Tunnelling and Underground Space Technology 76 (2018) 177–193

Contents lists available at ScienceDirect

Tunnelling and Underground Space Technology

journal homepage: www.elsevier.com/locate/tust

Response of urban single and twin circular tunnels subjected to transversal T

ground seismic shaking
Grigorios Tsinidis1
University of Sannio, Department of Engineering, Piazza Roma, 21, 82100 Benevento, Italy

A R T I C LE I N FO A B S T R A C T

Keywords: The response of tunnels subjected to ground seismic shaking is commonly evaluated, disregarding the existence
Tunnel-soil-buildings interaction of other structures in the near area. However, significant interaction phenomena may take place between the
Dynamic FEM analysis tunnel and existing structures in dense urban environments, altering the seismic response of the tunnel compared
Dynamic earth pressures to the one predicted for ‘green field’ conditions. This study aims at investigating the effect of heavy buildings on
Dynamic shear stresses
the response of urban single or twin circular tunnels, when subjected to severe ground seismic shaking in the
Dynamic lining forces
transversal direction. For this purpose, a numerical parametric study was conducted on representative tunnel-
soil-buildings systems, employing the finite element code ABAQUS. Several configurations were examined, in-
cluding single or twin circular tunnels, embedded in clayey soil deposits. The high-rise buildings were simulated
as equivalent single degree of freedom oscillators (SDOFs), founded on rigid raft foundations. Salient parameters
that affect the seismic response of tunnels, namely (i) the burial depth of the tunnels, (ii) the number of
buildings-SDOFs and their position relative to the axis of the tunnel and (iii) the soil properties and response
during ground shaking were investigated within this parametric study. The computed responses of the in-
vestigated configurations, in terms of (i) lining deformations, (ii) dynamic earth pressures and seismic shear
stresses developed along the perimeter of the tunnels and (iii) dynamic lining forces, were compared with the
cases where the buildings-SDOFs at the ground surface were neglected. The effect of buildings-SDOFs was
quantified by means of response ratios μ, defined as the ratio of the maximum envelope value of the response
parameter computed at critical lining section at the presence of the buildings-SDOFs, to the relevant value
estimated for green-field conditions. The response ratios were plotted against the flexibility ratios, F, of the
investigated soil-tunnel systems, highlighting the effects of the salient parameters mentioned above on the in-
teraction phenomena. Generally, the consideration of the buildings-SDOFs at the ground surface resulted in an
increase of the response of the tunnels, compared to the one predicted for ‘green field’ conditions, both in terms
of tunnel deformations and dynamic earth pressures and soil dynamic shear stresses around the tunnel, as well as
in terms of dynamic lining forces, with the effect being generally higher for shallow tunnels.

1. Introduction et al., 2010; Kontoe et al., 2011, 2014; Debiasi et al., 2013; Bilotta et al.,
2014; Parvanova et al., 2014a,b; Abate et al., 2015; Lanzano et al.,
Tunnels and embedded structures performed better than above 2015; Baziar et al., 2016; Tsinidis et al., 2016c; Lu and Hwang, 2017;
ground structures during past strong earthquakes. However, several Tsinidis, 2017; Tsinidis and Pitilakis, 2018 among others) and analy-
cases of severe damage or even collapses have been reported in the tical (Huo et al., 2006; Bobet et al., 2008; Park et al., 2009; Bobet, 2010,
literature for these types of structures (e.g. Wang, 1993; Iida et al., among others) researches over the last years. However, the above re-
1996), motivating a series of experimental (Chou et al., 2010; search efforts have not been clearly reflected in the current seismic
Shibayama et al., 2010; Chian and Madabhushi, 2012; Cilingir and design practice. Actually, conventional design specifications and
Madabhushi, 2011a, 2011b, 2011c; Lanzano et al., 2012; Chen et al., methods (e.g. Wang 1993; Penzien, 2000; Hashash et al., 2001; ISO,
2013; Tsinidis et al., 2015; Abuhajar et al., 2015a,b; Ulgen et al., 2015, 2005; Anderson et al., 2008; FHWA, 2009; Pitilakis and Tsinidis, 2014)
Hushmand et al., 2016; Tsinidis et al., 2016a,b, among others), nu- are usually used, mainly due to their easy implementation.
merical (Huo et al., 2005; Hashash et al., 2005; Hwang and Lu, 2007; The current seismic design methods disregard the potential effects
Anastasopoulos et al., 2007, 2008; Amorosi and Boldini, 2009; Hashash of other existing structures in the adjacent area on the seismic response

E-mail addresses: [email protected], [email protected].

1
Formerly in: Research Unit of Soil Dynamics and Geotechnical Earthquake Engineering, Department of Civil Engineering, Aristotle University of Thessaloniki, Thessaloniki, Greece.

https://doi.org/10.1016/j.tust.2018.03.016
Received 11 December 2017; Received in revised form 12 February 2018; Accepted 10 March 2018
Available online 26 March 2018
0886-7798/ © 2018 Elsevier Ltd. All rights reserved.
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

of tunnels. However, during severe ground seismic shaking, strong in- equivalent continuous rings with an effective stiffness computed fol-
teraction phenomena may be mobilized between the above ground lowing Wood (1975). The corresponding to the effective stiffness, ef-
structures, e.g. buildings, and the tunnel, passing few meters below or fective lining thickness was set equal to t = 0.30 m. Table 1 summarizes
close by the foundations of these buildings, which may alter the seismic the mechanical properties adopted for the tunnel linings.
response of tunnels, compared to the one predicted for ‘green field’ A number of above ground structures were assumed at the ground
conditions. These interaction phenomena are expected to amplify in surface, representing nine-storey buildings. In particular, two cases
dense urban environments. were examined, assuming either two buildings at the ground surface,
The effect of underground cavities or embedded structures on the i.e. buildings A and B in Fig. 1, or six buildings at the ground surface,
seismic loading of the ground surface has been investigated by several i.e. buildings A-F in Fig. 1, in order to investigate both sparsely or
researchers (e.g. Lee and Karl, 1992; De Barros and Luco, 1993; densely populated urban environments. The relative position of the
Manoogian and Lee, 1996; Yiouta-Mitra et al., 2007; Kouretzis et al., buildings to the tunnel axis was parametrically investigated, as follows;
2007; Smerzini et al., 2009; Baziar et al., 2014; Abuhajar et al., 2015b; for a first series of analyses, the buildings (two or six) were located
Alielahi et al., 2015; Alielahi and Adampira, 2016; Baziar et al., 2016). above the tunnel and were assumed to be in contact, representing a
Additionally, there are several studies on the effects of embedded continuous building system (i.e. s1 = 0 m in Fig. 1). In a second series of
structures (e.g. tunnels or underground stations) on the seismic re- analyses, the buildings were clustered in groups of one or three build-
sponse of adjacent surface structures (e.g. Wang et al., 2013; Guo et al., ings (i.e. when two or six buildings were considered, respectively).
2013; Wang et al., 2017). These clusters were set at a distance, s1 = 30.0 m, with each other,
On the contrary, the effect of buildings on the seismic response of representing the case where the tunnel (single or twin) is located below
embedded structures, e.g. shallow tunnels, has received less attention. an open area, e.g. a boulevard. In the latter cases, the distance between
Navarro (1992) investigated numerically the effect of large buildings on the tunnel axis and the edge of buildings footings was set equal to
the seismic response of an adjacent tunnel, highlighting the importance 7.5 m, while the distance between the subsequent structures (i.e. for the
of the distance between the tunnel and the building on the interaction six buildings cases) was set equal to zero. The buildings were simulated
effects. Pitilakis et al. (2014) investigated the effect of buildings on the as equivalent Single Degree Of Freedom (SDOF) oscillators founded on
seismic response of shallow circular tunnels in soft soils, revealing a raft foundations of 1.0 m height. This simulation approach, which
general increase of the tunnel response, when the buildings were con- constitutes an approximation of the real problem, was inspired by the
sidered, compared to the cases where these structures were neglected. direct displacement based design (DDBD) method (e.g. Priestley and
Dashti et al. (2016) examined the feasibility of investigating experi- Kowalsky, 2000). The selections regarding the simulation of the
mentally the seismic interaction effects between buildings and box type equivalent SDOFs were simplified compared to the DDBD method, since
tunnels. Abate and Massimino (2017a,b) examined the transversal the study focused on the seismic performance of the tunnels. In parti-
seismic response of horse-shoe-tunnel-soil-building systems, providing cular, the buildings-SDOFs had an effective stiffness, corresponding to a
insights on the effect of the building on the seismic response of the fix-based period Tfix equal to 0.9 s. The effective mass was set equal to
tunnel. 400 t, while an equivalent damping of 5% was assumed. The effective
The present study aims to shed light on the transversal seismic re- mass was introduced at an effective height of 18 m, i.e. 2/3 of the actual
sponse of shallow circular tunnels, as affected by the presence of heavy height of the adopted buildings. It should be noted that the effect of the
buildings in the adjacent area, extending significantly the study by basements of buildings on the dynamic interaction phenomena was not
Pitilakis et al. (2014). For this purpose, a parametric numerical study considered. The latter may be considered as a limitation of the present
was conducted on representative circular tunnel-soil-buildings config- study, since these effects might not be negligible for some cases, par-
urations by employing the finite element code ABAQUS. Both single ticularly when large basements exist. Despite the simplification of some
and twin circular tunnels were considered, while the effects of critical of the above assumptions, these are considered rational and in any case
parameters, including (i) the burial depth of the tunnels, (ii) the were necessary to circumscribe the complex problem in hand. The
number of buildings and their relative position to the axis of the tunnel properties of the SDOF oscillators and their foundations are summar-
and (iii) the soil properties and response during ground shaking, were ized in Tables 2 and 3, respectively. Additional analyses were carried
investigated. The effect of the presence of the buildings at the ground out, neglecting the buildings-SDOFs at the ground surface. The results
surface on critical response characteristics of the investigated tunnel of the latter analyses were used as benchmarks for the investigation of
configurations is discussed in the following. Response ratios are defined the effects of the buildings on the seismic response of the tunnels.
for critical lining sections, aiming to quantify these effects for the in- To investigate the crucial effect of soil conditions on the systems
vestigated tunnels. The present investigation is limited on the trans- dynamic response, four different clayey soil deposits were considered,
versal seismic response of tunnels, since this response is commonly corresponding to soil classes B (rather stiff soils), C (predominantly soft
related to the maximum stress states of the lining. However, the long- and loose soils) and D (very soft soils), according to the Eurocode 8
itudinal seismic response of tunnels might also be critical for the design (CEN, EN 1998-1, 2004). The shear wave velocity (Vs) and the un-
(e.g. Hatzigeorgiou and Beskos, 2010; Yu et al., 2017), particularly drained shear strength (Su) of the soil deposits were assumed to increase
when the tunnel is subjected to non-uniform seismic excitations (Yuan with depth, as shown in Fig. 2a and b, respectively. The density ρ of the
et al., 2016; Yu et al., 2016a,b, 2017). The effect of above ground investigated soil deposits ranged between 1.8 and 1.9 t/m3, for the soft
structures on the longitudinal seismic response of tunnels is an open and stiff soils, respectively, while the Poisson’s ratio, vs, was set equal to
issue that is not covered herein and calls for further investigation. 0.40. The non-linear response of the deposits due to seismic ground
shaking was described by means of G-γ-D curves, following Ishibashi
2. Numerical parametric analysis and Zhang (1993), as per Fig. 2c. Solid lines correspond to the upper
soil layers, i.e. the first 30.0 m from the ground surface (layers A), while
A comprehensive set of tunnel-soil-buildings configurations was dashed lines stand for the deeper soil layers, i.e. from 30.0 m to 50.0 m
examined in this study, as per Fig. 1. More specifically, a single (Fig. 1a) below ground surface (layers B). These curves were used to calibrate
or twin (Fig. 1b) segmental circular tunnel was embedded in a soil the soil constitutive models, as discussed in the following sections. All
deposit of 50 m in depth, resting on elastic bedrock. The tunnels had a soil deposits were assumed to be rested on elastic bedrock, with shear
diameter, d, of 6.0 m, while their burial depth, h, ranged between wave velocity Vsb = 1000 m/s and density ρsb = 2.2 t /m3.
10.0 m and 20.0 m. The distance between the two tracks in the case of The analyses were carried out by employing two medium to quite
twin tunnel, s2, was set equal to 15.0 m. In line with common analysis strong acceleration time histories at the base of the numerical models
practices, the segmental linings were simulated in the analyses as (Table 4). The selection of strong earthquakes was made, since

