infrastructures

Article
Two-Dimensional Numerical Analysis for TBM
Tunneling-Induced Structure Settlement: A Proposed Modeling
Method and Parametric Study
Rashad Alsirawan 1, *, Ashraf Sheble 2 and Ammar Alnmr 1

1 Department of Structural and Geotechnical Engineering, Széchenyi István University, 9026 Győr, Hungary;
[email protected]
2 Department of Geotechnical Engineering, Tishreen University, Latakia P.O. Box 2230, Syria;
[email protected]
* Correspondence: [email protected]

Abstract: The construction of tunnels in densely populated urban areas poses a significant challenge
in terms of anticipating the settlement that may result from tunnel excavation. This paper presents
a new and more realistic modeling method for tunnel excavation using a Tunnel Boring Machine
(TBM). This method is compared with other reference modeling methods using a validated model of
a subsurface tunnel excavated by a TBM with a slurry shield. A parametric study is conducted to
investigate the impact of key parameters, including structure width, foundation depth, eccentricity,
load on the structure, overburden depth, and tunnel diameter, on tunnel–soil–structure interaction
and the resulting structure settlements. The results reveal that the tunnel diameter, eccentricity,
and overburden depth have a significant impact on structure settlements, with values of 22.5%,
17%, and 7.1%, respectively. Finally, the paper proposes an equation for predicting the maximum
settlement of a structure, considering the critical parameters. The validity of the equation is evaluated
by comparing its results with the outputs from various case studies, including a newly validated
model, two real-life case studies, and centrifuge tests. The results indicate a high level of consistency
between the calculated and measured settlements.
Citation: Alsirawan, R.; Sheble, A.;
Alnmr, A. Two-Dimensional
Keywords: settlement trough; realistic modeling method; TBM modeling; TBM with slurry shield;
Numerical Analysis for TBM
parametric study; soil–structure interaction; equation of settlement estimation
Tunneling-Induced Structure
Settlement: A Proposed Modeling
Method and Parametric Study.
Infrastructures 2023, 8, 88.
https://doi.org/10.3390/ 1. Introduction
infrastructures8050088 The rapid growth of cities, followed by congestion and major traffic challenges, ne-
Academic Editor: Francesca Dezi
cessitates the extension of existing highways as well as the construction of new roads and
bridges. One of the solutions is to exploit the underground space through the construc-
Received: 6 April 2023 tion of tunnel networks. Tunnel excavation is associated with the deformations of the
Revised: 24 April 2023 soil surrounding the tunnel called ‘volume loss’. These deformations reach the surface,
Accepted: 3 May 2023 leading to a settlement trough, which causes damage to the surface structures [1]. These
Published: 5 May 2023 soil deformations in the case of Tunnel Boring Machine (TBM) tunneling are created by
soil movement towards the excavation face, soil movement in the radial direction resulting
from the over-excavating of the soil around the tunnel, and the gap left by the tail of the
shield, in addition to the subsequent deformation of the tunnel lining and the consolidation
Copyright: © 2023 by the authors.
that occurs in the case of clay soils [2].
Licensee MDPI, Basel, Switzerland.
This article is an open access article
By adopting a 2D analysis, several methods were presented to model the volume loss
distributed under the terms and
around the tunnel, which constitutes the main factor resulting in the surface subsidence
conditions of the Creative Commons caused by tunneling. Vermeer and Brinkgreve [3] hypothesized a two-stage method where
Attribution (CC BY) license (https:// the soil within the tunnel is deactivated and the tunnel lining is activated in the first stage.
creativecommons.org/licenses/by/ In the second stage, a uniform contraction of the tunnel lining is applied that results in
4.0/). deformation of the surrounding soil towards the inside of the tunnel. Moller [4] also

Infrastructures 2023, 8, 88. https://doi.org/10.3390/infrastructures8050088 https://www.mdpi.com/journal/infrastructures

Infrastructures 2023, 8, 88 2 of 20

suggested a two-stage method. In the first stage, the soil within the tunnel is deactivated
and a radial pressure is applied on the excavation surface equal to the grout pressure, and
with depth it increases linearly by the weight unit of the grout. In the second stage, the
tunnel lining is activated and the grout pressure is deactivated. Suwansawat et al. [5]
recommended a modification to the grout pressure method for Moller [4], where the
volume loss can be modeled in three stages. In the first stage, the soil within the shield
is deactivated and the face support pressure is activated. In the second stage, the tunnel
lining is activated and the gap between the lining and the shield is filled with soft grout
material. Then, the face support pressure within the tunnel is deactivated and replaced by
the grout pressure in the gap between the lining and the shield. In the final stage, the grout
pressure is deactivated, and the soft grout is replaced with a hardened one.
Numerous parametric studies were conducted on the issue of the interaction between
soil, tunneling, and surface structures. Maleki [6] studied the effects of stiffness, weight
and geometry of a structure, as well as the eccentricity between the axis of the tunnel and
the axis of the structure on the ground surface settlements. Katebi [7] investigated the
influence of tunnel depth, structure width and weight, and the eccentricity on the internal
forces and deformations of the tunnel lining. Giardina [8] conducted a numerical study
on the consequences of the stiffness, the weight of the structure, and volume loss, based
on the results of the Farrell [9] centrifuge test. Son [10,11] independently completed two
numerical studies of a brick structure that was located over sandy and clayey soils; the
researcher studied the effect of changing the volume loss, the depth, and the diameter of the
tunnel on the settlement under the brick structure. Boldini [12] conducted an investigation
to demonstrate the impact of the number of stories on the settlement trough. Specifically,
the study examined how the incremental stiffness of the structure with more stories affects
the settlement trough and also investigated the effect of the incremental structure’s weight
on the settlement trough.
Peck [13] presented an empirical formula to calculate the surface settlement caused
by tunneling by adopting the greenfield condition through the normal distribution curve
(Gauss). O’Reilly and New [14] proposed an equation to calculate the trough width
parameter, which signals the distance between the axis of the tunnel and the point of the
maximum slope of the surface settlement trough. Mair et al. [2] assumed that the volume
of the settlement trough at the surface is equal to the volume loss around the tunnel in the
case of clay soil, and they showed a connection between them based on their assumption.
Herzog [15] presented an equation to calculate the maximum settlement above the axis
of the tunnel based on experimental results, while Chakeri and Unver [16] developed
the work of Herzog [15] and considered the effect of soil cohesion, internal friction angle,
and the required supporting pressure on the drilling face of the Earth Pressure Balance
shields (EPB). Wang [17] suggested an empirical equation to calculate the subsidence in the
case of sandy soil, taking into account the change in the volume of the settlement trough
with depth.
The majority of previous studies have primarily focused on the analysis of settlement
through greenfield conditions. Furthermore, the parametric study has been limited in its
ability to fully comprehend the behavior of settlement under structures, and the proposed
equations used to calculate settlements have been primarily based on greenfield conditions.
This highlights the need for further research to better understand the complexities of
settlement behavior under various conditions, including those that arise when structures
are present.
In this paper, a validated model of the subsurface Second Heinenoord tunnel [18]
is utilized to calibrate a new and improved modeling method of shield tunneling using
the Finite Element Method (FEM). The proposed method is aimed at achieving a more
precise and realistic simulation of the tunneling process by accurately representing each
of its individual stages. Additionally, an extensive parametric study is conducted to
explore the impact of adjacent structure existence on the deformation caused by tunneling
in comparison with greenfield conditions and to examine the effects of various critical
Infrastructures 2023, 8, 88 3 of 20

parameters on the settlement under adjacent structures. Finally, a new and tailored equation
is proposed that can be applied in the elementary analysis to calculate the maximum
settlements of adjacent structures.

2. Proposed Method: Grout Hardening Method

The proposed 2D modeling method is designed to provide a comprehensive and
interconnected understanding of the stages involved in tunnel excavation using TBM with
shields. By representing the work progress within the tail of the shield in a manner that
closely approximates reality and accounts for the most crucial factors causing volume loss
around the tunnel mentioned in Section 1, this method leads to more accurate development
of stresses and deformations in the tunnel lining. Additionally, the method enables a more
precise modeling of the deformation of the soil around the tunnel and the propagation of
this deformation towards the surface, resulting in the creation of a subsidence trough at the
surface that aligns closely with the reference field measurements. The tunnel excavation
process is systematically divided into four stages as illustrated in Figure 1, allowing for a
step-by-step analysis of the changes occurring during the construction process:
1. In the first stage, the soil within the shield is deactivated, and the tail of the shield
is activated.
2. In the second stage, the tail of the shield is deactivated, and the grout pressure
is activated. This enables us to take into account the radial deformations of the
surrounding soil towards the tunnel interior, as a result of the gap produced by the
over-excavation of the cutting wheel, the void left by the tail of the shield, and the
insufficient grout pressure to resist the soil pressure from above, which is close to the
geostatic stresses before executing the tunnel.
3. In the third stage, the grout pressure is deactivated, and the tunnel lining is activated
while filling the gap around the tunnel with grout material of an initial stiffness.
The stiffness of the grout material is then adjusted until the required volume loss is
achieved, resulting in a subsidence trough on the surface that is closest to the reference
field measurements.
4. In the last stage, the grout material with initial stiffness is replaced by the grout
material with final stiffness. As a result, the radial deformations of the soil towards
Infrastructures 2023, 8, x FOR PEER REVIEW the interior of the tunnel come to a halt, thereby causing 4 ofthe
21 subsidence trough to

cease developing.

