SECTION GEOTECHNICS & UNDERGROUND STRUCTURES VOLUME: 18 | NUMBER: 1 | 2018 | JUNE

Assessment of Static and Dynamic Stresses

Horse-Shoe Tunnel with Connecting Gallery
using Finite Element Method: (As a Case Study)

Jaafar MOHAMMED 1 , Eva HRUBESOVA 2

1
Department of Geotechnics and Underground Engineering, Faculty of Civil Engineering,
VSB - Technical University of Ostrava, Ludvíka Podeste 1875/17, Ostrava - Poruba, Czech Republic
2
Department of Geotechnics and Underground Engineering, Faculty of Civil Engineering,
VSB - Technical University of Ostrava, Ludvíka Podeste 1875/17, Ostrava - Poruba, Czech Republic

[email protected], [email protected]

DOI: 10.31490/tces-2018-0004

Abstract. This paper illustrates the use of MIDAS terms of displacement and stresses acting on the tunnel
GTS NX to investigate the tunnel’s simulation in or- and rock mass.
der to highlight the effect of static and dynamic load
The materials and properties of tunnels and rock
on the behaviour of tunnel to compare this response ex-
mass were taken from the MIDAS GTS tutorials with-
pressed in terms of displacement and stresses acting on
out the experimental testing like tensile strength as
the tunnel and ground. Main tunnel is located in the
a case study, the Design Response Spectrum of UBC
ground having uniform property throughout its extent,
(1997) is mentioned in this manual also.Due to the
connecting gallery is located perpendicular to the main
application of the static load the stress-strain state
tunnel, the shotcrete and rock bolts for each tunnel will
around the tunnel periphery is changed, otherwise the
be installed.
result show that the applied dynamic stress is not negli-
gible for underground structure, but it is less dangerous
in comparison with the others.
Keywords The main application of the numerical methods in
the field of the tunnel engineering is to analysis of the
Tunnel, Numerical model, Static load, Dynamic stress, strain and deformations [4].Response spectrum
load, Earthquake, Displacement. analysis expresses the natural period, natural angu-
lar frequency or natural frequency at the maximum
physical quantity response as a function when a dy-
1. Introduction namic load is applied to the structure [1]. Shallow
tunnels through fractured zones exposed to increased
pressure, resulting in damage to the lining and defor-
According to recent studies and observed failures of un- mation. Tunnels crossing fault zones are exposed to
derground structures many researchers had been ad- increased stresses during earthquake loading, leading
dressed the design and construction of tunnel lining to complicated cracking and large differential displace-
against static/dynamic loads and earthquake vibration ments [12].
to get the best design use different methods include
empirical methods, analytical methods and numerical The dynamic responses of a buried tunnel in general,
methods. The 3D numerical models using finite ele- and its seismic responses, in particular, are of much
ment method software MIDAS GTS NX investigation interest [2].Although underground structures such as
will be carried out in order to highlight the effect of tunnels are more resistant to earthquakes than surface
the tunnel lining response under static and dynamic. structures [5].Finite element methods allow the compu-
tation of ground displacement at every point within the
The 3D numerical analysis prepared to simulate the ground. The geometry, initial conditions, excavation
static and dynamic behavior of a tunnel and connecting stages and ground behavior can be properly modeled
gallery, were undertaken to investigate the tunnel seis- using the numerical approach [3].
mic response and to compare this response expressed in

c 2018 TRANSACTIONS OF THE VSB-TECHNICAL UNIVERSITY OF OSTRAVA CIVIL ENGINEERING SERIES 18
SECTION GEOTECHNICS & UNDERGROUND STRUCTURES VOLUME: 18 | NUMBER: 1 | 2018 | JUNE

Stress and displacement balance of forces around the

tunnel are adversely affected and due to redistribu-
tion of these forces that create undue concentration
in some areas, instability occurs in the tunnel. Their
research also concludes that increasing the stiffness of
the support system can increase the effect of the seis-
mic loads [8].
The pressure on the lining of an underground space
or a tunnel increases with time. This is due to the
time-dependent mechanical behavior of the surround-
ing rocks or soil [6].The maximum change of internal
forces during an earthquake occurs in shear force, then
bending moment. Moreover, the normal force of tunnel Fig. 1: Tunnel profile showing the geometry of the model [1]
lining is less affected by seismic actions [13].
It is very likely that not only the shotcrete is af-
fected by deterioration. Also the contact between the
ground and the shotcrete shell and the contact between
the shotcrete shell and the inner liner can be affected.
Further, the properties of the ground around the open-
ing may change due to weathering [7].While the gen-
eral public is often skeptical about the performance
of underground structures, tunnel designers know that
underground structures are among the safest shelters
during earthquakes, based primarily on damage data
reported in the past [9], [10] and [11].

