International Society for Rock Mechanics and Rock Engineering ISRM International Symposium

Norwegian Group for Rock Mechanics Eurock 2025 – Expanding the Underground Space
Trondheim, Norway, 16-20 June
ISBN 978-82-8208-079-8

A simplified analytical method for bolt reinforcement of

the tunnel face in deep conditions

W. Mohamad, A. Saitta & G. Rousselot

Egis Structures et Environnement – Pôle Tunnels & Ouvrages Souterrains, Annecy, France
[email protected] (email of corresponding author)

M. Janutolo Barlet & B. Lecomte

Vinci Construction Grands Projets, Saint-Martin de la Porte, France

Abstract
This study introduces a tunnel face bolt reinforcement technique for deep rock tunnels, based on the
quantification of the plastic rock mass volume ahead of the tunnel face using the convergence-
confinement method. A closed-form solution is derived for scenarios involving homogeneous ground
with uniform face reinforcement. The methodology incorporates the reinforcement scheme and
evaluates stability under the most critical conditions. The procedure outlines the minimum bolt length
requirements and the required bolt density at the tunnel face.

A case study is conducted on the TELT (Tunnel EURALPIN LYON TURIN) project, examining
stability challenges and reinforcement needs for the SMP4 segment, now part of the CO7 operational
construction site in Saint-Martin-La-Porte. In this context, a tunnel was excavated in squeezing
ground, including a fault zone, at an approximate depth of 550 m.

Keywords
Bolt reinforcement, tunnel face stability, analytical method
 Eurock 2025, Trondheim, Norway

1 Introduction
Tunnel face stability is a fundamental aspect in the design and construction of underground structures.
Ensuring the stability of the tunnel face is crucial to prevent collapses and ensure the safety of workers
and equipment, particularly in poor ground conditions subjected to high overload. In such scenarios,
the ground experiences significant inelastic straining, resulting in substantial tunnel face deformations
(Lunardi 2008).

One of the most effective methods to enhance tunnel face stability in conventional tunnel excavations
is through face reinforcement techniques, specifically the installation of fiberglass bolts. Various
closed-form solutions for face reinforcement have been established for shallow tunnels, considering a
failure mechanism involving a wedge at the face and an overlying prism extending to the soil surface
(Anagnostou et Perazzelli 2015). However, these methods exhibit limited accuracy and applicability in
deep tunnels due to the differing failure mechanisms (Georgiou et al. 2021).

This paper presents an analytical method for face reinforcement in deep tunnels, based on the
assessment of the unstable rock mass volume ahead of the tunnel face. The analysis is conducted
under the assumption of a homogeneous isotropic ground, without considering groundwater effects.
The failure mechanism is governed by the development of the plastic zone around the tunnel face,
which is investigated using the convergence-confinement method with the Mohr-Coulomb failure
criteria.

The proposed method is applied to a case study on the TELT (Tunnel EURALPIN LYON TURIN)
project. In this context, a tunnel was excavated in a squeezing rock mass, including a fault zone, at an
approximate depth of 550 m. Monitoring results, including extrusions and encountered face
instabilities, are provided, and face reinforcements are verified for different project sectors.

2 Tunnel face reinforcement method

2.1 Failure mechanism for shallow and deep tunnels
In conventional tunnel excavations, reinforcement of the tunnel face is typically achieved using
fiberglass dowels. These reinforcement elements limit tunnel face deformation due to their passive
behaviour, as they are anchored ahead of the tunnel face. The density of bolt reinforcement is usually
determined to counterbalance the equivalent pressure required to ensure rock mass stability at the
tunnel face. For shallow tunnels, this pressure is established using the limit equilibrium approach,
which considers a complete cave-in of the ground up to the surface level (Fig. 1a). The three-
dimensional visualization of the static system (Fig. 1b) includes the wedge-prism failure body, where
the prismatic body exerts a load upon the wedge (Anagnostou et Kovári 1994). Failure occurs if the
applied load exceeds the bearing capacity of the wedge (Anagnostou et Perazzelli 2015). The load
representing the prism is calculated based on Terzaghi's silo theory. This method is commonly used
for shallow tunnels in soft grounds. However, in very deep tunnels, Terzaghi's vertical stress is
generally negative, resulting in a stable prism and wedge without any support pressure, even in
weathered rock masses or fault zones.

