4150兲 strain gages were attached
tional 共CPT and SPT兲 soundings and laboratory tests were per- on both sides of the main test pile surface at 17 levels. L-shaped
formed to fully characterize the physical and mechanical steel channels 共76 mm wide and 6 mm thick兲 were welded to both
properties of the soils found at the test site. sides of the piles to protect the strain gages from damage during
Fig. 1 shows schematically the location of the piles and sound- driving. Fig. 3 shows the distribution of the strain gages along the
ings at the test site. Table 1 summarizes the main properties of the pile length. The strain gages were spaced closer together near the
soils encountered. The second, fourth, sixth, and eleventh soil pile base to allow clear separation of QsL and Qb,ult.

464 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Table 1. Soil Properties
Grain size distribution
Total %
unit passing
Layer Depth wc LL PL PI weight D Rb cv ␴vp ␾c su No. 200 % % % %
No. 共m兲 共%兲 共%兲 共%兲 共%兲 共kN/ m3兲 共%兲 e0 Cc Cs 共m2 / sec兲 共kPa兲 OCR 共deg兲 共kPa兲 sieve gravel sand silt clay
1 0–2.9 96 138 49 89 13.4 — 3.56 0.67 0.82 1.66⫻ 10−6 — — — — — — — — —
a
2 2.9–3.7 15 — — — 22 78 — — — — — 31a — 6 3 91 6 —
3 3.7–5.7 19 18 10 8 21.6 — 0.47 0.11 0.03 2.45⫻ 10−7 265 5.6 — 78 71 — 29 42 29
4 5.7–7.3 — — — — 22a 52 — — — — — — 29a — — — — — —
a
5 7.3–7.8 — — — — 21 — — — — — — — — — 51 5 45 39 12
6 7.8–9.0 — — — — 22a 81 — — — — — — 29a — — — — — —
7 9.0–10.2 25 37 18 19 20.1 — 0.73 0.19 0.03 6.82⫻ 10−7 365 3.2 — 220 98 — 2 82 16
8 10.2–12.0 23 29 19 10 20.6 0.63 0.13 0.01 3.43⫻ 10−6 265 1.9 — 320 98 — 2 83 15
9 12.0–14.5 15 21 12 9 21.9 — 0.45 0.10 0.01 5.30⫻ 10−7 750 4.9 — 103 85 — 15 64 21
10 14.5–17 11 22 12 10 21.6 — 0.40 0.08 0.02 3.75⫻ 10−7 365 2.0 — 292 95 — 5 65 30
a
11 17–18.4 — — — — 21 95⬃ 100 — — — — — — 30a — 99 — 1 94 5
Note: wc = natural water content; e0 = initial void ratio; LL= liquid limit; PL= plastic limit; PI= plasticity index; Cc = compression index;
Cs = recompression index; cv = vertical coefficient of consolidation; ␴vp = preconsolidation pressure; OCR= overconsolidation ratio; ␾c = critical-state
friction angle; su = undrained shear strength; and DR = relative density.
a
These values are assumed.
b
Relative densities of the layers were estimated from cone resistance of C2 log using Salgado and Prezzi 共2007兲.

