Anda di halaman 1dari 13

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/323074459

. CHANGES IN STATE OF PRACTICE Two Methods for Development of Rock Shear


Strength Parameters for Risk-Informed Evaluation of Concrete Gravity Dams
Founded on Clean Discontinuous Ro...

Conference Paper · February 2018

CITATIONS READS
0 21

3 authors, including:

Michael S. Mccray
US Army Corps of Engineers
3 PUBLICATIONS   1 CITATION   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Bluestone Dam View project

All content following this page was uploaded by Michael S. Mccray on 09 February 2018.

The user has requested enhancement of the downloaded file.


ARMA 17-863

Two Methods for Development of Rock Shear Strength


Parameters for Risk-Informed Evaluation of Concrete
Gravity Dams Founded on Clean Discontinuous Rock
Michael S. McCray, P.G. and Kimberly D. Davidson, P.G.
USACE Dam Safety Modification Mandatory Center of Expertise, Huntington, WV, USA
Glenn A. Nicholson, Ph.D., P.E.
USACE Research and Development Center (Retired), Vicksburg, MS, USA

Copyright 2017 ARMA, American Rock Mechanics Association


This paper was prepared for presentation at the 51st US Rock Mechanics / Geomechanics Symposium held in San Francisco, California, USA, 25-
28 June 2017. This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and
critical review of the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any position of ARMA,
its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written
consent of ARMA is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 200 words; illustrations may not be
copied. The abstract must contain conspicuous acknowledgement of where and by whom the paper was presented.
ABSTRACT: Implementation of dam safety portfolio risk management requires risk assessments on all projects operated by the US
Army Corps of Engineers (USACE), including those previously reanalyzed and modified, to ensure modifications adequately reduce
risk within USACE tolerable risk guidelines. Examples of such projects include Delaware Dam, modified in 1993 for a stability
deficiency, and Bluestone Dam, which is currently being modified for stability and hydrologic deficiencies. The risk assessment
requires selection of upper and lower bounds and expected values of bedrock shear strength parameters representative of all potential
sliding failure planes. Because of improvements in laboratory testing equipment and changes in analysis methodology over the past
approximately thirty years, older data from direct shear testing of intact rock specimens used for analysis with the shear friction
method was discarded and new testing programs were conducted to estimate shear strength parameters for analysis by limit
equilibrium method. The shear strength parameters assigned to the structural wedge at both projects are represented by the sliding
strength of a clean discontinuity. The basic phi angle can be obtained from testing data, but selection of an appropriate inclination
angle (i-angle) is generally difficult. For Bluestone, site-specific first order i-angle measurements were made from excavations of
foundation rock downstream of the dam toe. Selection of the first order i-angle for Delaware proved more challenging because site-
specific first order i-angle measurements were not available. Instead, optical televiewer logs and data from direct shear tests of natural
fractures were used to inform the selection. This paper explains the rationale used in selecting upper and lower bounds and expected
shear strength parameters for the structural wedge of these two dams, and contrasts the methodology variations employed due to
differing levels of available site data.

1. CHANGES IN STATE OF PRACTICE original design, it was common to assess sliding stability
using the shear strength of the intact concrete-to-rock
In 2005, the U.S. Army Corps of Engineers moved from a contact and the intact shear strength of the foundation rock,
primarily standards-based approach for its dam safety rather than using the shear strength of imperfections in the
program (Dam Safety Assurance Program, or DSA) to dam foundation (discontinuities such as bedding planes and
safety portfolio risk management. Implementation fractures) which control shear strength of discontinuous
required risk assessments to be performed on all Corps rock in sliding analyses. The second problem is that testing
dams, including those which had been previously had been performed using older type shear machines that
evaluated under the DSA program with completed were known to produce inconsistent results.
modifications such as Delaware Dam and projects that
were undergoing modification like Bluestone Dam, to 1.1. Stability Analysis Methodology - “Shear-Friction”
ensure that breach risk was reduced within USACE’s and Limit Equilibrium Methods
tolerable risk guidelines. The risk assessment requires
(among other parameters) upper and lower bounds of shear The state of practice for external stability analysis of
strength parameters as well as expected values that are hydraulic gravity structures when Bluestone and Delaware
representative of all identified potential failure planes. Dams were designed in the late 1930s involved the use of
the “shear-friction” method, which was in general use from
Although foundation shear strengths had been previously approximately 1935 to 1981 by the Corps of Engineers.
evaluated during design and later studies of Bluestone and Early versions of the shear-friction method relied only on
Delaware Dams, two problems existed with using the the resisting shear strength of the structural wedge, which
design strength parameters and early laboratory testing was determined using test data from direct shears of intact
results for the new risk assessments. First, at the time of rock, then applied a factor of safety of 4.0. Discontinuities
in the foundation and scale effects of test samples were
ignored.