(a) (b)

Fig. 1 (a) A 3D numerical example of the deformed grid at collapse in a case of a shallow tunnel (Fernández, et al. 2021) (b)
Tunnel face failure mechanism of a shallow tunnel (Anagnostou et Kovári 1994)

2
 Eurock 2025, Trondheim, Norway

In very deep rock masses, the failure mechanism of the tunnel face is a multifaceted process
influenced by various factors such as in-situ stress conditions, rock mass properties, and tunnel
geometry. Primarily, stress-induced failures such as spalling and rock bursting are common, where
tensile stresses exceed the rock's tensile strength, causing slab formation or violent ejection of rock
fragments. Shear failure mechanisms, including slabbing and wedge failure, occur along pre-existing
weaknesses or newly formed fractures, resulting in the detachment of rock slabs or the dislodging of
rock blocks. Additionally, squeezing ground presents a significant challenge due to the time-dependent
plastic deformation of the rock mass under high in-situ stresses. This phenomenon occurs when the
stress on the tunnel walls exceeds the rock mass's strength, causing the rock to deform plastically and
gradually converge into the tunnel.

In the case of homogeneous rock mass behaviour without significant discontinuities or anisotropies,
the deformation pattern can be characterized by a symmetric, dome-like inward bulging of the tunnel
face. This failure morphology typically arises in environments where isotropic stress conditions
prevail, resulting in relatively uniform in-situ stresses in all directions. Fig. 2a presents borehole
recordings at the tunnel face in a shale-dominated fault zone of an SMP4 segment (Mathieu, et al.
2023), highlighting the penetration rate up to 10 meters ahead of the tunnel face. An unstable volume,
referred to as "dead ground", can be observed in the form of a spherical cap. This volume can be
correlated to the form and size of the plastic zone developing ahead of the tunnel face in deep isotropic
circular tunnel
rock masses (Fig. 2b).
tangential stress
A preliminary evaluation of the development of this plastic zone can be conducted using the concept
of the stability number N, which was first introduced by (Broms et Bennemark 1967) to assess the
short-term stability of the tunnel face in undrained clayey ground. However, this concept was later
extended by the following expression (Panet et Sulem 2021). The plastic zone develops ahead of the
tunnel face when N > 5. The face is considered stable when N <= 5.
2𝑃0 2𝛾(𝐻 + 𝑟0 )
N= =
MC failure strength 𝜎𝑐𝑚 𝜎𝑐𝑚 (2)

𝑐𝑜𝑠𝜙
𝜎𝑐𝑚 = 2c
1 − 𝑠𝑖𝑛𝜙 (3)

Where P0 In-situ stress

σcm Rock mass strength
r0 Radius of the excavated section
γ Rock mass unit weight
c, 𝜙 Rock mass cohesion and friction angle
H Cover depth

Based on this evaluation, the proposed face reinforcement approach will, in a simplified manner,
consider the entire plastic zone ahead of the tunnel face as the failure rock mass volume. The driving
force to be supported by the face reinforcement correspond to the total weight of the plastic volume.

------ 7
------ 5
Penetration rate
m/min
------ 2

------ 0

(a) (b) Plastic zone

Fig. 2 (a) Unstable rock mass volume identified through borehole recordings in a Shale-dominated fault zone of an SMP4
segment (Mathieu, et al. 2023) (b) Plastic zone around a deep tunnel using a 3D numerical model

3
 Eurock 2025, Trondheim, Norway

2.2 Analytical formulation of the proposed method

The self-weight Fvp of the plastic volume ahead of the tunnel is the sole driving force to be considered
in the calculation of the face reinforcement. For simplification purposes, the plastic volume ahead of
the tunnel face Vp0 is assumed to be a spherical cap with a base radius equivalent to the radial plastic
radius at tunnel face rp0 and a height h (Fig. 3a). The latter corresponds to the horizontal extent of the
plastic zone ahead of the tunnel face. It ranges between 50 and 75 % of the plastic radius at tunnel face
(Cantieni et al., 2011). We will consider the maximum ratio in this article, i.e. h = 0.75rp0. Therefore,
the corresponding self-weight Fvp can be estimated using to the following equation. The plastic radius
at tunnel face is determined using the convergence-confinement method.
1 57
𝐹𝑣𝑝 = 𝛾𝑉𝑝0 = 𝛾 𝜋ℎ(3𝑟𝑝0 2 + ℎ2 ) = 𝛾𝜋𝑟𝑝0 3
6 128 (4)

