Sie sind auf Seite 1von 14

ARTICLE IN PRESS

International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]


www.elsevier.com/locate/ijrmms

Shear load transfer for rock-socketed drilled shafts based on borehole


roughness and geological strength index (GSI)
Hoonil Seola, Sangseom Jeonga,, Chunwhan Chob, Kwangho Youc
a
Department of Civil Engineering, Yonsei University, Seoul 120-749, Republic of Korea
b
Technical Advisory Team, Samsung Corporation, Gyeonggi 463-824, Republic of Korea
c
Department of Civil Engineering, Suwon University, Gyeonggi 445-743, Republic of Korea
Received 8 December 2006; received in revised form 25 September 2007; accepted 25 September 2007

Abstract

The load distribution and deformation of rock-socketed drilled shafts subjected to axial loads were evaluated by a load-transfer
approach. The emphasis was on quantifying the shear load-transfer characteristics of rock-socketed drilled shafts based on constant
normal stiffness (CNS) direct shear tests performed by varying major factors influencing shaft resistance, including unconfined
compressive strength, borehole roughness, initial confining stress, pile diameter, and material properties. Based on the CNS tests and the
Hoek–Brown failure criterion, a nonlinear triple curve is proposed for the shear load-transfer function of rock-socketed drilled shafts. It
is presented in terms of borehole roughness and the geological strength index (GSI) so that structural discontinuity and surface
conditions of the rock mass can be considered. The proposed function was verified by the load test results of ten rock-socketed drilled
shafts subjected to axial loads. Seven piles were constructed in completely or moderately weathered rocks of granite-gneiss, and the
others were constructed in slightly weathered rocks of clayshale-limestone. Through comparisons with results of load tests, it was found
that the shear load-transfer function in the present study is in good agreement with the general trend observed by in situ measurements,
and this represents a significant improvement in the prediction of drilled shaft shear behavior.
r 2007 Elsevier Ltd. All rights reserved.

Keywords: Rock-socketed drilled shaft; Load transfer; Shaft resistance; Constant normal stiffness; Geological strength index; Borehole roughness

1. Introduction settlements rather than by ultimate bearing capacity. In


addition, because the ultimate shaft resistance is generally
The current design methods for rock-socketed drilled mobilized at smaller interface displacements between the
shafts are mainly based on local knowledge derived from shaft and surrounding rock than ultimate toe resistance,
the observation of load tests, or empirical methods related piles typically carry most of their working load in shaft
to the unconfined compressive strength (UCS) of intact resistance. Therefore, for optimum designs of rock-
rocks. However, it is known that this design approach for socketed drilled shafts, predicting the shear load transfer
piles is generally overly conservative by as much as an from the pile into the surrounding soil or rock is as
order of magnitude [1]. important as, or possibly more critical than, predicting the
According to studies by Reese and O’Neill [2] and ultimate bearing capacity.
Ghionna et al. [3], the bearing capacity of rock-socketed The bearing capacity and shear load-movement perfor-
drilled shafts should be determined by a serviceability limit mance of rock-socketed drilled shafts are critically depen-
capacity within the limit of allowable superstructure dent on construction details and installation geometry
conditions. Comprehensive studies of the details have been
Corresponding author. Tel.: +82 2 2123 2807; fax: +82 2 364 5300. reported by Horvath et al. [4], O’Neill et al. [5], and Seidel
E-mail addresses: geo_dr@yonsei.ac.kr (H. Seol),
and Collingwood [6]. They report that the shear behavior
Soj9081@yonsei.ac.kr (S. Jeong), chunwhan.cho@samsung.com of rock-socketed drilled shafts is highly influenced by the
(C. Cho), khyou@unitel.co.kr (K. You). following parameters: rock strength (drained intact and

1365-1609/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijrmms.2007.09.008

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
2 H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]

residual strength parameters are generally used), borehole f


roughness, rock mass modulus and Poisson’s ratio,
c A
discontinuity structure and surface condition of the rock
mass, pile diameter, initial normal stress between concrete
fmax
and rock prior to loading, and construction practices. Emass / 2.5D
D C
However, it is difficult to determine reliably, based on
empirical methods, the interaction between the above
Emass / 15D
mentioned factors in calculating the performance of a
socketed pile due to the complexity of the interaction.
A conservative approach to design is therefore pursued.
Although a few shear-load transfer functions have been B
proposed for rock-socketed drilled shafts [7–9], it appears
that none of the methods can reliably predict shear o w
behavior for overall rock types to a satisfactory level of
Fig. 1. Potential t–z relations for rock [8].
accuracy due to the intrinsic characteristics of individual
rock types, oversimplification of mechanisms, or subjective
empirical input data. Moreover, because study test sites are way as it would be for shaft friction in pile bearing capacity
mainly sandstone, limestone, and clayshale, less is known computations.
about the behavior of drilled shafts in weathered granite- O’Neill and Hassan [8] suggested potential t–z behavior
gneiss, which occupies two-thirds of the total land area of in rock, as shown in Fig. 1. If the pile–rock interface is
the Korean peninsula. clean so that the cement paste bonds to the rock, the
As a consequence, various influencing factors of the roughness pattern is regular, and the asperities are rigid, a
pile–rock interface, including the degree of weathering and t–z relation such as OABC can be obtained. In most cases,
rock type, should be quantified and taken into account however, the interface asperity pattern is not regular due to
when designing rock-socketed drilled shafts. Toward this some degree of smear; in addition, asperities are deform-
end, a theoretical methodology has been developed to able, which results in ductile, progressive failure among
provide a basis for a load-transfer criterion that would be asperities. Therefore, they proposed an interim criterion for
applicable to drilled shafts installed in rocks. The validity a hyperbolic t–z model in most rock types as described
of this study was tested through field case studies. below until better solutions become available:
w
2. Shear behavior of rock-socketed drilled shafts f ¼ , (1)
ð2:5D=E m Þ þ ðw=f max Þ
The load transfer from a pile to the surrounding soil where w is the pile movement, fmax the maximum unit
ought to be considered in order to achieve structural friction, D the pile diameter, and Em the effective Young’s
compatibility between loads and deformation. Basically, modulus of the rock mass.
the load-transfer method [10] and the continuum approach However, Johnston [16] warns that methods such as the
method [11,12] are used to calculate the load-deformation t–z curves elicit criticism because they do not explicitly
behavior of piles subjected to an axial load. Among them, consider failure mechanisms, random asperity patterns,
the load-transfer method follows a simple analytical rock stiffness, and effects of interface dilation on normal
procedure and can be applied to any complex composition stresses.
of soil layers with a nonlinear stress–strain relationship, In engineering practice, the pile–rock interface of drilled
inhomogeneous medium, and any variation in the sections shafts consists of irregular asperities of varying heights and
along a pile. This method calculates the load–deformation patterns, and hence various failure modes can be expected
relationship for drilled shafts on the basis of a load-transfer to occur, possibly simultaneously. For a pile–rock inter-
function utilizing subgrade reactions of the soil/rock face, shearing results in dilation as one asperity overrides
surrounding the shaft. The soil is modeled by a set of another. If the surrounding rock mass is unable to deform
localized springs, which are defined as a function of sufficiently, an inevitable increase in the normal stress, Dsn,
displacements at several discrete points along the pile occurs during shearing. Because of the increase in stress
including the pile tip: the unit shaft resistance vs local shaft normal to the interface produced by the constant normal
displacement relations (the t–z or f–w curves) and the unit stiffness (CNS) boundary condition, the frictional resis-
toe resistance vs local toe displacement relation (the q–z or tance between pile and rock increases. As a result, the shear
q–w curve). behavior of rock-socketed drilled shafts can be modeled
Several techniques are available for predicting the t–z better under CNS conditions than constant normal load
(or f–w) curves in soils [13–15] and rocks [7–9]. In all of (CNL) conditions.
these methods, the t–z curves are expressed by elastic– In order to take into account the effects of a CNS
plastic models having great differences in initial slope, condition at the interface of a rock-socketed pile or rock
although the ultimate value fmax is obtained in the same joint, much research have been carried out on the shear

