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

net/publication/257423092

Forensic rock mechanics, Ortlepp shears and other mining induced

structures. Keynote Lecture, RaSiM8

Conference Paper · September 2013

CITATIONS READS

6 1,247

1 author:

Gerrie Van Aswegen

Institute of Mine Seismology
40 PUBLICATIONS   460 CITATIONS   

SEE PROFILE

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

Dynamic fractures in mines View project

All content following this page was uploaded by Gerrie Van Aswegen on 21 May 2014.

The user has requested enhancement of the downloaded file.

 FORENSIC ROCK MECHANICS, ORTLEPP SHEARS AND OTHER MINING INDUCED STRUCTURES
Gerrie van Aswegen
Institute of Mine Seismology, Stellenbosch, South Africa
Simple ‘rules’ that help us to understand the most commonly observed mining induced structures can be listed. Rock tends
to deform in shear – it is simply the easiest way for rock to deform since it requires no volume change. Failure, however, is
mainly in tension, because rock is weakest in tension. The faster (more dynamic) the deformation, the finer the rock fragments
produced.
The most commonly observed mining induced structures are extension fractures. They serve as fossil traces of the 1
trajectories at the very instant of formation, yielding insights into the static and dynamic orientations of past stresses.
Fractographic striae indicate the direction of rupture propagation. In cases where such striae are relatively long and not
strongly radiating, care is required to distinguish them from slickensides which show the slip direction along slip surfaces. In
brittle shear zones the geometry of the internal extension fractures can show whether the shear failure was transpressional or
transtensional. The bending and smearing out of support units are better kinematic indicators of dynamic shearing than
slickensides.
Example cases, where the dynamic failure process was deduced from underground observations, include: the formation of
low angle fractures in Wits gold mine stopes as a result of the dynamic transpressional shears developed between extension
fractures during production blasting, extension fractures above a stope that curve in an anomalous fashion due to dynamic
shearing, the radial pattern of slickensides and smeared out rock anchors pointing back the source of a pillar burst, the
superimposition of slickensides on fractographic striae to explain the dynamic history of tunnel sidewall failure, slickensides
and anchor smearing indicating local sidewall failure being a sub-event of a proximal pillar burst, a 450 kg piece of rock face
ejected 2 m far at near 4 m/s, slab ejection from a stope face at ~3 m/s from a moment magnitude 0.1 faceburst, Ortlepp shears
without breakthrough shear ruptures indicating stiff loading systems, Ortlepp shears with significant breakthrough shear
ruptures and significant displacements indicating soft loading systems, the determination of the directions of rupture
propagation of Ortlepp shears and a particular case of spectacular pulverization in a transtensional dynamic shear.

INTRODUCTION it difficult to produce his monumental work (Ortlepp 1997) in

a modern massive mine where you hardly ever see the rock.
Regional setting
Some observations below were made in the surrounds
This contribution is based mainly on observations in
the gold mines of the Witwatersrand, but reference is also
made to observations in the platinum mines of the western
Bushveld, South Africa. The regional setting of these
formidable ore deposits (Figure 1) are conveniently
summarized for rock mechanics practitioners by Ryder and
Jager2002.
The auriferous conglomerate layers in the
Witwatersrand strata are referred to as ‘reefs’. Two main
reefs are mined in the West Wits area around Carletonville.
The deeper Carbon Leader Reef (CLR) has mainly
quartzites above and below apart from a ~1m shale band in
the hangingwall separated from the reef by approximately
1m. The Ventersdorp Contact Reef (VCR) occurs at the top
of the Witwatersrand sequence and is overlain by a great
thickness of Ventersdorp lava. In the Klerksdorp gold field
further west the two main reefs are the Vaal Reef and again
the VCR.
Because of the hard and brittle nature of the Wits Figure 1. A map of Southern Africa, schematically
quartzites and conglomerates and because the mining showing the geographic distribution of the
method requires mine workers to be within the stopes, Witwatersrand Supergroup and the Bushveld
Complex.
opportunities abound to observe details of rock failure
phenomena at close quarters. Dave Ortlepp may have found of the famous Merensky reef in the Bushveld Complex.

-1-
Despite the totally different origin as part of an Some overlap with two earlier papers, van Aswegen2005
extraordinary layered intrusive complex, the geometry of and van Aswegen and Stander2012 is necessary to allow a
this ore body is very similar to that of the gold reefs of the more comprehensive overview.
Wits Basin and the mining methods are essentially the
same. The Merensky reef (a pegmatitic, feldspathic In the sections below I firstly list and briefly define the
pyroxenite) is part of the 'critical zone' in the Rustenburg mining induced structures commonly observed in our
Layered Intrusion, the intrusion consisting mainly of up to mines. This is followed by several case studies of
7000 m (9000 m in the east) of norites and anorthosites. observations and interpretations to illustrate to some extent
what we can – and what we cannot - learn from
This paper
underground observations. The emphasis then shifts to
In this paper I point out some observations one can Ortlepp shears.
make in mines that may allow insight into the quasi-static
I do not address the dynamic closure of mine openings.
and dynamic rockmass response to mining, trying to answer
the question: can we learn something about the dynamic It is easily measured and is generally well documented.
Suffice to note here that the closure we measure is mainly
rockmass response to mining by physical observations
the non-elastic part (e.g. where the roof dynamically
underground? It is a valid question since detailed
underground observations can be time-consuming and delaminates) and can be more than an order of magnitude
greater than the elastic closure. I only very briefly refer to
possibly not cost-effective.
kinematic sequences. By careful observation of the
In earthquake seismology at least one author appears positions of pieces of rock or other items one can often
rather pessimistic about the value of field observations to decipher the sequence of damaging ground motions. A
give insight into the dynamic faulting processes. Cowan1999 simple example is where rock from a burst face lies
concludes that geologists lack experiment-based criteria underneath pieces of the shaken down hangingwall –
that enable them to distinguish in the field whether indicating that the face first burst and then the dynamic
particular parts of faults slipped dynamically or a- shaking of the roof caused a fall of ground.
seismically. Other authors, however, point to particular
COMMONLY OBSERVED MINING INDUCED
observables that would allow conclusions about the fault
STRUCTURES
slip dynamics other than pseudotachylite. From their high-
velocity friction experiments on siderite-bearing gouge Han General
et al.2007 show that mineral decomposition due to frictional We note in general that because simple shear yields
heating can leave evidence of paleoseismic events along volume constant plane strain (e.g. Ramsay1967, Chapter 3-
shallow crustal faults. In their preface to the 2012 Journal 4.3) and because rock has limited compressibility, rock
of Structural Geology special volume on Physico-chemical tends to deform in shear – it is simply the easiest way for
processes in seismic faults, Di Toro et al.2012 argue that rock to deform. Because rock is weakest in tension it tends
“Field, experimental and theoretical investigations to fail in tension.
complement seismological–geophysical studies of the
seismic cycle”. More recently Smith et al.2013 performed Another simple ‘rule’ says that the higher the strain
low- to high-velocity rock friction experiments that yielded rate, the smaller the fragments of rock at failure. At a very
localized dynamic recrystallization. They report that “ .. slow strain rate, there will be one pre-existing
localized dynamic recrystallization may be an easy-to- discontinuity/flaw with the lowest strength and the optimal
recognize microstructural indicator of seismic slip in orientation for failure relative to the principle stress
shallow, otherwise brittle fault zones”. direction. That weakness will start to grow and a single
crack or shear will result. At higher strain rates, the growth
Because of the short time span between the rate of that single flaw does not allow sufficient
development of particular dynamic features and the deformation in response to the imposed load and more
opportunity to observe them in mines we may have an flaws are mobilized and the rock breaks into smaller pieces.
advantage over field geologists to recognise phenomena of Ultimately extreme load rates yields the rock flour
dynamic origin since the field geologist has to look past the commonly found at rockburst sites (see discussion on rock
potentially destructive effects of later geological processes flour below).
like recrystallization and further deformation.
These simple ‘rules’ largely determine the evolution of
In a few instances here I compare quantitatively the extension fractures, shears and comminuted rock material
physical observations with the seismic source parameters, commonly observed in the deep gold mines of South
but most of the observations remain qualitative. Africa.

