Cogent Engineering

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/oaen20

Analytical and numerical assessment of a

preliminary support design – a case study

Sylvanus Sebbeh-Newton , Shaib Abdulazeez Shehu , Prosper Ayawah ,

Azupuri A. Kaba & Hareyani Zabidi |

To cite this article: Sylvanus Sebbeh-Newton , Shaib Abdulazeez Shehu , Prosper

Ayawah , Azupuri A. Kaba & Hareyani Zabidi | (2021) Analytical and numerical assessment
of a preliminary support design – a case study, Cogent Engineering, 8:1, 1869367, DOI:
10.1080/23311916.2020.1869367

To link to this article: https://doi.org/10.1080/23311916.2020.1869367

© 2021 The Author(s). This open access

article is distributed under a Creative
Commons Attribution (CC-BY) 4.0 license.

Published online: 11 Jan 2021.

Submit your article to this journal

Article views: 12

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=oaen20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

CIVIL & ENVIRONMENTAL ENGINEERING | RESEARCH ARTICLE

Analytical and numerical assessment of a
preliminary support design – a case study
Sylvanus Sebbeh-Newton1, Shaib Abdulazeez Shehu1, Prosper Ayawah2, Azupuri A. Kaba2 and
Hareyani Zabidi1*
Received: 18 October 2020
Accepted: 21 December 2020 Abstract: Underground tunneling results in redistribution of the in-situ stress state.
This process can create unstable underground conditions. Tunnel instability can
*Corresponding author: Hareyani
Zabidi, School of Materials and result in fatalities, property damages, project delays, high rehabilitation cost, and
Mineral Resources Engineering,
University Sains Malaysia, 14300
can compromise the overall functionality of the excavation. This study assesses the
Nibong Tebal, Penang Malaysia efficiency of the empirically recommended support design of the Pahang-Selangor
E-mail: [email protected]
Water Transfer Tunnel in Peninsular, Malaysia using convergence-confinement (CV-
Reviewing editor:
Ramon Fernando Colmenares,
CF) and 2-Dimensional two-dimensional Finite Element Method (FEM). To reduce the
Faculty of Engineering, Universidad effect of subjective uncertainties associated with the rock mass rating system, the
Cooperativa de Colombia, Medellin,
COLOMBIA lump rating system was replaced with the continuous rating system to classify
Additional information is available at
the rock masses and provide empirical support design for the tunnel. The maximum
the end of the article capacity and stiffness of the support systems were analytically determined using
the CV-CF method. The results of the maximum elastic deformation of the support
systems were compared to the actual critical internal pressures which show that
the empirically recommended support systems can satisfactorily stabilize the
underground excavation. The Reliability of the support design, the plastic zones, and
total deformations were determined using Phase2. The radius of plastic zones and
the maximum total deformations were significantly reduced after the supports were
installed. The plastic zones that remained were restricted within the reinforced
zones, signifying the effectiveness of the recommended support systems.

ABOUT THE AUTHOR PUBLIC INTEREST STATEMENT

Sylvanus Sebbeh-Newton is a PhD Candidate in A rock support system has a primary role of sup­
Geomechanics with a research focus in rock porting the rock mass around a tunnel after it has
mass quality modeling and stability investiga­ been disrupted by the excavation processes.
tions around underground excavations at Different tools; like empirical rock mass classifi­
Universiti Sains Malaysia. He received his B.Sc cation systems, analytical methods, and numer­
degree in Geological Engineering from Kwame ical modeling are used to assess the stability and
Nkrumah University of Science and Technology select optimal support for underground tunnels.
(KNUST), Ghana, in 2012 and his M.Sc degree in One of the serious concerns in the use of rock
Geological Engineering at University of Mines mass classification systems is that they are sub­
and Technology (UMaT), Ghana, in 2017. He also jective. The continuous rating system used for
holds a PgDip in Occupational Safety, Health, grading rock parameters reduces the subjective
and Environmental Management from Ghana uncertainties associated with the rock mass rat­
Institute of Management and Public ing (RMR) system. This method produces
Administration (GIMPA). He is a member of The a controlled RMR estimate between experienced
West African Institute of Mining, Metallurgy and and inexperienced rock engineers. The overall
Petroleum (WAIMM) and Ghana Institution of results obtained in this study show that the con­
Geoscientists (GhIG). tinuous rating system, convergence-confinement
method, and the 2-D Finite Element Method are
useful tools for preliminary assessments of
underground support designs.

© 2021 The Author(s). This open access article is distributed under a Creative Commons
Attribution (CC-BY) 4.0 license.

Page 1 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Subjects: Rock Mechanics; Tunnelling & Underground Engineering; Geomechanics

Keywords: rock mass rating; convergence-confinement method; numerical modeling;

tunneling; support systems

1. Introduction
Tunnel instability can result in fatalities, property damages, project delays, high rehabilitation cost,
and can compromise the overall functionality of the excavation. Tunnels are constructed in civil and
mining industries to serve several purposes including; mobility of people, water transfer, conveyance
of ore, irrigation, underground storage, military fortification, and hydro-electric power. Stability around
an underground excavation is affected by external and internal factors. These factors include; the type
of intact rock, type and nature of discontinuities, in-situ stresses, hydrological conditions, and the
excavation geometries (Kanik & Gurocak, 2018; Rehman et al., 2020; Xing et al., 2019). Discontinuities
such as bedding planes, joints, dykes, shear zones, and faults significantly reduce the strength of the
rock mass, complicate the mechanical behavior of the rock mass, and increases the rock deform­
ability. Instabilities may occur when the stress exceeds the rock strength or when a high differential
stress is encountered. During underground excavation, redistribution of stresses causes deformation
and failure around the openings which pose stability challenges in deep and shallow excavations.
Underground excavation challenges can be assessed by gathering site-specific rock engineering data.
Critical evaluation of data in rock engineering projects enables the selection of an appropriate
excavation method and the required support system for enhanced excavation efficiency and safety.

A rock support system in underground excavations has a prominent role of helping the rock
mass to support itself after it has been disrupted by excavation. Different tools; empirical, analy­
tical, and numerical modeling are used to assess the stability and select optimal support for
underground tunnels. Based on experience and extensive fieldwork, several researchers have
developed rock mass classification systems which are considered powerful tools for preliminary
support system design in civil and mining engineering tunnels. The commonly used systems are
the Q-system (Barton et al., 1974), Rock Mass Rating (RMR) (Bieniawski, 1989), Geological Strength
Index (GSI) (Hoek & Brown, 1997), and Rock Mass Index (RMi) (Palmstrøm, 1996).

The advent of sophisticated computers has proved numerical methods, which have the cap­
ability of simulating ground complexities, to be an efficient tool in rock and civil engineering.
Numerical modeling has been used by several researchers (Ghadimi Chermahini & Tahghighi, 2019;
Ur Rehman et al., 2019; Xing et al., 2018, 2019; Yalcin et al., 2016) to obtain the optimum support
design because of its’ ability to provide reliable information such as maximum displacement and
thickness of plastic zone which characterizes the rock mass behavior. Based on the thickness of
plastic zones and maximum deformation, Kanik et al. (2015) evaluated the effectiveness of
preliminary support obtained from RMR89 and RMR14. Similarly, Yalcin et al. (2016) determined
the effectiveness of supports recommended by RMR and RMi classification systems. Numerical
methods have the potential not only to solve complex tunneling problems, but to improve the
understanding and assessment of failure mechanisms, geotechnical risks, and more efficient
construction of rock reinforcement systems.

