You are on page 1of 10

Tunnelling and Underground Space Technology 26 (2011) 524533

Contents lists available at ScienceDirect

Tunnelling and Underground Space Technology


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

An equivalent beam model for the analysis of tunnel-building interaction


M. Maleki , H. Sereshteh, M. Mousivand, M. Bayat
Department of Civil Engineering, Bu-Ali Sina University, Hamedan, Iran

a r t i c l e

i n f o

Article history:
Received 7 September 2010
Received in revised form 9 January 2011
Accepted 15 February 2011
Available online 8 March 2011
Keywords:
Shallow tunnel
Adjacent structure
Ground movement
Numerical modeling

a b s t r a c t
The aim of this work is to study the effect of structural characteristics, including stiffness, geometry and
weight on tunneladjacent structure interaction. Ground materials, tunnel geometry and excavator
device are related to a part of metro tunnel of Tehran. To describe the ground behavior due to tunneling,
a 3D FE code with an elastoplastic soil model was used. The adjacent building was modeled in two ways:
one as an equivalent beam or shell and the other as a real geometry (3D frames). The obtained results
from this theoretical work indicate particularly that the stiffness of adjacent structure controls the
ground movement distribution induced by tunnel excavation which in agree with other researchers.
As it was predicatively, increasing in structure weight leads to create the large displacement components
in the ground. The structure width plays also a signicant role in displacement distribution of ground.
The comparison of the obtained results using two methods of structure modeling shows a very good conformity between them.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Underground transportation systems have been in demand in
many major cities. These systems require a tunnel which is constructed in urban areas, particularly in soft ground and in shallow
zones. Measurement, designing and performing of underground
structure can be known as the most important civil engineers
challenge (Bernat and Cambou, 1998; Liu et al., 2008).
Inuence on adjacent buildings is of major interest for tunneling operations in urban areas, due to the high interaction between
tunneling and existing structures (Pickhavar et al., 2010; Dimmock
and Mair, 2008). This problem/issue was previously analyzed using
a combination of in situ observations and numerical modeling.
Analysis of previous case histories paved the way for the establishment of various empirical relationships between tunneling induced ground movement and associated structure damage
(Burland and Wroth, 1974; Boscardin and Cording, 1989; Burland,
1995; Mair et al., 1996). These methods are widely used in practice.
In reality, a rigorous analysis of the tunneling-structure interaction problem is a hard task, due to (I) the high interaction between
tunneling and adjacent structure, (II) 3D nature of this problem
and (III) the non-linear geometrical behavior involved that leads
to use an appreciate numerical method (Mroueh and Shahrour,
2003). Different approaches have been used to represent the building with varying level of details in the numerical methods. According to the simplied operations are executed in two consecutives

Corresponding author. Tel.: +98 811 8257410; fax: +98 811 8257400.
E-mail address: Maleki@basu.ac.ir (M. Maleki).
0886-7798/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tust.2011.02.006

steps. The rst step addresses the determination of ground movement induced by tunneling using empirical (Peck, 1969), analytical
(Sagasta, 1987; Verruijt and Booker, 1996) or numerical methods
which was widely studied by many researchers (for instance Gonzlez and Sagaseta (2001) and ITA/AITES Report (2007) studies);
and in the second step structural analysis of building subjected
to the ground settlement calculated in the rst step is done.
According to the work of Mroueh and Shahrour (2003) the simplied approach can be considered as very conservative. In the fully
3D FE modeling, details of building can be modeled. The advantage
of such 3D model is that the building can be taken into account in
any geometrical conguration with respect to tunnel axis. Three
examples of such analyses are the works of Mroueh and Shahrour
(2003), Burd et al. (2000) and Keshuan and Lieyun (2008). In the
plane strain analysis, the building is described by its width and
height and details can be incorporated in the model. The obtained
results from 2-D nite element analyzes show that the narrowest
settlement troughs were predicted when the nonlinearity of soils
at small strains was modeled (Chow, 1994).
Furthermore, there is recommended another approach based on
soil-structure relative stiffness which takes into account easily the
effect of structure stiffness in tunnel-structure interaction (Franzius, 2003). In this method, the structure is modeled as an equivalent
elastic beam or shell having bending and axial stiffness.
In spite of various works existing in the literature, it misses yet
a clear parametric study concerning the intensity of inuence of
stiffness and the other structure characteristics in tunneladjacent
structure interaction problem.
The present work is dened in the framework of tunnel
adjacent structure subject. It focuses particularity on parametric

