Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 191 (2017) 646 – 655

Symposium of the International Society for Rock Mechanics

A Parametric Study Using Numerical Modelling to Assess

the Stability of Marble Quarries
Ahmet Turan Arslana,*, Bayram Kahramana, M. Kemal Özfırata, Thomas Frühwirtb,
Kemal Yıldızdağb, Halil Kösea
a
Faculty of Engineering, Dokuz Eylul University, İzmir, 35160, Turkey
b
Faculty of Rock Mechanics/Rock Engineering, TU Bergakademie Freiberg, Freiberg, 09599, Germany

Abstract

Marble quarries, are located at 3 km southeast of the city of Afyon İscehisar. The marble bed is lens shaped has a thickness of
300 meters. İscehisar is on south of the city center and marble bed extends through NW-SE direction for 6 km long and 1.5 km
wide. Present base elevation of İscehisar marble quarry is at 1020 m and it is planned to be excavated down to 980 m elevation.
The aim of this study is to determine the stability deep slopes at maximum depth (> 90 m) using numerical modelling. In
the finite element method, generalized Hoek-Brown and Mohr-Coulomb failure criteria together with a jointed rock mass model
are used. In marble quarries, particularly in this locations, the rock mass is severely affected by sliding blocks, caused by four
intersecting joint sets. Stability analysis were conducted under static and dynamic conditions using Phase2 V.9.016 software.
Shear Strength Reduction (SSR) technique that is built into the software used to determine the failure mechanisms and to suggest
and the necessary controls to ensure the stability of the slopes.
© 2017
2017Published
The Authors. Published
by Elsevier Ltd. by
ThisElsevier Ltd.
is an open access article under the CC BY-NC-ND license
Peer-review under responsibility of the organizing committee of EUROCK 2017.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of EUROCK 2017
Keywords: Open pit mine; Mohr-Coulomb parameters; Generalized Hoek-Brown Criterion; Finite element method; Slope stability

1. Introduction

Slope stability is one of the most important critical and mostly studied subjects of open pit mining. This is
because of the fact that slope instabilities can cause.

* Corresponding author. Tel.: +90-232-301-7273; fax: +90-232-453-0868.

E-mail address: ahmet.arslan@deu.edu.tr

1877-7058 © 2017 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.2017.05.228
 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655 647

Slope stability is controlled by many factors such as the local geological conditions, natural seismic activities,
ground water table and changes in pore pressures. The aim of the study is to determine the stability of deep slopes at
maximum depth using numerical modelling (> 90 m). The stability analyses were conducted using the methodology
suggested by Hoek et al. [1] and Hoek [2] using the generalized Hoek-Brown parameters. In dynamic loading
analyses, both Hoek-Brown and equivalent Mohr-Coulomb parameters were used [3–7]. The equations
recommended by Sofianos and Halakatevakis [3] were used to obtain the equivalent Mohr-Coulomb parameters for
the rock masses that have GSI value of more than 25. Sofianos [4]; Sofianos and Nomikos [5] suggested that when
estimating the equivalent Mohr-Coulomb parameters, the lower stress limit of a rock is changed to be equal to
the biaxial tensile strength, cohesion (c) and friction angle (ϕ) values obtained from these equations were compared
with estimated c and ϕ. Li et al. [6] suggested a rock slope sensitivity table using the equivalent Mohr-Coulomb
parameters and Hoek-Brown failure criterion in a limit-equilibrium analysis. They indicated that the factor of safety
values of slopes (higher than 45°) obtained from stability tables using the equivalent Mohr-Coulomb parameters are
somewhat higher due to the intervals wherein geometric discrepancy of curves from these two methods are high.
Therefore, they concluded that Mohr-Coulomb curves and Hoek-Brown criterion cannot be integrated into one
approach and then they proposed two different equations for slopes (with overall slope angles of t 45°) that
estimate minimum principal stress values. Nekouei and Ahangari [7] claimed that Hoek-Brown criterion is not
reliable due to the low correlation coefficient values using Mohr-Coulomb parameters and Hoek-Brown criterion in
slope stability tables proposed by [6, 9].