Although the empirical classification systems used by engineers make it easy to establish the
preliminary support systems, they do not provide any information on the plastic yield and total
maximum displacement needed to establish the optimal support system. One of the serious
concerns in the use of rock mass classification schemes is that they are subjective. Field engineers
with different experience levels classifying a given rock mass can result in significantly different
rock mass classifications. These inherent limitations of the empirical rock mass classification
system in underground support design are resolved by adopting both empirical rock classification
systems and numerical modeling techniques.

Page 2 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Convergence-confinement is an analytical method described by Sadeghiyan et al. (2016) as

a powerful tool for estimating the capacity of supports in underground excavations, the expected
plastic zone extent, radial displacement, and final convergence of the opening. The method is
a standard industry technique used for preliminary analysis of ground behavior after tunnel
excavation. The convergence-confinement technique as indicated by González-Cao et al. (2018)
is a simple and cost-effective solution for the preliminary design of excavation support. The
method gives rock engineers the opportunity to estimate the load imposed on a support installed
behind the face of a tunnel.

In this study, to reduce the effect of subjective uncertainties usually tagged with inexperienced
rock engineers, the quality of rock masses along the Pahang-Selangor Raw Water Transfer (PSRWT)
tunnel was classified using the widely accepted RMR system via the continuous rating system.
Convergence-confinement method was then utilized to evaluate the capacity of the recommended
rockbolt and shotcrete in stabilizing the tunnel. Finally, the effectiveness of the support design was
numerically verified based on the dimensions of the plastic region and maximum deformations.

2. Materials and method

2.1. Project description and geology of the study area

The Pahang Selangor Raw Water Tunnel (PSRWT) is the sole property of the Malaysian Ministry of
Energy, Green Technology, and Water and is located in Central Peninsula, Malaysia. The tunnel is
earmarked for transferring raw water (approximately 1890 million litres/day) from Pahang to
Selangor state to address the future water demand problems in Selangor and Kuala Lumpur
states. The tunnel has a maximum raw water discharge flow rate of 27.6 m3/sec, a total length
of 44.6 km, a diameter of 5.23 m and a longitudinal gradient of 1/1,900.

Geologically, Peninsular Malaysia is made up of four major tectonic zones, namely; the Western
Stable Shelf, the Main Range Belt, the Central Graben, and the Eastern Belt. Igneous rocks with
their associated metamorphic rocks are the most dominant rock types found along the tunnel.
Aside the Main Range granite of igneous origin, meta-sedimentary rocks of Palaeozoic and
Mesozoic age are also present.

The tunnel cuts through two major formations. The Karak Formation of Silurian-Devonian age
extends from the inlet portal to chainage 3.82 km consist of metasediments such as hornfels,
shale, metaschist, and phyllite. The Main Range granite is composed of coarse to very coarse-
grained, porphyritic biotite granite cut by minor porphyritic differentiates and fine-grained variety
of granite which represent a later phase of granite intrusion. Micro-granite, granodiorite, diorite,
monzonite, granite porphyry, quartz porphyry, felsite, vein quartz, megacrysts biotite granite,
megacrystic muscovite-biotite granite and equigranular tourmaline-muscovite granite are the
other foundation rocks within the Main Range Granite.

The Main Range Granite extends from chainage 3.82 km to the outlet in Langat, Selangor at
chainage 44.4 km. The main range granite is subdivided into three (3) sub-types, namely Bukit
Tinggi Granite, Genting Sempah micro-granite and Kuala Lumpur Granite. The Kuala Lumpur
Granite and the Genting Sempah Micro-granite isare separated by the Kongkoi Fault and the
Bukit Tinngi Fault also separates the Genting Sempah Microgranite from the Bukit Tinggi Granite.
While the Kuala Lumpur granite is megacrystic, the Genting Sempah Microgranite consistconsists
of microgranodiorite. The Bukit Tinggi Granite is constituted by very coarse grainedcoarse-grained
megacrystic biotite granite. The Main Range Granite formation is strongly folded due to the
intrusion of granitic rocks. The contact between the Main Range Granite and the Karak formation
lies approximately at chainage 3.82 km. The dips of the various rocks of the Karak formation
variesvary within a short distance due to folding within the formation. The Karak fault which trends
North-South cuts the tunnel at chainage 2.3 km. The rocks in the Karak formation have undergone
low to medium grade metamorphism and consist mainly of mudstone, siltstone, sandstone,

Page 3 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Figure 1. Geology of the major

rock formations along the
tunnel.

conglomerate, intraformational conglomerate, quartzite, schistose quartzite, quartz-mica schist,

and quartzo-hornfels. Figure 1 shows the general geology along the tunnel.

The tunnel route is divided into seven main sections, namely NATM-1, 2, & 3, TBM-1, 2, & 3, and
NATM-4, as schematically represented in Figure 2. This study was carried out in NATM-1 section of
the tunnel which falls within the Karak Formation. Based on geological and structural character­
istics, NATM-1 was sub-divided into six sub-units namely section 1, section 2, section 3, section 4
section 5, and section 6 made predominantly of sandstone, shale, andesite, phyllite, brecciated
metavolcanic, and metasiltstone, respectively.

2.2. Geotechnical investigations

For the entire project, scanline surveys and drilled cores according to ISRM (2015) were used to
describe the discontinuities and define the rock quality designation (RQD). The discontinuity
studies comprised width and infill type, degree of weathering, orientation, persistence, opening,
roughness, and spacing of measured discontinuities. Dip and dip direction measurements were

Figure 2. Section view showing

the 7 main sections along the
tunnel route.

Page 4 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Figure 3. Equal-Angle
Stereographic projection for
selected sections in the study
area.

processed with Dips 7.0 software based on equal-angle stereographic projection (Figure 3) to
identify the joint sets for each section. The joint sets in NATM-1 have close to very close spacing
with occasional wide spacing, medium persistence, very tight to tight opening, rough to smooth
planar roughness, slickenside, and moderate to high weathering characteristics. The attitudes of
the identified joint sets in the various sections are listed in Table 1. Laboratory experiments
including unit weight (γ), uniaxial compressive strength (σci), Young’s modulus (Ei,), Poisson’s
ratio, and triaxial test were conducted on the core samples drilled from rock blocks of all the
geological units based on techniques suggested by ISRM (2015).