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

study of structural characteristics effects of adjacent building on


ground movement induced by tunnel excavation. The adjacent
building is modeled rstly as an equivalent beam that allows us
easily to achieve parametric study operation. In the second part
of work, the structure is considered with its real geometry. The
study of interaction between tunnel and adjacent structure was
done using an indirect method developed in this work. PLAXIS
and SAP (powerful in structural problems analysis) codes were
used with the consideration of displacement eld compatibility
at soil media and structure interface. Tunnel geometry and ground
properties are corresponding to a part of metro tunnel of Tehran
city which was constructed using a slurry shield machine with
an outside diameter of 9.0 m. The obtained results by two methods
structure modeling were nally compared.

2. Modeling and parameters


PLAXIS 3D code only generates the triangle mesh, but it can use
the meshes in very ne size. Meshing is introduced in ve modes:
Very coarse, coarse, medium, ne and very ne. The important
ability of code is to make ner meshing regarding a region and
or surround of a line. However, the precision is increased by use
of ner mesh in a region but causes time to add for run problem
(PLAXIS code manual, 2005).
Medium mesh mode is used in present work and in more sensitive zones, mesh dimension gets ner. Selection of this size of mesh
is not worrying, because coarse meshes have been used in 3D settlement analysis by PLAXIS 3D code in some projects such as Rennsteig tunnel in Thuringia city. Also, for modeling Steinhaldenfeld
tunnel in Stutgart city, very coarse mesh with hardening elastoplastic constitutive model was used that had a good agreement with
real value from in situ information (Mair et al., 1996). Fig. 1 shows
FE mesh and also the lines of displacements measurement.
In this study, the most suitable constitutive model presented in
PLAXIS code was selected. This model is elastoplastic with the isotropic hardening mechanisms. It can be considered as development
of non-associated MohrCoulomb model. In fact, major limitations
of MohrCoulomb model are removed by adding a cap surface to
describe plastication under isotropic stress, and an isotropic hardening mechanism to express non-linear plastic behavior before the
failure. Evolution of yield surface in deviatoric mechanism is
controlled via deviatoric plastic strain. Volumetric plastic strain
controls the cap evolution. The plastic hardening and elastic
modulus are properly considered as function of conning pressure.
Basic properties of this model are:

525

Hardening plastic and elastic modulus is dependent on conning stress according to the exponential rule (exponential dependence of stiffness on stress).
Parabolic relationship between deviatoric stress and strain.
Separation of initial loading from unloadingreloading.
Coincidence of failure surface on MohrCoulomb criteria.
Nonetheless, this model is useful in monotonic loadings only
and some of important soil behavior aspects, such as failure surface
dependence on conning pressure and critical state concept are
not taken into account.
This model has eight parameters, fortunately all of which have
clearly physical meanings and are determined easily by the classical laboratory tests. Parameters of model are:
c: Soil cohesion.
u: Maximum internal friction angle.
w: Dilation angle.
E50
ref : Secant modulus in standard triaxial test at the reference
conning pressure (r3 = pref).
E50
ode : Tangent modulus related to the consolidation test.
Eref
ur : Modulus related to the unloading and reloading states.
m: Controls the dependence of plastic and elastic modulus on
conning stress.
mur: Poisson ratio in unloadingreloading state.
In the PLAXIS code, the mobilized shear strength in interface bond
is a function of shear strength of soil. This option is controlled
using the parameter Rinter that is equal to or less than 1.0, for real
soil-structure interaction the interface is weaker and more exible
than the associated soil layer, which means that the value of Rinter
should be less than 1.0. The Rinter in this study is supposed to be 0.7.
Because of the interface behavior before yielding is considered
elastic, the gapping or overlapping (i.e. relative displacements perpendicular to the interface) could be expected to occur. On the
other hand, the gap can be developed between the equivalent
beam and ground surface. In the present work, the gap in certain
case appeared, however, its value was very small without an effect
considerable on settlement prole in ground surface.
A section of line 1 of Teheran metro near 7tir square station was
modeled to achieve the aims of this study. Shield method was used
for tunnel construction. The information concerning the soil properties, tunnel geometry and tunneling device were taken from Tehran urban and suburban railway organization. Concerning the
geological aspects, 7tir station is located in the end part of nonhomogeneous alluvial formation in Tehran north and its lithologi-

