Turkey RCC Load Combi

DEVELOPMENT OF LOAD AND RESISTANCE FACTORS FOR
REINFORCED CONCRETE STRUCTURES IN TURKEY
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY
BY
FATİH KÜRŞAT FIRAT
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR
THE DEGREE OF DOCTOR OF PHILOSOPHY
IN
CIVIL ENGINEERING
SEPTEMBER 2007
Approval of the thesis:
DEVELOPMENT OF LOAD AND RESISTANCE FACTORS FOR

REINFORCED CONCRETE STRUCTURES IN TURKEY
submitted by FATİH KÜRŞAT FIRAT in partial fulfillment of the requirements

for the degree of Doctor Of Philosophy In Civil Engineering, Middle East
Technical University by,
Prof. Dr. Canan Özgen __________________

Dean, Graduate School of Natural and Applied Sciences
Prof. Dr. Güney Özcebe __________________

Head of Department, Civil Engineering
Prof. Dr. M. Semih Yücemen __________________

Supervisor, Civil Engineering Dept., METU
Examining Committee Members:
Prof. Dr. M. Yaşar Kaltakcı __________________

Civil Engineering Dept., Selcuk University
Prof. Dr. M. Semih Yücemen __________________

Civil Engineering Dept., METU
Assoc. Prof. Dr. H. Şebnem Düzgün __________________

Mining Engineering Dept., METU
Assist. Prof. Dr. Erdem Canbay __________________

Assist. Prof. Dr. Murat Güler __________________

Date: 07/09/2007
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: Fatih Kürşat Fırat
Signature :
iii
ABSTRACT
DEVELOPMENT OF LOAD AND RESISTANCE

FACTORS FOR REINFORCED CONCRETE
STRUCTURES IN TURKEY
Firat, Fatih Kürsat

Ph. D., Department of Civil Engineering
Supervisor: Prof. Dr. M. Semih Yücemen
September 2007, 293 pages
In this dissertation, a study is conducted to develop a probability based load and

resistance factor design criterion for structural members considering the local
conditions of Turkey. The Advanced First Order Second Moment (AFOSM)
procedure is utilized as the probabilistic method of analysis. Various sources of
uncertainties associated with concrete compressive strength, yielding and ultimate
strength of reinforcing steel bars and the dimensions of beams, columns and shear
walls are analyzed and quantified. The resistance statistics for different failure
modes of different types of reinforced concrete structural members are computed by
using the resistance parameters within the framework of reliability analysis.
Structural load effects of dead, live, wind, snow and earthquake loads are analyzed
considering the uncertainties in these loads.
iv
For different load combinations, the safety levels corresponding to the current
design practice are evaluated in terms of the reliability indexes for reinforced
concrete beam, column and shear wall design in flexure and shear, and also column
design in combined action of flexure and axial load. Depending on this evaluation
and the reliability index values reported from other countries, target reliability
indexes are selected for different load combinations and different failure modes of
structural members. Finally, a new set of load and resistance factors corresponding
to selected target reliabilities and levels of uncertainties are proposed for each
different failure modes of the structural members considered in this study,
separately.
Keywords: Uncertainty, Load and Resistance Factors, Reliability, Reliability Index,

Safety Level, Structural Member, Load Combination
v
ÖZ
YÜK VE DAYANIM KATSAYILARININ

TÜRKİYE’DEKİ BETONARME YAPILAR İÇİN
BELİRLENMESİ
Fırat, Fatih Kürşat

Doktora, İnşaat Mühendisliği Bölümü
Tez Yöneticisi: Prof. Dr. M. Semih Yücemen
Eylül 2007, 293 sayfa
Bu tez çalışmasında, Türkiye koşullarına göre istatistiksel yöntemler kullanılarak

betonarme yapı elemanları için yük ve dayanım katsayılarına dayalı bir tasarım
kriteri geliştirilmiştir. “Geliştirilmiş Birici Mertebe İkinci Moment” (GBMİM)
metodu istatistiksel analiz metodu olarak kullanılmıştır. Beton basınç
dayanımındaki, beton çelik çubuklarının akma dayanımındaki ve kolon, kiriş ve
perde duvar boyutlarındaki değişik belirsizlik kaynakları incelenmiş ve bu
belirsizlikler sayısallaştırılmıştır. İncelenen yapı elemanlarının değişik göçme
durumları için, istatistiksel dayanım parametreleri “güvenilirlik analizi”
çerçevesinde hesaplanmıştır. Sabit, hareketli, kar, rüzgar ve deprem yüklerindeki
belirsizliklerin sayısallaştırılması ve bu yüklerin yapıya olan etkisi ile ilgili bir dizi
vi
analizler yapılmıştır.
Türkiye’deki geçerli tasarım kriterine uygun olarak değişik yük kombinasyonları

için emniyet dereceleri, güvenilirlik indeksleri dikkate alınarak, bazı göçme
durumlarına göre ayrı ayrı incelenmiştir. Bu göçme durumlarıyla ilgili olarak, basit
eğilme ve kesme etkisindeki kiriş, kolon ve perde duvarların yanısıra eksenel basınç
ve eğilme altındaki kolonlar dikkate alınmıştır. Bu çalışmada hesaplanan emniyet
dereceleri ve yabancı kaynaklarda verilen hedef güvenilirlik indeksleri göz önünde
tutularak, yeni hedef güvenilirlik indeksleri farklı yük kombinasyonları ve farklı
göçme durumları için belirlenmiştir. Son olarak, belirlenen hedef güvenilirlik
indekslerine uygun, yeni yük ve dayanım katsayıları her bir yapı elamanının her bir
göçme durumu için ayrı ayrı önerilmiştir.
Anahtar Kelimeler: Belirsizlik, Yük ve Dayanım Katsayıları, Güvenilirlik,

Güvenilirlik İndeksi, Emniyet Derecesi, Yapı Elemanı, Yük Kombinasyonu
vii
ACKNOWLEDGMENTS
I would like to express my deepest appreciation to my thesis supervisor Prof. Dr. M.

Semih Yücemen for his professional guidance, support, criticism, and
encouragement. It was a great pleasure for me to work with him.
I owe special thanks to Assist. Prof. Dr. Erdem Canbay and Assoc. Prof. Dr. H.
Şebnem Düzgün for their helpful suggestions and guidance throughout the study.
I also would like to send my thanks to Turkish Meteorological Department,

