Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 191 (2017) 631 – 637

Symposium of the International Society for Rock Mechanics

A Review of Uncontrolled Pillar Failures

O. Vardar*, F. Tahmasebinia, C. Zhang, I. Canbulat, S. Saydam
School of Mining Engineering, University of New South Wales, Sydney NSW 2052, Australia

Abstract

Coal burst is a sudden and violent rock/coal failure that occurs in underground coal mines. It is considered to be a highly
catastrophic phenomenon which can cause significant damage to mine workings and equipment as well as result in multiple
fatalities. Throughout the history of underground pillar design, the relation between the post-peak behavior of pillars and stiffness
of the surrounding strata has been extensively studied. These two concepts play an important role in determining the failure mode
of the coal pillars and the amount of potential energy that can be converted to kinetic energy, which is the cause of coal burst. In
this paper, the post-peak behavior of pillars and surrounding strata stiffness are reviewed and the criterion developed to
investigate the instability of the pillar failures is explained. It is concluded that, as the pillar width to mining height (w/h) ratio
increases, its post-peak modulus increases; and a pillar exhibits different failure modes for various w/h ratios.
©
© 2017
2017TheTheAuthors. Published
Authors. by Elsevier
Published Ltd. Ltd.
by Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of EUROCK 2017.
Peer-review under responsibility of the organizing committee of EUROCK 2017
Keywords: uncontrolled pillar failure; pillar burst; coal burst

1. Introduction

Experiences in both hard rock and coal mines reveal that pillar failures are still one of the major hazards in
underground mining and forecasting the mode of the pillar failure can be vital. Therefore, it is essential to fully
understand the mechanics behind these failures in order to predict and control them. As suggested by Tincelin and
Sinou [1], pillar failures can be classified into two categories:
i. Slow, progressive deterioration of the pillars that causes relatively delayed surface subsidence and even damage
if the pillars fail,
ii. Sudden, violent collapse of pillar causing immediate surface damage and mostly associated with fatal accidents.

* Corresponding author. Tel.: +61-432-443-844.

E-mail address: o.vardar@unsw.edu.au

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of EUROCK 2017
doi:10.1016/j.proeng.2017.05.227
632 O. Vardar et al. / Procedia Engineering 191 (2017) 631 – 637

The first type of pillar failures is also named as controlled pillar failures which occur gradually and typically over
long periods of time (i.e. pillar spalling). These pillar failures are also termed as creep and squeeze [2]. Uncontrolled
pillar failures, on the other hand, take place in a sudden and violent manner and fall into the second group of pillar
failures as described above. Since the uncontrolled pillar failures occur rapidly and may not be preceded by any
deterioration of the pillars, they cause significant health and safety risks (e.g. coal burst, entrapment,
windblast etc.) [2].
Of note is that not all the sudden pillar failures are uncontrolled failures. Particularly in bord and pillar mines,
sometimes, controlled and time dependent failures may develop and can reach a level that large scale multiple pillar
failure may take place. These are named as ‘massive pillar collapses’ or ‘cascading pillar failures’ and may include
both controlled and uncontrolled collapses. Such massive pillar collapses are usually associated with undersized
pillars with low residual strength values [2–5]. These failures in this mode are usually observable due to high levels
of deformations.
It is, therefore, important to gain an improved understanding of mechanics contributing to the uncontrolled pillar
failures in order to sustain safe and productive mining practices. In this paper, the two main factors that govern
the failure mode of a pillar: post-peak pillar behavior and surrounding strata stiffness, are reviewed and their
interactions are discussed. It is also of note that, the principles covered in this paper are applicable for any mining
system that involves implementation of coal pillars, whether it is a longwall mine or a bord-and-pillar mine.