Forensic rock mechanics -2-

Extension fractures Slickensides: Kinematic indicators
It is hardly necessary to define extension fractures Slickensides on faults and fresh failure surfaces are
since these are the most common mining induced kinematic indicators of slip (Doblas1998) and are used
structures. It may, however, be overlooked that extension routinely to judge the sense and direction of fault slip. For
fractures are fossil traces of the σ1σ2 plane. We thus get rock mechanics practitioners and mine seismologists it is
insight into the state and orientation of past stress, whether important to distinguish between pre-existing striations on
old fault surfaces and new ones that developed very
the past be minutes or millennia. Note also that the fracture
recently as a result of slip under mining induced stress. This
set is a record of the stress orientation at the very instant of
is normally not too difficult since geologically old
fracturing. So it does not reflect a long history of the static slickensides are characterised by the recrystallization of
stress state. mineral grains thus creating fossil slickensides that are
Normally in rock mechanics we relate the extension relatively strong - they cannot rubbed off.
fracture orientation and curvature to the variations in the Fresh slickensides in mines are soft, being simply the
static stress field, the classic example being the bow-wave grooves left within fault gouge (the gouge itself could be
geometry of extension fractures forming around the ends of ancient or recently formed) as the one surface slid across
developing tunnels in stressed ground. The variations of the other, with the striations indicating the direction of
stress trajectories through zones of variable elastic shear movement. In fault slip the slickenside lineation is parallel
strain in terms of simple elastic shear strain theory and as to the line of intersection of the fault plane and the plane
reflected by extension fracture patterns, were discussed in containing the maximum and minimum principle stress
some detail recently (van Aswegen and Stander2012) and axes, i.e. the σ1σ3 plane. In mine openings, however, they
summarized here as Example Case 1. often simply show the direction in which an unstable block
Fractographic striae slid into the free space.

Fractographic features such as striae, mirror zones and Although, in theory, asymmetries in the groove
hackles reflect dynamic failure and are indicative of local geometries could suggest the sense of slip, this is quite
conditions at the time of failure. They can be useful to difficult to recognise in fresh gouge. In mining practice it is
decipher the details of rockbursting. These fractographic easier to judge the sense of slip from deformed man-made
features have been studied in some detail (see Bahat et al., objects like disrupted anchors as shown in Figure 2. In this
1995
, Sagy et al., 2006) so we generally know how to interpret case the bending of the sheared off remains of an anchor
particular patterns. The fractographic lineations map the through a bedding plane fault leaves no doubt about the
direction and sense of rupture propagation. direction and sense of slip. Where rock broke away around
such anchors and the remaining parts hang in the air they
Since the direction of rupture propagation and the are referred ‘naked tendons' (e.g. Ortlepp et al.2004) and the
direction of slip need not coincide (a simple notion from orientations of these can also be kinematic indicators - see
seismology applied to in-mine fractures) it is quite feasible Example Cases 3-5 below.
to find slickensides (see below) superimposed on - and at
an angle to fractographic lineations as part of a single Shear Zones
dynamic event as explained in Example Case 2. A brief review of shear zone nomenclature here allows
easier descriptions below.
The basic principles for the evolution of the macro
geometry and internal features of shear zones are scale
invariant. A shear zones may start as minor brittle shear and
evolve over time into a continental scale fault zone with
displacement of hundreds of km. Geological descriptions of
shear zones are generally complicated and cover different
structural features over scales that vary several orders of
magnitude. Here we are only interested in newly formed
shear zones that vary from conditions of infinitesimal strain
to shear displacements of tens of cm.
The geometric relations between extension fractures
that form within shear zones under conditions of
infinitesimal shear strain was considered in detail by van
Aswegen and Stander2012. I summarize here. The
Figure 2. Normal slip along a bedding plane fault development of the internal fabrics of brittle shears at the
with slickensides along a sheared off anchor. stage of infinitesimal strain is illustrated in Fig. 3.

3
 Figure 3. Geometric relationships between shear
zones, extension fractures and Riedel shears. Note
that extension fractures form under conditions of Figure 4. Low angle extension fractures (E2) formed by
infinitesimal strain and the schematic depiction of transpressional dynamic shear across primary
strain is here exaggerated. (From van Aswegen and extension fractures (E1). E3 fractures formed under
Stander2012). transtensional shear, thus the sigmoidal shapes. The
thin lines accentuate fractographic striae. (From van
We note that, apart from the shear couple acting on the Aswegen and Stander2012).
shear zone, there is also a normal load that could vary,
allowing transpressional or transtensional shear which
controls the angles between the shear zone walls and the events, up to moment magnitude 1.5, have been recorded in
structures that develop within the shear zone. the immediate surrounds.

From Fig. 3 one can see that the stress trajectories The primary extension fractures (E1) develop some
through zones of variable elastic shear strain may not be distance ahead of the advancing mining face (mining
straight. The fabric formed by closely spaced extension advances left to right in Figure 4) with a spacing of around
fractures may resemble schistosity, i.e. the ductile ‘s’-fabric 1 metre. When a set of E1 fractures end up close to the
normal to the direction of finite compressive strain. In advancing mining face a transpressional shear zone results
ductile shear zones we also distinguish a ‘c’-fabric which is with secondary low angle E2 extension fractures that form
parallel to the shear zone wall. The smaller the angle at a high angle to the E1 fractures. This most likely happens
between s and c, the larger the strain. The orientation of the under dynamic shear load during the blasting of the face
s-fabric is exactly opposite to that of the extension that cause the instantaneous shear couple as indicated in the
fractures, being normal to direction of minimum insert in Figure 4.
infinitesimal compressive strain. It is therefore important to At a later stage, when the set of E2 fractures are
distinguish between ductile shear fabrics and brittle partially (or completely) undermined, a tertiary set of
extension fabrics. Generally the evolution of a ductile shear extension fractures form within the shear zones defined by
fabric is not possible in hard rock under the the E2 fractures. In this case, however, the shear loading is
pressure/temperature/time conditions of mining induced transtensional and the extension fractures are curved,
structures. merging asymptotically with the extension fractures where
EXTENSION FRACTURES, FRACTOGRAPHIC the normal loading approaches zero (Figure 4).
STRIAE AND SLICKENSIDES – EXAMPLE CASES Similar structures have in the past been interpreted as
Example Case 1 – Low angle extension fractures ductile s-c fabrics (Sellers et al.1998).