Ministry of Public Works and Settlement, Turkish Chamber of Civil Engineers,
General Directorate of State Hydraulic Works, Techno-test, Vetaş and Armada
firms, Habaş, İçdaş, Ekiciler, Çolakoğlu, Egeçelik, Kroman and Yeşilyurt iron and
steel factories, and the materials laboratories of METU, ITU and Selçuk University
for providing the data used in this study.
Special thanks are due to Ali Faik Ulusoy, Seval Pınarbaşı, Mustafa Can Yücel,
Yusuf Baran and Ayşem Karadağ for their great friendship and help.
Finally, I am deeply grateful to my wife Emine and my daughter Ayça for their
endless support and altruism. I also would like to convey my deepest thanks to my
parents Nurcihan and A.Turan Fırat for their support and encouragement.
viii
TABLE OF CONTENTS
ABSTRACT ..........................................................................................................iv
ÖZ …………………………………………………………………………………..vi
ACKNOWLEDGMENTS................................................................................... viii
TABLE OF CONTENTS .......................................................................................ix
LIST OF TABLES...............................................................................................xiv
LIST OF FIGURES .............................................................................................xxi
LIST OF SYMBOLS .........................................................................................xxiv
CHAPTERS
1. INTRODUCTION............................................................................................1
1.1 GENERAL VIEW....................................................................................1
1.2 REVIEW OF RELATED WORK.............................................................3
1.3 AIM AND SCOPE OF THE STUDY .......................................................5
2. STRUCTURAL RELIABILITY MODELS ......................................................8
2.1 THE CLASSICAL RELIABILITY THEORY..........................................8
2.2 FIRST ORDER SECOND MOMENT METHOD ....................................9
2.2.1 Mean Value Method ....................................................................10
2.2.2 Advanced First Order Second Moment Method ...........................12
2.3 MODELING AND ANALYSIS OF UNCERTAINTIES........................14
3. QUANTIFICATION OF UNCERTAINTIES RELATED TO
RESISTANCE PARAMETERS .....................................................................17
3.1 CONCRETE ..........................................................................................17
ix
3.1.1 Evaluation of Data.......................................................................19
3.1.2 Uncertainty Analysis of Concrete Compressive Strength .............22
3.2 REINFORCING STEEL BARS .............................................................25
3.2.1 Evaluation of Data.......................................................................28
3.2.2 Uncertainty Analysis of Reinforcing Steel Bars ...........................30
3.3 DIMENSIONS .......................................................................................33
3.3.1 Data Collection and Analysis.......................................................34
3.3.2 Beam Width and Depth ...............................................................36
3.3.3 Beam Effective Depth .................................................................38
3.3.4 Column Width and Depth ............................................................40
3.3.5 Column Effective Depth ..............................................................42
3.3.6 Shear Walls .................................................................................44
3.3.7 Reinforcement Area.....................................................................46
4. MODELING OF CAPACITY OF REINFORCED CONCRETE
MEMBERS IN DIFFERENT FAILURE MODES..........................................48
4.1 CAPACITY OF REINFORCED CONCRETE BEAMS IN
DIFFERENT FAILURE MODES ..........................................................48
4.1.1 Flexural Capacity of Beams.........................................................49
4.1.2 Shear Capacity of Beams.............................................................57
4.2 CAPACITY OF REINFORCED CONCRETE COLUMNS IN
4.2.1 Combined Action of Flexure and Axial Load for Columns ..........64
4.2.2 Shear Capacity of Columns .........................................................72
4.3 CAPACITY OF REINFORCED CONCRETE SHEAR WALLS IN
4.3.1 Flexural Capacity of Shear Walls.................................................75
4.3.2 Shear Capacity of Shear Walls.....................................................78
5. MODELING OF LOADS...............................................................................80
5.1 INTRODUCTION..................................................................................80
5.2 DEAD LOAD ........................................................................................83
5.3 LIVE LOAD ..........................................................................................85
x
5.3.1 Arbitrary Point-in-Time Live Load (Lapt).....................................86
5.3.2 Maximum Live Load (L) .............................................................89
5.4 SNOW LOAD........................................................................................91
5.4.1 Roof Snow Load..........................................................................93
5.4.2 Ground Snow Load, Pko ...............................................................95
5.4.3 Annual Extreme Roof Snow Load ...............................................97
5.4.4 Maximum Roof Snow Load (S)...................................................99
5.4.5 Mean to Nominal Ratios of San and S.........................................101
5.5 WIND LOAD.......................................................................................102
5.5.1 Analysis of Wind Speed ............................................................104
5.5.2 Maximum, Yearly Maximum and Daily Maximum Wind Loads106
5.6 EARTHQUAKE LOAD.......................................................................110
5.6.1 Determination of Total Equivalent Lateral Earthquake Load......113
5.6.1.1 UBC-1994 ...................................................................114
5.6.1.2 TEC-2006 ....................................................................116
5.6.1.3 Case 1..........................................................................121
5.6.1.4 Case 2..........................................................................124
5.6.1.5 Case 3..........................................................................127
6. ASSESSMENT OF THE SAFETY LEVELS INHERENT IN THE
CURRENT DESIGN PRACTICE ................................................................132
6.1 INTRODUCTION................................................................................132
6.2 LOAD COMBINATION......................................................................133
6.2.1 Turkstra’s Rule..........................................................................134
6.2.2 Safety Criterion .........................................................................135
6.2.3 Load Statistics ...........................................................................136
6.3 RESISTANCE STATISTICS ...............................................................139
6.4 COMPUTATION OF RELIABILITY INDEXES.................................141
6.4.1 Reliability Indexes for Reinforced Concrete Beams in the
Flexural Failure Mode ...............................................................142
6.4.1.1 Gravity Loads ..............................................................142
6.4.1.2 Gravity and Wind Loads ..............................................145
xi
6.4.1.3 Gravity and Earthquake Loads .....................................147
6.4.2 Reliability Indexes for Reinforced Concrete Beams in the Shear
Failure Mode .............................................................................149
6.4.2.1 Gravity Loads ..............................................................149
6.4.3 Reliability Indexes for Reinforced Concrete Columns in the
Combined Action of Flexure and Axial Load Failure Mode.......155
6.4.3.1 Gravity Loads ..............................................................155
Shear Failure Mode ...................................................................159
6.4.5 Reliability Indexes for Reinforced Concrete Shear Walls in the
Flexural Failure Mode ...............................................................160
Shear Failure Mode ...................................................................161
6.4.7 Assessment of Target Reliability Indexes ..................................162
7. SELECTION OF LOAD AND RESISTANCE FACTORS...........................171
7.1 INTRODUCTION................................................................................171
7.2 LOAD AND RESISTANCE FACTORS CORRESPONDING TO
THE SELECTED TARGET RELIABILITIES .....................................172
7.2.1 Load and Resistance Factors for Reinforced Concrete Beams in
the Flexural Failure Mode..........................................................173
7.2.2 Load and Resistance Factors for Reinforced Concrete Beams in
the Shear Failure Mode..............................................................179
7.2.3 Load and Resistance Factors for Reinforced Concrete Columns
in the Combined Action Failure Mode .......................................185
7.2.4 Load and Resistance Factors for Reinforced Concrete Columns
in the Shear Failure Mode..........................................................191
7.2.5 Load and Resistance Factors for Reinforced Concrete Shear
xii
Walls in the Flexural Failure Mode............................................197
7.2.6 Load and Resistance Factors for Reinforced Concrete Shear
Walls in the Shear Failure Mode................................................197
7.3 OPTIMAL LOAD AND RESISTANCE FACTORS ............................198
7.4 MODIFIED OPTIMAL LOAD AND RESISTANCE FACTORS ........202
7.5 RECOMMENDED LOAD AND RESISTANCE FACTORS ...............209
8. SUMMARY AND CONCLUSIONS............................................................213
REFERENCES ...................................................................................................221
APPENDICES
A. STATISTICAL PARAMETERS OF 7 AND 28 DAY COMPRESSIVE

STRENGTH DATA ACCORDING TO CONCRETE CLASS
AND REGION……………………………………….…………………….. 229
B. STATISTICAL PARAMETERS OF BCIII(A) REINFORCING