The supporting force of a single bolt is determined based on several factors, including the material
properties of the bolt, the quality of the surrounding rock, the length and diameter of the bolt and the
installation pattern. The bolts are typically inserted into boreholes and then anchored in place using
grout or mechanical means. The calculation typically involves determining the bolt’s tensile strength,
the bond strength at the bolt-grout interface and the bond strength between the grout and the
surrounding rock. The design of face reinforcement should be conducted considering the most critical
bolt layout. The set of bolts with an initial length L is generally installed at regular intervals l
(installation interval), resulting in an effective bolt length L’ at the tunnel face that varies with each
excavation step. The most unfavourable effective length to be considered in the calculation is L’ = L-l.

Fig. 3b illustrates a simplified load transfer curve for a single bolt installed at the tunnel face. The bolt
accommodates its load along a distance dpl before transferring it to the rock mass over the anchoring
length La. The anchoring length and the load accommodating distance vary over the height of the
tunnel face depending on the extent of the plastic zone in the vicinity of each bolt. For computational
simplicity, a uniform distance dpl is considered, derived from the equivalent cylindrical volume of the
plastic zone. The volume of this cylinder is given by dpl(πrp02), leading to a distance dpl that can be
expressed according to Eq. 5. The initial bolt length L and the installation interval l should be selected
to ensure a minimal effective length that satisfies the condition L’ > 57 rp0/128.

57
𝑉𝑝0 𝜋𝑟 3
𝑑𝑝𝑙 = = 128 𝑝0 = 57 𝑟 (5)
𝜋𝑟𝑝0 2 𝜋𝑟𝑝0 2 128 𝑝0
Based on the previous assumptions, the supporting force of each bolt (modified from (Anagnostou et
Perazzelli 2015)) reads as follows:

𝐹𝑏 = min [𝐹𝑡 , 𝜋𝑑𝑝𝑙 min(𝑑𝜏𝑚 , 𝑑𝑏 𝜏𝑔 ) , 𝜋𝐿𝑎 𝑚𝑖𝑛 (𝑑𝜏𝑚 , 𝑑𝑏 𝜏𝑔 )] (6)

Where Ft Bolt tensile strength

d, db Borehole and bolt diameter, respectively
τm Shear strength of the grout-rock interface, assumed constant along the entire bolt
τg Shear strength of the grout-bolt interface, assumed constant along the entire bolt
La Anchoring length of the bolt = L’ - dpl

Eq. 6 accounts for the tensile failure of the bolt as well as the shear failure at the various interfaces
both within and outside the plastic zone. The number of bolts n required to ensure tunnel face stability
is determined by dividing the total weight of the plastic zone Fvp by the supporting force of a single
bolt Fb, so that n = Fvp / Fb. The required bolt density is given by n/πr02.

3 Determination of the plastic radius at tunnel face

The plastic radius is analysed through the application of the convergence-confinement method,
employing the Mohr-Coulomb failure criterion as specified below:
𝜎1 = 𝜎𝑐𝑚 + 𝑘𝜎3 (7)
Where σ1, σ3 Maximum and minimum principal stress, respectively

4
 Eurock 2025, Trondheim, Norway

Plastic zone
Equivalent cylindrical plastic volume at tunnel face
(a)
rp0 rp bolts
L’

Tunnel r0
h

La

Spherical cap plastic volume at tunnel face dpl

(b) Force transfer to the stable rock mass Force accumulating zone near the tunnel face

Bolt Force
Tunnel face
La dpl
L’
Point of zero shear stress

Fig. 3 (a) Unstable rock mass volume at tunnel face (b) Simplified load transfer curve for a single bolt

