-
Draft
Evaluation of the peak side resistance for rock socketed shafts
in weak sedimentary rock from an extensive
database of published field load tests: a limit state
approach
Journal: Canadian Geotechnical Journal
Manuscript ID cgj-2018-0590.R1
Manuscript Type: Article
Date Submitted by the Author: 02-Oct-2018
Complete List of Authors: Asem, Pouyan; University of Illinois
at Urbana-Champaign, Department of Civil and Environmental
EngineeringGardoni, Paolo; University of Illinois at
Urbana-Champaign
Keyword: Peak side resistance, weak rock mass, load test
database, resistance factor, rock socket
Is the invited manuscript for consideration in a Special
Issue? :Not applicable (regular submission)
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
1
Evaluation of the peak side resistance for rock socketed shafts
in weak sedimentary rock
from an extensive database of published field load tests: a
limit state approach
Pouyan Asem, Ph.D., A.M. ASCE, M. JGS Adjunct lecturer in Civil
and Environmental Engineering
University of Illinois at Urbana-Champaign 205 N. Mathews
Ave.
Urbana, IL 61801 [email protected]
and
Paolo Gardoni, Ph.D.
Professor of Civil and Environmental Engineering University of
Illinois at Urbana-Champaign
205 N. Mathews Ave. Urbana, IL 61801
[email protected]
Corresponding author: Pouyan Asem
January 19, 2019
Page 1 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
2
ABSTRACT
This paper presents the analyses of the measured peak side
resistance of rock sockets constructed
in weak claystone, shale, limestone, siltstone, and sandstone.
The peak side resistance is obtained
from the in situ axial load tests on drilled shafts, anchors,
and plugs. The parameters that affect
the development of the peak side resistance are determined using
the in situ load test data. It is
found that the peak side resistance increases with the
unconfined compressive strength and the
deformation modulus of the weak rock, and decreases with
increase in the length of the shear
surface along the rock socket sidewalls. The increase in the
socket diameter also slightly
decreases the peak side resistance. Additionally, it is found
that the initial normal stresses do not
significantly affect the measured peak side resistance in the in
situ load tests. The in situ load test
data are used to develop an empirical design equation for the
determination of the peak side
resistance. The proposed model for the peak side resistance, and
the reliability analysis are used
to determine the corresponding resistance factors for use in the
load and resistance factor design
framework for the assessment of the strength limit state.
KEYWORDS
Peak side resistance, weak rock mass, load test database,
resistance factor, rock socket.
Page 2 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
3
INTRODUCTION
The near-surface soils usually have low shear strength, are
compressible, and usually do not offer
the required bearing capacity to carry the loads of the heavy
structures. In such conditions, a
shallow footing where the ratio of the embedment depth (measured
from the ground surface DGS)
to the foundation diameter (B) is less than 2 (Terzaghi 1943)
may not be a feasible foundation
solution. Accordingly, a deep foundation (DGS/B > 4, after
Peck et al. 1974; Coduto 2001) is
typically used to transfer the load of the heavy structures to
the more competent strata, which are
often found at greater depths. Among the available deep
foundation options, drilled shafts are
frequently used, and are typically extended through the upper
incompetent and compressible
residual soils into the underlying weak rocks (Rosenberg and
Journeaux 1976; Horvath and
Kenney 1979; Rowe and Armitage 1987; Kulhawy and Phoon 1993;
Hassan and O’Neill 1997;
Seidel and Collingwood 2001; Vu 2013; Asem 2018) to support the
heavy loads. Weak rocks are
commonly encountered in the zone between the near-surface
residual soils, and the more
competent and unweathered bedrock at greater depths. Socketing
drilled shafts into weak rocks
has been used more frequently in the recent years (Horvath and
Kenney 1979; Horvath and Chae
1989; Hassan and O’Neill 1997) because weak rocks often provide
the necessary bearing layer to
support the heavy loads, and at the same time, result in more
economical and effective
construction compared to unweathered rocks found at greater
depths.
The base resistance, the side resistance, or a combination of
both contribute to the overall axial
resistance of the drilled shafts that are embedded (socketed) in
weak rock formations (Horvath
and Kenney 1979; Zhang and Einstein 1998; Lo and Hefny 2001;
Seidel and Collingwood 2001).
The in situ load test measurements, and the analyses based on
the theory of elasticity, however,
Page 3 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
4
show that a substantial portion of the axial service loads on
the rock sockets is carried using the
side resistance (Osterberg and Gill 1973; Horvath and Kenney
1979; Goodman 1980) because
the side resistance is mobilized at significantly smaller axial
displacements compared to the base
resistance (Carter and Kulhawy 1988; Seidel and Collingwood
2001). Therefore, when designing
a rock socketed drilled shaft in weak rock, it is necessary to
accurately determine the load
transfer to the rock socket sidewalls. The maximum shear stress
mobilized along the rock socket
sidewalls is represented by the peak side resistance (fsp). The
peak side resistance (fsp) may be
measured using the in situ axial load tests (Rosenberg and
Journeaux 1976; Horvath and Kenney
1979; Williams 1980; Rowe and Armitage 1987; Abu-Hejleh et al.
2003; Miller 2003; Kulhawy
et al. 2005; Vu 2013). An in situ load test, however, is costly
and time consuming. Therefore,
predictive models, which are commonly developed based on the
results of in situ load tests,
laboratory interface tests, or numerical models, have been used
to estimate the peak side
resistance (fsp) (Horvath and Kenney 1979; Rowe and Armitage
1987; Kulhawy and Phoon
1993; Hassan and O’Neill 1997; Seidel and Collingwood 2001; Vu
2013).
The work reported by Williams (1980), Williams and Pells (1981),
and Seidel and Collingwood
(2001) showed that in addition to the unconfined compressive
strength (qu), parameters such as
the rock mass deformation modulus (Em) may also be needed to
obtain a more accurate estimate
of the values of peak side resistance (fsp) for rock sockets
embedded in weak rocks. In the
following sections, we first develop a database consisting of
axial load tests on drilled shafts,
anchors, and plugs that were constructed in weak rocks. In the
drilled shaft and some of the plug
load tests, the components of side and base resistances were
separated using strain gauges that
Page 4 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
5
were placed along the rock socket sidewalls, or using a load
cell at the base of the drilled shafts
or the plugs. Occasionally, a compressible base or a void was
created near the base of the drilled
shafts or plugs to eliminate the base resistance to directly
measure the side resistance. In the
anchor load tests, no component of the base resistance was
present, and thus the side resistance
was measured directly. The load testing methods will be
discussed in more detail in the
subsequent sections of the paper. These load test data are then
used to investigate the parameters
that affect the peak side resistance (fsp) in the rock sockets
that are constructed in weak rocks.
Next, the method of Gardoni et al. (2002), and the in situ load
test data are used to develop a
predictive model for the peak side resistance (fsp) of the rock
sockets in weak rocks. A reliability
analysis is then used to determine the corresponding resistance
factors (φ) for use in the Load
and Resistance Factor Design (LRFD) framework.
DEFINITION OF THE WEAK ROCK
The stress relief in the rock masses that follows the removal of
the overburden geomaterials by
the glacial movements and other similar geologic events causes
the opening of the structural
discontinuities such as joints and fissures that may be
originally closed (Terzaghi 1936), and
subsequently results in increase of the water content (w) of the
rock materials exposed on the
face of open joints and fissures by absorbing water (Terzaghi
1936; Skempton 1948; Skempton
and Hutchinson 1969; Mesri and Shahien 2003). The increase in
the water content (w) leads to
the swelling of the rock materials under zero pressure along the
exposed walls of the open cracks
(Terzaghi 1936), and it results in the softening and
deterioration of the rock blocks that are
separated by the said structural discontinuities. The resulting
softening causes additional
movements in the rock formation, and more fissures and joints
will open (Terzaghi 1936;
Page 5 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
6
Skempton 1948), which further reduces the shear strength and
increases the compressibility of
the rock mass.
It follows from the discussion above that a weak rock mass is an
assemblage of the relatively
weathered and softened rock blocks, which are separated by the
individual structural
discontinuities (Hoek 1983; Singh and Rao 2005). The weathered
rock block refers to the
“unfractured blocks which occur between structural
discontinuities in a typical rock mass” (Hoek
1983). The structural discontinuities include “fold structures,
joints, faults, shears and
slickensided surfaces” (Hoek 1983; Goodman 1980; Goodman
1993).
Because a weak rock mass consists of relatively weathered and
softened rock blocks and
structural discontinuities that are also subjected to the
weathering and softening effect, and both
components affect the response of the weak rock to external
loads, the definition of weak rock
should reflect how weathering and softening affects the rock
blocks and the discontinuity
surfaces. Accordingly, the values of qu that are obtained from
the representative samples are
used to characterize the effect of the mineralogy and the
weathering state of the rock blocks, and
the Geological Strength Index (GSI) is used to reflect the
blockiness of the rock mass, and the
degree of the alteration of the natural structural
discontinuities such as joints and fissures. The
review of the technical literature suggests that for weak rocks,
the values of the unconfined
compressive strength (qu) ranges from 0.5 MPa to slightly
greater than 30 MPa (Deere and
Miller 1966; Barton et al. 1978; Rowe and Armitage 1987;
Cepeda-Diaz 1987; Kanji 2014), and
the geological strength index (GSI) is commonly less than 70
(Hoek and Brown 1997).