The Corps replaced the shear-friction concept with limit


equilibrium analyses in 1981 [1]. The basic principle of
this method is that the safety factor relates mobilized shear
strength for stability to available shear strength along the
failure plane. In addition, the minimum factor of safety
against structural sliding was decreased to 2.0 for normal
static loading conditions.

Limit equilibrium also incorporated the available shear


strength of the passive wedge when applicable, and the
shear strength of the structural and passive wedges were
based on the shear strength of discontinuities when
appropriate, instead of the shear strength of intact rock
alone. Fig.1. Typical approximate bilinear and real curvilinear failure
envelopes from modeled discontinuous rock, from Patton [2].
1.2. Shear Strength of a Clean Discontinuity
The basic phi angle (φu) for Eq. (1) and Eq. (2) is calculated
The maximum shear strength envelope (encompassing the from direct shear testing of smooth sawn surface data using
full range of normal stresses) along a clean continuous the least squares best fit equation:
discontinuity is actually curved but can be approximated
by two straight lines given by the two equations, (1) and φu = Σ(XY)/Σ(X2) (3)
(2) below, and shown graphically in Fig. 1 [2]. The shear
strength along a clean (i.e., without any filler material [3]), The sliding shear strength of a clean discontinuity also
continuous discontinuity with normal stresses high enough includes the additional strengthening influence of the of
to suppress asperity override is given by the following the 1st order i-angle (angle of asperity), also sometimes
equation: referred to as the joint roughness angle, which is caused by
irregularities along discontinuities. When possible, 1st
τf = σn∙ tanφu + ca (1) order i-angles are measured in the dam’s foundation rock
or very near the foundation location and measured in the
The shear strength along a clean continuous discontinuity direction of sliding. The primary difficulty in selecting
with low normal stresses is given by the equation below. design friction values lies in the selection of an appropriate
i-angle. Discontinuity surfaces or outcrop traces of
τf = σn∙ tan(φu + i) (2)
discontinuities are not frequently available from which to
where base a reasonable estimate of 1st order i-angles. In such
τf = maximum (peak) shear strength at failure cases estimates of the i-angle must rely on sound
engineering judgment and extensive experience in similar
σn = stress normal to the shear plane (discontinuity) geology [3].
σt = normal stress transition line. Normal stress greater
than σt produces a shear strength envelope 1.3. Changes in Shear Strength Testing
represented by Eq. (1); normal stress less than σt
produces a shear strength envelope represented by Although, at the time risk assessments were performed,
Eq. (2). data for direct shearing of discontinuities existed for both
φu = the basic phi angle of a smooth planar sliding surface Bluestone and Delaware Dams, testing had been
i = angle of inclination of 1st order (major) asperities (i- performed using a shear machine of old design (see Fig. 2),
angle) which was still in use for testing samples as late as the early
1990s. Large variations in test results from shear machines
φr = the residual friction angle of the material comprising
of this type were documented in a paper by Nicholson [6].
the asperities (shown in Fig. 1)
The paper describes the results of a testing program done
ca = the apparent cohesion (shear strength intercept) with “standard” test samples and detailed testing protocols
derived from the asperities
to minimize the influence of human error. Testing results
Because normal stresses at Corps projects are typically varied and showed significant scatter because of the
below σt for most rock types, Eq. (2) is generally used to inability of the machines to maintain the line of shear
represent the sliding shear strength [4]. forces through the centroid of the intended shear surface,
the inability to hold a constant normal test load during
shearing, and the friction between the shear machine and
the moving half of the sample. As a result of the study a
standardized direct shear machine was recommended for
all future testing (Fig. 3), and test results produced from
the earlier machines are now generally regarded as suspect.

A
Fig. 3. Modern automated direct shear machine.

The relationship between strength and test specimen size


is commonly referred to as scale effects, and is especially
important for discontinuous rock [3]. Scale effects are the
primary cause of difficulty in selecting in-situ rock
strength parameters based on laboratory test results. “The
key factor behind the changing behavior and strength with
increasing scale is the involvement of different sizes of
asperities in controlling the peak behavior of different
lengths of joints” [8].