The slope k of the σ1 - σ3 curve is defined by:

1 + 𝑠𝑖𝑛𝜙
𝑘= (8)
1 − 𝑠𝑖𝑛𝜙
We consider a circular tunnel of radius r0 subjected to a hydrostatic stress field P0 and a fictive support
pressure Pi as illustrated in Fig. 4. Failure of the rock mass occurs when the fictive pressure Pi falls
below a critical pressure Pcr defined as follows (Hoek 2023):
2𝑃0 − 𝜎𝑐𝑚
𝑃𝑐𝑟 = (9)
1+𝑘
The behaviour of the rock mass is elastic when the fictive pressure is greater than the critical pressure
(Pi > Pcr). In this case, the radial displacement ui is given by:
𝑟0 (1 + 𝜈)
𝑢𝑖𝑒 = (𝑃0 − 𝑃𝑖 ) (10)
𝐸
Where E, ν Rock mass Young modulus and Poisson ratio, respectively.

When failure occurs (Pi < Pcr), the plastic zone starts to develop around the circular tunnel with a
corresponding plastic radius rp that expresses as follows:
1
2(𝑃0 (𝑘 − 1) + 𝜎𝑐𝑚 ) 𝑘−1
(11)
𝑟𝑝 = 𝑟0 [ ]
(1 + 𝑘)((𝑘 − 1)𝑃𝑖 + 𝜎𝑐𝑚 )

The radial displacement in the plastic zone is given by:

𝑟0 (1 + 𝜈) 𝑟𝑝 2
𝑢𝑖𝑝 = [2(1 − 𝜈)(𝑃0 − 𝑃𝑐𝑟 ) ( ) − (1 − 2𝜈)(𝑃0 − 𝑃𝑖 )] (12)
𝐸 𝑟0
The plastic radius at tunnel face is correlated to the radial displacement at tunnel face. Considering the
longitudinal displacement profile develepped by (Vlachopoulos et Diederichs 2009), the radial
displacement at tunnel face u0 is determined by the following equation :
𝑢𝑚 −0.15(𝑟 ⁄𝑟 )
𝑢0 = 𝑒 𝑝𝑚 0 (13)
3

Where rpm Maximum plastic radius, calculated by Eq. 11 for Pi = 0

um Maximum radial displacement, calculated by Eq. 12 for Pi = 0

The calculated radial displacement at tunnel face u0 allows for the determination of the corresponding
fictive pressure using Eq. 12. This pressure can then be used to determine the plastic radius at tunnel

5
 Eurock 2025, Trondheim, Norway

face rp0 using Eq. 11. To simplify the calculation, a direct relationship between rp0 and u0 is given
below (deduced through a series of calculations for different geomechanical rock mass properties):
𝑟𝑝0 𝑢0
= 1.24( )0.59
𝑟𝑝𝑚 𝑢𝑚 (14)

The determination of the plastic radius using Eq. 14 is applicable when the stability number N > 5. If
the equation gives rp0 < r0, we consider rp0 = r0. In such cases, the face reinforcement is considered
"contingent," meaning it is installed only if required due to excavation difficulties.

Plastic zone rpi rpm

Elastic zone rp0 Plastic zone
u0
ui um

Fig. 4 Convergence-confinement method (modified from (Hoek 2023))

4 Saint-Martin-La-Porte case study

The Lyon-Turin railway project involves constructing a 67 km line linking Saint-Jean-de-Maurienne
(France) to Susa (Italy), including a 57 km twin-tunnel called the Mont Cenis Base Tunnel. An
exploratory tunnel (SMP4-3B) was built using conventional methods along the southern tube axis
(Fig. 5) of the base tunnel (Janin, et al. 2022), to study rock mass behaviour and construction
feasibility. The sector comprises sandstones and carbonaceous shale with a 550 m tunnel cover.

The support system has two phases (Fig. 5b): Phase A (top heading excavation) includes eventual
Forepoling, face bolting, shotcrete, radial bolting and sliding ribs. Phase B (benching down
excavation) comprises a thicker shotcrete layer and yielding blocks. Monitoring included face
extrusions and radial convergences, with observed deformations between 0.5% and 10%. A back-
analysis was performed to assess the rock mass properties along the southern tube. The short-term
Mohr-Coulomb characteristics are presented in Table 1. Two classes (4 and 6) are excluded from the
analysis as they are not represented within this sector of the project.