Fig. 2: Modified Response Spectrum using UBC (1997); Damp-

2. Conditions of Surrounding ing Ration = 0.05; Seismic Coefficient: Ca = 0.06 Cv =
0.06; Normalized Acceleration [1]
Ground

Adam McIntyre [?] clears up in short the condition of

surrounding ground due to seismic condition listed as: 3. Difination of Ground and
Structural Materials
• Seismic waves propagate faster in hard/dense ma-
terials.
This paper studies the 3D model with gravity direc-
– Less energy will be released at places where tion in Z direction. Design response spectrum of UBC
the ground is harder than the tunnel struc- (1997) is used as seismic response spectrum. The di-
ture, mension of the model are (x = 64, y = 60, z = 80) m.
After tunnel model is performed in the MIDAS GTS
– tunnels in harder ground will tend to deform NX (geometry, input data, mesh, etc.) the model is
with the ground and suffer less damage, run to analyze the tunnel stability and deformation
– conversely, tunnel structures in weaker under static and dynamic conditions by calculated the
soil/rock will absorb larger amounts of seis- value of each mesh node based on 3D finite element
mic energy and suffer greater damage. method to simulate the effect of earthquake on tunnel
and ground. Tab. 1 and Tab. 2 shows the ground ma-
• Disturbance to the plastic zone between the tun- terials, shotcrete and rock bolt, the material property
nel wall and reinforcement/lining structure could of shotcrete and rock bolt and the section property of
cause excessive vibration due to passing seismic shotcrete and rock bolt respectively.
waves.

– Disturbance caused by cave-in/collapse dur-

ing tunneling, squeezing ground, inclined
stresses, etc.

c 2018 TRANSACTIONS OF THE VSB-TECHNICAL UNIVERSITY OF OSTRAVA CIVIL ENGINEERING SERIES 19
SECTION GEOTECHNICS & UNDERGROUND STRUCTURES VOLUME: 18 | NUMBER: 1 | 2018 | JUNE

Name Hard Rock

Material model Mohr- Coulomb
Modulus of Elasticity (E) [kPa] 5,000,000
Poisson’s Ratio (ν) [-] 0.2
Unit Weight (γ) [kN/m3 ] 25
Unit Weight Saturated (γ) [kN/m3 ] 25
Cohesion (c) [kPa] 300
Friction Angle (ϕ) [◦ ] 40
Tensile Strength [kPa] 2,942
Coefficient of Lateral Pressure (Ko ) [-] 0.25

Tab. 1: Rock MassMaterials

Name Shotcrete Rock Bolt

Geometrical type Plane Line
Element Type Shell (2D element) Line (1D element)
Modulus of Elasticity (E) [kPa] 15,000,000 200,000,000
Poisson’s Ratio (ν) [-] 0.2 0.3
Weight Density (γ) [kN/m3 ] 24 77
Thickness [m] 0.16 -
Radius [m] - 0.025

Tab. 2: The material property of shotcrete and rock bolt of both main tunnel and connecting gallery