Fig. 1. FE mesh and lines of soil movement measurement.

526

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

Fig. 2. Schematic cross section of geometry and material of line 1 of Teheran metro near 7tir square station.

Table 1
Soil physical properties.
2
ref
ref
2
2
Rinter Eres
mur
ur (kN/m ) Eoed (kN/m ) E50 (kN/m )

0.7

1.7E5

5.6E4

5.67E4

w () u () C (kN/m2)

0.2 10

40

0.25

Table 2
Mechanical parameters of tunnel lining.
EA (kN/m)

EI (kN/m)

Tunnel lining (cm)

8.05E6

8.218E4

35

Table 3
Parameters of modeled structure.
Equivalent structure
W (kN/m/m)

EAstruct (kN/m)

EIstruct (kN m2/m)

20
40
80
160

1.035E7
1.725E7
3.105E7
5.865E7

7.97E7
3.989E8
2.393E9
1.627E10

Structure

Row

2-Storey
4-Storey
8-Storey
16-Storey

1
2
3
4

Poison ratio for equivalent beam element to load modeling assume to be 0.25.

cal composition consists of sand, gravel, cobblestone and clay. Formation of this area is of a good permeability and depth of groundwater table is 74 m. Geotechnical data of this station shows the
in situ alternative layers of GP, GW, GC, SC and SM. Values of geotechnical parameters are obtained based on jacking and direct
shear tests (Fig. 2).
According to geotechnical information, the parameters of constitutive model were estimated. The values of parameters are listed
in Table 1.
Thickness and mechanical parameters of tunnel lining are presented in Table 2.
3. Equivalent beam consideration
The structure of adjacent building is considered by an equivalent elastic beam with length of L and width of B (Fig. 1). Bending

stiffness (EI) and axial stiffness (EA) represent the overall stiffness
of the structure. The advantages of this method are; simplicity in
considering adjacent building stiffness according to structural system and weight of building and also, in 2D conditions, the small
amount of computational resources is required and therefore the
ability to perform extensive parametric studies can be achieved.
The second moment of area for the equivalent beam was then, calculated using the parallel axes theorem (Appendix A). Considered
structures in analyses were as 2, 4, 8 and 16-storeys. Diverse
parameters of structures have been presented in Table 3.
Different steps of FE calculations performed in the rst part of
this work can be concluded as:

Analysis of ground for considering the gravity.


Introducing the equivalent beam and ground analysis.
Effacing the deformation eld engendered by the beam.
Performing the tunnel and capturing the deformation.