STEEL BARS USED in TURKEY……………………………………………234
C. ANNUAL MAXIMUM SNOW DEPTHS FOR DIFFERENT

LOCATIONS………………………………………….…………..…….….....253
D. ANNUAL MAXIMUM WIND SPEEDS FOR DIFFERENT

LOCATIONS………………………………………………………………….256
E. DAILY MAXIMUM WIND SPEEDS OBSERVED DURING THE YEAR

2004 IN DIFFERENT LOCATIONS…………………………………………259
F. FIGURES SHOWING THE VARIATION OF RELIABILITY INDEX

OBTAINED ACCORDING THE TURKISH DESIGN PROVISIONS…...…272
G. FIGURES SHOWING THE VARIATION OF SAFETY LEVEL

CORRESPONDING TO THE RECOMMENDED LOAD AND
RESISTANCE FACTORS……………………………………..……………..280
CURRICULUM VITAE……………………………………………….…………293
xiii
LIST OF TABLES
Table 3.1 Statistical parameters of compressive strength data according to years

for Turkey .......................................................................................................21
Table 3.2 Statistical parameters of 28 day compressive strength data according
to concrete class for Turkey.............................................................................21
Table 3.3 Statistical parameters of 7 day compressive strength data according to
concrete class for Turkey.................................................................................21
Table 3.4 Different country standards related to reinforcing steel bars...................27
Table 3.5 Mean value and c.o.v. of yield strength of BCIII(a) reinforcing steel
bars produced by different steel and iron plants in Turkey ...............................28
Table 3.6 Mean value and c.o.v. of ultimate strength of BCIII(a) reinforcing
steel bars produced by different steel and iron plants in Turkey .......................29
Table 3.7 Mean value and c.o.v. of elongation of BCIII(a) reinforcing steel bars
produced by different steel and iron plants in Turkey.......................................29
Table 3.8 Statistical parameters of yield strength of BCIII(a) reinforcing steel
bars according to years ....................................................................................29
Table 3.9 Results of the statistical analysis of beam external dimensions for .........38
Table 3.10 Results of the statistical analysis of beam internal dimensions before
the placement of concrete ................................................................................39
Table 3.11 Results of the statistical analysis of column external dimensions for
different regions ..............................................................................................42
Table 3.12 Results of the statistical analysis for the column internal dimensions ...44
Table 3.13 Results of the statistical analysis of shear wall external dimensions
for different regions.........................................................................................45
xiv
Table 3.14 Results of the statistical analysis of reinforcement areas ......................47
Table 4.1 Statistics of the basic variables involved in the calculation of flexural
and shear capacities of reinforced concrete beams ...........................................49
Table 4.2 Mean to nominal ratios and coefficients of variation of the beam
flexural capacity (b=200 mm and b=300 mm) .................................................56
flexural capacity (b=400 mm and b=1000 mm)................................................57
shear capacity..................................................................................................62
Table 4.5 Statistics of the basic variables involved in the calculation of the
combined action and shear capacities of columns ............................................64
Table 4.6 Mean to nominal ratios and coefficients of variation of the axial and
flexural capacities for eccentrically loaded columns (b/h=0.6) .........................68
flexural capacities for eccentrically loaded columns (b/h=1)............................69
flexural capacities for eccentrically loaded columns (b/h=2)............................70
flexural capacities for eccentrically loaded columns (b/h=0.5) .........................71
and shear capacities of shear walls...................................................................75
Table 4.11 Mean to nominal ratios and coefficients of variation for the flexural
capacity of shear-walls ....................................................................................77
Table 4.12 Mean to nominal ratios and coefficient of variations of resistances
for shear-walls in the shear failure mode..........................................................79
Table 5.1 Statistics of dead load (from Ellingwood et al., 1980)............................85
Table 5.2 Statistics of arbitrary point-in-time live loads for offices reported in
different studies (from Ellingwood et al., 1980)...............................................87
Table 5.3 Results of arbitrary point-in-time live load surveys for office buildings
which was reported in different studies (from Kumar, 2002a)..........................89
Table 5.4 Statistics of maximum live loads as reported in different studies............90
xv
Table 5.5 Statistical parameters of the annual extreme ground snow load..............97
Table 5.6 Statistical parameters of the annual extreme roof snow load ..................99
Table 5.7 Statistical parameters of the 50-year maximum roof snow load............ 101
Table 5.8 Statistical parameters of the mean to nominal ratio for annual extreme
and 50-year maximum snow loads................................................................. 102
Table 5.9 Mean values and coefficients of variation of wind speeds for different
locations........................................................................................................ 106
Table 5.10 Mean values and the total uncertainties of wind loads for different
locations........................................................................................................ 108
Table 5.11 Mean to nominal wind load ratios and associated total uncertainties
for different locations .................................................................................... 109
Table 5.12 Type I distribution parameters of mean to nominal wind load ratios
for different locations .................................................................................... 109
Table 5.13 Geographical coordinates and seismic zones of selected locations,
and corresponding peak ground acceleration values for different return
periods .......................................................................................................... 111
Table 5.14 Parameters of Type II distribution for peak ground acceleration for
different locations.......................................................................................... 112
Table 5.15 Mean value and total variability of peak ground acceleration for
different locations.......................................................................................... 113
Table 5.16 Seismic zone factor in UBC-1994...................................................... 115
Table 5.17 Response modification factor (RW) for reinforced concrete buildings
in UBC-1994 ................................................................................................. 116
Table 5.18 Effective ground acceleration coefficients (A0) in TEC-2006............. 118
Table 5.19 Spectrum characteristic periods (TA, TB) in TEC-2006 ...................... 119
Table 5.20 Structural system behavior factor (R) in TEC-2006 for cast-in-situ
reinforced concrete buildings......................................................................... 120
Table 5.21 Mean to nominal ratios of earthquake load in terms of different local
site classes and building heights (RW=8, Ra(T)=7) ......................................... 122
xvi
site classes and building heights (RW=6, Ra(T)=6)......................................... 125
site classes and building heights (RW=6, Ra(T)=6)......................................... 126
Table 5.27 The average mean to nominal ratios of earthquake load obtained
from UBC 1994 and TEC-2006 for different locations................................... 130
Table 5.28 Mean to nominal ratios of earthquake load for different locations
where both values are computed based on TEC-2006 .................................... 130
Table 5.29 Statistical parameters of the mean to nominal ratio for earthquake
load ............................................................................................................... 131
Table 6.1 Statistics of dead, live, wind and earthquake loads............................... 138
Table 6.2 Resistance statistics for different reinforced concrete members in
different failure modes................................................................................... 140
Table 6.3 Relative frequency distribution of the ratio of a given load to dead
load ............................................................................................................... 142
Table 6.4 Reliability indexes and design situations for the flexural failure mode
of reinforced concrete beams subjected to D+L combination ......................... 143
of reinforced concrete beams subjected to D+S combination ......................... 144
of reinforced concrete beams subjected to D+Lapt+W combination ................ 146
of reinforced concrete beams subjected to D+L+Wapt combination ................ 147
of reinforced concrete beams subjected to D+Lapt+E combination.................. 149
Table 6.9 Reliability indexes and design situations for the shear failure mode of
reinforced concrete beams subjected to D+L combination ............................. 150
xvii
reinforced concrete beams subjected to D+S combination.............................. 151
reinforced concrete beams subjected to D+Lapt+W combination .................... 152
reinforced concrete beams subjected to D+L+Wapt combination .................... 153
reinforced concrete beams subjected to D+Lapt+E combination...................... 155
Table 6.14 Reliability indexes and design situations for reinforced concrete
columns in the combined action of flexure and axial load subjected to D+L .. 156
columns in the combined action of flexure and axial load subjected to D+S... 156
columns in the combined action of flexure and axial load (for D+Lapt+W) ..... 157