4. Simulation and Calibration

of the Numerical Model

The tunnel model consists of two parts: one is the main

tunnel running through the ground along Y - Axis and
the other part is the gallery which has a short exten-
sion along X - Axis and its located perpendicular to the
main tunnel. The depth of main tunnel under ground
surface is (z = 36.8 m) and located in the ground hav- Fig. 3: Static: Distribution of total displacement (including
ing uniform property throughout its extent. In each max. and min. value)in 2D and 3D of rock mass
tunnel, shotcrete and rock bolts will be installed. Con-
struction stage will be defined at each stage of excava-
tion and the process of construction stage of connecting ometry and the acceleration of gravity it can be easily
gallery start after the end of main tunnel construction set by inputting a scale factor of direction. The to-
stages. The type of lining for both the main tunnel tal length of the main tunnel along y direction is 60
and gallery are shotcrete. The rock bolts method used meters. The excavation of the tunnel is executed in
to support tunnels which are connected to the ground. full face excavation. The construction stages have 66
The rock bolts and shotcrete of the main tunnel will stages. The considered section length of tunnel exca-
be installed right after the excavation. The subsoil has vated in a single cycle is assumed as 2 m. In 3D model
been meshed with tetrahedral solid elements and the analysis, displacement constraints are applied in x di-
tunnel mesh consists of curved shell elements of uni- rection for left/right, y direction for front/back, xyz
form thickness 0.16 m. The rock bolts composed of direction for the bottom part as shown in Figs. 3 -
1D reinforced bar elements of uniform cross - section 7. The initial stage of model corresponds to the self
area (0.001963) m2 . The Eigenvalue analysis reveals - weight only and the ground boundary conditions are
7th and 15th to be the most influential modes for the activated. GTS NX automatically recognizes and sets
vertical vibration of the tunnel - rock mass system. the boundary conditions for the model. The maximum
Subsequently, a response spectrum analysis has been and minimum value of Max.principal stresses and Min.
performed with an input of a design response spectrum principal stresses (σ1 and σ2 ) respectively are shown in
applied in horizontal X direction as shown in Fig. 2. Figs. 8 - 11, for both static and dynamic result analysis
The subsoil is assumed to have a damping ratio of 0.05. and are increased at dynamic analysis than static.
Gravity is calculated automatically by multiplying the
inputted unit weight of the ground, the structure ge-

c 2018 TRANSACTIONS OF THE VSB-TECHNICAL UNIVERSITY OF OSTRAVA CIVIL ENGINEERING SERIES 20
SECTION GEOTECHNICS & UNDERGROUND STRUCTURES VOLUME: 18 | NUMBER: 1 | 2018 | JUNE

Fig. 9: The Max. and Min. values of minimum principal

Fig. 4: Dynamic: The Max. and Min. value of total displace- stresses of static and dynamic analysis
ment in 2D and 3D of rock mass

Fig. 5: Static analysis: Displacement distribution of rock mass

in direction X, Y and Z
Fig. 10: The Max. and Min. value of maximum principal
stresses (vector) distribution of static and dynamic
analysis

5. Conclusion
After tunnel model is created in MIDAS GTS NX, the
model is run to analyze the tunnel stability and defor-
Fig. 6: Static analysis: Displacement distribution of rock mass mation in static and dynamic conditions by calculated
in direction X, Y and Z the value of each mesh node based on 3D finite ele-
ment method to simulate the effect of earthquake on
tunnel and ground. From the results of model analy-
sis, it can be noticed that three dimensional analyses
which simulates the progress of works, as mentioned in
Tab. 3 and Tab. 4, and Figs. 3 - 5, above shows the
result of the maximum and minimum values. In this
study, the following conclusions are drawn: Figs. 3 - 7,
shows the displacement in static and dynamic condi-
tions. The maximum and minimum principle stresses
as shown in Figs. 8 - 11 for both static and dynamic
Fig. 7: The total displacement in static and dynamic analysis analysis. Fig. 12, in static analysis show the stresses
at TZ Translation(V)

Fig. 11: The Max. and Min. value of minimum principal

Fig. 8: The Max. and Min. values of maximum principal stresses (vector) distribution of static and dynamic
stresses of dynamic and static analysis

c 2018 TRANSACTIONS OF THE VSB-TECHNICAL UNIVERSITY OF OSTRAVA CIVIL ENGINEERING SERIES 21
SECTION GEOTECHNICS & UNDERGROUND STRUCTURES VOLUME: 18 | NUMBER: 1 | 2018 | JUNE

Displacement [m] Max. Principle stresses [kPa] Min. Principle stresses [kPa]
Max. value Max. value Max. value
Static 0.002397 236.037 - 61.8309
Min. value Min. value Min. value
0.0 - 626.886 - 4169.67

Tab. 3: The result in maximum and minimum values in static analysis

Displacement [m] Max. Principle stresses [kPa] Min. Principle stresses [kPa]
Max. value Max. value Max. value
Dynamic 0.000389 114.431 119.189
Min. value Min. value Min. value
0.0 6.74168 10.2978

Tab. 4: The result in maximum and minimum values in dynamic analysis

distribution on shotcrete and the axial force distribu-

tion on shotcrete are shown in Fig. 14, for static anal-
ysis. Fig. 13, in static analysis shows the axial forces
distribution in the Rock Bolts. Due to the application
of the static load the stress - strain state ground the
tunnel periphery is changed, the primary stress state is
disrupted and the potential of instability increases, oth-
erwise the result show that the applied dynamic stress
Fig. 12: Static analysis: Stresses distribution on shotcrete is not negligible for underground structure.