178
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

Fig. 1. Schematic illustration of the investigated tunnel-soil-buildings configurations.

Table 1 Table 4
Lining properties. Shaking motions characteristics.
Elastic modulus, E (GPa) Poisson ratio, v Density, ρ (t/m3) A/A Earthquake Country Station name Magnitude, Mw PGA (g)
name
32 0.2 2.5
EQ1 Friuli, 1976 Italy Tolmezzo-Diga 6.4 0.34
Ambiesta
EQ2 Kobe, 1995 Japan Takatori 6.9 0.61
Table 2
Equivalent SDOF properties.
Elastic Number of Effective Effective Fixed-based 3. Numerical analysis
modulus, E storeys, n mass, m (t) damping, D (%) period Tfix (s)
(GPa)
The analyses were carried out under plane strain conditions using
32 9 400.0 5 0.9 the finite element code ABAQUS (ABAQUS, 2012). A typical model
layout is presented in Fig. 4, corresponding to a case of a twin circular
tunnel, with six buildings at the ground surface. The length of the soil
Table 3 grid was selected equal 250 m for all the investigated cases, based on
Equivalent SDOF foundations properties. the results of a sensitivity analysis that was conducted to investigate
Height (m) Elastic modulus, E (GPa) Poisson ratio, v Density, ρ (t/m3) potential boundary effects on the computed response at the central area
of the numerical models, i.e. where the tunnels and the structures are
1.0 32 0.2 2.5 located. The analyses were conducted in total stresses, assuming un-
drained conditions, a valid and commonly adopted hypothesis for
clayey soils subjected to severe seismic ground shaking.
underground structures appear to be more vulnerable in such loading
More specifically, the soil was meshed with quadratic plane strain
conditions. Fig. 3 compares the acceleration and displacement response
elements. The elements size allowed the efficient reproduction of all the
spectra of the selected records.
waveforms of the whole frequency range under study (i.e.

Vs (m/s) Su (kPa)

0 200 400 600 800 0 150 300 450 600 1 25

0 0
0.8 20
12.5 12.5
Depth (m)

Depth (m)

0.6 15
D (%)
G/Go

25 25
0.4 10

37.5 37.5 0.2 5

50 50 0 0
í3 í2 í1 0
10 10 10 10
Soil C1 Soil C2 Soil C3 Soil C4
γ (%)

Fig. 2. (a) Small strain shear wave velocity (Vs) gradients of the investigated soil deposits, (b) undrained shear strength (Su) profiles of the investigated soil deposits,
(c) G-γ-D curves adopted for the clayey soil deposits, following Ishibashi and Zhang (1993).

179
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

5 0.8
Friuli (1976)
3.75 Takatori (1995) 0.6

PSA/PSAo
EC8 (type A)

PSD (m)
2.5 0.4

1.25 0.2

0 0
0 1 2 3 4 0 1 2 3 4
T (s) T (s)

Fig. 3. (a) Acceleration and (b) displacement response spectra of the selected records.

Fig. 4. Typical numerical model configuration in ABAQUS.

f = 0.2–10 Hz). More refined models, used during preliminary sensi- ABAQUS. The viscous damping was introduced in the form of the fre-
tivity analyses had a negligible effect on the computed results, in- quency dependent Rayleigh type. The double frequency approach was
creasing significantly the computational cost. The tunnel linings were used to calibrate the Rayleigh parameters. (Hashash and Park, 2002).
simulated using beam elements. With reference to the simulation of the The latter were properly tuned for the dominant frequencies of the soil
buildings; the footings were simulated with quadratic plane strain deposits. Table 5 summarizes the average shear strain moduli, G, and
elements, while the equivalent SDOFs were modelled by means of beam damping ratios, D, computed by the one-dimensional soil response
elements. The connection between the SDOFs and their footings was analyses of the investigated soil deposits under the selected ground
established on the basis or kinematic constraints, i.e. tie type in shaking motions.
ABAQUS. The effective mass of the structures was simulated using mass Despite its simplicity, the above visco-elastic material cannot re-
elements that were introduced atop of the SDOFs. plicate critical aspects of the actual soil response under ground shaking,
The interfaces between the tunnels linings and the soil, as well as including the soil yielding response, which might be critical for the
between the footings of the SDOFs and the soil were modelled, by seismic response of tunnels. Along these lines, additional analyses were
implementing a finite sliding hard contact model (ABAQUS, 2012). The carried out using a non-linear kinematic hardening model for the soil.
model allows the potential detachment and/or sliding between the in- The model, which is referred as KH model in the following, combines a
teracting elements during ground seismic shaking. The shear behaviour Von-Mises failure criterion with an associated plastic flow rule and can
of the interfaces was controlled by the classical Coulomb friction model, be considered as a reasonable approximation for the description of the
through the introduction of a friction coefficient, μ. The latter was set seismic response of clayey soils under undrained conditions
equal to 0.7 for the buildings foundations-soil interfaces, while a lower (Anastasopoulos et al., 2011). The evolution law of the model com-
value, i.e. 0.5, was set to the tunnel-soil interfaces. prises of an isotropic hardening component and a nonlinear kinematic
A linear elastic model was implemented for the simulation of the hardening component. The former describes the change in the equiva-
tunnel lining response. The model was calibrated adopting the me- lent stress, defining the size of the yield surface σo as a function of the
chanical properties presented in Table 1. The response of the SDOFs plastic deformation. The kinematic hardening component describes the
was simulated in a similar fashion, using the properties presented in translation of the yield surface in the stress space (defined through a
Table 2. The equivalent viscous damping ratio of 5%, which was ‘back-stress’ parameter α ). The evolution of the kinematic component of
adopted for the buildings, was introduced on the mass element in the the yield stress is described as follows:
form of Rayleigh damping.
1
The response of the soil under ground seismic shaking was initially ȧ = E (σ −α ) ε¯ ̇pl−γαε¯ ̇pl
σo (1)
simulated through a simplified visco-elastic material (referred as VE
model in the following), following the equivalent linear approximation. where ε̄ ̇pl is the equivalent plastic strain rate, E is the small-strain
This approach is commonly implemented in design practice, due to its elastic stiffness of the soil and γ is a parameter that describes the rate, at
easy application, while it is adopted by relevant design guidelines (e.g. which the kinematic hardening decreases with the increasing plastic
FHWA, 2009). One dimensional soil response analyses were conducted deformation. For the clayey soil deposits examined herein, the latter
in the frequency domain with the numerical code EERA (Bardet et al., parameter can be defined as (Anastasopoulos et al., 2011):
2000), using the G-γ-D curves, presented in Fig. 2c. The average shear
moduli and viscous damping, estimated for each soil layer and shaking γ = C /[ 3 × Su−σo] (2)
scenario, were then introduced in the soil-tunnel-buildings models in
Parameter σ0 that controls the initiation of the nonlinear behaviour

180
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

Table 5
Average shear moduli, G, and damping ratios, D, computed by one dimensional soil response analyses of the investigated soil deposits under the selected ground
shaking motions.
Shear moduli G (MPa) Damping ratios (%)
EQ1 EQ1

Layer Soil C1 Soil C2 Soil C3 Soil C4 Layer Soil C1 Soil C2 Soil C3 Soil C4

A 21.6 108.5 256.0 569.6 A 10.5 4.5 2.3 0.9

B 91.9 276.8 464.5 802.7 B 3.5 1.2 1.0 0.8

Shear moduli G (MPa) Damping ratios (%)

EQ2 EQ2

Layer Soil C1 Soil C2 Soil C3 Soil C4 Layer Soil C1 Soil C2 Soil C3 Soil C4

A 10.5 53.8 212.2 552.9 A 19.1 14.9 5.2 1.2

B 44.2 124.4 388.1 759.4 B 13.3 11.5 3.1 1.5

Soil C2 − Layer A Soil C2 − Layer B

1 40 1 30 300

0.8 30 0.8 150

20

τ (kPa)
0.6 0.6
D (%)

D (%)
G/Go

G/Go

20 0
0.4 0.4 γ = 0.1 %
10
10 í150 γ = 0.5 %
0.2 0.2
γ=1%
0 0 0 0 í300
í4 í3 í2 í1 0 í4 í3 í2 í1 0 í0.016 í0.008 0 0.008 0.016
10 10 10 10 10 10 10 10 10 10
γ
γ (%) γ (%)

Fig. 5. Calibration of the kinematic hardening model against the ‘target’ G-γ-D curves for (a) the upper layer (Layer A) and (b) the lower layer (Layer B) of soil deposit
C2; (c) representative stress-strain loops computed by numerical simulations of cyclic simple shear tests, corresponding to the various soil shear strains γ.