This case is fully described by van Aswegen and Example Case 2 – Curving extension fractures
Stander2012 and summarized in Figure 4. It is a photograph A special case was recently observed in another VCR
of hangingwall exposed after some 2 m deep failure in a stope in a West Wits mine. The mining configuration is
stope along the VCR ore body in the West Wits mining shown in Figure 5. A panel face was significantly damaged
district, ~3300 m below surface. Steeply dipping primary by a small seismic event. A crush pillar was left at the face
and secondary low-angle fractures, that combine to form position and mining initiated beyond the crush pillar,
hazardous key blocks, are well exposed. The fall was not allowing access to the damaged hangingwall of the stope.
associated with a particular seismic event, although during
the 6 months period up to the time the photograph was A striking feature observed here is the way the
taken, hundreds of very small and several small seismic extension fractures in the immediate hangingwall of the
stope curve the 'wrong' way as it hits an east dipping

Forensic rock mechanics -4-

discontinuity see Figure 6. The curving features are these extension fractures would wrap around the stope face
recognised as extension fractures by the prominent in a bow-wave fashion, i.e. rather curve downwards than
fractographic striae that decorate the surfaces. Normally upwards as in this case. So some anomalous local stress
field must have acted here, at least for a brief period.
The rotation of the stress field in the environs of planar
discontinuities is a known phenomenon and somewhat
variable explanations are found in literature. One example
is the orientation of the maximum principal stress in the
surrounds of the San Andreas fault, being close to normal to
the fault zone. This has been explained by Zoback et al.1987.
Considering only the horizontal stress, they refer to the
two-dimensional Mohr diagram: if σ1 is initially oriented at
an angle β to the fault, then, by reducing the shear stress
parallel to the fault and by keeping the mean horizontal
stress constant, a new orientation α for σ1 is given by
α = π/2 – 1/2 tan-1 {2Co/[(σmax - σhmin) cos( π – 2β)
Figure 5. Mine layout for Example Case 2. Stoping for β > 45º and
width is approximately 1 m. The grey dot is the site
of the damage. The "eye" shows the direction of view α = 1/2 tan-1 {2Co/[(σmax - σhmin) cos(2β)
of the photograph below. for β <45º, where Co is the shear strength of the fault.
A more general mathematical treatment of the stress
orientation around discontinuities is given by Rudnicki1979.
An experimental study by Gomez-Rivas and Griera2009
yielded empirical relations between the angle of incidence
of σ1 relative to a set of planar discontinuities, the strain
induced by the loading and the rotation of the stress field.
None of these works consider the stress rotations during
dynamic loading. It was postulated for Example Case 1
above that the extension fractures described was of
dynamic origin, i.e. the fractures represent a fossil trace of
the principle stress trajectories at the instance of dynamic
loading – the loading being the fast elastic closure
following stope advance during production blasting.
The geometry of the curving extension fractures in
Figure 6 may possibly be explained by the formation of the
fractures at the instant of shear failure along the east
dipping plane of weakness. The rupture leading to the
normal sense of slip indicated by the arrows likely initiated
below reef similar to the case of an Ortlepp shear overshoot
(see reference to ‘overshooting’ of Ortlepp shears below).
Figure 6. Extension fractures with prominent
The σ1 trajectories in the immediate footwall of the yielding
fractographic striae curves up and terminates
against a pre-existing plane of weakness at the site of slip plane turns orthogonal to the slip plane because of the
rockburst damage. The inset shows the sigma1 rotation of stress within the dynamic shear zone as shown
trajectories for the instance of normal slip along the in the inset in Figure 6. Further back (to the left in the
zone of weakness. figure) the fractures merge with the normal stope-induced
stress fractures as these dynamically develop further as a
result of the co-seismic stope closure. The moment tensor
analysis (Figure 7) of the best candidate ml 1.3 seismic
event that occurred at the general time of interest (the exact
date and time of the damage was not recorded) shows a
dominant normal slip mechanism with the one nodal plane
Figure 7. Moment tensor of the likely seismic source parallel to the shear surface shown in Figure 6. Note an
for the damage shown above. implosive component of the source as well, consistent with

5
some dynamic stope closure as shown by the freshly by elongate pillars and regularly spaced rectangular pillars
damaged timber support unit in Figure 6. completed the regional support system.
Example Case 3 - Pillar burst with radiating naked A moment magnitude 2.3 (ml 2.1) was recorded, being
tendons the largest in the surrounds at the time – and the first event
> ml 2.0 since 2004. A remarkable pattern of naked tendons
A particular case of an oversized crush pillar failure in
Rustenburg platinum was described recently (see Figure 12 radiating from a point near the middle of the damaged pillar
of Malovichko et al.2012, reproduced as Figure 4 in
Malovichko and van Aswegen2013). The pillar failed
dynamically (moment magnitude 1.9) with significant
damage in the surrounds. Part of the pillar was ejected
along a plane of weakness in the stope hangingwall and left

Figure 9. Naked tendons show the direction of the east

pillar wall ejection – the radial geometry as indicated in
Figure 8 could not be captured in one photograph.

Figure 8. Mine layout for Example Case 3. The

arrows schematically show the radial pattern of the
naked tendons (Figure 9). A-B is the section line for
Figure 10.

fresh slickensides. Roof and wall anchors were deformed as

the crushed rock was ejected away from the failed pillar. Figure 10. Schematic section along A-B in Figure 8 to
explain the dynamic failure mechanism. The circle
The resultant naked tendons pointed radially away from the
schematically shows the volume of co-seismic finite
seismic source and a range of 21‫‏‬° variation was measured. deformation.
A very similar case of pillar failure with naked tendons
(Figure 8) was noted by Scheepers2007 and I was escorted by
radiating from the seismic source occurred in 2006 in a
Joggie van Oort to the site a couple of days after the
gold mine in the Klerksdorp gold field. The mining layout
incident. In an internal report to the mine by the Rock
is given in Figure 8: The tabular VCR ore body dips to the
Engineering department the damage is described as pillar
west-south-west at approximately 35°. Mining had
wall ejection along a length of approximately 20 m. The
progressed towards the north-east along ~30 m wide panels,
depth of failure (portions of the pillar sidewall that were
each with a gully at the bottom. The panels were separated
ejected) was estimated as 0.7 m on the east side and 1.2 m

