AGSAnisotropy 2 D3 DLEModels Bar Weekes

AUSTRALIAN
GEOMECHANICS
JOURNAL AND NEWS OF THE AUSTRALIAN GEOMECHANICS SOCIETY ISSN 0818-9110
VOLUME 52: NO.4 DECEMBER 2017
INCLUDING
Natural hazards, risk, and the resilience of
transportation infrastructure: an example of
risk-based geotechnical asset management
Novel pile design for multi-level car park above
twin rail tunnels
Slope risk analysis supporting post-disaster
recovery: The 2016 Kaikoura Earthquake
Managing the risks associated with acid sulfate
rock in NSW road projects
Underpinning the concord road bridge under
traffic: Westconnex M4 East Project
Use of temporary anchors in reinforced soil wall
construction for the M4 Motorway widening
Directional shear strength models in 2D and 3D
limit equilibrium analyses to assess the stability
of anisotropic rock slopes in the Pilbara Region
of Western Australia
A case study of deep excavation and retention
design for Sydney Metro Northwest
DIRECTIONAL SHEAR STRENGTH MODELS IN 2D AND 3D LIMIT
EQUILIBRIUM ANALYSES TO ASSESS THE STABILITY OF ANISOTROPIC
ROCK SLOPES IN THE PILBARA REGION OF WESTERN AUSTRALIA
Neil Bar
Gecko Geotechnics Pty Ltd, Cairns, Australia
Geoffrey Weekes
Red Rock Geotechnical Pty Ltd, Perth, Australia
ABSTRACT
The bedded iron ore and gold deposits in the Pilbara Region of Western Australia are hosted in highly anisotropic rock
masses. For iron ore, these comprise strong banded iron formations discretely interbedded with very weak shales. Gold and
other precious metals deposits in the same region are hosted by interbedded siltstones, sandstone and quartzites. Slope
instability mechanisms generally involve sliding along bedding planes combined with joints or faults acting as release
planes.
Slope stability modelling techniques applied to highly anisotropic rock masses have developed significantly over the years
from basic kinematic analysis in the 1990’,s through to two-dimensional limit equilibrium analysis and numerical modelling
in the 2000,s with the available software increasing in functionality and complexity over time. Limit equilibrium analysis
software now offer a range of options to model the behaviour of anisotropic rock masses. The results obtained by these
different models can vary significantly. It has been found that selecting either inappropriate anisotropic shear strength
models for a given rock mass or using poorly calibrated models typically result in overly conservative slope designs. This
paper presents case studies which illustrate the importance of geological interpretations, correct constitutive model
selection, the use of non-linear shear strengths, and 2D and 3D modelling approaches.
1 INTRODUCTION
The central Pilbara region of Western Australia is renowned for its abundance of economically extractable, bedded iron ore
deposits near the townships of Newman, Tom Price and Paraburdoo. It is also host to several gold and other precious metals
deposits in the vicinity of Port Hedland and Marble Bar townships. Hundreds of open cut mines are operated by major
mining companies across the region with single operations often having access to several individual pits. Due to the broad
regional expanse of the operations, particularly in the iron ore sector, a very high extraction rate is achieved despite vertical
development rates remaining relatively low (typically one to three benches or 10m to 30m per year in an individual iron ore
pit). Final pit depths generally range from less than 100m to 350m in iron ore and less than 450m in gold and other precious
metals deposits.
In the iron ore sector, access to several individual pits allows mining companies to relocate mining fleets during unexpected
events such as unpredicted mineralogy or geotechnical issues, enabling production to be safely maintained. This luxury
allows short to medium term mine schedules to be maintained in the event of slope instability in a single pit. With suitably
monitored and managed pit slopes, and a good understanding of geological and geotechnical conditions, quite aggressive
slope profiles can be considered to maximize ore recovery and reduce overburden removal but still maintain a relatively low
business risk profile in terms of mine schedule delivery.
Slope stability modelling techniques have significantly improved over the years from basic kinematic analysis in the 1990`s
through to two-dimensional limit equilibrium analysis and simple finite element modelling using PC’s in the 2000’s.
However, in both the iron ore and gold deposits in the Pilbara region, open pit mines are typically designed using relatively
limited data density and the routine use of more advanced numerical modelling techniques such as FLAC and UDEC is
generally not practicable, nor is it always warranted. The recent development and improvement of simple-to-use 3D limit
equilibrium analysis software provides an additional method for identifying potential issues in slope designs.
AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017
AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 91