2. Effect of pillar geometry on post-peak behavior

The performance of a pillar is controlled by a number of factors, including the inherent strength of coal, cleating,
fracturing, internal friction angle, cohesion, surrounding strata conditions, pillar geometry and roof/pillar/floor
contact conditions [6]. Additionally, due to the variation in geological and geotechnical conditions as well as mining
methods, the size and shape of the pillars are modified accordingly and can show drastic differences. For instance,
rectangular pillars normally have somewhat higher strength due to the increased confinement; length–to–width ratio
of 2, 3 and 4 can result in the increase of the pillar strength by 1.1, 1.2 and 1.3 respectively [7].
The variety in the pillar geometries, mainly in w/h ratios, may also change the mode of pillar failure. According
to Salamon [8], while pillars with intermediate w/h ratios (3 < w/h < 5–7) can fail suddenly and violently as a whole,
wide pillars (w/h > 7) would only suffer from side failures and pressure bursts but complete pillar failure would not
occur. This agrees with the experiences in Germany where coal bursts (possibly rib bursts) are observed only in
the pillars with w/h ratios between 8 to 20 [9]. However, as it was reported by Gates et al. [10] that in the tragic
Crandall Canyon Disaster (2007), pillars with w/h ratios up to 8 had completely failed within a few seconds over
a distance of 800 m. Therefore, it is reasonable to assume that the sudden failures of coal pillars are usually
determined by the loading mechanisms; if the pillars are subjected to high levels of loads, sudden failures are
possible even with large w/h ratios.
Mark [4] has also classified the pillar failure modes based on their w/h ratios. In his paper, three categories of
pillars are noted: slender pillars (w/h < 3); intermediate pillars (4 < w/h < 8) and squat pillars (w/h > 10). Mark
stated that slender pillars can experience sudden collapses due to their low residual strength values while
intermediate pillars are commonly squeezed slowly by the surrounding strata; squat pillars, however, can carry
extreme loads and are associated with pressure bursts and rib failures. Mark’s categorization partially complies with
observations of Gates et al. [10].
Researchers have conducted laboratory [11, 12] and in-situ tests [13–20] to investigate the effect of w/h ratio on
the pillar strength and post-peak behaviour. Das [11] carried out several laboratory tests to obtain full stress-strain
curves of specimens within a wide range of w/h ratios (0.5–13.5). He showed that peak and residual strength of
the material as well as its post-peak modulus varies as the w/h ratio changes. He also reported that at low w/h ratios,
specimens show strain-softening behaviour hence the stress-strain curve slopes downward in the post-peak region.
Furthermore, around the w/h ratio of 8, specimens exhibit elastic-plastic behaviour and their modulus converges to
zero. If the w/h ratio continues to increase, they show strain-hardening behaviour and gain a post-peak modulus
greater than zero. However, many cases of pillar failures have shown that there is no correlation between
the laboratory tested UCS and the actual pillar strength [4].
Empirical formulas have also been developed to calculate the post-peak stiffness of the full-scale pillars by
back-analysing the in-situ pillar tests:
 O. Vardar et al. / Procedia Engineering 191 (2017) 631 – 637 633

௪ ି଴Ǥଽଷଵ
ߣ ൌ ͳǤͺ͸ͳ ቀ ቁ GPa (after Van Heerden [17]), (1)
௛

௪ ିଵǤଵହଽ
ߣ ൌ ͳǤͲ͵͵ ቀ ቁ GPa (after Wagner [13]), (2)
௛

଴Ǥହ଺ଶ௪
‫ܧ‬௣ ൌ െ ʹǤʹͻ͵ GPa (after Van der Merwe and Madden [21], from Van Heerden’s data), (3)
௛