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]] 3

strength envelope priority in the form of a modified asperities were tested to analyze the basic mechanisms and
Mohr–Coulomb model [17–19]. In a classic study, the the effect of influencing factors on shaft resistance. The
effect of joint irregularities on shear strength can be mechanisms established from these CNS tests could then be
demonstrated by sliding an interface up a slope inclined at applied to predict the performance of the shear-load
an angle i in the direction of shear. In the absence of transfer of rock-socketed drilled shafts.
cohesion it can be easily shown that Since the study of Johnston and Lam [21], significant
progress using a CNS direct shear machine has been made
t ¼ sn tanðfb þ iÞ, (2)
by several researchers [18,19]. Based on their results under
where fb is the basic friction angle relevant to sliding on a CNS conditions [16,18–20], shear behavior of rock-socket
smooth planar interface and i the angle of asperity [17]. drilled shafts can be expressed as elastic portion (SP1), a
Eq. (2) shows that the effect of an irregularity is equivalent non-slipping or sticking state in which normal stress is
to an increase in the apparent friction angle. constant due to no dilation before a slip at the joint
Seidel and Haberfield [18] suggested, using energy interface; elasto-plastic portion (SP2), the slipping state in
considerations, that the displacement of rough disconti- which dilation of the joint interface occurs during shearing
nuities in a confined environment can be expected to cause and causes an additional normal stress and shear stress,
an increase in the normal stress, which may promote and plastic portion (SP3), the residual state in which
asperity degradation with further shearing. The dilation normal and shear stresses are maintained or reduced due to
under such conditions is expected to be lesser than that rupture of the asperities.
associated with the initial asperity angle. They showed that This paper presents the results of a laboratory investiga-
Eq. (2) may be rewritten as follows: tion into the effect of factors influencing on the shear
  behavior of regular sawtooth pile–rock joints with un-
tanðfb Þ þ tanðiÞ
tðh;CNSÞ ¼ ðsn0 þ Dsnh Þ , (3) bonded interfaces, such as roughness, normal stiffness,
1  tanðfb Þ tanðih Þ
initial normal stress, and UCS.
where t(h,CNS) is the joint shear stress at a horizontal
displacement of h, snh the corresponding normal stress, ih the 3.1. Quantification of borehole roughness
tangent to the dilation curve at a horizontal displacement of
h under CNS conditions, and sn0 the initial normal stress. Roughness plays a key role in governing the shaft
Indraratna et al. [19] carried out CNS and CNL direct resistance of rock-socketed piles. The importance of
shear tests with regular sawtooth rock joints produced by borehole roughness to shaft resistance has been well
casting plaster. Based on their results, they proposed a recognized [22–24]. Before performing a CNS test,
shear strength envelope for CNS testing of soft joints by quantitative analysis of borehole roughness should be
modifying Barton’s method [20] as follows: carried out to determine the objective roughness repre-
  "   # sented as a natural irregular profile.
t snðCNSÞ b Seidel and Haberfield [18] recommended that roughness
¼ tan fb þ io 1  , (4)
sn CNS sc be scale dependent and therefore all roughness statistics
must be accompanied by a meaningful measure of scale.
where snðCNSÞ ¼ ðsn0  K n d v =AÞ is the normal stress
Therefore, the Monash socket roughness model [18] was
corresponding to peak shear stress for a given sn0 under
applied to analyze statistics of socket roughness in this
a CNS condition, Kn the normal stiffness, dv the dilation
study. This method is based on an idealized joint interface
corresponding to peak shear stress, A the joint surface area,
as a series of interconnected chords with a constant length,
and b the surface coefficient which accounts for the
la, as shown in Fig. 2. It is assumed that chord angle y is
degradation of joints. To be more realistic, many
normally distributed with a mean my, and standard
researchers have, in addition, proposed nonlinear strength
deviation sy. Thus, the asperity heights Dr will vary in the
envelopes which can incorporate simultaneous sliding and
distribution, which can be approximated as Gaussian for
shearing of asperities, as an extension of Patton’s method.
reasonable socket roughness and can be represented as
Despite their convenience for predicting the shear
follows:
strength of a rock joint, however, all of these methods
based on the Mohr–Coulomb failure envelope have been 1X n
1X n
Dr ¼ jDri j ¼ jl a sin jðyÞjj. (5)
difficult to apply to the t–z relation for load transfer n i¼1 n i¼1
analysis because they are expressed in terms of shear and
normal stress.
Chord Length, la = Const.
Δr5
Δr1 Δr2
Δr3 Δrn
3. CNS direct shear test Pile

Since the mechanisms of shear load transfer in a rock- Δr4 Rock


socketed drilled shaft are very complicated, artificially
made pile–rock interfaces with a series of regular sawtooth Fig. 2. Monash socket roughness model [18].