DIRECTIONAL SHEAR STRENGTH MODELS IN 2D AND 3D LIMIT EQUILIBRIUM ANALYSES TO ASSESS THE
STABILITY OF ANISOTROPIC ROCK SLOPES IN THE PILBARA REGION OF WESTERN AUSTRALIA
BAR AND WEEKES
2 GROUND CONDITIONS
2.1 PILBARA IRON ORE DEPOSITS
Pilbara iron ore deposits occur within banded iron formations of the Hamersley Group which comprises Archaean to
Proterozoic marine sedimentary and volcanic rocks. Geological structures play a key role in the location, geometry and
preservation of high grade iron ore bodies. The structural evolution of the Hamersley province is considered to be well
understood and is documented in Dalstra (2014). In general terms, it comprises normal faulting and thick-skinned tectonics
in the west and more intense folding, minor thrust faulting and possible thin-skinned tectonics in the east.
The stratigraphic units of economic interest consist of banded iron formation (BIF) with interbedded carbonates and shales.
BIF can vary in thickness due to differing amounts of carbonate dissolution & silica replacement during iron ore enrichment
formation (Harmsworth et al. 1990) and typically contain thick interbedded shale bands, some of which make excellent
stratigraphic marker horizons in the mining areas as they are remarkably persistent across hundreds of kilometres
throughout the Hamersley province.
Planar sliding along adversely oriented shale bands is the most common mode of slope instability from bench to overall
slope scale in mining operations and also within natural slopes of the Pilbara region (Bar, 2012; Seery, 2015).
Rock mass and bedding shear strengths are typically well understood in the Pilbara region as relatively limited variation
exists across individual deposits that are situated within the same stratigraphic unit (Bar et al. 2016). Intact rock and
subsequently rock mass shear strength, particularly in BIF-dominated stratigraphic units, may vary quite significantly with
weathering, and to a lesser extent, alteration. No significant difference between the shear strength of bedding planes of shale
and BIF (banded iron formation) units across the Pilbara (Maldonado & Haile, 2015). Maldonado & Mercer (2015) further
indicated that the shear strength of shale bedding planes is independent of weathering grade. Failure back-analyses, site
specific drilling & laboratory testing provide a means of assessing variation in shear strengths and material density.
In iron ore deposits, two scales of anisotropy exist (Figure 1):
1. Bedding scale – between individual bedding planes (e.g. BIF-BIF or shale-shale bedding planes).
2. Banding scale – between known specific bands within stratigraphic layers (e.g. shale band DS9 and the bands of
BIF either side in the Dales Gorge Member).
Figure 1. Intricately folded bedding scale anisotropy (A: individual BIF-BIF bedding planes) and gently folded
banding scale anisotropy (B: white-light yellow layers are discrete 0.5m to 1.0m thick shale bands within anisotropic
banded iron formation in the Dales Gorge Member of the Brockman Iron Formation).
92 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

BAR AND WEEKES
2.2 PILBARA GOLD DEPOSIT

A large gold deposit occurs within Proterozoic stratigraphy in the Yeneena Supergroup. Rock types include calcareous and
argillaceous siltstones, sandstones and quartzites. Geological structure is complex and is the primary reason behind the
mineralization.
Rock mass and bedding shear strengths in the gold deposit are well understood. Intact rock and rock mass shear strength,
and to a lesser extent, bedding shear strength, vary with the degree of weathering and the type of alteration (clay or silica
enrichment).
Planar sliding along adversely oriented bedding planes within siltstone, sandstone and quartzite are the most common mode
of slope instability. Although both bedding and banding scale anisotropy exist, only bedding scale anisotropy has influenced
slope stability thus far (Figure 2).
Figure 2. Example of folded, near-laminated weathered siltstone (A: siltstone-siltstone bedding planes) and planar
failure in fresh quartzite (B: quartzite-quartzite bedding planes). Both have bedding scale anisotropy.
3 MODELLING CONSIDERATIONS
3.1 ANISOTROPY
Anisotropy, as defined in engineering geology, refers to a rock whose engineering properties (such as strength and
permeability) vary with direction. Anisotropy is very common and present everywhere. Isotropy is rare (Barton & Quadros,
2015). Anisotropy is produced as a consequence of the geological history of the rock, and generally has its origins in the
varying mineralogical composition of different layers, and/or preferred orientation of mineral grains. Distinctive bedding
planes are produced in sedimentary rocks due to depositional cycles; distinctive flows and flow-tops in basalts; foliation in
gneisses; schistosity in schists; cleavage in slates, and faults through all the above. Igneous dykes, sills, weathered horizons,
and dominant joint sets are also forms of anisotropy or inhomogeneity. Seismic velocity through rock masses is also
strongly influenced by anisotropy (Barton 2007). Methods of estimating rock mass quality also accommodate anisotropy,
and include:
• Rock quality designation, RQD (Deere, 1963; Deere et al. 1967).
• Q-system (Barton et al. 1974; Barton & Grimstad, 2014).
• Rock mass rating, RMR (Bieniawski, 1976; Bieniawski, 1989).
• Q-slope (Barton & Bar, 2015; Bar & Barton, 2016).
• Geological strength index, GSI (Hoek, 1994).
In geotechnical practice, it has been observed that planar failure along bedding, joints, foliations, etc. tends to occur only if
the dip direction of a plane is within a certain angular range of the slope face dip direction. Typically, a value of 20 to 30
degrees is used based on empirical observations (Goodman, 1980; Hudson & Harrison, 1997).