columns in the combined action of flexure and axial load (for D+L+Wapt) ..... 158
columns in the combined action of flexure and axial load (for D+Lapt+E) ...... 159
Table 6.19 Reliability indexes for reinforced concrete columns in the shear
failure mode .................................................................................................. 160
Table 6.20 Reliability indexes of reinforced concrete shear walls in the flexural
failure mode .................................................................................................. 161
Table 6.21 Reliability indexes of reinforced concrete shear walls in the shear
failure mode .................................................................................................. 162
Table 6.22 Reliability indexes corresponding to the current design practice and
the target reliability indexes for different load combinations and different
structural members according to different studies .......................................... 163
Table 6.23 Current and the target reliability indexes for different load
combinations considering beams in the flexural failure mode ........................ 165
combinations considering beams in the shear failure mode ............................ 166
xviii
combinations considering columns in the combined action failure mode........ 167
combinations considering columns in the shear failure mode ......................... 168
combinations considering shear walls in the flexural failure mode................. 169
combinations considering shear walls in the shear failure mode..................... 170
Table 7.1 Optimal load and resistance factors for reinforced concrete beams in
the flexural failure mode................................................................................ 199
Table 7.2 Optimal load and resistance factors for reinforced concrete beams in
the shear failure mode.................................................................................... 200
Table 7.3 Optimal load and resistance factors for reinforced concrete columns in
the combined action failure mode .................................................................. 200
Table 7.4 Optimal load and resistance factors for reinforced concrete columns in
the shear failure mode.................................................................................... 201
Table 7.5 Optimal load and resistance factors for reinforced concrete shear walls
in the flexural failure mode............................................................................ 201
Table 7.6 Optimal load and resistance factors for reinforced concrete shear walls
in the shear failure mode................................................................................ 202
Table 7.7 Modified optimal load and resistance factors for reinforced concrete
beams in the flexural failure mode................................................................. 203
beams in the shear failure mode..................................................................... 204
columns in the combined action failure mode ................................................ 204
columns in the shear failure mode.................................................................. 205
shear walls in the flexural failure mode.......................................................... 205
xix
shear walls in the shear failure mode ............................................................. 206
Table 7.13 Load and resistance factors in USA and Jordan for beams in the
flexural failure mode ..................................................................................... 206
Table 7.14 Load and resistance factors recommended by Kömürcü and
Yücemen (1996) for beams in the flexural failure mode................................. 207
Table 7.15 Average load factors for all structural members in different failure
modes............................................................................................................ 207
Table 7.16 Load and resistance factors and corresponding reliability index
values for different structural members and failure modes for φ=1.0.............. 208
Table 7.17 Recommended load factors for all structural members in different
failure modes for Turkey ............................................................................... 210
Table 7.18 Recommended load and resistance factors and the corresponding
reliability index values for beams in the flexural failure mode ....................... 210
reliability index values for beams in the shear failure mode ........................... 210
reliability index values for columns in the combined action failure mode ...... 211
reliability index values for columns in the shear failure mode........................ 211
reliability index values for shear walls in the flexural failure mode................ 211
reliability index values for shear walls in the shear failure mode.................... 212
xx
LIST OF FIGURES
Figure 2.1 Reliability Index β2 (from Thoft-Christensen and Baker, 1982) ............13
Figure 3.1 The map showing the locations of iron and steel plants founded in
Turkey (from http://www.dcud.org.tr/indextur.htm).........................................26
Figure 4.1 Stresses and forces in reinforced concrete beams..................................50
Figure 4.2 Beam sections of T and L shapes (Ersoy and Özcebe, 2004) ................52
Figure 4.3 Stresses and forces in reinforced concrete columns...............................66
Figure 4.4 a) Lateral loads b) Isolated wall c) Shear diagram d) Moment
diagram ...........................................................................................................74
Figure 5.1 The map showing the locations of selected cities..................................83
Figure 7.1 Variation of the load and resistance factors for RC beams in the
flexural failure mode (Turkey; D+L; βT=3.0)................................................. 173
flexural failure mode (Bursa; D+S; βT=3.0) ................................................... 174
flexural failure mode (Turkey; D+S; βT =3.0) ................................................ 174
flexural failure mode (Canakkale; D+Lapt+W; βT =3.0).................................. 175
flexural failure mode (Turkey; D+Lapt+W; βT =3.0) ....................................... 176
flexural failure mode (Canakkale; D+L+Wapt; βT =2.7).................................. 177
flexural failure mode (Turkey; D+L+Wapt; βT =2.7) ....................................... 177
xxi
flexural failure mode (Ankara; D+Lapt+E; βT =1.75) ...................................... 178
flexural failure mode (Turkey; D+Lapt+E; βT =1.75) ...................................... 178
shear failure mode (Turkey; D+L; βT =3.0).................................................... 179
shear failure mode (Bursa; D+S; βT =3.0) ...................................................... 180
shear failure mode (Turkey; D+S; βT =3.0).................................................... 180
shear failure mode (Canakkale; D+Lapt+W; βT =3.0)...................................... 181
shear failure mode (Turkey; D+Lapt+W; βT =3.0)........................................... 182
shear failure mode (Canakkale; D+L+Wapt; βT =2.7)...................................... 183
shear failure mode (Turkey; D+L+Wapt; βT =2.7)........................................... 183
shear failure mode (Ankara; D+Lapt+E; βT =1.75).......................................... 184
shear failure mode (Turkey; D+Lapt+E; βT =1.75) .......................................... 184
Figure 7.19 Variation of the load and resistance factors for RC columns in the
combined action failure mode (Turkey; D+L; βT =3.2) .................................. 185
combined action failure mode (Bursa; D+S; βT =3.2)..................................... 186
combined action failure mode (Turkey; D+S; βT =3.2) .................................. 186
combined action failure mode (Canakkale; D+Lapt+W; βT =3.2) .................... 187
xxii
combined action failure mode (Turkey; D+Lapt+W; βT =3.2) ......................... 188
combined action failure mode (Canakkale; D+L+Wapt; βT =3.0) .................... 189
combined action failure mode (Turkey; D+L+Wapt; βT =3.0) ......................... 189
combined action failure mode (Ankara; D+Lapt+E; βT =1.75)......................... 190
combined action failure mode (Turkey; D+Lapt+E; βT =1.75)......................... 190
shear failure mode (Turkey; D+L; βT =3.0).................................................... 191
shear failure mode (Bursa; D+S; βT =3.2) ...................................................... 192
shear failure mode (Turkey; D+S; βT =3.2).................................................... 192
shear failure mode (Canakkale; D+Lapt+W; βT =3.2)...................................... 193
shear failure mode (Turkey; D+Lapt+W; βT =3.2)........................................... 194
shear failure mode (Canakkale; D+L+Wapt; βT =3.0)...................................... 195
shear failure mode (Turkey; D+L+Wapt; βT =3.0)........................................... 195
shear failure mode (Ankara; D+Lapt+E; βT =1.75).......................................... 196
shear failure mode (Turkey; D+Lapt+E; βT =1.75) .......................................... 196
xxiii
LIST OF SYMBOLS
A Cross sectional area
A Peak ground acceleration
A0 Effective ground acceleration coefficient
A1 Influence area
As Reinforcement area
Asw Cross sectional area of stirrups
A(T1) Spectral acceleration coefficient
b Member width
C Coefficient related to the fundamental period of vibration of the structure
C1,C2 Tabulated coefficients of Type I distribution
Ce Exposure factor for snow load
Ci Influence coefficient of load “i”
Cs Snow load coefficient
Ct Thermal factor of snow load
c Air density constant
c Distance from the neutral axis to outer compressive fiber in a T cross-section
xxiv
cb Depth of neutral axis at the balanced case in reinforced concrete cross
section
D Dead load effect
Df Failure domain
Ds Safety domain
d Depth of the member
de Effective depth of the member
E Earthquake load effect (the lateral seismic base shear)
Es Modulus of elasticity of steel reinforcement
FX,fx Cumulative distribution function (CDF) and probability density function of