ா೛ ହ
െ ൌ ͲǤʹ͵ ቈ ೢ െ ͳ቉ (after Ryder and Ozbay [7]), (4)
ா ቀ೓ቁ

where ɉcorresponds to the post-peak stiffness; ୮ is the post-peak modulus;  is the Young’s Modulus; w and
h are the pillar width and height, respectively.
The main difference between the in-situ and laboratory tests is that, due to the scale effect, the post-peak modulus
of laboratory specimens may become positive when the w/h ratio is approximately 8 or 9, whereas the post-peak
modulus of laboratory size specimens may become positive at w/h ratio of about 4 to 5. These studies also revealed
that the elastic modulus is independent from the pillar geometry while the post-peak modulus is not an inherent
material property but is highly associated with the w/h ratio, hence a structural property [13, 17].
Nevertheless, it can be concluded from those investigations that the post-peak modulus of the pillars increases as
the w/h ratio becomes larger. If the w/h ratio is further increased, the pillars indicate strain hardening behaviour by
taking excessive load, acting as an abutment. It is because of the factor that the higher confinement provided to
the pillar core as the pillar size becomes larger [13, 22]. Due to the increased confinement, the pillar can be
compressed by extreme loads exceeding its maximum inherent strength. However, it is still not fully known that if
the strength of a pillar continuously increases as the w/h ratio becomes larger, or if there is a maximum limiting
strength value for the pillars with large w/h ratios [23]. Under common mining dimensions it is difficult to test this
failure mode as the pillar failures can be arrested by the spalling that build up in roadways.

3. Surrounding strata stiffness

Although the geometry of a pillar can be designed to provide a higher deformation resistance, the load acting on
the pillar is also crucial in terms of determining the stability of the pillar. Therefore, both concepts (post-peak pillar
behaviour and loading strata stiffness) must be examined in detail and their interaction must be well understood in
order to avoid sudden and violent pillar failures.
The mechanisms behind the uncontrolled pillar failures can be explained by considering wide horizontal panel
consists of a number of pillars. As the number of pillars in the panel increases in an incremental manner, the load
acting on the pillars in the centre also increases progressively until it reaches the tributary area load, in which all
pillar carries the proportional load up to the surface. As this process continues, the panel gets wider the surrounding
strata stiffness slowly converges to zero and the strata behaves as a soft loading machine [24, 25]. Once the local
stiffness of the strata applying load on those critical pillars becomes less than the post-peak stiffness of the pillars,
the deforming pillars undergo an unstable and sudden failure [2, 5, 26–28].
The local mine stiffness (loading system) characteristics are governed by many parameters, such as elastic
modulus, thickness and span (or extraction ratio). However, it is not always possible to obtain direct measurements
of the local mine stiffness due to the complexities in the rock mass and the fact that it is more of a mathematical
entity [5]. Therefore, analytical and numerical models have been employed to analyse the stiffness of the mine
strata [5, 24, 27, 29–38].
Ozbay [24] conducted numerical modelling to provide design guidelines for yield pillars in shallow and
moderated depths. He also studied both the local and the critical strata stiffness concepts in his research. He reported
that as the number of the pillars increase (regardless the extraction ratio or span–to–depth ratio), the stiffness of
the surrounding strata decreases. Moreover, he showed that for the extraction ratios of 0.75 and more, the strata
stiffness reaches its maximum value when the depth is approximately four times of the span; and it reaches its
minimum when the depth is approximately five times smaller than the span. These observations imply that,
634 O. Vardar et al. / Procedia Engineering 191 (2017) 631 – 637

the number of pillars in a panel determines the system stiffness; as the panel gets wider, the system becomes softer
hence the possibility of a violent failure increases [21, 25].
In practice, most of the time, the failure of the overburden layers is a pre-requisite for a pillar collapse [21]. In
such situations the loading strata stiffness becomes zero, as self-weighting of the overburden strata is applied on
the pillars. Furthermore, the discontinuities within the overburden such as faults and dykes cause a reduction in
the system stiffness and consequently an increase in the probability of a violent pillar burst occurrence. However,
the effects of these major geological discontinues on the surrounding strata stiffness still have not been quantified.

