Engineering Failure Analysis 111 (2020) 104495

Contents lists available at ScienceDirect

Engineering Failure Analysis

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

Impact of neighbouring deep excavation on high-rise sun plaza

T
building and its surrounding
⁎
Sadik Oztopraka, Safiye Feyza Cinicioglub, Namik Kemal Oztoruna, Cenk Alhana,
a
Civil Engineering Department, Istanbul University-Cerrahpaşa, 34320 Avcılar, Istanbul, Turkey
b
Civil Engineering Department, Özyeğin University, 34794 Cekmekoy, Istanbul, Turkey

A R T IC LE I N F O ABS TRA CT

Keywords: A deep excavation was carried out at the border of 38-story Sun Plaza Tower in Istanbul with a
Deep excavation shoring that made use of temporary anchorages. However, building construction license was
Soil-structure interaction cancelled due to later detected legal issues. Since the structure could not have been placed timely,
Temporary shoring useful lifetime of temporary anchorages expired and safety concerns appeared. It led to rumors
Seismic performance
such as cracks appearing in the Tower and even became headlines in newspapers which led to
loss of tenants. Here, technical and legal problems, which have arisen from this extraordinary
situation, are discussed. Site investigations, numerical modeling and analyses conducted, and
solution alternatives developed for this complex problem, which requires an interdisciplinary
study, are explained in detail. The actual application of the proposed solution and the positive
outcome in terms of both ensuring structural safety and gaining new tenants are also reported. It
is expected that this study will draw attention to potential technical and legal problems that may
arise in densely populated city centers where high-rise structures are continuously being con-
structed side by side.

1. Introduction

One of the biggest problems regarding new high-rise construction in densely populated large metropolitan areas is the necessity of
constructing significant shoring systems since very few parcels left in these areas are typically squeezed between existing high-rise
buildings. Many aspects of a deep excavation in terms of its effects on nearby structures are investigated by various researchers
previously (e.g. [21,13,23,17,14]). As such situations are faced with an ever increasing rate, it draws attention of more researchers
recently. Abdallah et al. [1] evaluated the importance of 3D modeling compared to 2D modeling using PLAXIS software in case of
deep excavation-diaphragm wall-soil-structure interaction. Adam et al. [2] presented the geotechnical relevant works including the
construction of pit supporting system for the excavation pit and 35 m deep foundation works for the 220 m tall Donau City Tower 1 in
Vienna. Castaldo et al. [8] proposed a simplified probabilistic methodology to evaluate excavation-induced damage to adjacent
existing buildings by making use of the results of a numerical model of the boundary value problem. Castaldo and De Iuliis [9]
showed the significancy of the effect of deep excavation on the seismic vulnerability of existing buildings by means of different
seismic damage indices and inter-story drifts via a finite element model of a real case study. Yeganeh et al. [28] investigated the
influence of the deep excavation on the seismic behavior of an adjacent high-rise building and showed that neighbouring deep
excavations can have important effects. Furthermore, they have pointed out that the building should be modeled as a frame rather
than surcharge, only, as this also significantly affects the outcome of the soil-structure interaction analysis. Akhtarpour and

⁎
Corresponding author.
E-mail address: [email protected] (C. Alhan).

https://doi.org/10.1016/j.engfailanal.2020.104495
Received 4 August 2019; Received in revised form 11 December 2019; Accepted 3 March 2020
Available online 05 March 2020
1350-6307/ © 2020 Elsevier Ltd. All rights reserved.
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Mortezaee [3] revealed the importance of considering excavation and soil structure interaction in the modeling and seismic analysis
of high-rise structures in the context of a 17-story building located next to a deep excavation in Mashhad, Iran by making use of Plaxis
2D.
Since shoring systems for deep excavations next to existing buildings directly affect the safety of nearby structures, they have to be
built with large safety factors and constitute a major expense item. In order to reduce both the construction and the maintenance cost,
they are commonly planned and built with temporary shoring systems by making use of temporary anchors which typically have
maximum two years lifetime. By the end of this time frame, the anchors loose their load carrying capacity as they corrode and wear
off and thus transfer the loads, they carry at the beginning of construction, to the newly constructed structure. Ordinarily, this
construction method and corresponding behavioral mechanism work very efficiently provided that the construction schedule can be
implemented timely. But if legal issues arise relating to zoning law, which may lead to increased construction time, potential disasters
may come into scene due to the potential failure of the temporary shoring system leading to the failure of nearby high-rise buildings.
The Sun Plaza Tower Case is a perfect example which proved the deficiency in the legal system against such situations. Fortunately,
however, it has been solved before a disaster has happened.
The 38-story Sun Plaza Tower that is located in the Şişli District of the City of İstanbul is composed of 5 Basement Floors, Ground
Floor, Gallery Floor, 30 Office Floors, and a Lift/Machinery Floor. It is a reinforced concrete structure which was built in 2002. For
another structure that was planned to be constructed on a land nearby the Sun Plaza Tower, an excavation was started in July 2011
and reached a depth of 30 m before the construction was stopped (and has never been continued since then) due to the later detected
legal issues which led to the cancellation of the construction license. Since the structure could not have been placed within the
planned time frame, the useful lifetime of the temporary anchorages have expired approximately by the year 2013 and safety
concerns appeared. It had even led to rumors such as cracks appearing in the Tower. Due to the massiveness of the structure and its
location, which is in the middle of one of the most crowded areas in Istanbul, this issue found itself a place in the newspapers [12,18].
This has finally led to loss of tenants and thus loss of significant money to the owner of the Tower.
The aforementioned deep excavation with a temporary shoring, which was opened adjacent to the Sun Plaza Tower while it was
in service, is shown in Fig. 1. As it is explained later in the forthcoming sections, the excavation was back filled partially as part of an
incomplete operation and the picture shown in Fig. 1 belongs to that time which is the same time when the authors were involved in
the investigation. The original depth of the excavation was, however, 15 m deeper than what is seen in Fig. 1.
The Sun Plaza Tower, as shown in Fig. 2, consists of three buildings that are separated by expansion joints: Tower, shopping
center and a multi-story parking lot. Although these structures are independent and separated from each other, they have a common
mat foundation with varying thickness. While the thickness of the mat is 2.10 m underneath the core of the tower, it is reduced to
70 cm and 50 cm underneath the shopping center and parking lot structures, respectively. The main anxiety of the owners was related
to the safety of Tower since it was closest to the excavation site. Thus tower, the high-rise structure, is included in this investigation.
The height of the tower from top of the mat foundation to the top of the roof is 141.20 m. According to the project documents,
concrete class is C35 and steel grade S420. The structure is designed by [20] according to first degree earthquake zone defined in
Turkish Earthquake Code [26]. The coefficient of effective ground acceleration is A0 = 0.4. Spectrum characteristic periods of
TA = 0.1 s and TB = 0.3 s are used, which correspond to the strongest soil type, soil class Z1.
The structural system of the tower can be summarized as follows: There exists a core at the centre of the building that is composed
of two C-shaped reinforced concrete shear walls facing each other with thicknesses of 50 cm at the lower floors changing to 40 cm at
the upper floors. They are connected to each other via reinforced concrete beams of depth 80 cm. Inside this core, there exists other
shear walls of thickness 20 cm which are also used to support the elevators and the stairs. There exist evenly spaced reinforced
concrete columns around the outer perimeter with sectional dimensions of 100 × 50 cm at the lower floors changing to 50 × 40 cm
at the upper floors. Additionally, L-shaped reinforced concrete columns of 235 × 235 × 50 cm whose thickness reduce to 40 cm at
the upper floors are placed at the corners of the tower. These columns are connected to each other by 80 cm deep reinforced concrete
beams whose widths are 50 cm at the lower floors and drop to 25 cm at the upper floors. The typical slab is a waffle slab made up of
beams of (15 ~ 25) × 40 cm sectional dimensions which are spaced at 120 cm and a plate of thickness 12 cm. The typical floor plan