Figure
Figure 1. Stages
1. Stages of the proposed
of the proposed method
method (grout (grout
hardening hardening
method). method).
3. Finite Element Modeling
In order to validate the proposed modeling method, a case study of the subsurface
Second Heinenoord tunnel [18] was utilized. This section of the study involves two key
aspects: (i) conducting a 2D numerical modeling of the tunnel using the conventional con-
traction method to determine the appropriate constitutive model, by comparing the pre-
Infrastructures 2023, 8, 88 4 of 20

3. Finite Element Modeling

In order to validate the proposed modeling method, a case study of the subsurface
Second Heinenoord tunnel [18] was utilized. This section of the study involves two key
aspects: (i) conducting a 2D numerical modeling of the tunnel using the conventional
contraction method to determine the appropriate constitutive model, by comparing the
predictions of numerical analysis using the Hardening Soil with Small-strain Stiffness (HSS)
and the Hardening Soil (HS), and (ii) comparing the results obtained from three distinct
modeling methods (grout pressure method, grout hardening method, and contraction
method) with field measurements, with the aim of validating the proposed method.

3.1. Description of the Second Heinenoord Tunnel

The Second Heinenoord tunnel was built in the Netherlands in 1996. The outer
diameter of the tunnel is Da = 8.3 m. A 7-m-long slurry shield with a permanent frontal
reinforcement was used to construct this tunnel. The slurry pressure at the top of the shield
was 230 kPa that increased linearly with depth by the weight unit of bentonite, which was
estimated to be approximately 15 kN/m3 . The grout pressure was 125 kPa that increased
linearly with depth by the same magnitude. The tunnel lining was made of prefabricated
reinforced concrete rings that were 0.35 m thick and 1.5 m long [18].
During the tunneling, several field measurement programs were carried out, including
the measuring of the settlements of the ground surface (Sv ) above the tunnel in a transverse
direction perpendicular to the tunnel axis (X), as shown in Table 1. The overburden depth
(H) = 12.5 m from the ground surface to the tunnel upper point, and the groundwater level
was 1.5 m. This tunnel is classified as a subsurface tunnel (H/D < 2). The ground surface is
horizontal. The properties and levels of the soil layers [18] are shown in Table 2.

Table 1. Measured transverse settlements [4].

X (m) 0 6.2 8.5 10.4 16.7 20

Sv (mm) 26.2 16.86 10.44 6.96 1.34 0.34

Table 2. Properties of soil layers surrounding the tunnel.

Layer Depth (m) γsat (kN/m3 ) ∅ (o ) c0 (kPa) ν

Fill 0–4 17.2 27 3 0.34
Sand 4–19.75 20 35 1 0.3
Sand 19.75–23.25 20 35 1 0.3
Sand-Clay 23.25–27.5 20 31 7 0.32
γsat : saturated unit weight, ∅ : internal friction angle, c0 : cohesion, ν: Poisson’s ratio.

For the purpose of validation, the field-measured settlements of the ground surface
during the construction of the Second Heinenoord tunnel had been compared to the cal-
culated settlements using the 2D numerical model of the tunnel. During this process,
two constitutive models, the Hardening Soil with Small-strain Stiffness (HSS) and the
Hardening Soil (HS), and three modeling methods had been applied, which are the grout
pressure method [19], the contraction method [20], and the proposed method (grout hard-
ening method).

3.2. Two-Dimensional Geometric Model

The continuous approach was implemented to model the tunnel and the surrounding
soil layers. Due to symmetry, only half of the model was simulated, with the geometric
dimensions of the 2D numerical model appropriately chosen to meet the German require-
ments [21]. Figure 2 shows the geometric dimensions of the model as well as the boundary
conditions, i.e., horizontal transitions are prohibited in the side boundaries, while vertical
and horizontal transitions are prohibited in the lower boundaries. Rotating and horizontal
transitions are also prohibited at tunnel lining nodes located on the symmetry axis. To
 The continuous approach was implemented to model the tunnel and the surrounding
soil layers. Due to symmetry, only half of the model was simulated, with the geometric
dimensions of the 2D numerical model appropriately chosen to meet the German require-
ments [21]. Figure 2 shows the geometric dimensions of the model as well as the boundary
conditions, i.e., horizontal transitions are prohibited in the side boundaries, while vertical
Infrastructures 2023, 8, 88 5 of 20
and horizontal transitions are prohibited in the lower boundaries. Rotating and horizontal
transitions are also prohibited at tunnel lining nodes located on the symmetry axis. To
determine the appropriate mesh size, a sensitivity analysis was conducted on the mesh.
determine
The the appropriate
simulation mesh mesh
utilized a coarse size, adistribution
sensitivity analysis was
with local conductedaton
refinement thethe mesh.line
surface
The simulation utilized a coarse mesh distribution with local refinement at the surface line
and tunnel line since the settlement behavior demonstrated little difference between
and tunnel line since the settlement behavior demonstrated little difference between coarse
coarse and medium distributions. A mesh of the model had 317 trigonometric elements,
and medium distributions. A mesh of the model had 317 trigonometric elements, each of
each of which consisted of 15 nodes.
which consisted of 15 nodes.

Figure Finite element

Figure 2. Finite elementmesh
meshofofthe
the2D
2Dmodel.
model.

In order to account for the impact of the joints between the prefabricated concrete
parts, the tunnel lining was modeled using plate elements with linear elastic behavior,
while the bend stiffness was reduced by dividing it by a reduction coefficient equal to
four [22]. Table 3 presents the characteristics of the lining.

Table 3. Lining properties.

Constitutive Model γ (kN/m3 ) EA (GN/m) EI (MNm2 ) ν t (cm)

Linear elastic 24 10.5 26.78 0.15 35
γ: unit weight, EA: axial stiffness, EI: flexural stiffness, t: thickness of lining.

3.2.1. Appropriate Constitutive Model

The proper validation of the problem is one of the most important issues to be taken
into consideration. In order to validate a model, the outputs of a numerical method are
assessed through a comparison with the field measurements or the results of software
packages that employ independent solutions [23]. The first crucial step in geotechnical
numerical modeling is selecting and calibrating a robust and accurate soil material model
(i.e., constitutive model). A less precise model may yield poor results and make estimating
the proper behavior in the field more complex [24].
The reference studies suggested the HS and the HSS models. Therefore, the authors
carried out a comparative study to simulate the behavior of the soil utilizing the contraction
method. Table 4 illustrates the parameters of both HS and HSS models applied in our
study. Compared with the field measurements, the HSS model yields the closest shape
to the transverse settlement trough as shown in Figure 3, which is consistent with the
results of Law [25]. Figure 4 demonstrates the vertical displacements in soil that are caused
by the contraction method and the HSS model. After having determined the optimal
constitutive model (HSS) that simulates the soil behavior more accurately, a comparison of
the predictions of the three modeling methods described above was conducted.
 3.2.1. Appropriate Constitutive Model
The proper validation of the problem is one of the most important issues to be taken
into consideration. In order to validate a model, the outputs of a numerical method are
assessed through a comparison with the field measurements or the results of software
Infrastructures 2023, 8, 88 packages that employ independent solutions [23]. The first crucial step in geotechnical 6 of 20
numerical modeling is selecting and calibrating a robust and accurate soil material model
(i.e., constitutive model). A less precise model may yield poor results and make estimating
Table 4. Parameters
the proper of HS
behavior inandtheHSS
fieldmodels.
more complex [24].
The reference studies suggested the HS and the HSS models. Therefore, the authors
ref ref ref ref
Eoed
carried out E50a comparative
Eur studyGto 0 simulate γ0.7the
(%) behavior m (-)of the soil utilizingRthe
Layer vur OCR intercontrac-
(MPa) (MPa) (MPa) (MPa)
tion method. Table 4 illustrates the parameters of both HS and HSS models applied in our
1 0.2 14study. Compared
14 with 42the field measurements,
52 0.0005 the HSS0.5 model yields 1 the closest 1 shape to
2 0.2 35 35 105 175 0.0005 0.5 1 0.67
3 0.2 35
the transverse35
settlement
105
trough175as shown 0.0005in Figure 3, which 0.5
is consistent
1
with the results
0.67
4 0.2 7of Law [25].
12 Figure 4 demonstrates
35 88 the vertical
0.0005 displacements
0.9 in soil
1 that are 1caused by
the contraction method and the HSS re f model. After having determined
re f the optimal consti-
vur : unloading/reloading Poisson’s ratio, Eoed : reference tangent stiffness, E50 : reference secant stiffness,
tutive
re f model (HSS) that simulates the re f soil behavior more accurately, a comparison of the
Eur : reference unloading/reloading stiffness, G0 : reference shear modulus at small strain, γ0.7 : reference strain
threshold, m: exponential power, OCR: over consolidationdescribed
predictions of the three modeling methods above reduction
ratio, Rint : strength was conducted.
factor.