BAR AND WEEKES
Anisotropy (or true dip) can only be correctly modelled in three-dimensions. When creating two-dimensional vertical cross-
sections of a slope, the representation of any anisotropy is inherently likely to result in ‘apparent dip’ (i.e. as seen from the
vertical cross-section, which is not completely perpendicular to the anisotropy, especially since geology and therefore,
anisotropy is variable). This usually means one of two things:
1. Apparent dip is shallower than true dip and discontinuities that do not daylight or do not have potential for planar
sliding in 3D, may appear to be problematic in 2D. Anisotropy applied to a numerical model where planar sliding
is not kinematically admissible can result in highly conservative results as is shown by Bar et al. (2016).
2. Apparent dip is shallower than true dip and discontinuities that do daylight and have potential for planar sliding
may appear slightly more favourable in 2D than in 3D (due to appearing “flatter” than the critical dip angle for
sliding).
In the case of the anisotropy in the Pilbara region:
• Bedding scale anisotropy can and should be modelled using directional shear strength models in 2D or 3D.
• Banding scale anisotropy is best modelled using discrete weak bands in both 2D and 3D models rather than with
directional shear strength models.
The use of directional shear strength models for banding scale anisotropy allows the weakness plane to appear ubiquitously
in this slope, rather than at its actual discrete location. This usually results in over-conservatism. Seery (2015) compared
using ubiquitous directional shear strength models against discretely located shale bands within the Dales Gorge Member of
the Brockman Iron Formation. He concluded that the directional shear strength model did not necessarily honour the
geology, particularly due to the widely spaced shale bands in the Dales Gorge Member. In order to achieve an optimal slope
design, it is necessary to understand the geology, and have the ability to discretely model banding scale anisotropy (i.e.
discretely model shale bands) coincidentally with bedding scale anisotropy, as shown in a 2D cross-section in Figure 3
(same principle equally applies in three-dimensions). It would certainly be appropriate to use a ubiquitous shale band
strength model however where the position of discrete shale bands is not known.
Figure 3. Example of banding scale anisotropy (weak shale bands modelled discretely) and bedding scale anisotropy
(Newman Member and Mineralised Newman rock masses defined using directional shear strength models) used in
coincidence in a detailed 2D model.
3.2 SHEAR STRENGTH CRITERIA

The Hoek-Brown failure criterion (Hoek & Brown, 1980; Hoek et al. 2002) assumes isotropic rock conditions and rock
mass behaviour. However, the introduction of the geological strength index (Hoek, 1994) and modification of the geological
strength index for heterogenous rock masses such as flysch (Marinos & Hoek, 2001) promoted its wider use, even in
anisotropic rock masses. A Modified Hoek-Brown failure criterion for anisotropic intact rock was proposed following
research work into the effects of anisotropy for unconfined compressive strength (σ c) as well as Hoek-Brown constants, ‘m’
and ‘s’ (Saroglou & Tsiambaos, 2008).