Forensic rock mechanics -6-

 ramp faults. The crushing of the ends of wooden support
elements indicated dynamic stope closure of several cm.
The closure was also confirmed by seismological analysis
(see below). The failure mechanism is shown in Figure 10.
An interesting bit of detail was noted below one of the
more shallowly dipping ejection planes. A set of extension
fractures, most likely of co-seismic origin, display
curvature reminiscent of the E3 fractures referred to under
Example Case 1 above – see Figure 11.
The underground observations were very important
since the event occurred outside of the seismic network
operational at the time and the hypocentre location through
routine processing was off by some 50 m in plan and,
Figure 11. Curved extension fractures between because of the planar configuration of sensor sets, by about
two fault surfaces in the style of the ‘E3’ fractures 600 m in 3D. Most of what we know about this event we
described in Example Case 1. learned from these observations. The identification of the
pillar served as a ‘calibration blast’. Using the middle of the
pillar as the source location I found the appropriate P- and
S-wave velocities that would return the correct hypocentre
location and then did a moment tensor inversion. A
reasonably stable result was obtained through the grid
search routine. A significant implosive component was
Figure 12. Output of the moment tensor analysis indicated and a relatively small normal slip deviatoric
for the ml 2.1 pillar burst – example Case 3. component. The nodal planes are not, however, parallel to
the pillar edges as could be expected – off by about 35° –
see Figure 12.
on the west side of the pillar. It is possible, however, that
the depth of failure was symmetric, but that the ejection to The source ‘size’ from standard processing is ~200 m
the west appeared deeper since the physical movement in and the Energy_S/Energy_P ratio is 5. The interpretation of
that directions would have been aided by gravity. the physical meaning of ‘size’ in this case is unclear. An
estimate of the volume of co-seismic finite deformation is
Bedding-plane parallel faults characterize the first shown in Figure 10.
several metres of the lava above the gold reef. Over large
areas the actual contact between the VCR and the overlying Example Case 4 - Bedding plane failure along a tunnel
lava is such a fault, leaving the stopes with smooth Significant rockburst damage to a 44 m long, NE
hangingwalls that often hide unstable domes defined by trending connecting cross cut between two parallel tunnels
ramp faults. The underground investigation revealed that, at a West Wits mine occurred with the blast on 2 March
because of a ramp set juxtaposing the lava with the reef at 2009. A moment magnitude 1.3 event was recorded and
the pillar position, the pillar consisted mainly of the located near the site of damage. The depth was ~3000 m
stronger lava. The design is for the pillars to be smaller and below surface. A plan of the area is shown in Figure 13.
to consist of the weaker VCR so they would crush non-
violently. The particular combination of factors described At site 3, Figure 13 naked tendons show the direction
here, however, allowed anomalous stress to build up in the of sidewall ejection along a weak bedding plane. The
pillar leading to the burst. orientations of such tendons were measured at several
places as indicated by the arrows in Figure 13.
The observable damage mechanism included the
ejection of pillar material into the stope along the set of

7
 Figure 13. Plan view of a connecting cross-cut damaged Figure 14. Naked tendons smeared along the direction of
(shaded area) by a rockburst - Example Case 4. The sidewall ejection along inclined bedding plane. The view
numbers indicate points of interest referred to in the is towards the north-east at site 3 in Figure 13 above.
text. The arrows show the orientations of naked
tendons. slip along the bedding planes (Figures 16 and 17). The
The source size is 48 m which, coincidentally, is the deviation from perfect normal slip (rake of -110°) is
same as the distance of tunnel damage. Note that in this consistent with the direction of slip and ejection indicated
case the implosive component would not be expected to be by the naked tendons.
as high as in the case of the pillar failure described above At site 1 in Figure 13, the tendons were broken, lying
simply because the total volume of potential closure in the on the floor. One particular piece of rock was useful to
tunnel environment would be an order of magnitude less reconstruct the dynamic failure. Faint but distinct
than that in the stoping environment. The event was slickensides are superimposed on prominently radiating
recorded mostly by distant sensor sets not useful for source fractographic striae. The slickensides are parallel to the slip
mechanism analyses at all. One sensor set at a distance of direction indicated by the bent naked tendon still fixed to
760 m was found useful to test various options using the the rock. The primary sedimentary bedding planes, also
“Seismic Mechanism Tester” or “SMT”, an in house IMS visible in the rock, allowed visualization of what its
software tool that allows the user to quickly scan through a position at the sidewall of the tunnel was before the damage
number of potential mechanisms and compare the recorded (Figure 15). Thus one can reconstruct the history of
waveforms with the test mechanism's synthetic waveforms. dynamic failure and deduce that the damage mechanism
(Malovichko and van Aswegen2013). A good fit was found involved compression of the sidewall that forced it to
for a mechanism of some implosion plus oblique-normal

Figure 15. Left: Textbook example of radiating fractographic striae, showing the direction of rupture propagation on a
loose rock slab from site 1, Figure 13. Right: A zoom in on the area demarcated by the rectangular outline in the top
photograph, re-orientated to show the original position along the tunnel sidewall. Faint slickensides are superimposed
on the fractographic striae parallel to the direction of sidewall ejection indicated by the bend in the naked tendon.

Forensic rock mechanics -8-

 seismological analysis we showed that the main source of
the event is pillar failure about 18 m above the tunnel. In
this case the spectacular damage reflects at best a sub-
source and does not shed much light on the main source of
the seismic event. Was it not for the access to the failed
pillar above, one would not have been able to reconcile the
tunnel damage with the seismological record.

Figure16. Best fit moment tensor for the damaging Example Case 6 - VCR face burst, Klerksdorp gold field
event - Example Case 4. The moment tensor solution Rockbursts are naturally violent. Many hundreds of
was established using the 'SMT' method – see text. rockbursts have occurred in the presence of miners, but
virtually no eye witness accounts have been recorded of
what happened in terms of rocks and equipment being
thrown about. A standard task during an underground
investigation of a rockburst site is to find evidence of the
directions and velocities of rock ejection.

Figure 17. Synthetic waveforms fit the recorded ones

fairly well.
fracture and eject towards the north-west – consistent with
the source mechanism referred to above.
Example Case 5 - Bushveld pillar burst 2
Figure 18 is from footwall haulage below the Merensky
Reef at a Rustenburg platinum mine. Mining is at about
1800 m below surface. The figure shows prominent
slickensides along two parallel and shallow-dipping planes Figure 19. General damage and the position of a large
rock ejected from the face at least 2 m far. - Example
of weakness about 2 cm apart in the roof of a tunnel of
Case 6.
which the sidewall was violently ejected. Secondary planes
sub-parallel to the tunnel sidewall (i.e. at high angle to the
shallow-dipping planes) also show slickensides. The
seismic event is described in some detail by Malovichko et
al.2012 . Through physical observation of the failed pillar and

Figure 18. Slip striations on multiple slip surfaces

resulting from local sidewall ejection under the dynamic
load of a 40 m distant pillar burst- Example Case 5. Figure 20. A shard of quarzite about 55 mm in length,
embedded into a timber prop about 1 m from the face
9 that burst.
 One extreme case was described by Ortlepp1993 who he The faceburst was associated with two very small
found a shard of rock embedded in the lining of a events that were located at the faceburst site 4 minutes apart
ventilation pipe in a tunnel damaged by an ml 1.8 event. - see Table 1. Their source mechanisms are given in Figure
Experimenting with an air gun shooting projectiles into a 22.
piece of the lining recovered from underground, he
The burst affected about 7 m of the wide heading with
estimated a velocity of ~50 m/s for the shard.
some roof collapse at the very top (east end) of the heading.
It is remarkable how relatively small seismic events Rock fragments of variable sizes had been ejected from the
can cause significant ejection velocities, as illustrated by face. Two large slabs are of interest - see Figure 23. The
Example Cases 6 and 7. first one, lying on the footwall, measured 850x580x80mm.
Comparing the calcite vein geometry within the slab with
An exceptionally violent faceburst occurred in 2007 in
that on the face it is clear that it had travelled about 1 m
a VCR stope in the Klerksdorp gold field. The development
away from the face from a height of about 600 mm. This
of an ~8 m wide, 70 m long raise (on reef excavation along
would require an ejection velocity of ~3 m/s.
dip) across an 80 m wide un-mined block had just been
completed and full production was about to start. Mining The second slab was caught in a vertical pose while
depth was about 1300 m below surface. The nearest travelling away from the face by the hangingwall closing a
existing stope was 40 m to the south. Stoping width (height)
was about 1.5 m with packs along the centre gully and two
rows of timber profile props as stope support. It was, of
course, still a 'wide raise' and not a stope per se.
An ml 0.9 event was recorded at the time close to the
site of damage. Other information about the event includes
a static stress drop of 1 bar, an E_S/E_P ratio of 5.
Hundreds of such events have occurred at the mine and
generally no damage is recorded. In this case one could
argue that most energy went into ejecting the face and not
so much radiated to the far field sensors – see Figure 19. Figure 21. Mine layout at site of Example Case 7