Page 6 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
7
METHODS FOR PREDICTING PEAK SIDE RESISTANCE IN ROCK SOCKETS
Methods based on in situ load tests
Rosenberg and Journeaux (1976), Horvath and Kenney (1979),
Williams (1980), Rowe and
Armitage (1984; 1987), Carter and Kulhawy (1988), Abu-Hejleh et
al. (2003), Miller (2003),
Kulhawy et al. (2005), Vu (2013), and the Canadian Foundation
Engineering Manual (2006)
(which recommend the methods of Horvath et al. 1983; Rowe and
Armitage 1984; Carter and
Kulhawy 1988) introduced predictive models for the peak side
resistance (fsp) of rock sockets in
weak rocks. These investigators related the measured peak side
resistance (fsp) of rock socket
sidewalls in the in situ axial load tests on drilled shafts,
anchors, and plugs to the unconfined
compressive strength (qu) of the rock. These models use the
following general mathematical
form
(1)
where α and n are obtained by fitting Equation (1) to the load
test data. Table 1 summarizes the
empirical parameters α and n for some of the existing models (α
and n in Table 1 are obtained
from O’Neill et al. 1996; Seidel and Collingwood 2001; Canadian
Foundation Engineering
Manual 2006; Asem 2018). The following observations are
noteworthy in relation to the models
based on the in situ load tests
1. Some of the existing models were developed based on a limited
number of sites with
similar rock formations and rock socket properties, which
covered different ranges of
rock weathering conditions. As a result, such models may be
considered as site- or rock
Page 7 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
8
type-specific. Examples include methods of Williams (1980) that
used the data from six
sites (i.e., Stanley Avenue, Middleborough Road, Westgate
Freeway, Eastern Freeway,
Flinders Street, and Johnson Street) in Melbourne Siltstone,
Abu-Hejleh et al. (2003) that
used the data from four sites (i.e., I-225, County Line,
Franklin, and Broadway) in
Colorado claystone, sandstone, and siltstone, Miller (2003) that
used the data from three
sites (i.e., Lexington, Grandview Triangle, and Waverly) in
Missouri shale, and Vu
(2013) that used the data from two sites (i.e., Frankford, and
Warrensburg) in Missouri
shale. Due to their site-specific nature, these methods may need
to be re-calibrated using
additional data before they can be used to predict the peak side
resistance (fsp) of the rock
sockets sidewalls that are constructed in other rock formations
with different rock mass
characteristics.
2. Some design methods relate the values of fsp to only qu of
the weak rocks (Rosenberg
and Journeaux 1976; Abu-Hejleh et al. 2003; Miller 2003; Kulhawy
et al. 2005; Stark et
al. 2013; Vu 2013) do not account for the effects of (i) the
secondary structure of the rock
mass such as joins and slickensided surfaces, and (ii) the rock
socket geometry such as
diameter (B) and the length of the shear surfaces (L) that form
along the rock socket
sidewalls, on the mobilized fsp. Williams and Pells (1981) and
Seidel and Collingwood
(2001) showed that the peak side resistance of the rock socket
sidewalls is also affected
by other parameters such as rock socket sidewall roughness and
rock mass fracturing.
3. The existing load test data for the peak side resistance
(fsp) of the rock socket sidewalls in
weak rocks (e.g., data presented in Fig. 6.1 of Stark et al.
2013) show that the relationship
between the values of fsp and qu for the weak rocks could best
be described using a linear
Page 8 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
9
function where the empirical parameter n = 1. Yet, some of the
existing methods use an
empirical parameter n that is less than unity (i.e., n <
1).
Methods based on laboratory tests or numerical models
In this approach (Hassan 1994; Hassan and O’Neill 1997; Seidel
and Collingwood 2001;
Wainshtein et al. 2008), the design models were developed based
on (i) the laboratory constant
normal stress or constant normal stiffness direct shear tests on
natural- or artificial-rock/concrete
interfaces (e.g., Seidel 1993; Seidel and Collingwood 2001), or
(ii) the numerical models (e.g.,
Hassan and O’Neill 1997). These methods commonly relate the
shear strength of the interface to
(i) the frictional properties of the interface such as interface
friction angle (φint), and the initial
average roughness height (∆r) (e.g., Seidel 1993; Seidel and
Collingwood 2001), (ii) initial
normal stress (σno) acting on the interface (e.g., Hassan and
O’Neill 1997), and (iii) the rock
mass deformation modulus (Em) (e.g., Seidel and Collingwood
2001; Hassan and O’Neill 1997).
The following should be noted in relation to the models based on
laboratory tests or numerical
models
1. Some of the existing predictive methods were developed based
on the constant normal
stiffness (CNS) direct shear tests on rock/concrete interfaces.
Some aspects of these tests
are noteworthy. For examples use of a CNS direct shear test
method by Seidel (1993) is
more realistic than previously used traditional constant normal
stress direct shear tests,
and it appears that it can properly model the dilation of the
interface and the resulting
changes in the normal stresses on the mobilized shear surface.
Recent studies by the
Author shows that the fsp measurements from constant normal
stiffness direct shear tests
Page 9 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
10
reported by Seidel (1993) are in reasonable agreement with fsp
measurements in the
drilled shaft, plug, and anchor load tests in weak rocks.
2. Seidel and Collingwood (2001) proposed a method for
prediction of fsp that accounts for
the average rock/concrete interface roughness height (∆r) in
rock sockets. Seidel and
Collingwood (2001) also proposed an approach for prediction of
the ∆r that is shown in
Fig. 1 as upper- and lower-bound solid lines. Figure 1 also
presents measured ∆r and qu
data that are reported in Collingwood (2000). While the proposed
model of Seidel and
Collingwood (2001) is in good agreement with measured rock
socket sidewall roughness
data, a wide range of ∆r is possible for any given value of
qu.
3. The model of Hassan and O’Neill (1997) relates the normalized
peak side resistance (i.e.,
fsp/qu) to (i) the corresponding values of the unconfined
compressive strength (qu) of the
weak (or “soft” according to Hassan and O’Neill 1997) rock, (ii)
to the initial normal
stress (σno) on the rock socket sidewalls, and (iii) to the
interface friction angle (φint).
While the values of σno in each drilled shaft, plug, and anchor
may be reasonably
estimated using the hydrostatic pressure of fresh concrete,
Vesic (1963) and, the data
presented in this paper (see Fig. 9) suggest that the values of
the σno and fsp are not
strongly related. This is because the normal stresses (acting on
the rock socket sidewalls)
at the time of mobilization of fsp are not necessarily the same
as the values of the σno
because the rock socket often (with some exceptions such as
carbonate rocks) dilates
against the normal stiffness (Kn, Equation 4) of the weak rock
mass.
IN SITU LOAD TEST DATA
Existing databases
Page 10 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
11
The existing load test databases (e.g., Horvath and Kenney 1979;
Williams 1980; Rowe and
Armitage 1984; Hassan 1994; Collingwood 2000; Stark et al. 2013)
were used in the
development of some of the relationships for the prediction of
the peak side resistance (fsp) in the
rock sockets. The predictive models developed based on these
databases, with the exception of
methods such as Williams and Pells (1981) and Seidel and
Collingwood (2001), often relate the
measured peak side resistance (fsp) of the rock socket sidewalls
to the unconfined compressive
strength (qu) of rock measured using the specimens obtained from
the adjacent boreholes. The
databases of Horvath and Kenney (1979), Williams (1980), Rowe
and Armitage (1984), Hassan
(1994), and Collingwood (2000) are summarized in Table 2. One or
more of the following
parameters were not reported in the existing databases: (i) the
rock mass properties such as the
deformation modulus (Em), and (ii) the rock socket
characteristics, such as the depth of
embedment from the ground surface and top of rock (DGS or DTOR,
respectively) and the length
of the shear surface (L) mobilized along the rock socket
sidewalls. Therefore, the effect of these
missing parameters on the peak side resistance (fsp) cannot be
readily examined using the such
databases.
Updated load test database
The updated database consists of the in situ axial load tests on
rock sockets in weak rocks, which
are found in the literature. The rock sockets analyzed are a
part of the in situ load tests on drilled
shafts, anchors, and plugs in weak rock masses. The database was
first reported by Asem (2018),
and is summarized in the Supplemental Data section. In addition
to the length of the shear
surface formed along the rock socket and the rock socket
diameter (L and B, respectively), the
database summarizes the load test and rock socket excavation
methods used at each load test site,
Page 11 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
12
the estimated deformation modulus of the rock mass (Em), the
initial shear stiffness (Ksi)
mobilized along the shear surfaces formed on the rock socket
sidewalls, the unconfined
compressive strength (qu) of the rock, the measured peak side
resistance (fsp), and the
corresponding vertical socket displacement (δp). The database is
briefly discussed blow.