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
4 H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]

Based on this approach, the natural irregular profile of a cement mortar, which consists of cement and sand, was
borehole can be simplified to a regular sawtooth pattern. substituted for concrete because concrete contains large
Here, roughness height depends on chord length, and thus, aggregates which can be difficult to fit into laboratory scale
chord length la should be determined on the basis of the test samples.
converged value of roughness height. Referring to the quantified borehole roughness, test
The quantified values of roughness reported by previous samples were molded in gypsum plaster and cement mortar
studies and measured in this study are summarized in with asperity angles of 4.61, 9.11, and 15.61, and a chord
Table 1. Seidel and Collingwood [6] present bounds on length of 25 mm. Each sample was about 175 mm long and
roughness height as a function of the UCS of intact rock 100 mm wide as shown in Fig. 3.
based on back analysis of the existing pile load tests. Nam
[25] evaluated the borehole roughness, which is constructed 3.3. Test boundary condition and procedure
separately with an auger and core barrel, of 4 sites in
clayshale and limestone using a chord length of 50 mm. Lee To study the influencing factors on the shaft resistance of
et al. [26] measured the borehole roughness of granite, rock-socketed drilled shafts, a total of 54 CNS direct shear
gneiss, sandstone, and andesite at 10 different sites tests were conducted on sawtooth samples under various
(15 boreholes) on the Korean peninsula. They report that
the representative chord length of borehole roughness in la = 25mm
Δr = 2mm
gneiss-granite is approximately 50 mm and roughness
height Dr ranges from 1 to 4 mm regardless of rock type. i = 4.6°
These ranges include measured roughness heights of 2 test
sites in this study.
Based on these results, roughness can be represented by a
regular sawtooth with a chord length la of 50 mm and
roughness heights Dr ranging from 1 to 16.2 mm, which Δr = 4mm
correspond to a roughness angles i between 1.11 and 18.91.
i = 9.1° Segment length ≈ 2·la = 50mm
3.2. Sample preparation

Natural rock blocks with quantified sawtooth roughness


are required for CNS tests. It is impossible to perform a
Δr = 7mm
large number of CNS tests under various boundary
conditions because it is difficult to prepare large rock
block samples and to make regular asperity patterns with i = 15.6°
the rock block that include discontinuities. Therefore, two
industrial gypsum plasters were used to make an idealized
sawtooth rock sample since they can be molded into any
shape when mixed with water, and the long-term strength is
Specimen length = 175mm
independent of time once the chemical hydration is
completed. Additionally, to prepare the pile sample, Fig. 3. Sawtooth specimen for CNS direct shear test (gypsum plaster).

Table 1
Quantified values of roughness for rock socket

Rock type UCS of intact rock Roughness Roughness Remark


(MPa) height (mm) angle (1)

This study Gneiss 5–50 1–7 1.1–8.0 Bit


Seidel and Claystone, sandstone, 5–10 1.7–16.2 1.9–18.9
Collingwood [6] shale, limestone, etc. 10–70 0.9–6.6 1.0–7.6 Back analysis

Lee et al. [26] Granite 100–150


Gneiss 30–130 1–4 1.1–4.6
Sandstone 75–77 RCD/all casing
Andesite 74 1–3.5 1.1–4.0
Nam [25] Clayshale 3.6–5.3 4.1–6.1 Auger
1–4 4.7–5.8 5.4–6.7 Core barrel
Limestone 3.2–3.7 3.7–4.2 Auger
10 4.3–5.1 4.9–5.8 Core barrel

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]] 5

Table 2
Summary of test boundary conditions 2 -12
τ−w
Variable Values i = 4.6 deg
i = 9.1 deg
UCS of artificial rock (MPa) 20a 35b SP2 SP3 i = 15.6 deg
1.5 -9
UCS of concrete (MPa) 42 ψ−w

Normal Displ., ψ (mm)


Roughness anglec i (1) 4.6 9.1 15.6

Shear Stress,τ (MPa)


i = 4.6 deg
Normal stiffness Kn (MPa/mm) 0.2 0.5 1.0 i = 9.1 deg
Initial normal stress, sn0 (MPa) 0.35 0.70 1.05 i = 15.6 deg
1 -6
a
Artificial rock A.
b
Artificial rock B.
c
la ¼ 25.
0.5 -3
normal stiffnesses and initial normal stresses. A summary
of test boundary conditions is given in Table 2. The normal SP1
stiffness, Kn, of a rock-socketed drilled shaft can be
0 0
determined conventionally by theoretical analysis of an
0 10 20 30 40 50
expanding infinite cylindrical cavity in an elastic half-space
Shear Displacement, w (mm)
[27] as follows:
Dsn Em 2 -6
Kn ¼ ¼ , (6) τ − σn
Dr rð1 þ nm Þ
i = 4.6 deg
where Dsn is the increased normal stress, Dr the dilation, r i = 9.1 deg
the radius of a pile, and Em and nm are the deformation i = 15.6 deg
1.5 ψ − σn
modulus and the Poisson’s ratio of the rock mass,

Normal Displ., ψ (mm)


i = 4.6 deg -4
Shear Stress,τ (MPa)

respectively. The normal stiffness used in this study varied


i = 9.1 deg
from 0.2 to 1.0 MPa/mm based on back calculation from i = 15.6 deg
general properties of the rock-socketed pile. 1
The initial normal stress is imposed on the side wall of a
rock socket by a head of wet concrete as it is placed and
depends on complex influencing factors such as the cast -2
velocity of concrete, arching effect of aggregates, hardening 0.5
rate, degree of compaction, and shrinkage rate of cement
[28]. Because it is difficult to conveniently incorporate, the
initial normal stress can simply be assumed to be a function
of the cast depth of concrete based on the theory of fluid 0 0
static mechanics. Research by Seidel and Collingwood [6] 0 0.5 1 1.5 2 2.5
Normal Stress, σn (MPa)
reports that, for sockets varying in depth from 5 to 30 m,
normal stress can be expected to range from 50 to 500 kPa. Fig. 4. Effect of roughness angle (rock specimen A, under normal stiffness
For the case of a pile socket in rock 30 m long or more 0.5 MPa/mm, initial stress 0.35 MPa). (a) Shear stress and normal
below ground, CNS tests were conducted under initial displacement with shear displacement. (b) Shear stress and normal
normal stresses of up to 1050 kPa. displacement with normal stress.