4. Discussion on equivalent beam results analysis


In this section, the results of analyses are presented. The focus is
on ground horizontal displacement and ground settlement distribution. There are presented, in Figs. 3 and 4, the proles of horizontal ground movement at 6 m offset from the tunnel center
line in the cases of with and without consideration of structure
stiffness, respectively. The depth of tunnel is 17.7 m and the width
of buildings has been considered to be 100 m. Each of curves corresponds to one building with specic storey and stiffness. In addition, soil movement corresponding to Greeneld (GF) analysis has
been included. The maximum horizontal movement in GF conditions along the vertical line is reached at z = 18.14 m. From there
it reduces towards the surface so that at z = 11.50 m horizontal displacement is zero. In continuing, the movement toward ground
surface increases so that horizontal displacement of 2.5 mm occurs
at the ground surface. The presence of building causes a reduction
in horizontal movement at ground surface. For example, it changes
from 2.5 mm corresponding to GF to 0.9 mm for a 2-storeys building. Although horizontal movement is less than GF at ground surface, it increases with depth and becomes larger than that obtained
for GF conditions. These two gures indicate that with the addition

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

527

of building storey, soil movement increases. With comparison of


curves in the cases of with and without consideration of stiffness,
it is observed that consideration of building stiffness leads to
decrease horizontal soil movement. It can be generally said that
structure stiffness is a factor that resists against horizontal deformation of soil induced by tunnel excavation. The presented results
in Figs. 3 and 4 indicate that the horizontal movement in the surface has the opposite direction in comparison with movement at
tunnel depth. That is physically real because the region near
ground surface is in contraction state due to tunnel excavation.
On the other hand, the horizontal movement is toward the tunnel
centerline, whereas in depth the ground movement is outward of
centerline. Settlement distribution of ground surface in two cases
of with and without consideration of building stiffness has been
Fig. 5. Vertical soil movement prole of ground surface without consideration of
the building stiffness.

Fig. 6. Vertical soil movement prole of ground surface with applying the building
stiffness.

Fig. 3. Horizontal movement of soil prole without applying the building stiffness
in distance of 6 m from tunnel axis.

Fig. 7. Horizontal soil movement prole in 6 m distance from tunnel axis for
building with different widths.

Fig. 4. Horizontal movement prole of soil in distance of 6 m from tunnel axis with
applying the building stiffness.

presented in Figs. 5 and 6. It is clear that settlement distribution


is effectively inuenced by building stiffness. In fact, building stiffness causes the uniform settlement beneath the foundation, so this
uniformity is increased by increase in stiffness. One of the most
important interests of equivalent beam method is that for a given
adjacent building, we can attribute the stiffness to building with
respect to its structural characteristics. For example, for a weighty
masonry building an insignicant stiffness can be considered.

528

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

Fig. 8. Vertical soil movement prole in ground surface for 4-storey building with
different widths.

Fig. 11. Vertical soil movement proles in ground surface for 4-storey structure
with different lengths of building.

Fig. 12. Structure columns plane.

Fig. 9. Horizontal soil movement proles in 6 m distance from tunnel axis where
the eccentricity is 7.5 m.

Fig. 13. Process of interaction between SAP and PLAXIS.

Fig. 10. Horizontal soil movement proles in 6 m distance from tunnel axis where
the eccentricity is 12.5 m.

Fig. 7 shows the effect of building width on ground horizontal


movement due to tunnel excavation. In this gure ground horizontal movement has been drawn versus depth for vertical line placed

at 6 m from tunnel center. The building is considered to be 4-storeys with 15, 30 and 60 m width. For comparative purposes, the results of green eld analysis for B = 100 m are included. These
curves indicate that horizontal movement is increased with increase in building width. The proles of soil vertical movement
in ground surface for different building widths have been presented in Fig. 8. These results indicate that for the great values of
width the maximum settlement is decreased but a large domain
of ground is inuenced due to tunnel excavation.
Fig. 9 shows the soil horizontal movement proles of vertical
line placed in 6 m distance from centerline of tunnel. These curves
are for a 4-storeys building with 15 m width and 7.5 m eccentricity
(building center in comparison with tunnel center). For comparative purposes, the results of green eld analysis are included. These