Pile Driving 共BOR兲兴 were performed at various times after pile driving. For
the PDA tests, two axial strain transducers and two piezoelectric
The piles were driven using an ICE 42S open-ended diesel ham-
accelerometers were attached to the upper portion of the piles.
mer with a ram weight of 18.2 kN and a maximum rated energy
Mobilized pile resistance values were estimated using the signal-
of 56.8 kN m at full 3.13 m stroke. One of the reaction piles
matching program CAPWAP.
共RCE-1兲 and one extra pile 共ECEP兲 were driven to the same depth
as the main test pile. All the other reaction H piles 共except for
RHP-1兲 were driven to 24.4 m to make sure the necessary reac-
tion capacity was available. The hammer performance was moni- Pile Load Test Results
tored using the pile driving analyzer 共PDA兲 during pile driving. A
64.7 kN drop hammer, provided by GRL Engineers, Inc., was Driving Resistance
used during restrikes.
The blow counts during driving of the test pile were recorded at
1 m increments down to a depth of 12 m, and then at 0.3 m
Pile Load Tests increments to a depth of 17.4 m. Fig. 4 shows a plot of the driv-
Two static compression load tests were performed on the main ing resistance with depth, expressed by the number of blow
test pile 共50 and 90 days after driving兲. As discussed previously, counts per unit length of pile penetration 共blows/m兲. As can be
one close-ended pipe pile 共RCEP-1兲 and five H piles 共RHPs兲 were seen in this figure, the organic clay layer did not offer any resis-
used as reaction piles. For the static axial load tests, a load cell tance, and the driving resistance starts to develop significantly at
with a capacity of 2,670 kN was used to measure the axial load about 16.0 m. The number of blows per unit penetration needed
applied to the pile head. Two dial gages 共one on each side of the to drive the H piles through the very dense silty soil layer were
test pile兲 attached to two reference beams were used for measur- about 170 关see Fig. 4共b兲兴, but even 200 blows/ m 关see Fig. 4共a兲兴
ing the pile head settlement. were not sufficient for the closed-ended pipe piles to penetrate
Axial load increments ranging from 178 to 356 kN were ap- this silty soil layer.
plied during most of each pile load test; as the load applied at the
pile head approached the plunging or limit load, the load incre- Young’s Modulus of the Closed-Ended Pipe Pile
ments were reduced to the 45–89 kN range. For each loading
step, the pile head settlement was recorded at 1, 2, 5, 10, 20, 30, The strain values obtained 共corresponding to loads of 355, 711,
40, 60, 80, 100, and 120 min. Each loading step was applied after 1,067, 1,334, 1,512, and 1,556 kN in the static load tests兲 from
the pile head settlement rate was less than 0.5 mm/ hr. Average the strain gages installed above the ground level were used to
strains measured from the two strain gages installed on opposite calculate the Young’s modulus E p 共E p = 关load/ 共area of the pile兲 /
sides of the pile at each depth were used to obtain the load trans- strain兴 of the test pile 共Fellenius 2001; Fellenius et al. 2004兲. The
fer curves. Young’s modulus of the test pile was found to be equal to
A total of 15 PDA tests, shown in Table 2, were performed for 210 GPa.
a period of 152 days after pile driving 关once on the main test pile
共MCPE兲, five times on one of the reaction piles 共RCEP-1兲, and Dynamic Load Test Results
nine times on the extra pile 共ECEP兲兴 to investigate setup effects
on pile capacity. Dynamic end-of-initial-driving 共EOID兲 and tests An increase of Qult with time 共pile setup兲 has been reported for
at the end of driving and restrike tests 关beginning of restrike driven piles by a number of researchers 共Skov and Denver 1988;

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009 / 465

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Fig. 3. Instrumentation details: 共a兲 test pile cross section; 共b兲 sche-
matics of instrumentation

ended pipe piles 共piles MCEP, RCEP-1, and ECEP兲 as estimated


from CAPWAP analyses. For pile MCEP, restrike tests after driv-
ing were not performed to avoid any adverse effects on the re-
sponse of the main test pile in the static axial load tests. For piles
ECEP and RCEP-1, restrike tests were performed for a period of
154 days and 127 days after pile installation, respectively. As in-
dicated earlier, ECEP is the pile that was not load-tested statically.
Fig. 5 shows examples of the wave matching analyses done for
pile ECEP. As shown in Figs. 5共a–d兲, the forces and velocities
computed by the CAPWAP analyses at EOID and BOR#6 共re-
strike 104 days after driving兲 agree well with the measured forces
and velocities. Fig. 5共e兲 gives the distribution with depth of the
unit shaft resistance assumed in the wave matching analyses. The
unit shaft resistance, particularly in the lower silty clay layers,
increased with time. The shaft resistance profile is overall in
agreement with the driving resistance profile shown in Fig. 4共a兲.
Fig. 5共f兲 compares the load-settlement curves obtained from
CAPWAP analyses with those from the static load tests. Note that
the mobilized pile resistance values estimated from CAPWAP
analyses 共see Tables 3 and 4兲, are, for each hammer blow, a func-
tion of the set 共typically much less than the 35.6 mm correspond-
ing to a relative settlement of 10% of the pile diameter兲. The
mobilized pile resistance 共879 kN兲 of the test pile at EOID esti-
mated from the CAPWAP analysis is smaller than the ultimate
capacity 共1,345 kN兲 measured in the first static load test 共corre-
sponding to a relative settlement of 10% of the pile diameter兲,
while the mobilized pile resistance 共1,840 kN兲 of the test pile at
BOR#6 is larger than the ultimate capacity 共1,500 kN兲 measured
in the second static load test 共corresponding to a relative settle-
ment of 10% of the pile diameter兲. The mobilized pile resistance
of pile ECEP estimated from the CAPWAP analyses, shown in
Table 3, is 879 kN 共where QsL = 184 kN兲 at EOID, while it be-
Fig. 2. Soil profile and results of in situ tests 共a兲 CPT; 共b兲 SPT comes 1,840 kN 共where QsL = 1,109 kN兲 104 days after pile in-
stallation 共at BOR#6兲. Note that the mobilized pile resistance
increased significantly with time and most of this gain came from
Svinkin and Skov 2000; Komurka et al. 2003, Rausche et al. shaft resistance. The same trend exhibited by ECEP is observed
1985; Lee et al. 2009; Matsumoto et al. 1995兲. Dynamic load tests for pile RCEP-1.
were performed 共1兲 to evaluate the change in load capacity with
time for the closed-ended pipe piles driven into the mixed soil
Load-Settlement Response
profile; and 共2兲 to compare estimates of load capacity from CAP-
WAP analyses with those from the two static axial load tests. Fig. 6共a兲 shows the applied axial load versus pile head settlement
Table 3 shows values of mobilized pile resistance for the closed- curves obtained from the first and second static axial load tests. In