BAR AND WEEKES
The Hoek-Brown failure criterion is the most commonly used shear strength model for rock mass shear strength. It derives a
non-linear relationship between normal stress and shear strength. Assuming isotropic conditions are present, three key
parameters need to be derived for its use:
• Unconfined compressive strength, σc, derived from field strength estimates and laboratory testing.
• Dimensionless constant, mi, derived from literature and laboratory testing.
• Geological strength index, GSI, which is also commonly derived from estimates of RMR or Q in drill core or
visually from slope or tunnel exposures.
The Barton-Bandis criterion (Barton & Bandis, 1980; Barton & Bandis, 1990) is a commonly used shear strength model for
discontinuities such as joints and bedding planes where rock-to-rock contact is present, or will be present after minimal
shearing. It derives a non-linear relationship between normal stress and shear strength, and requires three key input
parameters:
• Residual friction angle, Φr, laboratory testing.
• Joint roughness coefficient, JRCn, derived from assessing the roughness of discontinuities and appropriate
scaling.
• Joint wall compressive strength, JCSn, derived from field strength estimates and Schmidt hammer use; and
appropriate scaling.
When applicable, for example, in soil mechanics, or potentially in clay-gouge-infilled faults or shears, the use of the linear
Mohr-Coulomb failure criterion is also appropriate.
3.3 DIRECTIONAL SHEAR STRENGTH MODELS
Directional shear strength models allow the application of different shear strengths to different slip surface orientations in
numerical analyses. They can be applied in such a way so as to ubiquitously represent weaker bedding planes, foliations or
pervasive joints upon which shearing is expected to preferentially occur relative to the remainder of the rock mass. In order
to use a directional shear strength model, the following is required:
• Orientation of the anisotropy plane (or multiple anisotropy planes).
• Intact rock or rock mass shear strength .
• Shear strength of the anisotropy plane (e.g. bedding, foliation, joint, etc.).
• Definition of ‘transition’ from intact rock or rock mass shear strength to anisotropy plane shear strength.
Limit equilibrium analysis software SLIDE (two-dimensional) and SLIDE3 (three-dimensional) of RocScience Inc. (Canada)
offer a variety of directional shear strength models, which can be divided into three key models:
• Anisotropic Strength (linear shear strength inputs only – Mohr-Coulomb failure criteria).
• Anisotropic Linear, historically with linear shear strength inputs (Mercer, 2012) and more recently also non-
linear shear strength inputs in the form of normal-shear functions.
• Anisotropic Function with either linear or non-linear shear strength inputs.
Seery (2015) and Bar et al. (2016) compare the different types of directional shear strength models to illustrate how results
from their use can vary. Through the back-analysis of excavated, stable slopes, Bar et al. (2016) showed that the Anisotropic
Linear model was most appropriate for a particular open pit mine hosted in the Brockman Iron Formation, and that the use
of the Anisotropic Strength model leads to an overly-conservative analysis result. Figure 4 graphically illustrates how
different directional shear strength models (principally transition zones) can be applied for particular rock mass conditions
at a specific normal stress (500 kPa in this instance) and specified rock mass, bedding and joint shear strength (Bar et al.
2016).

BAR AND WEEKES
Figure 4. Left: Variances in transition zones for Anisotropic Strength (AS) and Anisotropic Linear (AL) directional
shear strength models; Right: Anisotropic Function (AF) directional shear strength model with bedding linearly
transitioning to rock mass and joints with a very narrow transition to rock mass; Note: radius indicates magnitude
of shear strength in a given direction (Bar et al. 2016).
Extensive work at several other iron ore mines and the aforementioned gold deposits also supports the use of Anisotropic
Linear models in the Brockman Iron and Marra Mamba Iron Formations as well as the Yeneena Supergroup. Typically in
these geological settings, Anisotropic Linear transition parameters can be as follows:
• A=5° is most commonly adopted (based on the variability in geological models, accuracy of geological
compasses and the resolution used for slip surfaces in limit equilibrium analysis software).
• A=10° has in some instances been adopted for highly variable joint orientations or low confidence geology
models.
• B=30° is most commonly adopted (primarily based on back-analyses of several stable and some collapsed slopes
as well as initial discontinuum modelling by Snowden Mining Industry Consultants).
• B=15° has been adopted in some instances where it was supported with the back-analyses of stable and collapsed
slopes.
• B=5-10° has been adopted for pervasive but discontinuous joints (based on back-analyses of stable and collapsed
slopes; shear strengths in the joint-anisotropy plane were scaled based on joint persistence relative to the
surrounding rock mass – a pseudo step-path approach).
3.4 LINEAR APPROXIMATIONS OF NON-LINEAR SHEAR STRENGTHS

Linear approximations (Mohr-Coulomb tangential and secant ‘fit’) of non-linear shear strength envelopes derived from
Hoek-Brown or Barton-Bandis failure criteria can be used to predict the behaviour of rock masses, bedding and other
discontinuities in numerical analyses.