Sun Plaza Tower

A Excavated
B Site B
C
G A

Excavated Site

Fig. 1. The Sun Plaza Tower and the neighbouring excavation zone.

2
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Tower / Shopping center Otopark / Otopark Tower / Otopark

Fig. 2. Structural Sections [20].

(office floor plan) is shown in Fig. 3.

2. Deep excavation and related problems

The details of the deep excavation carried out at the border of the Sun Plaza Tower and the shoring system are shown in Fig. 4. In
order to describe the problems observed in different regions and to explain the findings in the site investigations, different regions are
labelled using letter A-F. There exists a “reinforced concrete retaining wall” (shown with label D) which is connected to the “Sun
Plaza Tower” (shown with label A) by “connection beams” (shown with label E). There is another reinforced concrete wall named in
this paper as “service road retaining wall” (shown with label C) that is 30 cm thick and located 4.6 m away from the “reinforced
concrete retaining wall”. In between and supported by these walls is the “service road” (shown with label F). At the beginning, while
excavation was being carried out, significant displacements were recorded by the inclinometers which were placed to monitor the
performance of the deep excavation shoring system (shown with label G). In parallel to this observation, in the region right behind
the shoring system G which includes “service road – F”, “connection beams – E”, “service road retaining wall – C”, and “reinforced
concrete retaining wall – D”, problems such as separation, settlement, and cracks have been observed. The region where movements
and damage have been observed is shown in a frame in Fig. 4, explicitly. For visual convenience, the labels A-G are also given in
Fig. 1, and Figs. 5 and 6. The structure given with label B in Fig. 1 is the shopping center. The observations made during the initial site
visit on March 2014, regarding the aforementioned problems are given in Figs. 5 and 6. It should also be noted that no visible damage

Fig. 3. Typical floor plan (office floor plan) [20].

3
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Fig. 4. Details of the deep excavation and the shoring system.

Fig. 5. Settlement in Service Road-F and separation and cracks in Service Road Retaining Wall-C.

in the tower that could be related to the deep excavation was observed.
As seen in Fig. 5, due to the lateral movements towards the excavation, settlements and separation from the surrounding re-
inforced concrete retaining walls are observed in the service road which is underlain by crushed stone fill. In addition, visible cracks
and damage have been observed in the “service road retaining wall – C” due to the same reason. This very wall is clearly observed to
be separated from the shopping mall building – B. There exist longitudinal cracks in the “reinforced concrete retaining wall – D” as

4
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Fig. 6. Cracks in the RC retaining wall (label D) and in the connection beams (label E).

shown in Fig. 6. This wall extends from the mat foundation up to the 2nd basement floor and is not a main part of the building as
shown in Fig. 4. This wall mainly supports the lateral thrust of the crushed stone fill that makes up the service road. It is observed that
water coming from outside seeps through the cracks in this wall towards inside. In addition, cracks were observed in the “connection
beams – E” that connects this wall to the Tower structure. These cracks are thought to appear due to the lateral movements of the
upper shoring system towards the excavation site.
Various studies were conducted by expert witnesses and technical reports were written about the problems faced in Sun Plaza case
[4,6,7,11,15,16,27] before the authors have been included. Most of these were based on site observations and included assessment of
such observations. Although a relatively more comprehensive study was conducted in the scope of in the report of Guler [15], the
evaluation made in this report suffers from a mistaken assessment towards the interpretation of inclinometer readings. Because, the
inclinometer readings were assessed by assuming the upper and lower shoring systems as a single body although it is formed of two
separate levels. By doing so, the relative displacements were divided by the whole height to obtain drift angles whereas the upper
level displacements and thus drifts were much more significant in reality. Moreover, it was this distinction which gave the crucial
clue about the behavioral mechanism that was encountered. The details of the actual behavioral mechanism are put forth in the
proceeding sections.

3. Modeling and analysis

The authors have started their investigation by first conducting a study on the soil characterization and stratification and the
related parameters were obtained for use in numerical modeling and analyses. The results of the analyses using models with available
site investigation data are compared with the inclinometer readings and the soil parameters were revised by conducting back analyses
by Plaxis. Limit State equilibrium software was used to determine the location and the mechanism of failure by making use of the
parameters obtained. Then, the effect of the excavation on the structure is investigated by seismic risk analyses.

3.1. Soil characterization and stratification

Four new boreholes and seismic tests (MASW) were done around Sun Plaza (Fig. 7) in order to characterize the soils and back-fill

BH-4
BH-2
BH-3
BH-1

Fig. 7. Location of; (a) Boreholes, (b) Seismic profiles.

