PERFORMANCE BASED DESIGN OF R/C TALL BUILDINGS IN TWO

DIFFERENT SEISMIC ZONES

A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY

BY

MUSTAFA GÜNEŞ ÇİFTÇİOĞLU

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR
THE DEGREE OF MASTER OF SCIENCE
IN
CIVIL ENGINEERING

JANUARY 2023
 Approval of the thesis:

PERFORMANCE BASED DESIGN OF R/C TALL BUILDINGS IN TWO

DIFFERENT SEISMIC ZONES

submitted by MUSTAFA GÜNEŞ ÇİFTÇİOĞLU in partial fulfillment of the

requirements for the degree of Master of Science in Civil Engineering, Middle
East Technical University by,

Prof. Dr. Halil Kalıpçılar

Dean, Graduate School of Natural and Applied Sciences

Prof. Dr. Erdem Canbay

Head of the Department, Civil Engineering

Prof. Dr. Barış Binici

Supervisor, Civil Engineering Dept., METU

Examining Committee Members:

Prof. Dr. Ahmet Yakut

Civil Engineering Dept., METU

Prof. Dr. Barış Binici

Civil Engineering Dept., METU

Prof. Dr. Kağan Tuncay

Civil Engineering Dept., METU

Prof. Dr. Yalın Arıcı

Civil Engineering Dept., METU

Assoc. Prof. Dr. Alper Aldemir

Civil Engineering Dept., Hacettepe University

Date: 20.01.2023
I hereby declare that all information in this document has been obtained and
presented in accordance with academic rules and ethical conduct. I also declare
that, as required by these rules and conduct, I have fully cited and referenced
all material and results that are not original to this work.

Name, Last name : Mustafa Güneş Çiftçioğlu

Signature :

iv
 ABSTRACT

PERFORMANCE BASED DESIGN OF R/C TALL BUILDINGS IN TWO

DIFFERENT SEISMIC ZONES

Çiftçioğlu, Mustafa Güneş

Master of Science, Civil Engineering
Supervisor : Prof. Dr. Barış Binici

January 2023, 326 pages

Performance-based design of R/C tall structures has become mandatory with Turkish
Earthquake Code (2018). While some professionals think that linear capacity design
is sufficient, others have supported this new approach only in regions where
earthquake is more effective. Nonlinear design requires comprehensive and detailed
theoretical knowledge. The compatibility of Perform 3D software, widely used in
this design, with experiment results is unknown and calibration is not mandatory in
the regulation. It is another unknown to what extent calibration will change the
design results.

For this purpose, study was carried out on a tall structure with a height of 112 m
above basement floor with flat slab, core-wall group and link beams. The design was
made according to linear and nonlinear analysis for two different earthquake zones,
i.e., Ankara and Istanbul. Necessity of nonlinear analysis in tall structures for both
earthquake zones has been confirmed. In order to make the design compatible with
experiments, calibration work was carried out for rectangular, T-shaped and U-
shaped walls and diagonally and conventionally reinforced link beams. With
calibration study, it was seen that cyclic degradation factors were very effective and
numerical calibration results were given. The model concrete strain values of walls

v
were very small from the experimental values. For this, the correction coefficient "a”
and formulation for walls concrete strain results are proposed. Ideal wall mesh
modeling has been proposed horizontally and vertically.

The results of models with calibrated and 2 alternative uncalibrated cyclic

degradation factors were compared, with results that were up to 30% lower in the
absence of cyclic degradation factors. Low cyclic energy degradation factors are
appropriate and sufficient for the design. As a result of all this study, some
suggestions have been made for structural engineers and for some clauses of the
Turkish Earthquake Code (2018).

Keywords: Nonlinear performance-based design, linear design, reinforced concrete

tall structures, Perform 3D software cyclic degradation factors, wall and link beam
calibration study

vi
 ÖZ

FARKLI İKİ DEPREM BÖLGESİNDEKİ BETONARME YÜKSEK

YAPILARIN PERFORMANSA DAYALI TASARIMI

Çiftçioğlu, Mustafa Güneş

Yüksek Lisans, İnşaat Mühendisliği
Tez Yöneticisi: Prof. Dr. Barış Binici

Ocak 2023, 326 sayfa

Betonarme yüksek yapıların performansa dayalı tasarımı Türkiye Deprem

yönetmeliği (2018) ile zorunlu hale gelmiştir. Bazı uzmanlar geleneksel kapasite
tasarımının yeterli olduğunu düşünürken, bazılarıysa sadece depremin daha etkin
olduğu bölgelerde bu yeni yaklaşımını desteklemiştir. Doğrusal olmayan tasarım
kapsamlı ve detaylı teorik bilgi birikimine sahip olmayı gerektirmektedir. Bu
tasarımda yaygın olarak kullanılan Perform-3D yazılımının deney sonuçları ile
uyumu bilinmemektedir ve yönetmelikte de kalibrasyon zorunlu değildir.
Kalibrasyon çalışmasının tasarım sonuçlarını ne ölçüde değiştireceği bir başka
bilinmezliktir.

Bu amaçla kirişsiz döşemeli, çekirdek perde grubu ve bağ kirişli bodrum katı üzeri
112 m yüksekliğindeki yüksek bir yapı üzerine çalışma yapılmıştır. Ankara ve
İstanbul gibi iki farklı deprem bölgesi için doğrusal ve doğrusal olmayan analize
göre tasarım yapılmıştır. Her iki deprem bölgesi için yüksek yapılarda doğrusal
olmayan analizin gerekliliği doğrulanmıştır. Deneylerle uyumlu tasarımın yapılması
için dörtgen, T şeklinde ve U şeklinde perdeler ile çapraz ve düz donatılı bağ kirişler

vii
için kalibrasyon çalışması yapılmıştır. Bu kalibrasyon çalışması ile çevrimsel enerji
sönüm parametrelerinin çok etkin olduğu görülmüştür ve sayısal kalibrasyon
sonuçları verilmiştir. Perde yapılarının model beton birim kısalma değerleri deney
değerlerinden çok küçük çıkmıştır. Bunun için perde beton birim kısalma değerleri
için “a” artırım katsayısı ve formülasyonu önerilmiştir. İdeal perde örgü modelleme
aralıkları yatay ve düşeyde önerilmiştir.

Kalibrasyonlu ve seçilen 2 alternatif kalibrasyonuz çevrimsel enerji sönüm

parametre değerlerine sahip modellerin sonuçları karşılaştırılmış, enerji sönüm
parametreleri kullanılmaması durumunda %30’a varan düşük sonuçlar çıkmıştır.
Düşük enerji sönüm parametre değerleri tasarım için uygun ve yeterlidir. Tüm bu
çalışma sonucunda yapı mühendisleri ve Türkiye Deprem Yönetmeliği (2018) bazı
kaideleri için önerilerde bulunulmuştur.

Anahtar Kelimeler: Doğrusal olmayan performansa dayalı tasarım, doğrusal tasarım,

betonarme yüksek yapılar, Perform 3D çevrimsel enerji sönüm parametreleri, perde
ve bağ kirişi kalibrasyon çalışması

viii
To My Precious Wife

ix
 ACKNOWLEDGMENTS

This thesis has been a work that covers the whole of my twelve years of professional
engineering experience in the field of high-rise buildings and my academic
knowledge afterwards. I am very grateful to the major people in my life who have
been instrumental in my upbringing and sacrifices in the creation of this work.

First of all, it was a very important chance for me that Prof. Dr. Barış Binici, one of
the most competent instructors in Türkiye on reinforced concrete structure behavior,
which has an important place in the content of the thesis, became my supervisor. In
addition to the very important information, I learned from him, I would like to thank
him for his guidance, advice, encouragement and very valuable support during the
thesis phase.

I would like to thank to Ozkan Hakan, MSc. Civil Engineer, with whom I exchanged
information and did not spare his technical support.

Finally, during this thesis phase, I would like to give the greatest thanks to my biggest
supporter, the mother of my two daughters, my precious wife, who gave me the
opportunity to do these academic studies in my busy business life, who showed great
sacrifice and patience. The completion of this work is a fact that I cannot achieve
without my precious wife, who supports me, gives me strength and is always by my
side.

x
 TABLE OF CONTENTS

ABSTRACT ...............................................................................................................v

ÖZ ........................................................................................................................... vii

ACKNOWLEDGMENTS .........................................................................................x

TABLE OF CONTENTS ......................................................................................... xi

LIST OF TABLES ...................................................................................................xv

LIST OF FIGURES ............................................................................................... xix

LIST OF ABBREVIATIONS ............................................................................ xxxiii

LIST OF SYMBOLS ...........................................................................................xxxv

CHAPTERS

1 INTRODUCTION .............................................................................................1

1.1 Motivation .................................................................................................. 1

1.2 Research Questions .................................................................................... 3

1.3 Outline of Thesis ........................................................................................ 5

2 SOME PRELIMINARIES FOR PERFORMANCE BASED DESIGN

CHECK ACCORDING TO TEC (2018) ...................................................................9

2.1 Buildings Types and Target Building Performance Levels ....................... 9

2.2 Structural Member Actions ...................................................................... 12

2.3 Design Earthquake Load for PBD ............................................................ 14

2.4 Structural Member Modeling ................................................................... 17

2.4.1 Effective Stiffness ............................................................................. 18

2.4.2 Expected Material Strengths ............................................................. 19

xi
 2.4.3 Nonlinear Structural Member Model Types...................................... 20

2.5 Evaluating the Results of Performance Analysis ...................................... 25

3 LINEAR ANALYSIS AND DESIGN OF THE CASE STUDY BUILDING 29

3.1 CASE STUDY BUILDING...................................................................... 29

3.1.1 Building Description and Dimensions............................................... 29

3.1.2 Structural System of The Case Study Building ................................. 32

3.2 DESIGN CRITERIA ................................................................................ 35

3.2.1 Seismicity .......................................................................................... 35

3.2.2 Materials ............................................................................................ 35

3.2.3 Design Loads ..................................................................................... 37

3.2.4 Load Combinations ........................................................................... 40

3.3 MODELING of CASE STUDY BUILDING ........................................... 40

3.4 DESIGN FLOW CHART ......................................................................... 42

3.5 LINEAR ELASTIC ANALYSIS STAGE ................................................ 44

3.5.1 Base Shear Calculation ...................................................................... 44

3.5.2 Modeling Details ............................................................................... 49

3.6 DESIGN OF SHEAR WALLS AND LINK BEAMS .............................. 53

3.6.1 Shear Wall Design of Case Study Building ...................................... 53

3.6.2 Coupling Beam Design of Case Study Building ............................... 71

3.7 DRIFT RATIO RESULTS OF CASE STUDY BUILDING FOR

LINEAR ELASTIC DESIGN ............................................................................. 76

4 PERFORM 3D V7.0.0 SOFTWARE .............................................................. 79

4.1 MODELING PHASE ............................................................................... 79

4.1.1 Nodes ................................................................................................. 79

xii
 4.1.2 Components Properties ..................................................................... 80

4.1.3 Compound Components.................................................................... 92

4.1.4 Wall and Link Beam Elements ......................................................... 95

4.1.5 Other Modeling Phase Subjects ...................................................... 100

4.2 ANALYSIS PHASE .............................................................................. 101

4.2.1 Load Cases ...................................................................................... 101

4.2.2 Analysis Sub-Options and Damping ............................................... 102

4.2.3 Energy Balance ............................................................................... 105

5 STRUCTURAL WALL AND LINK BEAM CALIBRATON WORK ........107

5.1 CALIBRATION WORK FOR STRUCTURAL WALLS ..................... 108

5.1.1 Experimental Data for Sample Rectangular and T-Shape Walls .... 108

5.1.2 Experimental Data for Sample U Shape Wall ................................ 113

5.1.3 Simulation of Walls Test Samples at Perform 3D V7.0 ................. 119

5.2 CALIBRATION WORK FOR LINK BEAMS ..................................... 177

5.2.1 Experimental Data for Link Beams ................................................ 178

5.2.2 Simulation of Link Beam Test Samples at Perform 3D V 7.0 ........ 183

6 PERFOMANCE BASED DESIGN OF THE CASE STUDY BUILDING ..203

6.1 Performance Based Design Criteria ....................................................... 203

6.1.1 Design Earthquake Loads ............................................................... 203

6.1.2 Some Design Criteria for The Case Study Building ....................... 206

6.2 Model of The Case Study Tall Building ................................................ 208

6.3 Performance Based Design Results of The Case Study Buildings ........ 226

6.3.1 Interstory Drift Ratios ..................................................................... 227

6.3.2 Shear Force in the Walls ................................................................. 231

xiii
 6.3.3 Strain Demands ............................................................................... 236

6.3.4 Rotation Demand in Beams ............................................................. 249

6.3.5 Performance of Diagonally Reinforced Link Beams ...................... 258

6.3.6 Shear Load Capacity Check of Beams ............................................ 260

6.3.7 Energy Dissipation Results.............................................................. 265

6.3.8 Summary of Performance Based Design of The Case Study Building

269

7 RESULTS AND INFERENCES OF THE THESIS WORK ........................ 275

REFERENCES ...................................................................................................... 289

APPENDICES ....................................................................................................... 291

A. Shear Load Capacity Diagrams of Walls for Three Alternative Models of

the Case Study Building in Ankara (Performance Based Design) .................... 291

B. Shear Load Capacity Diagrams of Walls for Three Alternative Models of

the Case Study Building in Istanbul (Performance Based Design) ................... 299

C. Strain Check of Walls for Three Alternative Models of the Case Study
Building in Ankara (Performance Based Design) ............................................. 307

D. Strain Check of Walls for Three Alternative Models of the Case Study
Building in Istanbul (Performance Based Design) ............................................ 317

xiv
 LIST OF TABLES

TABLES

Table 2.1 Earthquake Design Load Levels in TEC-2018 ......................................... 9

Table 2.2 Structural Performance Levels and Physical Damage Definitions of
Overall Structure(“Seism. Eval. Retrofit Exist. Build.,” 2017) .............................. 10
Table 2.3 Physical Damage Explanation of Structural Members for Different
Damage Levels (“Seism. Eval. Retrofit Exist. Build.,” 2017) ................................ 11
Table 2.4 Building Types which must be designed with Performance Analysis
according to TEC-2018 ........................................................................................... 12
Table 2.5 Performance Target Levels and Analysis Types of Buildings, must be
designed according to PBD for different EQ levels according to TEC-2018 ......... 12
Table 2.6 Summary of Deformation and Forced Controlled Actions for Structural
Members ................................................................................................................. 14
Table 2.7 Reinforced Concrete Effective Stiffness Values for Service-Level
Models in ACI318-19 and TEC-2018..................................................................... 18
Table 2.8 Reinforced Concrete Effective Stiffness Values for MCER-Level Models
(DD1-DD2 Level) in ACI318-19 and TEC-2018 ................................................... 19
Table 2.9 Expected Material Strengths ................................................................... 20
Table 2.10 Strain and Rotation Limits for Different Performance Levels in TEC-
2018 for PBD .......................................................................................................... 26
Table 3.1 Summary Measurements About Case Study Building............................ 32
Table 3.2 Location Coordinates of Case Study Buildings ...................................... 35
Table 3.3 Reinforcement Classes and Features (TS 708) ....................................... 36
Table 3.4 Cover and Live Loads of Case Study Building ...................................... 37
Table 3.5 Live Load Reduction Factors .................................................................. 37
Table 3.6 Earthquake Ground Motion Level .......................................................... 38
Table 3.7 Response Spectrum Parameters .............................................................. 38
Table 3.8 Seismic Design Parameters ..................................................................... 39
Table 3.9 Rigidity Reduction Factors ..................................................................... 42

xv
Table 3.10 Mass Participation Ratio Results for Case Study Buildings ................. 44
Table 3.11 Natural Period Limit Factor (µ)............................................................. 47
Table 3.12 Parameters for Design Period Calculation ............................................ 47
Table 3.13 Base-shear Calculation Parameters and Results .................................... 49
Table 3.14 Minimum Vertical Reinforcements of Web of Shear walls .................. 55
Table 3.15 Lateral Reinforcement of Walls of Case Building in Ankara ............... 59
Table 3.16 Lateral Reinforcement of Walls of Case Building in Istanbul .............. 60
Table 3.17 The Coefficient of The Degree of Bond, Ω For Coupling Beams ........ 74
Table 3.18 Coupling Beam Design Results of Case Study Building in Ankara ..... 74
Table 3.19 Coupling Beam Design Results of Case Study Building in Istanbul .... 75
Table 5.1 Mechanical Properties of Reinforcement used at Wall Specimens ....... 116
Table 5.2 Wall-1 (X direction loading) Physical Observations through Cyclic Test
............................................................................................................................... 117
Table 5.3 Wall-2 (Y direction loading) Physical Observations through Cyclic Test
............................................................................................................................... 117
Table 5.4 RW2 Wall P3D Reference Model (1. Model) Material Parameters...... 125
Table 5.5 Sensitivity Level Rates (Change Rate in Area or Boundary Value of
Cyclic Response) ................................................................................................... 127
Table 5.6 P3D Models Differences from Reference RW2 Wall Model (Model-1)
............................................................................................................................... 128
Table 5.7 P3D Models Differences from Reference RW2 Wall Model (Model-1)
............................................................................................................................... 129
Table 5.8 P3D Models Differences from Reference RW2 Wall Model (Model-1)
............................................................................................................................... 130
Table 5.9 RW2 Wall P3D Model Material Parameters Sensitivity Summary Results
............................................................................................................................... 131
Table 5.10 RW2 Wall P3D Ideal Model Material Parameters (Model-31) .......... 132
Table 5.11 P3D Models Differences from Reference TW2 Wall Model (1. Model)
for Sensitivity ........................................................................................................ 140

xvi
Table 5.12 Sensitivity Evaluation in Terms of Different Fiber Modelling for RW2
P3D Models .......................................................................................................... 157
Table 5.13 Geometric and Material Properties of Test Sample Link Beams ....... 180
Table 5.14 CB33F and FB33 Link Beams Experimental Strength and Deformation
Limits (Naish et al., 2013a)................................................................................... 186
Table 5.15 Model Trials Explanations Different from “CB33F” Reference Link
Beam Shear Hinge Model ..................................................................................... 187
Table 5.16 “CB33F” Link Beam “Shear Hinge Model” Input Parameters of
Reference Model and Ideal Model ........................................................................ 189
Table 5.17 Model Trials Explanations Different from “CB33F” Reference Link
Beam Moment Hinge Model ................................................................................ 190
Table 5.18 “CB33F” Link Beam “Moment Hinge Model” Input Parameters of
Reference Model and Ideal Model ........................................................................ 192
Table 5.19 Model Trials Explanations Different from “FB33” Reference Link
Beam Moment Hinge Model ................................................................................ 194
Table 5.20 “FB33” Link Beam “Moment Hinge Model” Input Parameters of
Reference Model and Ideal Model ........................................................................ 196
Table 5.21 Diagonally Reinforced Coupling Beam Proposed Deformation Limits in
ACI41-06 Code (“Seism. Eval. Retrofit Exist. Build.,” 2017) ............................. 200
Table 5.22 Summary P3D Input Parameters of Diagonally and Conventionally
Reinforced Link Beam .......................................................................................... 201
Table 6.1 Selected Earthquake Records of Time History Analysis for Ankara ... 204
Table 6.2 Selected Earthquake Records of Time History Analysis for Istanbul .. 205
Table 6.3 Structural Members Actions and Behavior for Case Tall Building under
DD1 Level Earthquake Design ............................................................................. 207
Table 6.4 Design Expected Material Properties ................................................... 207
Table 6.5 Defined Strain Gage Limits .................................................................. 216
Table 6.6 Calibrated and Uncalibrated Concrete and Reinforcement Material
Parameter Assignments of Core-walls .................................................................. 219
Table 6.7 Beam Effective Rigidity Properties of Case Study Building in Ankara 221

xvii
Table 6.8 Beam Effective Rigidity Properties of Case Study Building in Istanbul
............................................................................................................................... 221
Table 6.9 Beam Rotation Limits ........................................................................... 224
Table 6.10 Inelastic Component, “Moment Hinge, Rotation Type”, Assigned
Values of Beams for Building in Ankara for Calibrated and Uncalibrated
Alternatives............................................................................................................ 224
Table 6.11 Inelastic Component, “Moment Hinge, Rotation Type”, Assigned
Values of Beams for Building in Istanbul for Calibrated and Uncalibrated
Alternatives............................................................................................................ 225
Table 6.12 Inelastic Component, “Shear Hinge, Displacement Type”, Assigned
Values of Beams for Building in Istanbul for Calibrated and Uncalibrated
Alternatives............................................................................................................ 225
Table 6.13 PBD Summary Result Sufficiency Situation of Case Study Building 271
Table 6.14 Summary of Performance Based Design of Case Study Building ...... 272
Table 6.15 Mean Maximum Results Comparison Ratios of Calibrated (A3) and
Uncalibrated Models (A1 &A2) for Case Study Building in Ankara & Istanbul . 273
Table 7.1 Linear Analysis Base-shear Values over Basement of CSB ................. 277
Table 7.2 Non-linear Analysis Base-shear Values over Basement of CSB .......... 278
Table 7.3 Linear and Nonlinear Base-shear Demand Ratio of CSB ..................... 278
Table 7.4 Nonlinear Shear Design Results of Walls after PBD for CSB .............. 278
Table 7.5 Proposed “a” Multiplication Factor for Concrete Fiber Strain.............. 281
Table 7.6 Cyclic Degradation Factors after Calibration Work and Recommendation
Values at PEER Center Report (Moehle et al., 2011) ........................................... 284
Table 7.7 A1, A2 ve A3 Model Cyclic Degradation Parameters .......................... 286
Table 7.8 Mean Maximum Results Comparison Ratios of Calibrated (A3) and
Uncalibrated Models (A1 &A2) for Case Study Building in Ankara & Istanbul . 287

xviii
 LIST OF FIGURES

FIGURES

Figure 2.1 Viscous Damping Ratio vs. Building Height (Pacific Earthquake
Engineering Center, 2017) ...................................................................................... 16
Figure 2.2 Nonlinear Modeling Types of Structural Members (PEER/ATC 72-1,
2010) ....................................................................................................................... 20
Figure 2.3 Lumped Plastic Hinge Model Subjects a) View of Lumped Plastic Hinge
b) Monotonic Moment-Rotation Response of Plastic Hinge c) Cyclic Response of
Plastic Hinge (PEER/ATC 72-1, 2010) .................................................................. 21
Figure 2.4 Distributed Fiber Model of Wall (PEER/ATC 72-1, 2010) .................. 22
Figure 2.5 Analytical Model Types of Nonlinear Model of Walls(NIST, 2014) ... 23
Figure 2.6 Material Models for Fibers a) Confined and unconfined concrete
material model b) Reinforcing steel material model ............................................... 24
Figure 3.1 Case Study Building Typical Plan View ............................................... 30
Figure 3.2 Basement Floor Plan.............................................................................. 30
Figure 3.3 Section View of Case Study Building ................................................... 31
Figure 3.4 Typical High-Rise Building Plan .......................................................... 33
Figure 3.5 Beams Linking Core Wall Sub-groups.................................................. 34
Figure 3.6 Case Building Locations: Ankara & Istanbul ........................................ 35
Figure 3.7 Horizontal Earthquake Ground Motion Spectrum ................................ 39
Figure 3.8 3D Model of Case Study Building at ETABS 18.1.1 ............................ 40
Figure 3.9 Natural periods of Case Study Buildings in Ankara and in Istanbul
calculated with modal analysis by ETABS software .............................................. 46
Figure 3.10 Response Modification Factor (Ra) vs. Period (T) Graph ................... 48
Figure 3.11 Model-A .............................................................................................. 51
Figure 3.12 Model-B ............................................................................................... 51
Figure 3.13 Model-C ............................................................................................... 52
Figure 3.14 Etabs Models Used for Shear wall and Link Beam Design ................ 52
Figure 3.15 Thickness and Labels of Walls of Case Building in Ankara ............... 55

xix
Figure 3.16 Thickness and Labels of Walls of Case Building in Istanbul .............. 56
Figure 3.17 Boundary, Web Placement and Boundary Reinforcement of Shear
walls of Building in Ankara .................................................................................... 57
Figure 3.18 Boundary, Web Placement and Boundary Reinforcement of Shear
walls of Building in Istanbul ................................................................................... 58
Figure 3.19 Axial Load Capacity Check Diagrams of W1 to W12 Walls (Ankara)
................................................................................................................................. 61
Figure 3.20 Axial Load Capacity Check Diagrams of W13 to W22 Walls (Ankara)
................................................................................................................................. 62
Figure 3.21 Shear Load Capacity Check Diagrams of W1 to W12 (Ankara) ......... 63
Figure 3.22 Shear Load Capacity Check Diagrams of W13 to W22 Walls (Ankara)
................................................................................................................................. 64
Figure 3.23 Axial Load Capacity Check Diagrams of W1 to W12 (Istanbul) ........ 65
Figure 3.24 Axial Load Capacity Check Diagrams of W13 to W22 (Istanbul) ...... 66
Figure 3.25 Shear Load Capacity Check Diagrams of W1 to W12 Walls (Istanbul)
................................................................................................................................. 67
Figure 3.26 Shear Load Capacity Check Diagrams of W13 to W22 (Istanbul) ...... 68
Figure 3.27 Shear Force Distribution Ratio at 5th Story .......................................... 69
Figure 3.28 Labels and Wall Groups for Flexural Design of Core-wall ................. 70
Figure 3.29 Flexural Design Results (Demand/Capacity Ratio vs. Height) of Wall
Groups of Building in Ankara and in Istanbul ........................................................ 70
Figure 3.30 Beam and Link Beam Dimensions of Case Study Building in Ankara 73
Figure 3.31 Beam and Link Beam Dimensions of Case Study Building in Istanbul
................................................................................................................................. 73
Figure 3.32 a) Beam (40/60) and b) Link Beam (40/150) Reinforcement Details fo
Case Study Building in Ankara ............................................................................... 75
Figure 3.33 a) Beam (40/60), b) Beam (60/60) and b) Link Beam (60/150)
Reinforcement Details of Case Study Building in Istanbul..................................... 76
Figure 3.34 Drift Ratio Results of Case Study Building in Ankara and in Istanbul 77
Figure 4.1 Perform 3D Component Properties Task ............................................... 81

xx
Figure 4.2 ”Inelastic Steel Material” Input Panel ................................................... 82
Figure 4.3 ”Inelastic 1D Concrete Material” Input Panel ....................................... 83
Figure 4.4 ”Elastic Shear Material for a Wall” Input Panel.................................... 83
Figure 4.5 Perform 3D “YULRX” Backbone Curve(Computers and Structures,
Inc., 2006) ............................................................................................................... 84
Figure 4.6 Perform 3D Hysteretic Loop Model(Computers and Structures, 2006) 85
Figure 4.7 Strength Loss Interaction Factor (SLIF=0 (a), 1 (b)) (Computers and
Structures, Inc., 2006) ............................................................................................. 86
Figure 4.8 Perform 3D Input Panel for Cyclic Degradation Factor and Unloading
Stiffness Factor ....................................................................................................... 87
Figure 4.9 a) Cyclic Degradation Factor b) Hysteric Loop with The Same Cyclic
Degradation Factor and Different Unloading Stiffness Factors (USF=+1 (A), -1
(B), 0 (C)) (Computers and Structures, 2006) ........................................................ 88
Figure 4.10 Fiber Model of Walls at Perform 3D (PEER/ATC 72-1, 2010) .......... 89
Figure 4.11 Wall Fiber Modeling Options at Perform 3D (Computers and
Structures, Inc., 2006) ............................................................................................. 89
Figure 4.12 Beam, Reinforced Cross Section Input Panel at P3D .......................... 90
Figure 4.13 Inelastic Components Input Panel of Link Beam at P3D .................... 91
Figure 4.14 Elastic Gage Component Input Panel at P3D ...................................... 92
Figure 4.15 Beam Models with Plastic Hinge and Curvature Hinge Components
(Computers and Structures, Inc., 2006) .................................................................. 93
Figure 4.16 Compound Components of Beam with Moment Hinges at P3D......... 93
Figure 4.17 General Approach for Compound Components of Link Beam
(Computers and Structures, Inc., 2006) .................................................................. 94
Figure 4.18 Compound Components Sample of Link Beam at P3D ...................... 94
Figure 4.19 Positive Sign Convention of Wall Element(Computers and Structures,
Inc., 2006) ............................................................................................................... 96
Figure 4.20 Deep Link Beam (1.0 < ln/h <2.0) Connected to Wall with Imbedded
Beam (Computers and Structures, Inc., 2006) ........................................................ 98

xxi
Figure 4.21 Link Beam (3.0 < ln/h <4.0) Connected to Wall with Imbedded Beam
(Computers and Structures, Inc., 2006) ................................................................... 98
Figure 4.22 Sign Convention of Beam at P3D (Computers and Structures, Inc.,
2006) ........................................................................................................................ 99
Figure 4.23 Chord Rotation of Beam with Stiff End Zone (Computers and
Structures, Inc., 2006) ........................................................................................... 100
Figure 4.24 Elastic Slab/Shell Component Sign Convention (Computers and
Structures, Inc., 2006) ........................................................................................... 100
Figure 4.25  and  Viscous Damper Arrangement in Model (Computers and
Structures, Inc., 2006) ........................................................................................... 103
Figure 4.26 Variation of M and K Damping Ratios with Periods (Computers and
Structures, Inc., 2006) ........................................................................................... 104
Figure 4.27 View of Dissipated Energy Diagram at P3D ..................................... 106
Figure 5.1 Rectangular and T-Shape, i.e. RW2 and TW2, Wall Test Samples
(Orakcal et al., 2004) ............................................................................................. 109
Figure 5.2 Specimen RW2 (Orakcal et al., 2004) ................................................. 109
Figure 5.3 Specimen TW2 (Orakcal et al., 2004).................................................. 110
Figure 5.4 Measured Stress-Strain Relation of Reinforcements of RW2 (Orakcal et
al., 2004) ................................................................................................................ 110
Figure 5.5 RW2 Test Setup (Thomsen and Wallace, 1995) .................................. 111
Figure 5.6 Different Instruments Placed on Specimens (Thomsen and Wallace,
1995) ...................................................................................................................... 112
Figure 5.7 RW2-Lateral Load vs. Top Flexural Displacement Cyclic Response Test
Result ..................................................................................................................... 112
Figure 5.8 TW2-Lateral Load vs. Top Flexural Displacement Cyclic Response Test
Result ..................................................................................................................... 113
Figure 5.9 U-Shape Wall Specimen Dimension and Reinforcement Detail (Ile &
Reynouard, 2005) .................................................................................................. 114
Figure 5.10 U-Shape Wall Specimen Section View (Ile & Reynouard, 2005) ..... 115

xxii
Figure 5.11 Loading Protocol (Average Displacement vs Data Point) for Wall-1 (X
Direction-Left) and Wall-2 (Y Direction-Right) (Ile & Reynouard, 2005) .......... 116
Figure 5.12 Cyclic Experimental Response of U-Shape Wall-1 (X Direction
Loading) ................................................................................................................ 118
Figure 5.13 Cyclic Experimental Response of U-Shape Wall-2 (Y Direction
Loading) ................................................................................................................ 118
Figure 5.14 Rectangular, T and U Shape Test Wall Model Views at P3D ........... 119
Figure 5.15 Views of Top Slaving Assignments of Walls......... 120
Figure 5.16 Axial Load Assignment Views of Test Walls ................................... 120
Figure 5.17 Views of Drift Reference Node and Direction Assignments of Walls
............................................................................................................................... 121
Figure 5.18 View of Fixed Size Wall Fiber Module for Boundary of RW2 Wall 122
Figure 5.19 View of Fibers for Boundary of RW2 Wall ...................................... 122
Figure 5.20 RW2 Wall Cyclic Response Comparison Results of Experiment and
P3D Reference Model (1. Model) ......................................................................... 124
Figure 5.21 Cyclic Response Comparisons b/w C1-C8 for RW2 P3D Models .. 134
Figure 5.22 Cyclic Response Comparisons b/w C9-C16 for RW2 P3D Models . 135
Figure 5.23 Cyclic Response Comparisons b/w C17-C24 for RW2 P3D Models 136
Figure 5.24 Cyclic Response Comparisons b/w C25-C32 for RW2 P3D Models 137
Figure 5.25 TW2 Wall Cyclic Response Comparison Results of Experiment and
P3D TW2 Reference Model (1. Model) ............................................................... 139
Figure 5.26 Cyclic Response Comparisons b/w C1-C8 for TW2 P3D Models .... 141
Figure 5.27 Cyclic Response Comparisons b/w C9-C12 for TW2 P3D Models .. 142
Figure 5.28 RW2, Most Ideal P3D Cyclic Response Result ................................ 143
Figure 5.29 RW2, Close to Ideal P3D Cyclic Response Results for Rapid Analysis
............................................................................................................................... 145
Figure 5.30 RW2 Lateral Displacement Profiles of Experiment and P3D Model
(Model-31) for Different Drift Ratio .................................................................... 146
Figure 5.31 RW2 Wall, Base Level Concrete Strain Profiles of Experiment and
P3D Model (Model-31) for Different Drift Ratios ............................................... 146

xxiii
Figure 5.32 RW2 Wall, Base Level Steel Strain Profiles of Experiment and P3D
Model (Model-31) for Different Drift Ratios ........................................................ 147
Figure 5.33 TW2, Ideal P3D Cyclic Response Result .......................................... 148
Figure 5.34 TW2, Close to Ideal P3D Cyclic Response Result for Rapid Analysis
............................................................................................................................... 149
Figure 5.35 TW2 Lateral Displacement Profiles of Experiment and P3D Model for
Different Drift Ratio .............................................................................................. 149
Figure 5.36 TW2 Wall, Base Level Concrete Strain Profiles of Experiment and
P3D Model (TModel-1) for Different Drift Ratios ............................................... 151
Figure 5.37 TW2 Wall, Base Level Steel Strain Profiles of Experiment and P3D
Model (TModel-1) for Different Drift Ratios ....................................................... 152
Figure 5.38 UX and UY, P3D Cyclic Response Result with Ideal Material
Parameters ............................................................................................................. 154
Figure 5.39 Different P3D Fiber Modeling Types for RW2 ................................. 156
Figure 5.40 Cyclic Response Comparisons b/w GC1-GC7 in Terms of Different
Fiber Modelling for RW2-P3D Models ................................................................ 158
Figure 5.41 P3D Two Fiber Modeling Options, “Fixed Size” & “Auto Size” for
Walls ...................................................................................................................... 159
Figure 5.42 “Fixed Size” and “Auto-Size” P3D Modeling Options of RW2 and
TW2 ....................................................................................................................... 160
Figure 5.43 “Fixed Size” and “Auto Size” Modeling Option Cyclic Response
Result of RW2 ....................................................................................................... 161
Figure 5.44 “Fixed Size” and “Auto Size” Modeling Option Cyclic Response
Result of TW2 ....................................................................................................... 161
Figure 5.45 Concrete and Steel Strain Profiles of RW2, Wall for “Fixed Size” and
“Auto Size” Option................................................................................................ 162
Figure 5.46 Vertical Mesh and Strain Gage Length Variations ............................ 164
Figure 5.47 Cyclic Responses of Experiment and Model with 8x4 Mesh ............ 165
Figure 5.48 Cyclic Responses of Experiment and Model with 4x4 Mesh ............ 166
Figure 5.49 Cyclic Responses of Experiment and Model with 2x4 Mesh ............ 166

xxiv
Figure 5.50 Concrete Strain Profiles of RW2 Wall for Different Mesh Variations
............................................................................................................................... 168
Figure 5.51 Steel Strain Profiles of RW2 Wall for Different Mesh Variations .... 169
Figure 5.52 RW2 Wall, Base Level Concrete Strain Profile (Positive Displacement)
for test strain gage length, SGL-1, 229 mm .......................................................... 170
Figure 5.53 RW2 Wall, Base Level Steel Strain Profile (Positive Displacement) for
test strain gage length, SGL-1, 229 mm ............................................................... 171
Figure 5.54 TW2 Wall, Base Level Concrete Strain Profile (Positive Displacement)
for test strain gage length, SGL-1, 229 mm .......................................................... 171
Figure 5.55 TW2 Wall, Base Level Steel Strain Profile (Positive Displacement) for
test strain gage length, SGL-1, 229 mm ............................................................... 172
Figure 5.56 TW2 Wall, Base Level Concrete Strain Profile (Negative
Displacement) for test strain gage length, SGL-1, 229 mm.................................. 172
Figure 5.57 TW2 Wall, Base Level Steel Strain Profile (Negative Displacement) for
test strain gage length, SGL-1, 229 mm ............................................................... 173
Figure 5.58 TW2 and RW2 Walls Concrete Strain Experiment-P3D Model
Comparison Coefficient “a” vs Drift Ratio Graph ................................................ 174
Figure 5.59 TW2 and RW2 Walls Steel Strain Experiment-P3D Model Comparison
Coefficient “b” vs Drift Ratio Graph .................................................................... 175
Figure 5.60 Link Beams of The Case Study Building .......................................... 178
Figure 5.61 CB33F-Diagonal Sample Link Beam and Reinforcement Details
(ln/h=3.33 ; 1 in.=25.4 mm) (Naish et al., 2013a) ................................................. 179
Figure 5.62 FB33-Conventional Sample Link Beam and Reinforcement Details
(ln/h=3.33; 1 in.=25.4 mm) (Naish et al., 2013a) .................................................. 179
Figure 5.63 Link Beam Laboratory Test Setup..................................................... 180
Figure 5.64 Loading protocol of link beam test, a) Load-controlled (kN) b)
Displacement-controlled ....................................................................................... 181
Figure 5.65 Link Beam, CB33F, Load Deformation Response ............................ 182
Figure 5.66 Link Beam, FB33, Load Deformation Response .............................. 182

xxv
Figure 5.67 Two different Link beam Modeling Approaches: a) Moment Hinge
Model b) Shear Hinge Model (Naish et al., 2013b) .............................................. 183
Figure 5.68 Perform 3D Hysteretic Loop Model(Computers and Structures, 2006)
............................................................................................................................... 184
Figure 5.69 “CB33F Link Beam” Force-Deformation Response of Experiment and
Reference P3D “Shear Hinge Model” (R)............................................................. 187
Figure 5.70 Cyclic Response Results Comparisons of P3D Model Trials for
Diagonal Link Beam, CB33F with “Shear Hinge, Displacement Type” Module . 188
Figure 5.71 Cyclic Response Results Comparisons of P3D Model Trials for
Diagonal Link Beam, CB33F with “Moment Hinge, Rotation Type” Module..... 191
Figure 5.72 “FB33 Link Beam” Force-Deformation Response of Experiment and
Reference P3D “Moment Hinge Model” (R) ........................................................ 193
Figure 5.73 Cyclic Response Results Comparisons of P3D Model Trials for
Conventional Link Beam, FB33 with “Moment Hinge, Rotation Type” Module 195
Figure 5.74 CB33F Link Beam Cyclic Response Calibration Work Result with
Shear Hinge Approach .......................................................................................... 197
Figure 5.75 CB33F Link Beam Cyclic Response Calibration Work Result with
Moment Hinge Approach ...................................................................................... 198
Figure 5.76 FB33 Link Beam Cyclic Response Calibration Work Result with
Moment Hinge Approach ...................................................................................... 198
Figure 5.77 Generalized Force-Deformation Relation in ACI41-06 Code (“Seism.
Eval. Retrofit Exist. Build.,” 2017) ....................................................................... 199
Figure 6.1 Design DD1 Level (Target Spectrum) and DD2 Level Response
Spectrum Curves for Ankara and Istanbul ............................................................ 204
Figure 6.2 Scaled Time history Data Set a) for Ankara & b) for Istanbul ............ 205
Figure 6.3 7 Scaled EQ Spectrums for a) Ankara & for b) Istanbul .................... 206
Figure 6.4 Nonlinear P3D Model of Case Study Tall Building ............................ 209
Figure 6.5 Periods of Case Study Building in Ankara & Istanbul at P3D ............ 210
Figure 6.6 First Three Basement Floors Model with Effective Beams of Slab .... 211
Figure 6.7 Ground Floor Plan Model with Shells of Slabs ................................... 211