(yield stress at zero plastic strain) is defined as a fraction λ of the yield Kinematic tie constraints were set between every pair of nodes of
stress σy , i.e. σ0 = λ × σy . In this context, the model is completely de- the side boundaries that shares the same y-coordinate (displacements
fined on the basis of three parameters, i.e. the soil elastic Young's constraints in Fig. 4), allowing for common lateral displacement pat-
modulus, E, the soil ultimate strength, σy and the soil yield stress, σ0 (the terns, throughout the analysis procedure. This boundary condition
latter defined through parameter λ ). The model has been extensively imitates the desirable ‘shear beam’ response of free-field soil during
validated against physical tests, referring to circular tunnels (Tsinidis ground seismic shaking.
et al., 2014; Bilotta et al., 2014; Tsinidis et al., 2016c) retaining To determine the soil to tunnel relative stiffness of the examined
structures (Anastasopoulos et al., 2010) and surface, embedded and pile cases, the flexibility ratios were computed for all the soil-tunnel con-
foundations (Anastasopoulos et al., 2011; Anastasopoulos et al., 2012; figurations, as per Wang (1993). The computations were made using
Giannakos et al., 2012). In the herein study, parameter λ was system- the ‘average’ degraded soil shear moduli, estimated at the corre-
atically calibrated for the adopted soil layers, through numerical ana- sponding depths of the tunnels by the one-dimensional soil response
lyses of cyclic simple shear tests, as per Anastasopoulos et al. (2011). analyses with EERA (see Table 5), as follows:
Fig. 5 illustrates representative comparisons between the ‘target’ and
the computed (through the calibration procedure) G-γ-D curves for the F = [Es (1−νl2 ) R3]/[6El Il (1 + νs )] (4)
soil layers of soil deposit C2. The selected values of λ ranged between where Es is the soil’s elastic modulus, vs is the soil’s Poisson ratio, El is
1/2 to 1/8, for the investigated soil deposits. A small fraction of the lining’s elastic modulus, vl is the lining’s Poisson ratio, R is the
damping, i.e. 2%, was assigned to all soil deposits to simulate the vis- circular tunnel radius and Il is the lining moment of inertia (per unit
cous damping at low strains. This damping was simulated in the fre- length). The flexibility ratios of the selected soil-tunnel configurations
quency depended Rayleigh form, as for the visco-elastic analyses. ranged between 1.3 and 68.3, representing a wide range of soil to
The analyses were carried out in steps. Initially, the gravity loads tunnel relative stiffnesses; i.e. relatively stiff tunnels compared to sur-
were introduced within a static step. The seismic loading was then in- rounding ground (i.e. low flexibility ratios) to very flexible tunnels
troduced within an implicit dynamic step. During the initial static step, compared to the surrounding ground (i.e. high flexibility ratios).
the base of the numerical model was fixed in both horizontal and ver-
tical directions. In the subsequent ‘earthquake’ dynamic step, the hor-
izontal direction of the base was released and the seismic shaking 4. Discussion of results
motions were applied as vertically propagated shear waves through
appropriate horizontal dashpots, the latter defined as per Lysmer and 4.1. Deformation mechanisms of the investigated systems
Kuhlemeyer (1969). The dashpots coefficients, c, were defined as the
product of the mass density, ρ, and shear wave velocity, Vsb, of the The tunnels experienced the theoretically expected ovaling distor-
underlying bedrock, including also the corresponding ‘soil’ surface, A, tion during ground seismic shaking. Fig. 6 illustrates representative
covered by each dashpot, in order to maintain proportional results for ovaling ratios, R, - flexibility ratios, F, relations, computed for single
any horizontal element size: tunnels embedded in various depths in the soil deposits investigated
herein. The ovaling ratios are computed as R = δstr/δff, where δstr is the
c = ρ × Vsb × A (3) maximum ovaling distortion of the tunnel section during ground
seismic shaking and δff is the equivalent distortion of the soil computed

181
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

Burial depth, h=10 m Burial depth, h=20 m tunnel being considered at a burial depth h = 10.0 m and a continuous
3 3
structural system being assumed for the buildings-SDOFs at the ground
2.25 2.25
surface, i.e. s1 = 0. The positive values of u in the u-θ response refer to
1.5 1.5
R

R
uplifting of the footing, whereas the negative values stand for settle-
0.75 0.75 ment. For visco-elastic soil response and shaking motion EQ1, a rather
elastic moment-rotation response is observed for the footing, regardless
0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 of the number of the existing structures-SDOFs at the ground surface.
F F The footing does not experience any settlement, while reduced uplifting
no SDOFs 2 SDOFs 6 SDOFs is only observed during ground shaking (e.g. Fig. 8b). On the contrary,
when the systems are subjected to the more severe shaking motion EQ2,
Fig. 6. R-F relations computed for single tunnels embedded in visco-elastic soil significant geometrical no linear phenomena are taking place along the
deposits, when considering or neglecting the buildings-SDOFs at the ground interface of the footing with the soil. In particular, significant uplifting
surface (results for s1 = 0 m). is observed (e.g. red lines in the u-θ response), which results in energy
dissipation at the foundation level, as can be seen in the M-θ loops.
at free-field conditions and at the corresponding depth with the tunnel. Some slight differences are observed between the computed responses
The results are plotted for a visco-elastic soil response (VE model) and of the footing, when two or six structures-SDOFs are considered at the
shaking motion EQ1, considering or neglecting the buildings-SDOFs at ground surface. Similar observations were made for other soil condi-
the ground surface, and correspond to the case where the buildings- tions, as well as for the cases where s1 = 30 m. It is worth noticing that
SDOFs are in contact, i.e. s1 = 0 m. Higher ovaling ratios are computed for the same seismic shaking motion, the rocking response of the in-
for the same flexibility ratios, when the buildings-SDOFs are considered vestigated buildings-SDOFs was amplified with the increase of the
at the ground surface. This is particularly evident for shallow tunnels, foundation soil stiffness, i.e. for soil deposits C3 and C4 (Table 5).
i.e. for h = 10 m, with the differences reaching a maximum of 30% for The moment-rotation (M-θ) and uplifting/settlement-rotation (u-θ)
quite flexible tunnels (i.e. for F = 68.3). The effect of the buildings- responses of the investigated footing are quite different when a visco-
SDOFs was less evident for deep tunnels, i.e. for h = 20 m. elasto-plastic soil response is considered (i.e. KH model). The computed
Fig. 7 portrays typical deformed shapes of twin tunnels-soil-SDOFs M-θ loops indicate energy dissipation, associated with the yielding of
systems computed directly by the analyses at the time frame when the the soil under the footing. This yielding response, which naturally is
ovaling deformation of the tunnels is maximized. The presented de- more evident for the more severe shaking motion EQ2, results in an
formed shapes correspond to either a visco-elastic (i.e. VE model) or a accumulation of soil settlements under the footing, as can be seen in the
visco-elasto-plastic soil response (i.e. KH model) and are plotted for the u-θ diagrams and for some cases in permanent rotation of the footing, as
soil deposit C2 and for both examined shaking motions. For a visco- well (e.g. θ ≠ 0 in cases of six buildings-SDOFs at the ground surface in
elastic soil response, the buildings-SDOFs are exhibiting a rocking-up- Fig. 8b). The number of the structures at the ground surface affects the
lifting response (Fig. 7a and c). This is particularly evident in the case of stress state of the foundation soil during ground seismic shaking, which
the shaking motion EQ2, where significant uplifting of the footings of in turns affects the extent of soil yielding, as discussed in the ensuing.
the buildings-SDOFs is actually observed (Fig. 7c). The rocking-up- This rationale explains the higher dispersion between the predicted u-θ
lifting response of buildings-SDOFs is reduced in the cases where a and the Μ-θ responses of the examined footing, when two or six
visco-elasto-plastic soil response is considered (KH model, Fig. 7b and buildings-SDOFs are considered at the ground surface (compared to the
d). In the latter cases, a settlement of the buildings-SDOFs footings is responses computed for visco-elastic soil response).
actually observed, associated with the yielding of the foundation soil The above distinct deformation patterns, which generally depend on
caused by the oscillations of the SDOFs. The yielding phenomena and the characteristics of the foundation and the foundation soil (e.g. safety
the associated settlements of the footings are more evident in the case factor of the foundation), the existence of basements, as well as on the
of shaking motion EQ2 and for the softer soil deposits. seismic response of the buildings (idealized herein in a simplified
The above observations are also verified in Fig. 8, where the mo- fashion through a visco-elastic response), affect the stress state of the
ment-rotation (M-θ) and the uplifting/settlement-rotation (u-θ) re- soil and therefore the seismic response of the tunnels that are crossing
sponses of the footing of building-SDOF B (as per Fig. 1) are compared underneath or nearby. One critical effect refers to the yielding response
under various assumptions regarding the soil response (i.e. visco-elastic of the soil in the adjacent area of the structures, where the tunnels are
or visco-elasto-plastic response) and the number of buildings-SDOFs at located. Fig. 9 highlights this effect, by comparing contour diagrams of
the ground surface. The results refer to soil deposit C2, with one single the soil dynamic plastic strains (i.e. plastic strains caused by ground

Fig. 7. Deformed shapes of twin tunnels-soil-SDOFs

systems computed for (a) visco-elastic soil response
(VE model) and shaking motion EQ1, (b) visco-
elasto-plastic soil response (KH model) and shaking
motion EQ1, (c) visco-elastic soil response (VE
model) and shaking motion EQ2, (d) visco-elasto-
plastic soil response (KH model) and shaking motion
EQ2 (soil type C2, scale factor × 5).