1. Deformation modulus of weak rock mass (Em): the value of Em
is estimated using the
slope of the tangent line drawn to the initial part of the
relationship between the shear
stress (fs) and axial rock socket displacement (δ) (Fig. 2) that
were measured for each
load test. The initial slope of the fs-δ relationship is
represented by the initial shear
stiffness (Ksi). Equation (2), which is based on the theory of
elasticity, may be used to
relate the values of Ksi to Em that will eventually be used to
estimate the values of Em
(2)
In Equation (2), P is the load on the rock socket, and I is an
embedment influence factor
that is obtained from Pells and Turner (1979). Dividing both
sides of Equation (2) by the
circumferential area of the shear plane ( ), we obtain Equation
(3) for the
estimation of Em based on the in situ load test measurements
(3)
Page 12 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
13
Figure 3 shows the estimated values of Em using Equation (3)
that are plotted on the
available data from literature that include (i) the estimated Em
values from the base
pressure-base displacement relationships in the rock sockets
that are reported by Asem
(2018), and (ii) the results of the plate load tests on weak
rocks that are reported by Chern
et al. (2004). Figure 3 shows that the values of Em estimated
using Equation (3) follow a
similar trend as the values of Em reported in the published
literature for weak rocks. It is
noted that the use of the method of Pells and Turner (1979)
(that uses the theory of
elasticity) has resulted in meaningful estimates of Em (see Fig.
3) because the values of
Ksi, which are used to estimate the corresponding values of Em
are obtained at the early
stages of the load testing where the rock socket behaves
elastically, and the relative
displacements between the rock mass and the rock socket concrete
are small (Lo and
Hefny 2001).
2. Peak side resistance (fsp) and the corresponding displacement
(δp): the values of the
peak side resistance (fsp), and the corresponding displacement
(δp) are obtained from the
measured fs-δ relationships (e.g., Fig. 2) that are reported by
Asem (2018). The
definitions of fsp and δp are shown in Fig. 2. The fsp is the
shear stress where the slope of
the tangent line to the shear stress and displacement
relationship for the sidewalls of the
rock sockets first approaches zero, i.e., , and the reported
values of δp are the
corresponding displacements.
3. Rock socket geometry: the diameter for each rock socket (B),
the depth of embedment
from the ground surface (DGS) and from the top of rock formation
(DTOR) to the center of
rock socket, and the length (L) of the shear surface formed
along the rock socket
Page 13 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
14
sidewalls for each load test are summarized in the database. The
quantity L is equivalent
to the length of the rock socket for most cases. When strain
gauges are used to measure
the load transfer from the shaft to the surrounding weak rock
mass, L represents the
distance between the strain gauges.
4. Load test approach: Figure 4 shows the methods of load
testing used in the load test
database. Figures 4(a), (b) and (c) represent the top-down
(compression) load test method
where the loads are applied to the drilled shaft butt, Fig. 4(d)
is an Osterberg load test
method where the loads are applied bi-directionally using a load
cell embedded in the test
shaft at a desired depth, and Fig. 4(e) is a tension plug or
anchor load test method. In Fig.
4(a), a load cell is placed at the base of the drilled shaft to
separate the side and base
resistances. In Figs. 4(b) and (d), the values of the side and
base resistances are separated
using strain gauges. In Fig. 4(c) a void or a compressible base
is provided at the base of
the drilled shaft to eliminate the base resistance to directly
measure the side resistance.
Lastly, in Fig. 4(e), the loads are applied in tension in the
plug or anchor load tests.
The load test data are first used to examine the factors that
affect the peak side resistance (fsp).
The load test data are then used for the development of a
predictive model for the peak side
resistance (fsp) of rock sockets in weak rocks, and for the
development of the corresponding
LRFD resistance factors (φ).
PARAMETERS AFFECTING PEAK SIDE RESISTANCE
Rock type
Page 14 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
15
The variation of the measured peak side resistance (fsp) in the
drilled shaft, anchor, and plug load
tests with different rock mass and rock socket properties are
shown in Figs. 5 to 8. Different rock
types are shown with different symbols. It appears that the rock
type does not significantly affect
the measured values of fsp and may be of secondary importance
for load tests in weak
sedimentary rocks. However, due to the significant scatter in
the load test data, a definitive
conclusion on the effect of rock type requires additional high
quality data that are not available at
the present time, and this conclusion may need to be updated as
more data become available.
Unconfined compressive strength
Figure 5 shows that the peak side resistance (fsp) and the
unconfined compressive strength (qu)
are related, which is in agreement with Rosenberg and Journeaux
(1976), Horvath and Kenney
(1979), Williams (1980), Rowe and Armitage (1984; 1987), Carter
and Kulhawy (1988), Abu-
Hejleh et al. (2003), Miller (2003), Kulhawy et al. (2005), Vu
(2013), Stark et al. (2013), and
Asem (2018). The relationship between the values of fsp and qu
may be explained as follows
1. A jointed rock mass typically consists of weathered rock
blocks, which are separated by
several joint sets (Hoek 1983). The shear strength of the
weathered rock blocks that
affects the rock mass shear strength and deformational
properties, may be represented by
qu. Williams (1980), Williams and Pells (1981), and Hassan and
O’Neill (1997)
suggested that the shear surface mobilized on the rock socket
sidewalls is commonly
formed within the adjacent rock mass and not entirely at the
rock/concrete interface.
Therefore, the measured values of fsp are expected to be
affected by the shear strength of
the rock mass and therefore its qu.
Page 15 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
16
2. As the drilled shaft, anchor, or plug displaces with respect
to the adjacent rock mass, it
commonly dilates against the normal stiffness (Kn) of the rock
mass due the roughness of
the mobilized shear surface. It should be noted, however, that
in some occasions it has
been observed by previous investigators (e.g., Seidel 1993) that
the interface can show a
contractive behavior. However, for most rocks in the database
presented in this paper, a
dilative behavior is expected. The normal stresses that act on
the shear surface are
proportional to the dilation times the normal stiffness (Kn) of
the rock mass and increase
with axial displacement of the rock socket. As the compressive
strength of the shear
surface increases, (i) the magnitude of the dilation will
increase because the rock socket
sidewall asperity crushing decreases because the sidewall has a
greater compressive
strength due to larger qu, and (ii) the stiffness of the rock
mass increases because the
values of qu and Em are related (see Fig. 3), and Em governs Kn
(see Equation 4).
Therefore larger normal stresses are generated at failure that
leads to larger values of fsp.
Consistently, Fig. 5 shows that fsp increases with qu. The
scatter in the fsp-qu data in Fig. 5 can
be explained as follows
1. The construction method of rock sockets, the load testing
methods used in the drilled
shaft, anchor, and plug load tests, the sampling approach of the
rock mass to obtain rock
core for unconfined compression tests, and the rate of shearing
of rock specimens in the
laboratory for measurement of the unconfined compressive
strength (qu) are not similar
for all cases in the database and will contribute to the scatter
of fsp-qu data in Fig. 5.
Page 16 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
17
2. As will be shown in the subsequent parts of this section, the
peak side resistance (fsp) of
the rock socket sidewalls is not only a function of the
unconfined compressive strength
(qu). Other parameters such as the deformation modulus (Em) and
the length of the shear
surface (L) also affect the measured peak side resistance (fsp).
These missing parameters
in Fig. 5 contribute to the data scatter.
Deformation modulus of rock mass
As the rock socket displaces in the axial direction, the shear
surface formed along the rock socket
sidewalls dilates against the normal stiffness (Kn) of the
adjacent weak rock mass, and the
normal stresses on the shear surface increase accordingly. The
value of the peak side resistance
(fsp) for rock socket sidewalls in weak rocks is affected by the
change in the normal stresses that
act on the mobilized shear surface on the rock socket sidewalls.
The mobilized normal stresses
are affected by the normal stiffness (Kn) of the rock socket
sidewalls, which can be written as
(4)
where ν is the Poisson’s Ratio of the weak rock. Equation (4)
(Seidel and Collingwood 2001)
shows that Em and Kn are dependent. As Em increases, Kn also
increases, and larger normal
stresses are generated as the rock socket displaces axially and
dilates with respect to the adjacent
rock mass. Therefore, fsp is greater for a rock mass with larger
Em as shown in Fig. 6 that is also
in agreement with previous published work by Williams (1980),
Williams and Pells (1981), and
Seidel and Collingwood (2001).