3.4. Test results and discussion displacement response and normal displacement–normal
stress response have the normal displacement plotted on
A total of 54 individual tests were conducted under the right-hand vertical axis. As the roughness angle i
various boundary conditions described in the previous increases, the increment of normal displacement increases
sections (Table 2). In this paper, only a selection of typical in the elasto-plastic portion (SP2) as shown in the normal
test results have been presented. displacement–shear displacement (c–w) response of
Fig. 4 shows the results from three typical tests on Fig. 4(a), and thus normal stress increases as illustrated
artificial rock sample A with different roughness angles in the normal displacement–normal stress (c–sn) response
(4.61, 9.11, and 15.61) under the conditions of 0.5 MPa/mm of Fig. 4(b). Consequently, the increased normal stress
normal stiffness and 0.35 MPa initial normal stress with induces increased shear stress proportionately as shown in
factors of shear stress–shear displacement (t–w), normal the shear stress–normal stress (t–sn) response of Fig. 4(b).
displacement–shear displacement (c–w), and shear stress– These results are similar to those for a roughness angle of
normal stress (t–sn), normal displacement–normal stress 9.11 under the same initial normal stress (0.35 MPa) with
(c–sn). It is noted that the normal displacement–shear three normal stiffnesses (0.2, 0.5, and 1.0 MPa/mm) as

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
6 H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]

3 τ-w ψ-w -12 2 -6


Kn = 0.2MPa/mm τ-w ψ-w
Kn = 0.2MPa/mm
σn = 0.35 MPa σn0 = 0.35 MPa
Kn = 0.5MPa/mm Kn = 0.5MPa/mm 0
2.5 σn0 = 0.70 MPa σn = 0.70 MPa
Kn = 1.0MPa/mm Kn = 1.0MPa/mm 1.6
0

-9 σn = 1.05 MPa σn = 1.05 MPa


0 0

Normal Displ., ψ (mm)


Shear Stress, τ (MPa)

Normal Displ., ψ (mm)


Shear Stress, τ (MPa)
-4
1.2
1.5 -6 Initial Peak Strengh

0.8
1 -2
-3
0.5 0.4

0 0 0 0
0 10 20 30 40 50
0 10 20 30 40 50
Shear Displacement, w (mm)
Shear Displacement, w (mm)

2.5 -6
τ - σn 1.6 -6
τ - σn
Kn = 0.2MPa/mm
σn0 = 0.35 MPa
Kn = 0.5MPa/mm
2 σn = 0.70 MPa
Kn = 1.0MPa/mm 0
1.2 σn0 = 1.05 MPa
ψ - σn
Normal Displ., ψ (mm)
Shear Stress, τ (MPa)

-4

Normal Displ., ψ (mm)


Kn = 0.2MPa/mm
Shear Stress, τ (MPa)

ψ - σn -4
1.5 Kn = 0.5MPa/mm σn = 0.35 MPa
0
Kn = 1.0MPa/mm σn0 = 0.70 MPa
0.8 σn = 1.05 MPa
0
1
-2
-2

0.5 0.4

0 0
0 0
0 0.4 0.8 1.2 1.6 2 2.4
0 0.4 0.8 1.2 1.6 2
Normal Stress, σn (MPa)
Normal Stress, σn (MPa)
Fig. 5. Effect of normal stiffness (rock specimen B, under roughness angle
9.11, initial stress 0.35 MPa). (a) Shear stress and normal displacement Fig. 6. Effect of initial stress (rock specimen A, under roughness angle
with shear displacement. (b) Shear stress and normal displacement with 4.61, normal stiffness 0.5 MPa/mm). (a) Shear stress and normal
normal stress. displacement with shear displacement. (b) Shear stress and normal
displacement with normal stress.

shown in Fig. 5. The peak shear stress tends to increase as as well as roughness angle, and normal stiffness also
normal stiffness increases. This can be explained by the fact increases. However, there are some differences in shear
that the normal stress tends to be increased by normal behavior according to these factors. Generally, the rough-
stiffness in spite of the same roughness. However, it is ness angle and normal stiffness have a significant effect on
observed that while the slopes of shear stress vs normal the shear stiffness and strength of SP2 and SP3 following
stress curves are the same regardless of normal stiffness, slip between rock and concrete. On the other hand, the
they differ slightly with different roughness angles. initial normal stress has an effect on the shear stiffness and
Fig. 6 shows the shear responses under the condition of strength of SP 1 before slipping occurs. It should also be
0.5 MPa/mm normal stiffness and three different initial noted that initial peak strengths of CNS tests are smaller
normal stresses (0.35, 0.70, and 1.05 MPa) with a roughness than those of a natural rock-socketed pile because samples
angle of 4.61. Similarly, the results demonstrate that peak of rock and concrete are molded separately, so no bonding
shear stress tends to increase with the initial normal stress exists between them. Once slip occurs, however, cohesion

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]] 7

-8
5 Artficial rock A
Kn = 0.2 MPa / mm Kn = 0.5 MPa / mm i = 4.6 deg
σn0 = 0.35 MPa σn0 = 0.35 MPa i = 9.1 deg
σn0 = 0.70 MPa σn0 = 0.70 MPa i = 15.6 deg
σn0 = 1.05 MPa σn = 1.05 MPa before Test
4 0 -6
Artficial rock B

Normal Displacement (mm)


Kn = 1.0 MPa / mm i = 4.6 deg by elastic compression
Peak Shear Stress (MPa)

σn0 = 0.35 MPa i = 9.1 deg or roughness wear


3 σn0 = 0.50 MPa i = 15.6 deg
σn0 = 1.05 MPa
-4 during Test

-2

0
0
0 10 20 30 40 50
0 10 20 30
Shear Displacement (mm)
Roughness Angle (deg)
Fig. 8. Variation of surface profiles with material strengths (under normal
stiffness 0.5 MPa/mm, initial stress 0.35 MPa).
25

displacement smaller than 25 mm because of the asperity


20 shearing.
Fig. 8 shows surface profiles and the results obtained
from six typical CNS tests with different material strengths
of artificial rocks A and B under identical normal stiffness
15 (0.5 MPa/mm) and initial normal stress (0.35 MPa). Rock
wmax (mm)

Kn = 0.2 MPa / mm sample B has higher dilation than rock sample A because
σn = 0.35 MPa less elastic compression or less asperity wear occurred
0
10 σn0 = 0.70 MPa during the shear process. Under a CNS condition, the
σn0 = 1.05 MPa higher the joint material strength is, the higher the shear
Kn = 0.5 MPa / mm strength becomes. In addition, the critical movement wmax
σn = 0.35 MPa Kn = 1.0 MPa / mm of artificial rock B, at which peak shear stress is mobilized,
0
5 σn0 = 0.70 MPa is on the whole higher than that of rock A as illustrated in
σn0 = 0.35 MPa
σn = 1.05 MPa σn0 = 0.50 MPa Fig. 9. Consequently, this higher wmax increases the dilation
0
σn0 = 1.05 MPa of joints so that peak shear strength becomes higher. This
0
implies that the strength of the weaker material between
rock and concrete is also a significant factor influencing the
0 10 20 30
shear strength of joints.
Roughness Angle (deg)