466 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Table 2. Schedule of Static Load Tests and Dynamic Restrike Tests
Piles ECEP RCEP-1 MCEP
Penetration depth 共m兲 17.7 17.4 17.4
Elapsed time after driving (day)
BOR#1 1 1 —
BOR#2 2 8 Two static load tests: 50 days and 90 days after
BOR#3 7 107 driving
BOR#4 10 127
BOR#5 35 —
BOR#6 104 —
BOR#7 134 —
BOR#8 154 —
Remarks All three piles were tested during driving. BOR= beginning of restrike. For restrike tests up to 35 day tests after driving,
a diesel hammer was used; after that, a 64.7 kN drop hammer was used instead.

the case of the first static load test, the pile head settlement was
32.9 mm for an applied axial load of 1,334 kN, and it was almost
twice as much for a load of 1,512 kN. The settlement at the pile
head then increased to 122.6 mm under a load of 1,556 kN. Fi-
nally, the pile almost plunged into the ground with an applied
axial load of 1,646 kN and a corresponding pile head settlement
of 221 mm. Similar axial load response was observed in both load
tests. For the second static load test, the settlement at the pile
head was equal to 53.6 mm 共or a cumulative 260.3 mm measured
from the outset of the first load test兲 for an applied axial load of
1,601 kN and 169.5 mm 共or a cumulative 376.2 mm measured
from the outset of the first load test兲 for a load of 1,690 kN.
The ultimate axial loads corresponding to a pile head settle-
ment of 10% of the pile diameter for the first and second static
load tests are 1,345 and 1,500 kN, respectively. The correspond-
ing cumulative settlement, measured from the outset of the first
load test 共and, thus, including the settlement that occurred during
both the first and second load tests兲 is equal to 241.6 mm. The
value of Qult estimated using the hyperbolic method proposed by
Chin 共1970兲 for the first and second static axial load tests is 1,678
and 1,744 kN, respectively.
Fig. 6共b兲 compares the load-settlement curves for the first and
second static axial load tests; these are plotted together with the
same origin to investigate time-dependent effects. It can be ob-
served from Fig. 6共b兲 that the load-settlement curve from the
second static load test is slightly stiffer than that of the first static
load test.

Residual Loads
As a pile is driven into the ground with repeated hammer blows,
compression/tension waves travel through the pile. As a result of
the interaction between the pile and the surrounding soil, residual
loads remain locked in the pile at the end of driving when the
pile-soil system reaches static equilibrium 共the summation of the
residual base load and the resultant of the residual shaft loads is
equal to zero兲.
Many researchers 共Darrag and Lovell 1989; Briaud and Tucker
1984; Alawneh and Malkawi 2000; Alawneh et al. 2001; Paik and
Salgado 2003兲 have attempted to evaluate the development of
residual loads in driven piles. The residual load distribution with
depth is influenced by the relative stiffness of the pile-soil system
共Darrag and Lovell 1989; Alawneh and Malkawi 2000; Alawneh
et al. 2001兲, the total capacity of the pile, the ratio of shaft to total
Fig. 4. Driving resistances with depth: 共a兲 closed-ended piles driven capacity, and the pile material 共Darrag and Lovell 1989兲.
to a depth of 17.4 m; 共b兲 H piles driven to a depth of 24.4 m Assuming, conservatively, that negative 共downward兲 residual