B C

Fig. 2. 1940s direct shear machine similar to those used through


the early 1990s [7]. Intact samples are held in shear blocks (A)
with grout, with a 0.5 to 1.3 cm (3/16 to 1/2 in) gap between
shear blocks. Normal stress is applied by compressing a spring
between the stationary shear block and retaining plate (B). Once
the desired stress is reached, the nuts are tightened against the
retaining plate and the coil spring maintains the normal stress.
The frame is rotated 90 degrees to apply the shearing load (C).

The shear machine shown in Fig. 3 is completely computer


controlled with a data acquisition unit. Normal stress is
applied by a computer controlled, air driven Bellofram
pneumatic actuator (mounted on top of the shear frame)
that keeps normal stress constant even if the sample dilates
or contracts. Shearing rate is computer controlled and the Fig. 4. Idealized model of the scale effects on the shear behavior
of non-planar joints, after Bandis [8].
data acquisition unit monitors the shear stress load cell.
Linear variable differential transformers (LVDT) are used
to measure vertical and horizontal movements of the
Fig. 4 graphically depicts the effect that the length of a
sample, which are plotted with shear stress to show both
discontinuity has on the shear strength test result. The
peak and residual strengths. The location of the shear box
angle of inclination in Eq. (2) is the 1st order i-angle and is
and its shear load attachment points ensures that the line of
best represented by the shear stress of the largest sample
action of the shear forces passes through the centroid of the
(sample 4) in Fig. 4. The angle of inclination in a small lab
intended shear surface and the pillow blocks (slide rails)
sample is known as the 2nd order asperity (2nd order i-angle)
along which the shear boxes move are nearly frictionless.
and is best represented by the smallest sample (sample 1).
Second order asperities typically have higher angles of
inclination than 1st order asperities, and therefore small
scale lab tests can over-predict the in-situ shear strength of section (sliding strengths for the fault were evaluated by a
the same (much longer) discontinuity. separate study and are not discussed in this paper). In the
vicinity of the faults, horizontal to low angled slickensided
fractures are common.
2. SHEAR STRENGTH PARAMETER SELECTION -
BLUESTONE DAM

2.1. Bluestone Dam – Project Description

Bluestone Dam is located near the town of Hinton in


Summers County, WV. The dam forms Bluestone Lake
which covers 8 km2 (2,040 ac) and controls a 12,000 km2
(4,620 mi2) drainage area in the Kanawha River Basin.
The structure is a straight concrete gravity dam with an
overall length of 628 m (2,060 ft) and a height of 50 m (165
ft) above the natural stream bed (top of dam El. 467.9
m/1,535 ft). Discharge is accomplished via sixteen gated Fig. 5. Bluestone Dam, prior to modification (1990s).
sluices and twenty-one vertical lift gates at the crest within
the 241 m (790 ft) wide ogee spillway section. Additional 2.2. Bluestone Dam – Project Background
discharge capacity is provided through six 5.5 m (18 ft)
diameter penstocks through a non-overflow section of the Bluestone Dam is currently being modified under
dam. USACE’s DSA Program to correct hydrologic and stability
deficiencies identified during the 1990s [9]. In 2008,
Bluestone Dam is situated within a maturely dissected during a study by a team of Corps and US Bureau of
section of the Appalachian Plateau physiographic Reclamation (USBR) personnel [10], USBR indicated that
province. The dam is situated in the New River Valley, rock strength parameters used for DSA design could
which is cut from the shales and sandstones of the Hinton significantly contribute to project risk even after
Formation, within the Mauch Chunk Group of Upper completion of modifications. In particular, USBR was
Mississippian Age. The main dam rests on the Stony Gap concerned that cohesion had been assigned to the structural
Sandstone of the Hinton Formation which forms the bed of wedge sliding shear strengths of the orthoquartzite (phi=30
the New River at the site and ranges from 16 to 30 m (52 degrees, cohesion=0.77 kg/cm2 / 11 psi) and the
to 100 ft) thick below the present level of the stream. This interbedded orthoquartzite and shale (phi=32 degrees,
sandstone is very hard, gray, dense, well-cemented cohesion=0.42 kg/cm2 / 6 psi) on which most dam
siliceous quartzitic sandstone (orthoquartzite) with monoliths are founded, after it was identified that the two
interbedded zones and occasional lenses of black shale. rock units contained continuous, near horizontal, open
Bedding is sometimes stylolitic although there are locally bedding planes. USBR team members indicated that
cross-bedded zones with bedding planes dipping up to cohesion would not be used in similar situations for their
about 45° and zones showing ripple-like features. The projects.
Stony Gap Sandstone is underlain by a dense, moderately
hard, maroon and gray claystone called the Coney Shale. Rock strength parameters used in the DSA study [9] for
The rocks forming the dam abutments consist of shales and design of modifications were originally published in the
siltstones interbedded with sandstones. The bedrock units mid-1980s at the time when the Corps was changing from
at the site are somewhat lenticular, so that variations in the shear-friction method to limit equilibrium and when it
rock types occur horizontally within individual units. was common to include the apparent cohesion value.
Strength parameters had been selected for the
Bedrock in the abutments dips gently into the valley as a orthoquartzite and interbedded orthoquartzite and shale
result of the presence of the Bellepoint Syncline which is from exploration and testing performed in the 1970s-1980s
aligned perpendicular to the dam axis and intersects the when all testing was done with the old style shear machine.
dam in the right abutment. Beds in the valley bottom are It was decided to conduct a new drilling, sampling, and
nearly horizontal. Vertical to near vertical joints are testing program for the bedrock in question using the
present throughout the foundation and have been measured newer automated shear machine (rock of the abutments
spaced as closely as two feet apart in outcrop. The primary had been sampled and tested between 1999-2001).
joint set is oriented between N66°E and N85°E, with a
secondary set oriented N39°W to N12°W. Two faults have 2.3. Subsurface Investigation and Lab Testing
been identified at the project and underlie portions of six
monoliths on the right side of the dam in the non-overflow
Drilling, sampling, testing and field studies in the valley interbedded orthoquartzite and shale. Because of ongoing
bottom were conducted in 2008 [10]. The drilling program construction in the area of the thrustblock downstream of
consisted of four 4-inch core borings. The testing program the penstock section of the dam, a unique opportunity
evaluated shear strengths of natural fractures and smooth existed to obtain actual 1st order i-angle measurements in
sawn surfaces for both the massively bedded orthoquartzite multiple locations in the valley bottom.
and the shale interbeds which control the strength of the
interbedded orthoquartzite and shale.