Fig. 7. Variation of peak shear stress and critical movement with 4. Proposed shear load transfer function
roughness angle for rock sample A. (a) Peak shear stress. (b) Critical
movement.
Based on the CNS tests, it is observed that the shear
stiffness of the elastic portion (SP1) depends on initial
at the interface disappears so that the bonding effects may normal stiffness, while that of the elasto-plastic portion
be ignored. (SP2) depends on both rock socket roughness and normal
Fig. 7 shows variations in peak shear stress fmax stiffness. It has also been known that a reduction in joint
(Fig. 7(a)) and critical movement wmax (Fig. 7(b)) accord- shear strength is influenced by rock mass qualities such as
ing to roughness angle, normal stiffness, and initial stress joint structure and weathering of the joint wall as
from 36 CNS tests with rock samples A. It is observed that mentioned in the previous section. The Hoek–Brown
the peak shear strength fmax increases gradually as the failure criterion [29] is more useful for estimating the
roughness angle, normal stiffness, and initial normal stress ultimate state or failure load of the rock mass, rather than
increase, but not proportionally, and occurs at a shear the Mohr–Coulomb criterion.

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
8 H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]

In this study, a new method for the shear load transfer as follows:
function of rock-socketed drilled shafts is proposed  
GSI  100
based on the Hoek–Brown failure criterion defined as mb ¼ mi exp , (8)
28
follows:
 a  
s0 GSI  100
s01 ¼ s03 þ sci mb 3 þ s , (7) s ¼ exp , (9)
sci 9
where s0 1 and s0 3 are the major and minor effective where mi is a material constant for the intact rock and
principal stresses at failure, respectively, sci is the UCS of depends only on the rock type (texture and mineralogy).
an intact rock, mb the reduced value of the material Eq. (7) expressed in terms of the major and minor principal
constant mi for a rock mass, and s and a are constants that stresses can be rewritten as a nonlinear relationship
depend on the rock mass characteristics. Based on a GSI, (see Fig. 10(a)) between shear and normal stresses as
the strength parameters of the rock mass can be calculated follows [29]:
 0 
s  stm B
t ¼ Asci n , (10)
30
sci
where t is the shear strength for which unit shaft resistance
f can be substituted, A and B are regression constants, s0n is
25 the effective normal stress, and stm the tensile strength of
the rock mass ( ¼ ssci/mb).
Once this envelope (Eq. (10)) is transferred in relation to
20 shear stress and shear displacement, the shear load transfer
function of the rock-socketed pile can be expressed as a
wmax (mm)

nonlinear triple curve consisting of three parts as shown in


15 Fig. 10(b); a linear pre-slip portion (SP1), a nonlinear slip
portion (SP2), and a post-slip portion (SP3). Three shear
mechanisms are required for each portion of the load
10 transfer function.
The parameter s0n can be obtained by summing up the
Artficial rock A initial normal stress sn0 and increments of normal stress
5
Artficial rock B (Dsn) as follows:
Upper & lower bound of artficial rock A
s0n ¼ sn0 þ Dsn ¼ sn0 þ K n ðw  wst Þ tan i, (11)
Upper & lower bound of artficial rock B
0
where sn0 can be a function of cast depth z of the concrete
0 10 20 30 based on the theory of static fluid mechanics and Dsn is
Rouhness Angle (deg)
obtained from the product of normal stiffness and normal
Fig. 9. Variation of critical movement at which peak shear stress is displacement, which is calculated from relative displace-
mobilized with roughness angle. ment of the pile–rock interface and asperity angle. Kn can
Shear stress

Shear stress

CNS Test
f = c + σ· tan

Reduced

shear strength
fmax fmax

σ'n − σtm B
f = A · σci
σci

fst fst Nonlinear triple f-w Curve


c

σn0 σn,max Normal stress wst wmax Shear displ.

SP1 SP2 SP3

Fig. 10. Proposed shear load transfer function: (a) shear stress vs normal stress curve and (b) shear stress vs shear displacement curve.

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]] 9

be determined conveniently from Eq. (6). The relative (σn + Δσn)-σtm


log [ 0 ]
displacement of the pile–rock interface is calculated by σci
subtracting the current movement w and maximum move- -1.2 -1 -0.8 -0.6 -0.4
ment wst of the elastic portion (SP1). The parameter wst is -0.4
closely related to rock mass modulus and geological Artficial rock A
conditions [8]. In this study, it lies in the range 0.5–2 mm Artficial rock B
according to the results of both CNS tests and field load
tests. Therefore, wst can be conveniently proposed to be the Y = 2.33*X + 0.55 -0.8
following linear function of GSI:
wst ¼ 2  1:5GSI=100 ðmmÞ. (13)
τmax
Consequently, the shear transfer function, considering -1.2 log
σci
the influencing factors of the shaft resistance of rock-
socketed drilled shafts, can be proposed as the following
nonlinear triple curve comprising SP1, SP2, and SP3:
-1.6
 
sn0  ððssci =mb ÞÞ B w
f ¼ Asci ðw ¼ wst Þ, (14a)
sci wst
  -2
½sn0 þ K n tan iðw  wst Þ  ððssci =mb ÞÞ B
f ¼ Asci Fig. 11. Result of regression analysis based on CNS test.
sci
ðwst ow ¼ wmax Þ, ð14bÞ
The proposed function is validated through field case
f ¼ f max ðwmax owÞ, (14c)
histories for practical estimation of parameters A and B.
where f is the unit shaft resistance, sci the weaker UCS To this end, a total of 10 large-diameter drilled shafts
between rock and concrete, and A and B are strength socked in rocks with various degrees of weathering are
parameters which depend on the GSI of the rock mass. critically analyzed. The test piles under review range
The strength parameters A and B can be obtained from 0.76–3.0 m in diameter and 6.4–43.8 m in length. Among
regression analysis. By normalizing and taking logarithms, the 10 piles, 6 tests are examined by the Osterberg-cell load
Eq. (14) will be a linear line with slope B and intercept testing method. Details of all the tests are given in Table 3,
log A as follows: where the following parameters are listed: rock type, pile
length L, pile diameter D, elevation of estimated t–z curve
Y ¼ log A þ BX , (15)
(E.L), UCS, rock mass modulus Em, rock quality
where X and Y can be expressed by designation (RQD), rock mass rating (RMR), geological
    strength index (GSI) which can be correlated with RMR,
fwst sn0  stm
Y ¼ log ; X ¼ log ðw ¼ wst Þ, roughness angle i, ultimate unit shaft resistance fmax,
sci w sci
material constant for the intact rock mi, and initial normal
(16a) stress sini.
 Based on the data transformed by the normalizing
f functions (Eq. (16)), linear regression analysis is performed
Y ¼ log ,
sci to estimate parameters A and B. As mentioned before, A
 