2. Geology of marble quarries

The surface rocks named Afyon Metamorphics are mostly originated from sedimentary rocks and underwent
metamorphosis during intermediate to high temperature and pressure. Consisting of different lithologies, they are
divided into sub-groups called upper and lower Metamorphics. The uppermost unit on the metamorphic foundation
is the Iscehisar marble (Fig. 1). These marble formations are originally limestone (carbonate) and show different
geological characteristics in colour, grain size and mineralogical composition so that they are divided into three
different formations. During the metamorphism, rocks in that region were exposed to diverse deformation processes
and temperatures. They have been folded and bended through plastic deformations. Structural analyses revealed that
marbles in this region had been affected by at least three different folding phases [10]. Subsequent to
metamorphism, brittle deformations occurred. Marbles in this region is believed to contain joints with different
orientations [10].

Fig 1. Marble quarries and near surroundings in the geological map.

Metamorphism of Afyon rocks was active until the Palaeozoic era (appr. 400 million years B.C.) and after that
they were overlained by sedimentary rocks. Such as Degirmendere recrystallized limestone, Elmacik polygenic
conglomerates and Gözsüzlü limestone are those sedimentary rocks which were not affected by metamorphism.
Moreover, lava flow and pyroclastic serials were also observed in this region [10].
648 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655

3. Seismicity

There are tectonically active major structures in this region, which are Gediz depression system, Dinar1 and
Aksehir fault systems. The rock mass around this region exhibits a geological age that is nearly 200 million years
old, and located where West Anatolia Depression system and Middle Anatolia plain regime intersects. Aksehir fault
zone is the significant mechanism which dominates the tectonics at this region [11, 12]. Utku et al. [13] studied
the earthquakes in this region and stated that Afyon city and its vicinity is inside of an active tectonic system which
can trigger an earthquake with a minimum magnitude of approximately 4 once every 5 months. The same study also
found that this system may trigger earthquakes with magnitude 6 once every 30 years with a probability of 78%.

4. Physical and mechanical properties

The hydro-mechanical properties of marbles at Iscehisar Region is summarised in Table 1. Considering

the values in Table 1 and using the classification proposed by Anon [14], specific bulk density and porosity values
are categorized as “high” and “very low”. This marble is defined as “unweathered – slightly weathered” rock [15].
Based on the classification proposed by Deere and Miller [16] these marbles are also classified as “low modulus
ratio”; however, taking the uniaxial compressive strength into account, it is concluded that rock mass to exhibit
“high strength” [16].

Table 1. Summarized hydro-mechanical properties of Iscehisar marbles (www.oguzmermer.com).

Parameter Value
Hardness (Mohs) 3
Weight per unit of volume (gr/cm3) 2.73
Specific weight 2.75
Water absorption percent by weight – under atmospheric condition (%) 0.10
Water absorption percent by volume – under atmospheric condition (%) 0.20
Water absorption percent by weight – in boiling water (%) 0.10
Water absorption percent by volume – in boiling water (%) 0.20
Porosity (%) 0.20
Uniaxial compressive strength (MPa) 70
Compressive strength after frost (MPa) 59
Impact strength (kg.cm/cm3) 23
Flexural strength (MPa) 15
Young’s Modulus of elasticity (MPa) 5.9×103
Compactness ratio [-] 99.3
Porosity degree (%) 0.70
Average abrasion resistance (cm3/50 cm2) 25.4
Average tensile strength (MPa) 3.9

5. Measurement of discontinuities

The engineering geomechanical properties of discontinuities present at the marble quarries were determined by
using the methods from suggested by ISRM [15]. The discontinuity properties, which are summarized in Table 2
were measured by line survey on selected levels. The discontinuity spacing, which can influence the rock mass
strength and behaviour, is defined as the distance between adjacent discontinuities or two discontinuities in
a discontinuity set consists of parallel joints. Although marble deposits at Iscehisar are partly fissured and faulted
block cutting is still the best possible method.
 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655 649

Table 2. Iscehisar marble geomechanical input data.