Infrastructures 2023, 8, x FOR PEER REVIEW 7 of 21

Figure EffectofofHS
Figure3.3.Effect andHSS
HSand HSSonon
thethe
settlement trough.
settlement trough.

Figure 4.
Figure Vertical displacements
4. Vertical displacementsininsoil.
soil.
3.2.2. Comparison of the Modeling Methods
Table 4. Parameters of HS and HSS models.
Figure 5 shows that the curves of the transverse settlement trough, which were yielded
𝒓𝒆𝒇 𝒓𝒆𝒇 𝒓𝒆𝒇 𝒓𝒆𝒇
𝑬𝒐𝒆𝒅 using the𝑬three
𝟓𝟎 methods, 𝑬𝒖𝒓 𝑮𝟎 field measurements. However, the grout pressure
are close to the
Layer 𝒗𝒖𝒓 𝜸𝟎.𝟕 (%) m (-) OCR Rinter
(MPa)method (MPa)
gives a narrower transverse settlement
(MPa) (MPa) trough, while the contraction method offers
1 0.2 14 a wider one.
14 Finally, the proposed
42 method52 provides0.0005
the closest transverse
0.5 settlement
1 trough.
1
2 0.2 35 35 105 175 0.0005 0.5 1 0.67
3 0.2 35 35 105 175 0.0005 0.5 1 0.67
4 0.2 7 12 35 88 0.0005 0.9 1 1
𝑣 : unloading/reloading Poisson’s ratio, 𝐸 : reference tangent stiffness, 𝐸 : reference secant
stiffness, 𝐸 : reference unloading/reloading stiffness, 𝐺 : reference shear modulus at small
strain, 𝛾 . : reference strain threshold, 𝑚: exponential power, 𝑂𝐶𝑅: over consolidation ratio, 𝑅 :
strength reduction factor.

3.2.2. Comparison of the Modeling Methods

 3.2.2. Comparison of the Modeling Methods
Figure 5 shows that the curves of the transverse settlement trough, which were
yielded using the three methods, are close to the field measurements. However, the grout
pressure method gives a narrower transverse settlement trough, while the contraction
Infrastructures 2023, 8, 88 7 of 20
method offers a wider one. Finally, the proposed method provides the closest transverse
settlement trough.

Infrastructures 2023, 8, x FOR PEER REVIEW 8 of 21

4. Results and Discussion

4.1. Parametric Study
Figure 5. Comparison of different modeling methods.
FigureThe
5. Comparison
purpose of of different
the modeling
parametric studymethods.
is to understand the behavior of the surface
structure
4. Resultsduring the subsurface tunnel excavation and the interaction issue of tunneling,
and Discussion
soil, and surface
4.1. Parametric structure, as well as to obtain an equation to calculate the maximum set-
Study
tlementTheofpurpose
the structure.
of the parametric study is to understand the behavior of the surface
structure during the study
The parametric has not
subsurface included
tunnel a studyand
excavation of the internal forces
the interaction and
issue of the defor-
tunneling,
mation of the tunnel lining because the contraction coefficient (C%)
soil, and surface structure, as well as to obtain an equation to calculate the maximum directly represents the
volume
settlement lossofin thethe contraction method. However, the proposed method represents the
structure.
volume loss indirectly
The parametric study through
hasthe
notdeformation duringofgrout
included a study pressureforces
the internal application
and the and the
defor-
deformation of grout with initial stiffness (E, initial), which are not
mation of the tunnel lining because the contraction coefficient (C%) directly represents often used in research
studies.
the volume Becauseloss of
in this, the proposed
the contraction methodHowever,
method. makes it harder to obtainmethod
the proposed and calibrate this
represents
equation. Consequently, the contraction method and the HSS
the volume loss indirectly through the deformation during grout pressure applicationconstitutive model, which
provided a good agreement
and the deformation with
of grout withthe initial
field measurements, were adopted
stiffness (E, initial), which are in the
not parametric
often used
study that investigated
in research studies. Becausethe variables
of this,of the
the factors shown
proposed method in Figure
makes6, it where
harderDtoisobtain
the tunnel
and
diameter,
calibrate thise is eccentricity between the axis
equation. Consequently, the of the tunnelmethod
contraction and thatand of the
thestructure, B is the
HSS constitutive
structure
model, which width, Df is theafoundation
provided good agreement depth,with
P is the field
structure load, and Hwere
measurements, is the overbur-
adopted in
den depth.
the parametric study that investigated the variables of the factors shown in Figure 6, where
D is In
thethis parametric
tunnel diameter,study
e isand for the sake
eccentricity of safety,
between the groundwater
the axis of the tunnel table under
and that of the
the
greenfield
structure, Bcondition is considered
is the structure width, D tof be extremely
is the low depth,
foundation (27.5 m).
P isThe
themaximum settlement
structure load, and H
increases by 30% relative
is the overburden depth. to the real case of the groundwater level depth at 1.5 m.

Figure 6.
Figure Cross section
6. Cross section to
to clarify
clarify the
the studied
studied parameters.
parameters.

4.1.1.In
A this parametric
Surface Structurestudy and for
Modeling the sake of safety, the groundwater table under the
Method
greenfield condition is considered to be extremely low (27.5 m). The maximum settlement
The equivalent
increases beamto
by 30% relative method was
the real applied
case of the to model the structure,
groundwater which
level depth at 1.5includes
m. the
modulus of elasticity (E), the moment of inertia (I), and the cross-sectional area (A) (Figure
7). This method was used by several researchers [6,26,27]. Maleki et al. [6] compared the
internal forces, specifically, the bending moment of the tunnel lining, the ground surface
settlement, and the ground horizontal movement profile, which had been obtained by two
methods of structure modeling (e.g., real geometry model and equivalent beam model).
The authors found that the results were nearly identical, so they concluded that the equiv-
 Infrastructures 2023, 8, 88 8 of 20

4.1.1. A Surface Structure Modeling Method

The equivalent beam method was applied to model the structure, which includes
the modulus of elasticity (E), the moment of inertia (I), and the cross-sectional area (A)
(Figure 7). This method was used by several researchers [6,26,27]. Maleki et al. [6] compared
the internal forces, specifically, the bending moment of the tunnel lining, the ground surface
settlement, and the ground horizontal movement profile, which had been obtained by two
methods of structure modeling (e.g., real geometry model and equivalent beam model). The
authors found that the results were nearly identical, so they concluded that the equivalent
beam method can be used as a simple method to model the adjacent structure for practical
purposes. The characteristics of the flexible beam were calculated for a structure with a
number of stories (m), as shown in (Figure 7). It was assumed that the structure was made
up of (m + 1) slabs with a vertical spacing between them (3.4 m). The thickness of each
slab is assumed (tslab = 0.15 m), with dependence (L) in the direction perpendicular to
the slab section. The following Equations (1) and (2) govern the moment of inertia and
cross-sectional area:
t3 L
Islab = slab (1)
Infrastructures 2023, 8, x FOR PEER REVIEW 12 9 of 21

Aslab = tslab L (2)

Figure 7. Equivalent
Figure 7. Equivalentbeam
beammethod
method (on thebasis
(on the basisofof[6]).
[6]).

In the case of plane strain, values were given (Aslab = 0.15 m2 /m, Islab = 0.00028 m4 /m),
as well as the elastic modulus of the concrete
𝐼 (Ec𝑡= 23 𝐿
× 106 kN/m2 ). The equivalent beam (1)
12
stiffness, Equations (3) and (4), was calculated using the parallel axis theorem [28], with a
neutral axis in the middle of the structure height:
𝐴 𝑡 𝐿 (2)
( Ec A)structure = (m + 1)( Ec A)slab (3)
In the case of plane strain, values were given (Aslab = 0.15 m2/m, Islab = 0.00028 m4/m),
m +1
as well as the elastic modulus of the =
( Ec I )structure Ec ∑1 (E
concrete = 23
( Icslab +A 106h2m
× slab kN/m2). The equivalent(4)beam
stiffness, Equations (3) and (4), was calculated using the parallel axis theorem [28], with a
where (hm ) is the vertical distance between the neutral axis of the structure and that of the slab.
neutralThe
axisvalues
in theinmiddle of the structure height:
Table 5 were calculated, which represent the equivalent characteristics
of the structures with different stories. In this study, a ten-story structure with a width of
𝐸𝐴
(B = 13.5 m) and a uniformly distributed load of𝑚(150 1kPa)
𝐸 was
𝐴 used. (3)

𝐸𝐼 𝐸 𝐼 𝐴 ℎ (4)

where (hm) is the vertical distance between the neutral axis of the structure and that of the
slab.
The values in Table 5 were calculated, which represent the equivalent characteristics
 Infrastructures 2023, 8, 88 9 of 20

Table 5. Equivalent structure stiffness.