4. Uncontrolled pillar failures

As mentioned above, the mode of pillar failures (whether controlled or uncontrolled) depends on the behaviour of
the surrounding strata and the pillars. In other words, the post-peak stiffness of the pillars and the surrounding strata
stiffness govern the difference between uncontrolled and controlled failures [21]. As the pillars shed the load, when
they reach their peak capacity, the rejected load has to be transferred on to the unmined areas nearby. However, if
the surrounding rock strata acts as a soft loading system (see Figure 2) and continues to apply load on the pillar,
uncontrolled pillar failure is likely to occur [25, 27].
An interesting discussion on dynamic pillar failures has been presented by Salamon [8]. He postulated that, up to
a critical span, two zones can be defined within pillars: yielding zone and elastic core. Furthermore, at the critical
span, the state (two zones) transforms into four zones: slumped, crushed and yield zones and the elastic core. He
postulated that once the pillar sides are de-stressed, the transition from two zones to four zones takes place. This
transition is expected to be sudden and violent resulting in a coal burst which may not be avoided by the use of
support.
In addition, Salamon [8]noted that the failure progresses from the sides inwards to the core. Similar to Salamon,
Wagner [13] also reported that, during the in-situ tests, the circumferential portions of the pillars were highly
stressed at the pre-failure stage, yet the high stress zones mitigated towards the inner parts (into the centre/core of
the pillar) after failure of the outer portion.
Salamon [27] developed a pillar failure stability criterion based on the analogy between pillar systems and
laboratory compression tests. According to Salamon’s criterion to avoid unstable pillar failures, the absolute value
of the local loading strata stiffness must be higher than the absolute value of the post-peak stiffness of the pillar:

ȁߣ௦௧௥௔௧௔ ȁ ൐  ȁߣ௣௜௟௟௔௥ ȁ (5)

However, Salamon’s criterion can only predict the mode of the failure but not the rapidity or the intensity of
the process. Moreover, since the loading system is assumed to be an elastic continuum, the presence of major
discontinuities within the strata and the contact conditions between the roof and the pillar are also ignored in
the criterion.
In the situations where overburden layers fail and sit on a pillar as dead weight, the loading system stiffness
reduces down to zero. Hence, in order to prevent a violent uncontrolled failure to occur, the pillar post-peak
modulus must be at least equal or greater than zero. For instance, if this condition is applied on the Equation 3, it can
be reported that pillars with a w/h ratio of 4.08 or larger shouldn’t suffer from a sudden and violent failure due to
dead weight loading [21]. This number also complies with Ryder and Ozbay [7]’s finding that that pillars with a w/h
ratio of 5 or larger would only fail in a controlled manner.
In addition, Zipf [5] proposed a relationship between w/h ratio and post-peak modulus of the pillars based on
field test data [13, 14, 17, 39]. He concluded that the post-peak modulus of the pillars monotonically increases and
becomes greater than zero at a w/h ratio of 4.
In a similar manner, as part of this study another empirical formula is derived from the previous in-situ pillar tests
(Figure 1) [13, 17, 40]:

‫ܧ‬௣ ൌ ͳǤͳͺ͸ͻ݈݊ቀ‫ݓ‬ൗ݄ቁ െ ͳ.6845 GPa (6)

where the post-peak modulus is denoted as ‫ܧ‬௣ ; and w and h are the pillar width and height, respectively.
 O. Vardar et al. / Procedia Engineering 191 (2017) 631 – 637 635

Fig.1. Relationship between the post-peak modulus and w/h ratio of pillars based on field tests.