182
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

EQ1 - 2SDOFs EQ1 - 6SDOFs EQ2 - 2SDOFs EQ2 - 6SDOFs

26 26 40 40
M (MNm/m)

M (MNm/m)

M (MNm/m)
13 13

M (MNm/m)
20 20
0 0 0 0
-13 -13 -20 -20
-26 -26 -40 -40
-4 -2 0 2 4 -4 -2 0 2 4 -0.06 -0.03 0 0.03 0.06 -0.06 -0.03 0 0.03 0.06
(rad) 10 -3 (rad) 10 -3 (rad) (rad)

EQ1 - 2SDOFs EQ1 - 6SDOFs EQ2 - 2SDOFs EQ2 - 6SDOFs

70 35 300 300

0 0 0 0

u (mm)

u (mm)
u (mm)

u (mm)

-70 -35 -300 -300

-140 -70 -600 -600

-4 -2 0 2 4 -4 -2 0 2 4 -0.1 -0.05 0 0.05 0.1 -0.1 -0.05 0 0.05 0.1
(rad) 10 -3 (rad) 10 -3 (rad) (rad)

VE model KH model

Fig. 8. (a) Moment-rotation (M-θ) and (b) settlement-rotation (u-θ) response of the footings of building-SDOF B, computed for visco-elastic (VE model) and visco-
elasto-plastic (KH model) soil response (soil type C2).

seismic shaking), predicted for representative soil-tunnel configura- surface are found to affect the soil yielding patterns. Significantly
tions, when the buildings-SDOFs are considered or neglected at the higher plastic strains are predicted around the tunnels, when the
ground surface. The results, which are computed at the completion of buildings are located above them (i.e. when s1 = 0 m). The above
shaking motion EQ1, are plotted for the upper layer of soil deposit C2 yielding phenomena affect the response of the underlying tunnels, as
(layer A). Generally, the existence of the buildings-SDOFs at the ground discussed in the following sections.
surface result in a higher yielding response of the soil around the tun-
nels compared to the cases where a ‘green field’ condition is considered 4.2. Dynamic earth pressures
(i.e. higher soil plastic strains are distributed in significant soil areas
beneath the structures). It is worth noticing the significant soil yielding The effect of the buildings-SDOFs at the ground surface on the dy-
that is observed at the edges of the footings of the buildings-SDOFs, as a namic earth pressures, developed along the perimeter of the tunnels
result of their oscillating response during ground shaking. Additionally, during ground seismic shaking, is discussed in this section. The dis-
both the number and position of the buildings-SDOFs at the ground cussion is made on the basis of envelope values (i.e. maximum and

Fig. 9. Distributions of soil dynamic plastic strains

computed around single or twin tunnels, embedded
in Layer A of soil deposit C2, at the completion of
shaking motion EQ1; (a) single tunnel, h = 10 m,
(b) twin tunnel, h = 10 m, (c) twin tunnel,
h = 10 m, two SDOFs, s1 = 0 m, (d) twin tunnel,
h = 10 m, two SDOFs, s1 = 30 m, (e) twin tunnel,
h = 20 m, two SDOFs, s1 = 0 m, (f) single tunnel,
h = 10 m, two SDOFs, s1 = 30 m, (g) twin tunnel,
h = 10 m, six SDOFs, s1 = 30 m, (h) twin tunnel,
h = 10 m, six SDOFs, s1 = 0 m.

183
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

VE model, no SDOFs VE model, 2 SDOFs KH model, no SDOFs KH model, 2 SDOFs

200 200 200 200

100 100 100 100

(kPa/m)

(kPa/m)

(kPa/m)

(kPa/m)
0 0 0 0

-100 -100 -100 -100

-200 -200 -200 -200

0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360
(o ) (o ) (o ) (o )

Single tunnel Twin tunnel - left track

Fig. 10. Maximum and minimum envelope distributions of dynamic earth pressures computed along the perimeter of a single tunnel and the left track tunnel of a
twin tunnel, located in the same position, under shaking motion EQ1 (soil deposit C2, burial depth, h = 10.0 m).

minimum dynamic earth pressures computed regardless of the time tunnel. The effect is more evident in cases of shallow tunnels. It is worth
step), since these are commonly used for design purposes. noticing that the increases on the envelope earth pressures are quite
Fig. 10 compares representative maximum and minimum envelope comparable in both examined cases regarding the relative position of
distributions of dynamic earth pressures computed along the perimeter the buildings-SDOFs to the tunnel axis (i.e. s1 = 0 m, s1 = 30 m).
of a single tunnel and the left track tunnel of a twin tunnel, the latter The distributions of the maximum and minimum envelope dynamic
located in the same position with the single tunnel. The distributions earth pressures are somehow more complex, when a visco-elasto-plastic
refer to shallow tunnels (h = 10 m), embedded in the soil deposit C2, soil response is considered (KH model). This is attributed to the yielding
and are computed for shaking motion EQ1 for a visco-elastic (VE response of the soil, which causes stress redistributions within the soil
model) or a visco-elasto-plastic (KH model) soil response, considering around the tunnel and in turn results in residual values on the dynamic
or neglecting buildings-SDOFs at the ground surface. The computed earth pressures. These residual values on the dynamic earth pressures
distributions are almost identical, indicating a negligible effect of the lead to more complex forms for the pressures distributions around the
right track tunnel of the twin tunnel on the seismic response of the left tunnel. This response, which has been verified experimentally and nu-
one, for the investigated cases. merically for both circular and rectangular tunnels (e.g. Cilingir and
Fig. 11 portrays maximum and minimum envelope distributions of Madabhushi, 2011a,b; Tsinidis et al., 2015, 2016a), is affected by the
dynamic earth pressures computed along the perimeter of single tunnels presence of the buildings-SDOFs. Indeed, the existence of the buildings-
embedded in the soil deposit C2, when considering or neglecting the SDOFs results in additional yielding of the soil around the tunnels
buildings-SDOFs at the ground surface. The distributions computed for compared to the ‘green-field’ case (e.g. Fig. 9), which in turns lead to
a visco-elastic soil response (VE model), neglecting the buildings-SDOFs higher post-earthquake residual earth pressures and hence more com-
at the ground surface, are quite symmetric, with the dynamic earth plex dynamic earth pressures distributions. The above response is ob-
pressures envelopes being maximized at lining sections of θ = ± 45° (θ served in both examined cases regarding the positions of the buildings-
is defined as per Fig. 11). The consideration of the buildings-SDOFs at SDOFs at the ground surface (i.e. for both s1 = 0 and 30 m).
the ground surface results in a general increase of the envelope dis- To quantify the effect of the buildings-SDOFs on the maximum en-
tributions, as a result of the structures on the soil stress state around the velope dynamic earth pressures developed around the examined

s 1 =0m, h=10m s 1 =0m, h=20m s 1 =0m, h=10m s 1 =0m, h=20m

200 200 200 300
100 100 100 150
(kPa/m)

(kPa/m)

(kPa/m)

(kPa/m)

0 0 0 0
-100 -100 -100 -150
-200 -200 -200 -300
0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360
(o ) (o ) (o ) (o )

s 1 =30m, h=10m s 1 =30m, h=20m s 1 =30m, h=10m s 1 =30m, h=20m

200 200 200 200
100 100 100 100
(kPa/m)

(kPa/m)

(kPa/m)
(kPa/m)

0 0 0 0
-100 -100 -100 -100
-200 -200 -200 -200
0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360
(o ) (o ) ( ) o (o )

no SDOFs 2 SDOFs 6 SDOFs

Fig. 11. Maximum and minimum envelope distributions of dynamic earth pressures computed along the perimeter of the single tunnel under shaking motion EQ1 for
(a) visco-elastic soil response and s1 = 0 m, (b) visco-elasto-plastic soil response and s1 = 0 m, (c) visco-elastic soil response and s1 = 30 m, (d) visco-elasto-plastic
soil response and s1 = 30 m.

184
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

3 3 3 3

2.25 2.25 2.25 2.25

1.5 1.5 1.5 1.5

0.75 0.75 0.75 0.75

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

3 3 3 3

2.25 2.25 2.25 2.25

1.5 1.5 1.5 1.5

0.75 0.75 0.75 0.75

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 12. Response ratios μσ of maximum envelope dynamic earth pressures computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth h = 10.0 m
and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d) tunnel’s
burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elastic soil response (VE model).

tunnels, response ratios were computed, as follows: computed for s1 = 30 m (buildings or clusters of buildings set at a
θ = 225o θ = 225o
distance of 30 m) compared to the cases where the buildings-SDOFs are
μσ = max{env (σ|With SDOFs )}/max{env (σ|Without SDOFs )} (5) considered above the tunnels and in contact, i.e. s1 = 0 m. This ob-
θ = 225o
where max{env (σ|With SDOFs )} is the maximum envelope value of the servation should be attributed to the complex effect of the rocking os-
dynamic earth pressure computed at a critical section of the tunnel cillation of the buildings-SDOFs during ground shaking (especially in
lining (i.e. θ = 225°, as per Fig. 12), when the buildings-SDOFs are case of shaking motion EQ2) on the stress state of the ground, sur-
θ = 225o
considered, and max{env (σ|Without rounding the tunnels. Indeed, this rocking oscillation induces higher
SDOFs )} is the maximum envelope dy-
namic earth pressure estimated at the same location in the absence of stresses at the sides of the foundations compared to the middle sections,
the buildings-SDOFs. The response ratios were evaluated for all the creating complex stress fields in the soil that might be more critical for
examined single tunnels, as well as the left track tunnels of the twin tunnels passing near the buildings compared to those underpassing the
tunnels and were plotted against the relevant flexibility ratios, F. It is buildings.
recalled that the flexibility ratios, F, were evaluated, as discussed in Fig. 13 illustrates the response ratios μσ computed at the selected
Section 3, using Eq. (4). Through the following discussion, the effects of lining section for a visco-elasto-plastic soil response (KH model). The μσ
tunnel’s burial depth, soil response under ground shaking, number and – F relations, presented herein, are quite distinct compared to those
position of buildings-SDOFs and ground shaking motion characteristics, predicted for a visco-elastic soil response. In most of examined cases,
on the response ratios are highlighted. the response ratios are decreasing with the increase of the flexibility
Fig. 12 compares the response ratios μσ computed, as per Eq. (5), ratio. This is particularly evident for cases, where s1 = 0 (i.e. buildings-
assuming a visco-elastic soil response (VE model). In line with the SDOFs in contact, tunnel underpasses the buildings-SDOFs). Similar to
previous results, the consideration of the buildings-SDOFs at the ground the visco-elastic analyses results, higher response ratios are computed
surface results in a general increase of the envelope dynamic earth for the shallow tunnels (i.e. h = 10 m), as well as for more severe
pressures at the examined lining sections, for most of examined cases, shaking motions (i.e. shaking motion EQ2). Some minor differences (up
i.e. μσ > 1.0. Higher response ratios μσ are generally computed for the to 5%) are observed between the response ratios computed at the ex-
shallow tunnels (i.e. for h = 10 m, in Fig. 12a and b), when subjected to amined lining section of the single tunnel and the relevant section of the
shaking motion EQ2, reaching a maximum of 2.8 for quite flexible left track tunnel of the twin tunnel. These differences indicate a slight
tunnels. Although it is not possible to observe a clear tendency, a slight effect of the right track tunnel of the twin tunnel on the seismic re-
increase of the response ratios μσ with increasing flexibility ratio F is sponse of the left track one, when a visco-elasto-plastic soil response is
observed for most cases. Additionally, the response ratios computed for considered. Interestingly, the response ratios computed for s1 = 30 m,
the single tunnel and the left track tunnel of the twin tunnel are almost are in most of the examined cases lower than those computed for s1 =
identical, indicating again a rather negligible effect of the right track 0 m. Actually, there are cases where response ratios lower than unity
tunnel of the twin tunnel on the envelope dynamic earth pressures are computed for s1 = 30 m (e.g. for deep tunnels, h = 20 m), in-
developed on the left track tunnel. Generally, the consideration of six dicating a reduction of the envelope dynamic earth pressures at the
buildings-SDOFs at the ground surface results in higher response ratios examined section, when the buildings-SDOFs are considered. These
compared to the cases where two buildings-SDOFs are considered, with observations should be attributed to the effect of the position of the
the differences between the two cases being as high as 20–30%. The buildings-SDOFs on the soil yielding response patterns (see for example
relevant position of the buildings-SDOFs compared to the axis of the Fig. 9). Since the number of the buildings-SDOFs at the ground surface
tunnel is also found to affect the response ratios, to some extent. In- affects the extent of the soil yielding area around the tunnel and hence
terestingly, there are some cases, where higher response ratios are the residual dynamic earth pressures developed around the tunnel, it is