BAR AND WEEKES
Whilst appropriate for strength comparisons, such strength models do have limitations when applied to numerical models in
that they are specific to a relatively narrow normal stress range, and must be estimated as such. Mohr-Coulomb strength
models do however have the advantage of allowing probabilistic data sets to be used in some numerical models.
Figure 5 shows a set of rock mass (Hoek-Brown) and bedding (Barton-Bandis) non-linear shear strength envelopes that
have had linear (Mohr-Coulomb) approximations made between 0.2 and 1.2 MPa. This approximation is subsequently
extrapolated for numerical analyses. This type of approximation would historically have been carried out for the analysis of
inter-ramp or overall slopes, either to use certain directional shear strength models that cannot use non-linear shear strength
inputs or to reduce computational time, neither of which is necessary in most cases today. As illustrated in Figure 5, the
linear approximations overestimate apparent cohesion and total shear strength at low normal stresses (≤0.3 MPa), and
equally overestimate apparent friction angle and total shear strength at high normal stresses (≥1.1 MPa). At a normal stress
of approximately 0.7 MPa, the linear approximation underestimates shear strength.
Figure 6 is used to illustrate the aforementioned effects combined in a directional shear strength model with bedding
dipping at +30° as in Figure 4. In this case, the Anisotropic Linear model using transition parameters, A=10° and B=30°
was adopted. Shear strength is significantly overestimated by the linear approximation when normal stresses are either 0.2
MPa or 1.5 MPa, and slightly underestimated when normal stresses are 0.7 MPa in this example.
Figure 7 presents a practical application to illustrate the negative implications of using linear approximations of non-linear
shear strengths. Figure 7A is an isotropic model using the Hoek-Brown failure criterion (i.e. all anisotropy weaknesses for
bedding planes are ignored) and attains a minimum FoS (factor of safety) of 1.73. Adding anisotropy weaknesses for
bedding planes into the model with linear approximations of non-linear shear strengths (Figure 7B) attains a minimum FoS
of 1.75, which is higher than the isotropic case. This clearly demonstrates the overestimation of apparent cohesion and total
shear strength at low normal stresses, even with the addition of (favourably oriented) weakness planes. The use of
anisotropy weaknesses for bedding planes correctly with non-linear shear strengths identifies an even lower minimum FoS
of 1.42 in Figure 7C, highlighting a possible surficial slip surface not identified in the first two models.
Figure 5. Linear approximations of non-linear shear strengths overestimate shear strength at low and high normal
stresses.

BAR AND WEEKES
Figure 6. Anisotropic linear directional shear strength model (transition parameters: A=10°; B=30°) for the non-
linear shear strengths and linear approximations from Figure 5.
Figure 7. Comparison of 2D limit equilibrium analysis results for a cross-section in a proposed pit pushback; A:
Isotropic model – FoS=1.73; B: Anisotropic model using linear approximations of non-linear shear strengths –
FoS=1.75; C: Anisotropic model using non-linear shear strengths – FoS=1.42.
4 ANISOTROPIC LIMIT EQUILIBRIUM ANALYSIS (2D AND 3D)

Limit equilibrium analysis using directional shear strength models has been undertaken for three case studies using the
following techniques:
1. Selected 2D cross-section of a slope – cross-section of the slope is chosen by the geotechnical engineer as critical
or representative for assessing slope stability. Anisotropy is inherently almost always represented by ‘apparent
dips’ rather than ‘true dips’. The dip of anisotropy is typically modelled in incremental zones of 5-10°.
Depending on the complexity of the slope geometry and geology, anywhere from five to twenty cross-sections
are analysed for a given open pit slope design. This approach is considered standard practice.
2. Selected 3D-extruded cross-section of a slope. Same as above; however, the slope is extruded uniformly-laterally
and the results may give the geotechnical engineer an indication of the lateral extent of a potential slope
instability if conditions remain constant (laterally). This approach could be very useful for some iron ore and coal
deposits that have a long and straight strike.