Acknowledgment
The paper was prepared with the support of the Com-
petence Centers of the Technology Agency of the Czech
Republic (TACR) within the project Center for Effec-
tive and sustainable transport infrastructure (CESTI),
project number TE01020168.

References
[1] MIDAS. GTS NX On-line Manual [cit. 02.01.2017]
[EN] http://manual.midasuser.com/en_
Fig. 13: Static: Axial forces distribution in the Rock Bolts common/GTS

[2] XIAOQING, W., XIANLONG, J., PUYONG,

W. and Y. ZHIHAO. Parallel Dynamic Analy-
sis of a Large - Scale Water Conveyance Tun-
nel under Seismic Excitation Using ALE Finite-
Element Method. Received: 15 December 2015;
Accepted: 22 January 2016; Published: 28
January 2016. DOI: www.mdpi.com/journal/
applsci10.3390/app6020036.

[3] RALUCA, A., MOLDOVAN, A. and A. POPA.

The influence of different tunnel cross sections
Fig. 14: Static and dynamic: Axial forces distribution in the on surface settlement. Acta Technica Napocen-
shotcrete sis: Civil Engineering & Architecture. 2013,
vol. 56. http://constructii.utcluj.ro/
ActaCivilEng

c 2018 TRANSACTIONS OF THE VSB-TECHNICAL UNIVERSITY OF OSTRAVA CIVIL ENGINEERING SERIES 22
SECTION GEOTECHNICS & UNDERGROUND STRUCTURES VOLUME: 18 | NUMBER: 1 | 2018 | JUNE

[4] RANA M., SAJID, I. and S. KHAN/KABINDRA

MAN. Stability Analysis of Shallow Tunnel of Nor-
raLänken. OMaster of Science Thesis. Royal Insti-
tute of Technology (KTH), 2011, ISNN 1652-599X.

[5] TSERING T. The Impact of Earthquakes on Tun-

nels in different Rock Mass Quality Q: A numeri-
cal analysis Master Thesis in Geosciences. Univer-
sity of Oslo,2001. http://www.duo.uio.no

[6] KHOSHBORESH RAHIM A. A Study on Defor-

mation of Tunnels Excavated in Fractured Rocks.
Master of Science (MSc). Quebec, Canada, 2013.
[7] MÖDLHAMMER H. Numerical Methods for Tun-
neling using ABAQUS and Investigation of Long-
Time-Effects of the Shotcrete Shell and its Impact
on the Combined Support System. Master Thesis.
Montan universit ät Leoben/Österreich, 2010.
[8] HOSSEINI, N., ORAEE, K. and M. GHOLINE-
JAD. Seismic analysis of horseshoe tunnels under
dynamic loads due to earthquakes, in Aziz, N (ed).
In: 10th Underground Coal Operators’ Confer-
ence. University of Wollongong & the Australasian
Institute of Mining and Metallurgy, 2010, 140-145
[9] HOEK, E., CARRANZA - TORRES, C. and
B. CORCUM. Hoek-Brown failure criterion -
2002 Edition. In: NARMS-TAC Conference.
Toronto, 2002, 1, 267-273. Available at: www.
rocscience.com.
[10] YOUSSEF, M., HASHASH, A., HOOK JEF-
FREY, J., SCHMIDT, B. and J. I-CHIANG YAO.
Seismic design and analysis of underground struc-
tures. Tunnelling and Underground Space Tech-
nology. 16, 247-293, 2001, PII: S 0 8 8 6 - 7 7
9 8 Z 0 1. 0 0 0 5 1 - 7.
[11] AKHLAGHI, T. and A. NIKKAR. Effect of Ver-
tically Propagating Shear Waves on Seismic Be-
havior of Circular Tunnels. Hindawi Publishing
Corporation. Scientific World Journal. 2014, Ar-
ticle ID 806092, 10 pages. Received 8 August
2013; Accepted 19 October 2013; Published 30
January 2014. DOI: http://dx.doi.org/10.
1155/2014/806092.
[12] McINTYRE A. Effects of earthquakes on tunnels.
EOSC 547: Tunneling and Underground Design
[13] HAMDY, H.A., ABD-EL, R., ENIEB, M.,
KHALIL, A.A. and S.H.A.ABDOU. Seismic Anal-
ysis of Urban Tunnel Systems for the Greater
Cairo Metro Line. 2015, No.4. EJGE Vol. 20,
Bund. 10.

c 2018 TRANSACTIONS OF THE VSB-TECHNICAL UNIVERSITY OF OSTRAVA CIVIL ENGINEERING SERIES 23