Intact Rock- Scattered Systematically Intensely
Parameters
Mass Fractured Fractured Fractured
Uniaxial compressive strength (MPa) 70 70 70 70
Density (kN/m3) 26.78 26.78 26.78 26.78
RQD (%) 100 100–90 75–50 50–25
Discontinuities spacing – 0.3–1.0 m 60–20 cm 20–6 cm
Discontinuities aperture – 0.1–0.5 mm 0.5–2.5 mm 2.5–10 mm
Discontinuities persistency – 10–20 m 10–20 m 3–10 m
Discontinuities roughness – rough rough slightly rough
Discontinuities morphology – planar planar planar
Discontinuities infilling – none soft soft
Wall rock weathering – none moderately moderately
JRC – 12–16 12–16 8–12
Discontinuities hydraulic condition – dry dry dry
Excavation category (ESR) 1.6 1.6 1.6 1.6
Excavation method undisturbed undisturbed undisturbed undisturbed
RMR/GSI –/100 90/85 70/65 50/45

Discontinuity persistence defines the areal extent and dimensions of a discontinuity along a plane and it is
an important parameter as it affects the stability of slopes. The discontinuity persistency was estimated to be
“moderate – very high” (Fig. 2) using the criterion suggested by ISRM [15].

6. Kinematic analyses

In order to conduct the kinematic analyses, 300 observations of discontinuities were conducted at different
locations. The results were then used in contour diagram analyses. Three main discontinuity sets were observed in
rock slopes (Fig. 3). The orientation of these joint sets are 224/58 (J1), 321/66 (J2) and 146/63 (J3). Joints measured
at Iscehisar marble quarries show dip values of minimum of 45°, average of 72° and maximum of 90°, prevailing
dip direction is NE-SW and NW-SE (Fig. 3).

Fig. 2. Discontinuities observed at marbles.

650 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655

Fig. 3. Rose diagrams based on dip directions of joints and the contours generated by measurements at joints in İscehisar marbles.

In order to evaluate planar failure envelope, pole centre of bigger circles were used. The critical planar failures
were then determined at the zones where the intersection angle between slope and discontinuity dip directions are
within an interval of ±20°. To determine the critical wedge failure envelope, the intersection points were determined
using the same interval. Assessment of critical zone where toppling failure may occur was also conducted using
the same interval. In general the overall slope angles vary; the steepest slope angle of 88° has been used for
kinematic analyses. Kinematic analyses were conducted at two slope locations where the likelihood of slope failures
is considered to be highest. Group 1 and 2 include slopes with general dip direction and ultimate dip angle of
230°/88° and 65°/88°, respectively. During the kinematic analyses for the Group 1, it is found that J1 appears to
trigger a planar failure (Fig. 4A). Furthermore it is assessed that in Group 1 wedge failure is anticipated to occur
where J1–J2 and J1–J3 intersects with each other (Fig. 4B). Possible toppling failure may be observed at the
intersecting joint sets of J1–J2 and J2–J3 and at J1 (Fig. 4C). Planar failure is not expected for the Group 2 (Fig. 5A)
likelihood of wedge failure is insignificant (Fig. 5B). Similar to Group 1, it is possible that toppling failure may
occur in Group 2 where J1–J2, J1–J3 and J2–J3 intersects (Fig. 5C).

Fig. 4. Kinematic analyses for the slopes categorized under Group 1.

 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655 651

Fig. 5. Kinematic analyses for the slopes categorized under Group 2.

7. Slope stability analysis

In open pit mines, the design of slopes is critical for safety and economic extraction of reserves. With the recent
improvements in computing capabilities, numerical modelling has become the preferred method assessing
the stability of slopes in geotechnical engineering. A finite element model is also utilised in this study. Unlike
the limit equilibrium analysis, calculations in FEM do not only include the divided zone above the failure plane but
the entire modelled. The mesh resolution is important consideration in FEM as the error caused by analysis increases
with decreasing mesh density. Both the stress reduction factor (SRF), which represents the critical slope failure
conditions and shear strength reduction (SSR) method, which defines the factor of safety, can be utilised in Phase 2.
In this parametric study, generalised Hoek-Brown failure criterion was employed and the effect of input
parameters’ variation on slope stability was examined. GSI values of 100, 85, 65, 45; RQD values of 100, 85, 65,
35; earthquake magnitude of Mw: 6.5 and slope dip direction of 88° were used as input data in the models (Table 3).
The slope heights of 124 m and 100 metres along AA’ and BB’, respectively were analysed (Fig. 6). Since the slope
heights are different, the above mentioned parameters were varied for each slope height in the analyses.

Fig. 6. The sections where the calculations with the code Phase2 were performed.
652 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655

Table 3. Iscehisar marble Rock mass physical property values.