BAR AND WEEKES
3. Full 3D model of an open pit, or a portion of an open pit. Allows the analysis of complex 3D slope geometry
(e.g. circular pit or convex slope profiles) and full 3D geology. Anisotropy is fundamentally represented using
‘true dips’. The resolution or detail of anisotropy is generated from imported anisotropic wireframes and is
inherently higher than in 2D cross-sections (i.e. higher than 5-10° incremental zones).
For perspective, Figure 8 illustrates the above model types for the Yeneena Supergroup example. The three anisotropic
examples in the following sections include bedding dipping into slope (Yeneena Supergroup example), sub-horizontal
bedding (Brockman Iron Formation example) and bedding sub-parallel to slope (Marra Mamba Iron Formation example).
Figure 8. Limit equilibrium models for Yeneena Supergroup example; A: standard 2D cross-section (SLIDE); B: 3D-
extruded model (SLIDE3); C: Full 3D pit model (SLIDE3).
4.1 YENEENA SUPERGROUP EXAMPLE

The proposed 300m+ deep open pit shown in Figure 8 and Figure 9 is a pushback and deepening of an existing 200m high
pit. As illustrated in the cross-section Figure 9A, bedding in the siltstone, sandstone and quartzite dips into the slope (this is
more or less a ‘true dip’). The degree of weathering influencing intact rock strength decreases with depth. A total of seven
2D cross-sections were used to analyse the proposed open pit design. The full 3D pit model identified the lowest factor of
safety (FoS) in the pit (Figure 9C), which correlated well with the nearby 2D and 3D-extruded cross-sections. All models
used non-linear shear strengths and the Anisotropic Linear directional shear strength models. The full 3D pit model obtained
a slightly higher FoS, which is attributed to a combination of some lateral confinement provided by the gently-concaved
slope profile and the higher resolution and ‘true’ anisotropy modelled.

BAR AND WEEKES
Figure 9. Yeneena Supergroup: Siltstone (highly weathered: red, moderately weathered: orange, slightly weathered:
yellow), Sandstone (green) and Quartzite (light purple); A: 2D cross section (FoS=1.57); B: 3D-extruded cross-
section (FoS=1.55); C: Full 3D Pit Model (FoS=1.61); Lateral confinement of gently-concaved slope profile in full 3D
pit model shows marginally improved FoS.
4.2 BROCKMAN IRON FORMATION EXAMPLE

An early stage design of an open pit with relatively flat bedding that is not dipping in-line with the cross-section (i.e. an
apparent dip is shown in the 2D cross-section). Anisotropy in the 2D and 3D-extruded cross-sections is based on 5-10°
apparent dip increments whereas the full 3D pit model has a higher resolution and true dips.
Near identical slip surfaces are attained in terms of failure depths. However, slip surface shapes are significantly determined
by the 3D geometry (convex slope profile). Irrespective of modelling method, the actual slope design required significant
modification before excavation could commence.

BAR AND WEEKES
Figure 10. Brockman Iron Formation: Joffre Member (blues and cyans), Whaleback Shale (yellows) and Dales
Gorge (green); A: 2D cross section (FoS=0.76); B: 3D-extruded cross-section (FoS=0.81); C: Full 3D Pit Model
(FoS=0.87); True dip of bedding and higher resolution of bedding waviness in full 3D pit model help improve FoS.
4.3 MARRA MAMBA IRON FORMATION EXAMPLE

Bedding in the Marra Mamba Iron Formation dipping at moderate angles into the pit. Overall slope angle is 34°, for a slope
height of approximately 150m, such that bedding is not undercut. Ubiquitous shale model was required as little is known of
the position of the discrete shale bands as previously illustrated in Figure 3.
The 2D and 3D-extruded models give a similar result (Figure 11). However, the full 3D model indicates a 17% increase in
FoS over the 2D model and significantly limits the lateral extent of the potential slip surface compared with the 3D-extruded
model. This is attributed to having the true dip of bedding and higher resolution anisotropy in full 3D.

BAR AND WEEKES
Figure 11. Marra Mamba Iron Formation: A: 2D cross section (FoS=1.96); B: 3D-extruded cross-section (FoS=2.09);
C: Full 3D Pit Model (FoS=2.30); True dip of bedding and higher resolution anisotropy in full 3D pit model prevents
lateral length of slip surface and improves FoS.
5 KEY POINTS
This paper presented several examples and case studies to illustrate the importance of geological interpretations, correct
constitutive model selection, the use of non-linear shear strengths, and 2D and 3D modelling approaches. Readers are
encouraged to consider the following points when dealing with anisotropic (and isotropic) rock masses:
• Understanding the geology remains paramount – no matter how detailed, all models are wrong if the geology is
not understood and appropriately interpreted. Bands of weaker material, such as shale bands in iron ore deposits,
need to be modelled discretely where practicable. Shale bands discretely modelled will be superior to ubiquitous
shale (directional shear strength) models, however, the latter can certainly be applied where little or no
knowledge exists of the position of weak shale bands within the banded iron formation sequence. Geotechnical
engineers or engineering geologists should develop models with as true a representation of actual ground