2.4. Determination of the Basic Phi Angle

The expected basic phi angle (φu) was calculated from


direct shear testing of smooth-sawn surface samples at
three different normal loads using the Eq. (3). Because
direct shear testing of smooth-sawn surface data typically
has very little scatter, the upper and lower bound lines of
the shear strength envelope can be visually drawn. The
plots of smooth sawn surface data with selected upper
bound, lower bound and expected basic phi angles for Fig. 8. Dewatered thrustblock construction area showing
orthoquartzite and interbedded orthoquartzite and shale are locations of field measurements, 2008.
shown in Figs. 6 and 7.

Fig. 9. Placing the horizontal string line across the vertical rock
face at location 6.
Fig. 6. Plot of orthoquartzite smooth-sawn surface shear test data Field measurements were taken from six locations (Fig. 8)
with upper bound, lower bound and expected basic phi angles.
by placing a horizontal string line across the vertical rock
faces (Fig. 9) or using a laser level on horizontally exposed
bedding surfaces (Fig. 10) and taking offset measurements
of bedding inclination at 0.6 m (2 ft) intervals. The exposed
rock face and bedding plane surfaces allowed inspection of
multiple undulations and orders of scale in amplitude.
Results are plotted in Fig. 11, which show that as the
sampling length becomes longer the i-angle becomes
lower.

Fig. 7. Plot of interbedded orthoquartzite and shale direct shear


of smooth sawn surface test data with upper bound, lower bound
and expected basic phi angles.

2.5. Determination of the 1st Order i-angle

A field investigation was conducted in 2008 to ascertain


the 1st order i-angle of the orthoquartzite and the
Fig. 10. Self-zeroing laser level placed on top of a 4-foot bubble and recent excavations (see Fig. 8) in the penstock area
level at location 1. along with digital camera borehole logs show bedding
planes are open. This information showed that the use of
cohesion in the structural wedge sliding shear strengths for
these materials was not appropriate since fractures and
bedding planes of materials in the valley bottom are very
continuous and open, and because normal stresses of the
dam are insufficient to suppress asperity override.