xxvi
Figure 6.8 Typical Floor P3D Model Plan of Case Study Tall Building .............. 212
Figure 6.9 Rigid Diaphragms Assignment for Typical Floor of Case Building ... 213
Figure 6.10 Mass Assignment at Mass Center for Typical Floor of Case Building
............................................................................................................................... 213
Figure 6.11 Core-wall Leg Fibers Aspect Ratio View ......................................... 215
Figure 6.12 Wall Vertical Mesh View .................................................................. 215
Figure 6.13 Core-wall Strain Gage Positions........................................................ 216
Figure 6.14 Concrete Material Model of Wall fibers............................................ 217
Figure 6.15 Reinforcement Material Model of Wall fibers .................................. 218
Figure 6.16 Moment-Curvature Diagrams of Beams (Ankara) a) 40/60 Beams b)
40/150 Beams ........................................................................................................ 222
Figure 6.17 Moment-Curvature Diagrams of Beams (Istanbul) a) 40/60 Beams b)
60/60 Beams.......................................................................................................... 223
Figure 6.18 Story Shear Load Distribution for Columns & Walls Through X and Y
Direction (Ankara) ................................................................................................ 226
Figure 6.19 Story Shear Load Distribution for Columns & Walls Through X and Y
Direction (Istanbul) ............................................................................................... 227
Figure 6.20 X and Y Direction Drift Check (<0.03) for Three Model Alternatives
(A1,A2,A3) (Ankara) ............................................................................................ 228
Figure 6.21 X and Y Direction Drift Check (<0.03) for Three Model Alternatives
(A1, A2, A3) (Istanbul) ......................................................................................... 229
Figure 6.22 Calibrated (A3) and Uncalibrated Models (A1 & A2) Drift Ratio
Comparisons for X & Y Direction (Ankara) ........................................................ 229
Figure 6.23 Calibrated (A3) and Uncalibrated Models (A1 & A2) Drift Ratio
Comparisons for X & Y Direction (Istanbul) ....................................................... 230
Figure 6.24 Calibrated Model (A3) Mean Drift Check (<0.03-mean; <0.045-max)
with 7 Earthquakes for X and Y Direction (Ankara) ............................................ 230
Figure 6.25 Calibrated Model (A3) Mean Drift Check (<0.03-mean; <0.045-max)
with 7 Earthquakes for X and Y Direction (Istanbul) ........................................... 231

xxvii
Figure 6.26 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for All Walls (Ankara)........................................................... 232
Figure 6.27 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for All Walls (Istanbul) ......................................................... 233
Figure 6.28 Wall Labels and Walls Shear Load Pass Max. Shear Capacity (Red
Circle: in Ankara; Orange Circle: in Istanbul) ...................................................... 233
Figure 6.29 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for W1-W11 (Ankara) ........................................................... 234
Figure 6.30 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for W12-W22 (Ankara) ......................................................... 234
Figure 6.31 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for W1-W11 (Istanbul) .......................................................... 235
Figure 6.32 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for W12-W22 (Istanbul) ........................................................ 235
Figure 6.33 Strain Gage Labels of Core-Wall ....................................................... 236
Figure 6.34 Unconfined Parts Mean Concrete Strain D/C Ratio of Walls for Three
Model Alternatives (Ankara) ................................................................................. 237
Figure 6.35 Unconfined Parts Mean Concrete Strain D/C Ratio of Walls for Three
Model Alternatives (Istanbul)................................................................................ 237
Figure 6.36 Confined Parts (Rectangular Boundary) Mean Concrete Strain D/C
Ratio of Walls for Three Model Alternatives (Ankara) ........................................ 238
Figure 6.37 Confined Parts (Rectangular Boundary) Mean Concrete Strain D/C
Ratio of Walls for Three Model Alternatives (Istanbul) ....................................... 238
Figure 6.38 Confined Parts (Flange Boundary) Mean Concrete Strain D/C Ratio of
Walls for Three Model Alternatives (Ankara) ...................................................... 239
Figure 6.39 Confined Parts (Flange Boundary) Mean Concrete Strain D/C Ratio of
Walls for Three Model Alternatives (Istanbul) ..................................................... 239
Figure 6.40 Proposed (Multiplied by “a=4”) & Available Calibrated Unconfined
Parts Mean Concrete Strain D/C Ratio of Walls (Ankara).................................... 240

xxviii
Figure 6.41 Proposed (Multiplied by “a=6”) & Available Calibrated Unconfined
Parts Mean Concrete Strain D/C Ratio of Walls (Istanbul) .................................. 241
Figure 6.42 Proposed (Multiplied by “a=2”) & Available Calibrated Confined Parts
(Rectangular Boundary) Mean Concrete Strain D/C Ratio of Walls (Ankara) .... 241
Figure 6.43 Proposed (Multiplied by “a=2”) & Available Calibrated Confined Parts
(Rectangular Boundary) Mean Concrete Strain D/C Ratio of Walls (Istanbul) ... 242
Figure 6.44 Proposed (Multiplied by “a=4”) & Available Calibrated Confined Parts
(Flange Boundary) Mean Concrete Strain D/C Ratio of Walls (Ankara) ............. 242
Figure 6.45 Proposed (Multiplied by “a=6”) & Available Calibrated Confined Parts
(Flange Boundary) Mean Concrete Strain D/C Ratio of Walls (Istanbul) ............ 243
Figure 6.46 Strain Gages Exceed Concrete Strain Limits After Usage of “a”
Proposed Multiplication (Green Circle: In Istanbul; Blue Circle: In Ankara) ...... 244
Figure 6.47 Mean Reinforcement Strain D/C Ratio for Three Alternative Models in
Ankara ................................................................................................................... 244
Figure 6.48 Mean Reinforcement Strain D/C Ratio for Three Alternative Models in
Istanbul .................................................................................................................. 245
Figure 6.49 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))
and Uncalibrated (A1-(U)) Models (Ankara) ....................................................... 246
Figure 6.50 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))
and Uncalibrated (A1-(U)) Models (Istanbul) ...................................................... 246
Figure 6.51 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))
and Uncalibrated (A2-(U)) Models (Ankara) ....................................................... 247
Figure 6.52 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))
and Uncalibrated (A2-(U)) Models (Istanbul) ...................................................... 247
Figure 6.53 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A1-(U)) Models (Ankara) .............................................................. 248
Figure 6.54 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A1-(U)) Models (Istanbul) ............................................................. 248
Figure 6.55 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A2-(U)) Models (Ankara) .............................................................. 249

xxix
Figure 6.56 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A2-(U)) Models (Istanbul) ............................................................. 249
Figure 6.57 Label of Beams .................................................................................. 250
Figure 6.58 Beam Rotation Check (<0.041 (B40-60); <0,02 (B40-150)) for Three
Model Alternatives between B1 & B6 (Ankara) ................................................... 251
Figure 6.59 Beam Rotation Check (<0.041 (B40-60); <0,02 (B40-150)) for Three
Model Alternatives between B7 & B9 (Ankara) ................................................... 252
Figure 6.60 Beam Rotation Check (<0.046 (B40-60); <0.046 (B60-60); <0,02
(B40-150)) for Three Model Alternatives between B1 & B7 (Istanbul) ............... 253
Figure 6.61 Beam Rotation Check (<0.046 (B60-60)) for Three Model Alternatives
for B8 (Istanbul) .................................................................................................... 254
Figure 6.62 Mean Rotation D/C Ratio of Beams (Ankara) ................................... 254
Figure 6.63 Mean Rotation D/C Ratio of Beams (Istanbul) .................................. 255
Figure 6.64 Calibrated (A3) and Uncalibrated (A1) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A1(U) (Ankara) ......................................................... 255
Figure 6.65 Calibrated (A3) and Uncalibrated (A1) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A1(U) (Istanbul) ........................................................ 256
Figure 6.66 Calibrated (A3) and Uncalibrated (A2) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A2(U) (Ankara) ......................................................... 256
Figure 6.67 Calibrated (A3) and Uncalibrated (A2) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A2(U) (Istanbul) ........................................................ 257
Figure 6.68 Calibrated (A3) and Uncalibrated (A1) Beam (60/60) Rotation
Comparison Ratio, A3-(C)/A1(U) (Istanbul) ........................................................ 257
Figure 6.69 Calibrated (A3) and Uncalibrated (A2) Beam (60/60) Rotation
Comparison Ratio, A3-(C)/A2 (U) (Istanbul) ....................................................... 258
Figure 6.70 Shear Load Capacity Check (<3844 kN) of Diagonally Reinforced
Link Beam (Istanbul) ............................................................................................. 259
Figure 6.71 Shear Displacement Check (<39 mm (0.03 ln)) of Diagonally
Reinforced Link Beam (Istanbul) .......................................................................... 259
Figure 6.72 Beam Shear Load (D/C) Check between B1 & B3 (Ankara) ............ 260

xxx
Figure 6.73 Beam Shear Load (D/C) Check between B1 & B6 (Ankara) ............ 261
Figure 6.74 Beam Shear Load (D/C) Check between B1 & B7 (Istanbul)........... 262
Figure 6.75 Beam Shear Load (D/C) Check for B8 (Istanbul) ............................. 263
Figure 6.76 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (40/60) Shear
load D/C Ratio Comparison (Ankara) .................................................................. 263
Figure 6.77 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (40/150) Shear
load D/C Ratio Comparison (Ankara) .................................................................. 264
Figure 6.78 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (40/60) Shear
load D/C Ratio Comparison (Istanbul) ................................................................. 264
Figure 6.79 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (60/60) Shear
load D/C Ratio Comparison (Istanbul) ................................................................. 265
Figure 6.80 Energy Dissipation Comparisons of Calibrated (A3) & Uncalibrated
(A1-A2) Models (Ankara) .................................................................................... 266
Figure 6.81 Energy Dissipation Comparisons of Calibrated (A3) & Uncalibrated
(A1-A2) Models (Istanbul) ................................................................................... 266
Figure 6.82 Energy Dissipation Distribution of 14 Earthquakes for Calibrated
Model (A3) (Ankara) ............................................................................................ 267
Figure 6.83 Energy Dissipation Distribution of 14 Earthquakes for Calibrated
Model (A3) (Istanbul) ........................................................................................... 267
Figure 6.84 Mean Dissipated Inelastic Energy Distribution Between Beams and
Walls for Three Model Alternatives (Ankara) ...................................................... 268
Figure 6.85 Mean Dissipated Inelastic Energy Distribution Between Beams and
Walls for Three Model Alternatives (Istanbul) ..................................................... 269
Figure 6.86 Walls Having Insufficient Shear Capacity (Orange Rectangle: in
Istanbul; Red Rectangle: in Ankara) and Strain Fibers Passing the Limits After
Using Proposed Multiplication factor “a” (Orange Circle: in Istanbul; Red Circle:
in Ankara) ............................................................................................................. 271
Figure 7.1 TW2 (T-Shaped) and RW2 (Rectangular) Walls Concrete Strain
Experiment-P3D Model Comparison Coefficient “a” vs Drift Ratio Graph ........ 280
Figure 7.2 Check Strain Points of Selected Wall for Y- Direction Push .............. 282

xxxi
Figure 7.3 Strain Profile Observation for Selected Strain Points of Wall ............. 282
Figure 7.4 Cyclic Degradation Factor Alternatives for P3D Models .................... 285

xxxii
 LIST OF ABBREVIATIONS

ABBREVIATIONS

ACI American Concrete Institute

ATC Applied Technology Council
ASCE American Society of Civil Engineers
BRB Buckling Restrained Brace
C Comparison
CB Coupling Beam
CSI Computer and Structures, INC.
EQ Earthquake
EPP Elastic Perfectly Plastic
FB Frame Beam
LB Link Beam
MCE Maximum Considered Earthquake
MVLEM Multiple vertical line element model
NIST National Institute of Standards and Technology
P3D Perform 3D
PBD Performance Based Design
PEER Pacific Earthquake Engineering Research
PGA Peak ground acceleration
R/C Reinforced Concrete
RW Rectangular Wall
SGL Stain gage length
SLE Service Level Earthquake
SLIF Strength Loss Interaction Factor
TBI Tall Buildings Initiative
TEC Turkish Earthquake Code
TW T-Shaped Wall

xxxiii
USF Unloading Stiffness Factor
W Wall
WG Wall Group
YULRX Perform 3D Cyclic Energy Degradation Limits

xxxiv
 LIST OF SYMBOLS

SYMBOLS

a Proposed concrete strain multiplication factor

Ag Gross area
Ach Gross area of walls
Asd Diagonal reinforcement area
bw Width of beam
D Over-strength factor
EC Concrete elastic modulus
Ed(H) Horizontal design earthquake load
fce Expected concrete compressive strength
fck Characteristic concrete compressive strength
fctd Concrete design tensile strength
fye Expected reinforcement yield strength
fyk Characteristic reinforcement yield strength
fywd Lateral reinforcement yield strength
G dead and superimposed dead load
h Beam height
hk Height of beam
H Building height
H1 Perform 3d X direction coordinate
H2 Perform 3d Y direction coordinate
I Importance factor
Ig Gross moment of inertia
In Clear span of beam
Lp Plastic hinge length
Ls Shear length
mt Total mass of building over basement floor
My Yield moment capacity
Mp Plastic moment capacity
Mr Moment capacity
R Response modification factor

xxxv
Q Live load
V Perform 3d z direction coordinate
Ve Design shear load
Vcr Cracking shear capacity
Vr Total shear capacity
Vw Shear capacity only from reinforcement
VtE Total equivalent earthquake load
Vtx Total earthquake load according to modal analysis
c Concrete strain
cu Concrete ultimate strain
sy Reinforcement yield strain
s Reinforcement strain
su Reinforcement ultimate strain
θy Yield rotation
ζ Damping ratio
ωwe Confinement effect
 Unit weight
µ Natural period limit factor
βtE Magnification coefficients
βK Viscous damping ratio
βv Wall dynamic multiplication factor
ρsh Lateral reinforcement ratio
Ω Degree of coupling

xxxvi
 CHAPTER 1

1 INTRODUCTION

1.1 Motivation

The development of computer technology along with developments in structural

engineering allows complicated structures to be modeled in extensive detail. This
situation has led to new approaches in the design of buildings with an emphasis on
analysis under extreme loads such as earthquakes. Design of important structures is
now assessed with nonlinear analysis nowadays.

Nonlinear performance-based design approach was included in the Turkish

Earthquake Code in 2018. On the other hand, the traditional force-dependent
capacity design approach, which is widely used in the design of structures today, is
still actively continuing within Turkish Earthquake Code. As with every new
approach that is open to development and revisions, "nonlinear performance-based
earthquake design" is open to continuous development and improvement as a result
of developments in research and practice.

The subject of "nonlinear performance-based design" in the Turkish Earthquake

Regulation (2018) is an active area of research that requires continuous evolution of
structural engineers. Hence, this new design approach can be seen as a specialized
field that should be performed by structural engineers with serious and
comprehensive theoretical knowledge. Some of the structural engineers in Türkiye
are also keeping up with this new approach in terms of both knowledge and
adaptation to the software instruments used and developing themselves in this field.
The number of engineers specialized in this field is limited due to the limited number
of such projects in Türkiye. With the advent of new earthquake code, the number of
engineers who should adapt to this new design approach is increasing. But the

1
question is; "how uniform is the work done by engineers, who design such complex
buildings?" or "to what extent can structural engineers, who are the inspectors in this
job, verify the work done?".

Structural engineers model buildings using structural design and analysis computer
programs, which usually have some area of specialization. The most widely used
software in performance-based design for buildings is "CSI Perform 3D". Some
other software can be used for this design approach, but "CSI Perform 3D" is the
most popular program designed specifically for this job and preferred by the experts.
In a subject where the design has many complications and uncertainties for tall
buildings, constructing the right mathematical model, accurate analysis and
obtaining “true” appear to be impossible. The capabilities of software, parameter
selection, and its sensitivities on the results become key concern. As a result, a
competent structural engineer with sufficient theoretical knowledge in performance-
based structure design should pass through a complex and precise modeling stage
with an analysis software program and evaluate the results correctly. The inaccuracy
or inadequacy in any of these stages could raise a question mark in the results of this
laborious and competent design approach.

In Turkish Earthquake Code (2018), non-linear performance-based building design

is not required for all buildings. Such detailed design is conducted for tall buildings
and important structures having the height between 42 m and 70 m that should be
used just after earthquakes in severe earthquake zones. In the design of such
structures, the force-based traditional capacity design is conducted as the preliminary
design, while the performance analysis of the structure is then conducted to assess
the sufficiency of design. Hence, important structures are designed in detail in terms
of earthquake with a two-stage approach in the Turkish Earthquake Code (2018).

In the light of these explanations, the list of our motivations that led to this thesis
study are;

 In the latest earthquake code in Türkiye, this new complex design approach
is a matter of debate among both academic members and structural engineers.

2
  Practical contributions are needed for clauses of Turkish Earthquake
Regulation (2018) in terms of verification or revision in the field of nonlinear
performance-based design
 To demonstrate practical examples to structural engineers who will design
and work on nonlinear performance-based design
 To demonstrate this complex and laborious design approach with the widely
used "CSI Perform 3D" software program.

1.2 Research Questions

It has been discussed by many expert engineers and academicians whether nonlinear
structural design of tall buildings is necessary or not after the introduction of the
design of performance-based design come in New Turkish Earthquake Regulation
for buildings in low seismic zones. The necessity of this design should be questioned
because there are not many engineers with all the engineering knowledge such as
advanced concrete member behavior, structural dynamics, earthquake knowledge,
and structural analysis. Furthermore, the accuracy of such a complicated analysis is
not known compared to the actual system response. Engineers opposing nonlinear
structural design, prefer traditional forced based design approach due to the extensive
experience with it. This issue, the need for nonlinear analysis for tall buildings in
different seismic zones, is one of the research questions in this thesis.

It is mandatory to perform a nonlinear performance analysis for all high-rise

buildings in accordance with the Turkish earthquake code. In the content of this
thesis study, a case study high-rise building has been designed with the two-stage
approach according to TEC (2018), both in Ankara, which is not a severe earthquake
zone, and in Istanbul, which is a severe earthquake zone. The selected example high-
rise building is described in detail in Chapter 3. The case study high-rise building
has a flat slab system, and there is a core wall group and link beams to carry all
earthquake forces. The design of the case study high-rise building is accomplished
with traditional linear forced-based design method in stage one, including all

3
dimensions and reinforcement details. In the second stage, the adequacy of the core
wall group and link beams, which are designed as earthquake-resistant system, are
checked with performance-based nonlinear analysis. Design results comparison in
terms of size and reinforcement for the different earthquake zones in these two stages
is aimed to provide us an idea of whether nonlinear performance-based design is
required or not.

To compare the nonlinear performance analysis and linear forced-based design, it

must be ensured that the nonlinear performance analysis is performed correctly. First
stage of the performance analysis is that experimental lateral force-displacement
hysterical response results of structural elements, (in the selected case study high-
rise building, namely walls and link beams) that meet the earthquake force, should
be simulated with used software model. In other words, this model verification and
calibration should be performed in such a way that similar cyclic response results are
obtained between experimental hysterical results and model results. Such calibration
study is not mandatory in accordance with the regulations of the code. Hence, this
appears as one of the deficiencies in our current Turkish Earthquake Code (2018). In
this study, for case study high-rise building, prior to the building’s nonlinear
analysis, modeling has been conducted for walls and link beams modeled with
"Perform-3D" software to calibrate the modeling approach. With this study, which
we call calibration work, it has been determined that how 3D sample high-rise
structure performance analysis with "Perform-3D" program is realized. The effect of
modeling parameters on the ability of calibration to the test results is also one of the
research topics of this thesis.

The summary of the research questions of this thesis are;

 Do we need performance-based design for all tall buildings in different

earthquake regions?
 How much is the difference between results of linear and nonlinear design
and analysis of tall buildings for different levels of earthquake regions?

4
  How to correctly model wall and link beam in harmony with experimental
cyclic response results with Perform 3D software? Which Perfom-3D
software parameters are more effective on calibration work?
 What is the sensitivity of modeling parameters for performance analysis of
tall buildings? To what extent does material modeling perfectly consistent
with the results of the experiment, have an impact on the results?
 What is the importance of meshing and strain measurement gauge length
used in the evaluation of RC walls?

1.3 Outline of Thesis

This thesis consists of 7 main chapters. In each chapter, the following details are
given:

In Chapter 2, the concepts and design criteria related to the performance-based

design are presented. In this presentation, while design criteria and definitions are
mainly clarified according to the Turkish Earthquake Code (2018). Some
comparisons with American regulations are given and some recommendations are
proposed. Preliminary information is provided for the case study tall building’s
performance analysis in Chapter 6.

In order to find answers to the research questions of this thesis, linear elastic design
results of a high-rise case study building according to TEC-2018 is presented in
Chapter 3. Linear elastic design of case study high-rise buildings is fulfilled in two
different locations having different seismic characteristics, i.e., in Ankara and
Istanbul. In this chapter, first of all, all introduction and necessary information about
case study high-rise buildings are given. Core-wall group and link beams located in
the center of this high structure typical plan of a flat slab system are our two
important types of structural elements that meet all earthquake forces. In the
remaining sections of the thesis, all the studies will be focused on structural walls
and link beams. At the end of third chapter, dimensions related to core wall group

5
and link beams of the sample tall building under earthquake load and all
reinforcement detail information are presented for two different earthquake regions,
i.e., Ankara and Istanbul. This information will be input in, or in other words, a
preliminary design, in a nonlinear analysis of our case study tall building in Chapter
6.

Chapter 4 examines CSI Perform 3D software, which is considered to be the most

popular software for performance analysis and design of structures today.
Commonly used software and trusted by engineers working seriously on
performance analysis of structures in Türkiye is CSI Perform 3D, even if it is not a
very user-friendly software. So, performance-based design of case study tall
structure is realized with this software. Since this program is a very sophisticated and
not very user-friendly software, in order to use this software correctly, it is necessary
to give detailed information such as details of the software, important points to be
considered, working and modeling principles of this software. Perform 3D is
ultimately an instrument for structural analysis. It is important that this instrument is
used correctly. It is necessary to examine software in proper use. Therefore, Chapter
4 provides the necessary detailed information about software “CSI Perform 3D”.

Chapter 5 constitutes an important part of the thesis. In this section, a study is realized
on walls and link beams, which are two types of structural elements that meet the
earthquake load in case study tall structure. With this study, the results of the
experimental study related to nonlinear cyclic behavior of walls and link beams were
modeled with Perform 3D software and results were overlapped. This overlapping
study is carried out with experimental results of three types of walls, namely
rectangular, T and U-shaped walls, while in coupling beams it is made for
conventionally reinforced and diagonally reinforced link beams. In addition, for
classical frame beams between core wall groups, non-linear results of experiments
are matched with Perform 3D software results. With this calibration work, important
findings, such as how walls and link beams should be modeled in Perform 3D
software and which parameters should be used, are reached. It has been observed to
what extent it can be compatible with the results of the experiment. It has been

6
observed to what extent and which parameters and variables affect the nonlinear
behavior of walls and link beams. Within the scope of correct modeling data obtained
by this calibration work, an accurate base is created for performance analysis of 3D
sample tall structure.

In Chapter 6, the nonlinear design of case study tall building is fulfilled. Dimensions
of structural members and reinforcement details obtained according to linear elastic
design in Chapter 3 are used as input in this section. Taking into account the correct
modeling parameters obtained in Chapter 5 and without taking into consideration, a
total of 6 nonlinear analyzes of case study buildings are performed separately for
Ankara and Istanbul earthquake regions. It is observed how much it differed from
results of performance analysis in the section with results of linear analysis made in
Chapter 3. In addition, the effect of calibration work results on 3D structural
nonlinear design and total results are observed to what extent.

In Chapter 7, the answers sought to the research questions of the thesis and the
important conclusions drawn are indicated. Whether a nonlinear design approach is
required or not, the effect of calibration work with the results of PBD is detailed. The
calibration parameters of walls and link beams in correct modeling with Perfom-3D
software are summarized again. Proposed concrete strain multiplication factor “a” of
walls is also explained. New recommendations will be presented for some clauses of
TEC-2018.

7
 CHAPTER 2

2 SOME PRELIMINARIES FOR PERFORMANCE BASED DESIGN CHECK

ACCORDING TO TEC (2018)

A summary of the PBD check according to TEC (2018) guidelines is reviewed here
to clarify the steps followed in the later chapters of this thesis.

2.1 Buildings Types and Target Building Performance Levels

Structures are designed under four different levels of earthquake force. Definitions
of design earthquake force levels are shown in Table 2.1. Intensity of earthquake
force is named from DD-1 to DD4 earthquake level. Various levels of damage are
expected in structures under different force levels of the earthquake. These damage
levels and descriptions of overall damage to the structure are given in Table 2.2.
Descriptions of physical damage to each structural member according to structural
performance levels are also given in Table 2.3.

Table 2.1 Earthquake Design Load Levels in TEC-2018

EQ Design Level Probability of Exceedance Mean Return Period (years)

DD-1 2%/ 50 years 2475
DD-2 10%/ 50 years 475
DD-3 50%/ 50 years 72
DD-4 68%/ 50 years 43

9
 Table 2.2 Structural Performance Levels and Physical Damage Definitions of
Overall Structure(“Seism. Eval. Retrofit Exist. Build.,” 2017)

Structural Overall Damage

Overall Damage EQ Level
Performance Level of Structure
No permanent drift.
Structure substantially
retains original strength
and stiffness. Minor
cracking of facades,
O- partitions, and ceilings as
Very Light DD4
(Operational) well as structural elements.
All systems important to
normal operation are
functional. Continued
occupancy and use highly
likely
No permanent drift.
Structure substantially
IO-
Light retains original strength DD3
(Immediate Occupancy)
and stiffness. Continued
occupancy likely
Some residual strength and
stiffness left in all stories.
Gravity-load-bearing
elements function. No out
of plane failure of walls.
LS-
Moderate Some permanent drift. DD2
(Life Safety)
Damage to partitions.
Continued occupancy
might not be likely before
repair. Building might not
be economical to repair
Little residual stiffness and
strength to resist lateral
loads, but gravity load-
bearing columns and walls
CP-
Severe function. Large permanent DD1
(Collapse Prevention)
drifts. Some exits blocked.
Building is near collapse in
aftershocks and should not
continue to be occupied

10
 Table 2.3 Physical Damage Explanation of Structural Members for Different
Damage Levels (“Seism. Eval. Retrofit Exist. Build.,” 2017)

Structural IO (Light Damage LS (Moderate CP (Severe Damage

Members Level) Damage Level) Level)

Some boundary
Major flexural or
element cracking and
shear cracks and
spalling and limited
voids. Sliding at
buckling of
joints. Extensive
Minor diagonal reinforcement. Some
Walls crushing and
cracking of walls sliding at joints.
buckling of
Damage around
reinforcement.
openings. Some
Severe boundary
crushing and flexural
element damage
cracking
Extensive shear and
flexural cracks; some Shattered and
Coupling Experience
crushing, but concrete virtually
Beams diagonal cracking
generally remains in disintegrated
place
Major cracking and
Minor spalling in a
hinge formation in Extensive spalling
few places in
ductile elements. in columns and
ductile columns
Limited cracking beams. Limited
Columns & and beams.
or splice failure in column shortening.
Beams Flexural cracking
some nonductile Severe joint
in beams and
columns. Severe damage. Some
columns. Shear
damage in short reinforcing buckled
cracking in joints
columns
IO: Immediate Occupancy, LS: Life Safety, CP: Collapse Prevention

According to the Turkish Earthquake Code-2018, characteristics of new structures

to be built, for which performance analysis is mandatory, are presented in Table 2.4.
Accordingly, PBD is mandatory in design of all high-rise buildings in Turkey;
meanwhile, height limits to accept as high-rise building vary according to seismicity
of region. Building design with nonlinear analysis is also required in important
structures (i.e., importance factor is 1.5), that need to be used urgently after an
earthquake and are mid-height structures above 42 m in severe earthquake zones.

11
Structures to be analyzed for performance are also requested to be made according
to different earthquake levels and performance target levels. In performance-based
building design, design is necessarily started with linear elastic design and this
design is accepted as preliminary design. In Table 2.5, intended performance levels
under different levels of earthquake loads are indicated for buildings for which
performance analysis is mandatory.

Table 2.4 Building Types which must be designed with Performance Analysis
according to TEC-2018

Building Type SDS<0.33 0.33≤SDS<0.5 0.5≤SDS

70>HN>42 m & I=1.5 
HN>70 m 
HN>90 m 
HN>105 m 
SDS: Design Spectral Acceleration Coefficient for DD-2 stage EQ design level; HN: Building height
over podium or foundation; I: Importance Factor

Table 2.5 Performance Target Levels and Analysis Types of Buildings, must be
designed according to PBD for different EQ levels according to TEC-2018

Building Type DD4 DD3 DD2 DD1

70>HN>42 m & I=1.5
IO1 LS2 LS1
(0.50<SDS)
HN>70 m (0.50<SDS) IO1 LS2 LS1
HN>90 m
O2 LS2 CP1
(0.33<SDS <0.50)
HN>105 m
O2 LS2 CP1
(SDS <0.33)
1
Perfomance Based Design Target Level; 2Linear Elastic Capacity Design Target Level; O:
Operational, IO: Immediate Occupancy, LS: Life Safety, CP: Collapse Prevention

2.2 Structural Member Actions

Structural members should be grouped for PBD according to how they should behave
under EQ loads. They are divided into two main actions, i.e., deformation-controlled

12
actions and forced controlled actions. With deformation-controlled action, the
structural element will exhibit a ductile response with enough strength. Whereas a
forced controlled action refers to a brittle response dictated by member strength.

In tall building system behavior, it is expected that flexural deformation under

bending and axial load should occur along the critical wall length just above the
podium floors or at the base connected to the foundation. On the other hand, flexural
deformations are desired at the beam ends. Link beams mainly undergo shear
deformations due to their shorter shear span. It must provide enough strength in all
remaining regions. In Table 2.6, it is summarized how structural elements are
required to behave under EQ load types.

Although forced controlled action is divided into critical and non-critical in both
Turkish and American regulations, it is presented in Table 2.6 without separation to
include the whole. Difference between critical and non-critical action terms can be
defined that while a lack of strength that may occur in critical forced controlled
action can cause partial or total structural collapse, there is no situation that can cause
any or local collapse in non-critical forced action. In a high-rise building with a
podium, force effects on basement wall, foundation and ordinary slabs, except for
transfer floors can be thought of as a non-critical forced action class.

13
 Table 2.6 Summary of Deformation and Forced Controlled Actions for Structural
Members
DEFORMATION CONTROLLED ACTION- Inelastic
Behavior
STRUCTURAL
Under Moment Under Shear Load Under Axial Load
MEMBER
Shearwall 1   
Column 2   
Frame Beam 3   
Coupling Beam  4 
Slab 5   
Basement wall   
Foundation   
Transfer Girder   --
FORCED CONTROLLED ACTION- Elastic Behavior
STRUCTURAL
Under Moment Under Shear Load Under Axial Load
MEMBER
Shearwall   
Column   6
Frame Beam   
Coupling Beam  7 
Slab  8 9
Basement wall   
Foundation   --
Transfer Girder   --
1
P-M-M yielding of wall base (on top of foundation or basement podiums), 2 P-M-M yielding of
column base (on top of foundation or basement podiums),3 Flexural yielding of beam ends, 4 Shear
yielding of diagonally reinforced coupling beams, 5 Flexural yielding of slab-column connections, 6
Combined moment and axial load in gravity columns, 7 Shear of conventionally reinforced coupling
beams, 8 In-plane shear in transfer and other diaphragms,9 In-plane normal forces in diaphragms

2.3 Design Earthquake Load for PBD

In the design of structures, earthquake forces can be applied to a building model

using a number of different approaches such as equivalent lateral forces, response
spectrum-modal analysis, push-over analysis and time history analysis. Response
spectrum-modal analysis is widely used in the linear elastic design of 3D structures.

14
In the design, base shear in both directions resulting from modal analysis cannot be
lower than base shear found according to the equivalent earthquake load method.
Due to the long fundamental period of tall building, base shear is usually the
minimum base shear value found according to the equivalent earthquake load
method. In linear elastic analysis, structure design using the response spectrum
method according to DD2 earthquake level is preliminary design for PBD. In high-
rise buildings, time history analysis is mandatory according to Turkish Earthquake
Code in PBD used at DD1 or DD3 earthquake level. Push-over analysis is an
alternative as long as detail requirements are met in PBD, but push-over analysis is
not allowed in nonlinear analysis of high-rise buildings in TEC-2018.

11 horizontal ground motion records are required for the design check of structures
for time-history analysis method. The correct selection of horizontal earthquake
acceleration-time data sets containing both directions is important. In the selection
of horizontal earthquake data sets, ground motion should be selected by taking into
account similarities of the following headings:

 Similar spectral shape to target spectrum

 General tectonic regime
 Earthquake source mechanism properties, i.e., magnitude, fault mechanism,
fault distance, rupture surface distance (Rrup), rising time,
 Propagation path, i.e., distance b/w active fault to site, directivity,
 Site conditions, i.e., soil properties, Vs30,
 Effective duration of ground motion, D%5-D%95,
 PGV/PGA, arias intensity
 Scale factor closer to unity

Accordingly, 11 horizontal ground motion records are scaled according to the target
spectrum. With the scaling, mean of combined spectrums of 11 recordings should be
30% greater than target spectrum components between periods of 0.2Tp and 1.5Tp.
Effective duration of selected earthquake recordings should not be shorter than 5
times the first period of structure or 15 sec. Since damping ratio in high-rise buildings

15
is less than 5% and around 2.5%, damping ratio of 2.5% should be taken into account
in acquisition of scaling of earthquake records.

An appropriate damping ratio should be selected for structural models, after

obtaining scaled earthquake ground motion records for time history analysis. In
Turkish Earthquake Code-2018, it is recommended to choose a 2.5% damping ratio
in high-rise buildings analysis under DD4-DD3 and DD1 earthquake level forces.
For tall buildings, damping ratio approach is recommended in Figure 2.1. In this
figure, MCER and SLE represent DD1 and DD4 levels in Turkish Code, respectively.

Figure 2.1 Viscous Damping Ratio vs. Building Height (Pacific Earthquake
Engineering Center, 2017)

Equation-defining damping as a function of building height is shown in Equation

2.1. According to earthquake level analysis and building height, this damping ratio
can be selected by using the following equation (PEER, 2017).

0.2
𝜁critical = ≤ 0.05 (H in m) (2.1)
√𝐻

16
In nonlinear structural design, order of the load is important in analysis. Analysis is
fulfilled under gravity loads followed by earthquake loading with the assumed
gravity-loaded initial conditions. It should be mentioned that construction stage
approach should also be taken into account in the gravity loads in high-rise buildings.
In a nonlinear analysis, expected load and load combination that is closest to reality
is generated. Accordingly, expected live load is generally 30% of required live loads
in TS498. The load combination to be used generally in performance analysis is
shown in Equation 2.2.

(𝐻)
𝐺 + 0.3 𝑄 + 𝐸𝑑 (2.2)

Horizontal earthquake load appears in this load combination. The types of structures
from which vertical earthquake force is taken are much less. In structures with
vertical discontinuities and large spans, vertical earthquake forces can be taken into
account. In time history analysis, earthquake recording in both orthogonal directions
is defined simultaneously in modeling. After 11 earthquake records are given in
directions perpendicular to each other, analysis is repeated after load directions are
rotated to 90o. In other words, a total of 22 different analysis are taken into account
in PBD of tall buildings.

2.4 Structural Member Modeling

Nonlinear member behavior from test results and model results must be compatible
in order to accurately model nonlinear behavior of the structural elements. This
adaptation should ensure that correct modeling approach is obtained by harmonizing
the analysis and results of experiment such as expected strength, expected
deformation capacity, cyclic response, and strain profiles, energy consumption area,
stiffness and strength degradation of each structural element. In nonlinear behavior
of walls, it is also important whether uplift, rotation and effect of migration of neutral
axis issues are taken into account in the modeling.

17
Modeling aspects for stiffness, strength and nonlinear mechanical and material
aspects are presented next.

2.4.1 Effective Stiffness

The reduction of gross effective stiffness of structural elements compatible with

experiments are given in Table 2.7 and Table 2.8 for design of building under
different earthquake levels. In the models of building elements in linear or nonlinear
design under earthquake force, effective stiffness values will be defined according
to this table. In this way, reduction of lateral stiffness after cracking will be
considered for building seismic analysis.

Table 2.7 Reinforced Concrete Effective Stiffness Values for Service-Level

Models in ACI318-19 and TEC-2018

Effective Stiffness for Service-Level Models (DD3-DD4 Level)

Component Axial Flexural Shear
Structural walls
-- 0.25EcIg / (1.0EcIg )[1] --
(out of-plane)
Basement walls
1.0EcAg 1.0EcIg 0.4EcAg
(in-plane)
Basement walls
-- 0.25EcIg / (1.0EcIg ) [1] --
(out of-plane)
0.07 (l/h) EcIg ≤ 0.3EcIg
Coupling beams 1.0EcAg 0.4EcAg
/ (0.3EcIg ) [1]
Non-PT transfer
0.4 EcAg /
diaphragms 0.5EcAg 0.5EcIg
(0.32 EcAg) [1]
(in-plane only)
PT transfer
diaphragms 0.8EcAg / (none)1 0.8EcIg /(none) [1] 0.4EcAg/ (none) [1]
(in-plane only)
Beams 1.0EcAg 0.5EcIg /(0.7EcIg ) [1] 0.4EcAg
Columns 1.0EcAg 0.7EcIg [2] /(0.9EcIg ) [1] 0.4EcAg
[1] [2]
Effective stiffness in TEC-2018 different from ACI318-19; Stiffness in
columns dependent in ACI 318-19 on axial load level so that ≥0.5Agfc ; 0.7EcIg ,
≤0.1Agfc ; 0.3EcIg, 0.1Agfc≤ ,≤0.5Agfc; interpolation

18
 Table 2.8 Reinforced Concrete Effective Stiffness Values for MCER-Level
Models (DD1-DD2 Level) in ACI318-19 and TEC-2018

Effective Stiffness for MCER-Level Models (DD1-DD2 Level)

Component Axial Flexural Shear
Structural walls
-- 0.25EcIg --
(out of-plane)
Basement walls
1.0EcAg 0.8EcIg 0.2EcAg
(in-plane)
Basement walls
-- 0.25EcIg --
(out of-plane)
0.07 (l/h) EcIg ≤
Coupling beams 1.0EcAg 0.3EcIg 0.4EcAg
/ (0.15EcIg )[1]
Non-PT transfer
diaphragms 0.25EcAg 0.25EcIg 0.1EcAg
(in-plane only)
PT transfer
diaphragms 0.5EcAg/ (none)1 0.5 EcIg/ (none) [1] 0.2EcAg/ (none) [1]
(in-plane only)
Beams 1.0EcAg 0.3EcIg/ (0.35EcIg ) [1] 0.4EcAg
Columns 1.0EcAg 0.7 EcIg [2] 0.4EcAg
[1]
Effective stiffness in TEC-2018 different from ACI318-19; [2] Stiffness in columns dependent in
ACI 318-19 on axial load level so that for ≥0.5Agfc ; 0.7EcIg , for ≤0.1Agfc ; 0.3EcIg, for 0.1Agfc≤
,≤0.5Agfc; interpolation

If structural walls or R/C columns are modeled with distributed fiber model,
structural element deformation and cracking conditions are automatically defined
without the need to use effective stiffness reduction coefficients. So, there is no
effective stiffness reduction coefficients for the modeling of these members.

2.4.2 Expected Material Strengths

In nonlinear structure analysis, expected strength is used instead of concrete and

reinforcement characteristic strength values. The relationship between expected
material strength and characteristic strengths is given in Table 2.9. The expected
strength utilization is used in all nonlinear calculations as well as in the modeling
context.