variable X, respectively.
fc Concrete compressive strength
fct Tensile strength of concrete
fs Steel stress
fy Yield strength of steel bars
fyw Yield strength of shear reinforcement
G Gust factor of wind load
g(x) Limit state function
HN Building height
h Ground snow depth
h Member depth
I Importance factor
k Shape parameter of Type II distribution
k1 A dimensionless coefficient which is a function of strength of concrete
xxv
L Live load effect of maximum live load
lw Horizontal length of a shear wall
M Safety margin
Mr Bending moment capacity
N Correction factor
N Axial load
Pf Probability of failure
Pko Ground snow load
P(E) Probability of event E
Ps Survival probability (reliability)
pi The weight assigned to the ith load situation
R Generalized resistance
R Rate of loading
R Structural system behavior factor
Rw Response modification factor for the earthquake load
Ra(T1) Seismic load reduction factor
S Snow load effect
S Site coefficients of soil properties
S(T) Spectrum coefficient
s Spacing of stirrups
T Fundamental period of vibration of a building in a specified direction
TA,TB Spectrum characteristic periods
t Depth of the flange thickness
xxvi
U Effect of factored load
u The scale parameter of Type I distribution
V Wind speed
V Total design base shear
Vc Shear resistance of concrete
Vd Maximum design shear force
Vr Shear strength
Vw Resistance of shear reinforcement
W Wind load effect
W Weight of structure
X Basic random variable
Χ Mean value of X
Χ′ Nominal value of X
X∗ Design value of X
Χ̂ The model used to estimate X
Xapt Arbitrary point-in-time value of X
xi Distance between neutral axis and ith steel layer in reinforced concrete cross
section
Z Seismic zone factor
z Normalized parameter
α Direction cosine
α The local parameter of Type I distribution
β Reliability index
xxvii
βT Target reliability index
Γ(.) Gamma function
γ Generalized load factor
∆ Prediction uncertainty
δ Basic variability
εcu Ultimate strain in concrete
η Stress distribution coefficient
λ,ζ Parameters of lognormal distribution
µ Mean value
ν Characteristic parameter of Type II distribution
ρ Average density of snow load
ρ Steel ratio
ρb Balanced steel ratio
ρ′ Compression reinforcement ratio
σ Standard deviation
σs Steel stress
φ(.) Tabulated probability distribution function of the standard normal variate
Ω Total variability
ϕ Generalized resistance factor
AFOSM Advanced First Order Second Moment Method
APT Arbitrary point-in-time
a.p.t. Arbitrary point-in-time
xxviii
C.D.F. Cumulative distribution function
c.o.v. Coefficient of variation
FOSM First Order Second Moment Method
JCSS Joint Committee on Structural Safety
LRFD Load Resistance Factor Design
RC Reinforced Concrete
TEC Turkish Earthquake Code (Specification for Structures to be Built in

Earthquake Areas)
xxix
CHAPTER 1
INTRODUCTION
1.1 GENERAL VIEW
Due to the uncertainties in the properties of building materials and rising

construction costs, the structural design engineers have been working mainly on the
subject of safe, functionally reliable and economical structures. Structures and
structural elements must satisfy the following requirements with appropriate levels
of reliability:
-They should be serviceable till the end of their economic life (serviceability
limit state requirement).
-They should withstand extreme or frequently repeated effects taking place

during their construction and economical lifetime (ultimate limit state
requirements).
-They should not be damaged by natural hazards and accidental events like
earthquakes, floods, hurricanes, fires and explosions.
Thereby, modern structural design codes should provide a simple, economical and
safe way for the design of engineering structures. Design codes not only make easy
the work of engineers but also optimize the resources of society. Traditionally,
1
design codes take design equations as the basis, and the reliability of a given design
can be checked by a simple comparison of resistances and load effects. Due to the
fact that loads and resistances are subjected to uncertainties, structural design must
be made in the presence of uncertainties. Additionally loads and strengths of
materials are subjected to random variability due to the complex environmental and
structural mechanisms that govern them. Moreover, the data required for the
structural design are obviously not certain with 100 % confidence. These
uncertainties in structural design are taken into consideration by using probabilistic
methods. Therefore, in recent years, the trend in code development is evolving
towards the use of probabilistic techniques as a basis for developing design criteria.
The traditional approach does not take these uncertainties sufficiently into
consideration and their logical basis is not sufficiently developed. However, the
uncertainties in structural design are considered fully in probabilistic approach. In
this approach, loads and design parameters are treated as random variables and
safety is to be realized by using a tolerable risk or reliability index. Reliability level
can be chosen to reflect the consequences of failure. Besides, the reliability of
complete structural system is determined by considering the reliability of individual
structural system components. The combined effects of different potential failure
modes are also taken into consideration while applying this procedure. Additionally,
this method provides a way to update the present standards with additional
knowledge in a systematic way.
The safety formats of design codes, i.e. design equations, characteristic values, and
load and resistance factors may be chosen so that the level of reliability of all
structures designed according to relevant structural design codes and specifications
is uniform and independent of the choice of material properties, loading types,
operational conditions and environmental surroundings, within the scope of the
structural reliability methods (probabilistic methods). This procedure, covering the
choice at the preferred level of reliability or target reliability, is commonly referred
to as code calibration (Faber et al., 2003).
2
1.2 REVIEW OF RELATED WORK
First quarter of the 20th century is the beginning of introduction of the probability
concepts in the assessment of structural safety. ACI Building Code introduced the
ultimate strength design in which the design was based on ultimate limit states
using loads increased by load factors and strengths reduced by strength reduction
factors in the early sixties of the 20th century. These important factors were
established upon an elementary statistical analysis carried out by using the
incomplete data and insufficient information. Remarkable studies in this discipline
began in the late sixties and a growing interest has been shown in structural
reliability since then.
Wide ranging studies have been performed during seventies to derive load and
resistance factors for steel structures. These load factors differ from the ones used in
the concrete structures. Recognizing the possible confusion in design offices due to
use of different set of load and resistance factors in the design of steel structures and
reinforced concrete structures, ACI Building Code Committee passed a motion in
1976 endorsing the idea of almost identical load factors for all materials in a manner
that is agreed upon by all committee members.
So as to propose a set of universal load and resistance factors to be utilized in the

design of buildings, the Centre for Building Technology at the National Bureau of
Standards brought Drs. Cornell, Ellingwood, Galambos, and MacGregor together in
1979. The outcomes of this study were published in several papers (Ellingwood et
al., 1980; Galambos et al., 1982, Ellingwood et al., 1982). The basic set of load and
resistance factors was incorporated in the 1982 ANSI A58.1.
Rackwitz (2000) studied the code-making process by using the total cost
minimization as the decision tool to assess the target reliability. In his study, the
maintenance and reconstruction costs were also taken into consideration. Aktas
3
(2001) studied the structural design code calibration using reliability based cost
optimization, in which predefined target reliability levels were not used.
The Liaison Committee, that coordinated the activities of six international