5
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Fig. 8. Variation of elasticity modulus and limit pressure obtained from pressuremeter tests carried out along the boreholes.

materials. Two of the boreholes (BH-1 and BH-2) were carried out next to the tower to check the soil and rock parameters with the
previous works and latter (BH-3 and BH-4) were done in the excavation to characterize the back-fill material. Pressuremeter tests
were also carried out in the BH-1 and BH-2. Seismic profiles; SP-1, SP-2 and SP-3 were used for characterizing the fill material. In
relation to this, SP-4 profile was used to compare the natural ground condition with the back-fill. In order to understand the stra-
tification of the subsoil and the soil nearby the foundation of Sun Plaza, the sand-clay-silt units were cut up to 10–11 m in the BH-1
boring. The SPT N values in these units vary between 11 and 31. It is determined from the identification experiments that there exist
gravel and sand in the upper levels which turns into silt and clay at the deeper levels. At 11–13 m depth, there exist completely
weathered claystone/shale and from about 13 m depth average strength claystone/shale fracturing decreases as depth increases. The
variations of the limit pressure and the elasticity modulus with depth which are obtained from the soil stratification and pressure-
meter tests are shown in Fig. 8.
In the BH-2 boring, which was carried out in the section backfilled after the foundation excavation on the side next to the
building, an 18 m layer of crushed stone fill was passed and then a rock unit was encountered. The variation of the limit pressure and
modulus of elasticity obtained from the soil stratification and pressuremeter tests in BH-2 are presented in Fig. 8. In Fig. 9, the
transition from soil to rock can be seen between the piles.
It is possible to obtain the soil strength parameters from the limit pressure values which are obtained from pressuremeter ex-
periments via empirical approaches. In this context, Center d’Etudes Menard [10] proposed Eq. (1) to obtain the angle of shear
strength (ϕ′) in case of drained soils.

∅' − 24
pL∗ = b·2 4 (1)

Here, p*L is the net limit pressure and b is a coefficient depending on the condition of the soil which can be taken as 180 for wet
homogenous soil, 350 for dry heterogenous soil, and 250 on average. The effective shear strength angles (ϕ′) calculated with this
equation by using b = 250 for fill and clay are given in Table 1. The cohesion for claystone and shale (Table 1) is calculated by the
equation of Baguelin et al. [5]:

6
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Claystone/shale units just above the Weathered Clayey silty sands

foundation level of tower claystone (Belgrad formation)

Fig. 9. Natural soil-rock units just above the deep excavation fill.

Table 1
Initial parameters of soils/rocks for starting back-analysis.
Level (m) Soil C (kPa) ϕ (°) Eo = Eur (MPa) E50 (MPa)

52.5–48.5 Made fill 1–5 27–28 15–45 5–15

48.5–38 Clay-Gravel-Sand (Belgrad form.) 1–5 30–32 100–300 30–100
38–34 Weathered claystone 25–50 20–30 300–600 100–200
34-… Claystone-shale 200–400 30–35 1000–1500 300–500

Note: The pressuremeter experiments are used in obtaining strength values while pressuremeter and MASW experiments are used in obtaining
modulus of elasticities.

0.75
Su P∗
= 0.21·⎜⎛ L ⎟⎞
pa ⎝ pa ⎠ (2)

Owner provided some photos (Fig. 10) to understand the backfilling process. It can be seen in Fig. 10b that the excavation pit was
previously filled with a heterogeneous material with mainly coarse content. The borings and seismic tests were evaluated together in
order to determine the compaction degree of the backfill and the depth of the rock unit. When the Vs velocities obtained from P-1, P-2
and P-3 profiles in the backfilled area are compared with the Vs velocity of the natural soil obtained from P-4 profile, it was
determined that the average velocities were similar but the relatively loose region detected between about 4 m to 9 m within the
backfill (See Fig. 11 – red-hatched area) gave lower velocities indicating the presence of a looser soil.
This is an expected case for fills that do not qualify as engineering fills. As a matter of fact, existence of regions where Vs profiles

Fig. 10. (a) Backfilling of the excavation pit up to the second shoring level in the first stage, (b) Compaction process and used material during
backfilling.

7
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Elasticity Modulus Tomography Vs (m/s) Vs (m/s) Vs (m/s) Vs (m/s)

0 200 400 600 0 200 400 600 0 200 400 600 0 200 400 600
0
P-1 P-2 P-3 P-4

10

11

12
Max. Elasticity Modulus, E (MPa)

13

14

Fig. 11. Vs velocities in the backfilled area P-1, P-2, P-3 compared to P-4, in natural soil.

differentiate, the noticeable declines in certain regions at seismic speeds (especially at the depth of 4–9 m from the surface), and the
increase in drilling speed in the same regions suggest that the backfill has a heterogeneous structure. Table 2 shows the module values
of the backfill material, which is divided into three layers based on their compactness levels are given in Table 2.
In the field, three inclinometers were located at the beginning in order to monitor the condition of the excavation and their
locations are given in Fig. 12. Inclinometer readings were provided by the owner which were built through a standart procedure. The
inclinometer 1 (INC-1) and inclinometer 2 (INC-2) results are presented in Figs. 13 and 14, respectively. INC-1 readings were
recorded between November 2011 and January 2013. It can be seen from these readings that upper 12 m part is moving. INC-2
readings, from August 2011 to February 2013, indicate similar behavior. As the INC-2 was broken after 3 January 2013, the readings
from the top 16 m are available for the period between January 2014 and November 2014 (Fig. 14b).
Both inclinometers exhibit forward movement at the top (upper 13–14 m). The rate of this movement decreases by time and
stabilizes (refer to −2m or −3m in Fig. 13b and 14c, respectively). On the opposite, below Point f shown in Fig. 17, the bending at
the inclinometers stays almost constant (refer to −18 m in Fig. 13b and Fig. 14c).

Table 2
Module values of the backfill materials.
Level (m) Soil Eo = Eur (MPa) E50 (MPa)

33.3–29 Upper back-fil 250–350 80–120

29–24.6 Intermediate back-fill 180–300 60–100
24.6–18.3 Lower back-fill 500–750 170–250

8
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Fig. 12. Locations of inclinometers (INC) with respect to piles and the Sun Plaza Tower.

Inclinometer readings are the most important indicators in terms of reflecting time-dependent behavior and the resulting me-
chanism. Therefore, they were used during modeling for both obtaining the parameters via back analysis and to compare the behavior
with numerical analysis.