The burst was characterised by rock fragments of fraction. This slab displayed prominent fractographic
various sizes being ejected from the face in a direction lineations which, at first glance, appear contradictory -
approximately normal to the face. Scour marks and fresh Figure 24. The most prominent lineation set turns through
slickensides along the very smooth hangingwall mapped 180 degrees and runs towards what appears to be a collision
their paths. Minimal dynamic closure was evident from with a second set. Closer inspection (through the study of
some the timber props. Remarkably, apart from those in the
immediate path of the large rock slabs ejected from the
face, all the installed props were still in place, including the
one shown in Figure 20.
One particular piece of quartzite (Figure 19) was
measured to be approximately 0.17 m3 in volume, thus with
a mass of some 450 kg. It was ejected at least 2 m far,
having dropped a maximum 1.4 m. This works out to a
minimum value of the ejection velocity of 3.7 m/s.
A higher ejection velocity from a shard of quartzite is
indicated by the way it penetrated one of the timber props a
metre from the face - Figure 20.
Example Case 7 - Face burst in a dyke, Klerksdorp gold
field.
Another faceburst occurred at the same mine in 2011.
In this case at a depth < 1200 m. To cross a dyke a 15 m
wide excavation (referred to as a 'heading') was advanced
from within a 40 m wide panel (Figure 21). The wide
heading was not preconditioned and a burst occurred in the
heading face. Figure 22. Moment tensors for the two small
events associated with Example Case 7

Forensic rock mechanics - 10 -

 Table 1. Source parameters of two small events
associated with the faceburst: Example Case 7
Event 1 2

Time 162614 163046

Moment Magnitude 0.1 0.0

Seismic Moment 1.55E+009 1.29E+009 Nm

Radiated Energy 1.37E+002 3.20E+002 J

Corner Frequency 54.6 66.8 Hz

Static Stress Drop 2.11E+004 2.77E+004 Pa

Apparent Stress 7.40E+003 7.40E+003 Pa

Event Size 36 31 m

Energy_S / Energy_P 2.8 8.3

stereo-pair photographs) reveals that the two sets of

fractures are not on the same fracture plane. Figure 24. Fractographic striae on the ejected slab (top
slab in Figure 23) which was arrested 'in flight' by stope
We distinguished at least four parallel fracture planes closure.
case. Note that there was no indication of buckling failure
here.
At the site of maximum faceburst damage two
prominent calcite veins approximately 1 m apart strike
parallel to the face and dip shallowly southwards towards
the stope. The one vein is prominent in the face and the
other forms the hangingwall of the stope.
These structures have large strike and dip dimensions
and slip along them during the burst is suggested by the
slickensides (black lines, Figure 23) at high angle to the
face as indicated by the arrows. This may explain the
reverse slip component in the moment tensor solution of
Figure 23. Ejection trajectories of two slabs and shear Event 1 (Figure 22). Such reverse slip is opposite to the
direction along planes of weakness, Example Case 7.
usual stope closure mechanisms for South African mines
where the vertical gravitational loading controls the stress
over a slab thickness of approximate 6 cm, each with
field. This extent of damage for such a small event is also
prominent fractographic lineations. One may speculate that,
anomalous and it is likely explained by local stress rotation
if a significant number of parallel extension fractures all
associated with the strong dyke.
develop dynamically and simultaneously, the associated
dynamic volumetric strain may cause slab ejection as in this

11
 ORTLEPP SHEARS dynamic seismic event.
In this section I revisit those special structures In their type location they are normal faults because the
discovered and studied by Dave Ortlepp, but never properly principle stress is near vertical (Figure 25). Figure 26 is a
named by him. We simply call them “Ortlepp shears”. sketch from a photograph showing a conjugate set of shears
Ortlepp shears are a class of dynamic brittle shear that form ahead of a CLR stope in a West Wits mine. The particular
in deep mines where the stress is high enough to cause shear set was associated with an ml 2.5 event in a high
failure of strong and homogenous, intact rock. A complex stress environment. The black shading represents the zone
of intense fracturing with the prominence of ‘rock flour’ –
see below.
Internal features and the question of temporal evolution
Ortlepp shears consist mainly of two mechanistic
elements, namely en echelon extension fractures and
breakthrough shear ruptures (van Aswegen2008). No other
structural elements are required to describe them. Most of
the shears described by Ortlepp (see for example
Ortlepp1992, Ortlepp1997 and references therein) display
breakthrough shear ruptures (geometrically similar to R-
shears bands in the Riedel nomenclature). Due to the once-
off generation of the complete structure during one seismic
event in hard rocks with little or no micaceous minerals, we
do not see the development of the other classic elements of
brittle-ductile shears such R2 (Riedel), P- and Y-shears.

Figure 25. Reproduction of Figure 2 of Adams et al. 1981 The basic geometry repeats in a left stepping fashion
showing the distribution of fractures around a stope when viewed as a right lateral shear
face. Their ‘Type II’ we refer to as Ortlepp shears
Ortlepp2000 makes an interesting point on the genesis of the
internal structure evolves, for most part, during one

Figure 26. A conjugate set of Ortlepp shears with breakthrough shear rupture, rock flour (dark patches) and 110
mm displacement. The location of this particular shear zone is shown in Figure 30.

Forensic rock mechanics - 12 -

complete fracture event. He discusses two extreme less intuitively interpreted the direction of rupture
postulates namely a) the structure develops in an propagation as he demonstrated in section 3.2.4.3 of his in
evolutionary way “with the localisation and alignment of his monograph (Ortlepp1997), the relevant figure is re-
very small extension fractures (on a scale of a few tens of produced here as Figure 27. This interpretation suggests
millimetres). These, in due course, are suddenly linked up that the Ortlepp shear initiates away from the mine opening
into one or more relatively planar and discrete through- and propagates towards it. A simple argument that validates
going principal slip surfaces”. Alternatively b) ..” the shear his interpretation is as follows: The stress distribution
rupture spreads as a very rapidly extending fracture front around a shallowly dipping stope face in a deep South
through a pristine highly - stressed (but not yet failed) rock African gold mine, where the ambient stress field is largely
space”. After a discussion about the observable features defined by gravitational loading, prohibits normal faulting
favouring the one or the other postulate, he concludes: in the hangingwall of the stope along a plane dipping
towards the stope. Similarly normal faulting cannot initiate
“The main features of previously observed shear
in the footwall behind the face along a plan dipping away
rupture surfaces which reject the above evolutionary
from the face. I ascribed the occurrences of portions of
hypothesis, are:
Ortlepp shears in such environments to a process of
 the remarkable planarity and continuity of the ruptures ‘overshoot’ (van Aswegen2008). Like in the example of
on the large scale of tens of meters Figure 27, the shear initiates some distance away from the
 the complete lack of damage, even on a sub stope, propagates toward the stope, cuts through the reef
microscopic scale, in the rock fabric within centimetres plane just ahead of the stope and comes to rest in the
of the main shear surfaces and within millimetres of environment where the stress field works against the
the subsidiary fractures such as the pinnate joints. normal slip mechanism.
When visualised from a perspective which embraces
The relation between the sense of slip, direction of
the entire burst rupture extending over a few thousand
rupture propagation and the internal structure of the shear
square meters in area, it seems improbable that the
zone is shown in Figure 27. After detailed numerical
pre-conditioning process could be confined to such an
analyses, Fliss et al.2005 concluded that there is no simple
extraordinarily narrow zone only millimetres or
correlation between fault- and branch-fault geometry and
centimetres wide”.
rupture directivity. Ando and Yamashita2007 suggest that a
Direction of rupture propagation fault will branch along the extensional side of the
The direction of rupture propagation was an aspect of propagating fault tip. This would logically lead to right
stepping of a brittle right lateral shear zone, which is clearly
the shears that Dave Ortlepp paid particular attention to,
not the case as seen in Ortlepp shears and many others (e.g.
because of its importance to damage potential. He more or

Figure 27. The internal structure and direction of rupture propagation of the type Ortlepp shear. Redrawn from
Ortlepp1997, page 37, Figure (b).