Intact Rock-Mass Scattered Systematically Intensely
Parameters Fractured Fractured Fractured
RMR/GSI –/100 90/85 70/65 50/45
Modulus ratio (MR) 842 842 842 842
Intact modulus Ei (MPa) 59000 59000 59000 59000
Hoek&Brown index (mb) 9 3.95 1.32 0.44
Hoek&Brown index (s) 1 0.119 0.0063 0.0003
Hoek&Brown index (a) 0.5 0.5 0.5 0.5
Rock mass cohesion (MPa) 13 4.07 1.17 0.55
Rock mass friction angle 49 49 45 37
Rock mass uniaxial comp. strength (MPa) 70 23.59 5.49 1.22
Rock mass global strength (MPa) 64.72 25.64 11.32 6.00
Rock mass tensile strength (MPa) -7.78 -2.02 -0.33 -0.05
Deformation modulus (GPa) 37.0 31.0 16.0 5.0

Figure 7 shows the lower and higher of shear strength values by blue and green-red colours, respectively.

Fig. 7. The calculated shear strength deformation for the different values of the GSI in the Phase2.
 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655 653

Figure 8 shows the relationship between the calculated SRF values for different GSI values along the sections
A and B. The regression functions between SRF and GSI for the sections A and B are determined as
SRF = 24.502GSI-2.159 (R2 = 0.9834) and SRF = 27.18GSI-2,057 (R2 = 0.9707), respectively.

Fig. 8. The calculated SRF values versus GSI parameter values.

Under the dynamic loading conditions (i.e. earthquakes), it is important to determine the dynamic response
properties of a rock mass. Strength reduction and deformations should be investigated under the dynamic loading
conditions. It is well-know that the behaviour of a rock mass under dynamic loading depends on the rock mass
competence and deformations. Therefore stress – shear deformations, stress – volumetric deformations and
subsequent horizontal and vertical displacements along the sections A and B were analysed using Phase2 [17] under
the dynamic loading condition. The time interval of a possible earthquake is estimated by using the acceleration
– time data obtained from the Afyon Merkez station (Afyon Sultandağı earthquake in 03.02.2002, Mw: 6.5). For
different GSI values of 100, 85, 65 and 45, calculations were performed within this time interval (Fig. 9).

Fig. 9. Acceleration-time relation obtained from an earthquake in Afyon (left) and lateral displacement of a selected point on the slope surface.

Received points in the slope surface are calculated maximum horizontal displacements under seismic effect in
cross-section A and B analyses (Table 4; Fig. 10).

Fig 10. Contours of horizontal displacements calculated along sections A and B under dynamic loading together with selected points.
654 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655

Table 4. Calculated horizontal displacements of the selected points along the sections A and B under dynamic loading.
X displacement (m) X displacement (m)
Point Number Point Number
GSI GSI
1 2 3 4 5 1 2 3 4 5
100 0.039 0.035 0.022 0.033 0.037 100 0.017 0.011 0.007 0.011 0.012
Section A

Section B
85 0.078 0.064 0.035 0.053 0.058 85 0.029 0.015 0.009 0.016 0.017
75 0.20 0.15 0.06 0.10 0.12 75 0.06 0.027 0.014 0.028 0.031
65 0.50 0.37 0.12 0.35 0.39 65 0.13 0.05 0.025 0.076 0.094
55 1.59 1.12 0.45 1.76 1.96 55 0.64 0.12 0.055 0.42 0.54
45 2.00 1.46 0.99 3.54 3.82 45 1.38 0.34 0.13 1.35 1.73

The calculated horizontal displacements for different GSI values in a case of an earthquake with a magnitude of
6.5, indicate that higher displacements (> 0.25 m) can be where the GSI values 70–75. Consequently these higher
displacements may trigger regional failures (Fig 11).

Fig 11. The relation between calculated horizontal displacements and GSI values of selected points along the sections A and B for the earthquake
magnitude of 6.5.

8. Conclusion

In open pit mines, the design of slopes is critical for safety and economic extraction of reserves. In this study,
slope stability analysis is made by numerical modelling in marble pits. As a result, steepest marble slopes should be
controlled by FEM software in every production stage. Thus, production and work safety will be efficiency.