3.2. Plaxis model and back-analysis

First of all, the Plaxis model was constructed on the soil environment set out in Section 4.1 by using the initial soil parameters and
the superstructure parameters were determined by carrying out a procedure to find out the representative stiffness equivalent of the
superstructure. The aim of the superstructure modeling and the analyses presented here is to determine the superstructure stiffness as
precise as possible so that soil-structure interaction can be modeled realistically in the PLAXIS [24] analyses. Here, Probina [25]
structural analysis software was used to create the numerical model of the Sun Plaza Tower and conduct analyses under earthquake
loading. The top floor displacement obtained under earthquake loading retrieved from Probina analyses of the superstructure and the
total structural mass were used to calibrate the “continuous environment model” that is used to represent the structure in PLAXIS
analyses.
The 3-dimensional model created in Probina is shown in Fig. 15. In order to obtain the mass and stiffness of the superstructure to
be used in PLAXIS, the square shaped (27.7 m × 27.7 m) portion of the building, which is the main part that is continuous from the
bottom to the top, was modeled in such detail that it represents the dynamic characteristics and seismic behavior. The story heights
were taken as 3.60 m. Floors were modelled as rigid diaphragms. Mode superposition was used in the seismic analyses and cracked
sections were used to reflect the reduced flexural rigidities of the structural members.
Free vibration analyses revealed the translational periods in x and y directions as 3.56 s (1st mode) and 3.39 s (2nd mode). The
elastic design spectrum of the Turkish Earthquake Code [26] is constructed using the coefficient of effective ground acceleration of
A0 = 0.4 (corresponding to first degree earthquake zone), spectrum characteristic periods of TA = 0.1 s and TB = 0.3 s (corre-
sponding to the strongest soil type), structural importance factor of I = 1.0, and response modification factor of Ra = 6. Live load
contribution factor is n = 0.3. For the modeling of the soil springs under the mat foundation, the modulus of subgrade reaction is
taken as 200,000 kN/m3. As a result of the spectral analysis, the top floor displacements are obtained as 46 cm in x and y directions,
respectively. Parameters used for modeling in PLAXIS is given in Tables 3 and 4.
In order to perform the Plaxis 2D stress-deformation analysis, the parameters set out in Section 4.1 were used as initial para-
meters. However, a reverse analysis was required since there was a difference between these parameters and inclonometer readings.
By comparing the results of the analysis in each stage with the inclinometer readings of that stage, the parameters were revised and
the representable parameters were obtained that minimize the difference with the inclonometer readings. The parameters obtained
from the reverse analyses are given in Tables 5 and 6. Of these parameters, the most influential ones that dominated the reverse
analysis in the order of influence level were the elasticity modulus, the shear strength angle, and the effective cohesion of soils. The
PLAXIS finite element model and mesh (with 6-node elements) is presented in Fig. 16. Non-linear hardening soil model was im-
plemented for soils and rock. On the other hand, piles were modelled with plate elements and anchors were modelled with node-to-
node anchor elements. Geogrid elements provided the contact with anchor and soil. All the structural elements were considered as
elastic. It is shown in Fig. 17 that the inclinometer curve almost coincides throughout the entire depth with the reverse analysis curve
that is obtained using the parameters above. This shows that the reverse analysis study captures the real behavior in the field and the
solution proposals with these representable parameters can be evaluated.
As discussed in the section where inclinometer readings were evaluated, Plaxis analyses confirm the lateral expansion behavior in
the upper region and compression behavior in the lower region of the of the soil beneath the piles. It is understood that the point ‘f’
shown in Fig. 12 is markedly bent. This point, which was once the tip of the shoring pile of Sun Plaza, also coincides with the point
approximately 16 m below the surface where INC-2 is broken. The distribution of the plastic hinges obtained from the Plaxis analysis
of this stage is shown in Fig. 12. It can be seen from the distribution of the plastic hinges shown in Fig. 18, which is compatible with
the assessment made relating to the inclinometer curves and the results of the limit equilibrium analysis, that the upper part of the
soil in between the two piles reaches to the state of collapse by expanding, whereas in the lower regions there is compression. This
compression is in the form of hardening in the upper levels and is at the limit of tension cut-off in the lower levels.

9
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

(a)
Depth (m)

Cumulative Displacement (mm)

(b)
For -2m depth

For -18m depth

Fig. 13. Lateral displacement readings of Inclinometer-1 (INC-1); (a) By depth, (b) By date.

10
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

(a) (b)
Depth (m)

Cumulative Displacement (mm) Cumulative Displacement (mm)

(c) For -3.0m depth

For -18.0m depth

Fig. 14. Lateral displacement readings of Inclinometer-2 (INC-2); (a) By depth from August 2011 to February 2013, (b) By depth after January 2014,
(c) By date.

3.3. Limit state analyses

Limit state equilibrium software LimitState GEO [19] is used in order to define the possible failure mechanisms for the shoring
system and the excavation next to the Sun Plaza Tower. LimitState GEO conducts fast and effective numerical analyses by making use
of the optimization of discontinuity planes. By defining loads and material strength properties, solutions can be obtained through the
possible failure mechanisms which are composed of blocks that slide and/or rotate with respect to each other. The models and the
parameters used for the structure and the soil in the context of limit state analyses in this study are given in Tables 7 and 8. Analyses
are carried out to determine the failure mechanisms of the following stages:

11
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Fig. 15. Structural model – 3-dimensional view and floor plan.

Table 3
Input parameters for defining the floor loads in Plaxis.
Parameters For providing 70 t/m2 stress beneath For providing 52 t/m2 stress beneath For providing 8 t/m2 stress beneath
foundation foundation foundation

Material model Elastic Elastic Elastic

Material type Non-porous Non-porous Non-porous
Unit weight, γ (kN/m3) 5.0 4.2 4.0
Elasticity modulus, E (MPa) 175,000 175,000 175,000
Poisson ratio,ν 0.45 0.45 0.45

*If the displacement of the pile wall exceeds 2.0 cm displacement at static loads and additional 2.0.