19
 Table 2.9 Expected Material Strengths

MATERIAL EXPECTED STRENGTH

Concrete fce= 1.3 fck
Reinforcing Steel fye= 1.2 fyk

2.4.3 Nonlinear Structural Member Model Types

There are different modeling forms of building elements in nonlinear analysis of

structures. As the most basic distinction, nonlinear structural elements can be
modeled under two main headings as lumped plastic hinge or distributed fiber model.
Figure 2.2 shows representative views of modeling patterns suitable for a nonlinear
hinge or nonlinear fiber approach.

Figure 2.2 Nonlinear Modeling Types of Structural Members (PEER/ATC 72-1,

2010)

In Turkish Earthquake Code, use of lumped plastic hinge model is recommended for
columns, and beams, while distributed fiber model approach is the preferred choice
for wall elements.

20
 i. Lumped Plasticity Model (Inelastic Hinge Model)

Nonlinear hinge is defined at column or beam ends in lumped plastic models. Force-
deformation monotonic curves may be in the form of moment-rotation or force-
displacement in the plastic hinge definitions. In addition to defining the yield,
ultimate residual strengths and the corresponding rotations, cyclic response behavior
compatible with experimental results should be defined in plastic hinge sections.
Lumped plastic hinge model subjects are shown in Figure 2.3.

Figure 2.3 Lumped Plastic Hinge Model Subjects a) View of Lumped Plastic
Hinge b) Monotonic Moment-Rotation Response of Plastic Hinge c) Cyclic
Response of Plastic Hinge (PEER/ATC 72-1, 2010)

Plastic hinge positions of column and beam can be defined at the end of structural
elements or in the middle part of plastic hinge length. In TEC (2018), identification
at the exact end is allowed. The plastic hinge length (Lp) can be taken as half of the
cross-sectional length (h) in which it works (TBDY-2018, 2018)

In ASCE-SI 41, for structural elements modeled with frame elements such as
columns and beams, elastic section rigidity between plastic hinge parts allows
definition according to cracked section rigidity reduction coefficients specified in
Table 2.7 and Table 2.8. According to TEC (2018), effective stiffness should be
defined according to a value calculated according to Equation 2.3. The yield rotation
value in Equation 2.3 is determined by another empirical complex formula that
includes the entire flexural, shear and bond slip effect. Therefore, according to

21
ACI318, the definition of effective cross-sectional rigidities provides greater
convenience. American regulation also includes column effective rigidity reduction,
which varies according to the axial load level.

𝑀𝑦 𝐿𝑠
(𝐸𝐼)𝑒 = (2.3)
𝜃𝑦 3

ii. Distributed Plasticity

Distributed fiber model approach is obtained by both concrete and reinforcement

steel filaments in the members according to their exact locations, amount and feature
of fibers (Figure 2.4). While material model parameters are defined according to
whether concrete is confined or unconfined, both tensile and under compression
material parameter definitions are made in reinforcements. The model of behavior
of concrete under tensile force can also be optionally included. The advantages of
fiber models are direct inclusion of axial-bending interactions, combining section
and finite element analysis, and better accuracy due to better section description. The
key disadvantage is usually the cost of computations.

Figure 2.4 Distributed Fiber Model of Wall (PEER/ATC 72-1, 2010)

Wall structural members exhibit bending behavior in sections close to the foundation
or just above podium section surrounded by basement walls. Nonlinear behavior is
expected in this region, defined as the critical wall height, whereas elastic cracked

22
behavior occurs in the upper sections. For this reason, frequency of meshing in the
section that continues along the critical wall height is important for the accuracy of
results. Examples of fiber models along critical wall height are schematically shown
in Figure 2.4.

The analytical modeling approaches for distributed fiber modeling approach. One of
these analytical approaches is the phenomenological macroscopic or meso-scale
model approach named as, multiple vertical line element model (MVLEM) proposed
by Vulcano et al. (1988) (Figure 2.5a). The four-node wall panel element solution
method, which is also used in the commonly used CSI Perform 3D software solution
approach, is another method (Figure2.5b). Finally, according to the 2D microscopic
finite element method analytical solution method, nonlinear behavior analysis of
walls can be performed (Figure 2.5c).

Figure 2.5 Analytical Model Types of Nonlinear Model of Walls(NIST, 2014)

Regardless of the method by which solution is made, the fact that the physical
behaviors and subjects such as neutral axis migration, concrete tension stiffening,
progressive crack closure, nonlinear shear behavior, effect of fluctuating axial force
should be reflected in the nonlinear analysis of wall members in harmony with the
results of experiment in modeling shows the success of analytical solution. In content

23
of this thesis, our solutions have been made according to "four-node wall panel"
approach widely used by Perform 3D software.

There is no need to make effective stiffness adjustments within the in-plane behavior
of building walls modeled with distributed fibers. Effective stiffness is automatically
defined by assigned concrete and reinforcement steel fibers and material modeling
is defined for these fibers. Reduction coefficients in Table 2.7 and Table 2.8 can be
used for out-of-plane elastic behavior of walls. In addition, as if shear deformations
caused by shear force in wall behavior are not more prominent in deformations
caused by bending, shear behavior can be modeled as linear elastic.

Material models and material parameters used are an important issue that affects the
nonlinear behavior of the wall model. In the Turkish regulation, a concrete material
model proposed by Mander [Link] is proposed (Figure 2.6a). Also, a recommended
model for reinforcement is shown in Figure 2.6b. When different lack of issues such
as behavior of reinforcement model under tension, compression and behavior
embedded in concrete are considered, it is thought that it can be further elaborated
in the regulation in terms of material model of reinforcement.

Figure 2.6 Material Models for Fibers a) Confined and unconfined concrete
material model b) Reinforcing steel material model

Lastly, other issues to consider for nonlinear analysis of structures are P-delta effect
and accidental eccentricity. Accidental eccentricity should be taken into account in
the event of torsional irregularity.

24
2.5 Evaluating the Results of Performance Analysis

i. Drift Ratio

Maximum allowed relative story drift ratio for DD1 earthquake level and all
earthquake records (2x11=22 earthquake records) is 0.045, according to Turkish
Earthquake Code (2018) for tall buildings. In addition, the mean of absolute
values of peak story drift ratios for 22 earthquake records cannot exceed 0.03.
No story drift ratio is specified for non-linear structure design except for high-
rise structures under DD1-level earthquake force. This deficiency in regulation
is an important issue that needs to be eliminated.

For the service stage earthquake level (DD4), no story drift ratio limit is also
specified in the Turkish Earthquake Regulation. 0.5% may be recommended for
this (PEER (2017) recommendation).

ii. Strain & Rotation Limits

In nonlinear analysis of structures, strain check is performed under bending

behavior when wall structural elements are modeled with distributed fiber
approach for performance assessment. Design is satisfactory if both concrete and
reinforcement strain checks remain within desired limits. On the other hand,
under the bending behavior of structural elements such as columns and beams,
which are defined as frame elements, it is usually customary to check if plastic
rotation amounts remain within limits. Table 2.10 shows concrete and
reinforcement strain limits and plastic rotation limits for different performance
targets. In Equations 2.4, 2.5, and 2.6, concrete strain, reinforcement strain and
plastic rotation limits are defined for the collapse prevention performance target,
respectively. The concrete strain definition, cCP , in Equation 2.4 can be applied
for rectangular form and confined section. 𝜔𝑤𝑒 specifies contribution of
confined effect to concrete strain. The reinforcement strain, sCP, definition in
Equation 2.5 for collapse prevention limit is specified as 40% of ultimate
reinforcement strain. So, the reinforcement strain collapse prevention limit is

25
 0.032 for common reinforcement usage, grade B420C, with a 𝜖𝑠𝑢 value of 0.08.
This value is 0.05 according to American regulations and study reports (Pacific
Earthquake Engineering Center, 2017). In Turkish regulation, the value of 0.032
remains less as a more conservative value.

Table 2.10 Strain and Rotation Limits for Different Performance Levels in TEC-
2018 for PBD

c s θp
Immediate Occupancy (IO) 0.0025 0.0075 0
Life Safety (LS) 0.75 cCP 0.75sCP 0.75 θpCP
Collapse Prevention (CP) cCP sCP θpCP

cCP= 0.0035 + 0.04√𝜔𝑤𝑒 ≤ 0.018 (2.4)

sCP= 0.4 𝜖𝑠𝑢 (2.5)

2 0.5𝐿𝑝
θpCP= [(𝜙𝑢 − 𝜙𝑦 )𝐿𝑝 (1 − ) + 4.5𝜙𝑢 𝑑𝑏 ] (2.6)
3 𝐿𝑠

The plastic rotation limits in the columns and beams are calculated as in Equation
2.6. According to this equation, plastic rotation depends on in addition to ultimate
and yield curvature, shear span, plastic hinge length, and longitudinal reinforcement
diameter. According to this general definition, plastic rotation limits should be
determined at all column and beam nodes according to cross-sectional characteristics
and reinforcement configuration. According to ASCE 41-17, instead of such a
rotation limit recipe, there is an indication with charts based on equations and
numerical limits.

Plastic rotational capacities vary greatly between frame beams and coupling beams.
Therefore, rotational capacity defined by Equation 2.6 cannot be valid for link
beams. In Turkish Earthquake Regulation, it seems essential to define separate
plastic rotation capacity limits for coupling beams.

26
Strain and plastic rotation evaluations will be made according to the mean of the
absolute maximums of the 22 earthquake analysis results. Evaluations for the
response being ductile or brittle are completed according to the average values.

iii. Shear Stress Check

Strain and rotation limits are specified before for deformation-controlled inelastic
behavior. For forced controlled elastic behavior, shear force capacity checks of wall
and link beams that are active in earthquakes, must be made according to Equations
2.7 and Equations 2.8 according to TEC-2018.

𝑉𝑒 ≤ 0.85𝐴ch √𝑓ck (walls without openings) (2.7)

𝑉𝑒 ≤ 0.65𝐴ch √𝑓ck (walls with openings) (2.8)

In the performance analysis except for tall structures, mean of maximum absolute
shear force values of 22 earthquake analyzes are used as demand. In the case of tall
structures, shear force demand is determined by the value, 𝑉𝑒,𝑚𝑒𝑎𝑛+𝑠𝑡𝑎.𝑑𝑒𝑣. , in
Equation 2.9.

1.2 𝑉𝑒,𝑚𝑒𝑎𝑛 ≤ 𝑉𝑒,𝑚𝑒𝑎𝑛+𝑠𝑡𝑎.𝑑𝑒𝑣. ≤ 1.5 𝑉𝑒,𝑚𝑒𝑎𝑛 (2.9)

27
28
 CHAPTER 3

3 LINEAR ANALYSIS AND DESIGN OF THE CASE STUDY BUILDING

The structural design of a high-rise building under earthquake loads consists of two
main stages, i.e., linear preliminary elastic design and nonlinear performance-based
design according to TEC (2018). In the first stage, the structure is designed and
detailed under earthquake loads by using linear elastic analysis. This preliminary
stage is an input for the second stage, i.e., nonlinear performance-based design. Thus,
all detailed calculations are completed in these two-stage calculations, and so the
first preliminary design stage is verified and necessary additions for reinforcement
are completed. The first stage design is a minimum and cannot be reduced in the
second stage. In this section, we explain the preliminary design results of our case
study tall building according to Turkish Earthquake Regulation-2018. After
introducing our case study structure with flat slab system, the design results of shear
walls and coupling beams that effectively meet earthquake loads, and the results of
serviceability limits, i.e., drift ratios, are shown.

3.1 CASE STUDY BUILDING

3.1.1 Building Description and Dimensions

Selected case study tall building has a proper rectangular plan and does not have any
horizontal or vertical irregularity. Typical floor plan dimensions of upper and
basement floors are 51x42 m (Figure 3.1) and 77.6 m x 68.4 m (Figure 3.2)
respectively. Height of the building is 112.4 m from top of the basement and 129.4
m from the foundation level. There are totally 32 floors which include 4 basement
floors and 28 upper typical floors (Figure 3.3). Typical floor heights at basement and

29
upper part of the building are 3.4 m and 3.8 m respectively. Key dimensions
describing the structure are given in Table 3.1.

Figure 3.1 Case Study Building Typical Plan View

Figure 3.2 Basement Floor Plan

30
Figure 3.3 Section View of Case Study Building

31
 Table 3.1 Summary Measurements about Case Study Building

Subject Measurements
Floor plan dimensions for upper floors 51 m x 42 m
Floor plan dimensions for basement floors 77.6 m x 68.4 m
Height of the building above the foundation 129.4 m
Height of the building above the basement 112.4 m
Core wall plan dimensions 18 x 25 m
Max clear span at upper stories 9.4 m
Typical story height for the upper part of building 3.8 m
Typical story height at basement of building 3.4 m

3.1.2 Structural System of the Case Study Building

Case study building has a flat slab system and coupled with a core shear wall.
Maximum clear span of the upper typical stories is 9.4 m. Slab thicknesses are 30,
35, and 40 cm at the basement floors respectively and 30 cm at the upper portions.
The flat plate system is not considered as a lateral force resisting system, and it only
acts as a vertical load bearing part. For this reason, the total earthquake load is
assumed to be carried by reinforced concrete shear walls.

There is core wall group in the middle of the upper typical plan to resist earthquake
load effectively (Figure 3.4). Core wall plan dimensions are 18x25 m (Figure 3.1).
The ratio of the core wall length to the height of the building is a typical indicator
for their behavior in high-rise buildings. These ratios are almost 1/ (4.5) for the x-
direction and 1/ (6.2) for the y direction. Thickness of the walls inside the core varies
from 30 cm to 90 cm according to the two selected seismic spectral acceleration
spectrum, as will be described later in the thesis.

32
 Figure 3.4 Typical High-Rise Building Plan

One additional axis of columns is placed around the core wall (Figure 3.4). Columns
are conducted by using gravity loads. Columns dimensions for the outside zone of
superstructure projection are 80x80 cm at basement level. The dimensions of the
columns near the core wall range from 130x130 cm to 100x100 cm as they rise from
the bottom up.

Link beams are important and effective structural members to meet the seismic loads
by linking the core wall sub-groups. There are two types of link beams, LB1, LB2 in
terms of height and height dimensions are 60 cm and 150 cm (Figure 3.5). Width of
beams changes with wall width though height of building. Link beams can be
detailed by conventional reinforcement or diagonal reinforcement.

33
 Figure 3.5 Beams Linking Core Wall Sub-groups

In the general architectural formations of high-rise buildings, there are basements

and podium floors for usage such as car parking, mechanical area, shopping mall,
etc. The most important features of these floors are that they are very rigid compared
to the upper floors and are less ductile. In our case study structure, 4 basement floors
constitute significant changes in building behavior and ductility. For buildings
having very rigid basement floors, the design of building under the earthquake load
is specifically described in the regulations. The fact that the basement walls in the
case building are located on 3 sides constitutes a situation in which the structure will
not behave fully symmetrically in the basement section (Figure 3.2). The fact that
the basement walls are 3-sided and asymmetrical also causes asymmetrical behavior
on the core walls in the middle of the plan.

34
3.2 DESIGN CRITERIA

3.2.1 Seismicity

The case study building is assumed to be located in regions with different seismicity.
Two metropole cities namely Ankara and Istanbul, are selected for this purpose
(Figure 3.6). All analyses and designs are made separately for these two locations.
The coordinates of the locations, latitude and longitude, are shown in Table 3.2.

Table 3.2 Location Coordinates of Case Study Buildings

Location Latitude Longitude

Ankara 39,946536 32,874877
Istanbul 40,978182 29,100889

Figure 3.6 Case Building Locations: Ankara & Istanbul

3.2.2 Materials

[Link] Concrete

The following values are concrete grades for minimum concrete compressive
strength values (MPa) of 28 days cylindrical test specimens. According to TEC-2018
[Link], concrete cannot be selected below C25 concrete class. The selected concrete
classes acceptable according to TS EN 206 are given below;

35
  Foundation: C50
 Slab: C50
 Column and walls: C50

[Link] Reinforcement Class and Diameters

The selected reinforcement class is suitable according to TS708 and TEC-2018,

[Link].b. Mechanical properties of all reinforcement classes are shown at Table 2.3.

Table 3.3 Reinforcement Classes and Features (TS 708)

Class B 420C B 500B B 500C B500A

Yield Strength (min) Re
420 500 500 500
(N/mm2)

Ultimate Strength (min) Rm

- - - 550
(N/mm2)

Ultimate Strength/Yield Strength ≥1,15 1,08 ≥1,15

-
Ratio Rm /Re <1,35 (min.) <1,35

Experimental Yield
Strength/Characteristic Yield 1,30 - 1,30 -
Strength ratio Re act/Re nom (max)
Ultimate Strain (min) A5 (%) 12 12 12 5
Total strain under max loading
7,5 5 7,5 2,5
(min) Agt (%)

 Selected reinforcement class: B420C

 Used reinforcement diameters: 10, 12,14, 16, 18, 20, 22, 25, 26, 30

36
3.2.3 Design Loads

[Link] Gravity Loads

Self-weight is automatically calculated in the model by assigning γ=2.5 t/m3 for

reinforced concrete.

Cover and live loads are given in Table 3.4. These loads are selected based on TS498,
i.e., Turkish standard of design load for buildings. Live load reduction is applied
according to the coefficients of office usage in Table 3.5.

Table 3.4 Cover and Live Loads of Case Study Building

Slab
Cover Live
Level Thickness Usage
(t/m2) (t/m2)
(cm)
(+6,80/+112,
1-28 Floors 30 0,25 0,35 Office
40)
Ground Floor (G) (+0,00) 40 0,60 0,50 Common
[Link] Floor
(-6,80) 35 0,50 0,50 Shopping
(B1)
[Link] Floor Car
(-10,20) 30 0,25 0,50
(B2) Parking
[Link] Floor Car
(-13,60) 30 0,25 0,50
(B3) Parking

Table 3.5 Live Load Reduction Factors

a) Houses, offices, etc.

# of Floors 1 2 3 4 5 6 7 8 9 10 11 12
1 Reduction (%) 0 0 0 20 40 60 80 80 90 40 40 40
Reduction
2 1 1 1 0,95 0,88 0,8 0,71 0,65 0,6 0,6 0,6 0,6
Coeff. (β)
b) Factory, etc.
3 Reduction (%) 0 0 0 10 20 30 40 40 40 20 20 20
Reduction
4 1 1 1 0,98 0,94 0,9 0,86 0,83 0,8 0,8 0,8 0,8
Coeff. (β)

37
[Link] Earthquake Load

Earthquake ground motion levels are defined as DD1, DD2, DD3 and DD4 in the
TEC-2018. Table 3.6 specifies the definition of earthquake ground motion levels.
Linear elastic design, which is the 1st stage for tall buildings under earthquake load,
is accomplished by using the DD-2 ground motion level.

Table 3.6 Earthquake Ground Motion Level

DD-1 Peak ground acceleration (PGA) with an exceedance probability of %2 in

50 years and 2475 years return period
DD-2 Peak ground acceleration (PGA) with an exceedance probability of %10
in 50 years and 475 years return period
DD-3 Peak ground acceleration (PGA) with an exceedance probability of %50
in 50 years and 72 years return period
DD-4 Peak ground acceleration (PGA) with an exceedance probability of %68
in 50 years and 43 years return period

Case study building is designed with linear elastic analysis according to the response
spectrum curve data in Istanbul and Ankara for DD2 earthquake ground motion
level. Important parameters define these response spectrum curves are given in Table
3.7.

Table 3.7 Response Spectrum Parameters

S1 (g) SD1 (g) SS (g) SDS (g) PGA (g) TA (s) TB (s)
Ankara 0.123 0.178 0.317 0.454 0.152 0.081 0.405
Istanbul 0.259 0.389 0.944 1.133 0.389 0.069 0.343

38
 Horizontal Earthqıake Ground Motion Design Spectrum
Period, Tn (s) vs. Spectral Acceleration, Sae (g, m/s2)
1,4
1,2
1,0
Sae (g, m/s2)

0,8
0,6
0,4
0,2
0,0
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0
Tn, Period (s)

Ankara Istanbul

Figure 3.7 Horizontal Earthquake Ground Motion Spectrum

All the parameters for the design of the case building under earthquake load are
selected from proper places according to TEC-2018. Seismic force resisting system,
response modification factor (R) and overstrength factor (D) are selected from Table
4.1 at TEC-2018. According to Table 4.1, the seismic force resisting system is “A13.
All the earthquake loads are resisted by special reinforced shear walls”. Response
modification factor (R), over-strength factor (D) and all other important selected
design parameters are summarized in Table 3.8.

Table 3.8 Seismic Design Parameters

Ankara Istanbul
Site Class ZC ZC
Importance Factor (I) 1 1
Building Usage Class (BKS) 1 1
Earthquake Design Class (DTS) 3 1
Building Height Class (BYS) 1 1
Response Modification Coefficient (Rtop) 6 6
System Over-strength Factor (Dtop) 2.5 2.5
Response Modification Coefficient (Rbottom) at Basement 2.5 2.5
System Over-strength Factor (Dbottom) at Basement 1.5 1.5
Damping Ratio () 0.05 0.05

39
3.2.4 Load Combinations

According to TEC-2018, used load combinations are given below.

(𝐻) (𝑍)
𝐺 + 𝑄 + 0.2𝑆 + 𝐸𝑑 + 0.3𝐸𝑑 (3.1)

(𝐻) (𝑍)
0.9𝐺 + 𝐻 + 𝐸𝑑 − 0.3𝐸𝑑 (3.2)

3.3 MODELING of CASE STUDY BUILDING

Case study building is modelled with a three-dimensional (3D) model by using

ETABS 18.1.1 (Figure 3.8).

Figure 3.8 3D Model of Case Study Building at ETABS 18.1.1

The modeling approach considered herein is fully compliant with TEC-2018, and
with the following assumptions;

40
 Finite element method (FEM) is used in the analysis of the model.

 While columns and beams are modelled as frame elements, slab and shear

walls are defined by using shell elements.

 Shell and frame elements have all 6 degrees of freedom (ux, uy, uz, rx,ry,rz) at

the node.

 In-plane forces and deformations are taken into account in the modeling of

slab in order to accurately consider back-stay effects. (+/-) 0.05 % of the

central mass to central rigidity difference as the accidental eccentricity is

assigned automatically.

 Mass modelling is succeeded by ETABS automatically by assigning the load

participation coefficient. Mass participation ratio for live load is taken as 0.3.

 Second order geometric nonlinear analysis, i.e., P-Delta effect, is taken into

account in models by using moment magnification factor.

 Analysis type is the linear elastic analysis. The structural members’ rigidities

are reduced according to coefficients in Table 3.9 (TBDY-2018, 2018) when

structure is designed only under load combinations, includes earthquake

loads. For statical combinations excluding earthquake loads, the model

should not contain rigidity reduction factors for structural elements.

41
 Table 3.9 Rigidity Reduction Factors

Reinforce Concrete Structural Elements Rigidity Reduction Factors

Wall-Slab (In-Plane) Axial Shear

Wall 0.50 0.50
Basement wall 0.80 0.50
Slab 0.25 0.25
Wall – Slab (Out of Plane) Flexural Shear
Wall 0.25 1.00
Basement wall 0.50 1.00
Slab 0.25 1.00
Frame Element Flexural Shear
Link Beam 0.15 1.00
Frame Beam 0.35 1.00
Columns 0.70 1.00
Wall (equivalent frame) 0.50 0.50

3.4 DESIGN FLOW CHART

1-Determine Design Material Strengths

 Concrete Strengths
 Concrete Elastic Modulus
 Reinforcement Class

2-Determine Design Loadings

 Gravity Loads: Self-weight, live load, cover, partition walls, fillings, soil
loads for recreation areas, parapets, facades
 Lateral Loads: Wind loads, earthquake loads, soil loads,

3-Determine Flat Slab Thickness under Design Loadings

 Design for deflection limits

42
  Design for enough slab strength
 Design for punching

4-Generating Building Models-1. MODEL (Design for DD2-EQ)

 Rigidity reduction factors of DD2 design EQ level is used for structural

elements
 Determine EQ Base Shear: Period is very long for tall buildings so
earthquake load is determined according to minimum code base shear for
tall buildings.
 Design columns under load combinations includes EQ load
 Column axial stress should not exceed 0.35 fck under EO load combinations
to determine preliminary column dimensions
 Check drift limit (use R=1) and other irregularities
 Check slabs for punching design under load combinations includes EQ load

5-Generating Building Models-2. MODEL (Design for DD2-EQ)

 Rigidity reduction factors of DD2 design EQ level remains the same with 1.
MODEL
 For shear wall design, columns ends are assigned with moment releases
where columns show more flexural behavior like just above podium or
foundation.
 Column end moment release assignment causes period change and so if
model base-shear scale factor is determined according to model period,
rearrange the model base-shear scale factor to get minimum base shear. In
other words, be sure that minimum code base shear should be valid in the 2.
Model without any change.
 Shear wall axial stress should not exceed 0.4 fck under EO load combinations
to determine preliminary shear wall dimensions
 Design shear walls under shear load with different wall leg parts inside core
wall group.

43
  Determine shear wall boundaries and web regions and design shear walls
under moment with different core wall group parts

6-Generating Building Models-3. MODEL (Design for Wind and Statical Gravity
Load Combinations)

 No Rigidity reduction factors is used for structural elements or "0.5" factor

can be used for lateral structural elements (beams, slabs) flexural rigidity.
 All structural members should be designed under wind and statical gravity
load combinations

3.5 LINEAR ELASTIC ANALYSIS STAGE

3.5.1 Base Shear Calculation

Modal analysis method is mandatory in high-rise buildings (TBDY-2018, 2018). In

order to obtain the correct earthquake force, some basic rules and equations
requested by the regulation are specified in the details below.

Firstly, both x and y directions of earthquakes will be separately determined by the

rule that the sum of effective modal masses should not be less than 95% of the total
mass of the building (Equation 3.3).

YM YM
 mtxn
(X)
 0.95 mt ;  mtyn
(Y)
 0.95 mt (3.3)
n=1 n=1

Table 3.10 Mass Participation Ratio Results for Case Study Buildings

Mass Participation Ratio (%) X Y

Ankara 95.31 96.47
Istanbul 96.08 95.60

(𝑋)
Base shear 𝑉tx obtained from modal analysis method for any (X) earthquake
(𝑋)
direction cannot be less than base shear 𝑉tE gotten from equivalent lateral load

44
method for tall buildings. In the event that it is for any (X) earthquake
(𝑋) (𝑋)
direction,𝑉tx < 𝑉tE , all reduced internal forces and displacements obtained by
modal analysis method are multiplied by the β (X)
tE magnification coefficient. This

magnification factor is only applied for the upper part of the building over basement
level.

(𝑋)
(𝑋) 𝑉tE
𝛽tE = (𝑋) ≥1 (3.4)
𝑉tx

Base shear from equivalent lateral load is calculated according to Equation 3.5 to
compare base shear gotten from modal analysis method.

(𝑋) (𝑋)
𝑉tE = 𝑚𝑡 𝑆aR (𝑇𝑝 ) ≥ 0.04 𝛼𝐻 𝑚𝑡 𝐼𝑆DS 𝑔 (3.5)

𝒎𝒕 : Total mass of the upper part of building calculated with Equation 3.6. Total
mass of case building for the upper part is 81,636 tons.

N
mt =  mi (3.6)
i=1

𝜶𝑯 ∶ This coefficient is correction factor to calculate minimum base shear for tall
buildings and gotten from Equation 3.7. This coefficient is 0.93 for the case building
having 112.4 m height.

 H = 1.0 H N  105 m
 H = 2.05  0.01 H N 105 m < H N  155 m (3.7)
 H = 0.5 155 m  H N

(𝑿,𝒀)
𝑻𝒑 : First natural period of building for any direction that is calculated with modal
analysis by ETABS software (Figure 3.9). However, this period is restricted to
Equation 3.8. This equation is advised by Prof. Ahmet Yakut to revise current
regulation at TEC-2018. Periods are different for buildings in Ankara and in Istanbul
because of different rigidity and dimensions of core walls.

45
 Figure 3.9 Natural periods of Case Study Buildings in Ankara and in Istanbul
calculated with modal analysis by ETABS software

(𝑋)
𝑇𝑝 < µ 𝑇pA (3.8)

µ: This value is given Table 3.10.

46
 Table 3.11 Natural Period Limit Factor (µ)

DTS 1 1a 2 2a 3 3a 4 4a
µ 1.4 1.4 1.4 1.4 1.5 1.5 1.6 1.6

It is advised by Dr. Yakut that TpA is calculated by Equation 3.9, 3.10 and 3.11 for
flat plate systems or buildings that having shear walls to respond all earthquake
design load.

𝑇𝑝𝐴 = 𝐶𝑡 𝐻𝑁𝑥 (3.9)

0.1
Ct =  0.07 (3.10)
At

  wj  
2

At =  Awj 0.2       Awj (3.11)

j   N   j
H


Design period is calculated for the buildings with these procedures and more detail
information is shown in Table 3.12.

Table 3.12 Parameters for Design Period Calculation

ANKARA ISTANBUL
X Y X Y
TP,Etabsx,y (s) 5,26 3,95 4,85 3,41
Ct 0,07 0,07 0,07 0,07
x 0,75 0,75 0,75 0,75
x,y
TpA (s) 3,62 3,62 3,38 3,38
DTS 3 3 1 1
 1,5 1,5 1,4 1,4
x,y
Tp max (s) 3,62 3,62 3,38 3,38
Tpx,yDesign (s) 3,62 3,62 3,38 3,38

(𝑋,𝑌)
𝑆aR (𝑇𝑝 ) is the spectral accelerations 𝑆ae (𝑇) (Equation 3.12) reduced according

to 𝑅𝑎 (𝑇) (Equation 3.13, 3.14) depends on response modification factor (R) and
importance factor (I) for natural periods of both direction, x and y (Figure 3.10):

47
 Sae (T )
SaR (T )  (3.12)
Ra (T )

R
Ra (T )  T  TB (3.13)
I

R T
Ra (T )  D    D  T  TB (3.14)
I  TB

Ra(T) vs. T, Periyod (s)

7
6
5
4
Ra

3 Rax
2 Ray
1
0
0,0 2,0 4,0 6,0 8,0
T, Periyod (s)

Figure 3.10 Response Modification Factor (Ra) vs. Period (T) Graph

Design base shear forces are calculated for case buildings located in Ankara and in
Istanbul with all these regulations and Equations from 3.3 to 3.14. Summary results
are given in Table 3.13.

48
 Table 3.13 Base-shear Calculation Parameters and Results

ANKARA ISTANBUL
X Y X Y
mt (t) 87589 87589
SDS 0.454 0.454 1.134 1.134
x,y
Tp modal (s) 5.26 3.95 4.85 3.41
Tpx,ydesign (s) 3.62 3.62 3.38 3.38
Rax (Tpx,y) 6 6
Sae (Tpx,y) 0.049 0.049 0.115 0.115
SaR (Tpx,y) 0.008 0.008 0.019 0.019
h,x,y 0.93 0.93
Vtx,y (ton-f) 461 791 1221 2277
VtEx,y (ton-f) 1473 1473 3679 3679
tE (x,y) 3.20 1.86 3.01 1.62
Vt,d x,y (ton-f) 1473 1473 3679 3679

3.5.2 Modeling Details

Case study buildings are structures with both having flat plate systems and relatively
more rigid basement parts. For these two exceptions, special place is reserved for
both loading and analysis in the TEC-2018.

For buildings with flat plate systems, it is stated that all earthquake loads should be
carried by ductile shear walls. In order to achieve this situation, a two-stage
earthquake calculation is applied. In stage one, the upper and lower parts of the
columns will be released, while in stage two, these joints are canceled and a model
having monolithic joints will be obtained and two types of models are obtained, and
these models are shown in Figure 3.11 and Figure 3.12. The most important issue to
be considered here is that the behavior and periods of the structure changes if the
columns joints are released from the top and bottom parts of them. This causes a
change in the base shear of the building for the model with released joints of
columns. Arrangement should be made in the Model-B (Figure 3.12) with released
joints of columns so that the base shear force is not lower than the first determined
base shear force. In our case structures, column joint release is made only for over

49
basement level. Because the structure behaves in a ductile behavior from this point
on. Due to the basement walls, no release is applied in any joints of the columns in
the basement part of the building, which has a rigid and low-ductile behavior. While
the internal forces in structural elements are selected as the most unfavorable in both
models in terms of design, the relative floor displacement should be controlled in the
model where the columns joints are monolithic. The values are taken into account as
the reactions required for the design of the core walls and link beams create more
unfavorable reactions in the model in which the columns joints are released. In the
column design, internal forces in the Model-A (Figure 3.11) in which columns joints
are not released are taken into account.

Another special issue in case study building is that while there is a very rigid
basement part in the lower sections of the buildings, the structure has a more ductile
behavior over the basement level. This situation requires the earthquake force to be
obtained by applying different system over-strength factor (D) and response
modification coefficient (R) to masses above and below the basement level of the
structure. Two different models have been created to achieve this, i.e., Model-B
(Figure 3.12) and Model-C (Figure 3.13). In Model-C, the masses in the basement
part of the structure are only entered into dynamic modal analysis. In the Model-B,
only the masses in the upper parts of the structure enter dynamic modal analysis.
While response modification coefficient (R) is 2.5 and system over-strength factor
(D) is 1.5 for the Model-C, response modification coefficient (R) is 6 and system
over-strength factor (D) is 2.5 at the Model-B. In addition, the previously mentioned
column joint release principle is valid at the Model-B because of the flat plate
system.

50
ETABS models and usage conditions created according to the regulations are more
clearly explained individually with figures below.

 Monolithic column joints

 Tx and Ty are firstly determined
 Modal analysis base shear result just
over basement is compared with
minimum equivalent base shear
 Results of G+Q vertical load
combination part is taken from this
model for G+Q+Ex,y design
combination because of column joint
releases of other model.
 Story drift checks are carried out with
R=1.
 Column design is done with this model
Figure 3.11 Model-A
and R=6 assignment

 Upper part is assigned with

mass and basement part is
massless.
 Only column joints at the
upper part are released
 R/I=6 and D=2.5
 Tx and Ty change. Base shear
is revised with new
(𝑋,𝑌)
𝛽tE magnification
coefficients according to the
first model results.

Figure 3.12 Model-B

51
  The upper part is massless and
basement part is only assigned
with mass.
 All column joints are
monolithic.
 R/I=2.5 and D=1.5
 Tx and Ty change. Base shear
calculated with modal analysis
is valid and no need
(𝑋,𝑌)
𝛽tE magnification coefficients
usage.

Figure 3.13 Model-C

Shear walls and link beams design are carried out with the sum of reactions gotten
from Model B and Model C (Figure 3.14).

Figure 3.14 Etabs Models Used for Shear wall and Link Beam Design

52
3.6 DESIGN OF SHEAR WALLS AND LINK BEAMS

As mentioned before, the main structural elements respond to all earthquake loads
are shear walls and link beams for buildings having flat slab system. This portion’s
combination of shear-wall groups and link beams are mainly defined as “core tube”
of tall buildings. The tube of case study building is designed according to TEC-2018.

3.6.1 Shear Wall Design of Case Study Building

[Link] Design Regulations of Ductile Shear Walls in TEC (2018)

Behaviors of structural elements in tall are mainly classified as linear behavior and
nonlinear ductile behavior according to capacity design principles. While shear walls
and link beams just over basement level are expected to exhibit nonlinear ductile
behavior under axial load and moments (A-Mx-My), on the other hand; axial load
and shear forces of columns, slabs, basement walls and foundations show linear
brittle behaviors. According to the capacity design approach, all brittle and linear
designs like punching, shear load transfers between slab and walls under shear and
axial loads should be avoided to prevent the collapse of a building.

All vertical load bearing structural elements are designed to be ductile for tall
buildings (TBDY-2018, 2018). Key aspects of shear wall design are limiting the
axial load, providing sufficient shear capacity and enabling ductile response through
detailing.

Shear walls dimensions are mainly determined by the axial load limitation rule and
shear design of walls. Axial load, N dm determined by design load combination

includes earthquake load (G+Q+E), should supply the condition at Equation 3.15.

𝐴𝑐 ≥ 𝑁dm /(0.35𝑓ck ) (3.15)

Design shear load, 𝑉𝑒 , should be less than shear capacity, Vr (Equation 3.16).

53
Ve ≤ Vr (3.16)

Maximum shear capacity is limited by Equations 3.17 and 3.18.

𝑉𝑒 ≤ 0.85𝐴ch √𝑓ck (Walls without spacing) (3.17)

𝑉𝑒 ≤ 0.65𝐴ch √𝑓ck (Walls with coupling beams) (3.18)

Lateral reinforcement amount is determined according to Equation 3.19.

𝑉r = 𝐴ch (0.65𝑓ctd + 𝜌sh 𝑓ywd ) (3.19)

Although the wall design under shear load is carried out according to Equation 3.19,
frictional shear capacity of the wall should also be checked by using Equation 3.20.
In the case of an inadequacy, it is the most practical approach to increase the capacity
of frictional shear capacity by increasing the vertical reinforcements of the shear-
walls.

𝑉𝑒 ≤ 𝑓ctd 𝐴𝑐 + 𝜇𝐴𝑠 𝑓yd

𝑉𝑒 ≤ min[0.2𝑓ck 𝐴𝑐 ; (3.3 + 0.08𝑓ck )𝐴𝑐 ] (3.20)

Shear loads are multiplied by system over-strength factors Dbottom=1.5 and Dtop=2.5
(𝑀𝑝 )𝑡
for design shear loads. Term, 𝛽v , at Equation 3.21 can be taken by 1.0 for
(𝑀𝑑 )𝑡

structural system having shear walls met all earthquake loads. However, over-
strength factor D is used in our design.

(𝑀𝑝 )𝑡
𝑉e = 𝛽v 𝑉𝑑 (3.21)
(𝑀𝑑 )𝑡

[Link] Shear Wall Elastic Design Results of Case Study Building

Dimensions and labels for shear design of core wall of case building in Ankara and
in Istanbul are shown in Figure 3.15 and Figure 3.16 respectively. Boundaries of
shear walls are arranged according to code regulations. Boundary, web arrangements
and longitudinal reinforcement of them are shown in Figure 3.17 and Figure 3.18 for

54
the case study buildings in Ankara and in Istanbul, respectively. According to
minimum vertical reinforcements (min=0.0025) in TEC-2018, practically used
longitudinal reinforcement for the web of shear walls is shown in Table 3.14. In
addition, lateral design reinforcements of walls of case study building in Ankara and
in Istanbul are shown in Table 3.15 and Table 3.16.