529

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533


Table 4
Reaction forces of structure in initial step.
Force (kN m) step 1

1
2
3
4
5

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

8.4
6.6
7.1
7.3
1.8

1262
1596
1638
2180
1446

8.9
10.3
94.6
120.1
16.2

7.8
4.7
5.3
3.6
3.1

1670
2076
2160
2721
1910

11.3
6.2
6.3
2
12.8

9.5
129.5
138.1
5.3
1.7

1806
2297
2413
2914
2029

11.4
5.9
6.5
2.9
13.9

6.7
130.1
134.9
3.2
12.6

2196
2711
2828
3620
2507

14.4
2.4
10.4
0.8
13.5

1.6
5.9
5.7
11
13.5

1451
1840
1941
2524
1660

16.5
1.5
124.2
149
5.8

Table 5
Reaction forces of structure in nal step.
Force (kN m) step 5

1
2
3
4
5

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

10.3
7.9
8.4
8.8
3.1

1242
1600
1715
2143
14277

5
6.3
90.3
99.9
13.4

6.6
3.8
4.7
3
3.5

1644
2075
2187
2731
1897

7.7
2.5
4.5
3.4
10.5

10.1
135.1
142.5
5.3
1.7

1775
2305
2442
2928
2012

7.1
2
4.7
4.3
11.4

7.1
13.4
143.2
2.6
11.8

2167
2700
2852
3631
2494

10.9
1.3
8.7
0.4
11.4

3
7.2
7.1
12.3
15.2

1433
1840
2006
2501
1649

13.4
15.1
119.7
12.3
15.2

Table 6
Displacement of beneath a structure columns in initial step.
Displacement (mm) step 1

1
2
3
4
5

Ux

Uy

Uz

Ux

Uy

Uz

Ux

Uy

Uz

Ux

Uy

Uz

Ux

Uy

Uz

0.27
0.31
0.35
0.41
0.47

13.10
14.41
14.81
14.12
11.99

0.34
0.34
0.33
0.35
0.35

0.27
0.30
0.35
0.42
0.48

14.71
16.43
16.41
15.72
13.55

0.38
0.37
0.36
0.38
0.4

0.26
0.3
0.35
0.44
0.50

16.3
17.73
18.11
17.36
14.99

0.41
0.40
0.40
0.41
0.43

0.26
0.30
0.36
0.45
0.51

17.38
18.85
19.25
18.46
15.99

0.43
0.42
0.42
0.43
0.45

0.25
0.3
0.36
0.46
0.55

18.95
20.52
21.01
20.09
17.41

0.48
0.46
0.46
0.46
0.49

Table 7
Displacement of the beneath a structure columns in nal step.
Displacement (mm) step 5

1
2
3
4
5

Ux

Uy

Uz

Ux

Uy

Uz

Ux

Uy

Uz

Ux

Uy

Uz

Ux

Uy

Uz

0.23
0.27
0.32
0.39
0.46

14.08
15.33
15.71
14.96
12.75

0.02
0.004
0.01
0.003
0.006

0.23
0.27
0.32
0.4
0.45

14.72
16
16.33
15.62
13.43

0.03
0.02
0.02
0.03
0.05

0.24
0.27
0.33
0.41
0.46

15.4
16.75
17.1
16.35
14.04

0.06
0.06
0.05
0.06
0.08

0.24
0.28
0.33
0.41
0.47

15.86
17.24
17.62
16.83
14.46

0.08
0.07
0.08
0.08
0.09

0.24
0.29
0.33
0.42
0.5

16.52
18
18.46
17.58
15.03

0.11
0.1
0.11
0.11
0.13

Fig. 14. 3D FE mesh for PLAXIS code.

results show that eccentricity of building inuences the soil horizontal movement around tunnel. In fact, an asymmetrical geometry of structure with respect to tunnel centerline results in
asymmetrical displacement eld of soil. When eccentricity takes

the great values, the effect of building on ground movement due


to tunneling will be negligible (Fig. 10).
Possibility of accurate studies on tunnel front behavior with
respect to construction methods is one of the most important