As evident in the Figure 1, the post-peak modulus becomes zero at w/h ratio of 4.13. Therefore, it can be
concluded that beyond this value, an uncontrolled pillar failure is not expected [27].
Furthermore, it is stated that when the pillars are deeply fractured due to abutment stresses and already have
yielded and reached their residual strength, the ultimate stability will be sustained as the pillar post-peak pillar
stiffness (ߣ) becomes a positive value due to the further compression of the pillars resulting in an increase to
the load. In other words, when a pillar in a crushed state will never have a negative ߣ value, and based on this
argument, it is not possible for a sudden and violent uncontrolled failure to take place [41].
The energy storage and release associated with uncontrolled pillar failures is also of great significance. During
an unstable pillar failure, the strain energy stored in the loading system is suddenly transferred to the pillar. In
Figure 2, the highlighted area is considered as the available ‘excess energy’ at a given strain of the coal/rock. It is
postulated that the unstable failure intensity is governed by the angle ፽; if the angle ፽ increases, the intensity of
the unstable failure also increases [26].
The relation between the amount of the total available energy and released energy has been investigated by many
researchers [34, 42–47]. It is noted that if the far-field stresses kept constant and the rock mass is unsupported,
nearly up to half of the energy generated by the source can be stored as elastic energy while the rest must be
dissipated in fracturing process. However, it is beyond the scope of this paper to review the energy considerations
regarding the coal burst occurrences.

Fig. 2. Effect of loading system stiffness on pillar failure mode (after Gu [48]).
636 O. Vardar et al. / Procedia Engineering 191 (2017) 631 – 637

5. Conclusions and recommendations

In this paper, the two controlling factors of the pillar failure mode, post-peak stiffness of the pillars and
the surrounding strata stiffness, have been reviewed. Additionally, the mechanisms regarding their interaction have
been described.
As it has been suggested in the literature review, the failure mode (controlled or uncontrolled) of a pillar is
strongly associated with two main factors: post-peak behavior of pillars and surrounding strata stiffness. The w/h
ratio is the main controllable parameter that governs the post-peak behavior of the pillars while the strata stiffness
may vary due to many reasons: mine layout, major geological structures, extraction ratio etc.
Both laboratory and field scale experiments have shown that as the w/h ratio increases, the absolute value of
the slope of the pillar (or the test specimen) stress-strain curve decreases. Moreover, in some cases where this ratio
is sufficiently high, the pillar may show strain hardening behavior and take extreme loads due to the high
confinement applied to the elastic core of the pillar. However, in the event of an uncontrolled pillar failure, where
the surrounding strata behaves as a soft loading system, the intact and elastic core of the pillar also bursts, which
results in the loss of whole pillar integrity. This interaction between the stiffness of the pillar in the post-peak regime
and surrounding strata can be explained by the stiffness criterion developed by Salamon [27].
It should be noted that the analytical explanations of the mechanism involved tend to be conservative. Therefore,
quantification of the concepts like local mine stiffness and pillar post-peak stiffness is crucial in order to conduct
sufficient mine designs. Within this scope, numerical models enhanced with well-quantified geotechnical parameters
can provide useful insights. These numerical models can also be utilized to assess other controlling parameters of
pillar stability such as contact conditions, floor behavior, effect of horizontal discontinuities, water and/or gas
content in the rock mass. Furthermore, sophisticated numerical modeling applications based on the energy
considerations (energy storage and release mechanisms) and back-analysis of the case histories can quantify
the likelihood of such dynamic pillar failures. Laboratory experiments can also provide reliable initial inputs for
the numerical models to extrapolate the results to the field scale and to increase the confidence level of the analyses.
Therefore, once better understanding of the sudden and violent uncontrolled pillar failures is gained, different
geological and mining conditions can be modeled to evaluate such risks, and strategies to prevent these incidents can
be developed.