Table 1. Joint sets obtained from each section of NATM-1

Sections Section 1 Section 2 Section 3 Section 4 Section 5 Section 6
Dips/Dip 79/202 65/334 87/049 74/022 77/028 83/062
direction 76/158 57/151 77/183 71/157
58/171 81/093
77/066 81/079 73/217
64/322 87/235
77/110 82/180

Page 5 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

2.3. Classification of rock masses along the tunnel section

Rock Mass Rating (RMR) also called geomechanics classification was used to classify the rock
masses along the tunnel. RMR is a supervised rock classification system, detailed and first
published by Bieniawski in 1973. The RMR system relies primarily on six rock mass parameters
namely: uniaxial compressive strength of rock material, RQD, spacing of discontinuities, condi­
tion of discontinuities, groundwater conditions, and orientation of discontinuities. The ratings of
individual parameters are summed to give a final RMR value. The final RMR value ranges from 0
to 100. The total rating is then used to categorize the rock mass into classes. The system has
five rock mass classes; I, II, III, IV, and V. In terms of rock mass competence, it decreases from
class I through class V, with class V been the least competent rock mass. In tunnel excavation,
these rock mass groups are used to determine the ground support system required to stabilize
the tunnel walls.

The RMR system has been refined by different researchers after the introduction of the RMR89
version. Geocontrol (2012) modified the basic RMR (RMRb) by introducing fracture frequency
defined as the number of joints per meter in the excavation face to replace RQD and discontinuity
spacing. These changes eliminated the difficulty in determining the RQD from excavation faces and
to obtain a good assessment of the condition of the discontinuities in the ground. Celada et al.
(2014) updated the RMR89 version after its existence for 25 years and named it RMR14 to distin­
guish it from the earlier versions. The revision includes the addition of new parameters, a revised
rating, and a final structure. The three new parameters include; intact rock alterability due to water
(swelling), an adjustment factor for the excavation method (Fe), and an adjustment factor related
to the stress-strain behavior of tunnel face.

Şen and Bahaaeldin (2003) developed the continuous grading system to reduce personal judg­
ment in the existing classical lump-rating system developed by Bieniawski (1973). Best-fit lines and
� intact rock
equations for rating of rock quality designation (RQD), average discontinuity spacing (X),
strength (σ), and groundwater conditions (G) were developed as given in equations 1, 2, 3, and 4
respectively. With this modification, only the ratings for discontinuity conditions and orientation of
discontinuities are affected by subjective uncertainties. This method generates a controlled RMR
estimate with a maximum gap of less than 10 % between experienced and unexperienced rock
engineers (Şen & Bahaaeldin, 2003).

In this study, the continuous grading system was used to grade the RMR parameters. The
continuous grading system only rates the RMR parameters, so the RMR89 was used to assign the
corresponding support design for the obtained rock classes. The recommended support systems
are summarized in Table 2.

rRQD ¼ 0:2RQD (1)

�
�
rX� ¼ 24 þ 151log X (2)

rσ ¼ 0:075σ (3)

rG ¼ 10 2:9logðGÞ; (4)

where rRQD represents the rating for RQD, rX� is the rating for average discontinuity spacing, rσ is the
rating for the intact rock strength, rG is the rating for the groundwater conditions.

Page 6 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Table 2. Empirical tunnel support design for the rock masses along the tunnel section based on
RMR
89
Support Section 1,2,3,4, and 5 Section 6
Average RMR 36, 29, 28, 33, 38 43
Class Poor rock Fair rock
Construction Phase Top heading and bench Top heading and bench
Excavation method Drill and Blast Drill and Blast
Advance length 1.0–1.5 m advance in top heading 1.5–3 m advance in top heading
Timing of support Install support concurrently with Commence support after each
excavation blast
Bolting Systematic bolts 4–5 m long, Systematic bolts 4 m long, spaced
spaced 1–1.5 m in crown and walls 1.5–2 m in crown and walls with
with mesh wire mesh in crown
Shotcreting 100–150 mm in crown and 50–100 mm in crown and 30 mm
100 mm in sides in sides
Steel sets Light to medium ribs spaced 1.5 m None
where required

2.4. Estimating of rock mass parameters for the analytical and numerical modeling

2.4.1. In-situ stresses

In underground excavations, in-situ stresses play an important role in the design and construction
stages. Results acquired from the calculation of in-situ stresses at different places indicate that the
magnitude of horizontal stress is usually greater than vertical stress at shallow depths (less than
500 m), while it occurs as hydrostatic at depths of approximately 1000 m below the surface (Brady
and Brown, 2004). In-situ stresses in rocks can be obtained either directly by the use of expensive
and difficult techniques such as flat-jack, over-coring, under-coring, and hydraulic fracturing or
indirectly using empirical approaches and through experiences gained from other nearby under­
ground projects (Rehman et al., 2020). In this study, the vertical stress for each section was
estimated from the relation (equation 5);

σv ¼ γH; (5)

whereσv = vertical stress (MPa), γ = average unit weight of rock mass (MN/m3), H = overburden
thickness in m.

The horizontal stressðσh Þ at depth H is much more difficult to determine than the vertical
stresses. The equation for estimating the mean horizontal in-situ stress was formulated by
Sheorey et al. (2001) and is based on the elastic constants of the rock mass, the geothermal
gradient due to the cooling of the crust, and the thermal expansion coefficient. In this study, the
horizontal stress for each geological unit was obtained based on the elastic constant of the rock
mass using equation 6 as established by Sheorey et al. (2001).

v βEG
σh ¼ γH þ ðH þ 1000Þ (6)
1 v 1 v

where,where σh = horizontal stress, H = depth of cover; E = elastic modulus; v = Poisson’s ratio; γ =

unit weight of the rock mass; β = coefficient of linear thermal expansion of rock (8.0 × 10−6/°C
appears to be a reasonable representative value for different rock types but not for coal); and G =
geothermal gradient (for crustal rocks = 0.0240°C/m, for coal measure rocks = 0.030°C/m).

Page 7 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

2.4.2. Hoek-brown constants and rock mass deformability

Hoek and Brown (1980) developed the Hoek-Brown failure criterion as one of the non-linear criteria
for defining the strength of rock mass. The failure criterion is generally recommended for intact
rocks and moderately to heavily jointed rock masses (Hoek and Brown, 2019). The Generalised
Hoek-Brown constants (mb, s, and a) were determined using the following equations;

� �
GSI 100
mb ¼ mi exp (7)
28 14D

� �
GSI 100
s ¼ exp (8)
9 3D

1 1� GSI=15 20=3
�
a¼ þ e e (9)
2 6

where mi is the intact rock parameter and can be obtained from the triaxial tests or a published
chart. GSI is the Geological Strength Index, D which ranges from 0 for undisturbed and 1 for very
disturbed rock mass, is the disturbance factor that depends on the degree of disturbance from
blast damage and stress relaxation (Marinos et al. 2007). D was set to 0.1, representing the
situation of low disturbance from blasting. In this study, RocData 5.0 software (Rocscience Inc.
2016) was used to acquire the mi of each dominant rock material along the tunnel section.

Tunneling is known to disrupt the initial in-situ conditions, thereby altering the pre-excavation
parameters (Satici & Ünver, 2015). Most rock masses at lower confinement levels reach their
residual strength when strained significantly by exhibiting some post-peak strength losses. The
rock mass parameters would change after excavation, so for a reliable numerical model, the
residual load-bearing capacity of the rock mass is an important input parameter since it influences
the tunnel stability, and attained only after noticeable plastic deformation. For this study, no post-
peak behavior test was conducted. Therefore, the Residual Geological Strength Index (GSIr Þ and
the Residual Hoek-Brown parameters (mbr ; sr ; andar Þwere calculated using equations 10 to 13 (Cai
et al., 2007).