associations in civil engineering created a joint committee on structural safety
(JCSS) in 1971. This committee’s basic goal was improve the general knowledge
related to structural safety. Thus, JCSS has developed a model code in 2001. This
code is related to only probabilistic design and gives no information about
mechanical models like buckling, shear capacity, flexure failure.
The study of Nowak and Szerszen (2003a,b) documents the calibration of the
Building Code Requirements for Structural Concrete (ACI 318). Their study is
presented in two articles. The first article addresses the topic of statistical model for
resistance (i.e. the models of strength of materials). The developed resistance
models served as a basis for the selection of resistance factors (strength reduction
factors). Accordingly, the second article focuses on the reliability analysis and
selection of resistance factors.
In Turkey, the probabilistic design approach is not well known. Therefore, there is
not much research related to the probabilistic design. Nonetheless, there are a few
studies, such as Keskinel (1971), Yücemen and Gülkan (1975), Gündüz (1988) in
which simple probabilistic concepts were used, but the conception utilized is far
from developing a reliability based design code. In this area, the first significant
research was done by Yücemen and Gülkan in 1989. In their research, a set of
reliability-based load and resistance factors were proposed for the reinforced
concrete beams. Later, Kömürcü (1995) developed a reliability-based design
criterion to be used in reinforced concrete beams in the flexural failure mode by
considering the local conditions and the design practice in Turkey, which is the
most comprehensive study conducted in this field in the past.
4
The research work presented in this dissertation is to a certain extent a follow up of
the research work conducted by Kömürcü (1995). However, in the present research
compared to the study of Kömürcü (1995), much more comprehensive data sets are
compiled with respect to the basic resistance and load variables, and also the load
statistics are computed for many different cities. Additionally, more recent and up
to date codes are taken into consideration in this dissertation. The most significant
difference is the fact that Kömürcü (1995) has considered only reinforced concrete
beams in the flexural failure mode whereas in this study, reinforced concrete beams,
columns and shear walls in the flexural and shear failure modes as well as
reinforced concrete columns in the combined action of flexure and axial load failure
mode are taken into consideration.
1.3 AIM AND SCOPE OF THE STUDY
This study aims at developing a set of load and resistance factors for the design of
reinforced concrete structural elements by considering the local conditions of
Turkey and by utilizing probabilistic methods. Resistance and load parameters are
assumed as random variables in this study. The data necessary for the assessment of
these parameters are collected in Turkey and are unified with the data published in
the international literature.
In this study, structural load effects resulting from dead, live, wind, snow and
earthquake loads are evaluated. For calibration purposes, the ratios of mean to
nominal load values are determined. Probability based resistance criterion for the
design of reinforced concrete beams, columns and shear walls in the flexural and
shear failure modes as well as the design of columns in the combined action of
flexure and axial load failure mode are examined by using the First Order Second
Moment Method (FOSM). In order to propose a set of load and resistance factors,
the Advanced First Order Second Moment Method (AFOSM) is adopted as the
structural reliability model.
5
The present building codes in Turkey were developed without taking various
sources of uncertainties into consideration in a direct way. In this study, various
sources of uncertainties associated with the load and resistance parameters are
analyzed and quantified based on the data available in Turkey and in the foreign
literature. The proposed criterion and the resulting load and resistance factors will
form a rational basis for the reflection of the effects of uncertainties directly on the
design.
After this chapter, the relevant structural reliability models and a probabilistic
method to be used for the assessment of uncertainties are introduced in Chapter II.
In Chapter III, the mean to nominal ratios and the total variabilities of basic
resistance variables, namely compressive strength of concrete, yield strength of
reinforcing steel bars and dimensions of the structural elements are quantified using
the local data and reported results. In Chapter IV, using the values proposed in
Chapter III, the mean to nominal ratios and the total uncertainties of resistances for
different failure modes of three basic types of reinforced concrete structural
members, namely beams, columns and shear walls, are computed within the
framework of structural reliability. In Chapter V, the load modeling is carried out
with the published data in the literature and the local data collected in Turkey. In
Chapter VI, reliability based resistance criterion for the failure modes of reinforced
concrete structural members are examined, and target reliabilities are selected. In
Chapter VII, a set of new load and resistance factors for different failure modes of
beams, columns and shear walls are computed according to AFOSM method by
using two computer programs prepared for this purpose. Finally, the summary and
conclusions of this dissertation are provided in the last chapter.
In Appendix A, statistical parameters of 7 and 28 day compressive strength data

according to concrete class and region are given. Appendix B displays the statistical
parameters of reinforcing steel bars used in Turkey. In Appendices C, D and E, the
meteorological data related to annual maximum snow depth, annual maximum wind
speed and daily maximum wind speed for 12 different locations are given,
6
respectively. The reliability index values for load combinations considered in this
study are displayed in graphical forms in Appendix F. Lastly, in Appendix G,
figures showing the variation of safety levels corresponding to the recommended
load and resistance factors are given.
7
CHAPTER 2
STRUCTURAL RELIABILITY MODELS
2.1 THE CLASSICAL RELIABILITY THEORY
The problems related to the reliability of structural engineering systems may be

considered as a problem of capacity against demand. When the safety of structure is
considered, the strength of the structure, R, should be adequate to resist the lifetime
maximum applied load, S. If the reliability of a single member is taken into account
for introducing the basic concepts of the structural reliability theory, the member
failure occurs when "R<S", or equivalently when "M<0", where M is the safety
margin (M=R-S). The probability of failure, Pf, is defined by the following
relationship:
Pf = P(R≤S)=P(R-S≤0)=P(M≤0) (2.1)
where P (.) represents the probability of occurrence of the event in brackets.
In the case of independent and normally distributed R and S with mean R and S ,
standard deviation, σR and σS , the failure probability is given in Eq. (2.2):
M  R−S 
Pf = 1 − φ  = 1 − φ 
 (2.2)
σ 2
 σ R + σS
2

8
where φ(.) is the probability distribution of the standard normal variate and M is the
mean value of the safety margin.
However, in practice, resistance (R) and load (S) are dependent on other basic
variables, e.g. the material properties, geometrical quantities and load effects.
2.2 FIRST ORDER SECOND MOMENT METHOD
The reliability of a structure can be determined as the probability of operating its

proposed function. In order to obtain a generalized formulation, a mathematical
model is first derived concerning the resistance and load variables according to the
~
limit state under consideration (Ellingwood et. al., 1980). Let M=g( Χ )=g(Xi,...,Xn)
~
be the limit state function, in which Χ =(Xi,...,Xn) is the vector of basic variables
concerning the structure. The failure surface can be described by Eq. (2.3). In this
equation, positive values of g indicate safe sets of variables (the safe region) and
non-positive values of g indicate unsafe sets of variables (the failure region) (Toft-
Christensen and Baker, 1982).
~
g (X ) = g ( X 1 ,..., X n ) = 0 (2.3)
If the safety margin M is linear and basic variables are normal, the reliability index
β1 is defined by:
β1 = µM / σM (2.4)
where β1 is a nonparametric measure of safety, µM and σM are the mean and