530

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

Fig. 15. Bending moment distribution in tunnel lining, (a is front section plan of structure b is middle section and c is rear section of structure).

interests of 3D FE analyses. In the present work, the effect of building length on ground movement around tunnel front has been
studied. To do this, a 4-storeys building with 15 m width was considered. The results of analyses for two different lengths of 10 and
50 m of building have been presented in Fig. 11. It can be seen that
ground vertical movement in direction of tunnel excavation is
inuenced by building length. In fact, for the case of building with
smaller length, the distribution of longitudinal vertical movement
is sharper than the case of building with the bigger length.

5. Consideration of structure as real geometry


Consideration of structure as real geometry and stiffness in soilstructure interaction problems can give more realistic response in
comparison with the equivalent beam. Although such consideration necessitates us to have the accurate information from present
adjacent building. This information is related to materials, geometry and structural system that are generally difcult to obtain.
The preparation of data for existing old buildings will be more

Fig. 16. Vertical displacement in the ground surface (a is front section plan of structure b is middle section and c is rear section of structure).

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

difcult. Therefore, it is reasonable to use other methods to model


adjacent buildings in tunneling or excavation problems. These
methods must be simple and have a good agreement in comparison
with the results, while the adjacent structure is modeled as its real
form. In the previous sections the adjacent building was modeled as
an equivalent beam that allowed us to study the effect of various
building properties such as stiffness, geometry, weight, width,
length and eccentricity of building with respect to tunnel axis, on
the tunnel-structure interaction problem easily. The question arising is how much equivalent beam properly faces to interaction
problems. To answer this question, one 10-storeys steel structure
is considered as real structure, and its results are compared with
the results of equivalent beam. The resistance system of structure
is bending frames in two orthogonal directions, and the connection
of columns to foundation is considered as pin. Footing is considered
to be as mat with 20 m  20 m  1.2 m dimensions. The column
plane of structure has been shown in Fig. 12.
The SAP code was used for the analysis of structures. The dead
and live loads combination is only considered in this study.
The excavation of tunnel creates a displacement eld for the
ground. For shallow tunnels, the created displacement eld inuences the adjacent building and urban services. On the other part,
the presence of adjacent building depending on geometry, weight,
and its stiffness controls the ground displacement eld due to tunneling. To study this important interaction between tunnel and

531

adjacent structure, it is recommended to use a unique nite element calculation code in which the behavior of ground material
and so the behavior of structure are properly described. At least,
in the practical works in geotechnical engineering there is rarely
existence of such general nite element code. In this paper, the
interaction between tunnel and real adjacent structure is studied
using an indirect method for which two nite element codes SAP
(strong in structural analysis) and PLAXIS (strong in geotechnical
engineering problems) have been performed alternately. To
achieve this aim, it will be necessary to use an iterative process between two codes. In Fig. 13, the analysis process using two codes
schematically is presented. As can be seen, the process is based
on transportation of total forces (Fx, Fy, Fz) and total displacements
(Ux, Uy, Uz) between two software i.e. SAP and PLAXIS 3D respectively; this process is updated in each steps. In the rst stage, the
structure is analyzed by SAP code and the forces of support points
are saved to be sent to the PLAXIS code. Now PLAXIS code is executed and gives displacements distribution for mat foundation lied
on ground surface. The displacements at the columns support
points are saved and sent to SAP code. These displacements are induced to support points of structure then SAP code is executed and
gives a set of new forces at the support points. The necessary condition for stopping the iteration process is to satisfy the displacement eld in the interface of structure and ground. On the other
hand, this denes the compatibility condition applied between

Fig. 17. Ground horizontal movement proles (a is front section plan of structure b is middle section and c is rear section of structure).