ðsn0 þ Dsn Þ  stm and B can be obtained by measuring the intercept and
X ¼ log slope of the regression line. Fig. 12 shows the relationship
sci
between strength parameters, A and B, and GSI. It is
ðwst ow ¼ wmax Þ. ð16bÞ
observed that values of A are constant at about 0.23 and
Fig. 11 shows the variations of peak shear stress tmax values of B range 0.48–0.82. Parameter B tends to decline
against normal stress in transformed coordinates (Eq. (15)) logarithmically as the GSI of the rock mass increases
based on the results obtained in a total of 54 CNS tests. In (y ¼ 0.293ln x+0.593), and can be approximated for the
Fig. 11, GSI and s were applied as 100 and 1, respectively, sake of simplicity by a bi-linear curve as follows:
because test samples are considered to be intact rock, and
B ¼ 0:008GSI þ 0:94 ðGSIo45Þ, (18a)
the value of mi was determined directly from rock triaxial
compressive tests on the intact rock. It can be concluded
B ¼ 0:002GSI þ 0:67 ðGSI ¼ 45Þ. (18b)
that the proposed function suitably represents peak shear
strength of joints, and can therefore properly predict the Finally, the proposed shear transfer function of drilled
shear strength of regular sawtooth joints with respect to shafts socketed in rocks can be obtained by substituting A
their roughness, normal stiffness, and initial normal stress. and B into Eq. (14).

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
10 H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]

Table 3
Material properties and input parameters

Site Pile Rock L D E.L UCS Em RQD RMR GSIb i fmax mid
no. typea (m) (mm) (m) (MPa) (MPa) (deg) (kPa)

Gyeonggi D2 Gneiss (CW) 13.5 1000 12.9 48 905 0 22 17 4.6c 4670 33


D3 Gneiss (CW) 13.5 1000 12.2 48 974 0 22 17 4.6c 720 33
D4 Gneiss (HW) 13.5 1000 12.1 48 1203 9 31 26 4.6c 41100 33
Gneiss (MW) 12.9 48 1932 40 42 37 1600
D5 Gneiss (MW) 13.5 1000 12.9 48 2748 52 45 40 4.6c 1830 33
f c
Incheon W8 Granite (MW) 45.1 2400 44.5 35 2130 8 – 40 4.6 1400 33
47.9 35 2300 – 45f 1750
E7 Granite (MW) 40.0 2400 49.0 30 1480 18 – 35f 4.6c 1400 33
50.7 30 1930 – 45f 1720
E5 Granite (MW) 40.1 3000 49.5 54 1630 25 – 45f 4.6c 2370 33
51.5 54 1300 – 40f 1950
52.5 54 1300 – 45f 1630
Houston H Clayshale (SW) 3.0e 760 9.2 1 70 95 – 95f 4.7 102 9
D Clayshale (SW) 5.8e 760 7.0 4 200 82 – 90f 6.2 405 9
R Limestone (SW) 4.5e 760 4.0 10 900 88 – 95f 3.7 1545 10
a
CW: completely weathered; HW: highly weathered; MW: moderately weathered; SW: slightly weathered.
b
GSI ¼ RMR76 ¼ RMR895 (where, RMR76418, RMR89423).
c
The value of the Hoek–Brown constant mi for intact rock.
d
Moderate magnitude of hole roughness angle (where, Dr ¼ 4 mm, la ¼ 50 mm).
e
General reference value presumed without performing field tests.
f
Length of test socket.

1 D2 D4 D5

Coefficient A
Coefficient B
0.8

Fill Fill Fill

Y = -0.293*ln (X) + 0.593


0.6 - 10.4m - 10.6m
- 11.7m
Res.Soil
A, B

- 12.5m Gneiss [HW] - 12.7m Res. Soil - 12.4m

- 13.8m Gneiss [MW] - 13.8m Gneiss [MW] - 13.8m


Gneiss[HW]
0.4
1m 1m 1m
Y = 0.23
Fig. 13. Subsurface profile and shaft embedments for test piles
0.2 (Gyeonggi).

0
test piles), only comparisons for nine t–z curves are
included in this paper.
0 20 40 60 80 100
Geological strength index
5.1. Gyeonggi case
Fig. 12. Strength constants, A and B, according to GSI.
The load-transfer characteristics of three instrumented
drilled shafts (D2, D4, and D5) reported by Kwon [30]
5. Validation with case histories are compared with the predicted values of the proposed
shear load-transfer function. These piles were founded
The proposed function was validated by comparing the in completely to moderately weathered gneiss. Fig. 13
present approach and other methods [7,8] with typical shows an idealization of the subsurface profile and shaft
results of three case histories as summarized in Table 3. embedments for the test piles. All of the test piles are
Although predictions of shear load transfer have been 1000 mm in diameter and 13.8 m in length. The layers
carried out for all tests (which include fifty t–z curves of ten and material properties of the piles were chosen to

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]] 11

shaft resistance did not reach the ultimate state and


800 continued to increase as the displacement increased.
This observation agrees with the observation of Williams
et al. [31], who report that the more rock mass is
Unit shaft resistance (kPa)

600 weathered, the more shaft displacement occurs before the


ultimate state is reached, so that propagation failure occurs
gradually.
400 The proposed function generally does a better job of
predicting the measured shaft resistance than other load
transfer functions. In particular, the proposed method is in
200 Measured (CW) reasonably good agreement with t–z curves measured by
Pred. This Study (CW) loading tests in highly weathered rock.
Pred. O'Neill & Hassan
Pred. Baguelin
0 5.2. Incheon case
0 4 8 12 16 20
Displacement (mm) The load-transfer characteristics of three large-diameter
instrumented drilled shafts (W8, E7, and E5) were
compared to the proposed shear load-transfer function.
1200
These piles were installed in moderately weathered granite
for real construction projects in Incheon.
Unit shaft resistance (kPa)