13
 Logic dictates that the shear, as shown in Figure 27,
would initiate in the region of maximum shear stress,
propagate upwards towards the stope and simultaneously
downwards into the footwall.
Breakthrough shear rupture
Breakthrough shear ruptures accommodate significant
displacement and are the sources of most rock flour. They
are the geometric equivalents of R-shears (Riedel
nomenclature) or ‘Principles Shear Zones’ termed by
Skempton as quoted by Ahlgren1999. In many cases,
however, they are not present at all. For example, during
the investigation of the damage in Example Case 4 above I
noticed a couple of old Ortlepp shears now visible due to
the breaking away of the tunnel sidewall and thus the
removal of the shotcrete and other support elements that
normally hide the rock. Figure 29 shows the most
prominent of these shears, being the 'fault' shown at site 4,

Figure 28. The end of an Ortlepp shear where it was

arrested by the facture zone ahead of a stope face.

Ahlgren1999, Katz et al.2003). The stepping geometry of the

brittle shear relates to the sense of slip (left-lateral or right
lateral) and is independent of the direction of rupture
Figure 29. An older Ortlepp shear exposed by younger
propagation.
damage at site 4, Figure 13, Example Case 4. The shear
Since there is no simple rule to link the direction of is accentuated in the main picture and a zoom-in shows
rupture propagation to internal structure of an Ortlepp the details of its appearance.
shear, the above overshoot argument has to suffice. An
example of such an overshoot occurrence observed in the Figure 13.
reef footwall was described when explaining the origin of None of the Ortlepp shears noted in the area display
the strong linear fabric along the shear zone walls (van breakthrough shear ruptures. The actual displacement is of
Aswegen2008, Figure 2). That particular exposure was the order of millimetres, not visible in the photograph. The
showed no breakthrough shear rupture. bend of the prominent bedding plane in the photograph is
The way an Ortlepp shear ends can also indicate the coincidental and is not related to the shear displacement.
direction of rupture propagation. The likelihood of In rock mechanics tests and earthquake research it is
exposing the beginning or end of the shear is very low common knowledge that it is difficult to build a press that is
during normal mining. When the rupture path gets too close stiff enough not to allow the destruction of a rock sample
to the stope, however, the rupture may be arrested by the once failure has advanced to form a shear from the one end
softness of the fracture zone ahead of the stope. The way of the sample to the other. Special tricks are required to
the shear zone splits up into the existing fractures then stop the loading once failure starts, e.g. by monitoring
allows one to see where the rupture came from. Figure 28 is acoustic emissions (e.g. Lockner et al.1992). Nature,
an example from the Hartebeestfontein mine (see Bosman however, has no problem creating an environment where a
and van Aswegen2003). In this case the rupture ended by brittle shear would initiate and then freeze forever, leaving
diffusing into the stope’s fracture zone – this rupture clearly interesting outcrops for us to wonder about.
propagated towards the stope from the hangingwall.

Forensic rock mechanics - 14 -

 The longwall mine layout of past Wits gold mines panel revealing a prominent Ortlepp shear, part of a
presented relatively soft loading system, likely explaining conjugate set as shown in Figure 26. The sketch in Figure
the fact that most Ortlepp shears, once they got going, were 30 explains the relative locations of the conjugate set and
not easily stopped. Modern mine designs are, of course, less the reverse shears.
soft, explaining the relative scarcity of large Ortlepp shears
The reverse slip shear is explained as a sub-event under
found recently.
the dynamic load from the initial failure as illustrated in the
Compared to a production stoping environment, the sketch. This concept has subsequently been confirmed by
loading system stiffness around the tunnel system at numerical modelling (Murphy2012).
Example Case 4 is very stiff. This explains the lack of
The pulverization shown in Figure 30 yielded a fault
breakthrough shear ruptures along those Ortlepp shears.
fabric with no flux features typical of dynamic failure under
The case from my 2008 paper referred to above was extensional dynamic loading. The rock flour infiltrated
different since there was no breakthrough shear rupture, but large and very small cracks that opened during the event
the mining environment was longwall mining at a large similar to the way ultra-cataclasite and pseudotachylite do.
scale and the general loading system was soft. One
Powder from the shear zone was kindly examined by
explanation for this paradox could be that the shear initiated
Dr. Arno Zang of the GFZ German Research Centre for
relatively far from the face, out in the solid rock. Here, the
Geosciences. In a sample from which the larger grains were
stress had to be high enough to cause failure, but the
sieved out, he found that 10% of the cumulative number of
rupture far enough from the advancing stope not to be
grains was smaller than 4 µm, 50% of grains smaller than
affected by the proximal loading system softness. So: three
14 µm and 90% of the cumulative grain number smaller
important conclusions from one simple observation: a) the
than 25.6 µm. This does not reflect the ultrafine powder
stress far from the face was very high, b) the resultant
reported by Reches and Dewers2005 and Wilson et al.2005
Ortlepp shear crossed the path of the advancing stope a
They analysed material from an Ortlepp shear at the
relatively long distance ahead and, c) the direction of
Hartebeestfontein mine. They found that most of the
rupture propagation was towards the stope.
individual fragments in the rock flour can be as small as
Rock flour 0.3±0.2 µm and they estimated volumetric strain rates of
105/s to produce the material. They used special techniques
One of the characteristics of the rockbursts in the Wits
to ensure disaggregation, possibly explaining the difference
mines is the presence of rock flour in the zones of extreme
in their results compared to that of Arno Zang.
deformation. When the failure zone daylights in a stope the
air is instantly filled with fine white powder. A particular Total displacement
example case is re-visited here with more detail.
One of the most direct (and seemingly simple)
A ‘standard’ face parallel shear slip event caused observations we can make underground is fault
significant damage to a CLR stope in the West Wits gold displacement after a large seismic event. South African
mining area. During underground investigations I found a gold mine examples come to mind since a) we have many
significant reverse slip shear zone filled with rock flour, large faults that occasionally slip dynamically, b) most
explaining a report from the mine overseer that the large faults are “strike” faults in the sense that their strike is
damaged stope was white with rock flour when they entered approximately parallel to the strike of the planar ore body
it after the incident. necessitating tunnels to cross the faults allowing for good
exposures, c) we paint 'grade lines' along the tunnel walls
The rockburst was associated with an ml 2.5 seismic
that serve as markers, d) we have water furrows along the
event of complex nature. The moment tensor analysis based
tunnels that yield 'rapids' after fault displacement and e) we
only on the lowest frequency components of the waveforms
have train tracks along the tunnels that allow easy
confirmed the standard normal faulting mechanism typical
surveying of large scale displacements. One particular
of Ortlepp shears plus an implosive component consistent
example presented many times is that of Art McGarr
with the dynamic stope closure all too evident underground.
caressing the source of a 37 cm displacement earthquake
After the rockburst, up-dip mining allowed a series of along the Brand fault in Welkom in 1989 (Ortlepp 2005).
cross sections to be investigated ahead of the damaged