standard deviation of M, respectively. The higher the value of β1 is, the safer the
structure will be.
9
If the safety margin M is linear in the basic variables X1,…,Xn as expressed below:
M=a0+a1.X1+…an.Xn (2.5)
where, a0, a1,…,an are constants, then the mean and variance of M can be obtained
from the following equations:
µM= a0+a1.µ1+…an.µn (2.6)
σM2=a12.σ12+…+an2.σn2+ ∑ ∑ ρij.ai.aj.σi.σj
i =1 j =1i ≠ j
(2.7)
Here, the last term is related to the correlation between any pair of basic variables
and ρij is the correlation coefficient between Xi and Xj.
For the case of a linear safety margin, M, the calculation of the mean safety margin
(µM), the standard deviation of safety margin (σM) and also the reliability index (β1)
is easy. In the case of nonlinear failure functions, the mean value method and
advanced methods are used, which are explained briefly in the following sections.
2.2.1 Mean Value Method
~
When the safety margin M is nonlinear in Χ =(X1,…,Xn), then approximate values
for µM and σM can be obtained by using a linearized safety margin. The safety
margin can be expressed by the following formula:
~
M =g( Χ )≅g (X1,X2,...,Xn) (2.8)
If Eq. (2.8) is expanded in a Taylor series about ((X1,...,Xn) =(µ1,...,µn) and if only the
linear terms are kept, then:
10
~  ∂g 
M =g( Χ )≅ g( µ 1,…, µ n) + ∑  ∂Χ
i =1 
 ( Χ i − µ i ) (2.9)
i 0
 ∂g 
where   is evaluated at ( µ 1,…, µ n).
 ∂Χ  0
Approximate values of µ M and σM are determined by:
µ M ≅ g( µ 1,…, µ n) (2.10)
 ∂g   ∂g 
σ2M= ∑∑     Cov( Χ i , Χ j ) (2.11)
i =1 j=1  ∂Χ i
 
 0  ∂Χ j  0
where Cov(Xi, Xj) is the covariance between Xi and Xj.
Ellingwood et al. (1980) stated that the mean value method has the following two
basic shortcomings:
-If the failure function is nonlinear, significant errors may be introduced at increasing
distances from the linearized point by neglecting higher order terms.
-The mean value method fails to be invariant to different mechanically equivalent

formulations of the same problem. In effect, this means that reliability index depends
on how the limit state is formulated. This is a problem not only for nonlinear forms
of failure functions but even in certain linear forms, e.g., when the loads counteract
one another.
11
2.2.2 Advanced First Order Second Moment Method
Due to shortcomings in the FOSM and the mean value methods, Hasofer and Lind
(1974) proposed a new reliability index denoted by (β2). The process of calculating
β2 is generally referred as the advanced first-order second-moment (AFOSM)
method.
In this method, first the basic variables Xi's are converted into the standardized
variables, Zi's, including zero means and unit variances, as shown below:
Χi − Χi
Zi = , i = 1,2,..., n (2.12)
σ Χi
The failure surface in the X-coordinate system is mapped into a failure surface in
the z-coordinate system. The failure surface in the z-coordinate system divides the z
space into a failure region and a safe region. Reliability index β2 is defined as the
shortest distance from the origin to the failure surface in the z-coordinate system.
The design point is defined as the point having the shortest distance to the origin on
the failure surface. In Figure 2.1, the reliability index, β2, and the design point are
illustrated for the case of two basic variables.
The definition of reliability index, β2 can be formulated as:
1/ 2
 n 
β 2 = min  ∑ zi 2  (2.13)
 i =1 
12
Figure 2.1 Reliability Index β2 (from Thoft-Christensen and Baker, 1982)
When the failure surface is nonlinear, the searching of the design point and the
calculation of the reliability index can be carried out by the iterative algorithm.
According to the iterative procedure, the design point (z*i,…,z*n) on the failure
surface is determined by solving the following system of equations (Thoft-
Christensen and Baker, 1982):
− ∂g / ∂z i
αi = , i=1,2,…,n (2.14)
(∑ (∂g / ∂z ) ) i
2 1/ 2
∗
Zi = α iβ 2 (2.15)
∗ ∗
g (z i ,..., z n ) = 0 (2.16)
where αi’s are directional cosines that minimize β2. The derivatives are evaluated at
the design point. Since the computed design point is related to standardized
13
coordinate system, the basic variables on the design point must be converted to
original variable space. This procedure can be carried out by using the following
formulation:
∗
Χ i = Χ i (1 − α i β.Ω i ) (2.17)
where Ω i is the coefficient of variation reflecting the total uncertainty in Xi. The
load and resistance factors, γi, corresponding to a prescribed reliability index, β,

may then be determined by the following expression:
∗
Χi
γi = (2.18)
′
Χi
where Χ ′ is the nominal or design value of the ith load or resistance parameter
specified in the code ( Ellingwood et al., 1980)
If all of the Xi’s are independent, normally distributed random variables and the
*
failure function g is linear in Χ i , then the reliability index, which may be related to
the probability of failure (Pf), can be determined by using FOSM method described
above. This relationship between β and Pf is given in Eq. (2.19):
β = φ −1 (1 − Pf ) , Pf = 1 − φ(β) (2.19)
2.3 MODELING AND ANALYSIS OF UNCERTAINTIES
The assessment of safety and reliability in the probabilistic approach first of all
requires the quantification of uncertainties, which can be achieved by the
assessment of standard deviation, variance or coefficient of variation (c.o.v.) of
basic variables. In the evaluation of structural reliability, it is highly important to
14
quantify the uncertainties as accurately as possible. The model for each limit state
contains a specified set of basic variables, e.g. material and geometrical properties,
physical quantities that characterize actions and environmental loads. The
evaluation of reliability must include the uncertainties related to these basic
variables as well as the prediction models.
Uncertainties in civil engineering problems can be divided into two main groups:
aleatory uncertainty and epistemic uncertainty. Aleatory uncertainty is due to
inherent variability and this type of variability, which is a state of nature, cannot be
controlled or reduced. On the other hand, epistemic uncertainty is due to lack of
knowledge and can be reduced by additional information and data. Prediction,
modeling and statistical errors are sources of epistemic uncertainty. The
uncertainties related to discrepancy between laboratory conditions and in-situ
conditions are examples for prediction uncertainty. Modeling uncertainty is owing
to simplifying assumptions, indefinite boundary conditions on unknown effects of
other variables that are not included in the model. Statistical uncertainty is due to
the limited number of observations utilized in the estimation of the mean of a basic
variable. Engineering judgment has an important role in reducing epistemic
uncertainties. Comprehensive uncertainty analysis requires inputs from experience-
based professional engineering judgments.
The different sources of uncertainties can be modeled by using the following

relationship for a basic variable (Ang and Tang, 1984):
Χ i = N i Χ̂ i (2.20)
where:
Χ i : the true (but unknown) value of the ith basic variable
Ν i : a random correction factor to account for epistemic uncertainties
Χ̂ i : the model used to estimate Xi
15
The mean of Ni will be denoted by Ν i and its coefficient of variation by ∆ X , which
i
respectively quantify the mean bias and epistemic uncertainty. Χ is the mean of Χ i
and its coefficient of variation is denoted by δ Χ i , which also quantifies the inherent
variability, i.e. aleatory uncertainty. According to FOSM method, the total

uncertainty can be formulated as fallows:
Ω X = δ 2X + ∆2X (2.21)
i i i
If Ni is expressed as the product of component correction factors as

N i = N i1 N i 2 ,..., N in , then based on FOSM method the following relations can be
written for Ν i and ∆ Χ i , assuming all Ni’s mutually independent:
Ν i = Ν i1 Ν i 2 ,..., Ν in (2.22)
∆ X = ∆2 X i1 + ∆2 X i 2 +,...,+ ∆2 X in (2.23)
i
16
CHAPTER 3
QUANTIFICATION OF UNCERTAINTIES RELATED