0:0134GSI
GSIr ¼ GSIe (10)

GSIr 100
mbr ¼ mi eð 28 Þ (11)

GSIr 100
sr ¼ eð 9 Þ (12)

1 1 � GSIr =5 20=3
�
ar ¼ þ e e (13)
2 6

The deformability of rock masses can be determined directly or indirectly. The indirect methods
were used since it depends on empirical methods which make use of the rock mass classification
indices and the equivalent continuum approach. The equivalent continuum approach treats the
rock as a continuous material and its deformability reflects the deformation properties of both the
intact rock and the discontinuities (Zhang, 2017). The deformation modulus Em, and the

Page 8 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

compressive strength of the rock mass ðσrm Þ for all the sections were obtained empirically using
equations 14 and 15 developed by Hoek and Diederichs (2006) and Hoek et al. (2002) respectively.

!
1 D=2:0
Em ¼ Ei 0:020 þ ð60þ15D GSIÞ
(14)
1þe 11

�a 1
ðmb þ 4s aðmb 8sÞÞ m4b þ s
σrm ¼ σci : (15)
2ð1 þ aÞð2 þ aÞ

where Em is the deformation modulus, Ei is Young’s modulus of the intact rock, σci is the uniaxial
compressive strength of the intact rock.

All the parameters used for both analytical and numerical assessment of the supports are listed
in Table 3.

2.5. Analytical methods (convergence-confinement)

The Convergence-confinement (CV-CF) method is a useful analytical tool for preliminary estimation
of support capacities, the extent of the plastic zones, and deformations around an excavation. The
method was implemented in two stages. The ground reactions around the tunnel was first
analyzed using the closed-form solution developed by Carranza-Torres and Fairhurst (2000). This
method is based on the Hoek-Brown failure criterion and assumes a cylindrical tunnel of radius R,
exposed to a uniform far-field stress (σ0) and internal pressure σR. The uniform internal pressure
and far-field stress are “scaled” to give the scaled internal pressure Pi and scaled far-field stress So,
respectively, in MPa (equations 16 and 17).

σR s
Pi ¼ þ (16)
mb σci mb 2

Table 3. Rock mass and material Properties used for Numerical and Analytical modeling
Property/ 1 2 3 4 5 6
Section
Elastic type isotopic isotopic isotopic isotopic isotopic isotopic
Strength of 6.568 1.644 7.414 5.103 5.422 3.957
rock mass,
MPa
Deformation 1258.3 369.9 1434.1 2525 3174.1 1888.2
modulus of
rock mass,
Mpa
Poisson’s 0.298 0.416 0.289 0.32 0.314 0.343
ratio of rock
mass
Vertical 2.295 2.349 2.214 2.7 1.485 1.485
stress, MPa
Horizontal 1.35 1.804 1.32 2.054 1.619 1.357
stress,MPa
GSI 31.8 28.5 27.75 31.85 34.4 38.48
mi 13 4 20 7 7 7
mb 1.138 0.311 1.518 0.617 0.672 0.778
s 0.000512 0.000355 0.000328 0.000523 0.000683 0.00108

Page 9 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

σo s
So ¼ þ (17)
mb σci mb 2

where the uniform internal pressure is represented by σR (MPa), σo obtained from the relation (σo =
(σv + σh)/2) mb and s are the Hoek-Brown’s constants. The applied internal support pressure on the
tunnel walls is given as σR ¼ ð1 λÞσo where λ is the de-confining rate which varies from 0 at the
initial excavation stage to 1 for a fully excavated tunnel (De La Fuente et al., 2019).

The critical internal pressure, pcr

i marks the elastic limit and is considered as the point where the
elasticity of the rock mass ceases and transitions into the plastic behavior. The elastic behavior is
maintained when σR ≥pcr cr
i . A plastic zone of radius Rp develops around the opening if σR <pi . The
scaled ðPcr cr
i ) and actual (non-scaled) critical internal pressure pi are obtained using the equations
below:

1 h pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffii2
Pcr
i ¼ 1 1 þ 16So (18)
16

� �
s
pcr cr
i ¼ Pi 2
mb σci (19)
mb

The relationship between the radial displacements (Ue Þ and internal pressure σR for σR ≥ pcr
i in the
elastic region is given by equation 20.

1þv
Ue ¼ R ðσ o σR Þ (20)
E

The radius of the plastic region Rp and the radial displacement Up that develops around the tunnel
when σR <pcr
i are obtained from the equations below:

h �qffiffiffiffiffiffi pffiffiffiffi�i
Rp ¼ exp 2 Pcr i Pi R (21)

"
pffiffifficrffiffiffi #� � � � ��2
1þv Pi � 1 Rp 22v 1 2v Rp
Up ¼ R σo Pcr
i þ 1 þ � ln
Em 2 So Pcr i R 4 So Pcr
i
R
pffiffifficrffiffiffi � � � �
1 2v Pi Rp
2 ln þ1 (22)
2 So Pcr i R

The equations and methods described above were developed specifically for circular tunnels. Upon
realization that most underground excavations are non-circular, Carranza-Torres and Fairhurst
(2000) suggested the use of equivalent radius, Re, for non-circular tunnels such as horse-shoe
cross-sections. The Equivalent Radius (Re) which is identical to a circular tunnel of radius R was
computed using equations 23 and 24 (Kabwe et al., 2020).

rffiffiffi
A
Re ¼ (23)
π

1 2
A¼ πR þ LR (24)
2

Page 10 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

where A is the cross-sectional area of the horseshoe tunnel, L is the height of the tunnel and π is
constant.

The input parameters used for the convergence-confinement analysis include the equivalent
tunnel radius Re, σo , σR, mb, s, σci ; Em and v. The equivalent tunnel radius, Re of 2.89 m was constant
for all the tunnel sections. The far-field stresses, σo was calculated for all the sections and the
uniform internal support pressure, pi was zero because the de-confining rate is taken as one (1) for
fully excavated unsupported tunnel cases. The Hoek-Browns constants, intact rock strength, and
the dynamic rock mass properties were changed for each section as described earlier.

In the second stage of the CV-CF analyses, the maximum support pressures and elastic stiffness
based on the recommended support systems were estimated for all the outlined sections.

�
The maximum support pressure Pmax
sh and the elastic stiffness (Ksh Þ provided by shotcrete are
given as:

" #
σcc ðR tc Þ2
Pmax
sh ¼ 1 (25)
2 R2

� �
Ec R 2 ðR tc Þ2
Ksh ¼ (26)
2ð1 vc 2 ÞðR tc ÞR2

where σcc the unconfined compressive strength of shotcrete in MPa, Ec is the Young’s Modulus of
the shotcrete, vc is the Poisson’s ratio of the shotcrete assumed to be 0.25, tc is the thickness of
shotcrete which depends on the surface roughness after blasting and scaling and usually varies
between 50 and 100 mm for temporary support. Singh and Bortz (1975) provided a table for the
selection of typical values of σcc and Ec for both dry and wet shotcrete mixture.