BAR AND WEEKES
conditions as the available data allows, and attempt to correlate the output of the model with real, monitored
slope performance before using it as a predictive tool. This process should also be revisited when new data
becomes available or as circumstances change.
• Non-linearity is real; and non-linear shear strengths should be used where appropriate. Using linear
approximations of non-linear shear strengths has its risks, primarily in over-estimating shear strength for rock
masses or discontinuities in shallow and very deep-seated failure mechanisms (e.g. future, deeper slope
cutbacks). However, linear approximations (Mohr-Coulomb) do provide intuitive shear strengths which can be
compared between project areas and material types. Additionally, they can be entered into currently available
probabilistic analysis simulations.
• Directional shear strength models that ubiquitously model planes of weakness or anisotropy are very useful,
particularly when calibrated against observed slope performance. Anisotropy is best modelled in 3D using the
true dip of the planes as well as an inherently higher resolution model due to the 5-10° incremental zoning that is
required in 2D cross-sections.
• Full 3D pit models remove some of the conservatism intrinsic to 2D modelling. They require slightly more
computational time; however, they can be used to identify zones with potential issues rather than relying solely
on ‘the correct selection of cross-sections’.
• Groundwater and pore water pressure remains the only aspect of the ‘ground’ that we can physically control (if
permeability and time permit). Although the case studies presented in this paper were not critically affected by
groundwater, pit dewatering and slope depressurization should remain a key focus in any form of open pit slope
design, and therefore, modelling.
The combined use of 2D and 3D models allows geotechnical engineers and engineering geologists to gain a better
understanding of rock mass behaviour, particularly in anisotropic rock masses. As a result, doing so can provide a more
robust solution for slope design optimization and aid in unlocking and maximizing value from a mineral deposit.
6 ACKNOWLEDGEMENTS
The authors sincerely thank Dr Thamer Yacoub and the RocScience team for their guidance and support with the SLIDE3
software.
7 REFERENCES
Bar N (2012) Performance driven slope management, design and optimization at Brockman Operation, Rio Tinto Iron Ore.
In Proc. 9th Young Geotechnical Professionals Conference (9YPGC), St Kilda, 11-14 July 2012, Australian
Geomechanics Society: 19-24.
Bar N, Barton N (2016) Empirical slope design for hard and soft rocks using Q-slope. Proc. 50th US Rock Mechanics /
Geomechanics symp. ARMA 2016, Houston, 26-29 June 2016. ARMA16-384.
Bar N, Johnson TM, Weekes G (2016) Using directional shear stress models to predict slope stability in highly anisotropic
rock masses. In Ulusay et al. (eds.), Rock Mechanics and Rock Engineering: From the Past to the Future; Proc.
ISRM int. symp. Eurock 2016, Cappadocia, 29-31 August 2016: 595-600.
Barton N (2007) Rock Quality, Seismic Velocity, Attenuation and Anisotropy, Taylor & Francis Group, London UK.
Barton N, Bandis S (1980) Technical Note: Some Effects of Scale on the Shear Strength of Joints. Int. J. Rock Mech. Min.
Sci. & Geomech. Volume 17: 69-73, Pergamon Press Limited, UK.
Barton N, Bandis S (1990) Review of predictive capabilities of JRC-JCS model in engineering practice. In Rock joints,
Proc. Int. Symp. on Rock Joints, Loen, Norway: 603-610.
Barton N, Bar N (2015) Introducing the Q-slope method and its intended use within civil and mining engineering projects.
In Schubert & Kluckner (ed.), Future Development of Rock Mechanics; Proc. ISRM reg. symp. Eurock 2015 & 64th
Geomechanics Colloquium, Salzburg, 7-10 October 2015: 157-162.
Barton N, Grimstad E (2014) Forty years with the Q-system in Norway and abroad. Fjellsprengningsteknikk, Bergmekanikk,
Geoteknikk, NFF, Oslo, 4.1-4.25, 25p.
Barton N, Lien R, Lunde J (1974) Engineering classification of rock masses for the design of tunnel support. Rock
Mechanics 6: 189-236.
Barton N, Quadros E (2015) Anisotropy is Everywhere, to See, to Measure, and to Model, Journal of Rock Mechanics and
Rock Engineering, Volume 48 Issue 4, July 2015: 1323-1339.
Bieniawski ZT (1976) Rock mass classification in rock engineering. In Bieniawski (ed.), Exploration for rock engineering;
Proc. of the symp., Cape Town, Balkema, 97-106.