Table 3.14 Minimum Vertical Reinforcements of Web of Shear walls

Width of Wall (cm) 30 40 50 60 80 90

Web Vertical
Φ12/20 Φ12/20 Φ14/20 Φ14/20 Φ16/20 Φ18/20
Reinforcement

Figure 3.15 Thickness and Labels of Walls of Case Building in Ankara

55
Figure 3.16 Thickness and Labels of Walls of Case Building in Istanbul

56
Figure 3.17 Boundary, Web Placement and Boundary Reinforcement of Shear
walls of Building in Ankara

57
Figure 3.18 Boundary, Web Placement and Boundary Reinforcement of Shear
walls of Building in Istanbul

58
Table 3.15 Lateral Reinforcement of Walls of Case Building in Ankara

STORIES
Width Length
LABEL 1--4 5--6 7--9 10--32
(cm) (cm)
W1 40 345 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W2 40 665 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W3 40 585 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W4 40 345 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W5 40 315 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W6 40 380 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W7 40 335 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W8 40 380 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W9 40 345 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W10 40 665 Φ12/19 Φ12/19 Φ12/20 Φ12/20
W11 40 715 Φ12/18 Φ12/14 Φ12/15 Φ12/20
W12 40 345 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W13 40 335 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W14 40 345 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W15 40 665 Φ12/20 Φ12/18 Φ12/20 Φ12/20
W16 40 585 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W17 40 1800 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W18 30 410 Φ10/20 Φ10/20 Φ10/20 Φ10/20
W19 30 1625 Φ12/20 Φ10/20 Φ10/20 Φ10/20
W20 30 710 Φ12/20 Φ10/20 Φ10/20 Φ10/20
W21 30 1800 Φ12/14 Φ10/20 Φ10/20 Φ10/20
W22 40 1800 Φ12/10 Φ12/20 Φ12/20 Φ12/20

59
 Table 3.16 Lateral Reinforcement of Walls of Case Building in Istanbul

STORIES
Width Length
Label
(cm) (cm)
1--3 4 5 6 7--9 10--12 13--16 17--20 21--32
W1 60 345 Φ14/20 Φ14/20 Φ16/15 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W2 60 665 Φ16/15 Φ14/20 Φ16/14 Φ16/15 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W3A 60 335 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W3B 60 250 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W4 40 345 Φ12/20 Φ12/20 Φ14/15 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W5 40 315 Φ12/15 Φ12/20 Φ12/14 Φ12/15 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W6 40 380 Φ12/13 Φ12/20 Φ12/15 Φ12/15 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W7 40 335 Φ12/20 Φ12/20 Φ12/15 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W8 40 380 Φ12/15 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W9 40 345 Φ12/20 Φ12/20 Φ14/15 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W10 50 665 Φ16/10 Φ16/15 Φ16/15 Φ16/10 Φ16/10 Φ16/15 Φ14/15 Φ14/15 Φ14/20
W11 60 715 Φ20/10 Φ18/10 Φ20/10 Φ20/10 Φ20/10 Φ18/10 Φ18/15 Φ16/15 Φ16/20
W12 40 345 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W13 40 335 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W14 60 345 Φ14/20 Φ14/20 Φ14/10 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W15 60 665 Φ16/10 Φ16/15 Φ16/10 Φ16/14 Φ14/15 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W16A 60 335 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W16B 60 250 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
60-
W17 1800 Φ18/15 Φ18/20 Φ18/15 Φ16/20 Φ16/20 Φ16/20 Φ16/20 Φ14/20 Φ14/20
80-90
W18 40 410 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20 Φ12/20
W19 50 1625 Φ16/10 Φ14/10 Φ14/15 Φ14/15 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W20 50 710 Φ16/13 Φ16/15 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
W21 60 1800 Φ20/10 Φ16/10 Φ14/15 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20 Φ14/20
60-
W22 1800 Φ25/11 Φ18/10 Φ18/20 Φ16/20 Φ16/20 Φ16/20 Φ16/20 Φ14/20 Φ14/20
80-90

Demand and capacity graphs in terms of design under axial load and shear load are
given in Figures 3.19 to Figure 3.26 for case study buildings in Ankara and Istanbul
separately.

60
Figure 3.19 Axial Load Capacity Check Diagrams of W1 to W12 Walls (Ankara)

61
Figure 3.20 Axial Load Capacity Check Diagrams of W13 to W22 Walls (Ankara)

62
Figure 3.21 Shear Load Capacity Check Diagrams of W1 to W12 (Ankara)

63
Figure 3.22 Shear Load Capacity Check Diagrams of W13 to W22 Walls (Ankara)

64
Figure 3.23 Axial Load Capacity Check Diagrams of W1 to W12 (Istanbul)

65
Figure 3.24 Axial Load Capacity Check Diagrams of W13 to W22 (Istanbul)

66
Figure 3.25 Shear Load Capacity Check Diagrams of W1 to W12 Walls (Istanbul)

67
 Figure 3.26 Shear Load Capacity Check Diagrams of W13 to W22 (Istanbul)

Shear force distribution according to wall labels just above the basement level is
shown in Figure 3.27. W2, W10, W11 and W15 walls for x direction and W17, W19,
W21 and W22 walls for y direction are dominant to respond to base shear reaction.

68
 Figure 3.27 Shear Force Distribution Ratio at 5th Story

Labels of flexural design of core wall are different from labels of shear design.
Because under flexural behavior in high-rise buildings, it is envisaged that the core-
wall works in groups, not separately, as in shear design. Because it is not possible to
separate each piece of wall with the common intersection boundaries having dense
vertical reinforcements, as in shear design. It is predicted that wall groups work as a
single section under flexural behavior in such high structures. For this reason,
vertical reinforcement design is carried out as WG1, WG2, WG3 sections (Figure
3.28) and the results are given in Figure 3.29.

69
 Figure 3.28 Labels and Wall Groups for Flexural Design of Core-wall

Figure 3.29 Flexural Design Results (Demand/Capacity Ratio vs. Height) of Wall
Groups of Building in Ankara and in Istanbul

70
3.6.2 Coupling Beam Design of Case Study Building

[Link] Design Regulations of Coupling Beams in TEC (2018)

The coupling beams between wall groups have high shear strength and allow the two
different wall groups to work together as a single wall. Whether a single shear wall
behavior or two separate shear walls is exhibited depends on whether the beam
between the walls serves as a coupling beam, through the coefficient of the degree
of coupling, Ω in TEC-2018, provided in Equation 3.24. If this condition is satisfied,
the beam is a coupling beam and two separate shear wall pieces are considered to
working as a single shear wall. In case Equation 3.24 is not provided, the walls are
designed separately. The coefficient of the degree of coupling, Ω, is calculated by
Equation 3.22 and Equation 3.23. M1 and M2 are moments of shear walls under only
earthquake load at the base level, where walls show maximum flexural behavior. Nv
and c are the sum of all coupling beams shear forces at the same beam joint and the
distance of center of gravity of wall sections, respectively.

M DEV  M1  M 2  c NV (3.22)

c NV c NV
  (3.23)
M DEV M1  M 2  c N V

1
 (3.24)
3

The shear design of the coupling beams is very important. For this reason, the
reinforcement detail varies in the coupling beams according to the magnitude of the
shear force. If both the Equation 3.25 and 3.26 conditions are not met, diagonal
reinforcement bars will be placed and the amount of these reinforcements are
calculated by Equation 3.27. If the shear force is not excessive and any of Equations
3.25 and 3.26 are satisfied, beam is designed under shear similar to a slender beam
according to Equations 3.28, 3.29 and 3.30. If the maximum shear force in the beam

71
does not meet the condition in Equation 3.30, the beam dimensions should be
changed.

ℓ𝑛 > 2ℎ𝑘 (3.25)

𝑉𝑑 ≤ 1.5𝑏𝑤 𝑑𝑓ctd (3.26)

𝐴sd = 𝑉d /(2𝑓yd sin𝛾) (3.27)

𝑉e = 𝑉dy ± (𝑀pi + 𝑀pj )/ℓ𝑛 (3.28)

𝑀pi ≈ 1.4𝑀ri 𝑀pj ≈ 1.4𝑀rj (3.29)

Ve ≤ Vr
𝑉𝑒 ≤ 0.85𝑏𝑤 𝑑√𝑓ck (3.30)

[Link] Coupling Beams Design Results

In our case building, there are both beams and coupling beams between core wall
groups. Beam and coupling beam dimensions are 40/60 cm and 40/150 cm
respectively in the case structure in Ankara (Figure 3.30). On the other hand, in the
case study building in Istanbul, there are beams in dimensions of 40/60 cm and 60/60
cm, and coupling beams in dimensions of 60/150 cm (Figure 3.31).

72
Figure 3.30 Beam and Link Beam Dimensions of Case Study Building in Ankara

Figure 3.31 Beam and Link Beam Dimensions of Case Study Building in Istanbul

73
Link beams, LB1 and LB2, are checked as that coefficient of the degree of bond, Ω
is greater than 0.33 or not. Results are shown in Table 3.17. Table shows that LB1
and LB2 shows coupling beam behavior.

Table 3.17 The Coefficient of the Degree of Bond, Ω for Coupling Beams

ANKARA ISTANBUL
LB1 (40/150) LB2 (40/150) LB1 (60/150) LB2 (60/150)
M1 (t.m) 372 350 940 829
M2 (t.m) 152 137 366 318
NV (t) 220 250 630 650
c (m) 4.1 4.1 4.1 4.1
Ω 0.63 0.68 0.67 0.70

Design results of coupling and frame beams are shown for case study buildings in
Ankara and In Istanbul in Table 3.18 and Table 3.19, respectively.

Table 3.18 Coupling Beam Design Results of Case Study Building in Ankara

ANKARA
LB1 (40/150) LB2 (40/150) Beam (40/60)
Vcr (t) 0,0 0,0 0,0
Vc (t) 0,0 0,0 0,0
Vw (t) 163,0 163,0 62,0
Vmax (t) 349,0 349,3 133,0
Ve (t) 150,0 118,7 50,6
Vr (t) 0,0 0,0 0,0
Md (t.m) 53,7 45,2 32,3
Stirrup Ф14/10 Ф14/10 Ф14/10
Rein. top, bot. 5Ф25 5Ф25 5Ф22
Conventionally Conventionally Conventionally
Reinforced Reinforced Reinforced

74
 Table 3.19 Coupling Beam Design Results of Case Study Building in Istanbul

ISTANBUL
LB1 (60/150) LB2 (60/150) Beam (40/60) Beam (60/60)
Vcr (t) 0 0 0 0
Vc (t) 0 0 0 0
Vw (t) 425.1 425.1 123.8 161.1
Vmax (t) 523.3 523.3 132. 5 198.0
Ve (t) 211 159 97.5 127.4
Vr (t) 0 0 0 0
Md (t.m) 187 139 56.2 81.3
Stirrup 2 x Ф16/10 2 x Ф16/10 2 x Ф14/10 2 x Ф16/10
Rein. top, bot. 6Ф30 6Ф30 5Ф30 7Ф30
Diagonal Bars 8Ф30 8Ф30 -- --
Diagonally Diagonally Conventionally Conventionally
Reinforced Reinforced Reinforced Reinforced

All design beam drawings are shown in Figure 3.32 and Figure 3.33 below.

a)

b)

Figure 3.32 a) Beam (40/60) and b) Link Beam (40/150) Reinforcement Details fo
Case Study Building in Ankara

75
 a)

b)

c)

Figure 3.33 a) Beam (40/60), b) Beam (60/60) and b) Link Beam (60/150)
Reinforcement Details of Case Study Building in Istanbul

3.7 DRIFT RATIO RESULTS OF CASE STUDY BUILDING FOR

LINEAR ELASTIC DESIGN

In TEC-2018, the story drift ratio is controlled according to the Equation 3.31.
According to this formula, κ is 0.5 for concrete buildings and  is 0.398 and 0.526
for buildings in Ankara and in Istanbul, respectively. According to these values, the
simplified version of Equation 3.31 is shown in Equation 3.32 and 3.33 for building

76
in Ankara and Istanbul, respectively. The story drift ratio check is shown in Figure
3.34.

(X)
δi,max
  0.008  (3.31)
hi

(X)
δi,max
≤ 0.01 (3.32)
hi

(X)
δi,max
≤ 0.0076 (3.33)
hi

Figure 3.34 Drift Ratio Results of Case Study Building in Ankara and in Istanbul

77
78
 CHAPTER 4

4 PERFORM 3D V7.0.0 SOFTWARE

Perform 3D program is the most widely used program for nonlinear inelastic
modeling and design check of tall buildings. It is important to know the working
principles of this program in order to conduct modeling correctly. The displacement-
based design is a very different approach from the strength-based design. The
nonlinear inelastic behavior of each structural element should be well understood by
the engineer who will use the program and accurately reflect the behavior
accordingly. In this complex modeling, it is important to simplify the model to
minimize the time of analysis and the accuracy of models. The correct modeling of
walls and link beams with Perform 3D is explained in detail in this chapter.

Perform 3D has basically 2 phases, i.e., the modelling phase and the analysis phase.
In the modelling phase; nodes, component properties, elements, loads, drifts,
structure sections, and limit states are defined. At the analysis phase; load cases,
analysis series, modal analysis results, energy balance results, deflected shapes, time
histories, hysteresis loops, moment-shear diagrams, push-over plots, usage ratio
graphs, combinations and envelopes parts are visualized.

4.1 MODELING PHASE

4.1.1 Nodes

The first step in the modeling phase is the definition of nodal points. Nodes can be
created in the nodes section with the desired number and position to connect the
building elements together to create the geometry of the structure. Coordinates are
defined in 3D with x, y and z coordinates as H1, H2 and V. The minimum distance
between the nodes can be determined by the user. In addition, restraints or supports

79
can be assigned to the desired nodes. Masses can only be assigned to nodes. Mass
assignment on an element basis is not possible. Mass assignment to 3 displacement
directions and 3 rotation directions can be made separately. With the slaving section;
horizontal rigid floor, eccentric connection in a rigid floor, full rigid link and simple
equal displacement options can be modeled. In the use of the horizontal rigid floor
option, all axial loads in beam models will be zero. However, when fiber sections or
concrete type P-M-M hinges models are used in the beam model, rigid floor
constraint, which prevents the beams extension, will cause compression forces.

4.1.2 Components Properties

The component properties section is one of the most time-consuming parts of

modeling in the P3D program. This section describes materials, cross sections, basic
structural components, strength sections, and compound components (Figure 4.1).
While material definitions help define element sections, basic structural components
are defined through both material and cross sections. Also, the strength check of the
building elements that will exhibit elastic behavior is possible, while all the
characteristics of the building elements are combined according to their behavior in
the compound component task.

80
Figure 4.1 Perform 3D Component Properties Task

81
[Link] Material Properties

The materials of the structural elements to be used can be defined first. In the
materials section, which is mainly elastically and inelastically distinguished, force-
displacement behavior curves of steel and concrete materials are obtained. In the
inelastic materials section, steel material is defined in 3 forms, i.e., inelastic steel
materials non-buckling, tension only and buckling. The remaining inelastic material
definitions are 1D concrete material, shear material and compression for a wall. On
the other hand, elastic material definitions are divided as elastic material for a fiber
section, elastic shear material for a wall and elastic material for a slab or shell. Slabs
are only defined as elastic material.

In our model, “inelastic steel material, non-buckling” (Figure 4.2) is used to define
the reinforcement of the walls, on the other hand, “inelastic 1D concrete material”
(Figure 4.3) is selected for the reinforced concrete fiber sections model. The option
of “elastic shear material for a wall” (Figure 4.4) is considered for the design of a
wall under shear force because it is assumed that the behavior of a wall under shear
force would be elastic in the walls of high-rise buildings where bending behavior is
governed.

Figure 4.2 ”Inelastic Steel Material” Input Panel

82
 Figure 4.3 ”Inelastic 1D Concrete Material” Input Panel

Figure 4.4 ”Elastic Shear Material for a Wall” Input Panel

83
Inelastic behavior can be created with various parameters and options to define
hysteric loop at P3D in the form of force-displacement, stress-strain, force-rotation,
and force-curvature curves, both on the material inelastic behavior level or the basis
of section inelastic behavior. With these definitions, the hysteric loop model
“YULRX” is created. Highlights of the backbone curve, which defines the outlines
of the hysteric loop, can be defined as yielding point (Y), ultimate strength point (U),
ductile limit point (L), residual strength point (R), and analysis stop point (X) (Figure
4.5 and Figure 4.6).

Figure 4.5 Perform 3D “YULRX” Backbone Curve(Computers and Structures,

Inc., 2006)

84
 Figure 4.6 Perform 3D Hysteretic Loop Model(Computers and Structures, 2006)

“YULRX” hysteric loop is defined by the parameters under the basic headings, that
define the shape of the force deformation relationship, cyclic degradation and
strength loss. In terms of “shape of relationship”, one of two options can be selected
as in the form of elastic perfectly plastic (EPP) or trilinear relationship. Dissipated
energy (the area of the loop) in the hysteric loop can be determined by cyclic
degradation energy factors in the “YULRX” model. This energy degradation also
affects reloading and unloading stiffness values. In addition, the change of stiffness
in the case of unloading behavior for the shape of the trilinear backbone curve can
be controlled by the unloading stiffness factor. Although the ductile limit point (L)
cannot be exceeded even by the collapse prevention limit in the regulations, the
strength loss situation in the hysteric loop can be considered in the modeling at P3D.
There is a strength loss interaction factor (SLIF) determining strength loss only in
the positive direction (SLIF=0) or both positive and negative direction symmetrically
(SLIF=1) of the cyclic loop (Figure 4.7). A value between 0 and 1 will give a strength
loss midway behavior in terms of loss direction at a hysteric loop. So, hysteric loop
behavior can be created in symmetrical order or asymmetrically with different SLIF
values while all other relationships are maintained.

85
This principled approach is the same both in the definition of inelastic behavior
within the material itself and in the definition of inelastic behavior of the section.

Figure 4.7 Strength Loss Interaction Factor (SLIF=0 (a), 1 (b)) (Computers and
Structures, Inc., 2006)

There can be a decrease in stiffness, strength and dissipated energy under cyclic
loading of structural elements. The reduction in stiffness and strength affects energy
dissipation. In the P3D program, a cyclic hysteric loop can be adjusted by using the
energy degradation factor and unloading stiffness factors (Figure 4.8). Energy
degradation factor is the ratio of reduced area of degraded hysterical loop to initial
area of undegraded hysterical loop (Figure 4.9 (a)). In Figure 4.9 (b), the energy
degradation factor for loops A, B, and C is the same in three and is 0.55 (1.0=no
degradation). But the difference in these three is adjusted by the unloading stiffness
factor. The unloading stiffness factor is +1, -1, and 0 for loop A, B, and C,
respectively. Elastic range is reduced in loop A, while stiffness decreases in loop B.
In loop C, however, there is a decrease in both elastic range and stiffness. Energy
degradation factor and unloading stiffness factor are two important parameters for
determining the hysteric loop. For this reason, the correct selection of these two

86
factors is important to reflect the correct structural element inelastic behavior
confirmed by experiments on the model. As in many high-rise buildings,
experimental behaviors of walls and link beams, that also exhibit inelastic behavior
in our sample structure, were reflected in the model by calibration of these two
parameters. In this study, the recommended values for these coefficients for different
types of walls and link beams were identified by conducting detailed parametric
studies.

Figure 4.8 Perform 3D Input Panel for Cyclic Degradation Factor and Unloading
Stiffness Factor

87
 a)

b)

Figure 4.9 a) Cyclic Degradation Factor b) Hysteric Loop with The Same Cyclic
Degradation Factor and Different Unloading Stiffness Factors (USF=+1 (A), -1
(B), 0 (C)) (Computers and Structures, 2006)

[Link] Cross Section Definition at Perform 3D

Cross-section definitions are made for walls and link beams. The wall structural
element cross sections can be modeled by fibers as both the reinforcement and the
concrete fibers are separately defined horizontally and vertically at P3D (Figure
4.10). The cracking or crushing behavior of the concrete regions and the yielding or
buckling of the reinforcement parts of the wall section under bending and axial load
can be modeled with the uniaxial material behavior define the fibers. The number of
fibers, area of concrete and reinforcement parts can be defined in the P3D program
either as “fixed size” or “auto-size” (Figure 4.11). While areas, amount and
coordinates of fibers are specified numerically in a computational inefficient way for
the fixed size option, the auto size option more easily specifies the number of

88
reinforcement and concrete fibers divided into equal areas within the same wall
width length. As a result, “auto-size” gives you a faster modeling option, while
“fixed size” gives you a more flexible modeling option.

Figure 4.10 Fiber Model of Walls at Perform 3D (PEER/ATC 72-1, 2010)

Figure 4.11 Wall Fiber Modeling Options at Perform 3D (Computers and

Structures, Inc., 2006)

“Beam, Reinforced Concrete Section” can generally be used for the link beams as
well. Various properties of the link beam sections can be defined in this selection.
(Figure 4.12). These are the dimensions of the beam, section stiffness properties,
axial and shear areas, inertias, and material stiffness properties like young’s modulus,
shear modulus and poison’s ratio. In addition to these properties, inelastic bending
strength, inelastic shear strength, elastic nominal bending or shear strengths of that

89
section can be defined. Maximum allowable axial force or axial stress can also be
defined in terms of elastic capacity.

Figure 4.12 Beam, Reinforced Cross Section Input Panel at P3D

There is a modeling sequence while using P3D. Defining materials of concrete and
reinforcement are the first stage of modeling for fiber wall elements. Cross-section
definitions are first conducted for frame elements like beams and columns at P3D.
After material definitions and section properties are determined, inelastic or elastic
basic components should be defined. Then, all the detailed components are created
with these definitions to exhibit structural behavior, and these components can be
properly combined with the compound component task at the last step of modeling.

[Link] Basic Components and Strength Section

Basic components can be defined in two separate sections: inelastic and elastic
components. A response curve for inelastic behavior, i.e., force vs. displacement

90
curves or derivatives such as moment-rotation, moment-curvature, axial force-strain,
shear force-displacement, can be defined. Such material models can be defined for
beams, columns, bars, connection panel regions, seismic isolator, fluid damper and
BRB (Buckling Restrained Brace) at P3D. End zones of column and beam
connections, linear hinge and release parts, elastic bars and springs, elastic
connection panel zones, strain and rotation gages are components that can be defined
elastically. Inelastic or elastic section definitions can be made with basic component
tabs, while section checks in terms of force-based behavior can be made with
definitions on the strength sections tab. All critical sections of structural elements
that do not have ductile behavior can be checked in terms of strength with moment,
shear and axial load capacity definitions.

In the modeling of the high-rise case study building in this study, “Moment Hinge,
Rotation Type” and “Shear Hinge, Displacement Type” components of the inelastic
section of link beams can be used in the model (Figure 4.13). In addition, the
components “Axial Strain Gage (2-node)”, “Rotation Gage, Beam Type (2-node)”
and “Rotation Gage, Wall Type (4-node)” from the elastic section can also be used
(Figure 4.14).

Figure 4.13 Inelastic Components Input Panel of Link Beam at P3D

91
 Figure 4.14 Elastic Gage Component Input Panel at P3D

4.1.3 Compound Components

The definitions of structural elements to be used in modeling are completed step by

step in P3D. In these definitions, many sub-parameters and model definitions are
made according to the behavior of the structural element. Many minor components
such as material models, cross-section properties, elastic and inelastic behavior
models, and strength definitions are defined. The formation of elements at P3D is
done by the proper combination of these components. The behavior of structural
materials is complemented by the proper join of the components according to the
structural elements that are columns, beams and walls.

Elastic parts of the beam elements are defined by “cross-section” components

including stiffness and strength properties. In addition, the shear strength definition
is defined separately. If the beam is not a link beam, the beam exhibits inelastic
behavior under bending and while elastic behavior is assumed for the shear response.
Plastic hinge behavior is expected at the ends of the beam. Plastic hinge definition
can be done in terms of rotation or curvature by defining a plastic hinge length
(Figure 4.15). After all the components required for beam element definitions are
completed, they are combined in the compound component task at P3D (Figure
4.16).

92
 Figure 4.15 Beam Models with Plastic Hinge and Curvature Hinge Components
(Computers and Structures, Inc., 2006)

Figure 4.16 Compound Components of Beam with Moment Hinges at P3D

While many beam behaviors are inelastic under bending and elastic under shear
demands, the opposite is true for link beams. Inelastic shear behavior is active in link
beams because of the dominance of shear force. The link beam model basically
consists of the plastic shear-hinge in the middle of the elastic segment of the beam

93
and ends where moment strength sections are shown in Figure 4.17. Also, shear
hinge can be modeled on the basis of strain or displacement in P3D. Typical link
beam element model is shown in Figure 4.18.

Figure 4.17 General Approach for Compound Components of Link Beam

(Computers and Structures, Inc., 2006)

Figure 4.18 Compound Components Sample of Link Beam at P3D

The compound components formation of walls includes combination of fiber-cross

section definitions for axial-bending properties and "Elastic/Inelastic Shear Material
for a Wall" component for shear properties. In the fiber-cross section, definition can
be made as shear wall or general shear wall components. Only vertical inelastic fiber

94
formation is allowed for the shear wall module, on the other hand both vertical and
horizontal inelastic fibers can be defined with the general shear wall module. If the
wall type is a squat wall, or if a wall has irregular openings, it is appropriate to use a
general shear wall module. But shear wall component is generally used for slender
common walls.

4.1.4 Wall and Link Beam Elements

The components of wall and link beam elements are initially created and combined,
then assignments of them are made to the elements in the model. All structural
elements in the model can be generally grouped to control the results of analysis
conveniently according to their types, sizes, positions and reinforcement ratios. In
addition, local axis directions should be defined individually at P3D. Geometric
nonlinearity identification can be considered for all elements. The "K" viscous
damping ratio for each element can be increased and reduced individually.

The working principles and true modeling of wall and link beam elements in P3D
should be known for correct modeling of 3D total structure. For this reason, some
important subjects related to wall element are stated in the following items below:

 Each wall element consists of 4 nodes (K-L-I-J) and each node has six
degrees of freedom.
 Common definition of three local axes is axis 1, 2 and 3 for normal, vertical
and horizontal directions respectively.
 Positive sign convention of wall element for axial force, shear force and
bending moment is shown in Figure 4.19.

95
Figure 4.19 Positive Sign Convention of Wall Element(Computers and Structures,
Inc., 2006)

 Using the minimum number of elements and fibers by providing real

structural behavior in the model will reduce analysis time and margin of
error. Therefore, walls should be modeled with an optimum number of fibers.
 Neutral axis shifts towards the compression zone after cracking and yielding
of reinforcement in the tension section of the walls. It is important to take
into account the shift of the neural axis in the wall under axial loads and
bending loads in order to accurately reflect inelastic wall structural behavior.
P3D takes this important behavior into account.
 The amount of mesh of the walls vertically and horizontally should be kept
at the optimum level, so that both accurately reflecting the structure behavior
and shortening the analysis time. In high-rise buildings, it may be enough to
model one wall element on each floor in a vertical direction. Curvature, axial
and shear strains are constant throughout each wall mesh element. For this
reason, the number of elements can be increased or selected at a reasonable
height in the hinge section, which has plastic behavior for the walls.
 Appropriate inelastic hinge section height is very important in the modeling
of walls because it affects the calculated strain and bending moment
capacities. The correct height selection of elements representing hinge length
is a very sensitive parameter that affects the accuracy of the calculated strain.

96
 According to ASCE 41, the hinge length of walls is minimum of one-half of
cross-section depth and story height.

 In high-rise structures, core wall groups exhibit elastic behavior as they

continue upward while inelastic behavior exists in the hinge section of the
wall where they are connected with the foundation. As an alternative
modeling, inelastic modeling can be used in the lower hinge sections, while
the elastic wall component can be used in the upper sections. So, this will
both shorten the analysis time and minimize the error. The point to note here
is that a sufficient number of inelastic elements should be defined to represent
the hinge section.

 There is no definition of in-plane rotational stiffness in the nodes of a wall

element. Therefore, the connection between the beam and the wall is pin
connection. For this reason, imbedded beam modeling is required in order to
have moment-resisting connection between wall and beam. As in our sample
structure, it is necessary to pay close attention to the use of imbedded beam
for the correct model of link beams and core wall groups. The wall element
can also be used in the modeling of link beams, but the use of frame elements
together with the imbedded beam allows for better control. The use of
imbedded beam for deep link beams (1.0< ln /h <2.0) is throughout vertically
linking with walls at beams own height (Figure 4.20). For conventionally
reinforced link beams (3.0 < ln/h <4.0), imbedded beams are connected
horizontally to walls (Figure 4.21). The bending inertia of imbedded beams
modeled for slender link beams is taken as 20 times more than the stiffness
of the connected link beams inertia, while the axial area and torsional
stiffness are close to zero. The bending rigidities of the imbedded beams
required for deep beams in the weak direction should also be close to zero.
The reason for reducing the rigidities too much is to prevent increased local
stiffness in a way that affects the behavior of the wall.

97
Figure 4.20 Deep Link Beam (1.0 < ln/h <2.0) Connected to Wall with Imbedded
Beam (Computers and Structures, Inc., 2006)

Figure 4.21 Link Beam (3.0 < ln/h <4.0) Connected to Wall with Imbedded Beam
(Computers and Structures, Inc., 2006)

 "Monitored" fiber sections can be used to observe concrete and steel strain
demands while identifying shear wall fiber cross-section. However, it is more
appropriate to define strain gages describing the gage length for the curvature
calculation and detect curvatures at the boundary of the walls. Because
average measurements, which are more accurate and suitable, can be
obtained with average strain measurements, while local stress concentrations
can be detected with the "monitored" fiber method. The important thing here
is to make the determination of the gage length correctly.
 It is appropriate to make shear strength demand/capacity check with the
"structure section" for each wall arm. Because “structure section” help to see
the average shear strength of walls.

98
  In-plane behavior of shear wall element is more effective and dominant than
out-of-plane behavior. For this reason, out of plane bending is defined as
elastic behavior, while in-plane behavior is defined as fiber inelastic cross-
section. The out-of-plane elastic bending rigidities of walls should be reduced
by one-quarter.

Walls and link beams can effectively transfer earthquake forces to the foundation.
Acceptances and necessary information in the modeling of wall elements in P3D are
listed in the items above. In addition, having the necessary information about link
beam will be important for accurate modeling and result evaluations. The important
subject related to the modeling of beams in P3D are presented in the following items:

 Sign convention of beams is shown in Figure 4.22. Tension axial load, shear
force along axis 2 and bending about axis 3 are positive signs.

Figure 4.22 Sign Convention of Beam at P3D (Computers and Structures, Inc.,
2006)

 There are lots of component alternatives when creating compound

components for beam elements. These can be auto stiff end zone, P/V/M
release or linear hinge, elastic cross-section segment, inelastic fiber cross
section, rotation or curvature moment hinges, semi-rigid moment
connections and strain/displacement shear hinge components.
 The auto end zone component is actually automatically defined in column
and beam connections at P3D. This zone has rigidity of 10 times more than
the body of components according to the size of the column and beam.

99
  Rotation measurements of the beams having the stiff end zones are made
according to the clear span of the beams (Figure 4.23). This desired approach
is also considered in P3D.

Figure 4.23 Chord Rotation of Beam with Stiff End Zone (Computers and
Structures, Inc., 2006)

4.1.5 Other Modeling Phase Subjects

Some of the remaining important notes on the modeling phase in P3D are as follows:

 Slabs can be only modeled elastically by “Elastic Slab/Shell Element”

component with in-plane and out-plane bending stiffness for deformable
floor diaphragms. The sign convention of slabs can be seen in Figure 4.24.
Slab/shell elements have not any dissipated energy function in terms of K
damping.

Figure 4.24 Elastic Slab/Shell Component Sign Convention (Computers and

Structures, Inc., 2006)

100
  Loads can be given on nodes and elements. Self-weight is automatically
calculated according to gravity nodal loads, and geometric properties of
elements by P3D.
 “Structure Section” module can be used to determine the total sum of
average forces of a number of elements. It is common to check the shear
strength of a wall with this function. This can be accomplished by checking
the ratios of walls strength with average internal force.
 Limit states of deformation (rotations, curvature, and displacements),
strength, drifts, deflection and shear strength of structure sections can easily
be defined at P3D, and the demand-capacity ratio of these states is generally
1.0. If you want to easily increase the capacity of a component with %20,
demand-capacity ratio can easily be selected as 1.2. It is possible to assign
several limit states for different element groups and performance levels, i.e.,
immediate occupancy, life safety and collapse prevention.

4.2 ANALYSIS PHASE

4.2.1 Load Cases

There are a number of analysis options namely, gravity, dynamic earthquake (time
history analysis), push-over and response spectrum analysis at P3D. Linear and
nonlinear options can be assigned with gravity load cases. Most of the structures can
be accepted as elastic under gravity loads. So, there is no need to use nonlinear
analysis under gravity loads.

Event to event solution strategy is used for the nonlinear analysis. P3D conducts
nonlinear analysis in a number of predetermined load steps. These load steps are also
divided into events. These numbers should be enough to have a stable solution for
nonlinear analysis. There can be unstable situation and analysis will stop at the end
of maximum number of load steps.

101
Dynamic earthquake analysis, i.e., time history analysis, is conducted with a step-
by-step integration method, i.e., Newmark-constant average acceleration method
with acceleration records of H1 and H2 directions. Any earthquake records can be
defined and assigned with placing acceleration-time pair data to records folder.

4.2.2 Analysis Sub-Options and Damping

Analysis time is very important for seismic performance nonlinear analysis. In

particular, it is even more important in the solutions of high-rise structures with high
analysis time. The amount of overshoot tolerances for a yielding element can be
adjusted to decrease run time. So, event overshoot default factor “1 (%1)” can be
increased to “5 (%5)” or “10 (%10)” for large structures. However, one should make
sure that there is not a significant difference between external and internal work
energy after analyzes. These should be almost the same.

Structure dissipates some of the total input energy with viscous damping. Correct
damping assignments affect the remaining part of energy release and analysis results.
Viscous damping can be modeled as modal damping or Rayleigh damping.

P3D calculates up to 50 mode shapes. However lower modes after 50 modes are also
important for high-rise buildings or great structures for correct total behavior in terms
of including damping. While all mode shapes are damped up to 50 modes, but lower
modes are not damped if program user only use “Modal Damping” module. Damping
in short period modes should also be included for high-rise buildings. So, it is strictly
advised that small amount of K Rayleigh damping should be added to modal
damping to damp higher mode displacements for high-rise buildings.

Rayleigh damping is used also for nonlinear analysis. This damping model is given
by Equation 4.1. M is structure mass matrix, K is the initial elastic stiffness matrix,
and  and  are multiplying factors.

C = M + K (4.1)

102
Velocity dependent viscous dampers are arranged as Figure 4.25. While K viscous
dampers are internally connected between masses, M viscous dampers are
connected to external masses (Figure 4.25). K and M viscous dampers complete
each other for different periods. K viscous dampers results less damping in lower
modes and more damping in higher modes, on the other hand, K viscous dampers
results more damping in lower modes and less damping in higher modes (Figure
4.26). Rayleigh damping solutions are uncoupled between mode shapes. There is no
effect of a damping force related to a mode with another mode.

Figure 4.25  and  Viscous Damper Arrangement in Model (Computers and

Structures, Inc., 2006)

103
 Figure 4.26 Variation of M and K Damping Ratios with Periods (Computers
and Structures, Inc., 2006)

K damping uses initial stiffness of the structure in a nonlinear calculation. However,

stiffness continuously changes after cracking of concrete at nonlinear analysis.
Inelastic concrete fibers cracks, neutral axis changes and initial high stiffness
decreases. The same situation is valid for a coupling panel, modeled with wall
module, between coupled shear walls. P3D automatically decreases initial total
stiffness to %15 of initial stiffness for damping calculation purposes. For other
elements which use inelastic fibers, user should have an awareness that P3D does
not decrease initial stiffness of that element like wall if Rayleigh damping option is
used. This causes the estimation of k type energy dissipation. k values should be
revised for some element groups like coupling panels and include concrete fiber with
“Scale Factor for Beta-K Damping”.

Both damping methods are suitable, but to decrease run time and to get enough mode
shapes, it is advised that both Rayleigh and modal damping method can be run within
an order for preliminary analyzes then after checking that both methods give similar
results, modal damping method can be used with enough number of mode shape and
optimum run time.

104
4.2.3 Energy Balance

External earthquake energy is converted on structure to internal energy as stored

elastic energy and dissipated inelastic energy. Stored elastic energy can be divided
as kinetic energy on masses and recoverable strain energy in the elements while non-
recoverable (damaged) inelastic energy dissipated in the elements and viscous
damping energy can be generally thought as dissipated inelastic energy (Figure
4.27). Earthquake energy can be consumed on structure with movement, elastic
deformations, yielding, cracking, damage and friction etc. P3D shows us the amount
of consumed energy at different forms mentioned above separately. Energy balance
plots are drawn only for dynamic analyzes (time history analyzes) at P3D.

Energy balance task helps the P3D users to assess the performance of the structure.
It can be determined that which element or element group contributes to the inelastic
dissipated energy with how much amount. In addition, it can be observed that relative
amount of dissipated energy is contributed by different structural element groups.
This shows us that which structural elements or portions are more effective under
the earthquake load. Also, the correctness of the model can be checked by comparing
external and internal energies shown in the ECHO file. Energy differences should
not be greater than %5. If Rayleigh damping model is used at the model, contribution
of K dissipated viscous energy can be determined for all elements groups
separately.

105
 Figure 4.27 View of Dissipated Energy Diagram at P3D

There is a small calculation error at P3D, when a model consists of components with
stiffness degradation. P3D does not consider stiffness degradation during energy
calculation. This causes that the elastic displacement part of strain energy is
underestimated and inelastic damage energy is overestimated. This error gets smaller
for dynamic earthquake analysis. To sum up, that small error is not very important
for the right evaluation of the model analysis results.

106
 CHAPTER 5

5 STRUCTURAL WALL AND LINK BEAM CALIBRATON WORK

The key structural elements in resisting lateral forces are structural wall connected
with link beams. In the performance analysis of high-rise buildings equipped with
walls, it is very important that walls and link beams are modelled correctly in
harmony with the horizontal load-displacement cyclic response results from the
experiments. There is no such requirement in Turkish Earthquake Code-2018 about
such a calibration and engineers are free to select any constitutive model, whose
sensitivity requires detailed investigations.

The first issue that needs to be done or verified in the performance analysis of the
structures, is the calibration of the horizontal load-displacement cyclic response with
experimental behavior, called herein “Model Calibration”. The working principle
and the modeling provided in an existing software used in modeling to performance
based nonlinear structural analyses should be verified. This confirmation is carried
out by comparing with results of the experiment and simulations at the simple
structural element level. Cyclic hysterical response curves of structural elements are
complex behavioral patterns that depend on many physical parameters. For this
reason, there are many input values affects these curves according to the working
principle of the programs used in modeling. This is closely related to one of our key
research questions: "How much error occurs in the 3D high-rise building PBD for
various different selection of material parameters?”

The comparisons of experimental and model results in terms of hysterical response

are conducted for a rectangular wall, a T-shaped wall, a U-shaped wall, a diagonally
detailed link beam and conventionally detailed link beam. After providing detailed
information on the experimental studies, modelling, calibration and input
recommendation results are presented in this chapter.

107
5.1 CALIBRATION WORK FOR STRUCTURAL WALLS

Structural walls are widely used in the form of rectangular form or U-shaped for fire
escapes or elevator perimeter walls. In addition, T-shaped wall layouts are common
in buildings depending on architecture. In high-rise buildings, as in our sample
building, there are wall groups, i.e., core wall, where architectural requirements such
as fire escape, elevator, mechanical shaft. Experimental studies have been carried
out for rectangular, T-shaped and U-shaped walls so far. The modelling results of
these three examinations are reflected in core wall in high-rise buildings. In high-
rise buildings, it is common practice to assume walls to work as a group under
moment on core wall, while each wall leg of a group is evaluated separately under
shear force. Since there is still no experiment to examine hysterical behavior of a full
core wall with realistic dimensions, core wall in our high building sample case is
modelled through calibration work results obtained from the wall shapes that have
been experimented with.

5.1.1 Experimental Data for Sample Rectangular and T-Shape Walls

Rectangular and T-shape specimens, RW2 and TW2, which were tested by Thomsen
and Wallace, are used for simulations. Test samples with plan dimensions of 102
x1219 mm and height of 3660 mm are presented in Figure 5.1.

108
 Figure 5.1 Rectangular and T-Shape, i.e. RW2 and TW2, Wall Test Samples
(Orakcal et al., 2004)

Compressive strength of concrete at the test day was 42.7 MPa. There were
longitudinal steel bars, i.e., 8-#3 bars (db=9.53 mm) and #2 (db=6.35 mm) @191
mm, at boundary and web part of wall respectively. While stirrups of the boundary
are placed with db=4.76 mm @ 76 mm, lateral reinforcement of the web of wall were
placed with #2 (db=6.35 mm) @191 mm (Figure 5.2, Figure 5.3). Although #3 bars
were typical Grade 60 (414 MPa), yield stress of #2 and 4.76 mm smooth wires was
approximately 448 MPa. Measured stress-strain profiles of reinforcements are
shown in Figure 5.4.

Figure 5.2 Specimen RW2 (Orakcal et al., 2004)

109
 Figure 5.3 Specimen TW2 (Orakcal et al., 2004)

Figure 5.4 Measured Stress-Strain Relation of Reinforcements of RW2 (Orakcal et

al., 2004)

110
Schematic test setup is shown in Figure 5.5. Firstly, an axial load of 0.07-0.075 *Ag
fc’ was applied continuously by using hydraulic jacks and high-strength post-
tensioning cables. After axial load was applied, displacement controlled reversed
lateral cyclic loads were applied by hydraulic actuator from the top of the specimen.