References

[1] E. Hoek, C. Carranza-Torres, B. Corkum, Hoek-Brown criterion – 2002 edition, in: Proceedings of North American Rock Mechanics
Symposium, Toronto, Canada, 2002, pp. 267–273. http://earthquake.usgs.gov/earthquakes/world/historical.php/
[2] E. Hoek, Rock Mass Properties, In Practical Rock Engineering, The University of Toronto Press, Canada, 2006, pp. 1–47.
[3] A.I. Sofianos, N. Halakatevakis, Equivalent tunnelling Mohr–Coulomb strength parameters for given Hoek–Brown ones, International Journal
of Rock Mechanics and Mining Sciences 39 (2002) 131–137.
[4] A.I. Sofianos, Tunnelling Mohr–Coulomb strength parameters for rock masses satisfying the generalized Hoek–Brown failure criterion,
International Journal of Rock Mechanics and Mining Sciences 40 (2003) 435–440.
[5] A.I. Sofianos, P.P. Nomikos, Equivalent Mohr–Coulomb and generalized Hoek–Brown strength parameters for supported axisymmetric
tunnels in plastic or brittle rock, International Journal of Rock Mechanics and Mining Sciences 43 (2006) 683–704.
[6] A.J. Li, R.S. Merifield, A.V. Lyamin, Stability charts for rock slopes based on the Hoek–Brown failure criterion, International Journal of
Rock Mechanics and Mining Sciences 45 (2008) 689–700.
[7] A.M. Nekouei, K. Ahangari, Validation of Hoek–Brown failure criterion charts for rock slopes, International Journal of Mining Science and
Technology 23 (6) (2013) 805–808.
 Ahmet Turan Arslan et al. / Procedia Engineering 191 (2017) 646 – 655 655

[8] E. Hoek, E.T. Brown, Practical estimates of rock mass strength. International Journal of Rock Mechanics and Mining Sciences 34 (8) (1997)
1165–1186.
[9] T. Koca Kadakçı, M.Y. Koca, Slope Stability Assessment of Rock Slopes in an Open Pit Albite Mine Using Finite Element Method (FEM),
Jeoloji Mühendisliği Dergisi 38 (1) (2014) 1–17.
[10] E. Sümer, A.Ü. Tolluoğlu, Y. Erkan, Importance of Geologic data in Marble Production and Three Dimensional Modelling of Afyon
Iscehisar Marbles (In Turkish), Turkish 2nd Marble Symposium Proceedings Book, Afyon, 1997, pp. 35–43.
[11] A. Koçyiğit, Inner plate New Tectonic development of South-west Turkey (In Turkish), Bulletin of Turkish Geology 27 (1984) 1–16.
[12] A. Koçyiğit, E. Ünay, G. Saraç, Episodic graben formation and extensional neotectonic regime in West Central Anatolia and the İsparta
Angle; a case study in the Akşehir-Afyon Graben, Turkey, in: E. Bozkurt, J.A. Winchester, J.A.D. Piper (Eds.), Tectonics and Magmatism in
Turkey and surrounding Area, Geol Soc. Spec: Publ. 173 (2000) 285–301.
[13] M. Utku, M.A. Danışman, N. Akyol, Z. Akçığ, The Seismicity of Afyon and its Surroundings: Isoseist Map of 03 February 2002 Çay
Earthquake, and Earthquake Risk Analysis, Jeoloji Mühendisliği Dergisi 27 (2) 2003 18–34.
[14] Anon, Classification of rocks and soils for engineering geological mapping, Part 1: Rocks and soil materials, Bull. Int. Assoc. Eng. Geol.
19 (1979) 364–371.
[15] ISRM, The complete ISRM suggested methods for rock characterization, testing and monitoring: 1974–2006, in: R. Ulusay and J.A. Hudson
(Eds.), Suggested Methods Prepared by the Commission on Testing Methods, International Society for Rock Mechanics, Compilation
Arranged by the ISRM Turkish National Group, Kozan Ofset, Ankara, 2007.
[16] D.U. Deere,. R.P. Miller, Engineering classification and index properties for intact rocks, Technical Report No, AFNL-TR, 65–116, Air
Force Weapons Laboratory, New Mexico, 1966.
[17] Phase 2 V 7.016, Two-dimensional finite element analysis program, 2016, Rocscience Inc.