Fig. 15 shows the subsurface profile and shaft


embedments of the test piles. Both test piles W8 and E7
800
are 2400 mm in diameter and 45.1 and 40.0 m in
length, respectively. Test pile E5 is 3000 mm in diameter
and 40.1 m in length. All test piles were performed
using a bi-directional loading test (Osterberg cell test)
400
suited for the high bearing capacity of a large-diameter
Measured (HW) pile. They were loaded sufficiently to quantify the ulti-
Pred. This Study (HW)
Pred. O'neill & Hassan mate shaft resistance fmax. Applied loads of test piles
Pred. Baguelin W8, E7, and E5 were 240, 170, and 284 MN, respectively.
0
The 284 MN measurement of test pile E5 corresponds
0 4 8 12 16 20
to a new world record for loading using the Osterberg
Displacement (mm)
cell test.
Fig. 16 shows the predicted and measured t–z curves of
2000
test piles. It is observed that initial slopes of t–z curves
gradually decrease as the displacement of the shaft
Unit shaft resistance (kPa)

increases, and thus, f approaches an ultimate shaft


1500
resistance fmax hyperbolically. The proposed shear load
transfer function of this study predicts the general trend of
the measured t–z curves fairly well.
1000

W8 E7 E5
500 Measured (MW)
Pred. This Study (MW) Sea bed - 5.5m
Pred. O'Neill & Hassan - 6.1m - 11.2m
Pred. Baguelin Sea bed - 14.1m
Sea bed
Marine - 12.3m
0 Deposit - 15.3m

0 4 8 12 16 20 - 36.0m Marine
Marine
Deposit
Granite [HW] Deposit
Displacement (mm) - 43.2m
- 38.7m
- 51.2m - 41.1m
Fig. 14. Shear load transfer function (Gyeonggi). (a) Pile D2. (b) Pile D4. Granite [MW] Granite [HW] - 48.2m Granite [HW] - 48.1m
(c) Pile D5. - 61.5m
- 52.3m Granite [MW]
Granite[MW]
- 54.1m
Granite Granite - 54.7m Granite
[Hard] [Hard] [Hard]
O-Cell

represent the pile–rock interface based on the results of a


soil investigation.
2.4 m 2.4 m 3.0 m
Fig. 14 shows the predicted and measured t–z curves
for the test piles. It was observed that the measured Fig. 15. Subsurface profile and shaft embedments for test piles (Incheon).

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
12 H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]

H D R

2000 Silty Sand - 2.2m - 2.3m


Clay - 1.7m
Clay - 2.8m - 2.7m
Sand Limestone
- 4.3m [MW] - 3.1m
1500
Shaft resistance (kPa)

- 6.1m Sand

- 7.7m T - 5.9m
Clay T
- 5.6m
T O-Cell
- 8.4m O-Cell
1000 Clayshale -10.8m
Clay shale Limestone
O-Cell
R
-14.1m R R
- 11.2m - 7.2m

500 Mesaured (MW, E.L - 47.9m)


Pred. This Study
Pred. O'neill & Hassan
Pred. Baquelin 0.76 m 0.76 m 0.76 m
0
0 4 8 12 16 20 Fig. 17. Subsurface profile and shaft embedments for test piles (Houston).

Displacement (mm)
compared with the present approach. These piles were
founded in slightly weathered clayshale and limestone. The
1500
loading tests of all test piles consisting of test and reaction
sockets were performed using the Osterberg cell. The
reaction socket is an instrument designed to prevent end-
Shaft resistance (kPa)

1000
bearing failure and thus estimate ultimate shaft resistance
of the test socket. Fig. 17 shows an idealization of the
subsurface profile and shaft embedments for test piles. Test
piles H, D, and R are all 760 mm in diameter and their test
500
sockets are 3.0, 5.8, and 4.5 m in length, respectively.
Measured (MW, E.L - 49.0m)
Fig. 18 shows the predicted and measured t–z curves for
Pred. This Study the test piles. The proposed method predicts well the
Pred. O'neill & Hassan
Pred. Baquelin general trend of the measured t–z curves of the test piles, H
0
and D, in clayshale. In the case of test pile R, however,
0 4 8 12 16 20 good agreement between the predicted curve by the
Displacement (mm) proposed method and the measured t–z curve was obtained
in the elastic portion (SP1), wile in the elasto-plastic
portion (SP2), the proposed method slightly underesti-
2500
mates shaft resistance in comparison with test results.
Discrepancy between the predicted curve and the measure-
2000 ment curve results from the fact that the rock mass
Shaft resistance (kPa)

modulus Em of limestone, in which test pile R was


1500 socketed, was underestimated despite the observation that
minimal weathering of the rock mass had occurred.
1000
6. Conclusion
Meausred (MW, E.L - 49.5m)
500 Pred. This Study The main objective of this study was to propose a new
Pred. O'Neill & Hassan shear load transfer function of rock-socketed drilled shafts
Pred. Baquelin
0 which can consider various factors influencing shaft
0 10 20 30 40 50 resistance. The proposed shear load transfer function can
Displacement (mm) consider borehole roughness and the geological strength
index (GSI) which indicates the discontinuity and surface
Fig. 16. Shear load transfer function (Incheon). (a) Pile W8. (b) Pile E7. condition of the rock mass. Through comparisons with
(c) Pile E5.
case histories, it is found that the proposed load-transfer
function is in good agreement with in situ measurements.
From the findings of this study, the following conclusions
5.3. Houston case are drawn:

The load-transfer characteristics of three instrumented (1) Based on results of CNS tests, the shear load transfer
drilled shafts (H, D, and R) reported by Nam [25] are curve of rock-socketed drilled shafts can be expressed

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]] 13

resistance. Peak shear strength and shear stiffness of the


120 elastic portion (SP1) increase as initial normal stress
and UCS increase, whereas those of the elasto-plastic
portion (SP2) increase as roughness angle, normal
90 stiffness, and UCS increase.
Shaft resistance (kPa)

(2) By taking into account various influencing factors of


shaft resistance, including borehole roughness and
60 weathering of the rock mass, the new function
proposed (a nonlinear triple curve) is appropriate and
realistically represents the shear load transfer charac-
30 Measured (SW) teristics of a drilled shaft socket in a rock mass.
Pred. This Study
Pred. O'Neill & Hassan (3) It has been shown that the more a rock mass is
Pred. Baguelin weathered, the greater the displacement of the shaft
0 needed to reach the ultimate state and the more the
0 5 10 15 20 initial slope of the t–z curve declines. Compared with
Displacement (mm) the results of the case histories, it is clear that the
proposed method can better predict shear behavior,
including initial slope, of a drilled shaft in rocks with
500 various degrees of weathering, than other methods.