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009 / 467

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Table 3. Capacities of Closed-Ended Piles from CAPWAP Analyses
Mobilized static pile
resistance from
CAPWAP 共kN兲 Dynamic soil parameters
Smith Quake
LP Blow EMX set damping 共s/m兲 共mm兲
PILE 共m兲 type Day Total Shaft Base 共kN-m兲 共mm/blow兲 共skin/toe兲 共skin/toe兲
MCEP 17.4 EOID 0.1 904 195 710 22.5 4.3 0.394/ 0.295 3.05/ 7.37
ECEP 17.4 EOID 0.1 879 184 695 31.4 4.5 0.586/ 0.079 7.70/ 6.39
17.4 BOR #1 1 1,053 333 719 19.6 3.2 0.739/ 0.267 2.03/ 6.10
18.1 BOR #2 2 1,220 510 710 31.4 3.6 0.663/ 0.235 1.78/ 7.62
18.1 BOR #3 7 1,202 648 554 19.6 1.1 0.689/ 0.493 1.78/ 5.33
18.1 BOR #4 10 1,242 546 695 17.6 3.5 0.788/ 0.853 4.57/ 8.89
18.1 BOR #5 35 1,486 816 670 27.4 3.6 1.149/ 0.492 5.33/ 5.59
18.1 BOR #6 104 1,840 1,109 732 64.7 6.3 1.150/ 1.151 7.11/ 6.35
18.1 BOR #7 134 2,082 1,284 799 75.5 3.3 1.049/ 1.246 6.86/ 12.70
18.1 BOR #8 154 2,283 1,483 800 59.8 2.2 1.214/ 0.696 6.35/ 9.91
RCEP-1 17.4 EOID 0.1 961 290 671 18.6 6.0 0.539/ 0.264 2.54/ 16.30
17.5 BOR #1 1 1,081 405 678 30.4 2.0 0.856/ 0.154 2.54/ 14.49
17.6 BOR #2 8 1,102 431 671 21.6 1.6 0.659/ 0.393 2.09/ 6.37
17.7 BOR #3 107 2,104 1,277 825 51.9 4.5 0.545/ 0.411 3.76/ 13.26
17.7 BOR #4 127 2,254 1,379 873 57.8 5.0 0.476/ 0.202 2.54/ 9.38
Note: EOID= end of initial driving; BOR= beginning of restrike. Note that the mobilized pile resistance values estimated from CAPWAP analyses for each
hammer blow are a function of the set.

shear stresses are fully mobilized along the entire length of the Measured Base and Shaft Capacity
pile and that the shaft resistance is independent of the direction of
shear, the magnitude of the residual shear stress acting at some The load-transfer curves corresponding to each loading step for
depth down the pile shaft cannot be larger than one-half the limit the first and second static axial load tests are shown in Fig. 8.
unit shaft resistance qsL measured there in an axial static load test Given that the last strain gages were located at a depth of 16.9 m,
共Fellenius 2002兲. This follows from the fact that some portion of the load-transfer curves of both load tests were obtained by ex-
the applied load during the pile load test is consumed to change trapolation down to a depth of 17.4 m. It can be concluded from
the direction of the pre-existing, negative residual shear stress Fig. 8 that most of Qult of the MCEP pile is due to the shaft
along the pile after driving. By making this assumption, which resistance mobilized by the soils along the lower 5 m of the pile.
leads to the largest residual stresses possible to develop along the Table 4 shows QsL, Qb,ult, and Qult values measured at the end
pile, we can obtain the residual load distribution with depth by of the two static axial load tests 共50 and 90 days after pile driv-
integrating the residual shear stresses acting along the pile-soil ing兲 as well as other values estimated using the 10% relative
interface over the pile length. settlement criterion, Chin’s hyperbolic criterion, 共Chin 1970兲, and
Residual loads can be estimated from strain gage readings CAPWAP analyses 共based on both the EOID test for pile MCEP
after pile driving 共the gages need to be zeroed before driving兲. In and the restrike tests for pile ECEP at 35 days and 104 days after
our study, an attempt was made to measure the residual loads by the end of driving兲. In this table, as restrike tests were not per-
zeroing all strain gages before driving of the pile and measuring formed for the MCEP pile, the values of QsL, Qb,ult, and Qult
strain values immediately after driving and right before the first measured for pile ECEP, driven to the same depth in the same soil
static load test. However, as the data indicated a drift of the strain profile as pile MCEP are used in the comparison. In each restrike
gage readings, the residual load distribution with depth obtained test, more than 20 hammer blows were applied to the pile head to
from the strain gages is questionable. Consequently, we decided mobilize the shaft and base capacities. Referring to the set
to estimate the residual loads using the Darrag and Lovell 共1989兲 amounts shown in Table 3, it can be noted that the penetration
method. 共see set amounts兲 of the base of the pile for each hammer blow
As shown in Fig. 7共a兲, using the Darrag and Lovell 共1989兲 was much smaller than 35.6 mm, the value corresponding to 10%
method, the maximum negative residual shear stress and the zero settlement of the pile diameter. This indicates that the base resis-
residual stress in the pile are found to develop at depths of 8.0 and tance of the pile was not fully mobilized in the restrike tests.
13.4 m below the ground surface, respectively. Half of the limit Caution is necessary when comparing mobilized resistance values
unit shaft resistance values 共1 / 2 qsL兲 measured at the end of the estimated from CAPWAP analyses with those measured from
first pile load test 关see the load-transfer curves shown in Fig. 8共a兲兴 static load tests.
were also plotted in Fig. 7共a兲 since these are upper bound values The value of QsL 104 days after the end of driving estimated
for the residual shear stress. Fig. 7共b兲 shows the residual loads, using CAPWAP is 4% greater than the QsL measured in the sec-
which were obtained by integration of the residual shear stresses ond static load test. The QsL and Qb,ult CAPWAP estimates
over the pile length. As shown in Fig. 7共b兲, the residual base load 104 days after driving were about 35% and 10% greater than
estimated by the Darrag and Lovell 共1989兲 method is 84 kN. those obtained 35 days after driving, respectively. In contrast, the
Since the estimated residual loads are small when compared with values of QsL and Qb,ult measured in the second static load test are
Qult of the test pile, they were neglected in this paper. only slightly greater 共4% for the shaft resistance and less than that