Fig. 12. Persistent and open discontinuities in orthoquartzite and


interbedded orthoquartzite and shale downstream of the non-
Fig. 11. Sampling distance vs. i-angle from field measurements
overflow section of the dam.
The appropriate expected 1st order i-angle was selected
2.7. Sliding Shear Strength Envelopes
based on the sampling length that is equal to half of the
wavelength (distance from crest to crest) of the bedrock
Structural wedge sliding shear strengths parameters were
undulation that controls sliding, which is the undulation
determined by adding the basic phi angle and 1st order i-
with the highest amplitude (not necessarily the highest
angle values that had been selected for upper bound, lower
angle) parallel to the direction of sliding over the surface
bound and expected strength parameters. The
area that was being analyzed. The controlling undulation
orthoquartzite shear strength envelope is shown in Fig. 13.
was determined to have a wavelength of 8.5 m (28 ft),
The upper bound phi angle of 39 degrees is the sum of
equating to a sampling distance of 4.3 m (14 ft). This
upper bound basic phi angle of the smooth sawn surface
produced a 1st order i-angle of 5 degrees which was
data (31 degrees) plus the upper bound 1st order i-angle (8
selected as the expected value. The upper bound 1st order
degrees). The expected value of 35 degrees equals the
i-angle was selected using a sampling distance of half of expected basic phi angle (30 degrees) plus the expected 1st
that used for the expected value, which produced an upper
order i-angle (5 degrees). The lower bound of 31 degrees
bound i-angle of 8 degrees. The lower bound 1st order i-
equals the lower bound basic phi angle of the smooth sawn
angle was selected by reducing the expected value by the
surface data (29 degrees) plus the lower bound 1st order i-
difference between the expected and the upper bound
angle (2 degrees).
value, 3 degrees, producing an i-angle of 2 degrees.

2.6. Cohesion for Sliding Shear Strength

Use of cohesion in selection of shear strength parameters


for a discontinuity is dependent on the continuity of the
feature and the normal stress on it. If the discontinuity does
not exist over the entire monolith and the rock is partly
intact then the use of cohesion may be appropriate
(although strain compatibility becomes an issue).

Information used to determine characteristics of


discontinuities in the Bluestone Dam foundation include
foundation reports, boring logs, pressure testing results,
and inspection of excavations made during ongoing
construction. Foundation reports showed that very thin
shale laminations are continuous over whole monoliths,
Fig. 14. Interbedded orthoquartzite and shale shear strength
envelope with smooth sawn surface and peak natural fracture
direct shear test data. The best fit line of the natural fracture data
Fig. 13. Orthoquartzite shear strength envelope with smooth is included for comparison with the upper bound. (Note:
sawn surface and peak natural fracture shear test data. The best additional natural fracture shear tests were performed at higher
fit line given by Eq. (3) of the natural fracture data which normal loads to confirm that the transition stress (σt in Fig. 1) is
includes the strengthening influence of the 2nd order i-angle is not exceeded and use of Eq. (2) is therefore appropriate.)
included for comparison with the upper bound which is
calculated with the 1st order i-angle.
3. SHEAR STRENGTH SELECTION – DELAWARE
The structural wedge sliding shear strength envelope for DAM
interbedded orthoquartzite and shale is shown in Fig. 14.
The upper bound value of 27 degrees equals the upper 3.1. Delaware Dam – Project Description
bound basic phi angle of the smooth sawn surface data (19
degrees) plus the upper bound 1st order i-angle (8 degrees). Delaware Dam is located in central Ohio along U.S. Route
The expected value, 23 degrees, equals the expected basic 23 in Delaware County on the Olentangy River, a tributary
phi angle (18 degrees) plus the expected 1st order i-angle of the Scioto River. Construction of the dam was
(5 degrees). The lower bound of 19 degrees equals the completed 1 July 1948. The primary purpose of the dam is
lower bound basic phi angle of the smooth sawn surface flood damage reduction but the dam also provides benefits
data (17 degrees) plus the lower bound 1st order i-angle (2 for water supply, recreation, and fish and wildlife
degrees). conservation.

To check the validity of the selected upper bound


parameter, shear strength envelopes were plotted with the
peak natural fracture data and a least squares best fit line
as shown in Figs. 13 and 14. The least squares best fit line
from Eq. (3) of the peak natural fracture data represents the
basic phi angle with the average 2nd order i-angle and
should normally plot above the upper bound basic phi
angle and 1st order i-angle.

Fig. 15. Delaware Dam concrete gravity spillway section. (Note:


light colored squares on spillway face are locations of anchors
installed in 1993 as part of a stabilization program.)

Delaware Dam consists of a homogeneous embankment


and a concrete gravity spillway section totaling 5,670 m
(18,600 ft) in length with a crest elevation of 291.7 m
(957.0 ft), providing flood control up to El. 288.6 m (947.0
ft). The concrete gravity spillway section is 70.7 m (232 classify joint sets [13] and identify potential failure planes,
ft) in length and includes the gated spillway and outlet labeled A thru E in Fig. 16.
works. The spillway crest is at El. 281 m (922.0 ft) and
has six radial tainter gates (crest gates). The outlet works
which are contained within the concrete gravity spillway
section consist of five gated sluices which are
hydraulically controlled.