Radial bolts

Forepoling
Shotcrete

Shotcrete

Yielding blocs Face bolts

Excavated
southern tube
SMP1 Sliding ribs

(a) (b) Phase B Phase A 5-30 m

Fig. 5 (a) Saint-Martin-La-Porte work layout (b) SMP4-3B systematic support system
Table 1 Short-term characteristics of the southern tube (excluding class 4 and 6)

Colour code Young modulus E (MPa) Cohesion c (MPa) Friction angle 𝛟

Class 0 7600 2 37
Class 1 2530 1.63 32.2
Class 2 1760 1.06 31.7
Class 3 1530 0.82 30
Class 5 1100 0.42 24.5

6
 Eurock 2025, Trondheim, Norway

4.1 Extrusions and tunnel face instabilities

The target excavated section “large section” has a tunnel radius ranging from 6 to 6.5 m. However,
under extremely squeezing conditions, initial excavation necessitated a reduced section (R = 4.4 m) or
a small section (R = 3.15 m) in fault zones. Extrusions along the southern tube varied between 0.2%
and nearly 10%, as illustrated in Fig. 6. Face reinforcement comprised 30 fiberglass dowels (L = 12 m,
diameter = 28 mm, tensile strength = 380 kN) installed every 4-6 m in the red fault zone (class 5) with
a small section, and 50 bolts installed every 2-6 m in other zones with a large section. This
reinforcement pattern proved sufficient for classes 0 to 2 and largely effective for class 5. However,
tunnel instability was observed in class 3 sections (collapse of the tunnel face), and shotcrete cracks (a
single incident at chainage 10655 m) were noted in the fault zone. Consequently, additional face bolts
(up to 20) were installed, particularly between chainage 10950 and 11200 m in the class 3 zone.

Fig. 6 Extrusions and tunnel face instabilities along the southern tube

4.2 Reinforcement verification using the proposed approach

The required bolt density will be studied for selected zones using the new simplified approach. The
rock mass unit weight γ = 27 kN/m3 and Poisson’s ration ν = 0.3. The calculation is performed under
the assumption of a full-face excavation. The resulting density will be used to determine the number of
bolts required exclusively for the top heading. The procedure involves the following steps:

 Stability number N. If N > 5, the face is considered unstable.

 Maximum plastic radius and radial displacement using Eq. 11 and 12 for Pi = 0, respectively,
 Tunnel face displacement using Eq. 13 and tunnel face plastic radius using Eq.14,
 Plastic volume total weight using Eq. 4.
 Supporting force of each bolt using Eq. 6, we consider a more critical shear failure possibility at
grout-bolt interface with a shear stress of 1 MPa.
 Bolt density = n/πr02 = (Fvp / Fb)/ πr02.
 Number of bolts for the top heading: (bolt density) x (excavated section of the top heading).

For a rock mass of Class 2, the required number of bolts was determined to be 34 (50 bolts were
installed), as shown in Table 2. The zone of class 3 between chainage 10950 and 11200 m required at
least 60 bolts, confirming the tunnel face instabilities and the need for additional face bolts in this area.
The fault zone necessitated 37 bolts according to the calculations, which is close to the installed value
of 30. This zone also included Forepoling installed every 1-2 m, which is not accounted for in the
proposed method. In some fault areas, fiberglass dowels were replaced by steel bars to enhance shear
strength at the bolt-rock interface.
Table 2 Face reinforcement for different sections of SMP4-3B

Chainage (m) 11320-11350 10950-11200 10630-10800

In-situ stress P0 (MPa) 15 15 15
Rock mass class 2 3 5
Stability number N 7.9 10.6 23
Tunnel radius r0 (m) 6.05 6.05 3.15
Total Excavated section St = πr02 (m2) 115 115 31
Excavated section of Phase A (m2) 93 93 27
Minimum effective bolt length L’ (m) 6 6 6