468 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Fig. 5. Examples of wave matching analyses for pile ECEP: 共a兲 time versus measured and computed forces at EOID; 共b兲 time versus measured
and computed velocities at EOID; 共c兲 time versus measured and computed forces at BOR#6; 共d兲 time versus measured and computed velocities
at BOR#6; 共e兲 distribution with depth of the unit shaft resistance assumed in the matching analyses; and 共f兲 static load-settlement curves derived
from the results of the wave matching analyses together with the results of the static load tests

for the base resistance兲 than those measured in the first static axial Pile Capacity Predictions
load test. These results show that the setup estimated from the
dynamic tests is greater than that obtained from the static load Several methods were used to estimate both QsL and Qb,ult for the
tests. This is an important observation, but one that is not easy to test pile 共MCEP兲. In the case of the direct methods, the CPT- and
generalize from the results of just one case history. SPT-based methods suggested by Aoki and Velloso 共1975兲, Bus-

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009 / 469

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Table 4. Summary of Measured and Estimated Load Capacities
First loading test Second loading test

Total Shaft Base Total Shaft Base


Sources of capacities 共kN兲 共kN兲 共kN兲 共kN兲 共kN兲 共kN兲
Load at the end of the static load test 1,646 1,158 488 1,690 1,201 489
Ultimate load for a settlement of 10% of the pile diameter 1,345 946 399 1,500 1,066 434
Limit load estimated from Chin’s criterion 1,678 1,007 671 1,744 1,062 682
Total Shaft Base
共kN兲 共kN兲 共kN兲
Mobilized pile resistance from the CAPWAP prediction based 904 195 710
on the restrike test at EOID 共MCEP兲
Mobilized pile resistance from the CAPWAP prediction based 879 184 695
on the restrike test at EOID 共ECEP兲
Mobilized pile resistance from the CAPWAP prediction based 1,486 816 670
on the restrike test at 35 days after driving 共ECEP兲
Mobilized pile resistance from the CAPWAP prediction based 1,840 1,109 732
on the restrike test at 104 days after driving 共ECEP兲

tamante and Gianeselli 共1982兲, UWA 共Lehane et al. 2005兲, Fugro


共Kolk et al. 2005兲, NGI 共Kalsrud et al. 2005; Clausen et al. 2005兲,
and ICP 共Jardine et al. 2005兲 were used together with CPT and
SPT data 共S2 and C2; see Fig. 2兲 collected at the test site. As the
UWA and Fugro methods are applicable to sandy soils, only the
base capacity was calculated with these methods.
The Poulos and Davis 共1980兲 method and the Fleming et al.
共1992兲 and API 共1993兲 methods were the indirect methods con-
sidered. Note that the Fleming et al. 共1992兲 method is applicable
only to sandy soils, so the API 共1993兲 method was used in com-
bination with it for the clayey soils. Values of the critical-state
friction angle of the sand 关for the method proposed by Fleming et
al. 共1992兲兴 or the undrained shear strength su of the clay are
required as input in the shaft resistance calculations by these
methods 共the values for these parameters shown in Table 1 were
used in the calculations兲. The ultimate base resistance according
to the Poulos and Davis 共1980兲 method can be calculated by
multiplying the effective stress at the pile base by Nq,ult, which is
a function of the peak friction angle of the sand 共a friction angle
of 32.5° was used in the calculations兲. A detailed description of
these methods can be found in Salgado 共2008兲. In order to calcu-
late QsL, the soil profile was divided into 11 soil layers; the sec-
ond, fourth, sixth, and eleventh layers of the profile were treated
as sandy soils, and all the other soil layers, as clayey soils. Table
5 contains the predictions by the methods discussed in the pre-
ceding paragraph.