185
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

2.2 2.2 2.2 2.2

1.65 1.65 1.65 1.65

1.1 1.1 1.1 1.1

0.55 0.55 0.55 0.55

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

2.2 2.2 2.2 2.2

1.65 1.65 1.65 1.65

1.1 1.1 1.1 1.1

0.55 0.55 0.55 0.55

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 13. Response ratios μσ of maximum envelope dynamic earth pressures computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth h = 10.0 m
and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d) tunnel’s
burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elasto-plastic soil response (KH model).

expected the response ratios computed for two or six buildings-SDOFs negligible, since similar response ratios are computed for the single
at the ground surface to differ. Indeed, in most cases, the consideration tunnel and the left track one (located at the same position with the
of six buildings-SDOFs at the ground surface results in higher response single tunnel). Both the location and the number of the buildings-SDOFs
ratios. at the ground surface affect the computed response ratios μτ. In line
with the observations made for the dynamic earth pressures, higher
4.3. Soil dynamic shear stresses response ratios are observed for s1 = 30 m compared to the cases of s1
= 0 m, as well as for the cases where six buildings-SDOFs are con-
The consideration of the buildings-SDOFs altered the soil dynamic sidered at the ground surface.
shear stresses, developed along the perimeter of the examined tunnels Fig. 15 portrays the relevant response ratios μτ computed for a visco-
during ground shaking, compared to those predicted for a ‘green-field’ elasto-plastic soil response. The μτ – F relations presented herein are
condition. In this section, the effect of buildings-SDOFs on the soil different compared to those computed for a visco-elastic soil response.
dynamic shear stresses is discussed in terms of response ratios, μτ, With some exceptions (e.g. for deep tunnels, h = 20 m, and s1 = 30 m,
computed as follows: Fig. 15c and d), the response ratios μτ are higher than unity, indicating
an increase of the envelope soil dynamic shear stress at the particular
θ = 225 o θ = 225 o
μτ = max{env (τ|With SDOFs )}/max{env (τ|Without SDOFs )} (6) lining section when the buildings-SDOFs are considered at the ground
surface. Generally, the response ratios are increasing with the decrease
θ = 225o
where max{env (τ|With SDOFs )}is the maximum envelope value of the soil of the tunnel embedment (i.e. higher response ratios for h = 10 m), as
dynamic shear stress computed at a critical section of the tunnel lining well as for more severe shaking motions (i.e. shaking motion EQ2). It is
(i.e. θ = 225°, as per Fig. 14) when the buildings-SDOFs are considered, not possible to identify clear tendencies for the variation of response
θ = 225o
and max{env (τ|Without SDOFs )} is the maximum envelope soil dynamic ratio μτ with increasing flexibility ratio, F. As mentioned above, the
shear stress evaluated at the same location in the absence of the location and number of the structures-SDOFs at the ground surface
buildings-SDOFs. The computed response ratios are plotted against the affect the extent of soil yielding around the tunnels, subsequently af-
soil-tunnel flexibility ratios, F, for all the examined single tunnels and fecting the development of residual shear stresses around the tunnel.
the left track tunnels of the twin tunnels. The latter affect the form of the soil dynamic shear stresses envelope
Fig. 14 summarizes the response ratios that were computed for a distributions around the perimeter of the tunnel and therefore the
visco-elastic soil response (VE model). In most cases, the consideration computed response ratios. This may explain the not clear tendencies of
of the buildings-SDOFs leads to an increase of the envelope soil dy- response ratios with increasing flexibility ratio, F, at least to some ex-
namic shear stress at the examined lining section, i.e. μτ > 1.0. This is tent.
particularly evident, in cases of shallow flexible tunnels (i.e. h = 10 m,
F > 15), when systems are subjected to shaking motion EQ2. The re-
sponse ratios computed for these cases can be as high as 2.2. Con- 4.4. Lining dynamic bending moment
siderably lower response ratios are computed for the deep tunnels (i.e.
h = 20 m), when the systems are subjected to shaking motion EQ1. The seismic forces developed on the tunnel lining during ground
Actually, for relatively rigid tunnels, a slight reduction of the envelope seismic shaking are of great importance for an adequate design and
soil dynamic shear stress at the examined lining section is observed detailing. In this section, the effects of buildings-SDOFs on lining dy-
when the buildings-SDOFs are considered (i.e. μτ = 0.9 – 0.95). The namic bending moment are highlighted and discussed.
effect of the right track tunnel of the twin tunnel on the envelope soil Comparisons of maximum and minimum envelope distributions of
dynamic shear stresses developed on the left track one is again rather dynamic bending moments, computed along the perimeter of single

186
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

2.2 2.2 2.2 2.2

1.65 1.65 1.65 1.65

1.1 1.1 1.1 1.1

0.55 0.55 0.55 0.55

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

2.2 2.2 2.2 2.2

1.65 1.65 1.65 1.65

1.1 1.1 1.1 1.1

0.55 0.55 0.55 0.55

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 14. Response ratios μτ of maximum envelope soil dynamic shear stresses computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth h = 10.0 m
and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d) tunnel’s
burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elastic soil response (VE model).

tunnels, embedded in the soil deposit C2, when considering or ne- the envelope distributions, especially for the cases of shallow tunnels
glecting the buildings-SDOFs at the soil surface, are presented in (i.e. for h = 10 m). It is worth noticing that the effect of the buildings-
Fig. 16. The results refer to both visco-elastic (VE model) and visco- SDOFs on the dynamic bending moment distributions exists in both
elasto-plastic soil response (KH model) and are computed for shaking examined cases regarding the position of buildings at the ground sur-
motion EQ1. Symmetric envelope distributions are observed around the face (i.e. for s1 = 0 m or 30 m).
perimeters of the tunnels, for a visco-elastic soil response, with the The yielding response of the soil, which is predicted around the
dynamic bending moments being maximized at θ = ± 45°, following tunnels when a visco-elasto-plastic soil response is considered, and the
the distributions of dynamic earth pressures. The consideration of the associated stress redistributions that are taking place within the soil,
structures-SDOFs at the ground surface results in a general increase of result in residual values on the dynamic bending moments, which in

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

1.6 1.6 1.6 1.6

1.2 1.2 1.2 1.2

0.8 0.8 0.8 0.8

0.4 0.4 0.4 0.4

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

1.6 1.6 1.6 1.6

1.2 1.2 1.2 1.2

0.8 0.8 0.8 0.8

0.4 0.4 0.4 0.4

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 15. Response ratios μτ of maximum envelope soil dynamic shear stresses computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth h = 10.0 m
and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d) tunnel’s
burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elasto-plastic soil response (KH model).

187
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

s 1 =0m, h=10m s 1 =0m, h=20m s 1 =0m, h=10m s 1 =0m, h=20m

100 100 200 200

M (kNm/m)
50 100 100
M (kNm/m)

M (kNm/m)
50

M (kNm/m)
0 0 0 0

-50 -50 -100 -100

-100 -100 -200 -200
0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360
(o ) (o ) (o ) (o )

s 1 =30m, h=10m s 1 =30m, h=20m s 1 =30m, h=10m s 1 =30m, h=20m

100 100 200 200
50 50 100
M (kNm/m)

M (kNm/m)

M (kNm/m)
100

M (kNm/m)
0 0 0 0
-50 -50 -100 -100
-100 -100 -200 -200
0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360
(o ) (o ) (o ) (o )

no SDOFs 2 SDOFs 6 SDOFs

Fig. 16. Maximum and minimum envelope distributions of lining dynamic bending moments computed along the perimeter of the single tunnel under shaking
motion EQ1 for (a) visco-elastic soil response and s1 = 0 m, (b) visco-elasto-plastic soil response and s1 = 0 m, (c) visco-elastic soil response and s1 =30 m, (d) visco-
elasto-plastic soil response and s1 = 30 m.

turn lead in more complex forms for the dynamic bending moment SDOFs to the tunnel axis is more evident for shallow tunnels. Following
distributions around the perimeter of the tunnels. This observation is in a similar pattern to the dynamic earth pressures, higher response ratios
line with recent experimental and numerical studies on the seismic are computed for s1 = 30 m compared to s1 = 0 m, in most examined
response of circular tunnels (e.g. Lanzano et al., 2012; Cilingir and cases.
Madabhushi, 2011a, 2011b; Tsinidis et al., 2014, 2016c). As mentioned Fig. 18 illustrates the response ratios μM computed for a visco-elasto-
in the previous sections, the above response is affected by the presence plastic soil response (KH model). The trends of response ratios μM with
of the buildings-SDOFs at the ground surface, since the buildings-SDOFs increasing flexibility ratio, F, are quite distinct compared to those
have a clear effect on yielding patterns of the soil. Generally, the con- computed for a visco-elastic soil response. Actually, it is not easy to
sideration of more buildings-SDOFs at the surface, results in higher draw a clear tendency for the variation of the response ratios with in-
envelopes of dynamic bending moments on the tunnel lining. creasing flexibility ratio. Generally, higher response ratios (as high as
The effect of buildings-SDOFs on the maximum envelope dynamic 4.2) are predicted for shallow tunnels (i.e. for h = 10 m), when sub-
bending moments, developed at the examined tunnels during ground jected to shaking motion EQ2. The number and location of the build-
shaking, was quantified by means of response ratios μM, computed as ings-SDOFs at the ground surface are found to affect significantly the
follows: computed response ratios. In line with the observations made for the
o o
response ratios rσ (referring to envelope dynamic earth pressures),
θ = 225 θ = 225
μM = max{env (M|With SDOFs )}/max{env (M|Without SDOFs )} (7) higher response ratios are computed for s1 = 0 m, compared to s1 =
θ = 225 o 30 m, as well as for the cases, where six buildings-SDOFs are considered
where max{env (M|With SDOFs )} is the maximum envelope value of the
at the ground surface.
dynamic bending moment computed at a critical section of the tunnel
lining (i.e. θ = 225°, as per Fig. 17) when the buildings-SDOFs are
θ = 225o
considered, and max{env (M|Without SDOFs )} is the maximum envelope 4.5. Lining dynamic axial forces
value of the dynamic bending moment, predicted at the same location
in the absence of the buildings-SDOFs. The computed response ratios The axial response of the lining of a circular tunnel under the stress
were plotted against the soil-tunnel flexibility ratios, F, for all the ex- field that is induced by the surrounding ground during ground shaking
amined single tunnels and the left track tunnels of the twin tunnels. is generally much higher compared to its bending response -if not
Fig. 17 portrays response ratios rM computed for a visco-elastic soil predominant. Along these lines, the effect of the buildings-SDOFs at the
response. Generally, the consideration of the buildings-SDOFs results in ground surface was found to be generally much higher on the lining
an increase of the envelope dynamic bending moment at the examined dynamic axial forces, compared to the lining dynamic bending mo-
section, i.e. μΜ > 1.0. In line with the previous results, higher response ments. This is clearly illustrated in Fig. 19, where representative com-
ratios are computed for shallow tunnels (i.e. for h = 10 m), as well as parisons of maximum and minimum envelope distributions of the lining
for shaking motion EQ2. An increasing trend of the response ratios with dynamic axial forces, computed along the perimeter of single tunnels,
increasing soil-tunnel flexibility ratios, F, is also observed. Actually, for in the absence or presence of buildings-SDOFs at the ground surface, are
very flexible shallow tunnels and six buildings-SDOFs at the ground presented and compared. The results refer to tunnels embedded in the
surface, the computed response ratios can be as high as 1.4 and 2.3, for soil deposit C2 and are computed for shaking motion EQ1. The dis-
ground shaking motions EQ1 and EQ2, respectively. Additionally, si- tributions computed in the absence of the buildings-SDOFs and under
milar response ratios are computed for the single tunnel and the left the assumption of visco-elastic soil response are symmetric and are
track tunnel of the twin tunnel. Furthermore, higher response ratios are maximized at θ = ± 45° (Fig. 19a and c). The consideration of the
computed for the cases where six buildings-SDOFs are considered at the buildings-SDOFs at the ground surface results in a significant increase
ground surface, while the effect of relevant position of the buildings- of the envelope distributions, particularly in cases of shallow tunnels