Structure (Ec I)structure (KN·m2 /m) (Ec A)structure (KN/m)

1 story 2× 107 6.9 × 106
3 stories 2 × 108 1.38 × 107
5 stories 6.96 × 108 2.07 × 107
10 stories 4.39 × 109 3.8 × 107

4.1.2. The Influence of Structure Width (B)

The increase in the structure width leads to an extreme change in settlements at the
end of
Infrastructures 2023, 8, x FOR PEER REVIEW the structure far away from the tunnel axis, where the impact of the tunneling is
10 of 21
smaller, so an increase in structure width can reduce settlements at this end. This reduction
Infrastructures 2023, 8, x FOR PEER REVIEW 10 of 21
can, in turn, impact the maximum settlement at the opposite end, closest to the tunnel
axis, causing it to decrease as well. Thus, when compared to the greenfield conditions,
struction, as demonstrated in Figure 8. Although the maximum settlement above the tun-
an increase in structure width can result in a widening of the area affected by tunnel
nel axis may
struction, only experience
as demonstrated a minor8.change
in Figure Althoughduethe
to maximum
an increasesettlement
in structure width,
above thethere
tun-
construction, as demonstrated in Figure 8. Although the maximum settlement above the
canaxis
nel be a significant increase in differential settlement an(𝛿increase
up to ainratio of structure
width,width
tunnel may only only
axis may experience a minor
experience change
a minor due todue
change structure
to an increase in structure there
width,
to tunnel
can be can diameter increase
a significant of approximately 2.7, as
in differential indicated(𝛿in up
settlement Figurea ratio
9. Thesestructure
findings width
are in
there be a significant increase in differential settlementto (δ) up to ofa ratio of structure
agreement
towidth
tunnel with earlier
diameter research conducted
of approximately by
2.7, as2.7,Maleki
indicated [6].
in Figure 9. These findings are are
in
to tunnel diameter of approximately as indicated in Figure 9. These findings
agreement with earlier research conducted by Maleki
in agreement with earlier research conducted by Maleki [6]. [6].

Figure 8. Influence of structure width on the transverse settlement trough.

Figure 8. Influence of structure width on the transverse settlement trough.
Figure 8. Influence of structure width on the transverse settlement trough.

Figure
Figure9.9.Influence
Influenceof
ofB/D
B/Dratio
ratioon
onmaximum
maximumand
anddifferential
differentialsettlements.
settlements.
Figure
4.1.3. 9.
TheInfluence of B/D
Influence ratio on maximum
of Foundation Depthand
(D differential
) settlements.
4.1.3. The Influence of Foundation Depth (Df)f
The resultsshow
showthat thatincreasing
increasing thethe foundation depthcancan
leadlead
to a to a decrease in
4.1.3.The
The results
Influence of Foundation Depth foundation
(D f) depth decrease in set-
settlements
tlements and a change
and a change in their
in their distribution.
distribution. Specifically,
Specifically, maximum
maximum settlements
settlements decrease
decrease
as Thefoundation
the results show that
depth increasing
increases, thethe
and foundation
transversedepth can lead
settlement to a decrease
trough narrows inasset-
the
as the foundation
tlements and a depth
change in increases,
their and the transverse
distribution. Specifically,settlement
maximum trough narrows
settlements as the
decrease
structure
structure approaches
approaches the top
theincreases, of the
top of theand tunnel,
tunnel, as shown
as shown in in Figure 10.
Figure 10.trough This is
This isnarrowsdue
due to the to the
de-
as the foundation depth the transverse settlement as the
crease in approaches
structure soil weight the
abovetopthe tunnel,
of the resulting
tunnel, as shownfromin the increased
Figure 10. Thisfoundation depth,
is due to the de-
which reduces the total stresses above the tunnel. However, the results
crease in soil weight above the tunnel, resulting from the increased foundation depth, also show a corre-
sponding
which increase
reduces in differential
the total settlement
stresses above with increasing
the tunnel. However,foundation depth,
the results also as depicted
show a corre-
in Figure increase
sponding 11. The increase in differential
in differential settlement settlement is due foundation
with increasing to the narrowing
depth,ofasthe trans-
depicted
 Infrastructures 2023, 8, 88 10 of 20

decrease in soil weight above the tunnel, resulting from the increased foundation depth,
which reduces the total stresses above the tunnel. However, the results also show a
corresponding increase in differential settlement with increasing foundation depth, as
depicted in Figure 11. The increase in differential settlement is due to the narrowing of the
Infrastructures 2023, 8, x FOR PEER REVIEW 11 of 21
transverse settlement trough, which causes the building weight to affect the area directly
Infrastructures 2023, 8, x FOR PEER REVIEW 11 of 21
above the tunnel and deformation to concentrate more directly above the tunnel than at the
far end.

Figure 10. Influence of foundation depth on the transverse settlement trough.

Figure 10.
Figure Influenceof
10. Influence offoundation
foundationdepth
depth on
on the
the transverse
transverse settlement
settlementtrough.
trough.

Figure
Figure 11.11. Influence
Influence ofof Df/D
Df/D ratio
ratio onon maximum
maximum and
and differentialsettlements.
differential settlements.
Figure 11. Influence of Df/D ratio on maximum and differential settlements.
4.1.4. The Influence of Eccentricity (e)
4.1.4. The Influence of Eccentricity (e)
4.1.4. The
The study
Influence has uncovered
of Eccentricitythe influence
(e) of eccentricity, defined as the distance between
theThe study
tunnel axis hasand uncovered
that of the thestructure,
influence of eccentricity,
on the transverse defined as thebehavior.
settlement distance be- The
tween The
the study has
tunnel axis uncovered
and that of the structure,
influence on of eccentricity,
the transverse defined as the
settlement distanceThe
behavior. be-
study found that the size of the transverse settlement trough decreases as the eccentricity
tweenfound
study the tunnel
that axis
the and
size of that
the of the structure,
transverse on thetrough
settlement transverse settlement
decreases as the behavior.
eccentricityThe
increases, while the maximum settlement above the top of the tunnel becomes less than
study found
increases, while that thethe size of thesettlement
maximum transverseabove settlement
theshown
toptrough
of in decreases
theFigure
tunnel12. as the eccentricity
becomes less thanin
what is observed in the greenfield conditions, as This reduction
increases,
what while
is observed
settlement can be the maximum
inattributed
the greenfield settlement
to theconditions, above
displacement the
as shown top of the
causedinbyFigure tunnel becomes the less
12. Thisofreduction
the presence than
in
structure
what
settlementis observed
can be in the
attributed greenfield
to the conditions,
displacement as shown
caused byinthe
away from the tunnel axis, which is distributed over a larger area, leading to a decreaseFigure 12.
presence This
of thereduction
structure in
settlement
away from can
the be
tunnel attributed
axis, to
which theis displacement
distributed caused
over a by
larger the
in settlements on the top of the tunnel. Furthermore, the study has also shed light on the presence
area, leading oftothe
a structure
decrease
inaway
effect from
settlements theontunnel
of eccentricity the top onaxis,
of which
thethe tunnel.is distributed
differential Furthermore,
settlement overthea larger
studyThe
behavior. area, leading
hasresults
also shed to a decrease
light
demonstrate on the
that
in
effect settlements
of eccentricityon theon top
the of the tunnel.
differential Furthermore,
settlement the
behavior. study
The
when the eccentricity is zero, the problem is symmetrical, and the differential settlement has also
results shed light
demonstrate on the
that at
effect
when of eccentricity
the eccentricity
the structure is almost on the
is zero, differential
the problem
non-existent. settlement behavior.
is symmetrical,
However, The results
and the differential
as the eccentricity demonstrate
increases, thesettlement that
problem
atwhen the asymmetric,
eccentricity
the structure
becomes is almostisand zero,thethe
non-existent. problem
maximum However,is symmetrical,
as the and
settlement and theincreases,
eccentricity
differential differentialthesettlement
settlement problem
increase,
at the
becomes structure
asymmetric, is almostand non-existent.
the maximum However,
settlement as the
and eccentricity
differential
consistent with the results of Maleki [6]. The peak in maximum settlement and differential increases,
settlement the problem
increase,
becomes
consistent
settlement asymmetric,
withoccursthe results
at e/D and the
=of0.5, maximum
Maleki
after [6].
whichThesettlement
peakofinthem
each and
maximum differential
returns settlementsettlement
and as
to a decrease, increase,
differential
illustrated
consistent
settlement with
in Figure occurs
13. the results of Maleki [6]. The peak in maximum settlement
at e/D = 0.5, after which each of them returns to a decrease, as illustrated and differential
insettlement
Figure 13.occurs at e/D = 0.5, after which each of them returns to a decrease, as illustrated
in Figure 13.
 Infrastructures 2023, 8, x FOR PEER REVIEW 12 of 21
Infrastructures
Infrastructures2023,
2023,8,8,x 88
FOR PEER REVIEW 12 11
of of2120

Figure 12. Influence of eccentricity on transverse settlement trough.

Figure12.
Figure Influenceofofeccentricity
12.Influence eccentricityonontransverse
transversesettlement
settlementtrough.
trough.