532

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

Fig. 18. (a): Geometrical idealization of structure, (b): consideration of structure as equivalent beam.

structure and ground. This iterative process is continued until the


differences of two sequence steps were about zero. The reactions of
columns and their displacements components for the rst and nal
step of analysis have been presented in Tables 47.
Fig. 14 shows the 3D FE mesh for PLAXIS code in which the positions of foundation and columns have been specied.
6. Comparison of two models
In this part of work the internal forces particularity bending
moment of tunnel lining, ground surface settlement and its distribution, and also ground horizontal movement prole, obtained by
two methods of building modeling are compared.
Bending moment distribution of lining versus central angle of
the tunnel sections, for two methods of building modeling has
been presented in Fig. 15.
The central angle is measured anticlockwise from horizontal
plane. There can be seen a very good conformity between obtained
results by two methods.
The ground surface settlement is an important factor that must
be controlled in interaction problems. The ground surface settlement versus mesh width for two methods has been shown in
Fig. 16.
It is clear that the response of equivalent beam method is very
close to the obtained results from analysis with real geometry of
structure. There is a small difference between two methods in location of structure foundation. This difference is due to uniform distribution of building weight in the rst method in comparison with
the concentrated load of columns in second method.
Ground horizontal movement proles obtained by two methods
have been presented in Fig. 17. These proles are located at 6 m
distance from centerline of tunnel. From this gure we can conclude a very good conformity between the equivalent beam and
the real geometry methods.
There is a small difference (less than 0.2 mm) between the obtained results by two methods particularly in section G that is due
to special manner of structure modeling.
7. Conclusion
In this paper a set of FE analyses were performed to study the
effects of adjacent building characteristics on interaction between
tunnel and adjacent structure. The adjacent structure was modeled
by two different methods. The analysis results obtained by these

methods were then compared with each other. The conclusions


from this study are summarized as following:
(1) Structure stiffness plays an important role in tunnel-structure interaction problem. In fact ground movement due to
tunneling is controlled by the structure stiffness, in particular, neglecting the structural stiffness yields to unrealistic
ground surface settlement.
(2) Weight of structure is a very fundamental factor in ground
movement caused by tunnel excavation. Ground movements
are generally increased due to increase in the structure
weight.
(3) The obtained results indicate that horizontal movement is
increased with increase in building width and also for the
great values of width the maximum settlement is decreased
but a large domain of ground is inuenced due to tunnel
excavation.
(4) Eccentricity of building from the tunnel centerline is also an
important factor. Asymmetrical deformations are the rst
effect of structural eccentricity.
(5) Adjacent building was modeled by two methods: equivalent
beam and real geometry. The comparison of obtained analysis results indicates that the equivalent beam method for
practical purposes can be used as a simple way for introducing the adjacent building characteristics in tunneladjacent
structure interaction problems.

Appendix A. Calculations of equivalent beam characteristics


The building is modeled by an equivalent elastic beam (in 2D
analyses) or shell (in 3D analyses) that lies on the ground surface.
Youngs modulus E, second moment of area I, and cross section A
are the structural properties of equivalent beam. Each storey of
building is considered as a slab, therefore, considering one slab
for footing, one m storey building can be modeled as m + 1 slabs
(Fig. 18). If the vertical mean distance between slabs is H and the
thickness of each slab is tslab then, using the parallel axes theorem
(Timoshenko, 1955), second moment of area of equivalent beam
can be calculated. Second moment of area I and the area A for slab
are dened as:

Islab

t 3slab L
12

Aslab tslab L

M. Maleki et al. / Tunnelling and Underground Space Technology 26 (2011) 524533

where L is out-of-plane dimension of the slab. Assuming the neutral


axis to be at the mid-height of the building, the bending stiffness for
the equivalent beam is then calculated as following:

Ec Ibeam EC

m
1 
X

Islab Aslab hm

In which hm is the vertical distance between the structures neutral axis and the mth slabs neutral axis. Axial stiffness for equivalent beam is obtained by the following expression:

Ec Abeam m 1Ec Aslab

Fig. 18 shows the geometry of structure with respect to tunnel


position (Franzius, 2003).
References
Bernat, S., Cambou, B., 1998. Soil-structure interaction in shield tunneling in soft
soil. J. Comput. Geotech. 22, 221242.
Boscardin, M.D., Cording, E.G., 1989. Building response to excavation induced
settlement. J. ASCE J. Geotech. Eng. 115 (1), 121.
Burd, H.J., Houlsby, G.T., Augarde, C.E., Liu, G., 2000. Modelling tunnelling-induced
settlement of masonry buildings. In: Proceedings of the ICE, Geotechnical
Engineering, vol. 143 (1), pp. 1729.
Burland, J.B., 1995. Assessment of risk damage to buildings due to tunneling and
excavation. In: Proceedings of 1st International Conference on Earthquake and
Geotechnical Engineering, IS-Tokyo, Japan, pp. 11891201.
Burland, J.B., Wroth, C.P., 1974. Settlements on buildings and associated damage. In:
Proceedings of Conference on Settlement of Structures, BTS, Cambridge, pp.
611654.
Chow, l., 1994. The Prediction of Surface Settlements due O Tunneling in Soft
Ground. MSc Thesis, Oxford University.

533

Dimmock, P.S., Mair, R.J., 2008. Effect of building stiffness on tunneling-induced


ground movement. J. Tunnel. Undergr. Space Technol. 23, 438450.
Franzius, J.N., 2003. Behaviour of Buildings due to Tunnel Induced Subsidence. PhD
Thesis, Department of Civil and Environmental Engineering, Imperial College of
Science, Technology and Medicine, London, UK.
Gonzlez, C., Sagaseta, C., 2001. Patterns of soil deformations around tunnels.
Application to the extension of Madrid Metro. Comput. Geotech. 28, 445468.
ITA/AITES Report, 2007. 2006 on settlements induced by tunnelling in soft ground. J.
Tunnel. Undergr. Space Technol. 22, 119149.
Keshuan, M., Lieyun, D., 2008. A full 3-D nite element analysis of the prediction
between river-crossing tunneling and adjacent building. In: Proceedings of
World Tunnel Congress on Underground Facilities for Better Environment and
Safety, India.
Liu, H.Y., Small, J.C., Carter, J.P., 2008. Full 3D modelling for effects of tunnelling on
existing support systems in the Sydney region. J. Tunnel. Undergr. Space
Technol. 23, 399420.
Mair, R.J., Taylor, R.N., Burland, J.B., 1996. Prediction of ground movements and
assessment of risk of building damage due to bored tunneling. In: Proceedings
of the International Symposium on Geotechnical Aspects of Underground
Construction in Soft Ground, Balkema, Rotterdom, pp. 713718.
Mroueh, H., Shahrour, I., 2003. A full 3-D nite element analysis of tunneling
adjacent structures interaction. J. Comput. Geotech. 30, 245253.
Peck, R.B., 1969. Deep excavation and tunnelling in soft ground. In: Proceedings of
7th International Conference on Soil Mechanic Foundation Engineering, Mexico,
State-of-the-Art Volume, pp. 225290.
Pickhavar, J.A., Burd, H.J., Houlsby, G.T., 2010. An equivalent beam method to model
masonry building in 3D nite element analysis. J. Comput. Geotech. 88, 1049
1063.
PLAXIS 3D Tunnel 2.00 Manual, 2005.
Sagasta, C., 1987. Analysis of undrained soil deformation due to ground loss.
Geotechnique 37, 301320.
Timoshenko, S., 1955. Strength of Material, Part 1, Elementary Theory and Problems,
third ed. D. Van Nostrand, New York.
Verruijt, A., Booker, J.R., 1996. Surface settlements due to deformation of tunnel in
an elastic half plane. Geotechnique 46, 753757.