188
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

2.3 2.3 2.3 2.3

1.725 1.725 1.725 1.725

M

M
1.15 1.15 1.15 1.15

0.575 0.575 0.575 0.575

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

2.3 2.3 2.3 2.3
1.725 1.725 1.725 1.725
M

1.15 1.15

M
1.15 1.15
0.575 0.575 0.575 0.575
0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 17. Response ratios μM of maximum envelope lining dynamic bending moments computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth
h = 10.0 m and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d)
tunnel’s burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elastic soil response (VE soil model).

(i.e. for h = 10 m). This increasing response is evident, regardless of the with findings of recent experimental and numerical studies (e.g.
relative position of the buildings at the ground surface to the tunnel axis Lanzano et al., 2012; Tsinidis et al., 2014, 2016c), is affected by the
(i.e. for both s1 = 0 m and 30 m). number and position of the buildings-SDOFs, as the latter affect the
Similar to the lining dynamic bending moments, the yielding re- extent of soil yielding around the tunnel. This explains the significant
sponse of the soil around the tunnel and the associated soil stress re- differences between the distributions predicted for various assumptions
distributions, result in residual values on the dynamic lining axial regarding the number and position of the buildings-SDOFs when an
forces, which in turn lead to complex forms for the distributions of the elasto-plastic soil response is considered (Fig. 19b and d).
envelope lining dynamic axial forces. This response, which is in line Similar to the previous response parameters, a series of response

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

4.2 4.2 4.2 4.2

3.15 3.15 3.15 3.15

M

2.1 2.1 2.1 2.1

1.05 1.05 1.05 1.05

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

4.2 4.2 4.2 4.2

3.15 3.15 3.15 3.15

M

2.1 2.1 2.1 2.1

1.05 1.05 1.05 1.05

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 18. Response ratios μM of maximum envelope dynamic bending moments computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth
h = 10.0 m and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d)
tunnel’s burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elasto-plastic soil response (KH soil model).

189
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

s 1 =0m, h=10m s 1 =0m, h=20m s 1 =0m, h=10m s 1 =0m, h=20m

500 500 200 450
250 250 100 225
N (kN/m)

N (kN/m)

N (kN/m)

N (kN/m)
0 0 0 0
-250 -250 -100 -225

-500 -500 -200 -450

0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360
(o ) o
( ) o
( ) (o )

s 1 =30m, h=10m s 1 =30m, h=20m s 1 =30m, h=10m s 1 =30m, h=20m

300 300 300 300
150 150 150 150
N (kN/m)

N (kN/m)

N (kN/m)

N (kN/m)
0 0 0 0
-150 -150 -150 -150
-300 -300 -300 -300
0 90 180 270 360 0 90 180 270 360 0 90 180 270 360 0 90 180 270 360
(o ) (o ) (o ) (o )

no SDOFs 2 SDOFs 6 SDOFs

Fig. 19. Maximum and minimum envelope distributions of lining dynamic axial forces computed along the perimeter of the single tunnel under shaking motion EQ1
for (a) visco-elastic soil response and s1= 0 m, (b) visco-elasto-plastic soil response and s1 = 0 m, (c) visco-elastic soil response and s1 = 30 m, (d) visco-elasto-plastic
soil response and s1 = 30 m.

ratios μN were computed for a critical lining section of the examined Fig. 20 portrays the response ratios μN computed for a visco-elastic
single tunnels and left track tunnels of the twin tunnels, as follows: soil response. The consideration of the buildings-SDOFs results in a
o o significant increase of the maximum envelope dynamic axial force at
θ = 225 θ = 225
μN = max{env (N |With SDOFs )}/max{env (N |Without SDOFs )} (8) the examined section, compared to the one predicted for a ‘green-field’
θ = 225o condition (i.e. μΝ > 1.0). Furthermore, a clear increasing trend of the
where max{env (N |With SDOFs )} is the maximum envelope value of the
computed response ratios with increasing flexibility ratios, F, is ob-
lining dynamic axial force computed at a critical section of the tunnel
served. For the cases where six buildings-SDOFs are considered at the
lining (i.e. θ = 225°, as per Fig. 20) when the buildings-SDOFs are
θ = 225o ground surface, the response ratios, computed for shaking motion EQ1,
considered, and max{env (N |Without SDOFs )} is the maximum envelope
reach a maximum of 7.5 and 2.4 for shallow (i.e. for h = 10 m) and
lining dynamic axial force computed at the same location in the absence
deep (i.e. for h = 20 m) flexible tunnels, respectively. Much higher
of the buildings-SDOFs.

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

10 10 30 30

7.5 7.5 22.5 22.5

N

5 5 15 15

2.5 2.5 7.5 7.5

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

3 3 20 20

2.25 2.25 15 15
N

1.5 1.5 10 10

0.75 0.75 5 5

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 20. Response ratios μN of maximum envelope lining dynamic axial forces computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth h = 10.0 m
and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d) tunnel’s
burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elastic soil response (VE soil model).

190
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

3 3 4 4

2.25 2.25 3 3
N

N
1.5 1.5 2 2

0.75 0.75 1 1

0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

2 SDOFs 6 SDOFs 2 SDOFs 6 SDOFs

2 2 2 2
1.5 1.5 1.5 1.5
N

1 1

N
1 1
0.5 0.5 0.5 0.5
0 0 0 0
0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70 0 17.5 35 52.5 70
F F F F

s1 = 0 m, single tunnel s1 = 0 m, twin tunnel - left track

s1 = 30 m, single tunnel s1 = 30 m, twin tunnel - left track = 1.0

Fig. 21. Response ratios μN of maximum envelope lining dynamic axial forces computed at a critical lining section (θ = 225ο) for (a) tunnel’s burial depth h = 10.0 m
and shaking motion EQ1 (b) tunnel’s burial depth h = 10.0 m and shaking motion EQ2, (c) tunnel’s burial depth h = 20.0 m and shaking motion EQ1, (d) tunnel’s
burial depth h = 20.0 m and shaking motion EQ2, assuming a visco-elasto-plastic soil response (KH soil model).

response ratios are computed for shaking motion EQ2, reaching a dynamic earth pressures and the soil dynamic shear stresses developed
maximum of 30. Slightly higher response ratios (as high as 10–15%) are along the perimeter of the tunnel, as well as the dynamic lining forces,
observed in cases of shallow single tunnels compared to those predicted was quantified on the basis of response ratios μ. The key findings of this
for the left track tunnel of the twin tunnel. The differences between the study are summarized in the following:
two tunnel cases are negligible for deep tunnels (i.e. for h = 20 m). The
number and position of the buildings-SDOFs on the ground surface is • For a visco-elastic soil response (VE model), a rocking-uplifting re-
found to affect the computed response ratios, to some extent. Similar to sponse of the buildings-SDOFs was observed, which was amplified
the response parameters discussed above, there are cases, where higher for more severe ground shaking (i.e. shaking motion EQ2) and for
response ratios are computed for s1 = 30 m compared to s1 = 0 m. stiffer soil deposits. This response was reduced when an elasto-
Considerably lower response ratios rN are computed at the in- plastic soil response was considered (KH model). In the latter case, a
vestigated lining section, when a visco-elasto-plastic soil response is settlement of the footings of the buildings-SDOFs was actually ob-
considered (KH model, Fig. 21). In particular, the response ratios reach served, associated with the yielding of the foundation soil. The
a maximum of 2.5 and 3.6 for shallow tunnels (i.e. for h = 10 m), when above deformation patterns, affected the stress state and the
subjected to shaking motions EQ1 and EQ2, respectively, while lower yielding patterns of the soil (the latter predicted by the elasto-plastic
values are predicted for the deep tunnels (i.e. for h = 20 m). In most analyses) in the adjacent area, generally leading to a higher seismic
cases, a slight increase of the response ratio is observed with increasing response of the adjacent tunnels, compared to the one predicted in
flexibility ratio, F. Additionally, slightly higher response ratios are the absence of the buildings. Indeed, higher ovaling distortions,
computed on the single tunnel compared to the left track tunnel of the dynamic earth pressures, soil dynamic shear stresses and dynamic
twin tunnel, indicating a slight effect of the right track tunnel on the lining forces were predicted on the tunnels, in the presence of the
envelope dynamic axial force developed at the investigated section of buildings-SDOFs, in most of examined cases.
the left track tunnel. Similar to the observations made for the dynamic • For a visco-elastic soil response (VE model), the response ratios
bending moments, the number and position of the buildings-SDOFs at computed for the dynamic earth pressures, the soil dynamic shear
the ground surface are found to affect the computed response ratios for stresses and the dynamic lining forces, were generally increased
the envelope lining dynamic axial forces, with higher values being re- with the increase of the flexibility ratio, F, as well as with the de-
ported in cases of s1 = 0 m and for six buildings-SDOFs at the ground crease of the tunnel’s burial depth and the increase of the shaking
surface. motion severity. The number and position of the buildings-SDOFs at
the ground surface was found to affect the response ratios. In most
5. Conclusions cases, the scenario of six buildings-SDOFs resulted in a higher re-
sponse for the tunnel. Interestingly, there were cases where higher
The effect of heavy buildings at the ground surface on the trans- response ratios were computed for s1 = 30 m, compared to s1 =
versal seismic response of adjacent shallow circular tunnels was in- 0 m. The latter, should be attributed to the complex effects of the
vestigated in this study by means of numerical analyses in ABAQUS. oscillation of the buildings-SDOFs at ground surface during ground
Several soil-tunnel-buildings configurations were examined, including shaking, on the stress state of the ground, surrounding the tunnels.
single or twin tunnels, embedded in clayey soil deposits, with the high- • Similarly, an increased tunnel response was observed in the pre-
rise buildings being simulated in a simplified fashion as equivalent sence of the buildings-SDOFs, when a visco-elasto-plastic soil re-
SDOFs, founded on raft foundations. The effect of buildings-SDOFs on sponse was considered. The residual values that were presented on
critical parameters of the dynamic response of tunnels, including the the dynamic earth pressures, the soil dynamic shear stresses and the