Limit Shaft Resistance Predictions


Table 5 shows values of QsL calculated using the direct and indi-
rect design methods discussed earlier. The design methods predict
limit shaft resistances in the range of 0.55 to 1.16 times the actual
QsL measured at the end of the of the first static axial load test.
Limit shaft resistances estimated with the CPT-based methods
proposed by Bustamante and Gianeselli 共1982兲 and Aoki and Vel-
loso 共1975兲 are much smaller 共about 55 to 62%兲 than the mea-
sured QsL. The recently proposed CPT-based methods—the ICP
and the NGI methods—predict better the limit shaft resistances
共in the range of 0.75 to 1.05 of the measured values兲. The SPT-
based method proposed by Aoki and Velloso 共1975兲 predicted the
closest limit shaft resistances in comparison to the measured QsL. Fig. 6. MCEP first and second static axial load tests: 共a兲 cumulative
The limit shaft resistances predicted by the indirect methods pile head load-settlement curves; 共b兲 load-settlement curves

470 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Fig. 7. Assumed distribution of 共a兲 residual shear stress; 共b兲 esti-
mated residual load

Fig. 8. Load-transfer curves: 共a兲 first static loading test; 共b兲 second
关Poulos and Davis 共1980兲, Fleming et al. 共1992兲, and API 共1993兲 static loading test
methods兴 are about 12 to 16% larger than the measured QsL.

Ultimate Base Resistance Predictions


taken in calculating the ultimate base resistance of a closed-ended
As discussed in the previous section, the limit shaft resistances pipe pile embedded in a very dense but thin silty layer overlying
predicted by the indirect and direct design methods are in reason- a softer clay layer. Based on the physics of a penetration process,
able agreement with the measured QsL. However, as shown in the sensing depth is no more than 2 pile 共or cone penetrometer兲
Tables 4 and 5, the indirect methods overestimate the ultimate diameters below the base of the pile 共or cone penetrometer兲 for
base resistance by about 3.3 times the measured Qb,ult, and the homogeneous soil. At working loads, this depth could be larger
CPT- and SPT-based methods overestimate 共by 3.6 to 7.4 times兲 than that, as it also would for soil with strong heterogeneities 共or
the measured Qb,ult. This appears to be due to 共i兲 the extremely with a much weaker layer within reach of the pile base兲.
high values of qc and NSPT measured in the very dense silty layer With appropriate correction for the sensing depth, 共i.e., with
found at the pile base; and 共ii兲 the relatively small thickness of appropriate selection of a representative penetration resistance兲,
this layer, which means that the load is transferred from the pile several theoretical and experimental studies have shown that qbL
base to the weaker and more compressible soils below the bottom is approximately equal to the cone resistance qc 共Randolph et al.
of the silty layer. This suggests that great caution needs to be 1994; Salgado 2008兲. In addition, White and Bolton 共2005兲