7
 Eurock 2025, Trondheim, Norway

Installed number of bolts for top heading 50 50 30

Max radial displacement um (mm) 193 311 870
Max plastic radius rpm (m) 12.1 14.6 16.1
Tunnel face radial displacement u0 (mm) 48 72 135
Tunnel face plastic radius rp0 (m) 6.6 7.6 6.6
Plastic volume weight Fvp (kN) 10 690 16 765 11 041
Corresponding pressure Fvp / St (kPa) 90 150 360
Bolt supporting force Fb (kN) 257 229 260
Required bolt density (Fvp / Fb)/ πr02 0.36 0.635 1.35
Required number of bolts for top heading 34 60 37

5 Conclusions and perspectives

A new simplified analytical method is presented for assessing face reinforcement in deep tunnels
within a homogeneous isotropic rock mass, excluding groundwater effects. This method employs the
convergence-confinement approach to determine the plastic radius at the tunnel face, which is then
used to estimate the plastic volume ahead of the tunnel face. The face reinforcement pattern is
incorporated to calculate the supporting bolt pressure. An application of this method is demonstrated
in the TELT project, where a tunnel is excavated in squeezing ground conditions, posing challenges to
tunnel face stability. The method shows good agreement with the reinforcement scheme implemented
in the project. The method can be extended to account for the support installed behind the tunnel face,
which directly impacts the extent of the plastic zone ahead of the tunnel face. This face stability
approach will be tested on the TELT CO7 works currently being excavated in the squeezing zone at
Saint-Martin-la-Porte.

6 Acknowledgments
We are grateful to TELT, the project owner responsible for the works discussed in this communication
and the S2IP consortium, the project manager of the site.

7 References
Anagnostou, G, and K Kovári. (1994). The face stability of slurry-shield-driven tunnels. Tunnelling and
Underground Space Technology: 165-174.
Anagnostou, G., and P. Perazzelli. (2015). Analysis method and design charts for bolt reinforcement of the
tunnel face in cohesive-frictional soils. Tunnelling and Underground Space Technology (47) 162-181.
Broms, B. B., and H. Bennemark. (1967). Stability of clay at vertical openings. Journal of the Soil Mechanics
and Foundations Division, ASCE (93): 71-94.
Cantieni, L., G. Anagnostou, and R. Hug. (2011). Interpretation of Core Extrusion Measurements When
Tunnelling Through Squeezing Ground. Rock Mech Rock Eng (44): 641-670.
Fernández, F., J. E. Rojas, E. A. Vargas Jr, R. Q. Velloso, and D. Dias. (2021). Three-dimensional face stability
analysis of shallow tunnels using Numerical Limit Analysis and Material Point Method. Tunnelling and
Underground Space Technology.
Georgiou, D., M. Kavvadas, and A. Kalos. (2021). Numerical investigation of the stability of tunnel excavation
faces in deep tunnels. Rockscience International Conference.
Hoek, E. 2023. Tunnels in weak rock. In Practical Rock Engineering.
Janin, JP., T. Rossi, A. Silvestre, F. Laigle, and S. Lione. (2022). Back-analysis on Saint Martin la Porte works
of Lyon-Turin base tunnel. ITA-AITES World Tunnel Congress, WTC2022 and 47th General
Assembly. Copenhagen. 22-28.
Lunardi, P. (2008). Design and construction of tunnels: Analysis of Controlled Deformations in Rock and Soils
(ADECO-RS). Springer Science & Business Media.
Mathieu, E., F. Martin, S. Festa, and S. Pelizza. (2023). New Lyon-Turin Transalpine Rail Link – SMP4
Worksite: A successful experimental crossing of the productive coal deposit. TUNNELS ET ESPACE
SOUTERRAIN - N° 286 - Octobre/Novembre/Décembre.
Panet, M., and J. Sulem. 2021. Le calcul des tunnels par la méthode convergence-confinement. Presses de
l’École Nationale des Ponts et Chaussées.
Vlachopoulos, N., and M. S. Diederichs. (2009). Improved Longitudinal Displacement Profiles for Convergence
Confinement Analysis of Deep Tunnels. Rock Mechanics and Rock Engineering (42): 131-146.