Page 17 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
18
Rock socket geometry
Figure 7 shows that the measured values of the peak side
resistance fsp decrease with the rock
socket diameter (B) (Fig. 7a) (Horvath and Kenney 1979), and the
length of the shear surface (L)
(Fig. 7b). The shear length (L) is the distance over which an
average fsp is calculated based on
the load test results. The value of L may correspond to the
overall length of the rock socket, or
the distance between the strain gauges if they are used in the
load test. Bandis (1980) and Bandis
et al. (1983) also showed that the shear strength of rock joints
decreases with increase in their
length, which is in agreement with the trends observed for the
drilled shafts, anchors, and plugs
socketed in weak rocks (Fig. 7b). The decrease in the measured
values of the peak side resistance
(fsp) with B and L is explained below
1. As B and L in a rock socket increase, the zone of the
influence of the rock socket also
increases (Lo and Hefny 2001), and the number of discontinuity
sets that interact with the
rock socket increases accordingly. Therefore, the stiffness and
shear strength of the rock
socket sidewalls decrease because the mobilized shear strength
and stiffness of weak rock
are scale dependent parameters (Bieniawski and Van Heerden 1975;
Goodman 1980;
Hoek 1983; Lo and Hefny 2001). As Em decreases, Kn decreases and
fsp decreases
because smaller normal stresses are generated on the rock socket
sidewalls. Data
presented in Fig. 7, however, show that the effect of L on fsp
is more significant than the
effect of B on the mobilized values of fsp.
2. The non-uniform stress distribution along the rock socket
sidewall can result in
progressive failure along the rock socket sidewalls as loads on
the rock socket increase.
This will result in the overall decrease in the measured side
resistance as L increases.
Page 18 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
19
Initial state of stress
The embedment depth (DGS and DTOR) may be used to estimate the
lateral pressures that are
exerted by the concrete, which may be related to the initial
normal stresses on the shear surfaces
that are formed on the rock socket sidewalls (Hassan and O’Neill
1997). The data presented in
Fig. 8 show that the embedment depth (DGS and DTOR) does not
significantly affect the
mobilized values of the peak side resistance (fsp). Therefore,
the initial normal stresses (σno) and
the peak side resistance (fsp) are not strongly related. The
peak side resistance (fsp), however, is
affected by the final normal stresses that are mobilized on the
shear surface at the time of failure
(Seidel 1993). These final normal stresses are different from
the initial normal stresses (Vesic
1963, Seidel 1993) due to the dilation of the rock socket
against the normal stiffness (Kn) of the
rock socket side walls (Williams and Pells 1981; Seidel and
Collingwood 2001).
Fig. 9 shows a plot of the normalized peak side resistance
(fsp/qu) and qu. In Fig. 9, the fsp/qu
data are separated based on the estimated normalized initial
normal stress (σno/σp). The trend
lines representing the relationship between fsp/qu, qu and
σno/σp, which are reproduced based on
the method of Hassan and O’Neill (1997) are also plotted on the
load test data in Fig. 9. In the
normalized initial normal stress (σno/σp), σp is the atmospheric
pressure, and its value is equal to
0.101 MPa, and σno is estimated based on the hydrostatic
pressure of fresh concrete at the time of
pour by Equation (5)
(5)
Page 19 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
20
where γc and hc are, respectively, the unit weight of concrete,
and the head of concrete above the
depth where the corresponding values of the σno were calculated.
Equation (5) results in
reasonable estimates of σno because high slump concrete was used
in the construction of most of
the rock sockets reported in this study. Bernal and Reese (1983)
showed that for high slump
concrete, the lateral pressure exerted by the concrete
approaches hydrostatic condition. The
scatter in Fig. 9 is significant, however, this figure shows
that the values of fsp/qu and σno/σp are
not strongly related.
Figure 10(a) shows the variation of the rock mass deformation
modulus (Em) with the
embedment depth from the ground surface (DGS) for drilled shaft
load tests in weathered
Melbourne Siltstone (after Williams 1980), and Fig. 10(b) shows
the variation of fsp with DGS,
for the same load test site. Figure 10(a) shows that Em
increases with DGS, which may be
explained by the decrease in the weathering with the embedment
depth at this particular load test
site. Larger values of Em results in a larger rock socket
sidewall Kn (see Equation 4) and causes
the mobilization of larger normal stresses at the time of
failure, which leads to mobilization of
greater fsp values as shown in Fig. 10(b). It is noted that
increase in the initial normal stresses
due to increase in the values of DGS may also be contributing to
some extent, however, it is clear
that increase in Em with DGS is also a contributing factor for
increase in fsp with DGS. Based on
the discussion presented above, it may thus be concluded that
the increase in fsp with embedment
depth could be more related to decrease in the weathering, and
the subsequent increase in Em
(and the corresponding values of Kn), and not necessarily due to
increase in initial normal
stresses (σno) acting on the shear surface.
Page 20 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
21
If, on the other hand, the weathering increases with DGS, the
values of Em and thus Kn will
decrease with depth accordingly, and fsp will also decease. An
example of such scenario is
shown in Fig. 11 that is obtained based on the load tests in
weathered Melbourne Siltstone,
performed by Williams (1980). The variation of fsp with DGS
shown in Fig. 11 may not be
explained, nor could it be predicted using the method of Hassan
and O’Neill (1997), which
suggests that fsp should tacitly increase with DGS, with no
exception.
Vertical socket displacement
The vertical displacement that is required to mobilize the
values of fsp in rock sockets in weak
rocks (δp) is plotted versus the corresponding values of the
rock socket diameter (B) in Fig. 12.
Figure 12 shows that, on average, a vertical displacement of 10
mm is required to mobilize fsp.
This observation is in agreement with Brown et al. (2010).
Figure 12 also shows that the
displacement required to mobilize fsp is not a discernible
function of B.
Roughness
The importance of the shear surface roughness on the measured
values of the peak side
resistance (fsp) has been emphasized by Williams (1980),
Williams and Pells (1981), and Seidel
(1993) and many others. Johnston and Lam (1989), Kodikara
(1989), Seidel (1993), and Seidel
and Collingwood (2001) developed design models that relate the
shear stress to the pre-concrete
pour average roughness height (∆r) of the rock/concrete
interfaces. The roughness of the
mobilized shear surface along the rock socket sidewalls cannot
be accurately predicted, and was
not included in the development of the proposed design model for
fsp for the following reasons
Page 21 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
22
1. The method of Seidel and Collingwood (2001) for the
prediction of ∆r does not consider
the following effects: (i) formation of the smeared and
disturbed materials (Williams
1980) on the rock socket sidewalls, and (ii) increase in the
water content of the rock
socket sidewalls beyond its initial condition due to migration
of the moisture from the
concrete to the adjacent rock (Meyerhof and Murdock 1953). The
formation of the
disturbed materials and the change in the water content of the
interface can result in post-
construction changes (i.e., softening) in the characteristics of
the socket wall roughness
and can decrease its resistance to wear and crushing which in
turn can affect the dilation
of the shear surface.
2. Actual rock socket sidewall roughness measurements are not
made in most of the load
tests in the database that is compiled herein.
3. The available roughness measurements are those of
rock/concrete interface and provide
little information on the roughness of the actual shear surface
which is mobilized.
PROPOSED PREDICTIVE EQUATION FOR PEAK SIDE RESISTANCE
The method of Gardoni et al. (2002) is used to develop a
predictive model for the peak side
resistance (fsp). The following general equation is adopted for
the proposed model
(6)
where fsp(x, Θ) represents a probabilistic model for peak side
resistance, γ (x,θ) is a set of terms
that are developed based on the properties of (i) the rock
socket and (ii) rock mass, which control
the development of the peak side resistance, Θ = (θ,σ) is a set
of model parameters used to fit the
Page 22 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
23
proposed model, of the form shown in Equation (6), to the
observed data, σε is an additive
model error (additivity assumption) where ε is assumed to be a
normal random variable
(normality assumption) with zero mean and unit variance, and σ
denotes the standard deviation
of the model error that is assumed to be constant
(homoskedasticity assumption). The validity of
these assumptions can be assessed using diagnostics plots
following Rao and Toutenburg (1997)
and Gardoni et al. (2002).
Following Gardoni et al. (2002), the γ (x,θ) term can be
expressed using Equation (7)
(7)
The term hj(x) is the jth explanatory function used to include
all parameters (represented by the
vector x) that affect the development of fsp in the design
model, and θj is the jth model parameter
that is used to fit the proposed model to the observed data.
The following procedure is used to estimate the model parameters
Θ = (θ,σ), which are obtained
by fitting Equation (6) to the in situ load test data
1. The load test data that are used in the analysis correspond
mostly to rock sockets with
relatively clean and rough rock socket sidewalls. Quantitative
measure of roughness of
shear surface was not available in most cases. The proposed
methods herein should be
Page 23 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
24
applied with caution to the design of rock sockets where
drilling method is anticipated to
result in production of smooth or smeared rock socket
sidewalls.
2. The quantities of interest (fsp, qu, L, B, and Em) are first
transformed using a natural
logarithm. This is to justify the following assumptions: (i) the
model variance (σ2) is
independent of x (homoskedasticity assumption), and (ii) ε has a
normal distribution
(normality assumption) (Gardoni et al. 2002).
3. Once all parameters of interest are transformed using the
proper transformation technique
(in this case natural logarithm), the Maximum Likelihood (Ang
and Tang 2007) method
is used to estimate the model parameters, Θ = (θ,σ).
4. Once the model parameters Θ = (θ,σ) are estimated, both sides
of the model are
transformed back into the original space using an exponential
transformation, e(.).