References

[1] E. Tincelin, P. Sinou. Collapse of areas worked by the small pillar method, Practical conclusions and an attempt to formulate laws for the
phenomena observed, Third International Conference on Strata Control, Paris, 1960.
[2] J. Galvin, Ground engineering-principles and practices for underground coal mining, Springer, 2016.
[3] R.K. Zipf Jr, C. Mark, Design methods to control violent pillar failures in room-and-pillar mines, Transactions of the Institution of Mining
and Metallurgy-Section A-Mining Industry, 106 (1997) A124-A132.
[4] C. Mark, State-of-the-art in coal pillar design, Transactions-Society for Mining Metallurgy and Exploration Incorporated, 308 (2000)
123–128.
[5] R.K.J. Zipf, Using a post-failure stability criterion in pillar design, Proceedings of the Second International Workshop on Coal Pillar
Mechanics and Design, NIOSH IC 9448, 1999.
[6] O. Gaede, C. Schrank, I. Canbulat, A. Karrech. A strain-based failure criterion for pillar stability analysis, Proceedings AusRock 2014: Third
Australasian Ground Control in Mining Conference, The Australasian Institute of Mining and Metallurgy, 2014
[7] J. Ryder, M. Ozbay, A methodology for designing pillar layouts for shallow mining, in: ISRM International Symposium, International Society
for Rock Mechanics, 1990.
[8] M. Salamon, Modes of pillar and ribside failure – development and longwall, Strata control for coal mine design – Advanced workshop,
Sydney, School of Mines, University of New South Wales, 1995.
[9] G. Brauner, Rockbursts in coal mines and their prevention, AA Balkema, 1994.
[10] R. Gates, M. Gauna, T. Morley, J. O’Donnell Jr, G. Smith, T. Watkins, C. Weaver, and J. Zelanko, Report of investigation—fatal
underground coal burst accidents, august 6 and 16, 2007, crandall canyon mine, genwal resources inc, huntington, emery county, utah,
Department of Labor, Mine Safety and Health Administration, ID. 472 (2008).
[11] M. Das, Influence of width/height ratio on post-failure behaviour of coal, International Journal of Mining and Geological Engineering,
4 (1986) 79–87.
[12] R.W. Seedsman, P. Hornby, Controlled and uncontrolled pillar collapse, Department of Primary Industries and Energy, 1991.
 O. Vardar et al. / Procedia Engineering 191 (2017) 631 – 637 637