Figure 5.5 RW2 Test Setup (Thomsen and Wallace, 1995)

A total of 20-22.5 lateral load cycles were applied with a sequence of 0.1%, 0.25%,
0.50%, 0.75%, 1.00%, 1.50%, 1.00%, 1.50%, 2.00%, 2.50% drift levels under
displacement-controlled approach.

Different instruments were placed at critical regions of specimens to measure forces,

displacements, and strains (Figure 5.6). Wire potentiometers were used to measure
horizontal, vertical displacements and shear deformations. Vertical linear variable
differential transducers (LVDTs) were placed at the bottom portion of specimens to
measure axial strains and to calculate section curvatures. Concrete strain gages were
placed to both end of specimen to measure the strains of concrete. Hollow pancake
load cells and hydraulic actuator were used to measure axial load and lateral cyclic
loads respectively.

111
 Figure 5.6 Different Instruments Placed on Specimens (Thomsen and Wallace,
1995)

According to the test results, RW2 and TW2-Lateral Load vs. Top Flexural
Displacement Cyclic Responses are given below in Figure 5.7 and Figure 5.8.

RW2-Cyclic Response Test Result

160
120
Lateral Load, P (kN)

80
40
0
-100 -80 -60 -40 -20 0 20 40 60 80 100
-40
-80
-120
-160

Top Flexural Displacement (mm)

Figure 5.7 RW2-Lateral Load vs. Top Flexural Displacement Cyclic Response
Test Result

112
 TW2-Lateral Load-Top Flexural Displacement Cyclic
Response Test Result
200

100
Lateral Load, P (kN)

0
-80 -60 -40 -20 0 20 40 60 80
-100

-200

-300

-400

Top Flexural Displacement (mm)

Figure 5.8 TW2-Lateral Load vs. Top Flexural Displacement Cyclic Response
Test Result

5.1.2 Experimental Data for Sample U Shape Wall

U-shaped form for shear-walls is widely used in structures such as fire escape,
elevator perimeter walls etc. For this reason, the cyclic behavior of this shear-wall
form under earthquake load is important and was tested by Pegeon et al. at Elsa
Laboratory. This test was conducted within the framework of the "Shear Wall
Structures" of the European Research Programs for the development of Eurocode 8.

Three U-shape wall specimens were used at this cyclic test. These specimens were
separately used for uniaxial cyclic test for X and Y direction, and biaxial cyclic test
in XY direction. Only uniaxial cyclic test results (X and Y direction) were used for
the simulation of U-shape walls. Wall 1 and Wall 2 specimens were used for X and
Y direction cyclic test respectively.

113
The lengths of U-shape walls’ two leg and section depth were 125 cm and 150 cm
respectively. Thickness of the wall was 25 cm. Dimensions of boundaries were 25
cm x 37.5 cm. Longitudinal reinforcements of boundaries were 912 (at corners) and
1012 (others). Web portion longitudinal reinforcements were 10/25. Horizontal
reinforcements were 8/12.5 and 8/7.5 at flanges and web of U-shape wall
respectively. Boundary stirrups were 8/9. All these details are shown in Figure 5.9.

375 750 375

250 250 250

250
 12  10  12
375

 8/75  8/90
250
500

 10  10
1000

 8/125
250
187.5

 8/90
375

187.5

 12  12

250 1000 250

Figure 5.9 U-Shape Wall Specimen Dimension and Reinforcement Detail (Ile &
Reynouard, 2005)

114
Height of the wall was 3.6 m (Figure 5.10). Foundation was clamped down to
laboratory strong floor with 22 anchorages. There was 2 MN continuous vertical
force on wall, which was applied at gravity center of wall with help of six distributed
post-tensioning bars goes through top slab. Uniaxial lateral loads were separately
applied through X and Y direction from on the top of the slab with two pistons with
a displacement-controlled approach.

Figure 5.10 U-Shape Wall Specimen Section View (Ile & Reynouard, 2005)

Average cylindrical strength of the concrete used at Wall-1 and Wall-2 specimens
was 23.73 MPa. Mechanical properties of reinforcements used at Walls are shown
at Table 5.1.

115
 Table 5.1 Mechanical Properties of Reinforcement used at Wall Specimens

Bar Yield strength Ultimate strength Ultimate strain

diameter fsy (MPa) fsu (MPa) su (%)
8 mm 557 642 25.0
10 mm 525 617 24.2
12 mm 516 615 24.8

X and Y direction uniaxial cyclic loading test were realized at very slow rate, i.e.
quasi-static. Loading protocol were applied to failure stage for both direction test
(Figure 5.11).

15.0

10.0
Displacement (cm)

5.0

0.0

-5.0

-10.0

-15.0
0 500 1000 1500 2000 2500 3000

Data points
15.0

10.0
Displacement (cm)

5.0

0.0

-5.0

-10.0

-15.0
0 500 1000 1500 2000 2500 3000
Data points

Figure 5.11 Loading Protocol (Average Displacement vs Data Point) for Wall-1 (X
Direction-Left) and Wall-2 (Y Direction-Right) (Ile & Reynouard, 2005)

116
First cracking of wall base, cracks through the whole wall height, concrete crushing,
buckling of longitudinal reinforcement, rupture of stirrups and longitudinal
reinforcement situations are observed at increasing drift ratios. These observations
are tabulated in Table 5.2 and Table 5.3 for X and Y direction loading below.
Displacement ductility factor of cyclic test for each direction was about 6.

Table 5.2 Wall-1 (X direction loading) Physical Observations through Cyclic Test

Displacement Drift
OBSERVATIONS
(cm) (%)
1 0,3% First inclined cracking at the base
2 0,5% Flange cracks through whole wall height
4 1,0% Web cracks through whole wall height
8 2,1% Bucking of some of flange longitudinal reinforcements
Severe buckling, rupture of some stirrups and
12 3,1%
longitudinal bars at flanges and web ends

Table 5.3 Wall-2 (Y direction loading) Physical Observations through Cyclic Test

Displacement Drift
OBSERVATIONS
(cm) (%)
2 0,5% First inclined cracking at the base
4 1,0% Web cracks through whole wall height
Strong damage at the base of flange ends but wall still
8 2,1%
performed very well
Bar buckling, rupture of some stirrups and
12 3,1%
longitudinal bars at flanges and web ends

Experimental results of cyclic responses of Wall-1 and Wall-2 are given in Figure
5.12 and Figure 5.13 respectively below.

117
 Lateral Load vs. Drift Ratio (%)

1000
800
Lateral Load, P (kN)

600
400
200
0
-4 -3 -2 -1 -200 0 1 2 3 4
-400
-600
-800
-1000
Drift Ratio (%)

Figure 5.12 Cyclic Experimental Response of U-Shape Wall-1 (X Direction

Loading)

Lateral Load vs. Drift Ratio (%)

800
600
Lateral Load, P (kN)

400
200
0
-4 -3 -2 -1 0 1 2 3 4
-200
-400
-600
-800

Drift Ratio (%)

Figure 5.13 Cyclic Experimental Response of U-Shape Wall-2 (Y Direction

Loading)

118
5.1.3 Simulation of Walls Test Samples at Perform 3D V7.0

[Link] General Modeling Approach for Rectangular, T and U Shape

Sample Walls at P3D

Rectangular, T and U-shaped walls in the test wall samples was modelled according
to their geometry and reinforcement details in Perform 3D (Figure 5.14). Foundation
connection of walls was defined as fixed connection. Equal displacement through
laterally was assigned to top nodes of walls (Figure 5.15). Continuous axial load on
top of walls were also assigned (Figure 5.16). Drift directions and drift reference
nodes were separately defined for each sample walls (Figure 5.17). Because of
displacement controlled test approach, all wall samples were defined with horizontal
displacement from the top of walls according to the load protocol in the appropriate
experiment. Horizontal and vertical mesh of wall were made taking into account
through web and boundaries of walls. The mesh sensitivity was also examined later.

Figure 5.14 Rectangular, T and U Shape Test Wall Model Views at P3D

119
Figure 5.15 Views of Top Slaving Assignments of Walls

Figure 5.16 Axial Load Assignment Views of Test Walls

120
Figure 5.17 Views of Drift Reference Node and Direction Assignments of Walls

Wall material models defined with the components of "Inelastic Steel Material, Non-
Buckling", "Inelastic 1D Concrete Material" and "Elastic Shear Material for a Wall".
With "Shear Wall, Inelastic Section" cross section component and the "Fixed Size"
fiber modeling module, concrete and reinforcement fiber definitions are made
according to predefined material properties (Figure 5.18 and Figure 5.19). Details of
the modal parameters are discussed in the next section. After these basic definitions,
wall components are ready for assignment of element identification with the "Shear
Wall Compound Component" section. Also, the displacement loading protocol is
finally provided, then cyclic response model results are obtained to compare with
experimental cyclic response of walls.

121
Figure 5.18 View of Fixed Size Wall Fiber Module for Boundary of RW2 Wall

Figure 5.19 View of Fibers for Boundary of RW2 Wall

122
[Link] P3D Material Parameters and Their Effects on RW2 and TW2
Sample Walls Cyclic Response Results

Concrete and reinforcement material models and the parameters of these models are
of great importance in wall modeling in Perform 3D. The cyclic response results are
mainly influenced by 3 main headings: material modeling parameters, cross-section
fiber definitions and wall meshing. Achieving the ideal modeling method under these
3 main headings is main goal for correct 3D main building wall solutions and design.

Using the cyclic response results of experiments as reference helps us to gel ideal
modeling results. The first P3D wall model with default values for all wall shapes
is accepted as a reference model without thinking the correctness of parameter
values. To understand how much and to what extent change of each main title, i.e.,
material model parameters, cross-section fiber modeling and model mesh, and sub
parameters influences the cyclic response results, only one parameters value is
changed and compared with first reference model results regardless of the accuracy
or inaccuracy of the model result. Wall model mesh is also remained as the same
with reference models. New models are obtained with only single feature or
parameter value changes. The cyclic response results of these new models are
compared with the first reference model each time to observe what changes from
cyclic response behavior. According to these observations, the ideal coefficient and
modeling method are finally obtained to achieve the closest result of the cyclic
response result of experiment.

a) P3D Material Parameters and Their Effects on RW2 Sample Walls’ Cyclic
Response Results

RW2 Wall is modeled with all default material parameter values in Table 5.4 and
reference model result (1. Model) is obtained. Figure 5.20 demonstrates cyclic
response result of the experiment and 1. Model (P3D Reference Model). As shown
Figure 5.20, according to the initial material parameter definitions, there is a
reasonable agreement of the experimental and simulation cyclic response results.

123
The first predicted values of both reinforcement and concrete material parameters
for force displacement curves and for the cyclic degradation factors values that will
affect these curves have been assigned, as shown in Table 5.4. Separate
reinforcement and concrete basic materials definitions are made for the boundary
and web sections of walls according to the experimental information. In cyclic
degradation factor definitions, the first predicted values are separately entered for the
tensile and compression behavior of the reinforcement, while for the concrete
compression cyclic degradation YULRX factor values, 1; 0.4 ; 0.4 ; 0.1 ; 0.1,
recommended by Lowes et al. are used in the initial assignment of the first reference
P3D model of RW2 (Lowes et al., 2016).

RW2 WALL CYCLIC RESPONSE RESULT COMPARISON

Force (kN) vs Displacement (mm)
EXPERIMENT RESULT
P3D REFERENCE MODEL ([Link])
200

150

100

50

0
-100 -50 0 50 100
-50

-100

-150

-200

Figure 5.20 RW2 Wall Cyclic Response Comparison Results of Experiment and
P3D Reference Model (1. Model)

124
 Table 5.4 RW2 Wall P3D Reference Model (1. Model) Material Parameters

MATERIALS
STEEL CONCRETE
Web Boundary Confined Unconfined
All 1 E (MPa) 2E+05 2E+05 31075 31075
2 FY (MPa) 336 395 1.9 1.9
Tension Compr Tens. Compr Tens.
Basic Relationship

Stresses

3 FU (MPa) 550 500 2 2

4 FY (MPa) 448 434 32.62 39.15
.

5 FU (MPa) 550 550 43.8 43.5

6 DU 0.03 0.03 0.0001 0.0001
Strains

7 DX 0.1 0.1 0.0015 0.0015

8 DU 0.03 0.03 0.00263 0.002
.

9 DX 0.1 0.1 0.003 0.02

10 DL 0.07 0.07 0.00012 0.00012
11 DR 0.08 0.08 0.001 0.001
Strength Loss

Strains

12 FR/FU 0.1 0.1 0.001 0.001

13 DL 0.07 0.07 0.0027 0.00202
Compr.

14 DR 0.08 0.08 0.0268 0.01

15 FR/FU 0.1 0.1 0.185 0.001
16 TSLX No No --- ---
All
17 SLI 0 0 --- ---
18 Y 1 1 --- ---
Energy Factors
Tension Strains

19 U 0.85 0.85 --- ---

20 L 0.85 0.85 --- ---
Cyclic Degradation

21 R 0.85 0.85 --- ---

22 X 0.85 0.85 --- ---
23 Y 0.9 0.9 1 1
Energy Factors
Compr. Strains

24 U 0.75 0.75 0.4 0.4

25 L 0.75 0.75 0.4 0.4
26 R 0.75 0.75 0.1 0.1
27 X 0.75 0.75 0.1 0.1
All 28 USF 0.5 0.5 --- ---
TSLX: Total Strength Loss at Point X; SLI: Strength Loss Interaction; USF: Unloading Stiffness Factor;
Y:Yielding; U:Ultimate; L:Ductile Limit; R:Residual Limit; X:Analysis Stop Limit

125
RW2 Wall hysterical response solution trials have been conducted with 34 different
P3D models. In the solution of many of the models, only one parameter is changed
from the 1st reference model, and it is tried to observe the extent to which this
parameter affects the cyclic response result and how much sensitivity it affects.
While investigating the effect of parameter values on cyclic response result, it is tried
to reach the correct result by comparing with the 1st model as well as by comparing
with other model results. Table 5.6 and Table 5.7 describe subject that is the
difference for each model from the first model and the degree of sensitivity that this
change creates on the cyclic response. All representations between Figure 5.21 and
Figure 5.24 provide cyclic response comparisons between models.

Various observations are made between Model-2 and Model-34 with different
material parameters, properties and modeling type changes (Table 5.6, Table 5.7,
and Table 5.8). Models 31, 32, 33 and 34 are some ideal optional modeling results
with material parameters that is as a result of the first 30 model comparison studies.
In the first 30 models, material parameters are varied to observe model sensitivity,
so different sensitivity levels of high, moderate, low and none are selected based on
the ratio of area under the cyclic response curve to area under the cyclic response
curve of the reference model. Sensitivity level definitions are shown in Table 5.5. It
is observed that energy factors of tension strains of reinforcement and unloading
stiffness factors of reinforcement have very high effects on results. Another effective
material parameters are yielding and ultimate strength values of reinforcement under
tension loading behavior with moderate sensitivity level. All remaining parameters,
in other words, material parameter changes that belong to concrete in general, have
low effects on the cyclic response results. Summary results of sensitivity of material
parameters on cyclic response is shown in Table 5.9. In addition, the last ideal
suggestion coefficient values appear in Table 5.10 according to the comparison of
simulation that fit the best with the test results.

126
Table 5.5 Sensitivity Level Rates (Change Rate in Area or Boundary Value of
Cyclic Response)

Change Rate Change Rate

Sensitivity Level
Top Boundary Bottom Boundary
%100> High >15%
%15 > Moderate >5%
%5 > Low >2%
%2 > None >0%

127
Table 5.6 P3D Models Differences from Reference RW2 Wall Model (Model-1)

Model # CHANGE FROM REFERENCE MODEL-1 SENSITIVITY

Model-1 Reference P3D Model --
Tension strength model of confined concrete material
Model-2 None
is omitted
Tension strength model of both confined and
Model-3 Low
unconfined concrete material is omitted
Unconfined part of boundary cross-section is not
Model-4 Low
modelled (this situation is valid from Model 5 to19)

Model-5 Both criteria of Model-3 and Model-4 is considered Low

Unloading stiffness factor "USF" is taken as "+1",
Model-6 Moderate
instead of "+0.5"
Unloading stiffness factor "USF" is taken as "-1",
Model-7 Moderate
instead of "+0.5"
Unloading stiffness factor "USF" is taken as "0",
Model-8 Low
instead of "+0.5"
YX+3 model is used with different energy factors
values for steel material. Energy factors of tension
steel strains are 1, 0.95, 0.9, 0.85, 0.8 for Y, 0.01, 0.02,
Model-9 Low
0.03, X strain stages respectively, and energy factors
of compression steel strains are 0.9, 0.85, 0.8, 0.75, 0.7
for Y, 0.01, 0.02, 0.03, X strain stages respectively
ULRX energy factors of tension & compression
Model-
strains of steel material model are 0.5 decreased from High
10
0.85.
Model- ULRX energy factors for compression strains of steel
Low
11 material model are decreased from 0.75 to 0.5

Model- ULRX energy factors of tension strains of steel

High
12 material model are decreased from 0.85 to 0.5
ULRX energy factors of tension strains of steel
Model- material model are decreased from 0.85 to 0.5 and
High
13 unloading stiffness factor "USF" is changed from
"0.5" to "-1"
YULRX energy factors of tension strains of steel
Model-
material model are changed as Y:1 U:0.85 L:0.85 None
14
R:0.5 X:0.5

128
Table 5.7 P3D Models Differences from Reference RW2 Wall Model (Model-1)

Model # DIFFERENCE FROM REFERENCE MODEL-1 SENSITIVITY

YULRX energy factors of tension strains of steel
Model-15 material model are changed as Y:1 U:0.85 L:0.5 R:0.5 Low
X:0.5
YULRX energy factors of tension strains of steel
Model-16 material model are changed as Y:0.5 U:0.5 L:0.5 R:0.5 High
X:0.5
Tension yielding stress of steel material of boundary is
Model-17 Low
decreased as an amount of 45 MPa
Tension yielding stress of steel material of web and
Model-18 Low
boundary is decreased as an amount of 45 MPa
Tension ultimate stress of steel material of web is
decreased from 550 MPa to 400 MPa and tension
Model-19 Moderate
ultimate stress of steel material of boundary is
decreased from 500 MPa to 400 MPa
YULRX energy factors of tension strains of steel
Model-20 material model are changed as Y:1 U:0.5 L:0.5 R:0.5 High
X:0.5
Compression yielding stress of steel material of web is
decreased from 434 MPa to 334 MPa and compression
Model-21 Moderate
yielding stress of steel material of boundary is
decreased from 438 MPa to 338 MPa
Tension yielding stress of steel material of web is
decreased from 336 MPa to 250 MPa and tension
Model-22 Moderate
yielding stress of steel material of boundary is
decreased from 395 MPa to 300 MPa
Confined concrete compression strength is increased
Model-23 43 MPa to 50 MPa and confined concrete None
compression strain is increased 0.00263 to 0.003
Confined and unconfined concrete compression
strength is increased to 50 MPa and also confined
Model-24 None
concrete compression strain is increased 0.00263 to
0.003
Energy factors of compression strains of confined and
Model-25 unconfined concrete are changed as Y:1, U:0.8, L:0.8, None
R:0.8, X:0.1

129
Table 5.8 P3D Models Differences from Reference RW2 Wall Model (Model-1)

Model
DIFFERENCE from REFERENCE MODEL-1 SENSITIVITY
#
Energy factors of compression strains of confined and
Model-
unconfined concrete are changed as Y:1, U:0.1, L:0.1, None
26
R:0.1, X:0.1
Energy factors of compression strains of confined and
Model-
unconfined concrete are changed as Y:0.1, U:0.1, L:0.1, None
27
R:0.1, X:0.1
Energy factors of compression strains of confined and
Model-
unconfined concrete are changed as Y:0.8, U:0.8, L:0.8, None
28
R:0.8, X:0.0.8
Model-
Cyclic degradation option of concrete is taken as "none". None
29
Ultimate compression strain of confined concrete is
Model-
decreased to unconfined ultimate compression strain Low
30
value
Tension yielding stress of steel material of web and
boundary is decreased as an amount of 45 MPa (0.85 fy),
Ideal-(The
Model- and YULRX energy factors of tension strains of steel
best fitted to
31 material model are changed as Y:0.65 U:0.65 L:0.65
Test)
R:0.65 X:0.65 and unloading stiffness factor "USF" is
changed from "0.5" to "-0.5"
Tension yielding stress of steel material of web and
boundary is decreased as an amount of 45 MPa (0.85 fy),
and YULRX energy factors of tension strains of steel Ideal-(The
Model-
material model are changed as Y:0.65 U:0.65 L:0.65 best fitted to
32
R:0.65 X:0.65 and unloading stiffness factor "USF" is Test)
changed from "0.5" to "-0.5" and unconfined part of
boundary cross-section is not modelled
Tension yielding stress of steel material of web and
boundary is decreased as an amount of 45 MPa (0.85
fy), and YULRX energy factors of tension strains of steel
material model are changed as Y:0.65 U:0.65 L:0.65
Ideal-(The
Model- R:0.65 X:0.65 and Unloading Stiffness Factor "USF" is
best fitted to
33 changed from "0.5" to "-0.5" and unconfined part of
Test)
boundary cross-section is not modelled, symmetry
model is used for steel compression and tension stress
models, tension strength of concrete is selected as "No",
cyclic degradation of concrete is selected as "none",
All materials properties are similar with Model-31. Ideal-(The
Model-
Cross-section definitions of web and boundaries are used best fitted to
34
as "auto size module" instead of "fixed size module". Test)

130
 Table 5.9 RW2 Wall P3D Model Material Parameters Sensitivity Summary
Results

MATERIALS
STEEL CONCRETE
Web Boundary Confined Unconfined
All 1 E (MPa) -- -- -- --
Tension Comp Tens. Comp Tens.

2 FY (MPa) Low Low None None

Basic Relationship
Stresses

3 FU (MPa) Moderate Moderate None None

4 FY (MPa) Low Low None None
r.

5 FU (MPa) Low Low None None

6 DU -- -- None None
Strains

7 DX -- -- None None
8 DU -- -- Moderate None
r.

9 DX -- -- None None
10 DL -- -- None None
11 DR -- -- None None
Strength Loss
Strains

12 FR/FU -- -- --- ---

13 DL -- -- None None
Compr.

14 DR -- -- None None
15 FR/FU -- -- --- ---
16 TSLX -- -- --- ---
All
17 SLI -- -- --- ---
Energy Factors Energy Factors

18 Y Moderate Moderate --- ---

Compr. Strains Tension Strains

19 U High High --- ---

20 L Low Low --- ---
Cyclic Degradation

21 R None None --- ---

22 X None None --- ---
23 Y Low Low None None
24 U Low Low None None
25 L Low Low None None
26 R Low Low None None
27 X Low Low None None
All 28 USF Moderate Moderate --- ---
TSLX: Total Strength Loss at Point X; SLI: Strength Loss Interaction; USF: Unloading Stiffness Factor;
Y:Yielding; U:Ultimate; L:Ductile Limit; R:Residual Limit; X:Analysis Stop Limit

131
 Table 5.10 RW2 Wall P3D Ideal Model Material Parameters (Model-31)

MATERIALS
STEEL CONCRETE
Web Boundary Confined Unconfined
All 1 E (MPa) 2E+05 2E+05 31075 31075
2 FY (MPa) 280 350 1.9 1.9
Tension Compr Tens. Compr Tens.
Basic Relationship

Stresses

3 FU (MPa) 550 500 2 2

4 FY (MPa) 448 434 32.62 39.15
.

5 FU (MPa) 550 550 43.8 43.5

6 DU 0.03 0.03 0.0001 0.0001
Strains

7 DX 0.1 0.1 0.0015 0.0015

8 DU 0.03 0.03 0.00263 0.002
.

9 DX 0.1 0.1 0.003 0.02

10 DL 0.07 0.07 0.00012 0.00012
11 DR 0.08 0.08 0.001 0.001
Strength Loss

Strains

12 FR/FU 0.1 0.1 0.001 0.001

13 DL 0.07 0.07 0.0027 0.00202
Compr.

14 DR 0.08 0.08 0.0268 0.01

15 FR/FU 0.1 0.1 0.185 0.001
16 TSLX No No --- ---
All
17 SLI 0 0 --- ---
18 Y 0.65 0.65 --- ---
Energy Factors
Tension Strains

19 U 0.65 0.65 --- ---

20 L 0.65 0.65 --- ---
Cyclic Degradation

21 R 0.65 0.65 --- ---

22 X 0.65 0.65 --- ---
23 Y 0.9 0.9 1 1
Energy Factors
Compr. Strains

24 U 0.75 0.75 0.4 0.4

25 L 0.75 0.75 0.4 0.4
26 R 0.75 0.75 0.1 0.1
27 X 0.75 0.75 0.1 0.1
All 28 USF -0.5 -0.5 --- ---
TSLX: Total Strength Loss at Point X; SLI: Strength Loss Interaction; USF: Unloading Stiffness Factor;
Y:Yielding; U:Ultimate; L:Ductile Limit; R:Residual Limit; X:Analysis Stop Limit

132
It is observed that there is no need to model the tensile strength of confined and
unconfined concrete material for wall elements. In addition, the modeling of
unconfined concrete cross sections in the wall boundary part does not change the
result much. This can be observed among comparisons b/w comparison 1 (C1) and
4 (C4) in the Figure 5.21. However, unloading stiffness factor, “USF” for
reinforcement material has a high effect on stiffness in general cyclic response
behavior. This can also be observed among comparisons b/w C5 and C7 in the Figure
5.21. Unloading stiffness factor value, “USF” ranging from "-1" to "+1" is
recommended as "-0.5" (Table 5.10). In addition, for reinforcement, the energy
factor, “EF” is one of the parameters that most influences cyclic response behavior
that can be seen b/w C9 and C15 at Figure 5.22, but concrete material energy factor
parameters have little effect on the results. This can be observed between
comparisons. Among energy factors levels, ultimate level value, “U”, changes the
result with a very dominant effect. After the ultimate limit energy factor "U",
yielding "Y" and ductile limit "L" have a more partial effect on the result. Energy
factor values "R" and "X" levels have almost no effect. "0.65" is recommended for
YULRX energy factor inputs (Table 5.10). According to Pacific Engineering
Research Center, YULRX energy factors are recommended as a value of 0.6-0.7
(Moehle et al., 2011). Also, it is observed in comparison C16, C17 and C21 that
reducing reinforcement tension yielding strength by 50-100 MPa decreases the
yielding force result in cyclic response by almost 10%, while reducing reinforcement
ultimate strength value under tension by 100-150 MPa causes decrease at ultimate
force limit by almost 10%-15% (Figure 5.23). As can be seen in the comparison 20,
(C20), reducing compression yielding strength by around 100 MPa increases the
pinching behavior of cyclic response behavior (Figure 5.23). There is a little effect
on cyclic response with change at parameters related to concrete material. As it
appears between 22. and 29. comparisons (C22 and C29) reducing or increasing the
compression strain and strength of concrete has little effect on the cyclic response
result. In addition, in these comparisons, it is seen that the value of energy factors
defined for concrete has little effect on the results.

133
Figure 5.21 Cyclic Response Comparisons b/w C1-C8 for RW2 P3D Models

134
Figure 5.22 Cyclic Response Comparisons b/w C9-C16 for RW2 P3D Models

135
Figure 5.23 Cyclic Response Comparisons b/w C17-C24 for RW2 P3D Models

136
Figure 5.24 Cyclic Response Comparisons b/w C25-C32 for RW2 P3D Models

The results of the first 30 models and 29 comparisons are observed and applied at
31. 32. 33. and [Link] models. There is a simplification in terms of modeling and

137
parameters through ideal Model-31 to ideal Model-33. The ideal model material
proposal parameter values are presented in Table 5.10 for Model-31. The important
another issue worth mentioned about that it has been observed that the yielding force
limit in P3D cyclic response results is above from that of experiment results. For this
reason, it is proposed that Belarbi and Hsu (1994) recommendation should be taken
into account in reinforcement model. If the average stress -strain relationship for
reinforcing bars embedded in concrete recommended by Belarbi and Hsu (1994) is
used in the reinforcement model, the reinforcement yielding strength is decreased by
about %10-%15 and P3D model results will be closer to the result of the experiment.
For this reason, the reinforcement yielding strength shown in Table 5.4 in our first
reference model is higher than the reinforcement yielding strength offered in Table
5.10 for Model-31. According to the reinforcement model proposed by Belarbi and
Hsu (1994), yielding stress of embedded bars in concrete is lower than bare steel bars
yielding stress. In other words, tension stiffening on reinforcement causes lowering
of yielding stress.

b) P3D Material Parameters and Their Effects on TW2 Sample Walls’ Cyclic
Response Results

Hysteric response study is performed for the T-shaped wall according to the
recommended material parameters obtained from rectangular wall work. The
reference model cyclic response results for T-shaped wall fit relatively well with
experiment results (Figure 5.26). Trials on similar subjects in the RW2 model are
carried out in a T-shaped walls in terms of cyclic response. For this purpose, 11
different trial models have been created. All comparison results are presented in
Figure 5.27 and Figure 5.28. Description details and sensitivity results for these trial
models are presented in Table 5.11. The sensitivity of the material parameters does
not differ from that of the work carried out for the rectangular walls, and similar
results are obtained in general. Firstly, removal of the tensile strength of the concrete
from the model has little effect on hysteric results (C1 and C2 in Figure 5.27). The
unloading stiffness factor, USF, is again dominant effect on results for T-shaped
wall-TW2 like rectangular wall-RW2 (C3, C4 in Figure 5.27). The reduction of 100

138
MPa in yielding and ultimate strength of reinforcement does not make a huge
difference to the behavior of the first reference model. In this regard, the 10-15%
capacity change, moderate effect, in rectangular walls is not be observed for T-
shaped wall works. This can be observed in comparisons, C5, C6 and C7 in Figure
5.27. However, the energy factor change for reinforcement under tension
significantly affects cyclic response behavior for T-shape walls like rectangular walls
(C8, C9, C10, C11 in Figure 5.27 and Figure 5.28).

Figure 5.25 TW2 Wall Cyclic Response Comparison Results of Experiment and
P3D TW2 Reference Model (1. Model)

139
Table 5.11 P3D Models Differences from Reference TW2 Wall Model (1. Model)
for Sensitivity

Model # CHANGE FROM REFERENCE 1. MODEL SENSIVITY

TModel- Ideal-(The best
Reference P3D Model
1 fitted to Test)
TModel- Tension strength model of both confined and unconfined
None
2 concrete material is omitted
Tension strength model of both confined and unconfined
TModel-
concrete material is omitted and Moderate
3
unconfined part of boundary cross-section is not modelled
TModel- Unloading stiffness factor "USF" is taken as "0", instead of
Low
4 "-0.5"
TModel- Unloading stiffness factor "USF" is taken as "+0.5", instead
Moderate
5 of "-0.5"
TModel- Confined concrete compression strain is decreased from
6 0.00263 to 0.002
None
Tension yielding stress of steel material of web is taken as
original value as 336 MPa instead of decreased value,250
TModel-
MPa and tension yielding stress of steel material of boundary Low
7
is taken as original value as 395 MPa instead of decreased
value,300 MPa
Tension ultimate stress of steel material of web is decreased
TModel- from 550 MPa to 500 MPa and tension ultimate stress of
Low
8 steel material of boundary is decreased from 500 MPa to 450
MPa
TModel- YULRX energy factors of tension strains of steel material
High
9 model are changed as Y:1 U:0.1 L:0.65 R:0.65 X:0.65
TModel- YULRX energy factors of tension strains of steel material
High
10 model are changed as Y:0.65 U:0.5 L:0.5 R:0.5 X:0.5
TModel- YULRX energy factors of tension strains of steel material
High
11 model are changed as Y:1 U:0.8 L:0.65 R:0.65 X:0.65
TModel- YULRX energy factors of tension strains of steel material
High
12 model are changed as Y:0.8 U:0.8 L:0.65 R:0.65 X:0.66
Tension yielding stress of steel material of web and boundary
is decreased as an amount of 45 MPa (0.85 fy), and YULRX
energy factors of tension strains of steel material model are
changed as Y:0.65 U:0.65 L:0.65 R:0.65 X:0.65 and
TModel- unloading stiffness factor "USF" is changed from "0.5" to "- Ideal-(The best
13 0.5" and unconfined part of boundary cross-section is not fitted to Test)
modelled, symmetry model is used for steel compression and
tension stress models, tension strength of concrete is selected
as "No", cyclic degradation of concrete is selected as "none",
unconfined concrete model parts are cancelled.
All materials properties are similar with TModel-1. Cross-
TModel- Ideal-(The best
section definitions of web and boundaries are used as "auto
14 fitted to Test)
size module" instead of "fixed size module".

140
Figure 5.26 Cyclic Response Comparisons b/w C1-C8 for TW2 P3D Models

141
Figure 5.27 Cyclic Response Comparisons b/w C9-C12 for TW2 P3D Models

[Link] Cyclic Response Results with Ideal Model Material Parameters for
Walls

a) RW2

Model-31 is found to be the most compatible with the test result for RW2 wall as a
result of modeling according to the ideal material model parameters obtained (Figure
5.28). According to the calibration work of RW2, P3D model results showed good
compatibility with experiment results in terms of hysteretic shape and at first glance,
energy absorption capacities are almost similar. In addition, there is a very good
agreement in terms of lateral load capacity, lateral displacement capacity, stiffness
degradation properties. Although agreement is not perfect in terms of pinching
behavior, lateral yielding capacity and plastic displacement, it is reasonable in
general. The fact of the lateral yielding force capacity being higher than the test result
may be explained by the insufficient simulation of yielding behavior of the
reinforcement embedded in the concrete.

142
 Figure 5.28 RW2, Most Ideal P3D Cyclic Response Result

After obtaining the ideal results with Model-31, in order to increase computer
analysis speed and facilitate modeling (especially for use in computationally
demanding tall building analysis), new ideal model works have been conducted i.e.,
Model-32 and Model-33, by disabling material parameters with a low sensitivity
effect. Firstly, unconfined part has been omitted in the modeling of Model-32.
Almost 5% decrease in the lateral load capacity of the wall was observed as a result
of this cancellation (Figure 5.29). In addition to the change in Model-32, the
reinforcement material is modelled as symmetrically for both tension and
compression, concrete tensile strength and energy factors of concrete are not taken
into account for Model-33 (Table 5.8). With Model-33, pinching behavior of the
cyclic response has increased compared to the experiment result and energy
absorption area is even reduced. Also, lateral ultimate load capacity, shown in
Model-33, reduced, as in the Model-32. Material modeling approach at Model-33
can be considered as fast and acceptable in complex modeling of core wall in 3D
dimensional structure analysis of high-rise buildings.

143
The lateral flexural displacement profile prepared on each floor and for different drift
ratios for Model-31 can be observed at Figure 5.30. Both the experiment and P3D
model lateral flexural displacement profile results seem to be very compatible on
each floor, along the wall height and for different drift ratios.

In addition to hysteric shape and lateral flexural displacement profile comparisons

to confirm the accuracy of modeling, the experimental and simulation of strain
profiles of wall at basement level is also conducted. Although the stain gage lengths
of experiment (229 mm) and PBD Model (457 mm) are different (it should be the
same), strain comparisons results are again given in Figure 5.31 and Figure 5.32. In
terms of different model strain gage length comparison investigation, more detailed
investigation is presented later. But for these strain gage lengths, i.e., 229 mm and
457 mm, experiment and P3D Model strain comparison also results in very good
compatibility (Figure 5.31 and Figure 5.32). These average strain profiles for both
concrete and steel are compatible with the test results. In addition, location of neutral
axis is also successful in this compatibility. For both concrete and reinforcement, the
tensile strains remain on the safe side according to the test results, however
compression strains for all drift ratios are below the measured compression strain
results. In terms of crushing strain values of concrete, it cannot be said that
successful modeling results are achieved due to the fact that lower strain values are
obtained with P3D Model from experiment values. The cause of the larger
compressive strain may be stress concentration and additional nonlinear shear
behavior due to geometric reasons. On the other hand, it is observed that the results
of tension strain of reinforcement in P3D Model are very compatible with the
experiment.

144
Figure 5.29 RW2, Close to Ideal P3D Cyclic Response Results for Rapid Analysis

145
Figure 5.30 RW2 Lateral Displacement Profiles of Experiment and P3D Model
(Model-31) for Different Drift Ratio

Figure 5.31 RW2 Wall, Base Level Concrete Strain Profiles of Experiment and
P3D Model (Model-31) for Different Drift Ratios

146
 Figure 5.32 RW2 Wall, Base Level Steel Strain Profiles of Experiment and P3D
Model (Model-31) for Different Drift Ratios

b) TW2

The ideal material parameters obtained as a result of the work on the RW2 sample
work models are also used in TModel-1 for the T-shaped wall as a first trial. With
various changes in material parameters, there is no need to change the ideal results
obtained from RW2 model trials. Figure 5.33 shows a comparison of the test's cyclic
response result with the P3D Model generated according to the ideal material
parameter assignments for the TW2 wall. Although the overlap result in TW2 is not
as good as like in RW2, it is reasonably acceptable. In positive displacement, the
wall flange is under compression, while the wall flange is under tension in the
negative displacement situation. In case of positive displacement, horizontal plastic
displacement capacity is less than experimental value, while in case of negative
displacement, experimental lateral load capacity and lateral load displacement
capacity are less than P3D Model results (Figure 5.33). It may be said that acceptable

147
good results have been achieved in energy absorbing areas, stiffness degradation,
and general hysteric shape.

Figure 5.33 TW2, Ideal P3D Cyclic Response Result

As with RW2, for T-Shaped wall, TW2, experiment is compared with P3D Model
having omitted parameters that are not very effective, simplified modeling
conditions (Figure 5.34). Simplified model (TModel-13) achieves more pinching
behavior, while the energy damping area is even smaller. For simpler and faster
analysis, it is relatively acceptable to perform wall element modeling in accordance
with the Table 5.10, TModel-13 definition.

148
Figure 5.34 TW2, Close to Ideal P3D Cyclic Response Result for Rapid Analysis

Figure 5.35 TW2 Lateral Displacement Profiles of Experiment and P3D Model for
Different Drift Ratio

149
Model and test results comparisons are checked in terms of lateral displacement
profile and reinforcement and concrete strain profile for TW2, T-Shaped wall. The
displacement profile comparison is acceptable for both positive drift and negative
drift ratios. This situation can be observed in Figure 5.35 through wall height and for
different drift ratios.

TW2 wall simulation and experimental results for concrete and steel strain profile
comparisons for different strain gage lengths 457 mm and 229 mm respectively are
presented in Figure 5.36 and Figure 5.37. In both figures, concrete and reinforcement
strain profiles are presented separately for positive and negative displacement.
According to Figure 5.36, in case of positive and negative displacement or drifts, the
concrete tensile strain values are in good condition, and overestimated as observed
in the RW2 wall sample strain profile, while concrete compression strain values on
P3D model are lower than the test results. In other word, for compression strain
values, outcomes are underestimated. This situation applies to all drift ratios. On the
other hand, the profiles may be generally successful in compatibility in terms of
neutral axis location and general average line views.

Steel strain profiles of test and P3D model comparison results are presented in Figure
5.37 for both positive and negative displacements. In the case of positive
displacement where the flange of T-wall is under compression, steel tension strain
values of P3D model are above the test results for drift values of 0.5% and 1.0%,
while partially below for the 2.0% drift ratio. On the other hand, in the case of
negative displacement where the flange is under tension, steel tension strain values
remain on the safe side and are higher than the test results for all drift values.

To sum up, for the TW2 wall sample, except perfectly agreement of strain profiles
and concrete compression strain values, other global variables are successfully
obtained by P3D modeling.