MODEL CALIBRATION
The review of the load test data shows that the peak side
resistance (fsp) depends on the values of
qu, L, B, and Em. The effect of such parameters has been
emphasized in the published literature
(Pells et al. 1980; Williams 1980; Williams and Pells 1981;
O’Neill et al. 1996).
In this study, we first develop a “Base Model” that relates the
measured values of the peak side
resistance (fsp) to the corresponding values of the unconfined
compressive strength (qu) of the
weak rock along the rock socket walls. Additional terms,
including the rock socket and rock
mass properties (i.e., explanatory functions including L, B, and
Em), are then added to this “Base
Model” to investigate any potential improvement in the model
accuracy. The “Base Model” is
Page 24 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
25
calibrated using the load test data and the method of Gardoni et
al. (2002). Equation (8) shows
the calibrated Base Model
(8)
where fsp and qu are in units of MPa. The standard deviation (σ)
of the model error for Equation
(8) is 0.825 MPa. The statistics of the model error and model
parameters are summarized in
Table 3.
Analysis of the in situ load test data further shows that the
rock socket geometry affects the
mobilized values of fsp. Therefore, we investigate the
contribution of additional information on L
and B to the overall accuracy of the “Base Model” by updating
the “Base Model” using the
values of B and L. The updated models are shown in Equations (9)
and (10)
(9)
and
(10)
where the values of B and L are in units of meter. The estimated
values of σ for the model error
in Equation (9) is 0.766 MPa, and in Equation (10) is 0.768 MPa.
These results indicate that
Page 25 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
26
updating the “Base Model” using information on the values of L
and B improves the model
accuracy as indicated by the reduced values of σ in Equations
(9) and (10) compared to the
value of σ in Equation (8). Analyses, however, show that the
effect of B on σ is insignificant
compared to the effect of L on σ. This conclusion may be reached
by comparing the values of σ
in Equations (9) and (10). The statistics of the model error and
model parameters for Equations
(9) and (10) are, respectively, summarized in Tables 4 and
5.
Analysis of load test data also shows that fsp and the estimated
values of rock mass Em are
related. To investigate possible contributions of Em to the
accuracy of the “Base Model,” the
“Base Model” is now updated using the additional information on
the values of Em and L. The
updated models are summarized in Equations (11) and (12)
(11)
and
(12)
In Equation (11), Em is added to Equation (8) and in Equation
(12), L and Em are added to
Equation (8). The estimated value of σ for the model error for
Equation (11) is 0.799 MPa, and
value of σ for Equation (12) is 0.645 MPa. The statistics of the
model error and model
parameters for Equations (11) and (12) are, respectively,
summarized in Tables 6 and 7.
Page 26 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
27
Comparison of the values of σ in Equations (8) to (12) indicates
that Equation (12) is the most
accurate model for the prediction of fsp. The parameters
required to evaluate Equation (12) may
be estimated for design purposes. The proposed model for fsp
includes qu, L, and Em (Equation
12). The following are noted in relation to this model
1. Equation (12) states that fsp increases with qu and Em but
decreases with L. These are in
agreement with the in situ load test data presented in Figs. 5
to 7 and published literature.
2. Load test data shows that B does not significantly affect fsp
(Fig. 7a), particularly for
large diameter rock sockets (i.e., B > 400 mm).
3. Initial state of stress is not considered in the updating
process presented above because
load test data (Fig. 8 which shows the variation of the fsp with
DGS and DTOR that are
assumed to crudely represent the initial state of stress in the
rock mass adjacent to the
rock socket walls) do not show a significant relationship
between these parameters and
fsp. This is because fsp is related to normal stresses at the
time of failure. DTOR and DGS
only crudely represent the initial normal stresses on the rock
socket walls.
4. Because the roughness of the actual shear surface for the
load tests reported in this study
are not available, the roughness is not included in the
formulation of the proposed model.
The existing models are evaluated using the load test database,
and the results are summarized in
Table 8. The models are compared in terms of the coefficient of
variation of their bias (δλ). The
bias is defined as the ratio of measured peak side resistance
(fsp,m) to predicted side resistance
(fsp,p), and is denoted by λ. Standard formulation (e.g., Ang
and Tang 2007) may be used to
Page 27 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
28
evaluate the mean (µλ), and the standard deviation (σλ) of the
bias, λ = fsp,m/fsp,p. The coefficient
of variation (δλ) for bias (i.e., fsp,m/fsp,p) may then be
defined using Equation (13)
(13)
Coefficient of variation for bias (δλ) is a measure of the
“dispersion” or “variability” of fsp,m/fsp,p
relative to its mean value (Ang and Tang 2007), and may be used
as a measure of the accuracy of
the predictive model. The statistics for predictive model in
Equation (12) are summarized in
Table 9. The results presented in Table 8 show that δλ for all
of the existing models are greater
than the proposed model (i.e., Equation 12). The results
presented herein indicate that addition of
L and Em to the “Base Model” has resulted in improved
performance of the model.
CALIBRATION OF RESISTANCE FACTOR
Traditionally, designers have used a factor of safety (FS) in
the Allowable Stress Design (ASD)
approach to account for the uncertainties both in resistances
and loads. The FS is selected
subjectively, and does not properly quantify the design
uncertainties (Kulhawy and Phoon 2006;
Roberts and Misra 2009; Paikowsky et al. 2010). As an
alternative to the ASD, designers have
used the Load and Resistance Factor Design (LRFD) framework to
ensure the design is safe
against each limit state (e.g., bearing capacity failure). The
LRFD accounts for the uncertainty in
loads and resistances separately. Additionally, the LRFD
framework quantifies the design
uncertainties more rigorously using the theory of
probability.
Page 28 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
29
In the design of foundations, two limit states should be
recognized (Meyerhof 1951; Allen 1975;
Phoon et al. 2000). The foundation should be safe against the
bearing capacity failure, and the
foundation settlements should be less than the tolerable values.
The evaluation of the strength
limit state is discussed in the following sections to ensure a
bearing capacity failure will not
materialize. In the context of the LRFD framework, the strength
limit state is determined using
Equation (14)
(14)
where φs and φb, respectively, are the LRFD resistance factors
for the side and base resistances,
Rs and Rb are, respectively, the total side and base
resistances, and the values of γi are the load
factors for the ith axial load effects (Qi) at the top of
drilled shaft, anchor, or plug.
To design a rock socket using the LRFD framework, φs and φb
should be calibrated to produce a
design that is consistent with the acceptable level of failure
probability in the current practice. In
this section, the predictive model for fsp (Equation 12) is used
to determine φs using the First-
Order Reliability Method (FORM) following the procedure proposed
by Briaud et al. (2013).
First-Order reliability method (FORM)
The limit state function (g), which is required for the FORM
analysis (Briaud et al. 2013;
Gardoni 2017) is introduced in Equation (15)
Page 29 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
30
(15)
where λR is the bias of the predictive model for fsp (i.e.,
Equation 12), λDL and λLL, respectively,
are the bias for dead and live loads, and DL and LL are,
respectively, the dead and live loads.
Following the recommendations of Paikowsky et al. (2004),
Paikowsky et al. (2010), and Briaud
et al. (2013), an HS-20 design truck is used to represent the LL
for typical bridge piers (i.e., LL =
445 kN). The calibration of the LRFD resistance factors also
requires an estimate of the dead
load to live load ratio (DL/LL). The actual value of the DL/LL
ratio is a function of the span
length (l) of the structure. For example, McVay et al. (2000)
provided the following values of
DL/LL based on the values of l: DL/LL = 0.52, 1.06, 1.58, 2.12,
2.64, 3.00, and 3.53,
respectively, for l values of 9, 18, 27, 36, 45, 50, and 60 m.
These values are used in the
calibration process. A lognormal distribution is used to
represent the variation of the bias for fsp
predictive model (i.e., Equation 12) (i.e., λR). Following
Briaud et al. (2013), the statistics for
λDL and λLL are also assumed to follow a lognormal distribution,
and are reported in Table 9.
The values reported in Table 9 are obtained from the structural
engineering literature to be
consistent with the existing practice for the design of the
superstructure.
Equation (15) is now used in the FORM analysis to iterate about
Rs until the calculated
reliability index (β) (Ditlevsen and Madsen 1996; Haldar and
Mahadevan 2000; Gardoni 2017)
approaches the target reliability index (βT) that is consistent
with the acceptable level of failure
probability that is commonly used for design of the
superstructure. Based on a review of
literature, βT is selected as follows
Page 30 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
31
1. ISO (2015) evaluated βT for different “consequences of
failure” and “relative cost of
safety measure.” The consequences of failure included minor,
moderate, and large. The
relative cost of safety measure included large, normal, and
small. The recommended βT
ranges from 3.1 (for large cost of safety measure and large
consequence of failure) to 4.7
(for small cost of safety measure and large consequence of
failure).
2. The resistance factors recommended by Brown et al. (2010)
(i.e., FHWA-NHI-10-016)
for the design of drilled shafts in rocks correspond to a βT of
3.0.
3. The redundancy and the possibility of load redistribution
(i.e., load sharing among
foundation components) in the event of failure of one of the
load bearing components in a
system is an important criterion in determination of βT (Allen
et al. 2005). Liu et al.