400
Shaft resistance (kPa)

References
300 [1] Osterberg J. The Osterberg load test method for bored and driven
piles the first ten years. In: Proceedings of the seventh international
conference on piling and deep foundations, Vienna; 1998.
200 p. 1.28.1–1.28.11.
[2] Reese LC, O’Neill MW. Drilled shafts: construction procedures and
Measured (SW) design methods. Washington: US Federal Highway Administration;
100 Pred. This Study 1988.
Pred. O'Neill & Hassan [3] Ghionna VN, Jamiolkowski M, Pedroni S, Salgado R. The tip
Pred. Baguelin
0 displacement of drilled shafts in sands. Vertical and horizontal
0 5 10 15 20 deformations of foundations and embankments. Géotechnique 1994;
2:1039–57.
Displacement (mm) [4] Horvath RG, Kenny TC, Kozicki P. Method of improving the
performance of drilled piers in weak rock. Can Geotech J 1983;
2000 20:758–72.
[5] O’Neill MW, Townsend FC, Hanssan KM, Buller A, Chan PS. Load
transfer for drilled shafts in intermediate geomatrials. Washington:
1600 US Department of Transportation; 1995.
Shaft resistance (kPa)

[6] Seidel JP, Collingwood B. A new socket roughness factor for


prediction of rock socket shaft resistance. Can Geotech J 2001;
1200 38:138–53.
[7] Baguelin F. Rules for the structural design of foundations based on
the selfboring pressuremeter test. In: Proceedings of the symposium
800 on pressuremeter and marine application, IFP, Paris; 1982.
p. 347–62.
Measured (SW) [8] O’Neill MW, Hassan KM. Drilled shaft: effects of construction on
400 Pred. This Study performance and design criteria. In: Proceedings of the international
Pred. O'Neill & Hassan
Pred. Baguelin conference on design construction of deep foundations. Washington:
US Federal Highways Administration; 1994. p. 137–87.
0 [9] Kim SI, Jeong SS, Cho SH, Park IJ. Shear load transfer
0 10 20 30 40 50 characteristics of drilled shafts in weathered rock. J Geotech
Displacement (mm) Geoenviron Eng 1999;125:999–1010.
[10] Coyle HM, Reese LC. Load transfer for axially loaded piles in clay.
Fig. 18. Shear load transfer function (Houston). (a) Pile H. (b) Pile D. (c) J Soil Mech Found Div ASCE 1966;92:1–26.
Pile R. [11] Poulos HG, Davis EH. The settlement behavior of single axially
loaded incompressible piles and piers. Géotechnique 1968;18:351–71.
[12] Mattes NS, Poulos HG. Settlement of single compressible pile. J Soil
by a nonlinear triple curve composed of an elastic Mech Found Div ASCE 1969;95:189–207.
[13] Kraft LM, Ray RP, Kagawa T. Theoretical t–z curves. J Geotech
(SP1), elasto-plastic (SP2), and plastic portion (SP3). It Eng Div ASCE 1981;107:1543–61.
has been shown that the shear behavior of the three [14] Castelli F, Maugeri M, Motta E. Analisi non lineare del cedimento di
portions depends on each of the factor influencing shaft un Palo Singolo. Riv Ital Geotech 1992;26:115–35.

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008
ARTICLE IN PRESS
14 H. Seol et al. / International Journal of Rock Mechanics & Mining Sciences ] (]]]]) ]]]–]]]

[15] Vijayvergiya VN. Load-movement characteristics of piles. In: [23] Johnston IW, Lam TSK, Williams AF. Constant normal stiffness
Proceedings of the fourth annual symposium on waterways, ASCE, direct shear testing for socketed pile design in weak rock.
Long Beach, CA; 1977. p. 269–84. Géotechnique 1987;37:83–9.
[16] Johnston IW. Movement of foundations on rock. In: Vertical and [24] Horvath RG, Kenny TC. Shaft resistance of rock-socketed drilled
horizontal deformations of foundations and embankments, vol. 2. piers. In: Proceedings of the symposium of deep foundations, ASCE,
Geotech spec pub no 40, 1994. p. 1703–17. Atlanta; 1979. p. 183–214.
[17] Patton FD. Multiple models of shear failure in rock. In: Proceedings [25] Nam MS. Improved design for drilled shafts in rock. PhD
of the first congress of the International Society of Rock Mechanics, dissertation, University of Houston, 2004.
Lisbon; 1966. p. 509–13. [26] Lee MH, Cho CH, Yoo HK, Kwon HK. A study on the surface rough-
[18] Seidel JP, Haberfield CM. Towards an understanding of joint ness of drilled shaft into rock in Korea. In: Proceedings of the Korean
roughness. Rock Mech Rock Eng 1995;28:69–92. Society of Geotechnical Engineering conference, Seoul; 2003. p. 431–8.
[19] Indraratna B, Haque A, Aziz N. Laboratory modeling of shear [27] Boresi AP. Elasticity in engineering mechanics. Englewood Cliffs, NJ:
behaviour of soft joints under constant normal stiffness conditions. Prentice-Hall; 1965.
Geotech Geol Eng 1998;16:17–44. [28] Taylor WH. Concrete technology and practice. 1st ed. Sydney: Angus
[20] Barton N. Review of a new shear strength criterion for rock joints. and Robertson; 1965.
Eng Geol 1973;7:287–332. [29] Hoek E, Brown ET. Practical estimates of rock mass strength. Int J
[21] Johnston IW, Lam TSK. Shear behaviour of regular triangu- Rock Mech Min Sci 1997;34:1165–86.
lar concrete/rock joints—analysis. J Geotech Eng 1989;115: [30] Kwon OS. Effect of rock mass weathering on resistant behavior of
711–27. drilled shaft socketed into weathered rock. PhD dissertation, Seoul
[22] Pells PJN, Rowe RK, Turner RM. An experimental investigation University, 2004.
into side shears for socketed piles in sandstone. In: Proceedings of the [31] Williams AF, Johnston IW, Donald IB. The design of socketed piles
international conference on structural foundations on rock, Sydney; in weak rock. In: Proceedings of the international conference
1980. p. 291–302. structural foundations on rock, Sydney; 1980. p. 327–47.

Please cite this article as: Seol H, et al. Shear load transfer for rock-socketed drilled shafts based on borehole roughness and geological strength
index.... Int J Rock Mech Mining Sci (2007), doi:10.1016/j.ijrmms.2007.09.008

Das könnte Ihnen auch gefallen