Table 4
Characteristic parameters of the structural elemens.
Parameters Mat Foundation Mat Foundation Wall Wall Footing Pile ϕ65/100 Pile ϕ100/125 Vertical structural Slabs
t=2m t = 0.5 m t = 0.3 m t = 0.8 m members

EA, kN/m 50,000,000 12,500,000 7,500,000 20,000,000 8,200,000 13,000,000 5,000,000 3,000,000
EI, kN/m 16,667,000 260,000 56,250 1,067,000 220,000 817,000 25,000 2,500
d, m 2.0 0.5 0.3 0.8 0.57 0.87 0.25 0.1
w, kN/m/m 0 0 7.5 20 1.7 3.5 0 0
v (nu) 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2

Stage 1 – First excavation stage (up to level +33.3 m)

Stage 2 – Final (second) excavation stage (from +33.3 m up to level +18.30 m)
Stage 3 – Partial back-fill stage (up to the top of mat foundation, level +33.3 m)
Stage 4 – Complete back-fill stage (up to level +48.2 m)
The failure mechanisms for the stages defined above are given in Fig. 18: In the first stage, it is seen that the upper region
including the first row of piles and the soil behind it moves towards the excavation site. In the second excavation stage, where the
final excavation level which corresponds to +18.30 m is reached, no movement is observed in the lower region stabilized by the
second (bottom) row of piles. Lateral movements do not reach the foundation of the tower. It is clear that the detrimental portion of
the lateral movements occur in the first row of piles whereas no adverse effect was created by the excavation of the lower part. This
observation is also compatible with the inclinometer readings. For the partial back-fill stage (where anchors are removed in the
model, as they are temporary and would completely loose their operational capacity) which corresponds to the stage at which the
authors were employed in the investigation, it is seen that the service road made up of crushed stone fill and the retaining wall of the
service road are also affected in addition to the region including the first row of piles and the soil behind it (see damaged zone in
Fig. 17). The movements do not reach the foundation of the tower in this stage, either. In case, the excavation is completely filled
back (stage 4), evidently, the reinforced concrete retaining wall by the tower side (shown with label D Figs. 4 and 6) finds a

12
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Table 5
Parameters used for soils/rocks in Plaxis.
Parameter Made soil Gravel backfill Weathered claystone Claystone

Material model HS HS HS HS
Material type Drained Drained Drained Drained
Unit weight, γ (kN/m3) 17.5 19 21 23
Initial void ratio, eini 0.6 0.5 0.45 0.35
Interface load reduction factor 0.60 0.62 0.62 0.70
ref
Deviator reference modulus, E50 (MPa) 10 15 50 400
ref
Compression ref. modulus, Eoed (MPa) 10 15 50 400

Loading-unloading modulus, ref

Eur (MPa) 30 45 150 1200
Dependence of stress level, m 1.0 1.0 1.0 1.0
Poisson ratio for loading-unloading, νur 0.2 0.2 0.2 0.2
Reference stress, pref, kPa 100 100 100 100
Failure ratio, Rf 0.9 0.9 0.9 0.9
Effective cohesion, c’ (kPa) 1 1 50 200
Shear strength angle, ϕ’ (°) 27 28 20 30

Table 6
Parameters used for backfill material in Plaxis.
Parameter Backfill-3 (33.3–29 m) Backfill-2 (29–24.5 m) Backfill-1 (24.5–18.3 m) New fill (33.3–46.3 m)

Material model HS HS HS HS
Material type Drained Drained Drained Drained
Unit weight, γ (kN/m3) 19 20 21 21
Initial void ratio, eini 0.6 0.7 0.5 0.4
Interface load reduction factor 0.62 0.62 0.62 0.70
ref
Deviator reference modulus, E50 (MPa) 120 80 150 200
ref
Compression ref. modulus, Eoed (MPa) 120 40 150 200

Loading-unloading modulus, Eur ref

(MPa) 360 240 450 600
Dependence of stress level, m 1.0 1.0 1.0 1.0
Poisson ratio for loading–unloading, νur 0.2 0.2 0.2 0.2
Reference stress, pref, kPa 100 100 100 100
Failure ratio, Rf 0.9 0.9 0.9 0.9
Effective cohesion, c’ (kPa) 5 5 5 5
Shear strength angle, ϕ’ (°) 35 32 38 40

Fig. 16. Finite element mesh of the model.

13
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Dilative
(active)
behaviour

Depth (m)
Compressive
(passive)
behaviour

Cumulative Displacement (mm)

Fig. 17. Inverse analysis of inclinometer 2 (INC-2) and corresponding soil behavior.

counterbalance effect from which the stability of the service road and service road retaining wall (shown with label C Figs. 5 and 6)
can also be affected (see Fig. 19).

3.4. Seismic risk analysis

As long as the representable parameters and an appropriate model are used, it is possible to define a realistic soil behavior and
soil-structure interaction in Plaxis. The required stiffness and mass distribution for representing the structural system of the Sun Plaza
Tower in Plaxis is obtained from the results of the superstructure analysis conducted in ProBina software. Models created in each
software are shown in Fig. 20(a and b). The unit weights of the volumetric elements are assigned and calibrated such that the soil
stress distribution obtained under dead load + live load (G + Q) in Plaxis is compatible with that obtained in ProBina. Soil stress
distribution obtained in ProBina under G and Q load cases are shown in Fig. 21. After calibration, the soil stress distribution obtained
in ProBina under G + Q loading results in 700 kN/m2 at the central part and 520 kN/m2 at the edges.
As the next step of calibration, behavior under lateral loading is considered. The stiffness and mass distribution is calibrated such
that the top floor displacement obtained under earthquake loading in Plaxis is compatible with the one obtained in ProBina. In Plaxis
analysis, earthquake loading is defined via a pseudo-dynamic approach which requires the use of coefficient of lateral acceleration
coefficient, Ch. The effective ground acceleration is A0 = 0.3 and Ch for the soil analyses is assumed to be 40% of this value,
Ch = 0.12. Thus, the model in ProBina is analyzed under 0.12 g and the top floor displacement is obtained as 46 cm using R = 1.
Approximately the same lateral displacement (47 cm) is obtained in the Plaxis model as shown in Fig. 22. Note that shopping center
which shares the same mat foundation with the tower is not included in this tower-only model.
For a more precise modeling that reflects actual conditions, this shopping-center building and the common mat foundation
situation is reflected in the Plaxis modeling and thus, the original lateral displacements (before the excavation and partial back-fill
operations) under earthquake loading is obtained as approximately 42 cm (42.1 towards shopping-center Building and 41.7 cm
towards opposite direction) as shown in Fig. 22. This case is used as the target case for developing solution proposals for the problem
caused by the excavation. That is, the best solution is expected to provide same lateral displacements as this original case. This is

14
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

(1) Model for the first excavation stage Expected failure mechanism for Stage 1

(2) Model for the final (second) excavation stage Expected failure mechanism for Stage 2

(3) Model for the partial back-fill stage Expected failure mechanism for Stage 3

(4) Model for the complete back-fill stage Expected failure mechanism for Stage 4

Fig. 18. Limit state failure mechanisms according to different excavation stages.