(2001) defined a redundant foundation system as one in which the
calculated β value for
the overall system is 0.5 higher than the reliability of the
individual foundation members
within the group. Barker et al. (1991) determined that βT for
drilled shafts ranges from
2.0 to 3.7 (from MVFOSM analysis), and 2.0 to 4.3 (from more
advanced analysis).
Barker et al. determined that βT is 3.5 for non-redundant
systems, 2.5 to 3.0 for drilled
shafts, and 2.0 to 2.5 for highly redundant systems. Barker et
al. (1991) relied heavily on
the implied level of safety by previous design practice to
provide their recommendations
for the values of βT and the resulting resistance factors. Allen
et al. (2005) proposed that
the ability of the soil to redistribute the load allows for the
foundation system to be
designed for a lower reliability index compared to the
superstructure. The only exception
to this rule is the case of a single drilled shaft that supports
the entire bridge pier where a
βT value of 3.5 is desirable (Allen et al. 2005).
Page 31 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
32
4. Paikowsky et al. (2004) recommended a βT of 3.0 when the
drilled shaft group contains
less than 5 drilled shafts, and recommended a βT of 2.3 for the
cases where the number of
shafts in a shaft group exceeds five. Therefore, the βT for
which individual drilled shafts
are designed is heavily affected by the redundancy of the
drilled shaft foundation system.
This conclusion is in agreement with the work of other
researchers (e.g., Liu et al. 2001;
Allen et al. 2005).
5. Briaud et al. (2013) used a βT of 3.5 for the calibration of
resistance factors for the
assessment of scour of the bridge foundations.
6. Phoon et al. (2000) recommended a βT of 3.2 for the
assessment of the strength limit state
in drilled shafts. Phoon et al. (2000) reported that the actual
annual probabilities of failure
for foundations is between 0.1% to 1%. These values of
probability of failure may be
used to back-calculate βT of 2.5 to 3.0 for these foundations
using the method of
Rosenblueth and Esteva (1972) (i.e., ).
7. Phoon and Kulhawy (2005) determined the reliability of
drilled shafts in different soils
and rocks, and stated that the value of βT varied from 2.6 to
3.4 for all drilled shafts in
their database that were subjected to an undrained compression
loading.
8. Kulhawy and Phoon (2006) adopted the recommendations of the
Canadian Building
Code and the AASHTO bridge specifications. The Canadian Building
Code recommends
a βT of 3.5 for the superstructure and the foundation and the
AASHTO specifications
recommends a βT of 3.5 for the superstructure and a βT of 2.0 to
3.5 for the foundation.
9. Nowak (1995) used a βT of 3.5 for the superstructure. Allen
et al. (2005) cited a βT of 3.5
for the superstructure and a βT of 3.0 for the foundation design
from past practice
Page 32 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
33
although Allen et al. (2005) indicates that it is desirable to
“maintain a consistent level of
safety across all limit states of a given type (e.g., strength
limit state).”
10. Ellingwood and Galambos (1982) recommended βT values of 2.5
to 4.0 for structural
steel elements, and 2.3 to 3.6 for reinforced concrete elements
in the building structures.
The βT values for foundations should be selected so as to
maintain a degree of
consistency between level of safety of the foundation and the
superstructure.
11. The typical range of βT for different load combinations is
reported by Allen (1975).
According to Allen (1975), βT often ranges from 3.0 to 4.0.
Therefore, the quantity Rs is calculated for βT values of 2.0,
2.5, and 3.0 that fall within the rage
suggested in the published literature. Once Rs is calculated
from the FORM analyses, Equation
(16) may be used to calculate the corresponding φs
(16)
Equation (16) and the FORM analyses are used to determine the
values of φs for fsp that is given
by Equation (12), and for a range of l and βT, which are
discussed in the previous sections.
The resistance factors obtained from the FORM analysis are
summarized in Fig. 13. The
calculated range of the resistance factors is in agreement with
the recommended resistance
factors by other investigators (e.g., Paikowsky et al. 2010).
Figure 13 shows that the resistance
factor increases with decrease in βT, and slightly decreases
with increase in l.
Page 33 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
34
CONCLUSIONS
In situ drilled shaft, anchor, and plug load tests are used to
develop a framework for the
prediction of peak side resistance (fsp) of rock socketed
drilled shafts in weak rocks. A
comprehensive database is compiled. The effects of rock mass,
and rock socket geometry on
peak side resistance (fsp) are evaluated. A “Base Model” is
developed that relates the measured
values of peak side resistance (fsp) to the unconfined
compressive strength (qu) of the weak rock.
The method of Gardoni et al. (2002) is used to update the “Base
Model” using additional data
gathered and reported in the new database introduced in the
Supplemental Data section.
Analysis results shows that a model that includes the unconfined
compressive strength (qu), the
rock mass deformation modulus (Em), and the length of the
mobilized shear surfaces (L)
provides a more accurate estimate of fsp than all other methods
reviewed herein. The values of qu
represents the shear strength of the intact blocks that make the
rock mass. The values of Em and
L reflect the size effect and the number of discontinuity
surfaces that interact with the rock
socket.
The proposed model is not a function of the initial normal
stress (σno) on the rock socket walls
because analysis showed that the σno does not affect the fsp.
The estimated values of σno are
related to the embedment depth. The proposed model for fsp is
also not a function of ∆r because
the average roughness height was not measured and are not
available for the drilled shaft, plug,
and anchor load tests reported in herein. Additionally, the
proposed model should be used when
Page 34 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
35
drilling method leads to development of relatively rough rock
socket sidewalls that are free from
smeared materials.
The First-Order Reliability Method (FORM) is then used to
develop the corresponding Load and
Resistance Factor Design (LRFD) resistance factors (φ). The
recommended resistance factors are
functions of structure span length (l), and the target
reliability index (βT). It is shown that the
calculated values of resistance factor are not sensitive to the
span length of structure (l).
Page 35 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
36
ACKNOWLEDGEMENTS
The Authors would like to thank and acknowledge the University
of Illinois at Urbana-
Champaign for providing access to the technical and
computational resources that were required
for completion of this work. The Authors thank Professor
Emeritus James H. Long for his
contributions to this paper. The Authors also thank Dr. Armin
Tabandeh for his assistance with
MATLAB code used for reliability analysis performed in this
study.
Page 36 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
37
REFERENCES
Abu-Hejleh, N., O’Neill, M. W., Hanneman, D., and Atwooll, W. J.
2003. Improvement of the
geotechnical axial design methodology for Colorado’s drilled
shafts socketed in weak
rocks. Final Contact Report No. CDOT-DTD-R-2003-6, Colorado
Department of
Transportation, Denver, C.O., United States.
Allen, D. E. 1975. Limit States Design-A Probabilistic Study.
Canadian Journal of Civil
Engineering, 2(1): 36-49.
Allen, T. M., Nowak, A. S., and Bathurst, R. J. 2005,
Calibration to determine load and
resistance factors for geotechnical and structural design,
Transportation Research Board,
Washington, D.C, United States.
Ang, A. H.-S., and Tang, W. H. 2007. Probability concepts in
engineering: emphasis on
applications in civil and environmental engineering, (Second
ed.), Wiley-Interscience, New
York, United States.
Asem, P. 2018. Axial behavior of drilled shafts in soft rock.
Ph.D. Thesis, Department of Civil
and Environmental Engineering, University of Illinois at
Urbana-Champaign, Urbana,
Illinois, United States.
Bandis, S. 1980. Experimental studies of scale effects on shear
strength and deformation of rock
joints. Ph.D. Thesis, Department of Earth Sciences, University
of Leeds, Leeds, England.
Bandis, S. C., Lumsden, A. C., and Barton, N. R. 1983.
Fundamentals of rock joint deformation.
International Journal of Rock Mechanics and Mining Sciences and
Geomechanics
Abstracts, 20(6): 249-268.
Barton, N., Bamford, W. E., Barton, C. M., MacMahon, B., Kanji,
M. A., Babcock, K., Boyd, J.
M., Cruden, D., Franklin, J. A. et al. 1978. Suggested methods
for the quantitative
description of discontinuities in rock masses. International
Journal of Rock Mechanics and
Mining Sciences and Geomechanics Abstracts, 15(6): 319-368.
Barker, R. M., Duncan, J. M., Rojiani, K. B., Ooi, P. S. K.,
Tan, C. K., and Kim, S. G. 1991.
Manuals for the design of bridge foundations, Transportation
Research Board, Washington,
D.C, United States.
Bernal, J. B., and Reese, L. C. 1983. Study of the lateral
pressure of fresh concrete as related to
the design of drilled shafts, Center for Transportation
Research, The University of Texas at
Austin, Austin, Texas, United States.
Page 37 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
38
Bieniawski, Z. T., and van Heerden, W. L. 1975. The significance
of in situ tests on large rock
specimens. International Journal of Rock Mechanics and Mining
Sciences and
Geomechanics Abstracts, 12(4): 101-113.