150
Figure 5.36 TW2 Wall, Base Level Concrete Strain Profiles of Experiment and
P3D Model (TModel-1) for Different Drift Ratios

151
Figure 5.37 TW2 Wall, Base Level Steel Strain Profiles of Experiment and P3D
Model (TModel-1) for Different Drift Ratios

152
 c) U-Shaped Walls, Loaded Symmetric and Asymmetrically

Necessary comparisons have been made and evaluated for hysteric shape, lateral
displacement profile and basement level strain profile results with the cyclic
response modeling study on RW2 and TW2 wall samples. While very good results
are obtained for the rectangular wall in the comparisons in the mentioned subjects,
relatively better results are obtained for the T-shape wall than the rectangular wall
results. The ideal modeling parameters of both sample walls are presented in Table
5.10. The conclusion of results that these material parameters are suitable and
acceptable, is reached as a result of the model trial of different variations for both
RW2 and TW2 walls. These ideal material parameters are used in the modeling for
U-Shape wall. Displacement controlled force-drift ratio curves for both X and Y
direction are obtained as a result of comparison with the test and P3D model results.
X and Y direction force-drift ratio curves are shown in Figure 5.38.

As the most basic comparison, model and experimental results successfully agree in
terms of hysteretic shape achieved in RW2 and TW2 but not to the degree
sufficiently achieved for U-shape wall. Y-direction force-displacement curve
compatibility is better than X-direction one. While wall behavior in the Y direction
is more symmetrical and identifiable, the X-direction displacement profile of wall is
more eccentric and complex. So, one of the reasons for increase in incompatibility
can be interpreted in this way. From a simple structural wall shape such as a rectangle
to a slightly more complex T-shape wall, and ultimately to the most complex U-
shaped wall, after experiment and P3D model hysteric shape comparisons are
explored, it may be concluded that there is a need for more investigation in terms of
modeling of U-shaped wall force-displacement behavior. It seems that U-shaped
wall work examined by Pegon et al. in Elsa Laboratory is not very detailed as the
investigation done for rectangular and T-shaped wall by Orakcal K. The reasons for
hysteric shape incompatibility in P3D model and experiment comparison for U-
shaped wall may be inadequate modeling information on the material properties
especially steel yield strength. For more complex wall shapes, such as the U and I
curtains, it is clear that more experimental studies and more detailed ones are needed.

153
Figure 5.38 UX and UY, P3D Cyclic Response Result with Ideal Material
Parameters

154
[Link] Mesh Sensitivity

In addition to accurately identifying material parameters in order to capture the

correct behavior and results by modeling wall elements of structures with the
Perform-3D program, it is important to make wall cross-section definitions and the
number of wall element mesh correctly. The correct and efficient result will be
obtained by getting an ideal modeling approach as a whole for accurate material
parameters and together with optimum cross-section and mesh modeling. For this
reason, necessary studies are carried out on wall cross-section modeling and mesh,
and validation studies are fulfilled with the results of the experiment. Besides to
working on optimum fiber numbers in cross-section definition, two different cross-
section identification modes available in P3D, i.e., the "Fixed size" and "Auto size"
modules, are compared. In addition to cross-section modeling, another work is also
done on ideal wall mesh work and strain gage length.

a) Rectangular Wall, RW2

The optimum selection of the reinforcement and concrete fiber number in the cross-
section of walls element is important for the rapid analysis of complex 3D structure.
It is necessary to model the reinforcements, which usually settle at intervals of 15-
20 cm through web portions of walls and more frequent in the wall boundary
sections, and concrete fibers corresponding to them at ideal intervals. Modeling each
reinforcement as fiber individually will greatly increase analysis time in complex
high-rise structures. For this reason, different number of fiber modeling options is
tried for the RW2 wall in order to look at the way to simplify the model so as not to
affect the result too much (Figure 5.39). Accordingly, Model-31 of RW2, that is a
very detailed model, is simplified and 7 new trial models is created from Model-31.
With these trial models, differences and sensitivity between models are indicated in
Table 5.12. In addition, hysteric shape comparisons of RW2 models having different
fiber cross section modeling are presented in Figure 5.40.

155
Figure 5.39 Different P3D Fiber Modeling Types for RW2

156
Table 5.12 Sensitivity Evaluation in Terms of Different Fiber Modelling for RW2
P3D Models

DIRFFERENCE FROM REFERENCE 31. MODEL

Model # SENSITIVITY
IN TERMS OF FIBERS
Model-31 Reference P3D Model (Model-31) --
4 reinforcement fibers of boundary are decreased to
GModel-1 Low
2 reinforcement fibers
4 reinforcement fibers of boundary are decreased to
GModel-2 High
1 reinforcement fibers
2 concrete fibers of confined concrete of boundaries
GModel-3 Low
are decreased to 1 concrete fiber
2 concrete fibers of confined and unconfined
GModel-4 concrete of boundaries are decreased to 1 concrete Low
fiber
Web part of wall is modelled with 4 concrete fibers
GModel-5 None
instead of 8 fibers
Web part of wall is modelled with 2 reinforcement
GModel-6 None
fibers instead of 4 fibers
4 reinforcement fibers of boundary are decreased to
2 reinforcement fibers and web part of wall is
GModel-7 Low
modelled with 4 concrete and 4 reinforcement
fibers.

As a result of comparisons with different fiber modeling options, it can be concluded

that, first of all, it is more important that the reinforcement should be modeled
accurately compared to the concrete fibers. The fact that the number of concrete
fibers has been reduced does not have much effect. In addition, it is observed that
unconfined sections in boundary parts of wall do not need to be modeled
additionally. Modeling reinforcement in boundary sections under high tension load
according to the 1x1 aspect ratio is ideal, as seen with the GModel-7 (Figure 5.39)
and GC7 (Figure 5.40). As can be observed in GModel-2 and GC2, inaccurate results
are obtained with the insufficient number of tensile reinforcement fibers in boundary
sections under high tension load.

157
Figure 5.40 Cyclic Response Comparisons b/w GC1-GC7 in Terms of Different
Fiber Modelling for RW2-P3D Models

158
b) P3D "Fixed Size" and "Auto Size" Modeling Option Comparison for RW2
and TW2

Perform 3D program offers 2 options for cross section fiber modeling of wall section,
i.e., "Fixed size" and "Auto-size". “Fixed-size” module supplies a slower and less
practical but more detailed modeling option, while “Auto-size” provides a more
limited but faster modeling option (Figure 5.41). In the solution of large building
systems, P3D users often prefer "Auto-size" option for fast and practical modeling.
There was a need for more detailed modeling for work to compare RW2, TW2 and
U-shaped sample walls with experiment results, so, "Fixed-size" model option was
used for investigations before.

Figure 5.41 P3D Two Fiber Modeling Options, “Fixed Size” & “Auto Size” for
Walls

It is questioned that how much difference in results occurred when RW2 and TW2
sample walls are modeled with the "auto-size" module, which engineers could
commonly use in modeling. The cross-section fiber appearance of the RW2 and TW2
sample walls in the "Fixed size" and "Auto-size" modules is given in Figure 5.42.

159
Figure 5.42 “Fixed Size” and “Auto-Size” P3D Modeling Options of RW2 and TW2

The results of the comparison of these two modules are examined on the basis of
both hysteric shape and strain profile topics for RW2 wall sample solutions (Figure
5.43 and Figure 5.45). However, In TW2 wall sample solution, two fiber modeling
shapes are compared only in terms of hysteric shape. In RW2 rectangular wall cross-
section fiber modeling, the result of both modules in terms of hysteric shape is almost
the same (Figure 5.43). In terms of strain profile match, modeling with the "Auto-
size" fiber module causes a strain increase of 5-10% compared to modeling with the
"Fixed-size" module, leaving the resulting evaluation on the safe side (Figure 5.45).
While the effect of "Auto-size" module on TW2 wall sample makes no difference in
the case of positive displacement, it causes an increase of up to 10% after a certain
drift ratio in the lateral load capacity for negative displacement where the flange is
under tension (Figure 5.44). Since usual building designs do not check capacity or
strain control at high drift ratios, both modules can be seemed similar in T-shape for
lower drift ratios.

160
Figure 5.43 “Fixed Size” and “Auto Size” Modeling Option Cyclic Response
Result of RW2

Figure 5.44 “Fixed Size” and “Auto Size” Modeling Option Cyclic Response
Result of TW2

161
Figure 5.45 Concrete and Steel Strain Profiles of RW2, Wall for “Fixed Size” and
“Auto Size” Option

162
c) Different Mesh and Gage Length Option for Cyclic Response Comparison

Using the minimum number of elements and fibers with providing real structural
behavior in model will reduce analysis time and margin of error. Therefore, walls
should be modeled as simple as possible.

The amount of mesh of the walls vertically and horizontally should be kept at the
optimum level, so that both accurately reflecting the structure behavior and
shortening the analysis time. In high-rise buildings, it may be enough to model one
wall element on each floor in a vertical direction. Curvature, axial and shear strains
are constant throughout each wall mesh element. For this reason, the number of
elements can be increased or selected a reasonable height in the hinge section, which
has plastic behavior for the walls.

Appropriate inelastic hinge section height is very important at modeling of walls

because it affects the calculated strain and bending moment capacities. The correct
height selection of elements representing hinge length is the very sensitive parameter
that affects the accuracy of the calculated strain. According to ASCE 41, the hinge
length of walls is minimum of one half of cross section depth and story height. In
Turkish earthquake regulation, it is considered to use Equation 5.1.

𝐿𝑔 = 𝑚𝑎𝑘𝑠(0.2 𝑙𝑤 + 0.03ℎ𝑤 ; 0.08ℎ𝑤 + 0.022𝑓𝑦𝑒 𝑑𝑏 ) (5.1)

Here, lw and hw shows length and height of wall respectively. fye and db is taken as
the average yielding resistance and the largest longitudinal reinforcement diameter.

In order to select hinge height correctly, investigation is done by P3D software on

RW2 and TW2 walls with 5 different mesh and hinge height options with auto-size
fiber module. Analyses are performed with 9x4, 8x4, 4x4 and 2x4 mesh options
(Figure 5.46). 5 different types of mesh options automatically cause 5 different types
of strain gage length options. Strain gage length-1 (SGL-1) is the experiment strain
gage value, 229 mm. On the other hand, strain gage length-2 (SGL-2) is 271 mm
calculated according to Equation 5.1 for RW2 and TW2. Also, strain gage length-3
(SGL-3) is half the height of the floor, while strain gage length-4 (SGL-4) refers to

163
floor height and strain gage length-5 (SGL-5) is 2 times the floor height (Figure
5.46). Strain gage lengths, SGL1, SGL2, SGL3, SGL4 and SGL5 are 229 mm, 271
mm, 457.2 mm, 914.4 mm and 1828.8 mm respectively.

Figure 5.46 Vertical Mesh and Strain Gage Length Variations

Cyclic response results for 3 different P3D models with different mesh, that are
matched with test results, are shown in Figure 5.47, Figure 5.48 and Figure 5.49 for

164
8x4, 4x4, 2x4 mesh respectively. These three models are generated with "Auto-size"
cross-section module. In three different models, where the number of mesh decreases
by simplifying from 8x4 mesh to 2x4 mesh, the agreement of the P3D model hysteric
shape with the result of the experiment is gradually decreasing. Cyclic response
resulted in the best model being the 8x4, while the 4x4 mesh model is acceptable but
the model with 2x4 mesh is largely removed from results of the experiment.

Figure 5.47 Cyclic Responses of Experiment and Model with 8x4 Mesh

165
Figure 5.48 Cyclic Responses of Experiment and Model with 4x4 Mesh

Figure 5.49 Cyclic Responses of Experiment and Model with 2x4 Mesh

166
It automatically consists of five different strain gage lengths with P3D model with
five different mesh variations. At Perform 3D program, strain measurement is carried
out between nodes closest to each other. When defining stain gage, a node cannot be
skipped and tied to another desired node. For this reason, bottom mesh hinge height
section of wall automatically determines the strain gage length. For five different
strain gage lengths and mesh identifications, experiment strain profile is compared
with P3D Models for concrete and reinforcement. Concrete and reinforcement strain
profile result comparisons are presented in Figure 5.50 and Figure 5.51 respectively.
Reinforcement tensile strain and concrete compressive strain checks are more
important in terms of structural design. According to the strain checks of
reinforcement under tension, the results of the 8x4 mesh model remain on the safe
side, while the results of the 4x4 mesh model show a compatible match, on the other
hand, the results with the 2x4 mesh model give incompatible and unacceptable
results much lower than the test results. In terms of concrete compressive strain
checks, model results with 8x4 mesh except 2% drift ratio give the most compatible
results, while model results with 4x4 mesh are less accurate and model with 2x4
mesh are not accurate when compared with test results.

167
Figure 5.50 Concrete Strain Profiles of RW2 Wall for Different Mesh Variations

168
 Figure 5.51 Steel Strain Profiles of RW2 Wall for Different Mesh Variations

Strain gage length value-SGL1 in experiment is 229 mm, and in order for comparison
between model and experiment in terms of strain profile, strain gage length in P3D
model must be 229 mm. For this reason, SGL-1 value is chosen as 229 mm. SGL2
is 271 mm, which is the strain gage length value that it is encountered as required by
regulation calculated according to Equation 5.1. SGL3, SGL4 and SGL-5 are created
to observe the extent to which mesh change affects total strain profile and cyclic
response outcomes. Strain gage lengths, SGL3, SGL4 and SGL5 results are observed
and presented them with figures between Figure 5.51 and Figure 5.57. Strain-based
result evaluation is the basic approach in evaluation of wall structures in
performance-based design. For this reason, it is of great importance to observe the
extent of test and model compatibility in the evaluation of strain results according to
strain gage length-1, SGL-1 measurement. For results of experiment strain gage

169
length, 229 mm, model and final strain profile evaluation are presented in Figure
5.52 to Figure 5.57 for both RW2 and TW2 wall test samples.

Figure 5.52 RW2 Wall, Base Level Concrete Strain Profile (Positive
Displacement) for test strain gage length, SGL-1, 229 mm

170
Figure 5.53 RW2 Wall, Base Level Steel Strain Profile (Positive Displacement)
for test strain gage length, SGL-1, 229 mm

Figure 5.54 TW2 Wall, Base Level Concrete Strain Profile (Positive
Displacement) for test strain gage length, SGL-1, 229 mm

171
Figure 5.55 TW2 Wall, Base Level Steel Strain Profile (Positive Displacement) for
test strain gage length, SGL-1, 229 mm

Figure 5.56 TW2 Wall, Base Level Concrete Strain Profile (Negative
Displacement) for test strain gage length, SGL-1, 229 mm

172
 Figure 5.57 TW2 Wall, Base Level Steel Strain Profile (Negative Displacement)
for test strain gage length, SGL-1, 229 mm

Test reinforcement tension strain values are lower than P3D model strain values
according to strain profile on the base of RW2 and TW2 wall structural elements. In
other words, a design is formed on the safe side in terms of reinforcement strain
value evaluations. On the other hand, experiment concrete compression stain values
are much higher than concrete strain values in P3D model. The fact that concrete
strain in the model is low from the test values causes an inadequate and inappropriate
evaluation in wall structural member design. Especially in the case of a positive drift
for TW2 wall sample, i.e., under compression of wall flange section, concrete strain
values are much lower from the model concrete strains. In terms of compatibility
with the experimental results, negative drift behavior for TW2 wall sample reveals
results more similar to that of RW2 wall.

A numerical study is conducted on Model-31 (RW2), TModel-1 (TW2) according to

coefficient a created according to Equation 5.2 on how low concrete strain value is
in the model according to the experiment. Coefficient “a” is the ratio of experimental

173
concrete strain,c,experiment, to the model concrete strain, c,P3D model. If coefficient “a”
is greater than 1, this means that experimental result is larger from P3D model result.
If coefficient “a” is 1, it means that both model and test result are the same, which is
the ideal solution targeted. The coefficient “b” in Equation 5.3 shows the correlation
between the experiment reinforcement strain,s,experiment, and model reinforcement
strain ,s,P3D model , just like the coefficient “a”.

a=c,experiment / c,P3D model (5.2)

b=s,experiment / s,P3D model (5.3)

Figure 5.58 TW2 and RW2 Walls Concrete Strain Experiment-P3D Model
Comparison Coefficient “a” vs Drift Ratio Graph

174
 Figure 5.59 TW2 and RW2 Walls Steel Strain Experiment-P3D Model
Comparison Coefficient “b” vs Drift Ratio Graph

According to Figure 5.58, the coefficient “a” is greater than 1 for RW2, TW2
negative and positive drifts, and this coefficient “a” is almost an average of 2 for
RW2 and TW2 negative drift, on the other hand coefficient “a” is between 4 and 16
for TW2 positive drift, in which flange is under compression. This shows that P3D
model concrete strain results are far and less from experiment results. While model
concrete strain results for RW2 and TW2 under negative drift are about 2 times less
than test strain results, however, it is between 4-16 times less than experiment
concrete strains for TW2 under positive drift.

The coefficient b is less than 1 for all walls samples and drift directions according to
Figure 5.59. In other words, the results of model reinforcement strains are higher

175
than experimental results under all these conditions. This is a safe-side situation to
nonlinear design of structure and does not pose a problem.

Strain gage length-1 (SGL-1), 229 mm and strain gage length-2 (SGL-2), 271 mm
are almost identical values, and as seen in Figure 5.57 and Figure 5.58, a and b values
are almost identical to both strain gage lengths. For this reason, all interpretations
and evaluations for a and b values are valid within strain gage length-2 (SGL-2) in
Equation 5.1 required by the regulation.

To sum up, in the study for five different mesh types and strain gage lengths, the
model containing 2x4 mesh and strain gage length-5 (SGL5) exceeds the multiple
height in terms of both mesh and strain gage length. In 8x4 and 4x4 mesh models,
the mesh and strain gage length dimensions are at the maximum floor height.
Although P3D models having 8x4 and 4x4 mesh give sufficient results in terms of
hysteric shape and strain profile matching, the 8x4 mesh model was more acceptable
and give satisfactory results. Model results with 2x4 mesh are not acceptable results
compatible with the test results. It is recommended not to exceed the floor height in
mesh work and strain gage length, so design code restriction in terms of strain gage
length, maximum strain gage length cannot be exceeded floor height, is also verified.
On the other hand, with SGL-1 and SGL-2 study, P3D model and test compatibility
of the strain profile is checked. Reinforcement strain in P3D model is higher than the
results in experiment, it creates a result on the safe side in terms of nonlinear design,
while concrete strain result is much lower than test results in the model. According
to this study, in the design evaluations made with Perform 3D software, it is
recommended to evaluate concrete strain results for the wall structural members by
multiplying it by 2 and evaluating it with the limit value of the regulation.

176
5.2 CALIBRATION WORK FOR LINK BEAMS

One of the most important structural elements in meeting the seismic demands in
high-rise buildings is the link beam. Earthquake force is effectively resisted together
with the core wall group and the link beams. The clear span and heights of the link
beams are generally available because of the fire escape, elevator usage and
mechanical shaft areas that are architecturally necessary within core wall group. It is
thought that accurate modeling of coupling beams, which are very effective in
absorbing earthquake energy, will significantly affect the accuracy of the result in
3D structural performance analyses. For this reason, verification work should be
carried out on experimental results for link beams similar to conducted for shear-
walls.

Aspect ratio (clear span over beam height) is the important decisive parameter for a
beam to be referred as a “frame beam" or a "coupling beam". Accordingly,
reinforcement details and calculations will vary significantly. According to the
Turkish Earthquake Regulation (2018), beams with an aspect ratio value of less than
2 and containing high shear force (Vn > 0.35 √fck Acv) are classified as coupling
beams, while those other than the specified limits are called frame beams. However,
it is important to verify with accurate modeling and experimental results within the
beams in the frame beam class, which have an aspect ratio value greater than 2 in
high structures because they are similar in terms of behavior and the load they are
exposed to. Because link beams with this type of high aspect ratio are widely used
in high structures. The majority of the link beam beams in our case study building
that we will analyze have high aspect ratio (Figure 5.60). LB1 type link beams have
an aspect ratio of 6-7, while the aspect ratio is approximately 1 for LB2.

177
 Figure 5.60 Link Beams of the Case Study Building

Experimental studies for commonly used link beams of case study building are
selected for modelling with Perform-3D. Model and experiment lateral load-rotation
response results are compared for link beam with 2 types of reinforcement details,
flexural straight reinforcement (used in conventional link beam) and diagonal
reinforcement (used in diagonal link beam). Firstly, necessary information about the
experimental work is given for link beams separated as conventional and diagonal
link beam.

5.2.1 Experimental Data for Link Beams

Link beam tested by Naish et al. is a reference for our calibration study (Naish et al.,
2013a). This study was made on link beams having aspect ratio larger than 2 but less
than 4. Modeling work was carried out on link beams labelled as CB33F and FB33

178
for diagonal and conventional link beams respectively (Figure 5.61 and Figure 5.62).
Both sample link beams were studied on a 1/2 experimental scale.

Figure 5.61 CB33F-Diagonal Sample Link Beam and Reinforcement Details

(ln/h=3.33 ; 1 in.=25.4 mm) (Naish et al., 2013a)

Figure 5.62 FB33-Conventional Sample Link Beam and Reinforcement Details

(ln/h=3.33; 1 in.=25.4 mm) (Naish et al., 2013a)

Diagonally reinforced link beam, CB33F, consists of 9.5 straight flexural

reinforcements and 22 diagonal reinforcement bundle design as seen in Figure 5.61.
While the diagonal reinforcement bundle is not confined with stirrups, stirrups and
ties of beam is used 34% more than necessary. On the other hand, FB33, consists of
conventionally reinforced link beam. There are 19 reinforcements in the lower and
upper sections of the beam, while 9.5 reinforcements are available in the beam web

179
(Figure 5.62). It is only confined with stirrups (without any ties). Table 5.13 shows
some geometric and material properties for these sample test link beams.

Table 5.13 Geometric and Material Properties of Test Sample Link Beams

Diagonal Transverse
Aspect Clear Beam
Bundles Reinforcement
Ratio Span Dimensions
Angles (mm)
ln/h , degrees ln (mm) b/h (mm) Full Section
CB33F 3,33 12,3 1520 305/457 9.5 / 7.6
FB33 3,33 0 1520 305/457 9.5 / 7.6

(Ash,act / (Ash,act /
fc' (MPa) fy' (MPa) fu' (MPa)
Ash,req)x Ash,req)y
CB33F 1.34 1.26 47 483 620
FB33 -- -- 41 483 620

A suitable experiment setup was established to provide both force and displacement
control without creating axial force in link beams (Figure 5.63). Both vertical and
horizontal actuators were used and necessary measures are taken to prevent negative
factors such as slippage, rotation, etc. Linear variable differential transformers
(LVDTs) were used for displacement measurements.

Figure 5.63 Link Beam Laboratory Test Setup

180
The link beam load protocol started with load-controlled approach and continues
with displacement-controlled approach (Figure 5.64). Load controlled approach was
thought according to yielding force through with 0.125, 0.25, 0.5, and 0.75 Vy where
Vy=2 My/ln, respectively (Figure 5.64). The load-controlled protocol to capture
yielding behavior continues with displacement-controlled protocol. In displacement
controlled, percent chord rotation, where θ = /ln (: relative lateral displacement; ln:
beam clear span) was taken into account (Figure 5.64).

(a) (b)

Figure 5.64 Loading protocol of link beam test, a) Load-controlled (kN) b)

Displacement-controlled

Load-deformation response curves for diagonally-detailed, CB33F, and

conventionally straight detailed, FB33, link beams are shown in Figure 5.65 and
Figure 5.66, respectively. Firstly, it is seen that energy absorbing area of CB33F link
beam is larger than FB33 and also FB33 exhibits more pinching behavior according
to CB33F. Approximately both beams start to yield at a rotation of %1 chord rotation,
while strength degradation starts at approximately 7-8% rotation for CB33F beam
and 4-5% rotation for FB33 beam. Flexural, shear and slip-extension cracks at beam
wall face contribute to the deformation failure mode in link beam. But the most
effective is the slip-extension at beam face. In addition, as the beam rotation
deformation increases, slip-extension deformation activity increases (Naish et al.,
2013a).

181
Figure 5.65 Link Beam, CB33F, Load Deformation Response

Figure 5.66 Link Beam, FB33, Load Deformation Response

182
5.2.2 Simulation of Link Beam Test Samples at Perform 3D V 7.0

[Link] General Views of Link Beam Perform 3D Models

Frame elements were used in the modeling of link beams with Perform 3D program.
The link beam model was formed by combining the cross-section and inelastic
component definitions in the component properties section in the "Frame member
compound component" section at P3D. The P3D component joining approach can
be done in 2 different ways. As the first approach, inelastic “moment rotation spring”
can be assigned at the beam ends and cracked elastic beam section in the middle
section. Secondly, inelastic “shear displacement hinge” component in the middle of
the beam can be defined and remaining portions of beam will be cracked elastic
portions (Figure 5.67).

Figure 5.67 Two different Link beam Modeling Approaches: a) Moment Hinge
Model b) Shear Hinge Model (Naish et al., 2013b)

It is appropriate to use a "Moment-hinge model" for conventionally reinforced link

beams whose active behavior is flexure and is not under very high shear force.
Modeling methods can be preferred for a link beam that is under high shear force
detailed with diagonal reinforcements. So, for CB33F link beam from selected
experimental samples, it is modeled in the P3D program with both modeling
approaches, i.e., “Moment-hinge Model” and “Shear-hinge Model” and the relevant
results were observed. On the other hand, for FB33 test sample beam detailed with
conventional reinforcement, only the "Moment-hinge model" approach was used.

"Section Stiffness" and "Material Stiffness" properties were defined through the
cross-section definition of beam sections. In addition, the effective stiffness ratio was

183
set to 0.15 EcIg. An effective stiffness ratio of 0.15 in link beams is the most
appropriate choice (Naish et al., 2013b). According to this coefficient, the moment
of inertia has been reduced.

Inelastic hinge behavior definitions in link beam can be created with the modules
"Moment Hinge, Rotation Type" and "Shear Hinge, Displacement Type" at P3D.
After cracked elastic beam section and inelastic hinge portions of link beam frame
element were defined, these components were merged as a "compound component"
at P3D according to two modeling methods mentioned earlier. After the appropriate
combination of cross-section and inelastic components, modeling was completed
with definition of link beam frame element before loading. Loading assignments
were made according to specified loading protocol (Figure 5.64) and the results were
evaluated.

[Link] P3D Model Parameters and Their Effects on Cyclic Response of

Link Beams

It is important to correctly define inelastic hinge sections of link beams. CB33F

diagonal link beam and FB33 conventional link beam were defined with the use of
"Moment hinge, Rotation Type" and "Shear Hinge, Displacement Type" modules in
the inelastic hinge sections.

Figure 5.68 Perform 3D Hysteretic Loop Model(Computers and Structures, 2006)

184
As it can be remembered from the typical backbone curve definitions in Figure 5.68,
the input values in the presented tables should be evaluated according to this figure.
The action vs. deformation curves inputs for the module "Moment hinge, Rotation
Type" and "Shear Hinge, Displacement Type" are in the form of moment ([Link])
vs. rotation (%) and shear Force (N) vs displacement (mm), respectively. However,
all cyclic response comparison results of link beams are in the form of lateral load
(kN) vs. beam chord rotation (%).

The most important parameters in the definition of link beam inelastic sections are
yield and residual strength capacities, yielding (Y), ductile (L) and residual
deformation limits. In addition to the correct definition of these parameters,
assigning true cyclic degradation factor value correctly can have the most impact on
cyclic response result. With the correct value of the cyclic degradation factor, the
energy damping area and amount in the beam are adjusted. Although the effect of
the stiffness degradation factor on the cyclic response result is not as much as the
cyclic degradation factor, it is another parameter to be considered. According to the
modeling approaches, the results of the cyclic response for diagonal and
conventional link beam are presented with graphs in which the effect of each
parameter is observed.

a) Calibration Work of “CB33F” (Diagonally Detailed Link Beam) with

Shear Hinge Model Approach

Diagonal link beam "CB33F" was modeled with the shear hinge model approach in
Perform-3D program. The reference model was created with appropriate, predicted
input parameter values. Reference model and experimental cyclic response result
comparison for the "CB33F" link beam is given in Figure 5.69. New model trials
were created by changing the parameters over the reference model. With these new
model trials, effect of the parameters was observed by comparing results of the cyclic
response of the experiment. Descriptions of the new model trials were provided in
Table 5.15.

185
In shear hinge model approach in diagonal link beams, yield shear force strength was
theoretically calculated with Equation 5.4 according to the Turkish Earthquake
Regulation-2018.

𝑉d = (2𝑓yd sin𝛾) 𝐴sd (5.4)

The value calculated by Equation 5.4 was compatible with the test value specified in
Table 5.14. In addition, yielding (DY), strength degradation (DL) and residual
strength (DR) deformation limits were proposed for this experimental sample as
1.0%, 6.0% and 9.0% respectively ((Naish et al., 2013b). For the CB33F test link
beam sample with a clear span of 1524 mm, according to these specified chord
rotation values, yielding (DY), strength degradation (DL) and residual strength (DR)
deformation limits according to the "shear hinge, displacement model" approach
were 15 mm, 90 mm and 130 mm respectively.

Table 5.14 CB33F and FB33 Link Beams Experimental Strength and Deformation
Limits (Naish et al., 2013a)

Vn
Mn+/-,
V@ Mn , (ACI), Vave, Vy, Vmax, y, u,
kN.m
kN kN kN kN kN mm mm
CB33F 408.5 536 479.7 526.4 479.3 551.8 15.2 137.2
FB33 163.9 215 - 250.5 213 258.5 7.8 76.2

186
Figure 5.69 “CB33F Link Beam” Force-Deformation Response of Experiment and
Reference P3D “Shear Hinge Model” (R)

Table 5.15 Model Trials Explanations Different from “CB33F” Reference Link
Beam Shear Hinge Model

Model # CHANGE FROM REFERENCE PD3 MODEL (R) SENSITIVITY

R Reference P3D Model --
Beam bending moment of inertia is taken as gross
S-Model-1 High
moment of inertia (I=Ig) instead of I=0.15Ig
YULRX energy factors are used as Y:1 U:0.65
S-Model-2 High
L:0.65 R:0.65 X:0.65
Strength loss limits, DL and DR, are changed as 90
S-Model-3 High
and 120 mm instead of 120 and 180 mm respectively.
Strength loss limits, DL, DR and DX, are changed as
S-Model-4 90, 120 and 130 mm instead of 120, 180 and 200 mm High
respectively.
Strength loss limits, DL, DR and DX, are changed as
90, 120 and 130 mm instead of 120, 180 and 200 mm
S-Model-5 High
respectively. Also, YULRX energy factors are used
as Y:1 U:0.65 L:0.65 R:0.65 X:0.65.

187
Each parameter change is indicated by the degree of impact in Table 5.15 together
with a description. Each new model trial and reference model cyclic response result
comparisons are given in Figure 5.70. According to these comparisons, strength and
deformation assignments determine the limits for cyclic backbone curve. In addition,
it can be easily determined from comparisons that the most important coefficient
affecting the result is the energy degradation factor. If diagonal link beam is modeled
with the shear hinge modeling approach, this value is determined to be 0.65. In
addition, for the "CBF33" link beam, Table 5.16 provides that ideal model results
are compared with reference model for each input parameter.

Figure 5.70 Cyclic Response Results Comparisons of P3D Model Trials for
Diagonal Link Beam, CB33F with “Shear Hinge, Displacement Type” Module

188
 Table 5.16 “CB33F” Link Beam “Shear Hinge Model” Input Parameters of
Reference Model and Ideal Model

Reference
SHEAR HINGE, Displacement Type Ideal Model
Model
Mater. 1 E (MPa) 3.26E+04 3.26E+04
Material and Section Stiffness

2 G (MPa) 1.36E+04 1.36E+04

3 Axial Area (mm2) 139385 139385
4 Shear Area (2) 0 0
Section

5 Shear Area (3) 0 0

6 Bending Inertia-I2 (mm4) 1.62E+08 1.62E+08
7 Bending Inertia-I3 (mm4) 3.64E+08 3.64E+08
8 Torsional Inertia-J (mm4) 2.73E+09 2.73E+09
9 FY (N) 450000 450000
Relationship

Pos.-Neg.

10 FU (N) 470000 470000

Basic

11 DU (mm) 15 15 (0.01 ln)

12 DX (mm) 200 130 (0.09 ln)
Deformations
Strength Loss

13 DL (mm) 120 90 (0.06 ln)

Pos.-Neg.

14 DR (mm) 180 120 (0.08 ln)

15 FR/FU 0.3 0.3
All 16 SLI 0 0
Cyclic Degradation

17 Y - 1
18 U - 0.65
Pos.-Neg.

19 L - 0.65
20 R - 0.65
21 X - 0.65
22 USF 0 0
SLI: Strength Loss Interaction; USF: Unloading Stiffness Factor; Y:Yielding; U:Ultimate; L:Ductile Limit;
R:Residual Limit; X:Analysis Stop Limit

189
 b) Calibration Work of “CB33F” (Diagonally Detailed Link Beam) with
Moment Hinge Model Approach

CB33F diagonal link beam was also modelled with moment hinge assignment at both
ends of link beam instead of modeling with the shear hinge model in the middle of
the beam. According to this approach, firstly, reference model was generated with
foreseen parameters. There were three different model trials that were changed
separately from the reference model and the necessary definition explanations were
made in Table 5.17. The calibration of the experimental sample cyclic response result
with the "M-Model-3" trial results is ideal.

Table 5.17 Model Trials Explanations Different from “CB33F” Reference Link
Beam Moment Hinge Model

Model # CHANGE FROM REFERENCE (R) SENSITIVITY

R Reference P3D Model --
Rotation limits, DL, DR and DX, are changed
M-Model-1 as 0.06, 0.09 and 0.1 rad instead of 0.08, 0.12 High
and 0.15 rad respectively.
Rotation limits, DL, DR and DX, are changed
as 0.06, 0.09 and 0.1 rad instead of 0.08, 0.12
M-Model-2 and 0.15 rad respectively. Also, YULRX High
energy factors are used as Y:1 U:0.65 L:0.65
R:0.65 X:0.65.
Rotation limits, DL, DR and DX, are changed
as 0.06, 0.09 and 0.1 rad instead of 0.08, 0.12
M-Model-3 and 0.15 rad respectively. Also, YULRX High
energy factors are used as Y:1 U:0.5 L:0.5
R:0.5 X:0.5.

Figure 5.71 shows cyclic response comparisons of reference models and trial models
of CB33F test sample link beam. Backbone curve limit definitions in the moment
hinge model approach are equivalent to values used in the shear hinge model
approach. However, it is more appropriate to use 0.5 for “moment hinge” model
approach instead of 0.65 used in the cyclic energy degradation factor "shear hinge"
model approach. According to Pacific Engineering Research Center, YULRX cyclic
degradation factors are recommended as 0.5, 0.45, 0.4, 0.35, 0.35 respectively for

190
shear hinge" model approach (Moehle et al., 2011). This value is almost compatible
with our calibrated value. The ideal definition parameters and reference model
parameter values are presented in Table 5.18.

Figure 5.71 Cyclic Response Results Comparisons of P3D Model Trials for
Diagonal Link Beam, CB33F with “Moment Hinge, Rotation Type” Module

191
 Table 5.18 “CB33F” Link Beam “Moment Hinge Model” Input Parameters of
Reference Model and Ideal Model

MOMENT HINGE, Displacement Type Reference Ideal

Mater.

1 E (MPa) 3.26E+04 3.26E+04

Material and Section Stiffness

2 G (MPa) 1.36E+04 1.36E+04

3 Axial Area (mm2) 139385 139385
4 Shear Area (2) 0 0
Section

5 Shear Area (3) 0 0

6 Bending Inertia-I2 (mm4) 1.62E+08 1.62E+08
7 Bending Inertia-I3 (mm4) 3.64E+08 3.64E+08
8 Torsional Inertia-J (mm4) 2.73E+09 2.73E+09
9 FY ([Link]) - -
Relationship

Pos.-Neg.

10 FU ([Link]) 3.89E+08 (1) 3.89E+08 (1)

Basic

11 DU (rad) - -
12 DX (rad) 0.15 0.1
13 DL (rad) 0.08 0.06
Deformations
Strength Loss

Pos.-Neg.

14 DR (rad) 0.12 0.09

15 FR/FU 0.3 0.3
All 16 SLI 0 0
17 Y - 0.5
Cyclic Degradation

18 U - 0.5
Pos.-Neg.

19 L - 0.5
20 R - 0.5
21 X - 0.5
22 USF 0 0
SLI: Strength Loss Interaction; USF: Unloading Stiffness Factor; Y:Yielding; U:Ultimate; L:Ductile Limit;
R:Residual Limit; X:Analysis Stop Limit; (1) Mu= Vy* ln / 2 ; Vy=2 fy sinγ As

192
 c) Calibration Work of “FB33” (Conventionally Detailed Link Beam)
with Moment Hinge Model Approach

The conventional link beam "FB33" test specimen, where the shear force does not
have a very dominant effect and the bending behavior is more dominant, should only
be modeled with the moment hinge model approach. In this direction, the necessary
modeling is made with moment hinge model approach in the P3D program. The
comparison of the experimental cyclic response result with the reference model
having first trial parameter assignments, is shown in Figure 5.72.

Figure 5.72 “FB33 Link Beam” Force-Deformation Response of Experiment and

Reference P3D “Moment Hinge Model” (R)

Different P3D trial models have been created and descriptions related to trail models
of which are given in Table 5.19. The cyclic response comparison results are given
in Figure 5.73. The fitting between "F-Model-5" and "FB33" conventional link beam
cyclic response result is obtained and thought as the most suitable calibrated work.

193
 Table 5.19 Model Trials Explanations Different from “FB33” Reference Link
Beam Moment Hinge Model

Model # CHANGE FROM REFERENCE (R) SENSITIVITY

R Reference P3D Model --
Trilinear Shape of Force-rotation relationship is
used instead of EPP (Elastic Perfectly Plastic)
F-Model-1 relationship. DU, DL, and FY are changed as High
0.04 rad, 0.041 rad and 1.8E8 [Link]
respectively.
Trilinear Shape of Force-rotation relationship is
used instead of EPP (Elastic Perfectly Plastic)
relationship. DU, DL, DR, and DX are changed
as 0.045, 0.047, 0.05 and 0.08 rad respectively.
F-Model-2 High
FY and FU are changed as 1.8E8 and 1.9E8
[Link] respectively. Also, YULRX energy
factors are used as Y:0.3 U:0.25 L:0.23 R:0.22
X:0.2.
All changed parameters are like F-Model-2. Only
F-Model-3 difference is that YULRX energy factors are used High
as Y:0.5 U:0.5 L:0.5 R:0.5 X:0.5.
All changed parameters are like F-Model-2. Only
F-Model-4 difference is that Unloading Stiffness Factors High
(USF), "0.5" is used instead of "0".
All changed parameters are like F-Model-2 Only
F-Model-5 difference is that Unloading Stiffness Factors High
(USF), "1.0" is used instead of "0".

DL and DR chord rotation definitions are to be between approximately 4.50-5.00%,

when making conventional link beam backbone limit definitions (Naish et al.,
2013b). These rotation values are suitable rotations obtained with a moment-
curvature analysis with section properties in Table 5.20. Ultimate rotation limits can
be easily obtained with the moment-curvature analysis, however to determine the
correct value for energy degradation factor and unloading stiffness factor is the main
subject for calibration work of conventionally detailed link beam with high aspect
ratio. As a result, energy degradation factors should be recognized between 0.2-0.3,
while unloading stiffness factor "USF" value is recommended to be "1.0". In terms
of energy degradation factor, according to Pacific Engineering Research Center,

194
YULRX cyclic degradation factors are recommended as 0.24, 0.23, 0.22, 0.21, 0.2
respectively for frame beams (Moehle et al., 2011). These values are compatible with
our calibrated value. In Figure 5.73, all model trial comparisons can be observed. In
addition, in Table 5.20, the ideal model parameters compared with the reference
model and the experimental results are presented.

Figure 5.73 Cyclic Response Results Comparisons of P3D Model Trials for
Conventional Link Beam, FB33 with “Moment Hinge, Rotation Type” Module

195
 Table 5.20 “FB33” Link Beam “Moment Hinge Model” Input Parameters of
Reference Model and Ideal Model

MOMENT HINGE, Displacement Type Reference Ideal

Mater.