Figure 13. Influence

Figure 13. of e/D
Influence of e/D ratio
ratio on
on maximum
maximum and
and differential
differentialsettlements.
settlements.
Figure 13. Influence of e/D ratio on maximum and differential settlements.
4.1.5. The Influence of Structure Load (P)
4.1.5. The Influence of Structure Load (P)
The Influence
4.1.5. The study revealed that theLoad
of Structure vertical
(P)displacements between the two ends of the bottom
The study revealed that the vertical displacements between the two ends of the bot-
of the structure vary less with high structure loads, resulting in atworeduced differential
tomThe study
of the revealed
structure varythat the
less vertical
with high displacements
structure loads,between
resulting the ends of
in a reduced the bot-
differential
settlement
tom of the and
structure a wider
vary settlement
less with trough.
high Conversely,
structure loads, when
resulting the
in astructure
reduced load is zero,
differential
settlement and a wider settlement trough. Conversely, when the structure load is zero, the
the maximum
settlement and settlements
a wider settlement change marginally
trough. Conversely,in comparison
whenwith with greenfield
the structure conditions,
maximum settlements change marginally in comparison greenfieldload is zero, the
conditions, and
and the stiffness
maximum of thechange
settlements structure without in
marginally any load affectswith
comparison onlygreenfield
the shape conditions,
of the transverseand
the stiffness of the structure without any load affects only the shape of the transverse set-
settlement
the stiffness trough,
of theas as depicted
structure in Figure
without any14. 14. affects
load These only
observations
the shape are
of consistent
the transversewithset-the
tlement
previous trough,
studies depicted
of Malekiin[6],in Figure
Law14. [25],These observations
andobservations
Mirhabibi [29]. are consistent
Furthermore, withthe the pre-
study
tlement trough,
vious studies asMaleki
of depicted Figure These are consistent with theinvesti-
pre-
investigated the impact [6], Law [25],load
of structure and Mirhabibi
on maximum [29].
andFurthermore,
differential the study
settlements. The
vious studies of
gated the demonstrateMaleki [6],
impact of structure Law [25], and
load on the Mirhabibi
maximum [29].
andloadFurthermore,
differential the study investi-
outcomes that increasing structure’s leads tosettlements. The out-
a greater maximum
gated
comes thedemonstrate
impact of structure
thatuniform load on
increasing themaximum
structure’s and
loaddifferential
leadsbetween settlements.
to a greater The settle-
out-
settlement and a more distribution of settlement the maximum
two ends of the
comes demonstrate
ment and thereby that
a more uniform increasing the structure’s load leads to a greater maximum settle-
structure, reducingdistribution of settlement
the differential settlement,between
as shownthe two ends of
in Figure 15.the structure,
ment and areducing
thereby more uniform distribution
the differential of settlement
settlement, between
as shown the two
in Figure 15.ends of the structure,
thereby reducing the differential
4.1.6. The Influence of Overburden Depth (H) settlement, as shown in Figure 15.
With the increase in the overburden depth, the results show that the maximum settle-
ment at the bottom of the structure decreases, while the width of the transverse settlement
trough increases. This behavior can be attributed to the fact that the collapse surface inter-
sects the ground surface at a greater distance from the tunnel axis as the overburden depth
increases. As a result, the values of the displacements reaching the surface decrease, and
they are distributed over a larger width, as depicted in Figure 16. Moreover, the research
investigated the effect of tunnel overburden depth on the maximum and differential settle-
ments of the structure. The outcomes demonstrate that as the depth of the tunnel increases,
the settlement of the structure decreases, and the differential settlement in the structure
decreases, as illustrated in Figure 17.
 Infrastructures 2023, 8, x FOR PEER REVIEW 13 of 21
Infrastructures
Infrastructures2023,
2023,8,8,x88
FOR PEER REVIEW 1312ofof21
20

Figure 14.
Figure14. Influence of
Influenceof
14.Influence structure
ofstructure load
structureload on
loadon transverse
ontransverse settlement
transversesettlement trough.
settlementtrough.
trough.
Figure

Infrastructures 2023, 8, x FOR PEER REVIEW 14 of 21

Figure 15.
Figure Influence of
15. Influence of structure
structure load
load on
on maximum
maximum and
and differential
differential settlements.
settlements.
Figure 15. Influence of structure load on maximum and differential settlements.

4.1.6. The Influence of Overburden Depth (H)

4.1.6. The Influence of Overburden Depth (H)
With the increase in the overburden depth, the results show that the maximum set-
With the increase in the overburden depth, the results show that the maximum set-
tlement at the bottom of the structure decreases, while the width of the transverse settle-
tlement at the bottom of the structure decreases, while the width of the transverse settle-
ment trough increases. This behavior can be attributed to the fact that the collapse surface
ment trough increases. This behavior can be attributed to the fact that the collapse surface
intersects the ground surface at a greater distance from the tunnel axis as the overburden
intersects the ground surface at a greater distance from the tunnel axis as the overburden
depth increases. As a result, the values of the displacements reaching the surface decrease,
depth increases. As a result, the values of the displacements reaching the surface decrease,
and they are distributed over a larger width, as depicted in Figure 16. Moreover, the re-
and they are distributed over a larger width, as depicted in Figure 16. Moreover, the re-
search investigated the effect of tunnel overburden depth on the maximum and differen-
search investigated the effect of tunnel overburden depth on the maximum and differen-
tial settlements of the structure. The outcomes demonstrate that as the depth of the tunnel
tial settlements of the structure. The outcomes demonstrate that as the depth of the tunnel
increases, the settlement of the structure decreases, and the differential settlement in the
increases, the settlement of the structure decreases, and the differential settlement in the
structure decreases, as illustrated in Figure 17.
structure decreases, as illustrated in Figure 17.

Figure 16. Influence

Figure 16. Influence of
of overburden
overburden depth
depth on
on transverse
transverse settlement
settlement trough.
trough.
4.1.7. The Influence of Tunnel Diameter (D)
It is evident that as the diameter of the tunnel increases, the area affected by the
tunnel construction expands due to the increase in deformations around the tunnel. This
is mainly caused by an increase in the volume loss resulting from tunnel construction, as
depicted in Figure 18. The settlement at the bottom of the structure and the dimensions
of the transverse settlement trough increase as the diameter of the tunnel increases. It is
also noted that the maximum settlement increases significantly, whereas the differential
settlement in the structure increases slightly, due to the spread of deformations caused by
the tunnel excavation over a larger area, as demonstrated in Figure 19.

Figure 17. Influence of H/D ratio on maximum and differential settlements.

Infrastructures 2023, 8, 88 13 of 20
Figure 16. Influence of overburden depth on transverse settlement trough.

Figure 17. Influence of H/D ratio on maximum and differential settlements.

4.1.7. The Influence of Tunnel Diameter (D)

It is evident that as the diameter of the tunnel increases, the area affected by the tun-
nel construction expands due to the increase in deformations around the tunnel. This is
mainly caused by an increase in the volume loss resulting from tunnel construction, as
depicted in Figure 18. The settlement at the bottom of the structure and the dimensions of
the transverse settlement trough increase as the diameter of the tunnel increases. It is also
noted that the maximum settlement increases significantly, whereas the differential settle-
ment in the structure increases slightly, due to the spread of deformations caused by the
tunnel17.
Figure
Figure17.
excavation
Influence over
Influenceof
ofH/D
aratio
larger
H/Dratio on area, as demonstrated
onmaximum
maximumand anddifferential
in settlements.
Figure 19.
differentialsettlements.

4.1.7. The Influence of Tunnel Diameter (D)

It is evident that as the diameter of the tunnel increases, the area affected by the tun-
nel construction expands due to the increase in deformations around the tunnel. This is
mainly caused by an increase in the volume loss resulting from tunnel construction, as
depicted in Figure 18. The settlement at the bottom of the structure and the dimensions of
the transverse settlement trough increase as the diameter of the tunnel increases. It is also
noted that the maximum settlement increases significantly, whereas the differential settle-
ment in the structure increases slightly, due to the spread of deformations caused by the
tunnel excavation over a larger area, as demonstrated in Figure 19.
Infrastructures 2023, 8, x FOR PEER REVIEW 15 of 21

Figure
Figure18.
18.Influence
Influenceofoftunnel
tunneldiameter
diameteron
ontransverse
transversesettlement
settlementtrough.
trough.

Figure
Figure19.
19.Influence
InfluenceofofH/D
H/Dratio
ratioon
onmaximum
maximumand
anddifferential
differentialsettlements.
settlements.

4.2.Equation
4.2. EquationofofElementary
ElementaryAnalysis
Analysis
AAsensitivity
sensitivityanalysis
analysiswaswas conducted
conducted to demonstrate
to demonstrate the the influence
influence of various
of the the various
pa-
rameters on the settlements. Table 6 shows this influence through an increase ofof2020%%ofof
parameters on the settlements. Table 6 shows this influence through an increase
eachparameter.
each parameter. ItIt is
is clear
clear that
that the
the most
most effective
effectiveparameters
parameterson onthe
themaximum
maximumsettlement
settlementof
the adjacent structure are the following: the tunnel diameter (D), the foundation
of the adjacent structure are the following: the tunnel diameter (D), the foundation eccentricity
eccen-
(e), and
tricity (e),the
andoverburden
the overburden (H). D
depthdepth c , means
(H). that that
Dc, means the diameter of the
the diameter of tunnel is constant.
the tunnel is con-
Similarly, Hc means that the overburden depth is constant, and E represents the elastic
stant. Similarly, Hc means that the overburden depth is constant, and E represents the
modulus of the equivalent soil layer surrounding the tunnel.
elastic modulus of the equivalent soil layer surrounding the tunnel.