The spillway structure is entirely founded on the Delaware


Limestone which is thin to thick bedded, hard, gray to
brown, dolomitic, occasionally cherty, fossiliferous, and
contains shale partings. Excavation of the overburden and
unsuitable limestone produced an average founding
elevation for the spillway structure at El. 209 ± 1.2 m (868
± 4 ft), resulting in an average embedment of about 0.6 m
(2 ft), which coincides with the average bedding thickness
exhibited by the limestone.

Bedding planes are continuous and nearly horizontal


Fig. 16. OTV log of boring C-09-10 showing labeled potential
across the spillway site, and generally oriented N74°E with failure planes (A through E) with aperture and degrees of dip
a dip of 5° NW. Low-angle fractures nearly parallel to which were correlated over all five borings logged by OTV [13].
bedding are also evident within the limestone and strike
N47°E with a dip of 2° NW. Two prominent near-vertical The laboratory testing program [14] included shear testing
and closely spaced joint sets cut the limestone. One joint of natural fractures and smooth sawn surfaces for
set strikes N23°E with a dip of 89° NW and the second limestone and shale.
strikes N50°W with a dip of 88° NE.
3.4. Determination of the Basic Phi Angle
3.2. Delaware Dam – Project Background
As with Bluestone, the expected basic phi angle (φu) was
In the late 1980s and early 1990s, stability of Delaware calculated from direct shear testing of smooth-sawn
Dam’s spillway was reanalyzed using current guidance [1] surface samples at three different normal loads using Eq.
and was documented in a reconnaissance level study [11]. (3) and the upper and lower bounds were visually drawn,
Then, during a screening level risk assessment in 2007, as shown in Figs. 17 and 18.
concerns were expressed about the concrete gravity
spillway section because stability analysis showed that
minimum factors of safety were not met. As with
Bluestone, shear strength parameters published in the 1991
report were based on data produced with the old style shear
machine and were considered unreliable, and therefore a
new subsurface exploration and testing program was
recommended.

3.3. Subsurface Investigations and Lab Testing

Subsurface investigation in 2009 included eleven core


borings in the area of the spillway section, specifically in
the upstream approach, the passive wedge, through the
spillway apron, and on the left and right banks downstream Fig. 17. Plot of limestone smooth-sawn surface shear test data
of the outlet channel. Two borings were angled 30 degrees with upper bound, lower bound and expected basic phi angles.
from vertical in order to intercept near vertical major and
minor joint sets. In an attempt to recover any filling within
the open bedding planes, all vertical borings were sampled
with 15 cm (6 in) core [angled borings were sampled with
5 cm (2 in) core]. Optical televiewer (OTV) logging was
performed on five borings and used to help identify and
The lower bound i-angle was the minimum i-angle,
rounded to 1 degree, while the expected was the average,
4 degrees and the upper bound was the maximum i-angle,
rounded to 7 degrees.

Fig. 18. Plot of shale smooth-sawn surface shear test data with
upper bound, lower bound and expected basic phi angles.

3.5. Determination of the 1st Order i-angle

As previously discussed, i-angles are ideally measured in


the dam’s foundation rock in the direction of sliding with
an increasing sampling distance until the 1st order i-angle
is reached, as was done for Bluestone, however no
outcrops were available at or near Delaware Dam where Fig. 19. Sampling distance vs. i-angle from OTV logs.
measurements could be made. Instead, data from the OTV
logs were used. Measurements of i-angles from the four 3.6. Cohesion for Sliding Shear Strength
vertical core borings and the one angled core boring were
used to estimate the lower bound, expected and upper After studying the original Delaware Dam foundation
bound 1st order i-angles. reports (Fig. 20), boring logs and OTV logs (Fig. 16), it
was decided that horizontal joints in the foundation are
Two issues exist with using i-angles from OTV logs. The continuous and that the structural wedge sliding shear
first is that i-angles are not taken in the direction of strength parameters would therefore use no cohesion.
potential dam sliding and therefore must be corrected using
equation (4):