1 E (MPa) 3.26E+04 3.26E+04

Material and Section Stiffness

2 G (MPa) 1.36E+04 1.36E+04

3 Axial Area (mm2) 550 550
4 Shear Area (2) 0 0
Section

5 Shear Area (3) 0 0

6 Bending Inertia-I2 (mm4) 1.62E+08 1.62E+08
7 Bending Inertia-I3 (mm4) 3.64E+08 3.64E+08
8 Torsional Inertia-J (mm4) 2.73E+09 2.73E+09
9 FY ([Link]) - 1.80E+08
Relationship

Pos.-Neg.

10 FU ([Link]) 2.19E+08 1.90E+08

Basic

11 DU (rad) - 0.045
12 DX (rad) 0.05 0.08
13 DL (rad) 0.04 0.047
Deformations
Strength Loss

Pos.-Neg.

14 DR (rad) 0.045 0.05

15 FR/FU 0.3 0.3
All 16 SLI 0 0
17 Y 1 0.3
Cyclic Degradation

18 U - 0.25
Pos.-Neg.

19 L 0.65 0.2
20 R 0.65 0.22
21 X 0.65 0.2
22 USF 0 1
SLI: Strength Loss Interaction; USF: Unloading Stiffness Factor; Y:Yielding; U:Ultimate; L:Ductile Limit;
R:Residual Limit; X:Analysis Stop Limit

196
[Link] Summary of Calibration Work for Link Beams

Calibration work were completed with experimental force-beam chord rotation

hysterical curve of CB33F diagonal link beam and FB33 conventional link beam by
modeling with the Perform-3D program. The modeling method and parameter inputs
that will obtain similar results with the experimental results are also presented
separately in advance. Figure 5.74, Figure 5.75 and Figure 5.76 displays calibrated
work results for CB33F and FB33 link beam test samples.

Figure 5.74 CB33F Link Beam Cyclic Response Calibration Work Result with
Shear Hinge Approach

197
Figure 5.75 CB33F Link Beam Cyclic Response Calibration Work Result with
Moment Hinge Approach

Figure 5.76 FB33 Link Beam Cyclic Response Calibration Work Result with
Moment Hinge Approach

198
These calibration results are studies for experimental link beam samples reduced to
1/2 scale. Backbone deformation limits will vary for the exact beam scale. For this
reason, a recommendation limits are established to be used in ACI 41-06 by editing
deformation limits used as a result of the calibration study in the experimental
samples (Naish et al., 2013b). According to the generalized force-deformation
notation in Figure 5.77, recommendation limits for diagonally coupling beams are
given in Table 5.21 according to shear force ratio and aspect ratio (ln/h). However,
the values of a, b and c for conventionally reinforced coupling should be calculated
with moment-curvature analysis for section of link beam. For residual strength
degradation ratio both diagonally reinforced and conventionally reinforced link
beam, it is optimal to for a value of 0.3. As a result of the final evaluations, the
proposal to define modeling in the Perform-3D program for coupling beams of 3D
structures is given in Table 5.22.

Figure 5.77 Generalized Force-Deformation Relation in ACI41-06 Code (“Seism.

Eval. Retrofit Exist. Build.,” 2017)

199
Table 5.21 Diagonally Reinforced Coupling Beam Proposed Deformation Limits
in ACI41-06 Code (“Seism. Eval. Retrofit Exist. Build.,” 2017)

200
 Table 5.22 Summary P3D Input Parameters of Diagonally and Conventionally
Reinforced Link Beam

Conventionally
Diagonally Reinforced LB Reinforced
LB
SHA MHA MHA
Mater.

1 E (Mpa) Ec Ec Ec
Material and Section

2 G (Mpa) 0.4Ec 0.4Ec 0.4Ec

Stiffness

4 Shear Area (2) 0 0 0

5 Shear Area (3) 0 0 0
6 Bending Inertia-I2 0.15 Ig 0.15 Ig 0.15 Ig
7 Bending Inertia-I3 0.15 Ig 0.15 Ig 0.15 Ig
9 FY Vy - My
Relationship
Pos.-Neg.

10 FU 1.05*Vy My(1) 1.05*My

Basic

11 DU (rad) 0.01(3) ln 0.01(3) θU(2)

12 DX (rad) 0.09(3) ln 0.09(3) θX(2)
Deformatio

13 DL (rad) 0.06(3) ln 0.06(3) θL(2)

Pos.-Neg.
Strength
Loss

0.08(3) ln 0.08(3) θR(2)

ns

14 DR (rad)
15 FR/FU 0,3 0,3 0,3
17 Y 1 0,5 0,3
Cyclic Degradation

18 U 0,65 0,5 0,25

Pos.-Neg.

19 L 0,65 0,5 0,22

20 R 0,65 0,5 0,2
21 X 0,65 0,5 0,2
22 USF 0 0 1
USF: Unloading Stiffness Factor; Y:Yielding; U:Ultimate; L:Ductile Limit; R:Residual Limit; X:Analysis Stop
Limit; SHA: Shear Hinge Approach; MHA: Moment Hinge Approach; LB: Link Beam; 𝑽y = (2𝑓yd sin𝛾) 𝐴sd ;
(1) My = Vy* ln / 2 ; (2) θ: Calculated with Moment-Curvature Analysis; (3):It can be selected with Table 5.2

201
202
 CHAPTER 6

6 PERFOMANCE BASED DESIGN OF THE CASE STUDY BUILDING

Performance based design of case study tall building that has been designed using
linear elastic analysis method as described in Chapter 3. Reinforcement details and
dimensions as a result of that design are preliminary requiring checks with PBD
approach. In the nonlinear design check, CSI Perform 3D software is used. In
Chapter 5, the structural walls and link beams models are calibrated using Perform
3D program to match the results of experimental cyclic response and strain profiles
with model ones. Thus, how to model the structural walls and coupling beams by
using Perform 3D software are studied in detail in Chapter 5. In this chapter,
nonlinear performance analysis is performed for two different locations with low and
high seismicity. Six different building models are created in this section. While the
first and second of these model groups in each region have modeling requirements
for wall and link beams created without model calibration, the third models are
analyzed with calibrated parameters. In this way, it is possible to observe the extent
to which the agreement with the results of calibration work affects the building
response. In addition, it is observed to what extent the reinforcement and dimensions
obtained by linear elastic design in low and high seismic areas, are affected with the
performance-based design.

6.1 Performance Based Design Criteria

6.1.1 Design Earthquake Loads

The time history analysis method is used for case study building in terms of design
earthquake load. According to the selected DD-2 level spectrum curves, preliminary
design is conducted according to linear analysis requirements detailed in Chapter 3.

203
However, for the nonlinear performance analysis, time history earthquake data,
matched with the DD-1 level spectrum curves as target spectrum curves, is used.
DD2 level and DD-1 level spectrum curves are shown in Figure 6.1 for two different
earthquake regions.

Figure 6.1 Design DD1 Level (Target Spectrum) and DD2 Level Response
Spectrum Curves for Ankara and Istanbul

Seven earthquake records to be scaled are selected for the target spectrum curves
(DD1 level) of both regions in Figure 6.1. Properties of selected scaled earthquake
records for ZC soil class are shown in Table 6.1 & Table 6.2

Table 6.1 Selected Earthquake Records of Time History Analysis for Ankara

5-95% Arias
Earthquake Vs30
# Year Magnitude Mechanism Duration Intensity
Name (m/sec)
(sec) (m/sec)
Reverse
1 Loma Prieta 1989 6.93 284.79 17.9 0.3
Oblique
2 Landers 1992 7.28 strike slip 324.62 25.7 0.3
3 Duzce 1999 7.14 strike slip 481.0 15.5 0.2
4 Duzce 1999 7.14 strike slip 638.39 14.9 0.4
5 Landers 1992 7.28 strike slip 436.14 27.3 0.3
6 Darfield 2010 7.0 strike slip 204.0 20.0 1.1
7 Darfield 2010 7.0 strike slip 280.26 27.9 0.2

204
 Table 6.2 Selected Earthquake Records of Time History Analysis for Istanbul

5-95% Arias
Earthquake Vs30
# Year Magnitude Mechanism Duration Intensity
Name (m/sec)
(sec) (m/sec)
1 Kocaeli 1999 7.51 strike slip 386.75 34.3 1.0
2 Hector Mine 1999 7.13 strike slip 379.32 14.6 0.6
3 Landers 1992 7.28 strike slip 368.2 32.9 1.0
4 Tottori 2000 6.61 strike slip 616.55 9.9 0.4
5 Tottori 2000 6.61 strike slip 469.79 12.7 5.2
6 Tottori 2000 6.61 strike slip 420.2 11.1 2.9
7 Darfield 2010 7.0 strike slip 422.0 15.7 4.1

a)

b)

Figure 6.2 Scaled Time history Data Set a) for Ankara & b) for Istanbul

According to the target response spectra (DD1 level) in Figure 6.1, scaled time
history records are obtained between 0.05-8 seconds (Figure 6.3). Scaled earthquake
records for target spectrum curves for Ankara and Istanbul earthquake zones are
shown in Figure 6.2.

205
a)
1.0

0.9

0.8

0.7
Acceleration (g)

0.6

0.5

0.4

0.3

0.2

0.1

0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8
Period (sec)

b)
2.2
2

1.8

1.6
Acceleration (g)

1.4

1.2

1
0.8

0.6

0.4
0.2

0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8
Period (sec)

Figure 6.3 7 Matched EQ Spectrums for a) Ankara & for b) Istanbul

6.1.2 Some Design Criteria for the Case Study Building

 Earthquake design level for PBD is DD1 design level, i.e., peak ground
acceleration (PGA) with an exceedance probability of %2 in 50 years and
2475 years return period.
 Structural performance level for case tall building for this earthquake design
is collapse prevention (CP) in Ankara and in Istanbul respectively.

206
  Behaviors expected from structural elements in the performance-based
design under DD1 level earthquake are shown in Table 6.3 for our case study
tall building.

Table 6.3 Structural Members Actions and Behavior for Case Tall Building under
DD1 Level Earthquake Design

DEFORMATION CONTROLLED ACTION- Inelastic

Behavior
STRUCTURAL
Under Moment Under Shear Load Under Axial Load
MEMBER
Shearwall 1  
Frame Beam 2  
Coupling Beam  3 
 FORCED CONTROLLED ACTION- Elastic Behavior
STRUCTURAL
Under Moment Under Shear Load Under Axial Load
MEMBER
Shearwall   
Frame Beam   --
Coupling Beam  4 --
1 2
P-M-M yielding of wall base (on top of foundation or basement podiums), Flexural
yielding of beam ends, 3 Shear yielding of diagonally reinforced coupling beams, 4 Shear of
conventionally reinforced coupling beams

 According to Turkish Earthquake Code-2018, damping ratio in high-rise

buildings is taken as 2.5% for MCER (maximum considered earthquake),
DD1 level.
 Used design characteristic and expected material properties are shown in
Table 6.4.

Table 6.4 Design Expected Material Properties

Design Material Characteristic Strength Expected Strength
Concrete fck=50 MPa fce=65 MPa
Reinforcing Steel fyk=420 MPa fye=504 MPa

207
6.2 Model of the Case Study Tall Building

Case study high-rise building was modeled with CSI Perform 3D program (Figure
6.4). In order to obtain the correct structural behavior, structure periods and mode
shapes in linear analysis must be compatible with the model in nonlinear analysis. In
this context, attention has been paid to mass definitions and modeling of building
elements. Periods of case study building is shown in Figure 6.5. Case study building
has core-wall connected by link beams in the middle and surrounded by perimeter
columns. Tower has the first 4 floors of the basement and 28 floors above the
basement, with flat slab system. Although all of the earthquake force is taken by
core-wall and link beams in linear analysis, only the core-walls and link beams of
the case study structure without slabs are not defined in nonlinear analysis. Tower
perimeter columns and floors were included in the P3D model and a modeling close
to the structure behavior in linear analysis was obtained. Flat slab was defined as
effective beams on all floors (Figure 6.6 & Figure 6.8), while on the ground floor
they were defined as shells only to observe the back stay effect (Figure 6.7). Effective
frame beams for slabs and basement walls were modelled as elastic cracked
members.

208
Figure 6.4 Nonlinear P3D Model of Case Study Tall Building

209
Figure 6.5 Periods of Case Study Building in Ankara & Istanbul at P3D

210
Figure 6.6 First Three Basement Floors Model with Effective Beams of Slab

Figure 6.7 Ground Floor Plan Model with Shells of Slabs

211
 Figure 6.8 Typical Floor P3D Model Plan of the Case Study Tall Building

In the modeling of slabs, modeling was made according to rigid diaphragm

acceptance with slaving module of P3D (Figure 6.9). The in-plane forces and
displacements of transfer slab (4th story) were taken into account by not being
modeled as rigid diaphragm on the ground floor defined only in the shell. In addition,
masses were defined in the mass center for all slabs (Figure 6.10).

212
 Figure 6.9 Rigid Diaphragms Assignment for Typical Floor of Case Building

Figure 6.10 Mass Assignment at Mass Center for Typical Floor of Case Building

In the modeling of the case study building core-wall, the vertical fibers were defined
according to the "fixed size" module approach of the P3D program. Accordingly,

213
concrete confined and unconfined areas and reinforcement areas were defined as
fibers according to their coordinates. The wall boundary sections were modeled as
reinforcement and concrete fiber partitions according to 0.5x1 or 1x1 mesh aspect
ratio, while the wall web parts fibers were modeled as 2x1 aspect ratio (Figure 6.11).
While the concrete and reinforcement fibers of the walls in the plan was modeled in
the plan, the walls were divided at each story and the nodal points are created
accordingly in the vertical direction (Figure 6.12). Taking into account the 60 cm
depth of beams in the core-wall groups, horizontal and vertical embedded beams
were defined to transfer the moments. For this reason, there has been an increase in
vertical wall meshing. In the first basement floor (6.8 m height) under the ground
floor, the first three floors (6.8 m and 3.8 m height) within the critical wall height,
i.e., between basement-1 and 3th story in Figure 3.3, walls height were divided into
two vertically. In the Perform 3D program, since the strain gages are defined only
between the nodes, the vertical meshes also determine strain gage length
automatically. Strain gage lengths were defined as full floor height except at the
critical wall height and basement floors (excluding 1st basement floor, 6.8 m height),
while within the critical wall height, strain gage lengths were defined as half of the
floor height. These lengths are less than half of the length of the core-wall legs in the
plan. Strain gages were laid out uniformly and properly in the plan to measure the
strains in all boundaries and webs of core-walls. Strain gage positions are shown in
Figure 6.13. Strain gage limits (Table 6.5) were defined according to TEC-2018 with
expected concrete and steel strengths. In addition, according to TEC-2018, the limits
created according to C50 and B420C characteristic material strength classes will be
examined for the "Collapse Prevention (CP)" target design limit for the structures in
Ankara and Istanbul respectively.

214
Figure 6.11 Core-wall Leg Fibers Aspect Ratio View

Figure 6.12 Wall Vertical Mesh View

215
 Figure 6.13 Core-wall Strain Gage Positions

Table 6.5 Defined Strain Gage Limits

Compression Limits
Unconfined Confined
Concrete Parts Concrete Parts
Tension Limits
(for Web of (for Boundary
Walls) of Walls)
Operational (O) 0.005 0.00125 0.00125
Immediate Occupancy (IO) 0.0075 0.0025 0.0025
Life Safety (LS) 0.024 0.002625 0.0075
Collapse Prevention (CP) 0.032 0.0035 0.01

Wall reinforcement and concrete material definitions are conducted with the
"Inelastic Steel Material, Non-Buckling" and "Inelastic 1D Concrete Material"
modules, respectively (Figure 6.14 & Figure 6.15). In Chapter 5, it was observed that
the reinforcement material parameters were very important in terms of the sensitivity

216
of response estimations. Accordingly, reinforcement yield strength, cyclic
degradation factors and unloading stiffness factor parameters were determined as the
most important sensitivity parameters. One of the main objectives of the case
building study in this section is to observe the extent to which these highly sensitive
parameters affect the nonlinear design results. For two locations with different
seismicity, (i.e., Ankara and Istanbul), case study building nonlinear analysis was
performed with the parameters obtained as a result of calibration (A3) and two
different uncalibrated (A1-A2) alternatives and the performance analysis results
were compared accordingly. The P3D parameter assignments of reinforcement and
concrete material of wall for one calibrated and two non-calibrated alternatives are
given in Table 6.6.

Figure 6.14 Concrete Material Model of Wall fibers

217
Figure 6.15 Reinforcement Material Model of Wall fibers

218
 Table 6.6 Calibrated and Uncalibrated Concrete and Reinforcement Material
Parameter Assignments of Core-walls

MATERIALS
UNCALIBRATED CALIBRATED
Rein. Rein. Concrete Rein. CONCRETE
(A1) (A2) Confined Unconfined (A3) Confined Unconfined
All 1 E (Mpa) 2E+05 2E+05 40310 40310 2E+05 40310 40310
FY
2 504 504 --- --- 440 --- ---
Tens.

(Mpa)
FU
Basic Relationship
Stresses

3 600 600 --- --- 600 --- ---

(Mpa)
FY
Compr. Tension Com Tens. Compr.

4 504 504 45,4 39 504 45,4 39

(Mpa)
FU
5 600 600 75,7 65 600 75,7 65
(Mpa)
6 DU 0,07 0,07 --- --- 0,07 --- ---
Strains

7 DX 0,08 0,08 --- --- 0,08 --- ---

8 DU 0,07 0,07 0,003 0,0018 0,07 0,003 0,0018
pr.

9 DX 0,08 0,08 0,02 0,02 0,08 0,02 0,02

10 DL --- --- --- --- --- --- ---
11 DR --- --- --- --- --- --- ---
Strength Loss
Strains

12 FR/FU --- --- --- --- --- --- ---

13 DL --- --- 0,0052 0,0022 --- 0,0052 0,0022
14 DR --- --- 0,01 0,0042 --- 0,01 0,0042
15 FR/FU --- --- 0,6 0,2 --- 0,6 0,2
16 TSLX --- --- --- --- --- --- ---
All
17 SLI --- --- --- --- --- --- ---
18 Y --- 0,35 --- --- 0,65 --- ---
Factors
Tension

19 U --- 0,35 --- --- 0,65 --- ---

Energy
Strains
Cyclic Degradation

20 L --- 0,35 --- --- 0,65 --- ---

21 R --- 0,35 --- --- 0,65 --- ---
22 X --- 0,35 --- --- 0,65 --- ---
Compr. Strains

23 Y --- 0,35 --- --- 0,65 1 1

Factors

24 U --- 0,35 --- --- 0,65 0,4 0,4

Energy

25 L --- 0,35 --- --- 0,65 0,4 0,4

26 R --- 0,35 --- --- 0,65 0,1 0,1
27 X --- 0,35 --- --- 0,65 0,1 0,1
All 28 USF --- 1 --- --- -0,5 --- ---
TSLX: Total Strength Loss at Point X; SLI: Strength Loss Interaction; USF: Unloading Stiffness Factor;
Y:Yielding; U:Ultimate; L:Ductile Limit; R:Residual Limit; X:Analysis Stop Limit

Three different reinforcement material parameters for walls are named as A1, A2
and A3 according to Table 6.5. A1 and A2 represent the uncalibrated material
parameters. Cyclic degradation and unloading stiffness factors are not taken into

219
account in the A1 uncalibrated wall reinforcement material parameter assignment.
According to the uncalibrated material parameter assignments A2, the cyclic
degradation factors are approximately half of the values in calibrated parameters of
A3.

The behavior under the out of plane and shear force was included as elastic in the
modeling of walls. In the out of plane behavior, the rigidities of walls were reduced
to one-fourth of the gross rigidity of the uncracked section according to TEC-2018.

Apart from the core-wall structural elements, the 60 cm deep beams between the
core-wall groups and the 150 cm deep link beams were conveniently included in the
models. For case study building in both Ankara and Istanbul, inelastic behavior
modeling was carried out with the module "Moment Hinge, Rotation Type" at both
ends of all frame beams with a height of 60 cm. For the 150 cm height conventional
reinforced link beams, inelastic behavior modeling was performed with the "Moment
Hinge, Rotation Type" module at both ends of the beam, while inelastic behavior
modeling was performed at the middle of beam length with the "Shear Hinge,
Displacement Type" module for the diagonally reinforced link beam at case study
building in Istanbul. The remaining elastic sections, except for the moment and shear
hinge sections of the beams, were reduced by calculating the effective cross-sectional
rigidities according to Equation 6.1.

𝑀𝑦 𝐿𝑠
(𝐸𝐼)𝑒 = (6.1)
𝜃𝑦 3

Calculated effective cross-sectional rigities after moment-curvature analysis

according to beam sectional properties are given at Table 6.7 and Table 6.8 for case
study building in Ankara and Istanbul respectively.

220
 Table 6.7 Beam Effective Rigidity Properties of Case Study Building in Ankara

L My db φy Ls I Ie
LABEL θy Ie/I
(mm) (kN.m) (mm) (1/m) (mm) (mm4) (mm4)
B40-60(1) 3500 485.4 22 0.00597 1750 0.00576 7.2E+09 1.22E+09 0.17
B40-60(2) 4200 485.4 22 0.00597 2100 0.00632 7.2E+09 1.33E+09 0.19
B40-150 1450 1969 25 0.00216 725 0.00668 1.13E+11 1.77E+09 0.02

Table 6.8 Beam Effective Rigidity Properties of Case Study Building in Istanbul

L My db ϕy Ls I Ie
LABEL θy Ie/I
(mm) (kN.m) (mm) (1/m) (mm) (mm4) (mm4)
B40-60(1) 3500 874.9 30 0.00658 1750 0.00611 7.2E+09 2.07E+09 0.29
B40-60(2) 4200 874.9 30 0.00658 2100 0.00675 7.2E+09 2.25E+09 0.31
B60-60(1) 3500 1222 30 0.00655 1750 0.00609 1.08E+10 2.9E+09 0.27
B60-60(2) 4200 1222 30 0.00655 2100 0.00673 1.08E+10 3.15E+09 0.29
B60-150 1300 5083 30 0.00237 650 0.00721 1.69E+11 3.79E+09 0.02

Moment-curvature analysis were performed for all beams with the beam
reinforcement details obtained according to the linear analysis design result.
Accordingly, yielding, ultimate moment capacities and yielding and plastic rotation
capacities of all beams were obtained (Figure 6.16 & Figure 6.27). The plastic
rotational limits were determined according to Equation 6.2 at TEC-2018 (Table
6.9). Within all these information, the assigned values of the beam and link beam
inelastic sections applicable to Ankara and Istanbul are shown in Table 6.10, Table
6.11 and Table 6.12. The rotational performance limits for the beams are shown in
Table 6.9.

2 0.5𝐿𝑝
θpCP= [(𝜙𝑢 − 𝜙𝑦 )𝐿𝑝 (1 − ) + 4.5𝜙𝑢 𝑑𝑏 ] (6.2)
3 𝐿𝑠

221
 a)

b)

Figure 6.16 Moment-Curvature Diagrams of Beams (Ankara) a) 40/60 Beams b)

40/150 Beams

222
 a)

b)

Figure 6.17 Moment-Curvature Diagrams of Beams (Istanbul) a) 40/60 Beams b)

60/60 Beams

223
 Table 6.9 Beam Rotation Limits

BEAM θp,CP
B40-60 0.041
ANKARA
B40-150 0.020
B40-60 0.046
ISTANBUL B60-60 0.046
B60-150 0.020

In the definitions in Table 6.10, Table 6.11 and Table 6.12, values for the beams are
given for three different alternatives, i.e. A1, A2 and A3, in the calibrated and
uncalibrated main headings as in the same wall modeling approach.

Table 6.10 Inelastic Component, “Moment Hinge, Rotation Type”, Assigned Values
of Beams for Building in Ankara for Calibrated and Uncalibrated Alternatives

MOMENT HINGE, IN ANKARA

Rotation Type 40/60 BEAM 40/150 BEAM
Uncalibrated Calibrated Uncalibrated Calibrated
(Trilinear Model)
(A1) (A2) (A3) (A1) (A2) (A3)
FY
1 485 485 485 1969 1969 1969
Relationship

(kN.m)
Pos.-Neg.

FU
Basic

2 542 542 542 2482 2482 2482

(kN.m)
3 DU (rad) 0,0135 0,0135 0,0135 0,014 0,014 0,0144
4 DX (rad) 0,1 0,1 0,1 0,1 0,1 0,1
5 DL (rad) 0,048 0,048 0,048 0,044 0,044 0,044
Pos.-Neg.
Deformations
Strength Loss

6 DR (rad) 0,051 0,051 0,051 0,046 0,046 0,046

7 FR/FU 0,2 0,2 0,2 0,3 0,3 0,3

All 8 SLI 0 0 0 0 0 0
Cyclic Degradation

9 Y --- 0,2 0,3 --- 0,3 0,5

10 U --- 0,2 0,25 --- 0,25 0,5
Pos.-Neg.

11 L --- 0,15 0,22 --- 0,22 0,5

12 R --- 0,15 0,2 --- 0,2 0,5
13 X --- 0,15 0,2 --- 0,2 0,5
14 USF --- 0,75 1 --- 1 0

224
Table 6.11 Inelastic Component, “Moment Hinge, Rotation Type”, Assigned Values
of Beams for Building in Istanbul for Calibrated and Uncalibrated Alternatives

MOMENT HINGE, IN ISTANBUL

Rotation Type 40/60 BEAM 60/60 BEAM
Uncalibrated Calibrated Uncalibrated Calibrated
(Trilinear Model)
(A1) (A2) (A3) (A1) (A2) (A3)
1 FY (kN.m) 875 875 875 1358 1358 1358
Relationship

Pos.-Neg.

2 FU (kN.m) 974 974 974 1222 1222 1222

Basic

3 DU (rad) 0,0178 0,0178 0,0178 0,017 0,017 0,0168

4 DX (rad) 0,1 0,1 0,1 0,1 0,1 0,1
Deformations

5 DL (rad) 0,05 0,05 0,05 0,05 0,05 0,05

Pos.-Neg.
Strength

6 DR (rad) 0,052 0,052 0,052 0,052 0,052 0,052

Loss

7 FR/FU 0,2 0,2 0,2 0,2 0,2 0,2

All 8 SLI 0 0 0 0 0 0
9 Y --- 0,2 0,3 --- 0,3 0,3
Degradation

10 U --- 0,2 0,25 --- 0,25 0,25

Pos.-Neg.
Cyclic

11 L --- 0,15 0,22 --- 0,22 0,22

12 R --- 0,15 0,2 --- 0,2 0,2
13 X --- 0,15 0,2 --- 0,2 0,2
14 USF --- 0,75 1 --- 1 1

Table 6.12 Inelastic Component, “Shear Hinge, Displacement Type”, Assigned

Values of Beams for Building in Istanbul for Calibrated and Uncalibrated
Alternatives

Shear Hinge, IN ISTANBUL

Displacement Type 60/150 BEAM
Uncalibrated Calibrated
E-P-P (Elastic Perfectly Plastic)
(A1) (A2) (A3)
Relationship

1 FY (kN) --- --- ---

Pos.-Neg.

2 FU (kN) 3844 3844 3844

Basic

3 DU (mm) --- --- ---

4 DX (mm) 180 180 180
Deformati

5 DL (mm) 78 (0.06 ln) 78 (0.06 ln) 78 (0.06 ln)

Strength

Pos.-
Neg.
Loss

6 DR (mm) 130 (0.1 ln) 130 (0.1 ln) 130 (0.1 ln)
ons

7 FR/FU 0,3 0,3 0,3

All 8 SLI 0 0 0
9 Y --- 0,5 1
Degradation

Pos.-Neg.

10 U --- 0,3 0,65

Cyclic

11 L --- 0,3 0,65

12 R --- 0,3 0,65
13 X --- 0,3 0,65

225
6.3 Performance Based Design Results of the Case Study Buildings

All performance-based design results for a total of 6 models are given with
comparisons along with three alternatives, A1, A2 and A3, the parameters of which
were given in the tables before in detail for the walls and beams for two different
earthquake zones, i.e., Ankara and Istanbul.

Case study building has a flat plate system. In Figure 6.18 and Figure 6.19, the extent
which structural members contribute to lateral resistance for Ankara and Istanbul
buildings respectively. Accordingly, in all models, results are compatible with linear
analysis and walls meet shear forces of 90% and columns share a level of 10%.

Figure 6.18 Story Shear Load Distribution for Columns & Walls through X and Y
Direction (Ankara)

226
Figure 6.19 Story Shear Load Distribution for Columns & Walls through X and Y
Direction (Istanbul)

6.3.1 Interstory Drift Ratios

Interstory drift ratios are presented in Figure 6.20 and Figure 6.21 for both regions.
The drift ratios are approximately 0.005 and 0.01 for Ankara and Istabul,
respectively. These values fall below the mean maximum value of 0.03 according to
TEC-2018. Drift check of the calibrated model (A3) including 7 earthquake results
is also presented in Figure 6.24 and Figure 6.25. Accordingly, the maximum drift
limit applicable to any individual earthquake is not above 0.045. The result
difference between the calibrated A3 and non-calibrated (A1-A2) models is seen by
numerical comparison in Figure 6.22 and Figure 6.23. Accordingly, there seems to
be more difference through X direction in which models show more ductile behavior.
In the Y direction, the drift results between the models are almost the same. The
uncalibrated model (A1), in which not taking into account energy reduction factors
and unloading stiffness factor, and the other two models (A2; A3), A1 model has

227
less than 20 % difference in the drift ratios compared to the other models in the X
direction where ductile behavior is active. The A2 model ,which has a lower energy
degradation factor than the A3-calibrated model (half the order of the A3), and the
A3 model, gave almost the same average drift results in both directions for both
regions.

The approximate mean drift level of 0.005 and 0.01 through all stories respectively
in Ankara and Istabul is important in determining the multiplication factor, “b” in
Equation 5.2 at Chapter 5 proposed by us for the concrete strain evaluation of the
wall structures.

Figure 6.20 X and Y Direction Drift Check (<0.03) for Three Model Alternatives
(A1, A2, A3) (Ankara)

228
Figure 6.21 X and Y Direction Drift Check (<0.03) for Three Model Alternatives
(A1, A2, A3) (Istanbul)

Figure 6.22 Calibrated (A3) and Uncalibrated Models (A1 & A2) Drift Ratio
Comparisons for X & Y Direction (Ankara)

229
 Figure 6.23 Calibrated (A3) and Uncalibrated Models (A1 & A2) Drift Ratio
Comparisons for X & Y Direction (Istanbul)

Figure 6.24 Calibrated Model (A3) Mean Drift Check (<0.03-mean; <0.045-max)
with 7 Earthquakes for X and Y Direction (Ankara)

230
Figure 6.25 Calibrated Model (A3) Mean Drift Check (<0.03-mean; <0.045-max)
with 7 Earthquakes for X and Y Direction (Istanbul)

6.3.2 Shear Force in the Walls

The shear force demand/capacity (D/C ratios) results of the walls are given in Figure
6.26 and Figure 6.27. Accordingly, as expected, the walls are forced under shear
along the critical height from the basement. In addition, according to the wall labels
in Figure 6.28, three different model alternatives for each wall and demand/capacity
(D/C) ratios for each wall are given between Figure 6.29 and Figure 6.32 for each
wall. Within both earthquake regions, there are some walls with D/C ratios exceeding
1. By about 5 % in Ankara building, while it was about 20 % with a maximum of 50
% in Istanbul. For the case study building in Ankara, W17, W20, W21 walls in the
Y direction exceed the maximum shear stress (0.85√𝑓ck ) at the order of 5%. For case
study building in Istanbul, W10, W11, W15 and W16 walls in the X direction exceed
the limit by up to 50 %, while the W19, W20, W21 and W22 walls in the Y direction
exceed the limit up to 20 % (Figure 6.28).

231
The wall horizontal reinforcement designs carried out according to linear analysis in
both earthquake zones were insufficient compared to nonlinear analysis. In
Appendices A and B, the wall capacity and demand curves along all floors separately
for Ankara and Istanbul are given in comparison with linear analysis capacity curves.
Linear capacity curves are the calculations of the walls horizontal reinforcements
obtained as a result of linear analysis according to the expected material values.

According to figures between Figure 6.29 and Figure 6.32, the D/C ratios of Model
A1 are lower than those of A2 and A3. Avoiding the energy degradation factors in
the nonlinear analysis tends to results in unsafe demand estimations for shear force
design. The use of low values, as in the A2 model, gives safe results for the design
of the walls under shear force. The D/C ratios of the calibrated A3 model and the
non-calibrated A2 model are similar.

Figure 6.26 Shear Load Demandmean,max/Capacity Ratio Check for Three

Alternative Models for All Walls (Ankara)

232
 Figure 6.27 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for All Walls (Istanbul)

Figure 6.28 Wall Labels and Walls Shear Load Pass Max. Shear Capacity (Red
Circle: in Ankara; Orange Circle: in Istanbul)

233
Figure 6.29 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for W1-W11 (Ankara)

Figure 6.30 Shear Load Demandmean,max/Capacity Ratio Check for Three

Alternative Models for W12-W22 (Ankara)

234
Figure 6.31 Shear Load Demandmean,max/Capacity Ratio Check for Three
Alternative Models for W1-W11 (Istanbul)

Figure 6.32 Shear Load Demandmean,max/Capacity Ratio Check for Three

Alternative Models for W12-W22 (Istanbul)

235
6.3.3 Strain Demands

The reinforcement tensile and concrete compression average strain values of the
walls are measured and checked with 52 strain gages shown in Figure 6.33. In the
concrete compression strain measurements from the strain gage in Figure 6.33, the
"unconfined" concrete limit values are taken into consideration with strain gages in
red circle, while the orange one’s measure concrete compression strains of the
"rectangular confined boundary" sections. The remaining strains gages of the red and
orange circles are classified as strain gages of "flanged confined boundary" sections.
In the figures between Figure 6.34 and Figure 6.39, Dmean,max/Capacity (D/C) values
are given for all layers for strain gage values for concrete compressions in 3 different
categories as "unconfined", "rectangular confined boundary" and "flanged confined
boundary" for both earthquake zones and 3 model alternatives. “Unconfined”,
“rectangular confined boundary” and “flanged confined boundary” D/C ratios are
maximum 0.3, 0.2, 0.2 respectively.

Figure 6.33 Strain Gage Labels of Core-Wall

236
Figure 6.34 Unconfined Parts Mean Concrete Strain D/C Ratio of Walls for Three
Model Alternatives (Ankara)

Figure 6.35 Unconfined Parts Mean Concrete Strain D/C Ratio of Walls for Three
Model Alternatives (Istanbul)

237
Figure 6.36 Confined Parts (Rectangular Boundary) Mean Concrete Strain D/C
Ratio of Walls for Three Model Alternatives (Ankara)

Figure 6.37 Confined Parts (Rectangular Boundary) Mean Concrete Strain D/C
Ratio of Walls for Three Model Alternatives (Istanbul)

238
Figure 6.38 Confined Parts (Flange Boundary) Mean Concrete Strain D/C Ratio of
Walls for Three Model Alternatives (Ankara)

Figure 6.39 Confined Parts (Flange Boundary) Mean Concrete Strain D/C Ratio of
Walls for Three Model Alternatives (Istanbul)

In the calibration study conducted in Chapter 5, modeling with Perform 3D showed

that the concrete strain results are much lower than the experimental results.
According to the value of "a" in Equation 6.3, Figure 5.59 shows the variability of
the value "a" according to drift ratios.

a=c,experiment / c,P3D model (6.3)

239
When evaluated according to this calibration results, it is proposed according to
Figure 5.59 to multiply the P3D model concrete strain results by this coefficient “a”
and "a" value should be taken "2" for the "rectangular confined" sections, as in the
RW2 test sample, while in the case of a more distributed compression stress of the
compression zone in the "flanged" form, such as TW2, it is recommended to use an
increase coefficient between "4-16" with the value of "a" according to the structure
drift ratio. While the dominant average drift rate in Ankara is 0.005, drift ratio
demand for case study building in Istanbul is almost 0.01. Accordingly, the
"unconfined" and "flanged confined" concrete sections concrete strains are
magnified by multiplying the model strain gage values by the coefficient "a" as “4”
and “6” in Ankara and in Istanbul respectively. Between Figure 6.40 and Figure 6.45,
both earthquake zones and only calibrated A3 model concrete strain values are
increased by the coefficient "a" and the demand/capacity (D/C) values along the
stories are shown. Although the assumption that there is the same drift ratio on all
floors in the magnification with the same coefficient "a" is not exactly valid for the
basement floors, but it is accepted as such for convenience in terms of observation.
However, what should happen is to determine and apply the correction coefficient
"a" according to the story drift ratio determined according to Figure 5.59.

Figure 6.40 Proposed (Multiplied by “a=4”) & Available Calibrated Unconfined

Parts Mean Concrete Strain D/C Ratio of Walls (Ankara)

240
Figure 6.41 Proposed (Multiplied by “a=6”) & Available Calibrated Unconfined
Parts Mean Concrete Strain D/C Ratio of Walls (Istanbul)

Figure 6.42 Proposed (Multiplied by “a=2”) & Available Calibrated Confined

Parts (Rectangular Boundary) Mean Concrete Strain D/C Ratio of Walls (Ankara)

241
 Figure 6.43 Proposed (Multiplied by “a=2”) & Available Calibrated Confined
Parts (Rectangular Boundary) Mean Concrete Strain D/C Ratio of Walls (Istanbul)

Figure 6.44 Proposed (Multiplied by “a=4”) & Available Calibrated Confined

Parts (Flange Boundary) Mean Concrete Strain D/C Ratio of Walls (Ankara)

242
 Figure 6.45 Proposed (Multiplied by “a=6”) & Available Calibrated Confined
Parts (Flange Boundary) Mean Concrete Strain D/C Ratio of Walls (Istanbul)

Both concrete and reinforcement strains for all strain gages are shown throughout all
stories in Appendices C and D according to the proposed "a" coefficient correction
for the case study building in Ankara and Istanbul respectively. As a result of the
increases with the coefficient "a", concrete strain values of some strain gages
exceeded the limit values in both Ankara and Istanbul. Since the limit value is lower
in the "unconfined" sections than “confined” concrete parts, “unconfined” concrete
fibers are more critical. So, in the building in Ankara, the concrete strain values
multiplied with the "a" coefficient exceed the limit (Figure 6.46) while the concrete
strain in other "unconfined" concrete sections approaches the limit strain values.
Although the concrete strain is increased in the confined sections for the structure in
Ankara, the D/C ratio is maximum 0.4 for "rectangular boundary" fibers and 0.6 for
"flanged boundary" fibers.

The concrete strains at 16., 17., 19. and 35. strain gages of the "unconfined" concrete
sections for the building in Istanbul pass the limit after the use of the coefficient "a”.
On the other hand, the concrete strains at 5. and 9. strain gages of the "flanged
confined" concrete sections pass the limit after the use of the proposed "a"
coefficient. Strain gages exceeding concrete strain limits after the use of coefficient
"a" are shown in Figure 6.46.

243
 Figure 6.46 Strain Gages Exceed Concrete Strain Limits after Usage of “a”
Proposed Multiplication (Green Circle: In Istanbul; Blue Circle: In Ankara)

Wall reinforcement strain D/C ratios are around 0.15 for the building in Ankara
and 0.2-0.25 for the structure in Istanbul (Figure 6.47 & Fiure 6.48).

Figure 6.47 Mean Reinforcement Strain D/C Ratio for Three Alternative Models
in Ankara

244
 Figure 6.48 Mean Reinforcement Strain D/C Ratio for Three Alternative Models
in Istanbul

Concrete and reinforcement strain value comparisons are made along the building
height between 3 alternative models for wall structural member in the graphs
between Figure 6.49 and Figure 6.56. Accordingly, in both concrete and
reinforcement strains, the difference between the calibrated A3 model and the
uncalibrated A1 model seems to be much greater than the difference between the
calibrated A3 model and the uncalibrated A2 model. A2 model results are 5-10 %
higher than A3 model values in both concrete and reinforcement strains. The much
lower intake of energy degradation factors gives results that remain on the safe side
of the design in terms of vertical strain results. On the other hand, in the A1 model,
where energy degradation factors are not taken into consideration at all, it has given
less results in the concrete strain, up to 40 % in the reinforcement strain and up to
15-20 % in concrete strain compared to the calibrated A3 model, and has given lesser
results that are not on the safe side in terms of design.