15
 Figure 30. Example of pulverization during
dynamic transtensional shearing as part of a
complex rockburst. The sketch on the left shows
the position of the main photograph in the stope
hangingwall. The broken line circle schematically
indicates the seismic source region. The zone of
advanced pulverization has been enhanced in the
main photograph. Sub-area 1 shows transtensional
shear fabric reminiscent of the E3 fractures of
Example Case 1. Sub-area 2 shows pulverization
under similar shear load – clearly dynamic. Sub-
area 3 shows pulverization associated with reverse
shear (solid white lines are the associated extension
fractures), the latter superimposed on an internal
fabric indicating older normal shear (extension
fractures indicated by broken lines) along the same
shear zone.

Forensic rock mechanics - 16 -

 Table 2: Measured displacements (u) and umax for fault slip events and Ortlepp shears.
Estimated
Measured
Date Ref Mag Case from Ratio
displacement
moment
1970 1 2.2 Original Ortlepp shear – A – ERPM 100 26 5.6
1970 1 3.4 Original Ortlepp shear – B – ERPM 100 80 1.2
1976 2 5.1 Dagbreek fault, Western Holdings 220 567 0.4
1978 3 3.0 Ortlepp shear – Hartebeestfontein 100 51 2
1982 4,5 4.6 Weselia fault, Free State Geduld 400 319 1.3
1986 5 4.8 Dagbreek fault, St Helena 200 401 0.5
1986 6 4.8 No 2 Shaft fault, Vaal Reefs 200 401 0.5
1987 6 4.9 No 5 Shaft fault, Vaal Reefs 200 450 0.4
1989 5 4.5 Brand fault, Pres. Brand 370 284 1.3
1990 5 4.8 Stuirmanspan fault, Pres. Brand 200 401 0.5
1992 5 4.7 Saaiplaas fault – Saaiplaas 150 357 0.4
1997 3,7 3.7 Ortlepp shear – Hartebeestfontein 370 113 3.3
1999 5 5.1 Dagbreek fault, Matjhabeng 400 567 0.7
2001 5 4.1 No 5 Shaft fault, Vaal Reefs 200 179 1.1
2005 5 2.8 Ortlepp shear – 97W -TauTona 110 30 3.7
2006 5 2.5 Ortlepp shear – Bank area, TauTona 90 20 4.5

1 McGarr et al1979. Note: There is uncertainty about whether shear B was a magnitude 3.4 or 1.7
2. Patchet1977
3. Bosman and van Aswegen2003
4. Lawrence1984
5. Measurement by GvA
6. Brummer and Rorke1990
7. Reches and Dewers2005

In the case of the Ortlepp shears, marker bands are not African Witwatersrand gold mines and because the mining
always at hand to allow displacement to be assessed, but method requires mine workers to be within the stopes,
several measurements have been made as listed below. It is opportunities abound to observe details of rock failure
interesting to compare slip displacements along re-activated phenomena at close quarters.
faults with that along Ortlepp shears. In Table 2 I list
Simple ‘rules’ that help us to understand the most
available (and reasonably reliable) information about
commonly observed mining induced structures can be
dynamic shear slip events in South African mines. The
listed: Rock tends to deform in shear while failure is mostly
measured displacements are compared with the theoretical
in tension. The faster (more dynamic) the deformation, the
maximum displacements estimated from seismic moment
finer the rock fragments produced.
(having converted the magnitudes to moment).
The most commonly observed mining induced
It is clear that the displacements in the case of the
structures are extension fractures. They serve as fossil
Ortlepp shears are anomalous, while those along the large
traces of the 1 trajectories at the very instant of formation,
faults are similar to those of natural earthquakes. This is
yielding insights into the static and dynamic orientations of
interpreted to be as a result of the much softer loading
past stresses. Fractographic striae indicate the direction of
system in the immediate surrounds of stopes (where the
rupture propagation. In cases where such striae are
Ortlepp shears occur) compared to that of the large faults
relatively long and not strongly radiating, care is required to
that are re-activated under mining induced stress. On the
distinguish them from slickensides which show the slip
scale of the large faults, the stopes are small and the overall
direction along slip surfaces. In brittle shear zones the
loading system working on the structures are closer to that
geometry of the internal extension fractures can show
of natural faults.
whether the shear failure was transpressional or
CONCLUSIONS transtensional. The bending and smearing out of support
units like naked tendons are better kinematic indicators of
Because of the hard and brittle nature of the quartzites
dynamic shearing than slickensides.
and conglomerates that form the host rocks of the South

17
 Example cases where the dynamic failure process was Technology roadmap for rock mechanics, SAIMM,
deduced from underground observations allowed Johannesburg, 2003. pp. 143 – 149.
conclusions about:
BRUMMER, R.K. and RORKE, A.J.. Case studies on large
 The origin of particular low angle fractures in West rockbursts in South African gold mines. In: Fairhurst, C
Wits stopes. (ed.). RaSiM2, Minneapolis. AA Balkema. 1990. pp.
 The mechanism of failure at a site of strangely curving 323-329.
extension fractures.
COWAN, D.S. Do faults preserve a record of seismic slip?
 The source location and damage mechanism associated
A field geologist's opinion. Journal of Structural
with a pillar failure.
Geology, vol. 21. 1999. pp. 995-1001.
 The source mechanism of a tunnel failure.
 The difference between the primary source mechanism DOBLAS, M. Slickenside kinematic indicators.
and secondary damage in the near field of a pillar burst. Tectonophysics, vol. 295, 1998. pp.187–197
 The high velocity of face ejection associated with a
DI TORO, G., MITTEMPERGHER, S., FERRIE, F.,
faceburst. MITCHEL, T.M. and PENNACCHIONI, G. The
 The complex source mechanism of a faceburst
contribution of structural geology, experimental rock
associated with a very small seismic event.
deformation and numerical modelling to an improved
 The direction of rupture propagation of an Ortlepp understanding of the seismic cycle: Preface to the
shear.
Special Volume “Physico-chemical processes in
 The origin and mechanism as secondary source of an
seismic faults”. J. Struct. Geol., vol 38, 2012. pp. 3-10.
anomalous reverse shear with thick rock flour.
 The recognition of relatively soft and stiff loading FLISS, S., BHAT, H.S., DMOWSKA, R and RICE, J.R.
systems driving Ortlepp shears. Fault branching and rupture directivity. J. Geophys.
Res., vol. 110, 2005. pp. 1-22.
ACKNOWLEDGEMENTS
LOCKNER, D.A., BYERLEE, J.D., KUKSENKO, V.,
I am grateful for the assistance and collaboration I
PONOMAREV, A., and SIDORIN, A. Observations of
enjoyed over a decade of underground investigations by the
quasi-static fault growth from acoustic emissions. In:
mine employees- and consultants providing rock mechanics Evans, B., Wong, T.-F. (Eds.), Fault Mechanics and
services to the mines of AngloGold Ashanti Limited,
Transport Properties of Rocks. Academic Press, San
Harmony Gold Mining Company Limited, Anglo Platinum
Diego, 1992. pp. 3 – 31.
Limited and (recently) Village Main Reef Limited.
MURPHY, S.K. Linear elastic numerical modelling for
Thanks to Gerhard Hofmann’s comments, significant
failure prediction – an assessment. The Journal of The
improvements were made to the manuscript.
Southern African Institute of Mining and Metallurgy.
REFERENCES Vol. 112. 2012. pp. 737-748