Table 6. The sensitivity analysis of the various parameters on the settlements.

Parameter B/D Df/D e/D P (kN) H/Dc Hc/D E (kPa)

Percentage change (%) 20 20 20 20 20 20 20
𝑺𝐯,𝐦𝐚𝐱 % 1.1 0.2 17 1.7 7.1 22.5 5
Infrastructures 2023, 8, 88 14 of 20

Table 6. The sensitivity analysis of the various parameters on the settlements.

Parameter B/D Df /D e/D P (kN) H/Dc H c /D E (kPa)

Percentage change (%) 20 20 20 20 20 20 20
Sv,max % 1.1 0.2 17 1.7 7.1 22.5 5

In this part, a newly proposed equation was developed based on the validated model
of the tunnel. An extended parametric study of the independent variables (D, e, H, C), where
C% denotes the contraction coefficient, was conducted using a total of 392 observations to
derive the equation for maximum settlement (i.e., dependent variable). The presumptive
values of the independent variables are listed in Table 7. The contraction method was
adopted to derive the equation for the reasons explained above in Section 4.1. The EUREQA
Version 1.2 software [30] was applied to find the mathematical connections between the
dependent and the independent variables, based on the available database. This equation
is recommended for usein the elementary analysis of tunnels.

Table 7. The presumptive values of the independent variables.

H/D 0.5 0.75 1 1.25 1.5 1.75 2 -

e/D 0 0.28 0.54 0.8 1.2 1.6 2 -
C% 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

The following Equation (5) is used to calculate the maximum settlement (Sv,max ), in
terms of the independent variables:
Z
Sv,max (mm) = 3.042 + − 5.91 × C − 11.23 × cos[(Y ) + cos( X ) × sin(1.131 × X ) − 1.131 × X ] (5)
Y + X2 − X

where: Z = 17.14 × X × C − 59.4 × C, X = De , Y = D H

For this analysis, the foundation width (B)= 13.5 m, the foundation depth (Df ) = 2 m,
and the applied load (P) = 150 kPa.
Many fundamental issues should be considered with the use of this equation. This
equation can be considered eligible for subsurface tunnels with the following constraints:
H/D ≤ 2, the values of the contraction coefficient should range between (0.2–1.6) %, the
shallow foundation of the structure should be rigid, and the angles should be in radians.

Validation of the Proposed Equation

This paper utilized three cases to validate the proposed equation: (i) the Second
Heinenoord Tunnel: a comparison between the numerical analysis outputs of the validated
model and the equation solutions; (ii) the Milan Twin Tunnel: a validated model is adopted
for the comparison with equation solutions; and (iii) three case studies (the Thessaloniki
subway, the Naples tunnel, and centrifuge tests) are used to compare the real measurements
and the equation solutions.
• The Validated Model of the Second Heinenoord Tunnel [18]
A parametric study was conducted in terms of e⁄D and H⁄D to validate the proposed
equation. Figure 20 shows a consistent comparison between the numerical model results
and the equation solutions.
• The Validated Model of the Milan Twin Tunnel
With an overall distance of 12.6 km, the metro-line 5 in Italy is composed of a twin
tunnel with an outer diameter (D) = 6.7 m, and it stretches between the north and the west
of Milan [31]. The twin tunnel height (H) = 11.65 m and the span between tunnel axes is
15 m. The tunnels were excavated in a gravely sand soil, whose properties are illustrated in
Table 8. The groundwater level is 15 m, which means that the tunnels were constructed
 nenoord Tunnel: a comparison between the numerical analysis outputs of the validated
model and the equation solutions; (ii) the Milan Twin Tunnel: a validated model is
adopted for the comparison with equation solutions; and (iii) three case studies (the Thes-
saloniki subway, the Naples tunnel, and centrifuge tests) are used to compare the real
Infrastructures 2023, 8, 88 measurements and the equation solutions. 15 of 20

• The Validated Model of the Second Heinenoord Tunnel [18]

A parametric
partially underneathstudy was conducted
the water table. An in
EPBterms of was
shield e⁄D and
usedH⁄D
on to
sitevalidate thethe
to reduce proposed
ground
equation.
movements in these highly inhabited regions. The tunnel lining was modeled as aresults
Figure 20 shows a consistent comparison between the numerical model beam
and the equation
element solutions.
with linear elastic behavior, and the properties of the lining are listed in Table 9.

Figure 20. Comparison between the numerical model results and the equation solutions.
Figure 20. Comparison between the numerical model results and the equation solutions.
Table 8. The soil properties surrounding the Milan Twin Tunnel.
• The Validated Model of the Milan Twin Tunnel
ref ref ref ref
E50
With an overall Eoed
distance of 12.6
Eur km, the G0 5 in Italy is composed of a twin
m metro-line
γsat
∅◦ ψ◦ c (kPa) γ0.7 (%) vur Rinter
(kN/m3 ) (kN/m 2 ) (kN/m 2 ) (kN/m2 ) (kN/m 2)
tunnel with an outer diameter (D) = 6.7 m, and it stretches between the north and the west
20 33 0 of Milan
0 [31].48,000
The twin48,000
tunnel height
144,000(H) = 0.4
11.65 m250,000
and the span between
0.0001 0.2tunnel axes
0.67 is
15 m. The tunnels were excavated in a gravely sand soil, whose properties are illustrated
in Table 8. The groundwater level is 15 m, which means that the tunnels were constructed
Table 9. The properties of the Milan Twin Tunnel lining.
partially underneath the water table. An EPB shield was used on site to reduce the ground
movements
Parameterin these highlyt (m)inhabited regions.
γ (kN/m 3 ) tunnel lining
The v was modeled as a beam
E (GPa)
element with
Tunnel lining linear elastic behavior,
0.3 and the25properties of the lining
0.15 are listed in35Table 9.
E: Young’s modulus.
Table 8. The soil properties surrounding the Milan Twin Tunnel.

𝜸𝐬𝐚𝐭 The field measurements

𝑬𝐫𝐞𝐟 𝑬𝐫𝐞𝐟 of settlements of the transversal ground section S16 under
𝟓𝟎 𝐨𝐞𝐝
∅𝒐 𝝍𝒐 c (kPa) the greenfield conditions 𝑬 𝐫𝐞𝐟
were
𝐮𝐫 (kN/m 2)
conducted m the
after 𝐆𝟎𝐫𝐞𝐟excavation
(kN/m2) of𝜸the 𝒗𝒖𝒓 and
(%) tunnel
𝟎.𝟕 first 𝐑 𝐢𝐧𝐭𝐞𝐫
both
(kN/m3) (kN/m2) (kN/m2)
tunnels, respectively. Plaxis 2D software was adopted in this study to validate the twin
20 33 0 0 48,000 48,000 144,000 0.4 250,000 0.0001 0.2 0.67
tunnel (for additional information, see [31]).
HSS was employed to simulate the behavior of the surrounding soil of the twin tunnel.
The contraction coefficients are depicted in Figure 21. This study [31] targeted the ground
surface settlements that resulted from the excavation of the first tunnel.
Based on the back analysis method, a comparison was performed between the mea-
sured surface subsidence in the field and those predicted by the HSS model. The field
measurements were performed in two stages: (1) after the complete excavation of the
first tunnel, and (2) after the excavation of the second tunnel. Figure 21 indicates a good
agreement between the field data and the HSS results.
After the validation of this model, a structure represented by a distributed load (P)
of 150 kPa was added to the model, where the foundation width (B) was 13.5 m and the
foundation depth (Df ) was 2 m. Figure 22 illustrates the comparison between the proposed
equation results and the numerical model outputs for the maximum settlements of the
structure. The presented equation results are relatively compatible with those of FEM.
 The field measurements of settlements of the transversal ground section S16 under
the greenfield conditions were conducted after the excavation of the first tunnel and both
tunnels, respectively. Plaxis 2D software was adopted in this study to validate the twin
tunnel (for additional information, see [31]).
HSS was employed to simulate the behavior of the surrounding soil of the twin tun-
Infrastructures 2023, 8, 88 nel. The contraction coefficients are depicted in Figure 21. This study [31] targeted 16 of 20
the
ground surface settlements that resulted from the excavation of the first tunnel.

Figure 21. Comparison between the field measurements and the results of numerical modeling us-
ing HSS constitutive model of the Milan Twin Tunnel.

Based on the back analysis method, a comparison was performed between the meas-
ured surface subsidence in the field and those predicted by the HSS model. The field meas-
urements were performed in two stages: (1) after the complete excavation of the first tun-
nel, and (2) after the excavation of the second tunnel. Figure 21 indicates a good agreement
between the field data and the HSS results.
After the validation of this model, a structure represented by a distributed load (P)
of 150 kPa was added to the model, where the foundation width (B) was 13.5 m and the
foundation depth (Df) was 2 m. Figure 22 illustrates the comparison between the proposed
equation
Figure 21. results
Figure 21.
and the
Comparison numerical
between
Comparison between fieldmodel
thefield
the
outputsand
measurements
measurements
forthe
and the
the maximum
results
results ofof
settlements
numerical
numerical
ofusing
modeling
modeling
the
us-
structure.
ing HSS The presented
constitutive model equation
of the results
Milan Twin
HSS constitutive model of the Milan Twin Tunnel.
are relatively
Tunnel. compatible with those of FEM.