191
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

dynamic lining forces in these cases, as a result of the soil yielding Geotech. Geoenviron. Eng. 133 (9), 1067–1090.
around the tunnel, led to more complex forms for the envelope Anastasopoulos, I., Gerolymos, N., Drosos, V., Georgarakos, T., Kourkoulis, R., Gazetas,
G., 2008. Behavior of deep immersed tunnel under combined normal fault rupture
distributions of the above response parameters around the tunnel, deformation and subsequent seismic shaking. Bull. Earthq. Eng. 6 (2), 213–239.
compared to the visco-elastic analyses predictions. The existence of Anastasopoulos, I., Georgarakos, T., Georgiannou, V., Drosos, V., Kourkoulis, R., 2010.
buildings-SDOFs at the ground surface resulted in a general increase Seismic performance of bar-mat reinforced-soil retaining wall: shaking table testing
versus numerical analysis with modified kinematic hardening constitutive model.
of the soil yielding patterns, which further add to the complexity of Soil Dyn. Earthq. Eng. 30 (10), 1089–1105.
the above envelope distributions. This explains, at least to some Anastasopoulos, I., Gelagoti, F., Kourkoulis, R., Gazetas, G., 2011. Simplified constitutive
extent, the complex correlations observed between the computed model for simulation of cyclic response of shallow foundations: validation against
laboratory tests. J. Geotech. Geoenviron. Eng. ASCE 137 (12), 1154–1168.
response ratios μ and the soil-tunnel flexibility ratios, F. Generally Anastasopoulos, I., Loli, M., Gelagoti, F., Kourkoulis, R., Gazetas, G., 2012. Nonlinear
higher response ratios were computed for six buildings-SDOFs at the soil–foundation interaction: numerical analysis. In: Proceedings 2nd International
ground surface, as well as for s1 = 0 m. Conference on Performance-based design in Earthquake Geotechnical Engineering,

• The comparisons between the response parameters, computed for Taormina, Italy, 28–30 May 2012.
Anderson, D.G., Martin, G.R., Lam, I., Wang, J.N., 2008. NCHPR611: Seismic analysis and
the single tunnel and the left track tunnel of the twin tunnel (located design of retaining walls, buried structures, slopes and embankments. National
at the same position), indicated a negligible effect of the response of Cooperative Highway Research Program, Transportation Research Board,
the right track tunnel of the twin tunnel on the left track, for a visco- Washington, DC, USA.
Bardet, J.B., Ichii, K., Lin, C.H., 2000. EERA: A Computer Program for Equivalent-linear
elastic soil response. For a visco-elasto-plastic soil response, a minor Earthquake Site Response Analyses of Layered Soil Deposits. University of Southern
effect was observed, particularly for the dynamic lining forces. California, Department of Civil Engineering, Los Angeles.
Baziar, M.H., Moghadam, M.R., Kim, D.-S., Choo, Y.W., 2014. Effect of underground
tuennl on the ground surface acceleration. Tunn. Undergr. Space Technol. 44, 10–22.
The main conclusion of this work is that the seismic response of Baziar, M.H., Moghadam, M.R., Choo, Y.W., Kim, D.-S., 2016. Tunnel flexibility effect on
circular tunnels in urban areas can be affected by the presence of the the ground surface acceleration response. Earthq. Eng. Eng. Vibration 15, 457–476.
buildings at the ground surface. The shallower the tunnel the higher the Bilotta, E., Lanzano, G., Madabhushi, S.P.G., Silvestri, F., 2014. A numerical Round Robin
on tunnels under seismic actions. Acta Geotech. 9 (4), 563–579.
effects are expected to be. Numerical models, similar to those presented Bobet, A., 2010. Drained and undrained response of deep tunnels subjected to far-field
herein, may be used during final stages of analysis, in order to quantify shear loading. Tunn. Undergr. Space Technol. 25 (1), 21–31.
the effect of existing buildings on the seismic response of a newly de- Bobet, A., Fernandez, G., Huo, H., Ramirez, J., 2008. A practical iterative procedure to
estimate seismic-induced deformations of shallow rectangular structures. Can.
signed tunnel, particularly when the latter crosses densely urbanized Geotech. J. 45, 923–938.
areas. Although it is common in practice to use a visco-elastic model to CEN, 2004. Eurocode 8: Design of structures for earthquake resistance - part 1: General
describe the soil response under ground shaking, its implementation rules, seismic actions and rules for buildings. EN1998-1: 2004, European Committee
for Standardization, Brussels.
may not capture the significant effect of soil yielding on the tunnel
Chen, G., Wang, Z., Zuo, X., Du, X., Gao, H., 2013. Shaking table test on seismic failure
response. Hence, the implementation of an adequately calibrated non- characteristics of a subway station structure in liquefiable ground. Earthq. Eng.
linear soil model seems to be more rational, particularly in cases where Struct. Dyn. 42, 1489–1507.
strong soil yielding is anticipated. Chian, S.C., Madabhushi, S.P.G., 2012. Effect of buried depth and diameter on uplift of
underground structures in liquefied soils. Soil Dyn. Earthq. Eng. 41, 181–190.
Chou, J.C., Kutter, B.L., Travasarou, T., Chacko, J.M., 2010. Centrifuge modeling of
Acknowledgements seismically induced uplift for the BART transbay tube. J. Geotech. Geoenviron. Eng.
137 (8), 754–765.
Cilingir, U., Madabhushi, S.P.G., 2011a. A model study on the effects of input motion on
The author sincerely thanks the State Scholarships Foundation of the seismic behavior of tunnels. Soil Dyn. Earthq. Eng. 31, 452–462.
Greece (IKY) for providing financial support for this research, within Cilingir, U., Madabhushi, S.P.G., 2011b. Effect of depth on the seismic response of square
the IKY Fellowships of Excellence for Postgraduate Studies in Greece - tunnels. Soils Found. 51 (3), 449–457.
Cilingir, U., Madabhushi, S.P.G., 2011c. Effect of depth on the seismic response of circular
SIEMENS Program (2016–2017). Additionally, he gratefully acknowl- tunnels. Can. Geotech. J. 48 (1), 117–127.
edges Professor Kyriazis Pitilakis (Aristotle University of Thessaloniki, Dashti, S., Hashash, Y.M.A., Gillis, K., Musgrove, M., Walker, M., 2016. Development of
AUTH) for the fruitful discussions over the course of this study, as well dynamic centrifuge models of underground structures near tall buildings. Tunn.
Undergr. Space Technol. 86, 89–105.
as Mrs. Agoritsa-Maria Nteri (Civil Engineer, MSc, AUTH) and Mrs.
De Barros, F.C.P., Luco, J.E., 1993. Diffraction of obliquely incident waves by a cylind-
Varvara Kyriakidou (Civil Engineer, MSc, AUTH) for their help in the rical cavity embedded in a layered viscoelastic half space. Soil Dyn. Earthq. Eng. 12,
development of some of the numerical models presented herein. 159–171.
Debiasi, E., Gajo, A., Zonta, D., 2013. On the seismic response of shallow-buried rec-
tangular structures. Tunn. Undergr. Space Technol. 38, 99–113.
References FHWA (Federal Highway Administration), 2009. Technical manual for design and con-
struction of road tunnels-Civil elements. Publication No. FHWA-NHI-10-034,
ABAQUS, 2012. ABAQUS: theory and analysis user’s manual, version 6.12. Dassault Department of transportation, Federal Highway Administration, Washington D.C.,
Systèmes SIMULIA, Providence, RI, USA. U.S.
Abate, G., Massimino, M.R., Maugeri, M., 2015. Numerical modelling of centrifuge tests Giannakos, S., Gerolymos, N., Gazetas, G., 2012. Cyclic lateral response of piles in dry
on tunnel-soil systems. Bull. Earthq. Eng. 13 (7), 1927–1951. sand: finite element modeling and validation. Comput. Geotech. 44, 116–131.
Abate, G., Massimino, M.R., 2017a. Numerical modelling of the seismic response of a Guo, J., Chen, J., Bobet, A., 2013. Influence of a subway station on the inter-story drift
tunnel–soil–aboveground building system in Catania (Italy). Bull. Earthq. Eng. 15 (1), ratio of adjacent surface structures. Tunn. Undergr. Space Technol. 35, 8–19.
469–491. Hashash, Y.M.A., Hook, J.J., Schmidt, B., Yao, J.I.-C., 2001. Seismic design and analysis
Abate, G., Massimino, M.R., 2017b. Parametric analysis of the seismic response of cou- of underground structures. Tunn. Undergr. Space Technol. 16 (2), 247–293.
pled tunnel–soil–aboveground building systems by numerical modelling. Bull. Hashash, Y.M.A., Park, D., 2002. Viscous damping formulation and high frequency mo-
Earthq. Eng. 15 (1), 443–467. tion propagation in non-linear site response analysis. Soil Dyn. Earthq. Eng. 22 (7),
Abuhajar, O., El Naggar, H., Newson, T., 2015a. Seismic soil-culvert interaction. Can. 611–624.
Geotech. J. 52, 1–19. Hashash, Y.M.A., Park, D., Yao, J.I.C., 2005. Ovaling deformations of circular tunnels
Abuhajar, O., El Naggar, H., Newson, T., 2015b. Experimental and numerical investiga- under seismic loading, an update on seismic design and analysis of underground
tions of the effect of buried box culverts on earthquake excitation. Soil Dyn. Earthq. structures. Tunn. Undergr. Space Technol. 20 (5), 435–441.
Eng. 79, 130–148. Hashash, Y.M.A., Karina, K., Koutsoftas, D., O'Riordan, N., 2010. Seismic design con-
Alielahi, H., Adampira, M., 2016. Seismic effects of two-dimensional subsurface cavity on siderations for underground box structures. In: Proceedings of Earth Retention
the ground motion by BEM: amplification patterns and engineering applications. Int. Conference - Earth Retention Conference 3, ASCE, 620-637, Bellevue, WA, USA.
J. Civ. Eng. 14 (4), 233–251. Hatzigeorgiou, G.D., Beskos, D.E., 2010. Soil-structure interaction effects on seismic in-
Alielahi, H., Kamalian, M., Adampira, M., 2015. Seismic ground amplification by unlined elastic analysis of 3-D tunnels. Soil Dyn. Earthq. Eng. 30 (9), 851–861.
tunnels subjected to vertically propagating SV and P waves using BEM. Soil Dyn. Huo, H., Bobet, A., Fernández, G., Ramírez, J., 2005. Load transfer mechanisms between
Earthq. Eng. 71, 63–79. underground structure and surrounding ground: evaluation of the failure of the
Amorosi, A., Boldini, D., 2009. Numerical modeling of the transverse dynamic behavior of Daikai station. J. Geotech. Geoenviron. Eng. 131 (12), 1522–1533.
circular tunnels in clayey soils. Soil Dyn. Earthq. Eng. 59 (6), 1059–1072. Huo, H., Bobet, A., Fernández, G., Ramírez, J., 2006. Analytical solution for deep rec-
Anastasopoulos, I., Gerolymos, N., Drosos, V., Kourkoulis, R., Georgarakos, T., Gazetas, tangular structures subjected to far-field shear stresses. Tunn. Undergr. Space
G., 2007. Nonlinear response of deep immersed tunnel to strong seismic shaking. J. Technol. 21 (6), 613–625.
Hushmand, A., Dashti, S., Davis, C., McCartney, J.S., Hushmand, B., 2016. A centrifuge