𝐶𝐶𝐶𝐶 = tan−1 [tan(𝑇𝑇𝑇𝑇𝑇𝑇) × cos(𝐷𝐷𝐷𝐷𝐷𝐷 − 𝑡𝑡𝑡𝑡𝑡𝑡)] (4)

where
𝐶𝐶𝐶𝐶 = apparent i-angle (dip angle) in the direction of dam
sliding Fig. 20. Foundation photo 262. View of west edge of rock
𝑇𝑇𝑇𝑇𝑇𝑇 = true dip angle (i-angle) taken from the OTV logs excavation north of centerline for the east abutment wall taken at
𝐷𝐷𝐷𝐷𝐷𝐷 = dam sliding azimuth = 357 degrees 0+45.5 north, showing a continuous near horizontal
discontinuity from downstream to upstream [15].
𝑡𝑡𝑡𝑡𝑡𝑡 = true dip azimuth, the dip direction azimuth taken
from the OTV logs 3.7. Sliding Shear Strength Envelopes
Second, OTV logs only provide i-angles for a sampling Structural wedge sliding shear strength parameters were
distance equal to the borehole diameter. For Delaware, determined by adding the basic phi angle and the 1st order
data from two different borehole diameters were available: i-angle values that had been selected for upper bound,
the 15 cm (6 in) core boring produced a 19.7 cm (7.75 in) lower bound and expected strength parameters. The shear
hole (sampling distance) and the 5 cm (2 in) core produced strength envelope for limestone is shown in Fig. 21. The
a 7.6 cm (3 in) hole (sampling distance). Typically, as the upper bound value of 38 degrees equals the upper bound
sampling distance increases, the i-angle decreases as seen basic phi angle of the smooth sawn surface data (31
in Fig. 11, but no reduction in i-angle was observed from degrees) plus the upper bound 1st order i-angle (7 degrees).
the Delaware data (see Fig. 19), possibly as a result of the The expected value, 34 degrees, equals the expected basic
small number of samples since only four i-angles are phi angle (30 degrees) plus the expected 1st order i-angle
available for the 7.6 cm sampling distance. (4 degrees). The lower bound of 30 degrees equals the
The absolute values of the corrected i-angles from the lower bound basic phi angle (29 degrees) plus the
larger sampling distance of 19.7 cm (7.75 in) were used to minimum 1st order i-angle (1 degree).
select the lower bound, expected and upper bound i-angles.
When re-evaluating an existing project, care must be taken
to check the validity and accuracy of existing shear
strength data prior to use in new analyses. Changes from
original analysis methodology (shear-friction method)
potentially render rock shear strength parameters from
design inappropriate for use in limit equilibrium analyses
since discontinuities within foundations may not have been
identified or evaluated. In addition, changes in testing
technology have greatly improved the accuracy and
reliability of direct shear test data, and older data can be
considered suspect, at best. The cases of Bluestone and
Delaware Dams presented in this paper highlight these
issues and offer methodologies for selecting the structural
wedge sliding shear strengths parameters with different
types of field and laboratory data.

When shear test results are available from previous


investigations, the type of shear machine used for testing
Fig. 21 Limestone shear strength envelope with smooth sawn
should be assessed. Often, photographs of test samples can
surface and natural fracture direct shear test results.
be used to identify the machine, as can be seen in Fig. 23.
The shear strength envelop for shale partings is shown in Any testing performed with an old style shear machine
Fig. 22. The upper bound value of 34 degrees equals the should be considered suspect for several reasons: the
upper bound basic phi angle of the smooth sawn surface inability of the these types of machines to hold a constant
data (27 degrees) plus the upper bound 1st order i-angle (7 test load during shearing; the inability to maintain the line
degrees). The expected value, 30 degrees, equals the of shear forces through the centroid of the intended shear
expected basic phi angle (26 degrees) plus the expected 1st surface; and the significant frictional losses within the
order i-angle (4 degrees). The lower bound of 26 degrees system during operation. Because Bluestone and
equals the lower bound basic phi angle (25 degrees) plus Delaware Dams both had testing programs performed with
the minimum 1st order i-angle (1 degree). this type of machine, new sampling and testing programs
were performed to re-evaluate foundation sliding
strengths.

Fig. 22. Shale parting shear strength envelope with smooth sawn Fig. 23. 1989 Delaware Dam direct shear samples cast into shear
surface and natural fracture direct shear test results. blocks like those shown in Fig. 2 [16].
Tables 1 and 2 summarize the structural wedge sliding
Again, to validate the upper bound, shear strength shear strength parameters for Bluestone and Delaware
envelopes were plotted with peak natural fracture data and Dams, respectively. Older design values for the projects
a least squares best fit line. have been included for comparison. Smooth sawn surface
test data and i-angle measurements were used to determine
bracketed (upper and lower bounds and expected)
4. SUMMARY AND CONCLUSIONS structural wedge sliding shear strength parameters for risk
assessment. For Bluestone, 1st order i-angles were directly
measured; for Delaware, 2nd order i-angles measured from
OTV logs were corrected for the direction of dam sliding is now being more widely used in the engineering geology
and, using engineering judgment, were used to estimate a field.
1st order i-angle range.