Based on the back analysis method, a comparison was performed between the meas-
ured surface subsidence in the field and those predicted by the HSS model. The field meas-
urements were performed in two stages: (1) after the complete excavation of the first tun-
nel, and (2) after the excavation of the second tunnel. Figure 21 indicates a good agreement
between the field data and the HSS results.
After the validation of this model, a structure represented by a distributed load (P)
of 150 kPa was added to the model, where the foundation width (B) was 13.5 m and the
foundation depth (Df) was 2 m. Figure 22 illustrates the comparison between the proposed
equation results and the numerical model outputs for the maximum settlements of the
structure. The presented equation results are relatively compatible with those of FEM.

Figure 22. Comparison between the proposed equation results and the numerical model outputs for
the maximum settlements of the structure.

• Case Studies
- The Thessaloniki subway, Greece
The northbound and southbound lines are the designations of the twin tunnels that
make up the Thessaloniki city subway in Greece [32]. The two lines pass through densely
populated areas at relatively shallow depths, ranging from 8 to 20 m, and are in proximity
to adjacent structures. Two TBM-EPBs were utilized to decrease the surface settlements
when the excavation passed through a sandy clay deposit. The groundwater table is located
5 meters beneath the ground surface. Settlements of the stiff raft foundation of the adjacent
structure D91 were measured after the excavation of the southbound tunnel (first executed
tunnel), which consists of seven stories. The site conditions and the design parameters
are as follows: the tunnel diameter (D) = 6.2 m, the overburden depth (H) = 14.5m, the
eccentricity (e) = 22.5 m, the raft width (B) = 11 m, the over excavation (20 mm), and the
contraction coefficient (C) = 0.65%.
Figure 23 shows a comparison between the proposed equation and the field measure-
ments for the maximum settlements of the structure raft foundation. The result of the
present equation is somewhat consistent with the measured settlement.
- Line 6 of the Naples Underground
The Naples Underground Line 6 is a part of the public railway network provided
by the Municipal Plan of Transportation for the Metropolitan Area of Naples (Italy) [33].
The Santa Maria Della Vittoria church with a width of 17.3 m is located next to the path
 make up the Thessaloniki city subway in Greece [32]. The two lines pass through densely
populated areas at relatively shallow depths, ranging from 8 to 20 m, and are in proximity
to adjacent structures. Two TBM-EPBs were utilized to decrease the surface settlements
when the excavation passed through a sandy clay deposit. The groundwater table is lo-
cated 5 meters beneath the ground surface. Settlements of the stiff raft foundation of the
Infrastructures 2023, 8, 88 17 of 20
adjacent structure D91 were measured after the excavation of the southbound tunnel (first
executed tunnel), which consists of seven stories. The site conditions and the design pa-
rameters are as follows: the tunnel diameter (D) = 6.2 m, the overburden depth (H) = 14.5m,
of the
the Line 6 tunnel,
eccentricity which
(e) = 22.5 m, was excavated
the raft width (B)between
= 11 m,2009 and 2011.
the over Field (20
excavation measurements
mm), and
were conductedcoefficient
the contraction on the front
(C) façade
= 0.65%.of the church. The overburden depth (H) = 13.225 m,
the tunnel
Figurediameter
23 shows(D) = 8.15 m, and
a comparison the eccentricity
between (e) =equation
the proposed 12.6 m, while thefield
and the contraction
meas-
coefficientfor
urements was
the(C) = 0.3%.settlements
maximum The equation gives
of the a result
structure raftthat is fairly The
foundation. close to the
result field
of the
measurements
present equationofisthe maximum
somewhat settlement
consistent of the
with the measured
church, see Figure 23.
settlement.

Figure
Figure 23.
23. Comparison
Comparisonof
ofthe
themeasured
measuredand
andcalculated
calculatedvalues
valuesofofthe
thetwo
twocase
casestudies.
studies.

- -Comparison
Line 6 of the
Based Naples Underground
on the Centrifuge Tests
Theresults
The NaplesofUnderground
the centrifugeLine test 6series
is a part of the
for the tunnelpublic railwaydense
in regular network
dry provided
silica sand
by the Municipal
performed by Planetofal.Transportation
Ritter [34,35], at 75 for
g, the Metropolitan
were considered. A Area of Naples
structure on (Italy)
strip [33].
footings
Infrastructures 2023, 8, x FOR PEER REVIEW 19 of 21
The Santa Maria Della Vittoria church with a width of 17.3 m is
affected by tunneling was tested in 1/75th scale models. The dimensions correspondinglocated next to the path ofto
the Line 6 tunnel, which was excavated between 2009 and 2011. Field
the prototype scale (1/75th) are as follows: the tunnel diameter (D) = 6.2 m, the cover-to- measurements were
conducted
diameter ratioon (H/D)
the front façade
= 1.3, of the
and the church. The
overburden depthoverburden
(H) = 8.2 m.depth (H) = 13.225
The contraction m, the
coefficient
diameter
tunnel ratio
diameter (H/D)
(D) =
= 1.3,
8.15 and
m, the
and overburden
the eccentricitydepth
(e) =(H)
12.6 = 8.2
m, m.
while
is (C) = 1% [36]. Table 10 shows the parameters used for each test. Figure 24 confirms Thethecontraction
contraction coeffi-
coef-
that
cient is
ficient (C)
was = 1%
(C) =[36].
0.3%.Table
The 10 shows
equation the
gives parameters
a result used
that is for each
fairly
the equation results are in good agreement with the results of the centrifuge test. test.
close toFigure
the 24
field confirms
measure-
that
mentstheofequation resultssettlement
the maximum are in goodofagreement
the church,with see the
Figureresults
23. of the centrifuge test.
Table 10. -The parameters
Comparison Based
used on the
for each testCentrifuge Tests
(compiled by the authors after Franza et al., 2020 [36]).
Table 10. The parameters used for each test (compiled by the authors after Franza et al., 2020 [36]).
The results of the centrifuge test series for the tunnel in regular dense dry silica sand
Test Number B (m) e (m) H/D e/D
Testby
performed Number B (m)
Ritter et al. [34,35], at e (m)
75 g, were considered. H/D on strip e/D
A structure footings
1 1
affected by tunneling was tested15 12 12 The dimensions
15 scale models.
in 1/75th 1.31.3 corresponding
1.935 to
1.935
2 19.5
the prototype 2scale (1/75th) are as follows:
19.5 9.75 1.3
9.75diameter (D)1.3
the tunnel 1.573
= 6.2 m, the 1.573
cover-to-

Figure24.
Figure Comparisonofofthe
24.Comparison themeasured
measuredand
andthe
thecalculated
calculatedvalues
valuesofof the
the centrifuge
centrifuge tests.
tests.

5. Conclusions
Tunneling is a critical aspect of infrastructure development, with accuracy and effi-
ciency being essential for ensuring the safety and stability of surrounding structures. This
study presents a more realistic 2D modeling method that predicts ground movement and
Infrastructures 2023, 8, 88 18 of 20

5. Conclusions
Tunneling is a critical aspect of infrastructure development, with accuracy and ef-
ficiency being essential for ensuring the safety and stability of surrounding structures.
This study presents a more realistic 2D modeling method that predicts ground movement
and settlements accurately. Additionally, an extensive parametric study was conducted
to examine how the presence of adjacent structures impacts the deformation resulting
from tunneling, in comparison with deformation under greenfield conditions. A new
equation for calculating maximum settlement under adjacent structures simplifies the
design process and provides more accurate predictions, leading to safer and more efficient
infrastructure development.
Plaxis 2D software was used to investigate the interactions of a tunnel–soil–structure
system. The following conclusions can be drawn:
1. A new and more realistic modeling method is proposed for simulating the tunnel
excavation process. Although the reference modeling methods, namely the contraction
method and grout pressure method, have provided settlement troughs that were fairly
close to the measured value, the grout hardening method proposed in this study has
demonstrated a greater level of consistency with the field measurements.
2. The Hardening Soil Model with Small-Strain Stiffness (HSS) has yielded a transverse
settlement trough closer to the field measurement than that of the Hardening Soil
model (HS).
3. Increasing structure width, foundation depth, and overburden depth reduces the
maximum settlement under the structures. In contrast, the structure load and tunnel
diameter have negatively affected the maximum settlements.
4. Given the different values of the eccentricity between the tunnel and the structure, the
maximum settlements increase as the ratio (e/D) increases to be 0.5, and then decrease.
5. Each parameter studied has a unique impact on the settlement behavior. In addition,
the tunnel diameter, overburden depth, and eccentricity have a noticeably greater
effect on the settlement at the ground surface.
6. The increase in the structure width and the foundation depth has corresponded to
an increment in the structure differential settlements, whilst the structural load, the
overburden depth, and the diameter have reduced the differential settlements to
varying degrees.
7. Based on the finite element method (FEM), a new equation is developed for the
elementary analysis of the tunnels that calculates the maximum settlements beneath
the adjacent structures. A comparison of the results of the proposed equation with
the field measurements has been conducted for several case studies. The results of the
current equation are in good agreement with the measured values.
The proposed method and equation can serve as valuable tools for engineers and
researchers in the field of tunneling and infrastructure development, enabling them to
make informed decisions and minimize the risks associated with these complex projects.
The contributions of this study are expected to have a significant impact on the field of
tunneling and to pave the way for future research and development in this critical area of
infrastructure engineering.