Briaud, J. L., Gardoni, P., and Yao, C. 2013. Statistical, Risk,
and Reliability Analyses of Bridge
Scour. Journal of Geotechnical and Geoenvironmental Engineering,
140(2): 04013011.
Brown, D. A., Turner, J. P., and Castelli, R. J. 2010. Drilled
shafts: construction procedures and
LRFD design methods, FHWA-NHI-10-016, Washington, D.C., United
States.
Canadian Geotechnical Society. 2006. Canadian Foundation
Engineering Manual. Richmond,
British Columbia, Canada.
Carter, J. P., and Kulhawy, F. H. 1988. Analysis and design of
drilled shaft foundations socketed
into rock. Electric Power Research Institute, (EL-5918), Palo
Alto, California, United
States.
Cepeda-Diaz, A. F. 1987. An experimental investigation of the
engineering behavior of natural
shales. Ph.D. Thesis, Department of Civil and Environmental
Engineering, University of
Illinois at Urbana-Champaign, Urbana, Illinois, United
States.
Chern, J. C., Chang, Y. L., Lee, K. R., Yu, C. W., Li, T. J.,
Li, J. Y., Xu, D. J., Zhang, G., Min,
H., and Yuan, C. H. 2004. Correlation study on the deformation
modulus and rating of rock
mass. Sinotech Engineering Consultants, Inc. Report R-GT-97-04,
Taipei, Taiwan.
Coduto, D. P. 2001. Foundation design: principles and practices.
2nd ed., Prentice Hall, New
Jersey, United States.
Collingwood, B. 2000. The effect of construction practices on
the performance of rock socketed
bored piles. . Ph.D. Thesis, Department of Civil Engineering,
Monash University,
Melbourne, Australia.
Deere, D. U., and Miller, R. P. 1966. Engineering classification
and index properties for intact
rock. University of Illinois at Urbana-Champaign/Air Force
Weapons Laboratory Report
AFWL-TR-65-116, Urbana, Illinois, United States.
Ditlevsen, O., and Madsen, H. O. 1996. Structural reliability
methods. Wiley-Interscience, New
York, United States.
Ellingwood, B. and Galambos, T. V. 1982. Probability-based
criteria for structural design,
Structural Safety, 1(1): 15-26.
Page 38 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
39
Gardoni, P., Der Kiureghian, A., and Mosalam, K. M. 2002.
Probabilistic Capacity Models and
Fragility Estimates for Reinforced Concrete Columns based on
Experimental Observations.
Journal of Engineering Mechanics, 128(10): 1024-1038.
Gardoni, P. 2017. Risk and Reliability Analysis: Theory and
Applications. Paolo Gardoni (ed.),
Springer, Urbana, United States.
Goodman, R. E. 1980. Introduction to rock mechanics. 1st ed.,
John Wiley and Sons, New York,
United States.
Goodman, R. E. 1993. Engineering geology: rock in engineering
construction. 1st ed., John
Wiley and Sons, New York, United States.
Haldar, A. and Mahadevan, S. 2000. Probability, reliability and
statistical methods in
engineering design. John Wiley and Sons, Inc., New York, United
States.
Hassan, K. M. 1994. Analysis and design of drilled shafts
socketed into soft rock. Ph.D. Thesis,
Department of Civil and Environmental Engineering, University of
Houston, Houston,
Texas, United States.
Hassan, K. M., and O'Neill, M. W. 1997. Side load-transfer
mechanisms in drilled shafts in soft
argillaceous rock. Journal of Geotechnical and Geoenvironmental
Engineering, 123(2):
145-152.
Hoek, E. 1983. Strength of jointed rock masses. Géotechnique,
33(3): 187-223.
Hoek, E., and Brown, E. T. 1997. Practical estimates of rock
mass strength. International Journal
of Rock Mechanics and Mining Sciences, 34(8):1165-1186.
Horvath, R. G., and Kenney, T. C. 1979. Shaft resistance of
rock-socketed drilled piers. In
Proceedings of the Symposium on Deep Foundations, ASCE, New
York, NY, United
States, pp. 182-214.
Horvath, R. G., Kenney, T. C., and Kozicki. P. 1983. Methods for
improving the performance of
drilled piers I weak rock. Canadian Geotechnical Journal, 20(4):
758-772.
Horvath, R. G. and Chae, K. J. 1989. Long-term settlement of
model rock-socketed piers.
Canadian Geotechnical Journal, 26(3): 348-58.
International Organization for Standardization (ISO). 2015.
General principles on reliability for
structures, (ISO 2394; Switzerland: International Organization
for Standardization).
Johnston, I. W. and Lam, Thomas S. K. 1989. Shear Behavior of
Regular Triangular
Concrete/Rock Joints-Analysis. Journal of Geotechnical
Engineering, 115(5): 711-27.
Page 39 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
40
Kanji, M. A. 2014. Critical issues in soft rocks. Journal of
Rock Mechanics and Geotechnical
Engineering, 6(3): 186-195.
Kodikara, J. K. 1989. Shear behaviour of rock-concrete joints
and side resistance of piles in
weak rock. Ph.D. Thesis, Department of Civil Engineering, Monash
University,
Melbourne, Australia.
Kulhawy, F. H., and Phoon, K. K. 1993. Drilled shaft side
resistance in clay soil to rock. In the
Proceedings of the Design and Performance of Deep Foundations:
Piles and Piers in Soil
and Soft Rock, New York, United States.
Kulhawy, F. H., Prakoso, W. A., and Akbas, S. O. 2005.
Evaluation of Capacity of Rock
Foundation Sockets. In Proceedings of the 40th U.S. Symposium on
Rock Mechanics.,
American Rock Mechanics Association, Anchorage, Alaska, United
States.
Kulhawy, F. H., and Phoon, K. K. 2006. Some Critical Issues in
Geo-RBD Calibrations for
Foundations. In Proceedings of GeoCongress 2006, Atlanta,
Georgia, United States, pp. 1-
6.
Liu, W. D., Neuenhoffer, A., Ghosn, M., and Moses, F. 2001.
Redundancy in highway bridge
substructures, Transportation Research Board, Washington, D.C.,
United States.
Lo, K. Y. and Hefny, A. M. 2001. Foundations on rock. ed. R. K.
Rowe (Geotechnical and
Geoenvironmental Engineering Handbook).
McVay, M., Birgisson, B., Zhang, L., Perez, A., and Putcha, S.
2000. Load and Resistance
Factor Design (LRFD) for Driven Piles Using Dynamic Methods-A
Florida Perspective.
Geotechnical Testing Journal, 23(1): 55-66.
Mesri, G., and Shahien, M. 2003. Residual Shear Strength
Mobilized in First-Time Slope
Failures. Journal of Geotechnical and Geoenvironmental
Engineering, 129(1): 12-31.
Meyerhof, G. G. 1951. The Ultimate Bearing Capacity of
Foundations. Géotechnique, 2(4): 301-
332.
Meyerhof, G. G. and Murdock, L. J. 1953. An investigation of the
Bearing Capacity of Some
Bored and Driven Piles in London Clay. Géotechnique, 3(7):
267-82.
Miller, A. D. 2003. Prediction of ultimate side shear for
drilled shafts in Missouri shales. M.S.
Thesis, Department of Civil and Environmental Engineering,
University of Missouri-
Columbia, Missouri, MO, United States.
Page 40 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
41
Nowak, A. S. 1995. Calibration of LRFD Bridge Code, Journal of
Structural Engineering,
121(8): 1245-51.
O'Neill, M. W., Townsend, F. C., Hassan, K. M., Buller, A., and
Chan, P. S. 1996. Load transfer
for drilled shafts in intermediate geomaterials. U.S. Department
of Transportation, Federal
Highway Administration, Mountain View, California, United
States.
Osterberg, J. O., and Gill, S.A. 1973. Load transfer mechanism
for piers socketed in hard soils or
rock. In Proceedings of the 9th Canadian Rock Mechanics
Symposium, Montreal.
Paikowsky, S. G., Birgisson, B., McVay, M. C., Nguyen, T., Kuo,
C., Baecher, G., Ayyub, B.,
Stenersen, K., O’Malley, K., et al. 2004. Load and resistance
factor design (LRFD) for
deep foundations (NCHRP 507). Washington, D.C., United
States.
Paikowsky, S. G., Canniff, M. C., Lesny, K., Kisse, A., Amatya,
S., and Muganga, R. 2010.
LRFD Design and Construction of Shallow Foundations for Highway
Bridge Structures.
Transportation Research Board, Washington, D. C., United
States.
Peck, R. B., Hanson, W. E., and Thornburn, T. H. 1974.
Foundation Engineering. 2nd ed., John
Wiley and Sons, New York, United States.
Pells, P. J. N., and Turner, R. M. 1979. Elastic solutions for
the design and analysis of rock-
socketed piles. Canadian Geotechnical Journal, 16(3):
481-487.
Pells, P. J. N., Rowe, R. K., and Turner, R. M. 1980. An
experimental investigation into side
shear for socketed piles in sandstone. In Proceedings of the
International Conference on
Structural Foundations on Rock, Sydney, Australia, pp.