BAR AND WEEKES
Bieniawski ZT (1989) Engineering Rock Mass Classifications: A Complete Manual for Engineers and Geologists in
Mining, Civil, and Petroleum Engineering, New York, Wiley, pp272.
Dalstra HJ (2014) Structural evolution of the Mount Wall region in the Hamersley Province, Western Australia and its
control on hydrothermal alteration and formation of high-grade iron deposits. Journal of Structural Geology Volume
67, Elsevier: 268-292.
Deere DU (1963) Technical description of rock cores for engineering purposes. Felsmechanik und Ingenieurgeologie (Rock
mechanics and engineering geology) 1: 16-22.
Deere DU, Hendron AJ, Patton F, Cording EJ (1967) Design of surface and near surface excavations in rock. In Fairhust
(ed), Proc. 8th U.S. symp. Rock Mechanics: Failure and Breakage of Rock, AIME, New York: 237-302.
Goodman RE (1980) Introduction to Rock Mechanics, Chapter 8, John Wiley, Toronto, Canada: 254-287.
Harmsworth RA, Kneeshaw M, Morris RC, Robinson CJ, Shrivastava PK (1990) BIF-derived iron ores of the Hamersley
Province. Geology of the Mineral Deposits of Australia and Papua New Guinea Volume 1, Australasian Institute of
Mining and Metallurgy Monograph 14: 617-642.
Hoek E (1994) Strength of rock and rock masses, ISRM News Journal Vol 2(2): 4-16.
Hoek E & Brown ET (1980) Empirical strength criterion for rock masses. J. Geotech. Engng Div., ASCE 106 (GT9): 1013-
1035.
Hoek E, Carranza-Torres CT & Corkum B (2002) Hoek-Brown failure criterion – 2002 edition. In: Proceedings of the 5th
North American Rock Mechanics Symposium, Toronto, Canada: 267-273.
Hudson JA & Harrison JP (1997) Engineering Rock Mechanics – An Introduction to the Principles, Pergamon Press.
Maldonado A, Haile A (2015) Application of ANOVA and Tuckey-Cramer statistical analysis to determine similarity of
rock mass strength properties across Banded Iron Formations of the Pilbara region in Western Australia. In Proc.
SAIMM Slope Stability 2015, 12-14 October 2015, Cape Town, South African Institute of Mining and Metallurgy:
15-33.
Maldonado A, Mercer KG (2015) An investigation into the shear strength of bedding planes in shale materials from the
Hamersley Group rocks in the Pilbara region of Western Australia. In Proc. SAIMM Slope Stability 2015, 12-14
October 2015, Cape Town, South African Institute of Mining and Metallurgy: 47-62.
Marinos P & Hoek E (2001) Estimating the geotechnical properties of heterogenous rock masses such as Flysch. Bull.
Engng. Geol. Env. Volume 60: 85-92.
Mercer KG (2012) The History and Development of the Anisotropic Linear Model: Part 1. Australian Geomechanics
Newsletter July 2012, Nedlands, ACG.
Saroglou H & Tsiambaos G (2008) A modified Hoek-Brown failure criterion for anisotropic intact rock. International
Journal of Rock Mechanics & Mining Sciences Volume 45 (2): 223-234.
Seery JM (2015) Limit equilibrium analysis of a planar sliding example in the Pilbara Region of Western Australia –
comparison of modelling discrete structure to three anisotropic shear strength models. In Proc. SAIMM Slope
Stability 2015, 12-14 October 2015, Cape Town, South African Institute of Mining and Metallurgy: 681-696.

AGSAnisotropy 2 D3 DLEModels Bar Weekes

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AGSAnisotropy 2 D3 DLEModels Bar Weekes

Uploaded by

Copyright:

Available Formats

AUSTRALIAN

VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 91

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

92 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

2.2 PILBARA GOLD DEPOSIT

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 93

3.2 SHEAR STRENGTH CRITERIA

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

94 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 95

3.4 LINEAR APPROXIMATIONS OF NON-LINEAR SHEAR STRENGTHS

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

96 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 97

4 ANISOTROPIC LIMIT EQUILIBRIUM ANALYSIS (2D AND 3D)

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

98 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

4.1 YENEENA SUPERGROUP EXAMPLE

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 99

4.2 BROCKMAN IRON FORMATION EXAMPLE

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

100 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

4.3 MARRA MAMBA IRON FORMATION EXAMPLE

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 101

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

102 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017 103

AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

104 AUSTRALIAN GEOMECHANICS | VOLUME 52: NO.4 DECEMBER 2017

You might also like