For equally spaced mechanically anchored rockbolts, the maximum support pressure Pmax
rb , and
elastic stiffness Krb are given as;

Tbf
Pmax
rb ¼ (27)
sc sl

Eb πdb 2
Krb ¼ (28)
4lsc sl

where d is the diameter of bolt (m), l is the free length of the bolt (m), Tbf is the ultimate load of
the bolt obtained from a pull-out test (MN), Q is a deformation-load constant for the anchor and
head [m/MN], Eb is Young’s modulus for the bolt (MPa), sl is the longitudinal bolt spacing (m) and
sc is the circumferential bolt spacing given as sc ¼ 2πR
n ; where n is the number of equally spaced
bolts.

For a section where two or more support systems are installed, Carranza-Torres and Fairhurst
(2000) emphasized that the elastic stiffness of each support must be summed to determine the
combined effects. For a case where both rockbolts and shotcrete with their maximum pressures
defined as Pmaxrb and Pmax
sh and their elastic stiffness as Krb and Ksh respectively, the combined
stiffness Ks is computed as Ks ¼ Ksh þ Krb . The computed combined effect value remains valid until
a maximum possible elastic deformation is achieved by one of the two installed support systems

Page 11 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

where the combined support system is assumed to fail. The maximum possible elastic deformation
is given as:

Pmax
Umax
r ¼ s
(29)
Ks

The support with the lowest value of Umax

r defines the maximum support pressure required for the
two support systems operating together. The parameters of the recommended support in Table 2
were substituted into the above equations to obtain the maximum support pressure, elastic
stiffness, and the maximum possible deformation for the combined support systems.

2.6. Development of numerical model

Numerical modeling techniques are valuable tools in underground excavation designs since open­
ing geometry, strength parameters, and in-situ stresses are considered. In civil and mining
engineering, numerical modeling is used to analyze the rock mass behavior and its impact on
infrastructure and support mechanisms. To determine the displacement and plastic region around
an opening and to verify the empirically recommended support designs, numerical methods such
as Discrete Element Method (DEM), Finite Element Method (FEM), Boundary Element Method (BEM),
and Analytic Element Method (AEM) have been implemented by several researchers (Ghadimi
Chermahini & Tahghighi, 2019; Kanik & Gurocak, 2018; Morris & Johnson, 2009; Rehman et al.,
2020; Xing et al., 2019). For this study, Phase2 V. 7.0, a FEM software, was used to model the
deformations and plastic zones around the tunnel. Phase2 was selected because of its plastic
modeling capabilities, ability to handle multiple materials, and automatic mesh generation func­
tion. This code only allows 2-D analysis of non-linear deformations.

In rock engineering, the two commonly used numerical modeling approaches are the continuum
and discontinuum methods. The choice of a continuum or discontinuum approach depends on
a variety of problem-specific conditions; on the scale of the problem and the geometry of the
fracture system. Continuum approach may be used if there are only a few fractures and if the
opening of the discontinuity and the complete block detachment are not vital components.
Discontinuum approach is suitable for moderately to highly jointed rocks where the properties of
discontinuities have a bigger impact on deformation and failure modes or where there is
a potential for large-scale displacement of individual blocks (Backers, 2010; Jing, 2003; Ma & Fu,
2014). Continuum numerical approach considers the rock as continuous. Though discontinuities
could explicitly be added to the model, the continuum approach is only useful in instances where
the length of the material’s microstructure is significantly smaller than the object under considera­
tion (Xing et al., 2017). The added joint elements only produce limited deformation without any
detachment (Kulatilake & Shu, 2015). Continuum modeling assumes that the rock unit cannot be
opened or broken and that the joint spacing is close or extremely close (Satici & Ünver, 2015). The
majority of the discontinuities spacing in the study area varied between close to very close spacing.
This qualifies the surrounding rock mass behavior to be modeled as a continuum model which
assumes a continuous material throughout the body (Wyllie, 2018; Wyllie & Mah, 2017).

Due to low overburden thickness, the upper boundary was set free but the bottom, left and right
sides were fixed to prevent movement in X and Y directions. Parametric studies were carried out to
determine the optimal external boundary which is not influenced by the edge and corner effects,
by setting different boundary limits. No edge and corner effects were observed when the excava­
tion boundaries were set at eight times the tunnel width. To obtain reliable results, the external
boundaries were set to 10 times the span of the tunnel for all sections. A six node triangular finite
element mesh was automatically generated for all the sections and finer zoning was adopted for
the tunnel periphery.

Page 12 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Since the interest of this study is to observe the failure and deformations that would occur
around the excavation, the plastic material type was chosen. Elastic material type enables the
estimation and plotting of strength factors but does not model the failure especially the yielded
zone around the opening. The Hoek-Brown failure criterion is recommended when the rock mass is
heavily jointed or contains more than two joint sets (Hoek & Brown, 2019). In this research, all the
sections contain more than two joint sets (see Table 1), so the Hoek–Brown failure criterion was
used to calculate the yielded elements and plastic regions around the tunnel. Tunneling is
a 3-D problem, but in this study, a 2-D analysis was conducted, so the load splitting option of
the software code was used to simulate the 3-D deformations before the implementation of the
support system.

In the final stage of the numerical modeling, the performance of the recommended support
systems by the empirical method was investigated by considering the various changes in the
maximum deformations and width of the developed plastic zones.

3. Results and discussion

3.1. Convergence-confinement method

From the convergence-confinement analysis, the actual critical internal support pressures pcr i
calculated for all the sections are greater than the internal support pressure for the fully excavated
tunnel providing a valid argument of the expected plastic zone of varying extent around the
opening. The analytically determined radius of the plastic zone is 6.33 m, 15.39 m, 5.81 m,
9.84 m, 5.99 m, and 6.18 m for sections 1, 2, 3, 4, 5, and 6, respectively. The results of actual
critical support pressure, maximum deformation, and the radius of the plastic zone obtained using
the convergence-confinement method for each section are given in Table 4. The strain values for
all the sections were empirically calculated using equation 30 (Evert Hoek & Marinos, 2000). All the
calculated percentage strain values are less than 1%. According to Evert Hoek and Marinos (2000),
a strain value of less than 1% implies only few instability problems are to be expected. Therefore
simple tunnel support designs as recommended by the rock mass classification systems could
stabilize the tunnel.

� � p
pi σcm ð2:4poi 2Þ
ε ¼ 0:0020 0:00250 (30)
po po

where ε is the strain, pi is the internal support pressure, po is the vertical in-situ stress and σcm is
the strength of the rock mass.

The results of the maximum possible elastic deformation (see Table 4) were compared to the
actual critical internal pressures as was similarly done by Kaya and Bulut (2019) and it was realized
that the Umax
r of the combined supports for each section is greater than the actual critical internal
pressure pcri which signifies that the empirically recommended support systems can satisfactorily
stabilize the tunnel.