[13] H. Wagner, Determination of the complete load-deformation characteristics of coal pillars, Proceedings of the Third International Congress
on Rock Mechanics, National Academy of Sciences, 1974.
[14] Z. Bieniawski and U. Vogler, Load-deformation behaviour of coal after failure. National Mechanical Engineering Research Institute,
Council for Scientific and Industrial Research, 1969.
[15] Z. Bieniawski, The effect of specimen size on compressive strength of coal. in International Journal of Rock Mechanics and Mining
Sciences & Geomechanics Abstracts, 1968.
[16] Z.T. Bieniawski, Estimating the strength of rock materials, National Mechanical Engineering Research Institute, Council for Scientific and
Industrial Research, 1974.
[17] W. Van Heerden, In situ complete stress–strain characteristics of large coal specimens, Journal of the South African Institute of Mining and
Metallurgy 75 (1975) 207–217.
[18] H. Maleki, In situ pillar strength and failure mechanisms for us coal seams, Proceedings of the workshop on coal pillar mechanics and
design, Pittsburgh, PA: US Department of the Interior, Bureau of Mines, IC, 1992.
[19] A. Starfield, W. Wawersik, Pillars as structural components in room-and-pillar mine design, The 10th US Symposium on Rock Mechanics
(USRMS), American Rock Mechanics Association, 1968.
[20] J.A. Hudson, E.T. Brown, C. Fairhurst, Shape of the complete stress-strain curve for rock, Stability of Rock Slopes, ASCE, 1972.
[21] J. Van der Merwe, B. Madden, Rock engineering for underground coal mining, The South African Institut of Mining and Metallurgy.
Special Publication Series 7 (2002) 233.
[22] C.T. Holland, Cause and occurrence of coal mine bumps, Min Eng. (1958) 994–1004.
[23] C. Mark, A. Iannacchione, Coal pillar mechanics: Theoretical models and field measurements compared, Proceedings of the Workshop on
Coal Pillar Mechanics and Design. Pittsburgh, PA: US Department of the Interior, Bureau of Mines, IC, 1992.
[24] M. Ozbay, The stability and design of yield pillars located at shallow and moderate depths, Journal of the South African Institute of Mining
and Metallurgy 89 (1989) 73–79.
[25] M. Salamon, K. Oravecz, Rock mechanics in coal mining, Chamber of Mines of South Africa, 1976.
[26] N. Cook, A note on rockbursts considered as a problem of stability, J South Afr Inst Min Metall. 65 (1965) 437–446.
[27] M. Salamon, Stability, instability and design of pillar workings, International Journal of Rock Mechanics and Mining Sciences
& Geomechanics Abstracts, 1970.
[28] M. Salamon, H. Wagner, Role of stabilizing pillars in the alleviation of rock burst hazard in deep mines, in: 4th ISRM Congress,
International Society for Rock Mechanics, 1979.
[29] R.K. Zipf, Analysis of stable and unstable pillar failure using a local mine stiffness method, 1992.
[30] K. Heasley, A new laminated overburden model for coal mine design, Proceedings: New Technology for Ground Control in Retreat Mining
97–133 (1997) 60–73.
[31] K.A. Heasley, Numerical modeling of coal mines with a laminated displacement-discontinuity code, Dept. of Mining Engineering, Colorado
School of Mines, 1998.
[32] R.K. Zipf, Mulsim/nl application and practitioner's manual, US Department of the Interior, Bureau of Mines, 1992.
[33] R.K. Zipf, Mulsim/nl theoretical and programmer's manual, US Department of the Interior, Bureau of Mines, 1992.
[34] M. Salamon, Energy considerations in rock mechanics: Fundamental results, Journal of the South African Institute of Mining and Metallurgy
84 (1984) 233–246.
[35] M.D. Salamon, Some applications of the frictionless laminated model, The 30th US Symposium on Rock Mechanics (USRMS), 1989.
[36] M.D. Salamon, Subsidence prediction using a laminated linear model, The 30th US Symposium on Rock Mechanics (USRMS), 1989.
[37] A. Starfield, C. Fairhurst, How high-speed computers advance design of practical mine pillar systems, Engng Min. J. 169 (1968) 78–84.
[38] M.U. Ozbay, Design considerations for mining of hardrock tabular deposits situated at moderate depths, publisher not identified, 1987.
[39] W. Skelly, J. Wolgamott, and F.D. Wang, Coal mine pillar strength and deformation prediction through laboratory sample testing, in: The
18th US Symposium on Rock Mechanics (USRMS), 1977.
[40] Z.T. Bieniawski, In situ large scale tests on square coal specimens measuring two meters in width, CSIR: Pretoria, South Africa, 1968.
[41] M. Ozbay, M. Roberts, Yield pillars in stope support, in: Rock Mechanics in Africa, SANGORM Congress, 1988.
[42] N. Cook, The basic mechanics of rockbursts, Journal of the South African Institute of Mining and Metallurgy 63 (1963) 71–81.
[43] N. Cook, Seismicity associated with mining, Engineering Geology 10 (1976) 99–122.
[44] J. Whyatt, Dynamic failure in deep coal: Recent trends and a path forward Proceedings 27th International Conference on Ground Control in
Mining, 2008.
[45] N. Cook, Origin of rockbursts, Proc. Rockburst Prediction and Control, IMM, London, (1983) 1–9.
[46] M. Salamon, Rock mechanics of underground excavations, Advances in rock mechanics, Proc. 3rd Cong. ISRM, Denver B. 1 (1974)
951–1009.
[47] X. Zhang, G. Feng, L. Kang, S. Yang, Method to determine burst tendency of coal rock by residual energy emission speed, Journal of China
Coal Society. 34 (2009) 1165–1168.
[48] R. Gu, Distinct element model analyses of unstable failures in underground coal mines. 2013, Ph. D. Thesis, Colorado School of Mines.