Author Contributions: Conceptualization, A.S., R.A. and A.A.; proposed design method, A.S. inves-
tigation, A.A.; writing—original draft preparation, R.A. and A.A.; numerical modeling, R.A., A.S.
and A.A.; writing—review and editing, A.S., R.A. and A.A. All authors have read and agreed to the
published version of the manuscript.
Funding: This research received no external funding.
Data Availability Statement: Data will be made available upon request.
Conflicts of Interest: The authors declare no conflict of interest.
Infrastructures 2023, 8, 88 19 of 20

References
1. Franzius, J.N. Behaviour of Buildings due to Tunnel Induced Subsidence. Doctoral Dissertation, University of London, London,
UK, 2004.
2. Mair, R.J.; Taylor, R.N. Theme Lecture: Bored Tunnelling in The Urban Environment. In Proceedings of the Fourteenth
International Conference on Soil Mechanics and Foundation Engineering, Hamburg, Germany, 22–26 September 1997;
pp. 2353–2386.
3. Vermeer, P.A.; Brinkgreve, R. Plaxis Version 5 Manual; Delft University: Rotterdam, The Netherlands, 1993.
4. Möller, S.C. Tunnel Induced Settlements and Structural Forces in Linings. Doctoral Dissertation, Institute for Geotechnics,
University of Stuttgart, Stuttgart, Germany, 2006.
5. Likitlersuang, S.; Surarak, C.; Suwansawat, S.; Wanatowski, D.; Oh, E.; Balasubramaniam, A. Simplified Finite-Element Modelling
for Tunnelling-Induced Settlements. Geotech. Res. 2014, 1, 133–152. [CrossRef]
6. Maleki, M.; Sereshteh, H.; Mousivand, M.; Soleymani, H.; Ahmadi, M.; Rostami, J. An Equivalent Beam Model for The Analysis
of Tunnel-Building Interaction. Tunn. Undergr. Space Technol. 2011, 26, 524–533. [CrossRef]
7. Katebi, H.; Rezaei, A.H.; Hajialilue-Bonab, M. Assessment the Influence of Ground Stratification, Tunnel and Surface Buildings
Specifications on Shield Tunnel Lining Loads (By FEM). Tunn. Undergr. Space Technol. 2015, 49, 67–78. [CrossRef]
8. Giardina, G.; Dejong, M.J.; Mair, R.J. Interaction between surface structures and tunnelling in sand: Centrifuge and computational
modelling. Tunn. Undergr. Space Technol. 2015, 50, 465–478. [CrossRef]
9. Farrell, R.P. Tunnelling in Sands and the Response of Buildings. Doctoral Dissertation, University of Cambridge, Cambridge, UK, 2011.
10. Son, M. Response analysis of nearby structures to tunneling-induced ground movements in sandy soils. Tunn. Undergr. Space
Technol. 2015, 48, 156–169. [CrossRef]
11. Son, M. Response Analysis of Nearby Structures to Tunneling-Induced Ground Movements in Clay Soils. Tunn. Undergr. Space
Technol. 2016, 56, 90–104. [CrossRef]
12. Boldini, D.; Losacco, N.; Bertolin, S.; Fiamingo, A.; Grasso, P. Finite element modelling of tunnelling-induced displacements on
framed structures. Tunn. Undergr. Space Technol. 2018, 80, 222–231. [CrossRef]
13. Peck, R.B. Deep excavations and tunneling in soft ground. In Proceedings of the 7th International Conference on Soil Mechanics
and Foundation Engineering, Mexico City, Mexico, 17–21 August 1969; pp. 225–290.
14. O’Reilly, M.P.; New, B.M. Settlements above tunnels in the United Kingdom—Their magnitude and prediction. In Proceedings of
the Third International Symposium on Land Subsidence, London, UK, 14–17 November 1982; pp. 173–181.
15. Herzog, M. Surface Subsidence Above Shallow Tunnels. Bautechnik 1985, 62, 375–377.
16. Chakeri, H.; Ünver, B. A new equation for estimating the maximum surface settlement above tunnels excavated in soft ground.
Environ. Earth Sci. 2014, 71, 3195–3210. [CrossRef]
17. Wang, F.; Miao, L.; Yang, X.; Du, Y.J.; Liang, F.Y. The Volume of Settlement Trough Change with Depth Caused by Tunneling in
Sands. KSCE J. Civ. Eng. 2016, 20, 2719–2724. [CrossRef]
18. Van Jaarsveld, E.P.; Plekkenpol, J.W. Ground Deformations due to the Boring of the Second Heinenoord Tunnel. In Proceedings of
the Twelfth European Conference on Soil Mechanics and Geotechnical Engineering, Amsterdam, The Netherlands, 6–9 June 1999;
pp. 153–159. [CrossRef]
19. Maidl, U.; Ruse, N.; Maidl, B. Erfahrungen Mit Der Fem-Simulation Im Rahmen des Prozesscontrollings Beim Schildvor-
trieb/Experiences with Fem-Simulation In The Framework Of Process Controlling During Shield Tunnelling. Bauingenieur 2005,
80, 337–342.
20. Augarde, C.E.; Burd, H.J.; Houlsby, G.T. Some Experiences of Modelling Tunnelling in Soft Ground Using Three-Dimensional
Finite Elements. In Proceedings of the Fourth European Conference on Numerical Methods in Geotechnical Engineering
Numge98, Udine, Italy, 14–16 October 1998; pp. 603–612.
21. Meißner, H. Tunnelbau unter Tage—Empfehlungen des Arbeitskreises L. 6 Numerik in der Geotechnik. Abschnitt. Numer. Der
Geotech. EANG 1996, 19, 99–108. (In German)
22. Muir Wood, A.M. The Circular Tunnel in Elastic Ground. Géotechnique 1975, 25, 115–127. [CrossRef]
23. Alsirawan, R.; Koch, E. The Finite Element Modeling of Rigid Inclusion-Supported Embankment. Pollack Period 2022, 17, 86–91.
[CrossRef]
24. Alnmr, A. Material Models to Study the Effect of Fines in Sandy Soils Based on Experimental and Numerical Results. Acta Tech.
Jaurinensis 2021, 14, 651–680. [CrossRef]
25. Law, R.H. Effect of Existing Building on Tunneling-Induced Ground Movements. Master’s Thesis, Massachusetts Institute of
Technology, Cambridge, UK, 2012.
26. Pickhaver, J.A.; Burd, H.J.; Houlsby, G.T. An Equivalent Beam Method to Model Masonry Buildings In 3d Finite Element Analysis.
Comput. Struct. 2010, 88, 1049–1063. [CrossRef]
27. Franza, A.; Acikgoz, S.; Dejong, M.J. Timoshenko Beam Models for The Coupled Analysis of Building Response to Tunnelling.
Tunn. Undergr. Space Technol. 2020, 96, 103160. [CrossRef]
28. Timoshenko, S. Strength of Materials Part 1; Van Nostrand: New York, NY, USA, 1955; p. 422.
29. Mirhabibi, A.; Soroush, A. Effects of Surface Buildings on Twin Tunnelling-Induced Ground Settlements. Tunn. Undergr. Space
Technol. 2012, 29, 40–51. [CrossRef]
30. Dubčáková, R. Eureqa: Software Review. Genet. Program. Evolvable Mach. 2011, 12, 173–178. [CrossRef]
Infrastructures 2023, 8, 88 20 of 20

31. Fargnoli, V.; Boldini, D.; Amorosi, A. Twin Tunnel Excavation in Coarse Grained Soils: Observations and Numerical Back-
Predictions Under Free Field Conditions and In Presence of a Surface Structure. Tunn. Undergr. Space Technol. 2015, 49, 454–469.
[CrossRef]
32. Comodromos, E.M.; Papadopoulou, M.C.; Konstantinidis, G.K. Numerical Assessment of Subsidence and Adjacent Building
Movements Induced by Tbm-Epb Tunneling. J. Geotech. Geoenviron. Eng. 2014, 140, 04014061. [CrossRef]
33. Bilotta, E.; Paolillo, A.; Russo, G.; Aversa, S. Displacements Induced by Tunnelling Under a Historical Building. Tunn. Undergr.
Space Technol. 2017, 61, 221–232. [CrossRef]
34. Ritter, S.; Giardina, G.; Dejong, M.J.; Mair, R.J. Centrifuge Modelling of Building Response to Tunnel Excavation. Int. J. Phys.
Model. Geotech. 2018, 18, 146–161. [CrossRef]
35. Ritter, S.; Giardina, G.; Dejong, M.J.; Mair, R.J. Influence of Building Characteristics on Tunnelling-Induced Ground Movements.
Geotechnique 2017, 67, 1–12. [CrossRef]
36. Franza, A.; Ritter, S.; Dejong, M.J. Continuum Solutions for Tunnel–Building Interaction and a Modified Framework for
Deformation Prediction. Géotechnique 2020, 70, 108–122. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.