3.2. Results of numerical model

The developed deformation and plastic regions were assessed based on an elasto-plastic
study. The properties of the recommended support systems used are set out in Table 5. For
each section, the thickness of the developed plastic regions, yielded elements, and max­
imum deformation was obtained for both supported and unsupported cases and presented
in Table 6. The maximum total displacement values for the unsupported case for each
section are very low. Nevertheless, the plastic zone thickness around the excavation for
both crown and walls in all the sections indicates that the tunnel will encounter instability
problems.

Page 13 of 20
 Table 4. Convergence-Confinement analysis results for NATM-1 of PSRWT tunnel
Parameter/Section 1 2 3 4 5 6
https://doi.org/10.1080/23311916.2020.1869367

Uniform internal 0 0 0 0 0 0
pressure σR
Far field stress σo 1.822 2.076 1.767 2.377 1.552 1.421
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367

Scaled internal pressure 0.00040 0.00367 0.00014 0.00137 0.001513 0.001776

Pi
Scaled far field stress So 0.03242 0.27071 0.02342 0.07406 0.04509 0.05340
Scaled critical internal 0.003375 0.107088 0.00186 0.014289 0.006085 0.008331
pressure ðPcr
i )

Actual critical internal 0.169555 0.804105 0.130341 0.422312 0.162845 0.178485

pressure pcr
i

Strain ɛ % 0.024419 0.408104 0.017833 0.056005 0.014997 0.028172

Radius of plastic zone Rp 6.32837 15.38813 5.80923 9.84361 5.99276 6.18374
Pmax
sh 1.19 1.19 1.19 1.19 1.19 0.595
Ksh 0.380 0.380 0.380 0.380 0.380 0.187
Pmax
rb 0.087 0.087 0.087 0.087 0.087 0.087
Krb 0.019 0.019 0.019 0.019 0.019 0.019
Elastic stiffness of 0.399 0.399 0.399 0.399 0.399 0.206
combined support, Ks
(MPa/m)
Maximum support 1.250 1.250 1.250 1.250 1.250 0.657
pressure provided by
combined support, Ps
Umax
r (m) 3.132 3.132 3.132 3.132 3.132 3.192

Page 14 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Table 5. Characteristics of support elements used for numerical modeling of the tunnel
Properties Rockbolt Shotcrete Wire mesh
Young’s Modulus (GPa) 200 30 200
Poisson’s ratio - 0.2 0.35
Uniaxial Compressive - 35 500
Strength (MPa)
Residual Uniaxial - 3.5 -
Compressive Strength
(MPa)
Peak Tensile Strength - 3.1 500
(MPa)
Residual Tensile Strength - 0 -
(MPa)
Peak load (MN) 0.25 - -
Residual load (MN) 0.025 - -
Type Φ25 mm - Φ 6.5 mm
Fully bonded 150 × 150 mm

Table 6. Results of Numerical Analysis for NATM-1 of the tunnel

1 2 3 4 5 6
Unsupported Case
Yielded 356 574 328 416 331 373
elements
Radius of 8.354 14.613 8.155 8.907 7.775 8.344
plastic zone
(m)
Maximum 2.206 20.341 1.915 1.554 0.624 1.156
total
displacement
(cm)
Supported Case
Yielded 193 552 189 275 184 199
elements
Radius of 7.706 13.884 7.672 9.423 7.268 7.785
plastic zone
(m)
Maximum 1.857 19.576 1.522 1.506 0.558 0.954
total
displacement
(cm)

Figures 4 and Figures 5 show the Extent of developed Plastic zones around the sections for
both supported and unsupported cases. The maximum total displacements occurring along the
tunnel for the unsupported cases have a minimum value of 0.625 cm and a maximum value of
20.341 cm. Maximum displacement in an excavation is directly proportional to the strength of
the rock mass and inversely proportional to the in-situ stresses. Section 2, dominated by shale,
produced the highest maximum deformation value of 20.341 cm. Shale has the least rock mass
strength and deformation modulus of 1.64 MPa and 369.9 MPa, respectively, compared to the
other rock mass units along the tunnel. Section 5—dominated by phyllite, however, is ranked as
the section with the third highest rock mass strength and the highest deformation modulus of
5.42 MPa and 3174.1 MPa , respectively. The relatively higher rock mass strength and deforma­
tion modulus coupled with minimum vertical stress of 1.485 MPa accounted for the lowest

Page 15 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Figure 4. Extent of developed

Plastic zones for .sections 1, 2,
and 3

deformation of 0.625 cm for section 5. In numerical modeling, yielding is one of the most
important criteria when the elasto-plasticity analysis is employed. Yielding condition is said to
occur when the rock is loaded beyond its elastic limit.

The extent of the plastic zone around the unsupported sections pre-informs of a major instability
if the tunnel is left unsupported, especially in section 2 which recorded a maximum plastic zone of
14.61 m. Similar value (i.e 16 m) was obtained for the CV-CF method. Section 2 is dominated by
moderate to highly weathered shale and an estimated rock mass strength of 1.64 MPa. The
analytical approach used in this research tends to produce similar results with the 2-D FEM and
therefore could be deduced that both the 2-D FEM and convergence-confinement method could be
reliably used in preliminary underground support designs.

Page 16 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Figure 5. Extent of Developed

Plastic Zones for sections 4, 5,
and 6.

The differences in yielded elements, the extent of the plastic region, and the variation in
deformations after the implementation of the recommended support at each section were studied
and compared with the non-supported scenarios. The plastic zone defines the extent of damage
when the material yields. The number of yielded elements were reduced in all the sections.

The total maximum deformation reduced averagely by 11.62 % after the recommended support
systems were installed. Except for section 4, the radius of the plastic zones decreased significantly in
the other sections. The behavior of the opening is considered problematic if the yielded zones extend
beyond the length of the installed rockbolts (Li, 2017). Hoek and Brown (1997) indicated that to avert
serious tunnel instability problems, the ratio of maximum tunnel deformation to tunnel radius must be
maintained below 0.02. Section 2 produced a tunnel deformation to radius ratio of 0.08 which reflects
serious instability problems. Phase 2 is known to produce small strains in numerical modeling, there­
fore it is limited in accommodating very large strains. Therefore, relying on the obtained magnitudes of
deformation rather than the extent of plastic zones and yielded elements for stability assessment is

Page 17 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

misleading. Although the suggested supports reduced the thickness of plastic regions and yielded
elements, a substantial percentage of plastic zones still remained around the openings. As observed
from the supported cases in Figures 4 and Figures 5, the failure zones around the walls and crowns for
all the selected sections fall inside the supported zone which confirms the stability of the excavation
hence affirming the efficiency of the recommended support by the empirical rock mass rating system.

4. Conclusions
In this study, the stability and effectiveness of the empirically recommended support for the Pahang-
Selangor Raw Water Transfer tunnel were investigated using the convergence-confinement method
and numerical modeling. To minimize the subjective uncertainties associated with Rock Mass Rating
system, the continuous rating system developed by Şen and Bahaaeldin (2003) was used for rating the
parameters involved in the Rock Mass Rating system. The convergence-confinement method was
analytically utilized as a preliminary tool to estimate the load imposed on the supports and the ground
reactions around the excavation. A two-dimensional Finite Element Method was used to analyze the
effect of induced stresses and the effectiveness of the empirically recommended support systems.