Table 7
Models and parameters for the structure.
Structural Member Material Model Drainage Condition γn (kN/m3) cu (kN/m2)

Floor Load Rigid – 6,0 –

Concrete Mohr Coulomb Undrained 25,0 300–4000

15
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Table 8
Soil models and parameters.
Soil Type Soil Model Drainage Condition γn (kN/m3) γsat (kN/m3) c (kN/m2) ϕ°

01_Fill Mohr Coulomb Drained 17,5 19,0 1 27

02_Gravelly sand Mohr Coulomb Drained 19,0 20,0 0 28
03_Weathered claystone Mohr Coulomb Drained 20,0 21,0 50 20
04_Claystone Mohr Coulomb Drained 22,0 23,0 200 30
05_Crushed stone Mohr Coulomb Drained 19,0 20,0 0 28

Fig. 19. Verification of limit state analysis via Plaxis analysis.

(a) ProBina Model (b) Plaxis Model

Soil under the mat

Springs are used under foundation is modeled
mat foundation using volumetric elements

Fig. 20. Models created in ProBina and Plaxis.

16
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

(a) (b)

Stress Stress
(t/m2) (t/m2)

Fig. 21. Soil stress obtained in ProBina under (a) Dead load – G and (b) Live Load – Q.

46.5 cm 47.3 cm 42.1 cm 41.7 cm 41.2 cm 47.1 cm

(a) (b) (c)

Fig. 22. Top floor displacements (in −x and +x directions) obtained in Plaxis for model calibration; (a) for tower-only model (b) for actual model
(Tower + shopping-center building on the common mat foundation), (c) for the case of partially filled excavation.

investigated in the proceeding part of this section.

The original dynamic behavior shown in Fig. 22(b) is changed due to the excavation. It is possible to observe this phenomenon by
examining the top floor displacements for partially filled case and comparing it to the original case reported above. As seen in
Fig. 22(c), for the partially filled case, the displacements in + x and –x directions are different (about 6 cm difference) and the
displacement towards the excavation site is significantly increased compared to the original case.
Using the calibrated Plaxis model, the displacements under the mat foundation is calculated for the original and partially back-
filled cases in order to determine the effect of excavation and back-filling operations. As seen in Fig. 23, the total displacements
increased from 4.3 mm (original case) to 7.9 mm (partially back-filled case). Although the displacement under the mat foundation is
increased, it is clear that it is still quite below acceptable limits.

Fig. 23. Displacements under foundation for (a) original and (b) partially back-filled cases.

17
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Fig. 24. Deformed shape of the reinforced concrete retaining wall: (a) before excavation, (b) partially back-filled excavation, (c) fully back-filled
excavation – results of PLAXIS analysis.

3.5. Reinforced concrete wall analysis

Although the reinforced concrete shear wall (shown with label D in Figs. 4 and 6) is not a part of the main structural system of the
Sun Plaza Tower, safety of this wall is very important as it may cause loss of life and money in case it fails towards the tower
structure. Furthermore, it is vital for the safety of the service road as it directly supports the crushed stone fill forming the service
road. Therefore, also taking the field observations into account (see Fig. 6 for cracks in the wall and water seeping through the
cracks), this wall is assessed by conducting detailed analyses both in Plaxis and SAP2000 [22].
Results of the finite element analysis show that the deformations and related internal forces reached an unacceptable level in the
original case by the static loading due to crushed stone fill before the excavation. The deformed shape obtained in Plaxis corre-
sponding to original (before the excavation), partially back-filled and fully back-filled stages are shown in Fig. 24. Corresponding
lateral peak displacements are 24 mm, 19 mm and 20 mm, respectively.
In addition to Plaxis analysis, in order to investigate the impact of solution proposals regarding this wall, it is also modeled and
analyzed in [22]. In these independent analyses, the tower structure is used as a support and the deformations obtained are therefore

Fig. 25. Deformed shape of the reinforced concrete retaining wall under static loading case – relative deformations obtained in [22].

18
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Table 9
Top floor displacements of the tower under earthquake load for different solution alternatives.
Solution Alternatives Left (−x) Right (+x)
cm cm

Alternative-1: Removing the connection beams (forming expansion joint) 40.1 48.7
Alternative-2: Removing the connection beams (forming expansion joint) and remediating the crushed stone by injection 29.0 49.1
Alternative-3: Extending the foundation after removing the crushed stone and forming a frame before the back-fill to + 48.2 m level 28.9 48.7

Alternative-4: remediating the crushed stone by injection and back-filling to + 48.2 m level 37.1 41.7
Alternative-5: Adding additional connection beams/fixing the beams and back-filling to + 48.2 m level 39.2 44.3
Alternative-6: Adding additional connection beams/fixing the beams and back-filling to + 48.2 m level after additional piles 39.7 43.2

relative deformations that directly cause internal forces in the wall. The deformed shape obtained form this analysis is shown in
Fig. 25. As seen, the peak relative deformations are obtained as 12 mm.