ADAMS, G.R., JAGER, A.J., and ROERING, C. ORTLEPP, W.D. Note on fault-slip motion inferred from a
Investigations of rock fractures around deep level gold study of micro-cataclastic particles from an
mine stopes. In H.S. Einstein (comp.), Rock Mechanics underground shear rupture. Pageoph., 139, No. 3/4,
from Research to Application. Proceedings of the 22nd 1992. pp. 677-695.
U.S. Symposium on Rock Mechanics, Cambridge, MA,
ORTLEPP, W.D. High ground displacement velocities
29 June-2 July, 1981. Cambridge: MIT, pp. 227-236. associated with rockburst damage. RaSiM3, Kingston,
AHLGREN, S.G. The nucleation and evolution of Riedel Ontario, 1993. pp. 101-106.
shear zones as deformation bands in porous sandstone.
ORTLEPP, W.D. Rock fracture and rockbursts: an
A prepublication manuscript submitted to the illustrative study, Monograph series M9, S. Afr. Inst. of
Department of Geosciences in partial fulfillment of the
Min. Metall., Johannesburg, 1997. 98p.
requirements for the degree of Master of Science,
Graduate College, University of Arizona, 1999. ORTLEPP, W.D. A study of rockburst source mechanism.
Safety in Mines Research Advisory Committee, Final
BAHAT, D., RABINOVITCH, A and FRID, V. Tensile
Project Report, GAP 524, 2000
Fracturing in Rocks, Tectonofractographic and
Electromagnetic Radiation Methods. Springer Verlag, ORTLEPP W. D., MURPHY S. and VAN ASWEGEN G.
2005. 570p. The mechanism of a rockburst - an informative case
study, Proc. 2nd Int. Seminar on Deep & High Stress
BOSMAN, J.D. and VAN ASWEGEN, G. Dynamic failure
Mining, SA Inst. Min. Metall. Symposium Series, S37,
of pillar remnants – case studies. ISRM 2003– 2004. pp. 23-40.

Forensic rock mechanics - 18 -

ORTLEPP, W.D. RaSiM comes of age – a review of the static loading. J. Geophys. Res., vol. 111, 2006.
contribution to the understanding and control of B04406, doi:10.1029/2005JB003948.
rockburts. RaSiM6, 2005. pp. 3-20
SCHEEPERS. L.J. Dynamic failure of an oversized crush
GOMEZ-RIVAS, E. AND GRIERA, A. Influence of pillar. Unpublished presentation at the ISS
mechanical anisotropy on shear fracture development. International Seminar, Stellenbosch. 2007
Trabajos de Geología, Universidad de Oviedo, vol. 29,
SELLERS, E.J., BERLENBACH, J. and SCHWEITZER, J.
2009. pp. 305-311.
Fracturing around deep level stopes: Comparison of
HAN, R., SHIMAMOTO , T., ANDO, J-I AND REE, J-H.. numerical simulation with underground observations.
Seismic slip record in carbonate-bearing fault zones: In ROSSMANITH (Ed.) Mechanics of Jointed and
An insight from high-velocity friction experiments on Faulted Rock, Balkema, Rotterdam, 1998. pp. 425-430
siderite gouge. Geology, vol. 35, 2007, pp. 1131-1134.
SMITH, S.A.F., DI TORO, G., KIM, S. REE, J.-H.,
KATZ, Y. WEINBERGER, R AND AYDIN. A. Geometry NIELSEN, S., BILLI, A. and SPIESS, R. Coseismic
and kinematic evolution of Riedel shear structures, recrystallization during shallow earthquake slip.
Capitol Reef National Park, Utah. J. Struct. Geol., vol . Geology vol. 41, 2013, pp. 63-66.
26, 2004, pp. 491-501.
VAN ASWEGEN, G. Ortlepp Shears – dynamic brittle
LAWRENCE, D. A. Seismicity in the Orange Free State shears of South African gold mines. In: Potvin, Y (Ed.)
gold mining district. In: Gay, N.C. and Wainwright, 1st Southern Hemisphere International Rock
E.H. (eds.). RaSiM1, SAIMM Symposium Series No. 6, Mechanics Symposium, 16-19 September, Perth,
1984, pp. 121-130 Western Australia. 2008. pp. 111 – 120.
MALOVICHKO, D., VAN ASWEGEN, G. and CLARK, VAN ASWEGEN, G. and STANDER, M. Origins of some
R. Mechanisms of large seismic events in platinum fractures around tabular stopes in deep South African
mines of the Bushveld Complex (South Africa). The mines. The Journal of The Southern African Institute of
Journal of The Southern African Institute of Mining Mining and Metallurgy, vol.112, 2012. pp. 729-735.
and Metallurgy, vol. 112, 2012. pp. 419-429.
WILSON, B., DEWERS, T.A., RECHES, Z. and BRUNE,
MALOVICHKO, D. and VAN ASWEGEN, G. Testing of J. Particle size and energetics of gouge from earthquake
the source processes of mine related seismic events. rupture zones. Nature, vol. 434, 2005. pp.749-752.
RaSiM8. 2013.
ZOBACK, M.D., ZOBACK, M.L., MOUNT, V.S., SUPPE,
MCGARR, A., POLLARD, D. and ORTLEPP, W.D. J., EATON, J.P., HEALY, J.H., OPPENHEIMER, D.,
Observations and analysis of structures in exhumed RAESENBERG, P., JONES, L., RALEIGH, C.B.,
mine-induced faults. Proceedings of Conference VIII - WOMG, I.G., SCOTTI, O and WENTWORTH. C.
Analysis of Actual Fault Zones in Bedrock, U.S. New evidence on the state of stress of the San Andreas
Geological Survey Open-File Report 79-1239. Version fault system. Science, New Series. vol. 238, 1987. pp.
1.2, 1979. 1105-1111.
PATCHET, S.J. Notes in unpublished report on the
Welkom earthquake. Anglo American Corporation.
1977.
RAMSAY, J.G.. Folding and Fracturing of Rocks.
McGraw-Hill, 1967. 568p.
RECHES, Z. and DEWERS, T.A. Gouge Formation by
Dynamic Pulverization During Earthquake Rupture.
Earth and Planetary Science Letters, 235, 2005. pp.
361-374.
RYDER, J.A. and JAGER, A.J. A textbook on rock
mechanics for tabular hard rock mines. 2002
SIMRAC, Johannesburg.
RUDNICKI, J.W. Rotation of principal stress axes caused
by faulting. Geophys. Res. Lett., vol. 6, 1979. pp 135-
138.
SAGY, A., COHEN G., Z. RECHES, and J. FINEBERG.
Dynamic fracture of granular material under quasi-

19

View publication stats