Table 1. Shear strength envelopes for Bluestone Dam with REFERENCES


historical shear strength parameters for comparison
1st Order i-anglei + φ uii = φ (degrees) c (psi) 1. USACE. 1981. ETL 1110-2-256, “Sliding Stability for
Orthoquartzite Concrete Structures”.
Upper Bound 8 31 39 0
Expected 5 30 35 0 2. Patton, F. D. 1966. “Multiple Modes of Shear Failure in
Lower Bound 2 29 31 0 Rock and Related Materials”, Ph.D. thesis, University of
Illinois.
Interbedded Orthoquartzite and Shale
Upper Bound 8 19 27 0 3. Nicholson, Glenn A. 1983. GL-83-13, “Selection of Design
Expected 5 18 23 0 Shear Strengths for Clean Discontinuous Rock”. US Army
Lower Bound 2 17 19 0 Engineer Waterways Experiment Station.
Strengths used in DSA Analyses [9]
Orthoquartzite 30 11 4. USACE. 1994. EM 1110-1-2908, “Rock Foundations”.
Interbedded Orthoquartzite and Shale 32 6
i 5. Nicholson, Glenn A. 1989. “An Empirical Nonlinear Stress
Field measurement i-angles
ii
Dependent Constitutive Relationship for Rock Masses
Direct shear of smooth sawn surface test using Eq. (3) Based on Rock Mass Classification”. Ph.D. thesis, The
Pennsylvania State University.
Table 2. Shear strength envelopes for Delaware Dam with
historical shear strength parameters for comparison 6. Nicholson, Glenn A. 1994. “A Test is Worth a Thousand
i
1st Order i-angle + φ u ii
= φ (degrees) c (psi) Guesses – A Paradox”. American Rock Mechanics
Association, 1st North American Rock Mechanics
Limestone
Symposium.
Upper Bound 7 31 38 0
Expected 4 30 34 0 7. USACE-ORDL. 1944. “Physical Properties of Foundation
Lower Bound 1 29 30 0 Rock, Sutton Dam, Elk River, West Virginia”.
Shale
Upper Bound 7 27 34 0 8. Bandis, S. 1980. “Experimental Studies of Scale Effects on
Expected 4 26 30 0 Shear Strength, and Deformation of Rock Joints.” Ph.D.
Lower Bound 1 25 26 0 thesis, University of Leeds.
Strengths from 1991 Reconnaissance Report [11]
9. USACE. 1998 “Bluestone Lake Dam Safety Assurance
Rock-on-rock (natural fracture) 25 13
Program Evaluation Report”, USACE Huntington District
Grout-on-rock 45 200
internal report.
i
Corrected OTV log i-angles
ii 10. USACE. 2012. “Bluestone Dam Baseline Risk
Direct shear of smooth sawn surface test using Eq. (3)
Assessment”, USACE Huntington District internal report.

For both projects the upper bound shear strength envelopes 11. USACE. 1991. “Delaware Lake Olentangy River, Ohio,
(basic phi angle + 1st order i-angle) were compared to the Reconnaissance Report, Stability Investigation of Dam
least squares best fit line Eq. (3) of the peak natural fracture Spillway”, USACE Huntington District internal report.
data, (basic phi angle + 2nd order i-angle). Because of the
strengthening influence of the 2nd order i-angle on the 12. USACE. 2007. “Documentation Screening for Delaware
natural fracture data, the upper bound shear strength Dam”, USACE Huntington District internal report.
envelope should typically plot below the peak natural
13. Stantec. 2010. “Borehole Geophysical Investigation
fracture least squares best fit line. Performed by ARM”, USACE Huntington District internal
report.
Typically there are not opportunities to measure 1st order i-
angles within foundations of existing projects but, when 14. Stantec. Jun 2010. “Delaware Dam DSA Rock Core Testing
available, measurements increase confidence in the sliding Rock Core Testing Report”, USACE Huntington District
strength parameter estimates. Measurements of 1st order i- internal report.
angles may be done manually as they were for Bluestone,
however new technologies are available which can be used 15. War Department, Corps of Engineers, Huntington District.
to collect this type of data, such as photogrammetry which 1947. “Delaware Reservoir Project, Foundation Report for
Monolith EA-D”, File No. OVHWR-9.
16. US Army Engineer Division, Ohio River, Corps of
Engineers, Ohio River Division Laboratory. 1989.
“Delaware Lake, Results of Laboratory Testing of 4”
Diameter Rock Core”.

View publication stats

Anda mungkin juga menyukai