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009 / 471

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Table 5. Predicted Ultimate MCEP Pile Capacities resistance was mobilized in the lower third of the pile.
Total Shaft Base 2. The estimated residual loads were minimal when compared
Prediction methods 共kN兲 共kN兲 共kN兲 with the measured ultimate bearing capacity of the pile and
were, therefore, neglected.
Aoki and Velloso 共1975兲—CPTa 2,980 723 2,257
3. The direct and indirect design methods estimated reasonably
LCPC 共Bustamante and Gianeselli 1982兲—CPTc 2,816 638 2,178 well the limit shaft resistance, but overestimated significantly
UWA 共Lehane et al. 2005兲—CPTb 2,376 共by factors of 3.3–7.4兲 the ultimate base resistance of the
Fugro 共Kolk et al. 2005兲—CPTb 2,048 closed-ended pipe pile. The ultimate base resistance esti-
NGI 共Kalsrud et al. 2005, Clausen et al. 2,811 1,227 1,584 mated using qbL = qc was about 6.7–8.5 times larger than the
2005兲—CPTb measured Qb,ult. Therefore, in situations in which a closed-
ICP 共Jardine et al. 2005兲—CPTb 2,856 867 1,989 ended pipe pile is embedded in a very dense silty 共or, gener-
Aoki and Velloso 共1975兲—SPTa 4,319 1,104 3,215 alizing, sandy兲 layer of limited thickness overlying a softer
Poulos and Davis 共1980兲b 2,714 1,301 1,413 or much looser soil layer, the estimation of the ultimate base
Fleming et al. 共1992兲 and API 共1993兲b 2,758 1,347 1,411 resistance needs to be made with caution. The stress and
a
Van der Veen 共1953兲 criterion for ultimate base load. strain fields are affected by the weaker soil layer, and obser-
b
10% relative settlement criterion for ultimate base load. vations made for homogeneous bearing layers regarding
c
Ultimate base load criterion is not available. sensing depths and ratios of unit base resistance qb,ult to ei-
ther cone resistance or SPT blow counts may not apply.
4. The shaft resistance estimated by CAPWAP analysis
showed that the qbL / qc ratio is close to unity if a large pile load 104 days after pile installation was six times larger than the
test database they used is corrected for partial embedment, local value obtained at the end of pile driving, but no correspond-
heterogeneity, partial mobilization, and residual stresses. This has ing increase in the base resistance was observed.
also been demonstrated by additional research at the University of 5. The behavior of axially loaded piles driven in multilayered
Western Australia 共Xu 2007兲. The average cone resistances qc soils is quite complex and is not well understood because the
from the pile base to 1D and 2D below it are equal to 42 and interaction between the pile and the different soils will differ
33 MPa, respectively. Assuming that qbL = qc, the estimated base in each layer. Therefore, additional fully instrumented pile
capacities of the closed-ended pipe pile for the CPT data averaged load tests in conjunction with detailed site characterization
over a depth ranging from the pile base to 1D and 2D below it are are needed to better understand the behavior of piles driven
4,158 N and 3,247 kN, respectively; these values are about 6.7– in multilayered soils.
8.5 times larger than the measured Qb,ult. This implies that the
assumption of qbL = qc needs to be used with caution when esti-
mating the base resistance of closed-ended pipe piles embedded Acknowledgments
in thin dense layers overlying softer soils. Calculation of axial
pile capacity in these cases 共which obviously should be avoided The pile load tests presented in this paper were partially funded
in design but are of interest nonetheless, particularly given that by the Indiana Department of Transportation. This support is
engineers may be tempted to unload piles in stronger intermediate greatly appreciated. The writers acknowledge Kwangkyun Kim,
layers of limited thickness兲, as this case history suggests, is com- Hoyoung Seo, lrem Zeynep Yildirim, Grace Abou-Jaoude, and
plicated also by the fact that the stress and strain fields for a Mir Zaheer for assisting with the laboratory and pile load tests.
homogeneous bearing layer are different than those corresponding
to a situation in which a discontinuity in soil profile exists near
the pile base. Observations regarding sensing depths and ratios of References
qb / qc made for homogeneous soil may not apply, and so calcula-
tions using typical sensing depths and qb / qc or qb / NSPT ratios Alawneh, A. S., and Malkawi, A. I. H. 共2000兲. “Estimation of post-
may not necessarily lead to good estimates of pile-based load driving residual stresses along driven piles in sand.” Geotech. Test. J.,
capacity. Some authors 共e.g., Xu and Lehane 2008兲 argue in favor 23共3兲, 313–326.
of a larger depth of influence and suggest a way to calculate the Alawneh, A. S., Nusier, O., Malkawi, A. I. H., and Al-Kateeb, M. 共2001兲.
base resistance of the pile based on the cone resistance values, “Axial compressive capacity of driven piles in sand: A method includ-
distance of the weak layer to the pile base, and thickness of the ing post-driving residual stresses.” Can. Geotech. J., 38共2兲, 364–377.
weak layer. Such an approach does lead to improved estimates of American Petroleum Institute 共API兲. 共1993兲. “Recommended practice for
base resistance. planning, designing and constructing fixed offshore platforms.” Work-
ing stress design, API RP 2A, 20th Ed., American Petroleum Institute,
Washington, D.C.
Aoki, N., and Velloso, D. A. 共1975兲. “An approximate method to estimate
Conclusions the bearing capacity of piles.” Proc., 5th Pan-American Conf. of Soil
Mechanics and Foundation Engineering, Vol. 1, 367–376.
In this study, in order to better understand the behavior of closed- Briaud, J. L., and Tucker, L. 共1984兲. “Piles in sand: A method including
ended pipe piles driven into multilayered soil profiles, two static residual stresses.” J. Geotech. Engrg., 110共11兲, 1666–1680.
axial load tests and a number of dynamic axial load tests were Bustamante, M., and Gianeselli, L. 共1982兲. “Pile bearing capacity predic-
performed on a fully instrumented closed-ended pipe pile and tion by means of static penetrometer CPT.” Proc., 2nd European
other driven piles, respectively. The load test program included Symp. on Penetration Testing, Vol. 2, 493–500.
extensive in situ and laboratory testing for complete characteriza- Chin, F. V. 共1970兲. “Estimation of the ultimate load of piles not carried to
tion of the soils at the test site. The following lessons can be failure.” Proc., 2nd Southeast Asian Conf. on Soil Engineering, Vol. 1,
learned from this study: 81–90.
1. It was observed that a substantial portion of the limit shaft Clausen, C. J. F., Aas, P. M., and Karlsrud, K. 共2005兲. “Bearing capacity