4. Discussion and proposed solution alternatives

When the results of the analysis summarized above is evaluated together, it is found out that existing state of the partially back-
filled excavation does not pose a direct threat to the main structural system of the Sun Plaza Tower. Rather, it is detected that the
problem is localized in the part which is marked as “damaged zone” in Fig. 4 which encompasses (i) the service road, (ii) the retaining
walls supporting this service road, (iii) the shoring piles at the border of the excavation site, (iv) the soil in between these piles and (v)
the service road retaining wall. Regardless, the safety of this region had to be assured in order to prevent any potential total failure
which may be faced particularly in case of an earthquake leading to loss of lives and money. The solution alternatives developed by
the authors are therefore focused on this region and the criteria set was to reach global equilibrium conditions. In this context, the top
floor displacements and particularly the difference between the top floor displacements in +x and –x directions under earthquake
loading are monitored. The proposed solution alternatives and the top floor displacements under earthquake loading corresponding
to these cases in +x and –x directions are listed in Table 9. The target performance was to minimize the difference between the top
floor displacements in +x and –x directions. When Table 9 is assessed taking this criteria into account, it is concluded that 4, 5, and 6
are the best solution alternatives. However, alternative 4 is not preferred as it includes a grout injection operation which may be
uncontrollable and leakage through the crushed stone fill may also have occurred. Moreover, it adds extra lateral load on the
reinforced concrete retaining walls which are already cracked and damaged. Thus, alternatives 5 and 6 are the ones that would
provide the best solution. Although both can be used, alternative 6 where additional row of piles are used, may be a better option due
to the following reasons:

(1) Inclinometer readings showed that the upper (first) row of piles are displaced significantly which might have caused considerable
deformations at the bottom sections of these piles. With that purpose, a new row of piles is proposed to share the loads with the
existing piles.
(2) Proposed new piles are also helpful to reduce the loads transferred to the retaining walls from the back-fill.

4.1. Additional connection beam construction (lateral struts)

Before backfilling, safety measures need to be taken against possible additional loads which the retaining wall may be exposed to.
In this context, in order to reduce the loads on the existing beams and transfer the loads to the new beams, and also to reduce the
internal forces in the wall to reasonable and safer levels, an “additional link beam manufacturing” is proposed in this section.
Accordingly, 35 pieces of HE300B or equivalent struts with a length of 160 cm (as the gap distance is 160 cm) and a minimum of 215
tonf axial load capacity shall be provided. In the case of the proposed struts, the behavior of the system changes favourably where the
maximum deformations of the reinforced concrete wall decreases from 12 mm to 8 mm (See Fig. 26) and the internal forces are
reduced.

4.2. Compacted backfill

The deep excavation must be filled up to the level of +48.2 as shown in Fig. 12. Filling should be completed in between the front
of Sun Plaza, the shopping center, and two neighbouring buildings with an inclination that will not hold water (i.e. 1–2%). However,
3 horizontal: 1 vertical slope (see Fig. 27) should be kept to the extent that the land boundaries permit where it is necessary to be re-
acclimatized to the general land structure. The material to be used for filling must be special furnace material. In this regard,
0–60 mm well-graded crushed stone containing stone powder should be filled up to +46.0 m. Above this level, to prevent water
infusion, 0–40 mm well-graded crushed stone with a greater amount stone powder was recommended. Moreover, a peripherial
drainage system was also proposed.

19
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

Fig. 26. Application points of struts, details of struts and the deformation of the retaining wall.

+48.2 m
%1-2 slope
3H : 1V
3H : 1V

Fig. 27. Recommended level of filling.

4.3. Supplementary row of piles

Since it is determined that the existing piles became plasticized and unable to perform their duties based on the inclinometer
readings and analyses conducted in this paper, it is suggested to construct bored piles with diameters D = 0.8 m, at intervals,
s = 1.2 m and length L = 22.2 m in order to reduce the lateral forces that will be exerted to the structure by the new filling. Another
important task that can be undertaken by these piles is to contribute to the protection of the service road and the Sun Plaza retaining
walls if the excavation pit was reopened for some reason in the future. However, it should be noted here that the degree of this
contribution is limited and these piles cannot be used as a support system to such an excavation. In such a case, the necessary analysis
for the excavation and its environment should be conducted and all necessary measures –including the establishment of an appro-
priate support system- should be provided by those who undertake the excavation operation. The placement of the piles and the
drainage wells required is shown in Fig. 28. Although the pile construction level is 46.3 m, the piles can be constructed at any level
between 46.3 and 48.2 m.

20
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

(a) Section view (b) Plan view

Fig. 28. Placement of the new piles and the drainage wells.

Fig. 29. Current view from the excavation.

4.4. Construction stages

The order of construction processes is important. Following the order below is necessary in order to eliminate the risks: (1) the
formation of the connecting beams supporting the reinforced concrete retaining wall; (2) formation of the newly compressed filling
up to +46.3 m; (3) construction of new piles; (4) completion of the remaining fill; (5) opening of drainage wells (see Fig. 29).

5. Conclusions

This paper presents a comprehensive study that includes site observations, examination of the previous documents and reports,

21
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

modeling of the ground and the structure, and analytical evaluations considering the interaction of soil and the structure in in-
vestigating the impact of nearby deep excavation on the high-rise Sun Plaza building and the development of remediation solutions.
The analysis of the problem and the full process of the diagnosis, the development, and the implementation of the solution proposals
and the situation after the application are explained herein.
Previous studies conducted by various organizations and individuals were mainly based on site observations and included as-
sessment of such observations. However, it is shown in this study that a healthy solution could only be reached by a proper meth-
odology that includes (i) a complete soil characterization and stratification for use in numerical modeling and analyses, (ii) cali-
bration of the model parameters and determination of failure mechanism through instrumentation data and conducting back
analyses, (iii) consideration of the construction stages in the numerical modeling and analyses, (iv) taking the interaction of the soil
and structure into account in seismic analyses particularly in seismic prone regions.
Interaction of the soil/excavation and structure can be taken into account via various methods. In this study, a practical and
convenient approach, where a geotechnical software, Plaxis, and a structural analysis software, Probina, used in coupled way, is
implemented. The representative stiffness equivalent of the superstructure was obtained using Probina and Plaxis was used to model
the soil environment by using the initial soil parameters and the superstructure parameters. The top floor displacement obtained
under earthquake loading retrieved from Probina analyses of the superstructure and the total structural mass were used to calibrate
the “continuous environment model” that was used to represent the structure in Plaxis analysis. In this context, Plaxis was used to
obtain lateral displacement of top floor for various configurations and seismic actions. Numerical analyses revealed that the original
dynamic behavior was changed due to the excavation. The proposed solution alternatives were assessed in terms of their capability in
providing the same or close top floor displacement to the original configuration (before the excavation and partial back-fill opera-
tions) towards both sides, i.e to the excavation and away from the excavation site. That is, the best solution was expected to provide
same lateral displacements as this original case under seismic loading.
Use of additional structural supports for the retaining wall and back-filling the excavation to the top level after use of a row of
additional piles in front of the retaining wall was selected as the best solution. The actual application of the proposed solution and the
positive outcome of this, in terms of both ensuring structural safety and gaining new tenants, are reported.
One of the biggest problems regarding new high-rise construction in densely populated large metropolitan areas is the necessity of
constructing significant shoring systems since very few parcels left in these areas are typically squeezed between existing high-rise
buildings. In this examined case, since the placement of the new structure in the deep excavation could not have been completed
within the planned time frame due to the bureaucratic and legal disputes the useful lifetime of the temporary anchorages have
expired and safety concerns appeared. Furthermore, the construction license was permanently cancelled which made the situation
worse. In order to avoid such situations in the future, the attention of the city managers and law makers should be drawn to such
potential problems and necessary regulations including revision of codes which enforce numerical analysis that takes structure
excavation interaction into account should be considered. In other cases, such deep excavations may even lead to major failures
causing loss of lives and damage to property. As part of the remediation process, use of permanent anchorages, construction of
retaining walls or use of piled shoring systems may be considered.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgments

Authors would like to express their appreciation to Civil Engineer Mert Tiber and Lawyers Yesim Dere and Ceren Arslan for their
support and help and SunPlaza for providing necessary data.

References

[1] M. Abdallah, F. Hage Chehade, W. Chehade, A. Fawaz, Interaction deep excavation-adjacent structure: numerical two and three dimensional modeling, Adv.
Mater. Res. 324 (2011) 344–347.
[2] D. Adam, R. Markiewicz, J.-D. Deix, Donau City Tower 1 – Deep Foundation, Excavation and Dewatering Scheme for the 220 m Tall High-rise Building in Vienna.
Vplyv vody na geotechnické konštrukcie, Bratislava, 3–4 June, 2013.
[3] A. Akhtarpour, M. Mortezaee, Dynamic response of a tall building next to deep excavation considering soil-structure interaction, Asian J. Civil Eng. 20 (2019)
479–502.
[4] Arcadis (EC HARRIS ITALIA SRL), Geotechnical Advisory Report, Sun Plaza Building and Adjacent Plot of Land, 2014.
[5] F. Baguelin, J.F. Jézéquel, D.H. Shields, The Pressuremeter and Foundation Engineering, Trans Tech Publications, Series on Rock and Soil Mechanics, 1978.
[6] E. Berkman, Expert Witness Report submitted to T.C. İstanbul 14. Civil Court of Peace, File No: 2013/121 D. İş, 2013 (in Turkish).
[7] J. Birks, K.J. Chan, T.Y. Yap, Sun Plaza Building Impact Assessment Report – Effect of Excavation and Backfilling Operations in Map Sheet 2, Block 1, Parcel 136.
Mott MacDonald Report, 2014.
[8] P. Castaldo, M. Calvello, B. Palazzo, Probabilistic analysis of excavation-induced damages to existing structures, Comput. Geotech. 53 (2013) 17–30.
[9] P. Castaldo, M. Deluliis, Effects of deep excavation on seismic vulnerability of existing reinforced concrete framed structures, Soil Dyn. Earthq. Eng. 64 (2014)
102–112.
[10] Center d’Etudes Menard, Detérmination de la Possée Excercée par un Sol sur une Paroi de Souténement, Publication D/38/70, 1970.
[11] H. Demiroglu, Expert Witness Report submitted to T.C. İstanbul 6. Civil Court of Peace, File No: 2014/28 D. İş, 2014 (in Turkish).
[12] Emlakeki, News about Buildings: Sun Plaza, 2014. https://www.emlakeki.com/32-katli-plazada-cukur-korkusu-haberi-21339.
[13] R.J. Finno, F.T. Voss Jr., E. Rossow, T.J. Blackburn, Evaluating damage potential in buildings affected by excavations, J. Geotech. Geoenviron. Eng. 131 (10)

22
S. Oztoprak, et al. Engineering Failure Analysis 111 (2020) 104495

(2005) 1199–1210.
[14] K.H. Goh, R.J. Mair, Building damage assessment for deep excavations in Singapore and the influence of building stiffness, J. Geotech. Eng. 42 (2011) 1–12.
[15] E. Guler, Measures to be taken to eliminate the effects of the deep excavation on the neighboring Sun Plaza Building, Geotechnical Evaluation Report, 2013.
[16] E. Guler, Effects of backfill study on the safety of the Sun Plaza Building, Geotechnical Evaluation Report, 2014.
[17] G.A. Horodecki, E. Dembicki, Impact of deep excavation on nearby urban area, in: Proc 14th Eur. Conf. Soil Mech. Geotech. Eng., 2007, pp. 575–580.
[18] Istanbul Gercegi, News about Sun Plaza, 2014. https://www.istanbulgercegi.com/maslaktaki-insaat-cukuru-plazalarda-panik-yaratiyor_67993.html.
[19] LimitState GEO, User Manual, 2010.
[20] MPI Engineering, Static Report of Sun Plaza, 2002.
[21] Y.L. Mo, J.M. Hwang, Seismic response of reinforced concrete buildings close to deep excavation, Comput. Methods Appl. Mech. Eng. 145 (1–2) (1997) 117–134.
[22] SAP2000, Computers and Structures Inc., Linear and Nonlinear Static and Dynamic Finite Element Analysis and Design Of Three-Dimensional Structures,
Berkeley, USA, 2005.
[23] M.J. Schuster, C.H. Juang, M.J.S. Roth, D.V. Rosowsky, Reliability analysis of building serviceability problems caused by excavation, Géotechnique 58 (9) (2008)
743–749.
[24] Plaxis, Material Models. Delft University, 2015.
[25] Probina, Construction Design System. Prota Computer, 2013.
[26] Turkish Earthquake Code, Regulations of Structures Constructed in Disaster Regions, Ministry of Public Works and Settlement, Ankara, Turkey, 2007.
[27] Wiss, Janney, Elstner Associates, Inc., Sun Plaza Design Review, 2003.
[28] N. Yeganeh, J.B. Bazaz, A. Akhtarpour, Seismic analysis of the soil-structure interaction for a rise building adjacent to deep excavation, Soil Dyn. Earthq. Eng. 79
(2015) 149–170.

23