245
Figure 6.49 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))
and Uncalibrated (A1-(U)) Models (Ankara)

Figure 6.50 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))

and Uncalibrated (A1-(U)) Models (Istanbul)

246
Figure 6.51 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))
and Uncalibrated (A2-(U)) Models (Ankara)

Figure 6.52 Reinforcement Strain Comparisons of Walls for Calibrated (A3-(C))

and Uncalibrated (A2-(U)) Models (Istanbul)

247
Figure 6.53 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A1-(U)) Models (Ankara)

Figure 6.54 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A1-(U)) Models (Istanbul)

248
 Figure 6.55 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A2-(U)) Models (Ankara)

Figure 6.56 Concrete Strain Comparisons of Walls for Calibrated (A3-(C)) and
Uncalibrated (A2-(U)) Models (Istanbul)

6.3.4 Rotation Demand in Beams

The beams named between B1 and B9 were shown in Figure 6.57. The B3 and B9
link beams have 150 cm depth, while the remaining beams are 60 cm deep. Instead
of check of beam rotation limit for the diagonally detailled link beam in the building
in Istanbul, the shear displacement inelastic behavior was checked. For both

249
earthquake zones between Figure 6.58 and Figure 6.61, the maximum average
rotations of all beams over the building height except the diagonal reinforced link
beams were given for thre calibrated and uncalibrated alternatives. Accordingly,
there was no beam exceeding the collapse prevention (CP) rotation limits. In other
words, the rotation of all beams was below the limits. It should be noted that the
rotation of link beams (40/150) in Ankara were very small values on the upper floors.

Among the A1, A2 and A3 calibrated and uncalibrated models, the A2 uncalibrated
model with the smallest energy degradation factor values gave the highest rotation
values, while the A1 uncalibrated model gave the lowest rotation values, resulting in
unsafe design results. The numerical comparison between the models was given
between Figure 6.62 and Figure 6.69. The rotation result difference between the A1
non-calibrated model and the A3 calibrated model beams reached 30 %, while the
beam rotation result difference between A2 and A3 models did not exceed 5 %.

Figure 6.57 Label of Beams

250
Figure 6.58 Beam Rotation Check (<0.041 (B40-60); <0,02 (B40-150)) for Three
Model Alternatives between B1 & B6 (Ankara)

251
Figure 6.59 Beam Rotation Check (<0.041 (B40-60); <0,02 (B40-150)) for Three
Model Alternatives between B7 & B9 (Ankara)

252
Figure 6.60 Beam Rotation Check (<0.046 (B40-60); <0.046 (B60-60); <0,02
(B40-150)) for Three Model Alternatives between B1 & B7 (Istanbul)

253
Figure 6.61 Beam Rotation Check (<0.046 (B60-60)) for Three Model
Alternatives for B8 (Istanbul)

Figure 6.62 Mean Rotation D/C Ratio of Beams (Ankara)

254
 Figure 6.63 Mean Rotation D/C Ratio of Beams (Istanbul)

Figure 6.64 Calibrated (A3) and Uncalibrated (A1) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A1(U) (Ankara)

255
Figure 6.65 Calibrated (A3) and Uncalibrated (A1) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A1(U) (Istanbul)

Figure 6.66 Calibrated (A3) and Uncalibrated (A2) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A2(U) (Ankara)

256
Figure 6.67 Calibrated (A3) and Uncalibrated (A2) Beam (40/60) Rotation
Comparison Ratio, A3-(C)/A2(U) (Istanbul)

Figure 6.68 Calibrated (A3) and Uncalibrated (A1) Beam (60/60) Rotation
Comparison Ratio, A3-(C)/A1(U) (Istanbul)

257
 Figure 6.69 Calibrated (A3) and Uncalibrated (A2) Beam (60/60) Rotation
Comparison Ratio, A3-(C)/A2 (U) (Istanbul)

6.3.5 Performance of Diagonally Reinforced Link Beams

Diagonally reinforced link beam shear force and shear displacement results are given
in Figure 6.70 and Figure 6.71. According to these results, beam shear displacements
are far from the maximum collapse prevention (CP) limit of 39 mm (0.03 ln).
According to this limit, the demand/capacity (D/C) ratios are 0.04, 0.1 and 0.08 for
the A1, A2 and A3 models, respectively. The shear forces are almost at the inelastic
limit. The A2 and A3 model shear displacements are almost the same, with a
maximum difference of 8 % more than the A1 model displacements.

258
Figure 6.70 Shear Load Capacity Check (<3844 kN) of Diagonally Reinforced
Link Beam (Istanbul)

Figure 6.71 Shear Displacement Check (<39 mm (0.03 ln)) of Diagonally

Reinforced Link Beam (Istanbul)

259
6.3.6 Shear Load Capacity Check of Beams

In terms of the design of frame beam and link beams under the shear force, the
Demandmean, max/Capacity (D/C) ratios throughout stories are shown between Figure
6.72 and Figure 6.75. The D/C ratios of all the 60 cm deep beams in Ankara, except
for conventionally reinforced link beams, are between 0.4-0.6. For the link beams
(40/150 ) in Ankara, the shear laod D/C ratios exceeded limits and the preliminary
design stirrups obtained as a result of linear analysis are insufficient.

Numerical comparisons were made between the three model alternatives,i.e., A1, A2
and A3, in terms of the beam shear load D/C ratios. The D/C ratio results were almost
the same for all of beams with a height of 60 cm where shear force is not effective.
However, for the conventionally reinforced link beam where shear force was active
in the case study building in Ankara, the A1 uncalibrated model had 15-20 % less
demand than results of the A3 calibrated model. The shear load D/C ratios of link
beams (40/150) in the A2 non-calibration model and the A3 uncalibrated model were
almost the same. Numerical comparisons between models can be observed in figures
between Figure 6.76 and Figure 6.79.

Figure 6.72 Beam Shear Load (D/C) Check between B1 & B3 (Ankara)

260
Figure 6.73 Beam Shear Load (D/C) Check between B1 & B6 (Ankara)

261
Figure 6.74 Beam Shear Load (D/C) Check between B1 & B7 (Istanbul)

262
 Figure 6.75 Beam Shear Load (D/C) Check for B8 (Istanbul)

Figure 6.76 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (40/60) Shear
load D/C Ratio Comparison (Ankara)

263
 Figure 6.77 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (40/150)
Shear load D/C Ratio Comparison (Ankara)

Figure 6.78 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (40/60) Shear
load D/C Ratio Comparison (Istanbul)

264
Figure 6.79 Calibrated (A3) & Uncalibrated (A1-A2) Model Beams (60/60) Shear
load D/C Ratio Comparison (Istanbul)

6.3.7 Energy Dissipation Results

The dissipationf of earthquake energy provides engineers important information in

terms of the building behavior. Building in both Ankara and Istanbul, 55-60 % of the
total earthquake energy is dissipated through "damping", while 9-10 % is dissipated
as “kinetic energy”. The "strain enegy" parts resulted from elastic displacement and
"inelastic energy" dissipated by inelastic behavior differ as expected for buildings in
Ankara and Istanbul. While dissipated inelastic energy is at the level of 8-10 % for
the building in Ankara, this rate increases to about twice as high as 16-20 % in
Istanbul. On the contrary, "strain energy" is around 12 % in the structure in Istanbul,
while this value is around 24 % in the structure in Ankara. These expected results
are shown in Figure 6.80 and Figure 6.81 for three different alternative models and
two different earthquake zones. The energy ratio changes of the A3 (calibrated)
model for 14 different earthquakes are also shown in Figure 6.82 and Figure 6.83.

265
Figure 6.80 Energy Dissipation Comparisons of Calibrated (A3) & Uncalibrated
(A1-A2) Models (Ankara)

Figure 6.81 Energy Dissipation Comparisons of Calibrated (A3) & Uncalibrated

(A1-A2) Models (Istanbul)

266
Figure 6.82 Energy Dissipation Distribution of 14 Earthquakes for Calibrated
Model (A3) (Ankara)

Figure 6.83 Energy Dissipation Distribution of 14 Earthquakes for Calibrated

Model (A3) (Istanbul)

267
In the case study building examined, there are two basic types of structural member
types that meet the earthquake force, which are wall and beams. The sharing of the
dissipated inelastic energy between these two main structural member groups are
given in Figure 6.84 and Figure 6.85 for three different model alternatives and two
different earthquake zones. While 94 % and 6 % of the inelastic energy in the
structure in Ankara is dissipated in the beams and the walls respectively, 80 % and
20 % of the inelastic energy in the building in Istanbul is shared in beams and core-
wall, respectively.

While the A2 and A3 models give similar results in all energy sharing, the "inelastic
dissipated energy" values in the A1 model are 33 % higher than in A2 and A3
models. In addition, in the A1 model, the “inelastic dissipated energy” is less in the
walls and more in the beams than in the A2 and A3 models.

Figure 6.84 Mean Dissipated Inelastic Energy Distribution between Beams and
Walls for Three Model Alternatives (Ankara)

268
 Figure 6.85 Mean Dissipated Inelastic Energy Distribution between Beams and
Walls for Three Model Alternatives (Istanbul)

6.3.8 Summary of Performance Based Design of the Case Study Building

Case study building performance analysis results are given according to the topics
and summary information for both earthquake zones in Table 6.14. More succinctly,
Table 6.13 shows the insufficient design subjects. According to these, the wall
horizontal reinforcements in both earthquake zones are insufficient according to the
linear analysis. In both regions, there are walls exceeding maximum shear capacities
and it is necessary to change the wall dimensions. According to the designs for the
conventionally reinforced link beams in Ankara under shear force, the amount of
stirrup should be increased. As for the concrete strain limit, when no increase is
considered as a result of the P3D solution, all concrete strains are found sufficient
for both regions. However, if the proposed "a" multiplication coefficient is used,
fiber sections with insufficient concrete strain results are produced. Walls exceeding
the maximum shear capacity, i.e. requiring a change in size, and fiber positions

269
exceeding the concrete strain limit as a result of the use of the "a" proposed
coefficient are shown in Figure 6.86 on the same figure for both earthquake zones.

Table 6.15 provides numerical comparison results for the calibrated A3 model and
uncalibrated A1 and A2 models. The table summarizes the ratios of the mean
maximum D/C values between the models and summarizes the variations of the A1
and A2 uncalibrated values according to the calibrated A3 model. Accordingly,
although the majority of all results are almost the same between the A2 model, which
has very small energy degardation factors (0.2-0.3), and the calibrated A3 model,
whose energy factors are medium (0.5-0.6), the A2 results give 5-10 % more results,
pushing the limits more to give safe results in design. There is a difference of up to
10 % to 40 % between the results of the A1 uncalibrated model without the energy
degardation factor (1.0) and the A3 calibrated model, giving results that are further
smaller than the limits. Accordingly, it can be stated that energy degardation factor
must be considered in performance based design.

270
 Figure 6.86 Walls Having Insufficient Shear Capacity (Orange Rectangle: in
Istanbul; Red Rectangle: in Ankara) and Strain Fibers Passing the Limits after
Using Proposed Multiplication factor “a” (Orange Circle: in Istanbul; Red Circle:
in Ankara)

Table 6.13 PBD Summary Result Sufficiency Situation of Case Study Building

PBD Result Sufficiency

Ankara Istanbul
Drift Check  
Shear Check  
Concrete Strain  
Walls
Proposed Concrete Strain  
Reinforcement Strain  
Rotation Check (40/60)  
Rotation Check (60/60) -- 
Rotation Check (40/150)  --
Shear Displacement Check
Beams -- 
(60/150)
Shear Check (40/60)  
Shear Check (60/60) -- 
Shear Check (40/150)  --

271
Table 6.14 Summary of Performance Based Design of Case Study Building

SUBJECTS ANKARA ISTANBUL

Drift Check General story mean nonlinear drift ratio 0.005 0.01
# of walls whose linear analysis horizontal
21/22 22/22
Walls reinforcements revised after PBD
Shear Design # of walls pass max. shear capacity after PBD 3/22 8/22
Max Shear D/C Ratio of Walls 1.04 1.2-1.5
# of reinforcement strain gages passes the
none none
limit
Max. mean reinforcement strain D/C ratio 0.15 0.2-0.25
Max. mean unconfined concrete strain D/C
0.3 0.3
ratio
Max. mean confined rectangular fibers
0.2 0.2
concrete strain D/C ratio
Max. mean confined flange fibers concrete
0.2 0.2
strain D/C ratio
Max. mean unconfined concrete strain D/C
ratio with usage of "a" proposed 1.03 1.17
multiplication factor
Max. mean rectangular fibers concrete
Walls
strain D/C ratio with usage of "a" proposed 0.4 0.5
Strain Check
multiplication factor
Max. mean flanged fibers concrete strain
D/C ratio with usage of "a" proposed 0.6 1.1
multiplication factor
# of unconfined concrete strain gages passes
the limit with usage of "a" proposed 1 4
multiplication factor
# of confined rectangular concrete strain
gages passes the limit with usage of "a" none none
proposed multiplication factor
# of confined flanged concrete strain gages
passes the limit with usage of "a" proposed none 2
multiplication factor
# of beams pass max. rotation capacity
Beam none none
after PBD
Rotation Check
Max Mean Rotation D/C Ratio of Beams 0.35 0.55
# of beams with 60 cm height pass max. Shear
none none
Beam capacity after PBD
Shear Check Max Mean Shear D/C Ratio of Beams with
0.6 0.4
60 cm Height
# of link beams with 150 cm height pass max.
2/2 none
Link Beam shear capacity after PBD
Shear Check Max Mean Shear capacity or displacement
1.3 0.1
D/C Ratio of Link Beams with 150 cm Height
Mean Inelastic Energy Sharing (%) 8% 16%
Mean Strain Energy Sharing (%) 25% 12.5%
Earthquake
Mean Damping Energy Sharing (%) 58% 62%
Energy
Mean Kinetic Energy Sharing (%) 9% 9%
Dissipation
Inelastic Energy Dissipation at Beams 94% 80%
Inelastic Energy Dissipation at Walls 6% 20%

272
 Table 6.15 Mean Maximum Results Comparison Ratios of Calibrated (A3) and
Uncalibrated Models (A1 &A2) for Case Study Building in Ankara & Istanbul

ANKARA ISTANBUL
A3-(C) /A1-(U) A3-(C) /A2-(U) A3-(C) /A1-(U) A3-(C) /A2-(U)
min. max. mean min. max. mean min. max. mean min. max. mean
Drift
1.02 1.18 1.10 0.98 0.99 0.99 0.99 1.16 1.08 0.98 0.99 0.99
Check
Shear
0.95 1.11 1.03 0.97 1.00 0.99 0.93 1.17 1.05 0.96 1.00 0.98
Check
Walls

Concrete
0.90 1.18 1.04 0.83 1.08 0.96 0.88 1.19 1.04 0.93 1.08 1.01
Strain
Rein.
0.80 1.37 1.09 0.93 1.02 0.98 0.91 1.36 1.14 0.94 1.02 0.98
Strain
Rotation
Check 0.98 1.32 1.15 0.96 1.00 0.98 1.09 1.28 1.19 0.97 0.99 0.98
(40/60)
Rotation
Check -- -- -- -- -- -- 1.09 1.35 1.22 0.96 0.99 0.975
(60/60)
Rotation
Check 0.98 1.19 1.09 0.96 1.03 1.00 -- -- -- -- -- --
(40/150)
Shear
Beams

Displa.t
-- -- -- -- -- -- 0.70 1.00 0.85 1.00 1.39 1.20
Check
(60/150)
Shear
Check 1.00 1.02 1.01 1.00 1.00 1.00 1.00 1.01 1.01 1.00 1.00 1.00
(40/60)
Shear
Check -- -- -- -- -- -- 1.00 1.01 1.01 1.00 1.00 1.00
(60/60)
Shear
Check 0.98 1.18 1.08 0.97 1.00 0.99 -- -- -- -- -- --
(40/150)
Dissipated
Inelastic -- -- 0.73 -- -- 1.18 -- -- 0.80 -- -- 1.1
Dissipation

Energy
Energy

Strain
-- -- 1.02 -- -- 1.00 -- -- 1.04 -- -- 1
Energy
Damping
-- -- 1.02 -- -- 0.98 -- -- 1.03 -- -- 0.98
Energy

273
274
 CHAPTER 7

7 RESULTS AND INFERENCES OF THE THESIS WORK

Within the scope of this thesis work, both linear elastic design and performance-
based designs of the case study tall building for two different seismic zones (in
Ankara and Istanbul) were conducted according to TEC-2018 rules. With the linear
elastic design, core-walls and beam dimensions were determined and all
reinforcement details were calculated. Drift checks were made according to linear
elastic analysis. The linear elastic analysis design results are detailed in Chapter 3.
After linear elastic analysis of the case study tall building, selected load-deformation
cyclic response experiments for wall and link beams were simulated in CSI Perform-
3D program and calibration work was performed. With the calibration study,
accurate modeling methods and appropriate P3D software parameter values were
obtained. The sensitivity of the parameters is determined. Cyclic degradation factors,
i.e., energy degradation factors, have been found to be the most effective parameters
in the results. After the calibration study, performance-based designs of the case
study buildings were carried out for two different earthquake zones. Strains were
checked for walls and their designs were completed under shear force. Rotations
were checked for the beams and their designs were also made under the shear force.
Thus, differences between linear elastic design and performance-based nonlinear
design results of the case study buildings in two different seismic regions (in Ankara
and Istanbul) were observed. In addition, while performing performance-based
designs of case study tall buildings, three different cyclic degradation factors
including calibration values were selected and six different nonlinear models were
created with two different earthquake zones. The results of the nonlinear design of
models with three different cyclic degradation factors were compared. The PBD
results of the A3 model with calibrated cyclic degradation factor and the models (A1,
A2) with uncalibrated cyclic degradation factor were compared. In a performance-

275
based analysis in which calibration is not taken into account, the extent to which the
results deviated was determined. The detailed conclusions of the thesis study are
presented below in articles:

 Both linear and nonlinear performance analysis of the case study tall building
was performed for two different seismic zones (Ankara and Istanbul)
according to the Turkish Earthquake Code-2018 (TEC-2018). The
preliminary design results of linear analysis in both regions were different
and linear elastic design was found to be insufficient for both buildings. The
nonlinear analysis of the case study building led to results requiring to change
size of some of the walls in both earthquake zones because of insufficient
shear capacity. Also, it was necessary to increase the horizontal
reinforcements for the walls that did not require size changes for both
earthquake zones. Performance based design of case study building in this
thesis showed that nonlinear analysis is critical for the final design of walls.
 The design of the walls under shear force in the performance-based design in
the case study building (CSB) has resulted in higher reinforcement amount
than those obtained from linear analysis. Results are summarized in Table
7.1 and Table 7.2 respectively. Comparison of design base-shears for two
different design approach can also be observed in Table 7.3 with base-shear
demand ratios. While shear forces were increased with a dynamic
magnification coefficient of “2.5” for linear analysis (Table 7.1), shear forces
were increased with magnification coefficient “1.2-1.5” for the nonlinear
analysis (Table 7.2) according to standard deviation. Design demand base-
shear ratio of nonlinear analysis with respect to linear analysis was between
“1.7” to “4.3” for case study tall building (Table 7.3). Although the spectral
acceleration ratio of DD1 to DD2 was almost “2” for both seismic regions,
base-shear force ratios without any correction or amplification were between
“9” to”15”. (Table 7.3). Shear forces from modal analysis with cracked
member properties is lower than the mean base shear results of the nonlinear
time history analysis of tall buildings. Period, stiffness, rigidity and behavior

276
 of tall building continuously changes from initial to end time point of time
history analysis. At nonlinear design approach, expected material strength of
concrete and reinforcement was used. As an experience and trials, shear
capacity increase for nonlinear design value was almost %50 percent
compared to the shear force demand from linear analysis according to TEC-
2018 clauses. However, the base-shear design demand increase was between
“1.7” to “4.3”. This shows that in high-rise buildings, if the demand/capacity
(D/C) ratio is around to 1.0, it is likely that the horizontal reinforcements in
the walls will change or walls sizes will be insufficient and require redesign
based on the nonlinear design. Table 7.4 shows change and difference for
both design stage in walls design under shear load for both seismic regions.
Insufficient excessive shear capacity of walls at linear design stage led to the
change of horizontal reinforcement or size of walls in both seismic regions
after performance-based design of case study building. Based on this result,
it can be recommended to design the walls of tall structures by limiting the
demand/capacity (D/C) ratio not exceeding “0.5”. In addition, the design of
wall members with a constant shear rigidity reduction coefficient (0.5 G) in
the plane behavior according to TEC-2018 seems questionable. The cracked
rigidity values to compute shear demands should be further investigated in
the nonlinear analysis stage to yield more consistent results with the linear
elastic design stage. The choice of shear modulus rigidity reduction factor for
walls between 0.1-0.5G, incline with the experimental observations may
allow safer and suitable design.

Table 7.1 Linear Analysis Base-shear Values over Basement of CSB

Ankara (ton) Istanbul (ton)

Linear Analysis Base-shear over Basement X Y X Y
L-V1 Linear Modal Analysis Story Shear Force 461 791 1221 2277
L-V2 Minimum Limit Story Shear Force 1473 1473 3679 3679
L-Vd Increased Design Shear Force (x 2.5) 3683 3683 9198 9198

277
 Table 7.2 Non-linear Analysis Base-shear Values over Basement of CSB

Ankara (ton) Istanbul (ton)

Nonlinear Analysis Base-shear
X Y X Y
over Basement (ton)
NL-V1 Mean Nonlinear Story Shear Force 5432 11735 11387 22464
Mean Increased Design Story Shear
NL-Vd 7333 15842 15373 30326
Force (x 1.35)

Table 7.3 Linear and Nonlinear Base-shear Demand Ratio of CSB

Ankara Istanbul
X Y X Y
Design Base-shear Demand Ratio Nonlinear /Linear
2.0 4.3 1.7 3.3
(NL-Vd / L-Vd)
Base-shear Demand Ratio Nonlinear /Linear
11.8 14.8 9.3 9.9
(NL-V1 / L-V1)

Table 7.4 Nonlinear Shear Design Results of Walls after PBD for CSB

Walls Nonlinear Shear Design Results Ankara Istanbul

# of walls whose linear analysis horizontal reinforcements
21/22 22/22
revised after PBD
# of walls pass max. shear capacity after PBD 3/22 8/22
Max Shear D/C Ratio of Walls 1.04 1.2-1.5

 Perform 3D Simulation calibration work was performed to compare with

cyclic response test results for the most important building structural
members, i.e, walls and link beams before the case study building 3D
performance-based design. In this simulation study, the necessary work was
done for the load-deformation cyclic response calibration work for
rectangular, T-shaped and U-shaped walls. In the rectangular and T-shaped
walls, the sensitivity degree of P3D material parameters on the result of
cyclic response was investigated. Accordingly, the cyclic degradation factor

278
 values (YULRX) of the reinforcement and the unloading stiffness factors
(USF) are the sensitive parameters. Reinforcement yield and ultimate
strength values are not very important, but they affect the result to a certain
moderate extent. Incomplete or inadequate definitions in concrete material
definitions have little effect on the cyclic response result. However, the most
effective and dominant effect is the cyclic degradation factor (YULRX)
values of the reinforcement under tension. Taking into account the embedded
bar in concrete value proposed by Belarbi and Hsu (1994) for the
reinforcement yield strength value (0.85-0.9 fsy) is more compatible in terms
of the test result yield strength. Although the effect of this small reduction in
rebar models in 3D building performance-based design on the overall result
is not considered much, a reduction of 10-15% can be achieved in terms of
accurate modeling in the yield strength value in the reinforcement material
model.
 While the ideal material parameters obtained after the sensitivity analysis of
the material parameters and the cyclic response calibration results for the
rectangular and T-shaped walls were satisfactory, however the same success
could not be achieved for U-shaped. This may be because the selected U-
shaped experimental data is not as solid as the rectangular and T-shaped wall
experiment information. It is thought that experimental studies should further
be conducted on non-rectangular walls to perform calibration studies.
 With the rectangular and T-shaped test walls simulation study, successful
calibration results were obtained in cyclic response curves, lateral
displacement and base level reinforcement strain profile results. However,
the P3D base level concrete strain results are much lower than test results due
to the plane section remains plane hypothesis where the other results are
compatible with the test results. For estimating the concrete strain demands
more accurately, the coefficient "a", which expresses the ratio of the test
result to the model result, is proposed (Equation 7.1).
a=c,experiment / c,P3D model (7.1)

279
 The coefficient “a” is approximately "2" in the rectangular boundary section
for drift rates of between 0.5% and 2.0% (Figure 7.1). When a wider "flanged
boundary" condition occurs like the concrete compression section parts in the
T-shaped walls, this coefficient ranges from "4-8" for drifts between 0.5%
and 1.5%.

Design usually
remains in this region

Figure 7.1 TW2 (T-Shaped) and RW2 (Rectangular) Walls Concrete Strain
Experiment-P3D Model Comparison Coefficient “a” vs Drift Ratio Graph

Walls concrete strain results of the 3D case study tall building performance-
based analysis can be multiplied by the proposed "a" coefficient to
realistically estimate the seismic demands. According to the prevailing drift
ratios of 0.5 % and 1.0 % in Ankara and Istanbul, "rectangular boundary"
concrete fiber strains were increased by multiplying "2" and "flanged
boundary" concrete fiber strains were increased by "4" and "6". In both
earthquake zones, the concrete vertical strain values were found to be below
the strain limits, however in the case of using an increase with the proposed

280
 "a" coefficient, there were some concrete strain fibers that exceeded the
limits. These limit violations were especially in unconfined concrete fiber
sections with much lower limits than confined concrete. In the building
model in Istanbul, there were a notable number of concrete strains that
exceeded the limits. Since there has never been a situation that has exceeded
the concrete strain limit among the engineers who are commonly involved in
building design with Perform 3D, the issue of the accuracy of the model
calculation results has always been questioned among them. With numerical
data and calibration study, for drift levels below about 1.2%, a possible
simple and practical value of coefficient “a” is given in Table 7.5.

Table 7.5 Proposed “a” Multiplication Factor for Concrete Fiber Strain

Concrete Fiber Condition a

Confined rectangular boundary concrete fiber 2
Confined flanged boundary concrete fiber (in T-L shape region) drift*+2
Unconfined concrete fiber drift *+2
*
drift in “%”

To demonstrate the appropriateness of the proposed coefficient "a", a strain

profile examination just over basement level was performed on the core-wall
group to observe the values at the strain gage points numbered in Figure 7.2
for case study tall building. Strain profile result check was made for the drift
rate between 0.5-3.0% in the trial study for the structure in Ankara (Figure
7.3). Although there are meshed panel assignments in the wall vertically,
resulting in nearly "plane section remains plane" observed in the group wall
behavior. For the Perform 3D model results, concrete strain values remained
very low away from the test limits in requiring some adjustment in future
studies.

281
Figure 7.2 Check Strain Points of Selected Wall for Y- Direction Push

Figure 7.3 Strain Profile Observation for Selected Strain Points of Wall

282
 This "a" increment coefficient proposed for concrete vertical compression
strain was obtained for wall test samples under compression force "0.07-
0.075 fck Ag". Conducting similar studies in wall test samples under different
axial load stages will allow more accurate recommendations to be made in
this regard.
 Mesh sensitivity work was done for walls. Both vertically and horizontally
were modeled with different mesh options. For models with different
concrete and reinforcement fiber arrangement alternatives made in the plan
for rectangular walls, cyclic response results were compared with ideal model
results. According to the thickness of the walls, the concrete fiber model with
an aspect ratio of 1x1 is ideal for mesh work on the plan. In the wall boundary
sections, reinforcement fibers can be meshed densely as 0.5x1 aspect ratio
reinforcement fiber modeling will be appropriate according to the wall
thickness.
 In the vertical mesh study, the fact that the mesh height is not more than the
floor height is verified as a result of comparing both the base level strain and
cyclic response results of different mesh options with the test results. It is
also verified with test results that "hinge length of walls is minimum of one
half of cross section depth and story height" according to ASCE-41 in the
critical hinge sections of walls. In the case study building model, meshing of
walls was done with the half of the floor heights on the 6.8 m story height at
critical region of building, and so the strain gage length definition also was
taken as the half of the floor heights at this critical region.
 In the Perform 3D model, the results of the fiber modeling options of walls,
i.e., the "auto-size" and "fixed-size" modeling modules in the plan were
compared in terms of both cyclic response and strain profile. The results of
the "auto-size" module, which is easier to model, and the more laborious
"fixed-size" module, have similar results in both cases. Strains results were a
little more in the "auto-size" module, concluding in results that were on the
safer side in terms of design. As a result, the easy-to-use "auto-size" module

283
 of walls in the P3D program can be preferred for the fiber mesh option in the
wall plan.
 Calibration work was also performed in accordance with the selected
experiment for modeling in P3D software both for diagonally and
conventionally reinforced link beam. In addition to the modeling method for
both types of beams, cyclic degradation factor values compatible with the
tests were obtained. Diagonally reinforced link beam was modeled with both
"moment hinge approach" and "shear hinge approach" and calibrated. The
cyclic degradation factors obtained as a result of calibration are given in
summary in Table 7.6 for both beam and walls reinforcement material. In
addition, the cyclic degradation factors recommended in the report of Pacific
Earthquake Engineering Research Center, "Case Studies of the Seismic
Performance of Tall Buildings Designed by Alternative Means" were added.
Calibration study results and report recommendation seem to be compatible.
In the wall structural member, the cyclic energy degradation factor under
compression of concrete material and reinforcement are not very effective in
the results and their sensitivities are low.

Table 7.6 Cyclic Degradation Factors after Calibration Work and

Recommendation Values at PEER Center Report (Moehle et al., 2011)

PROPOSED CYCLIC DEGRATION FACTORS

CALIBRATION WORK PEER CENTER REPORT
Diagonally Diagonally
Walls Reinforced Link Walls Reinforced
Beam Frame Link Beam Frame
Shear Moment Beam Shear Beam
Reinfor. Hinge Hinge Reinfor. Hinge
Approach Approach Approach
Y 0.65 1 0.5 0.3 0.7 0.5 0.24
U 0.65 0.65 0.5 0.25 0.68 0.45 0.23
L 0.65 0.65 0.5 0.22 0.64 0.4 0.22
R 0.65 0.65 0.5 0.2 0.62 0.35 0.21
X 0.65 0.65 0.5 0.2 0.6 0.35 0.2
USF -0.5 0 0 1 0 0 0

284
 After obtaining the proper modeling methods and cyclic degradation factors
as a result of calibration, in order to observe the extent to which these factors
affect the 3D Case study building performance-based design results, A1, A2,
A3 models with 3 different energy degradation values were analyzed in 2
different earthquake zones and the design results of six models were
compared (Figure 7.4).

Figure 7.4 Cyclic Degradation Factor Alternatives for P3D Models

A1 and A2 are used with non-calibrated values in the case study building
models, while calibrated values are used in A3. The summary values used are
shown in Table 7.7. Table 7.8 shows the comparison of the performance-
based design results of the models. Accordingly, cyclic energy degradation
factors must be used in the models. The difference observed to be up to 30 %
between the A1 model not include energy factors than the other models (A2,
A3). A1 model results were less and unsafe side than the other models (A2,
A3). The difference between the A3 model with medium cyclic energy
degradation values, and the A2 uncalibrated model with low cyclic energy
degradation values was between 0.25-0.35 in half order of the A3, which
causes more energy reduction, is not very much. However, A2 model results
remains on the safe side of the design by giving 5-10 % more D/C results
than A3 model. As a results, in the performance analysis, it is more
appropriate to use low cyclic degradation factor values to stay in the safe side
of the design.

285
Table 7.7 A1, A2 ve A3 Model Cyclic Degradation Parameters

A1-(U) A2-(U) A3-(C)

Y 1.0 0,35 0,65
Reinforcement
U 1.0 0,35 0,65
Walls
L 1.0 0,35 0,65
R 1.0 0,35 0,65
X 1.0 0,35 0,65
USF --- 1 -0,5
Y 1.0 0,2 0,3
60 cm Height
Beams with

U 1.0 0,2 0,25

L 1.0 0,15 0,22
R 1.0 0,15 0,2
X 1.0 0,15 0,2
USF --- 0,75 1
Y 1.0 0,3 0,5
Conventionally

Link Beams
Reinforced

U 1.0 0,25 0,5

L 1.0 0,22 0,5
R 1.0 0,2 0,5
X 1.0 0,2 0,5
USF --- 1 0
Y 1.0 0,5 1
Link Beams
Diagonally
Reinforced

U 1.0 0,3 0,65

L 1.0 0,3 0,65
R 1.0 0,3 0,65
X 1.0 0,3 0,65
USF --- 1 -0,5

286
 Table 7.8 Mean Maximum Results Comparison Ratios of Calibrated (A3) and
Uncalibrated Models (A1 &A2) for Case Study Building in Ankara & Istanbul

ANKARA ISTANBUL
A3-(C) /A1-(U) A3-(C) /A2-(U) A3-(C) /A1-(U) A3-(C) /A2-(U)
min. max. mean min. max. mean min. max. mean min. max. mean
Drift
1.02 1.18 1.10 0.98 0.99 0.99 0.99 1.16 1.08 0.98 0.99 0.99
Check
Shear
0.95 1.11 1.03 0.97 1.00 0.99 0.93 1.17 1.05 0.96 1.00 0.98
Check
Walls

Concrete
0.90 1.18 1.04 0.83 1.08 0.96 0.88 1.19 1.04 0.93 1.08 1.01
Strain
Rein.
0.80 1.37 1.09 0.93 1.02 0.98 0.91 1.36 1.14 0.94 1.02 0.98
Strain
Rotation
Check 0.98 1.32 1.15 0.96 1.00 0.98 1.09 1.28 1.19 0.97 0.99 0.98
(40/60)
Rotation
Check -- -- -- -- -- -- 1.09 1.35 1.22 0.96 0.99 0.975
(60/60)
Rotation
Check 0.98 1.19 1.09 0.96 1.03 1.00 -- -- -- -- -- --
(40/150)
Shear
Beams

Displa.t
-- -- -- -- -- -- 0.70 1.00 0.85 1.00 1.39 1.20
Check
(60/150)
Shear
Check 1.00 1.02 1.01 1.00 1.00 1.00 1.00 1.01 1.01 1.00 1.00 1.00
(40/60)
Shear
Check -- -- -- -- -- -- 1.00 1.01 1.01 1.00 1.00 1.00
(60/60)
Shear
Check 0.98 1.18 1.08 0.97 1.00 0.99 -- -- -- -- -- --
(40/150)
Dissipated
Inelastic -- -- 0.73 -- -- 1.18 -- -- 0.80 -- -- 1.1
Dissipation

Energy
Energy

Strain
-- -- 1.02 -- -- 1.00 -- -- 1.04 -- -- 1
Energy
Damping
-- -- 1.02 -- -- 0.98 -- -- 1.03 -- -- 0.98
Energy

287
 A2 uncalibrated model with very low cyclic reduction factors showed results
at the safer side of the design than the other models (A1, A3). This showed
us that if an engineer using the Perform 3D software conducted performance-
based design according to the existing TEC-2018 rules without adjusting the
calibration of structural members, it is seen that obtained results will be on
the safe side with low cyclic degradation factor values and as if accurate
modeling in line with the limits given by the regulation, as long as the cyclic
degradation factor value is not taken as "none (1.0)".
 The beam rotation limit, θpCP in TEC-2018 depends on the ultimate curvature
calculation, u. While conducting the ultimate curvature calculation, u, code
clause restricts the engineers to use reinforcement strain limit as only 40 %
of the ultimate reinforcement strain (s=0.4 su) in TEC-2018 ([Link]). This
restriction causes the very low beam rotation limits, θpCP that is very limited
for seismic regions. According to ASCE 41-17, the beam rotation collapse
prevention limits vary between 0.02-0.05, while according to TEC-2018, the
ultimate curvature calculation, u, is calculated for beam rotation limits, θpCP
as on the order of 0.01. This value order was also valid for the case study
building beams. The experience of engineers performing nonlinear analysis
is in line with this opinion. For this reason, it is recommended to take
reinforcement strain, s= su in the calculation of beam rotation limit, θpCP.
Accordingly, the calculation values of beam rotation limits, θpCP are
compatible with ASCE 41-17. To determine the rotation limits (Table 6.9) of
the beams of the case study building, reinforcement strain was taken as s=
su instead of s=0.4 su.

288
 REFERENCES

Computers and Structures, Inc., C. (2006). Peform Components and Elements For
Perform-3D and Perform-Collapse Version 4. [Link]

Computers and Structures, I. . C. (2006). PERFORM-3D Nonlinear Analysis and

Performance Assessment for 3D Structures User Guide.
[Link]

Ile, N., & Reynouard, J. M. (2005). Behaviour of U-shaped walls subjected to

uniaxial and biaxial cyclic lateral loading. Journal of Earthquake Engineering,
9(1), 67–94. [Link]

Lowes, L. N., Lehman, D. E., & Baker, C. (2016). Recommendations for Modeling
the Nonlinear Response of Slender Reinforced Concrete Walls Using
PERFORM-3D. SEAOC Convention, December, 1–18.

Moehle, J., Bozorgnia, Y., Jayaram, N., Jones, P., Rahnama, M., & Shome, N.
(2011). Case Studies of the Seismic Performance of Tall Buildings Designed
by Alternative Means – Task 12 Report for the Tall Buildings Initiative: Final
Report to California Seismic Safety Commission and California Emergency
Management Agency. Pacific Earthquake Engineering Research Center, July.

Naish, D., Fry, A., Klemencic, R., & Wallace, J. (2013a). Reinforced Concrete
Coupling Beams — Part I : Testing. ACI Structural Journal, 110(6), 1057–
1066. [Link]
origsite=gscholar&fromopenview=true

Naish, D., Fry, A., Klemencic, R., & Wallace, J. (2013b). Reinforced Concrete
Coupling Beams — Part II : Modeling. ACI Structural Journal, 110(6), 1067–
1076.
[Link]
details/id/51686161

289
NIST. (2014). (For Thesis) Guidelines for Nonlinear Structural Analysis for Design
of Buildings Part I-General Applied Technology Council. NIST GCR 17-917-
46v1.

Orakcal, K., Wallace, J. W., & Conte, J. P. (2004). Nonlinear modeling and
analysis of slender reinforced concrete walls. ACI Structural Journal, 101(5),
688–698. [Link]

Pacific Earthquake Engineering Center. (2017). Guidelines for Performance-Based

Seismic Design of Tall Buildings. PEER Report 2017/06, May, 147.

PEER/ATC 72-1. (2010). Modeling and acceptance criteria for seismic design and
analysis of tall buildings. Applied Technology Council, 1.
[Link]

Seismic Evaluation and Retrofit of Existing Buildings. (2017). In Seismic

Evaluation and Retrofit of Existing Buildings.
[Link]

TBDY-2018. (2018). Türkiye Bina Deprem Yönetmeliği. 395.

[Link]

290
 APPENDICES

A. Shear Load Capacity Diagrams of Walls for Three Alternative Models

of the Case Study Building in Ankara (Performance Based Design)

Figure A.1 Wall Labels

291
292
293
294
295
296
297
298
B. Shear Load Capacity Diagrams of Walls for Three Alternative Models
of the Case Study Building in Istanbul (Performance Based Design)

Figure B.1 Wall Labels

299
300
301
302
303
304
305
C. Strain Check of Walls for Three Alternative Models of the Case Study
Building in Ankara (Performance Based Design)

Figure C.1 Strain Gage Labels

307
308
309
310
311
312
313
314
315
316
D. Strain Check of Walls for Three Alternative Models of the Case Study
Building in Istanbul (Performance Based Design)

Figure D.1 Strain Gage Labels

317
318
319
320
321
322
323
324
325
326