472 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
of driven piles in sand, the NGI approach.” Proc., 1st Int. Symp. on loaded steel pipe piles driven in soft rock.” J. Geotech. Engrg.,
Frontiers in Offshore Geotechnics, 677–681. 121共4兲, 305–315.
Darrag, A. A., and Lovell, C. W. 共1989兲. “A simplified procedure for Paik, K., and Salgado, R. 共2003兲. “Determination of bearing capacity of
predicting residual stresses for piles.” Proc., 12th Int. Conf. on Soil open-ended piles in sand.” J. Geotech. Geoenviron. Eng., 129共1兲,
Mechanics and Foundation Engineering, Vol. 2, 1127–1130. 46–57.
Fellenius, B. H. 共2001兲. “From strain measurements to load in an instru- Poulos, H. G., and Davis, E. H. 共1980兲. Pile foundation analysis and
mented pile.” Geotech. News, 19共1兲, 35–38. design, Wiley, New York.
Fellenius, B. H. 共2002兲. “Determining the resistance distribution in piles. Randolph, M. F., Dolwin, J., and Beck, R. 共1994兲. “Design of driven piles
Part 2: Method for determining the residual load.” Geotech. News, in sand.” Geotechnique, 44共3兲, 427–448.
20共3兲, 25–29. Rausche, F., Goble, G., and Likins, G. 共1985兲. “Dynamic determination
Fellenius, B. H., Harris, E. E., and Anderson, D. G. 共2004兲. “Static load- of pile capacity.” J. Geotech. Engrg., 111共3兲, 367–383.
ing test on a 45 m long pipe pile in Sandpoint, Idaho.” Can. Geotech.
Salgado, R. 共2008兲. The engineering of foundations, McGraw-Hill, New
J., 41共4兲, 613–628.
Fleming, W. G. K., Weltman, A. J., Randolph, M. F., and Elson, W. K. York.
Salgado, R., and Prezzi, M. 共2007兲. “Computation of cavity expansion
共1992兲. Piling engineering, Blackie, Glasgow and London.
pressure and penetration resistance in sands.” Int. J. Geomech., 7共4兲,
Jardine, R., Chow, F., Overy, R., and Standing, J. 共2005兲. ICP design
251–265.
methods for driven piles in sand and clays, Thomas Telford, London.
Skov, R., and Denver, H. 共1988兲. “Time-dependence of bearing capacity
Karlsrud, K., Clausen, C. J. F., and Aas, P. M. 共2005兲. “Bearing capacity
of driven piles in clay, the NGI approach.” Proc., 1st Int. Symp. on of piles.” Proc., 3rd Int. Conf. on Application of Stress-Wave Theory
Frontiers in Offshore Geotechnics, 677–681. to Piles, 1–10.
Kolk, H. J., Baaijens, A. E., and Senders, M. 共2005兲. “Design criteria for Svinkin, M. R., and Skov, R. 共2000兲. “Setup effects of cohesive soils in
pipe piles in silica sands.” Proc., 1st Int. Symp. on Frontiers in Off- pile capacity.” Proc., 6th Int. Conf. on Application of Stress-Wave
shore Geotechnics, 711–716. Theory to Piles, 107–111.
Komurka, V. E., Wagner, A. B., and Edil, T. 共2003兲. “Estimating soil/pile Van der Veen, C. 共1953兲. “The bearing capacity of a pile.” Proc., 3rd Int.
setup.” Final Rep. No. 0092-00-14, Wisconsin Highway Research Conf. on Soil Mechanics and Foundation Engineering, Vol. 2, 84–90.
Program, Wisconsin Dept. of Transportation, Madison, Wis. White, D. J., and Bolton, M. D. 共2005兲. “Comparing CPT and pile base
Lee, W., Kim, D., Salgado, R., and Zaheer, M. 共2008兲. “Setup of driven resistance in sand.” Geotech. Eng., 158共1兲, 3–14.
piles in mixed soil.” Soils Found., accepted. Xu, X. 共2007兲. “Investigation of the end bearing performance of displace-
Lehane, B. M., Schneider, J. A., and Xu, X. 共2005兲. “The UWA-05 ment piles in sand.” Ph.D. thesis, Univ. of Western Australia.
method for prediction of axial capacity of driven piles in sand.” Proc., Xu, X., and Lehane, B. M. 共2008兲. “Pile and penetrometer end bearing
1st Int. Symp. on Frontiers in Offshore Geotechnics, 683–689. resistance in two-layered soil profiles.” Geotechnique, 58共3兲, 187–
Matsumoto, T., Mich, Y., and Hirano, T. 共1995兲. “Performance of axially 197.

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / APRIL 2009 / 473

Downloaded 24 Mar 2009 to 192.248.40.6. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright

Potrebbero piacerti anche