The main conclusions of this study can be summarized as follows:

(1) The maximum elastic deformation of the combined supports (Umax r ) for each section was
�
greater than the actual critical internal pressure pcr
i which signifies the effectiveness of the
empirically recommended support design in stabilizing the tunnel.
(2) The radius of the plastic region and maximum deformations estimated by the convergence-
confinement method and 2-Dimensional Finite Element Method were very similar for all the
sections along the tunnel route. Thus the two methods can be reliably used for preliminary
underground support designs, especially for horse-shoe shaped tunnels.
(3) The radius of plastic regions and the maximum total deformations were significantly
reduced after the installation of the recommended support systems. The failure zones
that remained were restricted within the reinforced zone, hence confirming the effective­
ness of the recommended support systems.
(4) From a practical standpoint, the overall results obtained in this study show that the con­
tinuous rating system, convergence-confinement method, and the 2-D Finite Element
Method are useful tools for preliminary assessments of underground support designs. The
convergence-confinement method used in this research is only applicable for initial or
preliminary tunnel design stages.
(5) The rock mass contains joints, thus anisotropic behavior should be expected. As such, the
Hoek-Brown criterion may produce deceptive results. For accurate assessments of the
support system, non-isotropic models and discontinuum numerical modeling approach
such as the Distinct Element Method (DEM) should be explored for future research.

Acknowledgements Disclosure statement

Authors are grateful to the Ministry of Energy, Green The authors wish to confirm that there is no known con­
Technology, and Water and Shimizu-Nisihimatsu-UEMB flict of interest associated with this publication.
IJM Joint Venture (SNUIJV) for the data provided for
this research.
Funding
The authors received no direct funding for this research.
Author details
Sylvanus Sebbeh-Newton1 Citation information
Shaib Abdulazeez Shehu1 Cite this article as: Analytical and numerical assessment
Prosper Ayawah2 of a preliminary support design – a case study, Sylvanus
Azupuri A. Kaba2 Sebbeh-Newton, Shaib Abdulazeez Shehu, Prosper
Hareyani Zabidi1 Ayawah, Azupuri A. Kaba & Hareyani Zabidi, Cogent
E-mail: [email protected] Engineering (2021), 8: 1869367.
1
School of Materials and Mineral Resources Engineering,
Universiti Sains Malaysia, 14300, Penang, Malaysia. References
2
Geological Engineering Department, Rock Mechanics and Backers, T. (2010). Applications of fracture mechanics
Explosives Research Building, Missouri University of numerical modelling in rock engineering. First Break,
Science and Technology, Rolla, MO 65409, USA. 28(3), 53–62 . 10.3997/1365-2397.2010008.

Page 18 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

Barton, N., Lien, R., & Lunde, J. (1974). Engineering mechanics and rock engineering. International
classification of rock masses for the design of Journal of Rock Mechanics and Mining Sciences, 40(3),
tunnel support. Rock Mechanics Felsmechanik 283–353. 10.1016/S1365-1609(03)00013-3.
Mécanique Des Roches, 6(4), 189–236. 10.1007/ Kabwe, E., Karakus, M., & Chanda, E. K. (2020). Proposed
BF01239496. solution for the ground reaction of non-circular tun­
Bieniawski, Z. T. (1989). Engineering rock mass classifica­ nels in an elastic-perfectly plastic rock mass.
tions: A complete manual for engineers and geolo­ Computers and Geotechnics, 119. 10.1016/j.
gists in mining, civil, and petroleum engineering. In compgeo.2019.103354.
Engineering Rock Mass Classifications: A Complete Kanik, M., & Gurocak, Z. (2018). Importance of numerical
Manual for Engineers and Geologists in Mining, Civil, analyses for determining support systems in tunnel­
and Petroleum Engineering. Wiley. ing: A comparative study from the
Cai, M., Kaiser, P. K., Tasaka, Y., & Minami, M. (2007). trabzon-gumushane tunnel, Turkey. Journal of
Determination of residual strength parameters of African Earth Sciences, 143, 253–265. 10.1016/j.
jointed rock masses using the GSI system determi­ jafrearsci.2018.03.032.
nation of residual strength parameters of jointed Kanik, M., Gurocak, Z., & Alemdag, S. (2015). A comparison
rock masses using the GSI system. (February). of support systems obtained from the RMR89 and
International Journal of Rock Mechanics and Mining RMR14 by numerical analyses: macka tunnel project,
Sciences, 44(2), 247–265. 10.1016/j. NE Turkey. Journal of African Earth Sciences, 109,
ijrmms.2006.07.005. 224–238. 10.1016/j.jafrearsci.2015.05.025.
Carranza-Torres, C., & Fairhurst, C. (2000). Application of Kaya, A., & Bulut, F. (2019). Geotechnical studies and
the convergence-confinement method of tunnel primary support design for a highway tunnel: A case
design to rock masses that satisfy the Hoek-brown study in Turkey. Bulletin of Engineering Geology and
failure criterion. Tunnelling and Underground Space the Environment, 78(8), 6311-6334. 10.1007/s10064-
Technology, 15(2), 187–213. 10.1016/S0886-7798(00) 019-01529-8.
00046-8. Kaya, A., & Sayın, A. (2019). Engineering geological
Celada, B., Tardáguila, I., Varona, P., Rodríguez, A., & appraisal and preliminary support design for the
Bieniawski, Z. T. (2014). Innovating Tunnel Design by salarha tunnel, Northeast Turkey. Bulletin of
an Improved Experience-based RMR System. Engineering Geology and the Environment, 78(2),
Proceedings of the World Tunnel Congress 2014 – 1095-1112. 10.1007/s10064-017-1177-2.
Tunnels for a Better Life, 3, 1–9 Kulatilake, P. H. S. W., & Shu, B. (2015). Prediction of rock
De La Fuente, M., Taherzadeh, R., Sulem, J., Nguyen, X. S., mass deformations in three dimensions for a part of
& Subrin, D. (2019). Applicability of the an open pit mine and comparison with field defor­
convergence-confinement method to full-face exca­ mation monitoring data. Geotechnical and Geological
vation of circular tunnels with stiff support system. Engineering, 33(6), 1551-1568. 10.1007/s10706-015-
Rock Mechanics and Rock Engineering, 52(7), 2361– 9921-5.
2376. 10.1007/s00603-018-1694-8. Li, C. C. (2017). Principles of rockbolting design. Journal of
Ghadimi Chermahini, A., & Tahghighi, H. (2019). Rock Mechanics and Geotechnical Engineering, 9(3),
Numerical finite element analysis of underground 396–414. 10.1016/j.jrmge.2017.04.002.
tunnel crossing an active reverse fault: A case study Ma, G., & Fu, G. (2014). A rational and realistic rock mass
on the sabzkouh segmental tunnel. Geomechanics modelling strategy for the stability analysis of blocky
and Geoengineering, 14(3), 155–166. 10.1080/ rock mass. Geomechanics and Geoengineering, 9(2),
17486025.2019.1573323. 113–123. 10.1080/17486025.2013.871067.
González-Cao, J., Alejano, L. R., Alonso, E., & Morris, J., & Johnson, S. (2009). Dynamic simulations of
Bastante, F. G. (2018). Convergence-confinement geological materials using combined FEM/DEM/SPH
curve analysis of excavation stress and strain analysis. Geomechanics and Geoengineering, 4(1),
resulting from blast-induced damage. Tunnelling and 91–101. 10.1080/17486020902767354.
Underground Space Technology. 73, 162–169. Palmstrøm, A. (1996). Characterizing rock masses by the
10.1016/j.tust.2017.12.005. RMi for use in practical rock engineering. Tunnelling
Hoek, E., & Brown, E. T. (1997). Practical estimates of rock and Underground Space Technology, 11(2), 175–188.
mass strength. International Journal of Rock 10.1016/0886-7798(96)00015-6.
Mechanics and Mining Sciences, 34(8), 1165–1186. Rehman, H., Naji, A. M., Ali, W., Junaid, M., Abdullah, R. A.,
10.1016/S1365-1609(97)80069-X. & Yoo, H. K. (2020). Numerical evaluation of new
Hoek, E., & Brown, E. T. (2019). The Hoek–brown failure Austrian tunneling method excavation sequences:
criterion and GSI – 2018 edition. Journal of Rock A case study. International Journal of Mining Science
Mechanics and Geotechnical Engineering, 11(3), 445– and Technology, 30(3), 381–386. 10.1016/j.
463. 10.1016/j.jrmge.2018.08.001. ijmst.2020.03.009.
Hoek, E., Carranza, C., & Corkum, B. (2002). Hoek-brown Sadeghiyan, R., Hashemi, M., & Moloudi, E. (2016).
failure criterion – 2002 edition. Narms-Tac. 267–273. Determination of longitudinal convergence profile
10.1016/0148-9062(74)91782-3. considering effect of soil strength parameters.
Hoek, E., & Diederichs, M. S. (2006). Empirical estimation International Journal of Rock Mechanics and Mining
of rock mass modulus. International Journal of Rock Sciences, 82, 10–21. 10.1016/j.ijrmms.2015.10.011.
Mechanics and Mining Sciences, 43(2), 203–215. Satici, O., & Ünver, B. (2015). Assessment of tunnel portal
10.1016/j.ijrmms.2005.06.005. stability at jointed rock mass: A comparative case
Hoek, E., & Marinos, P. (2000). Predicting tunnel squeezing study. Computers and Geotechnics, 64, 72–82.
problems in weak heterogeneous rock masses. 10.1016/j.compgeo.2014.11.002.
Tunnels and Tunnelling International. (pp. 1–20). Şen, Z., & Bahaaeldin, B. H. (2003). Modified rock mass
https://www.rocscience.com/documents/hoek/refer­ classification system by continuous rating. Engineering
ences/H2000d.pdf%0Ahttp://www.rockscience.com/ Geology, 67(3–4), 269–280. 10.1016/S0013-7952(02)
hoek/references/H2000d.pdf 00185-0.
Jing, L. (2003). A review of techniques, advances and Sheorey, P. R., Murali Mohan, G., & Sinha, A. (2001).
outstanding issues in numerical modelling for rock Influence of elastic constants on the horizontal in situ