192
G. Tsinidis Tunnelling and Underground Space Technology 76 (2018) 177–193

study of the influence of site response, relative stiffness, and kinematic constraints on Eng. 67, 1–15.
the seismic performance of buried reservoir structures. Soil Dyn. Earthq. Eng. 88, Priestley, M.J.N., Kowalsky, M.J., 2000. Direct displacement-based seismic design of
427–438. concrete buildings. Bull. New Zealand Soc. Earthq. Eng. 33 (4), 421–444.
Hwang, J.H., Lu, C.C., 2007. Seismic capacity assessment of old Sanyi railway tunnels. Shibayama, S., Izawa, J., Takahashi, A., Takemura, J., Kusakabe, O., 2010. Observed
Tunn. Undergr. Space Technol. 22, 433–449. behavior of a tunnel in sand subjected to shear deformation in a centrifuge. Soils
Iida, H., Hiroto, T., Yoshida, N., Iwafuji, M., 1996. Damage to Daikai subway station. In: Found. 50 (2), 281–294.
Special issue on geotechnical aspects of the January 17 1995 Hyogoken–Nanbu Smerzini, C., Aviles, J., Paolucci, R., Sanchez-Sesma, F.J., 2009. Effect of underground
earthquake. Soils and Foundations, pp. 283–300. cavities on surface earthquake ground motion under SH wave propagation. Earthq.
Ishibashi, I., Zhang, X., 1993. Unified dynamic shear moduli and damping ratios of sand Eng. Struct. Dyn. 38, 1441–1460.
and clay. Soils Found. 33 (1), 182–191. Tsinidis, G., 2017. Response characteristics of rectangular tunnels in soft soil subjected to
ISO (International Organization for Standardization), 2005. ISO 23469: Bases for design transversal ground shaking. Tunn. Undergr. Space Technol. 62, 1–22.
of structures - Seismic actions for designing geotechnical works. International Tsinidis, G., Pitilakis, K., Trikalioti, A.D., 2014. Numerical simulation of round robin
Standard ISO TC98/SC3/WG10. Geneva, Switzerland: International Organization for numerical test on tunnels using a simplified kinematic hardening model. Acta
Standardization. Geotech. 9 (4), 641–659.
Kontoe, S., Zdravkovic, L., Potts, D., Mentiki, C., 2011. On the relative merits of simple Tsinidis, G., Pitilakis, K., Madabhushi, G., Heron, C., 2015. Dynamic response of flexible
and advanced constitutive models in dynamic analysis of tunnels. Geotechnique 61 square tunnels: centrifuge testing and validation of existing design methodologies.
(10), 815–829. Geotechnique 65 (5), 401–417.
Kontoe, S., Avgerinos, V., Potts, D.M., 2014. Numerical validation of analytical solutions Tsinidis, G., Pitilakis, K., Madabhushi, G., 2016a. On the dynamic response of square
and their use for equivalent-linear seismic analysis of circular tunnels. Soil Dyn. tunnels in sand. Eng. Struct. 125, 419–437.
Earthq. Eng. 66, 206–219. Tsinidis, G., Rovithis, E., Pitilakis, K., Chazelas, J.L., 2016b. Seismic response of box-type
Kouretzis, G., Bouckovalas, G., Sofianos, A., Yiouta-Mitra P., 2007. Detrimental effects of tunnels in soft soil: experimental and numerical investigation. Tunn. Undergr. Space
urban tunnels on design seismic ground motions. In: Proceedings of the 2nd Technol. 59, 199–214.
Japan–Greece Workshop on Seismic Design, Observation, and Retrofit of Tsinidis, G., Pitilakis, K., Anagnostopoulos, C., 2016c. Circular tunnels in sand: dynamic
Foundations. Tokyo, Japan, April 3–4, 2007. response and efficiency of seismic analysis methods at extreme lining flexibilities.
Lanzano, G., Bilotta, E., Russo, G., Silvestri, F., Madabhushi, S.P.G., 2012. Centrifuge Bull. Earthq. Eng. 14 (10), 2903–2929.
modeling of seismic loading on tunnels in sand. Geotech. Test. J. 35 (6), 854–869. Tsinidis, G., Pitilakis, K., 2018. Improved R-F relations for the transversal seismic analysis
Lanzano, G., Bilotta, E., Russo, G., Silvestri, F., 2015. Experimental and numerical study of rectangular tunnels. Soil Dyn. Earthq. Eng. 107, 48–65.
on circular tunnels under seismic loading. Eur. J. Environ. Civ. Eng. 19 (5), 539–563. Ulgen, D., Saglam, S., Ozkan, M.Y., 2015. Dynamic response of a flexible rectangular
Lee, V.W., Karl, J., 1992. Diffraction of SV-waves by underground, circular, cylindrical underground structure in sand: centrifuge modeling. Bull. Earthq. Eng. 13,
cavities. Soil Dyn. Earthq. Eng. 11, 445–456. 2547–2566.
Lu, C.C., Hwang, J.H., 2017. Implementation of the modified cross-section racking de- Wang, J.N., 1993. Seismic Design of Tunnels: A Simple State of the Art Design Approach.
formation method using explicit FDM program: a critical assessment. Tunn. Undergr. Parsons Brinckerhoff Inc., New York.
Space Technol. 68, 58–73. Wang, H.-F., Lou, M.L., Chen, X., Zhai, Y.M., 2013. Structure–soil–structure interaction
Lysmer, J., Kuhlemeyer, R.L., 1969. Finite dynamic model for infinite media. J. Eng. between underground structure and ground structure. Soil Dyn. Earthq. Eng. 54,
Mech. Div., ASCE 95 (4), 859–878. 31–38.
Manoogian, M.E., Lee, V.W., 1996. Diffraction of SH-waves by subsurface inclusions of Wang, G., Yuan, M., Ma, X., Wu, J., 2017. Numerical study on the seismic response of the
arbitrary shape. J. Eng. Mech. Div. ASCE 122, 123–129. underground subway station- surrounding soil mass-ground adjacent building
Navarro, C., 1992. Effect of adjoining structures on seismic response of tunnels. Int. J. system. Front. Struct. Civ. Eng.. (Online Press). http://dx.doi.org/10.1007/s11709-
Numer. Anal. Meth. Geomech. 16 (11), 797–814. 016-0381-7.
Park, K.H., Tantayopin, K., Tontavanich, B., Owatsiriwong, A., 2009. Analytical solution Wood, M., 1975. The circular tunnel in elastic ground. Geotechnique 1, 115–127.
for seismic-induced ovaling of circular tunnel lining under no-slip interface condi- Yiouta-Mitra, P., Kouretzis, G., Bouckovalas, G., Sofianos, A. 2007. Effect of underground
tions: a revisit. Tunn. Undergr. Space Technol. 24 (2), 231–235. structures in earthquake resistant design of surface structures. In Proceedings: Geo-
Parvanova, S., Divena, P., Manolis, G.D., 2014a. Elastic wave fields in a half-plane with Denver 2007 Conference. Denver, Colorado, Feb. 18–21.
free-surface relief, tunnels and multiple buried inclusions. Acta Mech. 225 (7), Yuan, Y., Yu, H.T., Li, C., Yan, X., Yuan, J.Y., 2016 Multi-point shaking table test for long
1843–1865. tunnels subjected to non-uniform seismic loadings – Part I: theory and validation. Soil
Parvanova, S., Divena, P.S., Manolis, D.G., Wuttke, F., 2014b. Seismic response of lined Dyn. Earthq. Eng., 2016. (Online Press), http://www.10.1016/j.soildyn.2016.08.
tunnels in the half-plane with surface topography. Bull. Earthq. Eng. 12 (2), 017i.
981–1005. Yu, H.T., Yuan, Y., Xu, G.P., Su, Q.K., Yan, X., Li, C., 2016a. Multi-point shaking table test
Penzien, J., 2000. Seismically induced racking of tunnel linings. Earthq. Eng. Struct. Dyn. for long tunnels subjected to non-uniform seismic loadings – Part II: application to the
29, 683–691. HZM immersed tunnel. Soil Dyn. Earthq. Eng., (Online Press), http://www.10.1016/
Pitilakis, K., Tsinidis, G., 2014. Performance and seismic design of underground struc- j.soildyn.2016.08.01.
tures. In: Maugeri, M., Soccodato, C. (Eds.), Earthquake Geotechnical Engineering Yu H.T., Cai C., Guan X.F., Yuan Y., 2016b. Analytical solution for long lined tunnels
Design, Geotechnical Geological and Earthquake Engineering 28. Springer subjected to travelling loads. Tunnelling and Underground Space Technology, 2016b,
International Publishing, Switzerland, pp. 279–340. 58: 209-215.
Pitilakis, K., Tsinidis, G., Leanza, A., Maugeri, A., 2014. Seismic behaviour of circular Yu, H., Yuan, Y., Bobet, A., 2017. Seismic analysis of long tunnels: a review of simplified
tunnels accounting for above ground structures interaction effects. Soil Dyn. Earthq. and unified methods. Undergr. Space 2, 73–87.

193