295-302.
Phoon, K. K., Kulhawy, F. H., and Grigoriu, M. D. 2000.
Reliability-based design for
transmission line structure foundations. Computers and
Geotechnics, 26(3-4): 169-185.
Phoon, K. K. and Kulhawy, F. H. 2005. Characterization of model
uncertainties for drilled shafts
under undrained axial loading. In Proceedings of Contemporary
issues in foundation
engineering (ASCE), GSP131, Austin, Texas, United States, pp.
1-13.
Rao, C. R. and Toutenburg, H. 1999. Linear models: least squares
and alternatives. 2nd ed.
Springer, New York, United States.
Roberts, L. A., and Misra, A. 2009. Load and resistance factor
design (LRFD) of deep
foundations using a performance-based design approach. Journal
of GeoEngineering, 4(3):
87-92.
Page 41 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
42
Rosenberg, P., and Journeaux, N. L. 1976. Friction and end
bearing tests on bedrock for high
capacity socket design, Canadian Geotechnical Journal, 13(3):
324-33.
Rosenblueth, E. and Esteva, L. 1972. Reliability Basis for Some
Mexican Codes, Special
Publication, 31.
Rowe, R. K., and Armitage, H. H. 1984. The design of piles
socketed into weak rock. GEOT-11-
84, Ottawa, Ontario, Canada.
Rowe, R. K. and Armitage, H.H. 1987. A design method for drilled
piers in soft rock. Canadian
Geotechnical Journal, 24(1): 126-142.
Seidel, J. P. 1993. Analysis and design of pile shafts in weak
rock. Ph.D. Thesis, Department of
Civil Engineering, Monash University, Melbourne, Australia.
Seidel, J. P., and Collingwood, B. 2001. A new socket roughness
factor for prediction of rock
socket shaft resistance. Canadian Geotechnical Journal, 38(1):
138-153.
Singh, M., and Rao, S. K. 2005. Empirical methods to estimate
the strength of jointed rock
masses. Engineering Geology, 77(1–2): 127-137.
Skempton, A. W. 1948. The rate of softening in stiff fissured
clays, with special reference to
London clay. In Proceedings of the 2nd International Conference
on Soil Mechanics and
Foundation Engineering (Rotterdam), pp. 50-53.
Skempton, A. W., and Hutchinson, J. 1969. Stability of natural
slopes and embankment
foundations. In Proceedings of the 7th International Conference
on Soil Mechanics, New
Mexico.
Stark, T. D., Long, J. H. and Assem, P. 2013. Improvements for
determining the axial capacity of
drilled shafts in shale in Illinois. Illinois Department of
Transportation Report No. FHWA-
ICT-13-017, Illinois, IL, United States.
Terzaghi, K. 1936. Stability of slopes of natural clays. In
Proceedings of the Harvard Conference
on Soil Mechanics (1).
Terzaghi, K. 1943. Theoretical soil mechanics. John Wiley and
Sons, New York, United States.
Vesic, A. S. 1963. Bearing capacity of deep foundations in sand.
Highway Research Record,
(39): 112-153.
Vu, T. T. 2013. Load and resistance factor design of drilled
shafts at the service limit state. Ph.D.
Thesis, Department of Civil and Environmental Engineering,
University of Missouri-
Columbia, Columbia, Missouri, United States.
Page 42 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
43
Wainshtein, I., Hatzor, Y. H., and Doktofsky, M. 2008. Does
shaft resistance of piles in rock
scale with RQD?. In Proceedings of 2nd U.S.-Canada Rock
Mechanics Symposium, San
Francisco, United States.
Williams, A. F. 1980. The design and performance of piles into
weak rock. Ph.D. Thesis,
Department of Civil Engineering, Monash University, Melbourne,
Australia.
Williams, A. F. and Pells, P. J. N. 1981. Side resistance rock
sockets in sandstone, mudstone, and
shale. Canadian Geotechnical Journal, 18(4): 502-13.
Zhang, L., and Einstein, H. H. 1998. End Bearing Capacity of
Drilled Shafts in Rock. Journal of
Geotechnical and Geoenvironmental Engineering, 124(7):
574-584.
Page 43 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
44
FIGURE CAPTIONS
Fig. 1. Relationship between the measured initial rock socket
concrete/rock interface
average roughness height (∆r) with the unconfined compressive
strength (qu) of
weak rock (after Collingwood 2000; Asem 2018). Solid lines
represent the
method of Seidel and Collingwood (2001) for the prediction of ∆r
based on qu.
Fig. 2. Definition of the peak side resistance (fsp) and initial
shear stiffness (Ksi) using
measured shear stress (fs) and displacement (δ)
relationships.
Fig. 3. Relationship between modulus of deformation of rock mass
(Em) from base
resistance database, plate load tests, and side resistance
database and the weak
rock unconfined compressive strength (qu) (after Asem 2018).
Fig. 4. Load testing methods used in the load test database
(after Asem 2018).
Fig. 5. Variation of peak side resistance (fsp) with unconfined
compressive strength (qu)
of weak rock (data from Asem 2018).
Fig. 6. Variation of peak side resistance (fsp) with deformation
modulus of weak rock
mass (Em) (data from Asem 2018).
Fig. 7. Variation of peak side resistance (fsp) with rock socket
diameter (B) and rock
socket length (L) (data from Asem 2018).
Fig. 8. Variation of peak side resistance (fsp) with rock socket
embedment depth from the
ground surface, and top of rock (DGS and DTOR, respectively)
(data from Asem
2018).
Page 44 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
45
Fig. 9. Variation of the normalized peak side resistance (α =
fsp/qu) with normalized
initial normal stress on rock socket walls (σno/σp), and
unconfined compressive
strength (qu) with method of Hassan and O’Neill (1997) (data
from Asem 2018).
Fig. 10. Variation of modulus of deformation of weak rock mass
(Em) and peak side
resistance (fsp) with depth of embedment from the ground surface
(DGS) for
drilled shaft load tests in weathered Melbourne Siltstone (data
from Williams
1980).
Fig. 11. Variation of modulus of deformation of weak rock mass
(Em) and peak side
resistance (fsp) with depth of embedment from ground surface
(DGS) for drilled
shaft load tests in weathered Melbourne Siltstone (data from
Williams 1980).
Data present an example of a case where fsp decreases with
embedment depth.
Fig. 12. Variation of the displacement required to mobilize the
peak side resistance (δp)
with the rock socket diameter (B) (data from Asem 2018).
Fig. 13. Variation of the calculated LRFD resistance factors
with structure span length and
target reliability index.
Page 45 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
1
FIGURES
Figure 1
Page 46 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
2
Figure 2
Page 47 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
3
Figure 3
Page 48 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
4
Figure 4
Top of rock
Groundsurface
B
Base loadcell
Test shaftReaction
shaftsDGS DTOR
Reaction frame
Top loadcell
Top of rock
Ground surface
B
Test shaft Reaction shaftsDGS DTOR
Reaction frame
Top load cell
Straingauge
Top of rock
Ground surface
B
Void or compressible
base
Test shaftReaction
shaftsDGS DTOR
Reaction frame
Top load cell
Top of rock
Ground surface
B
O-cell
Test shaftDGS DTOR
Strain gauge
Top of rock
Ground surface
B
Plug or anchor
Reactionshafts
DGS DTOR
Reaction frameT
Method (a) Method (b) Method (c)
Method (d) Method (e)
Page 49 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
5
Figure 5
Page 50 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
6
Figure 6
Page 51 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
7
Figure 7
Page 52 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
8
Figure 8
Page 53 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
9
Figure 9
Page 54 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
10
Figure 10
Page 55 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
11
Figure 11
Page 56 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
12
Figure 12
Page 57 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
13
Figure 13
Page 58 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
1
TABLES
Table 1 Existing empirical coefficients ( and n) for peak side
resistance (fsp) of rock
sockets in weak rocks.
Design method nRosenberg and Journeaux (1976) 0.362 0.52Horvath
and Kenney (1979) 0.20 0.50Pells et al. (1979) 0.20 1.00Williams
(1980) 0.43 0.37Horvath et al. (1983) 0.20-0.30 0.50Rowe and
Armitage (1984) 0.45 0.50Carter and Kulhawy (1988) 0.20 0.50Stark
et al. (2013) 0.30 1.00Stark et al. (2017) 0.31 1.00
Page 59 of 67
https://mc06.manuscriptcentral.com/cgj-pubs
Canadian Geotechnical Journal
-
Draft
2
Table 2 Summary of the existing drilled shaft load test database
(after Asem 2018).
Database Number of load test
Rock type Rock socket
diameter, B (mm)
Unconfined compressive
strength, qu (MPa)
Horvath and Kenney (1979)
76 (in situ load tests on small- and large-diameter rock
sockets)
Shale, claystone, mudstone, siltstone, sandstone, limestone and
chalk
13 to 1220 0.35 to 110 for shale and mudstone, 7 to 24 for
sandstone and 1 to 7 for limestone and chalk
Williams (1980) (in situ load tests, embedment depth, DGS <
2000 mm)
Weathered