Page 19 of 20
Sebbeh-Newton et al., Cogent Engineering (2021), 8: 1869367
https://doi.org/10.1080/23311916.2020.1869367

stress. International Journal of Rock Mechanics and tunnels in an underground mine in USA using
Mining Sciences, 38(8), 1211–1216. 10.1016/S1365- three-dimensional numerical modeling. Rock
1609(01)00069-7. Mechanics and Rock Engineering, 51(2), 579–597.
Ur Rehman, Z., Mohammad, N., Hussain, S., & Tahir, M. (2019). 10.1007/s00603-017-1336-6.
Numerical modeling for the engineering analysis of rock Xing, Y., Kulatilake, P. H. S. W., & Sandbak, L. A. (2019).
mass behaviour due to sequential enlargement of low­ Stability assessment and support design for under­
ari tunnel chitral khyber pakhtunkhwa, Pakistan. ground tunnels located in complex geologies and
International Journal of Geotechnical Engineering, 13(1), subjected to engineering activities: case study.
54–60. 10.1080/19386362.2017.1319615. International Journal of Geomechanics, 19(5),
Wyllie, D. (2018). Rock slope engineering: civil 05019004. 10.1061/(ASCE)GM.1943-5622.0001402.
applications. In CRC Press, Taylor & Francis Group, 13, Yalcin, E., Gurocak, Z., Ghabchi, R., & Zaman, M. (2016).
369–370. 10.2113/gseegeosci.13.4.369. Numerical analysis for a realistic support design:
Wyllie, D. C., & Mah, C. W. (2017). Rock slope engineering: case study of the komurhan tunnel in eastern Turkey.
Civil and mining, 4th edition. In Rock Slope International Journal of Geomechanics, 16(3),
Engineering: Fourth Edition, pp 1–432. CRC Press. 05015001. 10.1061/(ASCE)GM.1943-5622.0000564.
10.1201/9781315274980. Zhang, L. (2017). Evaluation of rock mass deformability
Xing, Y., Kulatilake, P. H. S. W., & Sandbak, L. A. (2017). using empirical methods – A review. In Underground
Rock mass stability investigation around tunnels in Space (China), 2(1), 1–15. 10.1016/j.
an underground mine in USA. Geotechnical and undsp.2017.03.003.
Geological Engineering, 35(1), 46–67. 10.1007/ Zhu, Z., Li, Y., Xie, J., & Liu, B. (2015). The effect of principal
s10706-016-0084-9. stress orientation on tunnel stability. Tunnelling and
Xing, Y., Kulatilake, P. H. S. W., & Sandbak, L. A. (2018). Underground Space Technology, 49, 279–286.
Investigation of rock mass stability around the 10.1016/j.tust.2015.05.009.

© 2021 The Author(s). This open access article is distributed under a Creative Commons Attribution (CC-BY) 4.0 license.
You are free to:
Share — copy and redistribute the material in any medium or format.
Adapt — remix, transform, and build upon the material for any purpose, even commercially.
The licensor cannot revoke these freedoms as long as you follow the license terms.
Under the following terms:
Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made.
You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
No additional restrictions
You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.

Cogent Engineering (ISSN: 2331-1916) is published by Cogent OA, part of Taylor & Francis Group.
Publishing with Cogent OA ensures:
• Immediate, universal access to your article on publication
• High visibility and discoverability via the Cogent OA website as well as Taylor & Francis Online
• Download and citation statistics for your article
• Rapid online publication
• Input from, and dialog with, expert editors and editorial boards
• Retention of full copyright of your article
• Guaranteed legacy preservation of your article
• Discounts and waivers for authors in developing regions
Submit your manuscript to a Cogent OA journal at www.CogentOA.com

Page 20 of 20