TO RESISTANCE PARAMETERS
3.1 CONCRETE
The quality control of concrete is quite important for achieving safety in reinforced
concrete structures. The following properties can be taken into account for the
quality of concrete from the viewpoint of mechanical characteristics:
-Compressive strength,
-Durability against physical and chemical effects,
-Time dependent deformations, i.e. creep and shrinkage.
Furthermore, the following properties can be sought for concrete produced in some
circumstances in addition to the three basic properties given above:
-Impermeability,
-Abrasion resistance,
-Unit weight,
-Resistance in flexure and tension.
The most appropriate way to evaluate the quality of concrete is through the
statistical analysis of data. Taking into consideration the properties listed above, the
17
mean value of the samples obtained must be high, while the coefficient of variation
must be as low as possible.
There are a number of factors that affect the quality of concrete, and these factors
can also affect each other considerably. To obtain good quality concrete, the
following recommendations can be made:
-The materials like sand, gravel, cement and admixtures, which constitute
concrete, should be of good quality.
-Concrete should be a carefully dosed mixture of cement, aggregate, water
and admixtures for the class of concrete desired.
-The technology used should be good.
-Knowledge and ability of personnel who performs actual production should
be sufficient to obtain good quality concrete in the stages of production.
-The techniques of production and maintenance should be arranged for the
concrete class selected, and the environmental conditions should be adjusted in the
application of these techniques.
In common construction sites in Turkey, it is well known that the conditions

required for high quality production were not efficiently taken into account until the
last few years and the issue of production of good quality concrete was not given
notable importance. Due to irresponsibility, unconsciousness and lack of control
and supervision in production of concrete, poor quality of production was generally
observed. Even for a concrete class sampled from the same construction site, it was
a reality that there were significant differences among the concrete samples
produced.
One of the aims in this section is to shed some light upon the quality of concrete
production in Turkey. The most important property that affects the quality of
production is compressive strength, which contributes to other properties
proportionately. Therefore, the quality control of concrete is examined by using the
18
compressive strength tests utilizing the standard cube specimens (Dislitas et al.,
2006). Considering the compressive strength of concrete and its variation according
to regions and years, the quality of concrete production is investigated from a
statistical point of view.
3.1.1 Evaluation of Data
Laboratory test results of 150x150x150 mm cube specimens, which are obtained

from different laboratories located in different cities of Turkey, form the database
considered in this study. The data that belong to 2000, 2001, 2002, 2003, 2004 and
2005 years are analyzed comprehensively. Besides, the data that belong to earlier
years are also taken into consideration for the sake of comparison.
The data used in this study, which are sampled from Ankara, Istanbul, Konya,
Corlu, Trabzon, Erzurum, Kayseri, Samsun, Bursa, Izmir, Gaziantep, Denizli,
Antalya and Malatya regions are obtained from the following institutions:
-The laboratories of Turkish Chamber of Civil Engineers (IMO) in Istanbul,

Konya and Antalya offices,
-The materials laboratory of Middle East Technical University (METU)
Department of Civil Engineering,
-The materials laboratory of Istanbul Technical University (ITU)
Department of Civil Engineering,
-The materials laboratory of Selcuk University (SU) Department of Civil
Engineering,
-The laboratories of Ministry of Public Works and Settlement in Malatya,
Trabzon, Erzurum, Kayseri, Samsun, Bursa and Gaziantep offices,
-The laboratory of General Directorate of State Hydraulic Works (DSI)
Konya office,
-The laboratory of Techno-test firm,
19
-The laboratory of Vetas construction materials firm,
-The laboratory of Armada concrete and construction materials firm.
In the following tables (Tables 3.1-3.3), an idea can be obtained about the quality of
concrete produced in Turkey. Just a certain building firm or a ready mixed concrete
production plant is not taken into consideration, and compressive strength test
results used are obtained from different construction sites. Since compressive
strength test results are not evaluated for each construction site separately, high
values of coefficient of variation are obtained for Istanbul region. However, in
Ankara, Konya, Corlu, Trabzon, Erzurum, Kayseri, Samsun, Bursa, Izmir,
Gaziantep, Denizli, Antalya regions, the values of c.o.v. are obtained for each
construction site individually. Therefore, all of the regions except Istanbul are taken
into account during the calculation of c.o.v. for Turkey, while all regions are
considered during the calculation of mean values of concrete classes. The results of
data analysis are summarized in Tables 3.1-3.3.
Statistical parameters of 7 day and 28 day compressive strength data are given
according to regions mentioned above in Appendix A. In Tables 3.1-3.3 and
Appendix A, fck is the nominal value of the concrete compressive strength and
“number of the values under the limit” refers to the number of samples whose
compressive strength is less than “fcmin= fck-4” N/mm2, as specified in TS EN 206-1
(2002).
As indicated in Table 3.1, the comparison of the mean compressive strength values
for the years 1994-1995 and the subsequent years covering the interval between
2000-2005, reveals that in Turkey, the mean compressive strength has increased
considerably over the years. Comparing the overall weighted average compressive
strength values it can be said that the 28 day compressive strength is 1.23 times
higher than the 7 day strength, as it is seen in Table 3.2 and Table 3.3. It is also to
be noted that the c.o.v. of the 7 day compressive strength (0.132) is 1.26 times more
than the c.o.v. of the 28 day compressive strength (0.105).
20
Table 3.1 Statistical parameters of compressive strength data according to years for
Turkey
Years Number of Mean Coefficient of Number of Percentage of

samples (N/mm2) variation values under values under
the limit the limit (%)
94/95 417 20.60 --- 58 13
2000 732 26.97 0.142 40 5.46
2001 535 30.97 0.107 23 4.30
2002 465 31.21 0.104 10 2.15
2003 644 30.78 0.131 36 5.59
2004 1283 28.87 0.123 30 2.34
2005 615 29.97 0.120 24 3.90

concrete class for Turkey
Concrete Number of fck, cyl Mean Coefficient Number of Percentage

Class samples (fck,cub) (N/mm2) of values of values
(N/mm2) variation under the under the
limit limit (%)
C14 137 14 (18) 20.04 0.143 1 0.83
C16 755 16 (20) 25.11 0.144 13 1.73
C18 739 18 (22) 25.82 0.120 23 3.11
C20 5817 20 (25) 28.46 0.104 118 2.7
C25 2767 25 (30) 32.48 0.100 53 2.81
C30 870 30 (37) 40.07 0.079 14 2.47
Overall 11085 29.87 0.105 222 2.65

concrete class for Turkey
Concrete Number of fck, cyl Mean Coefficient of
Class samples (fck,cub) (N/mm2) variation
(N/mm2)
C14 24 14 (18) 14.32 0.159
C16 418 16 (20) 20.26 0.154
C18 538 18 (22) 18.69 0.153
C20 3091 20 (25) 23.04 0.132
C25 1701 25 (30) 26.85 0.129
C30 463 30 (37) 32.63 0.100
Overall 6235 24.20 0.132
21
3.1.2 Uncertainty Analysis of Concrete Compressive Strength
After conducting a statistical analysis on the collected data, the mean value of cubic
compressive strength of overall concrete is obtained as 29.87 N/mm2 and c.o.v. of
compressive strength as 10.5% by considering the whole data set (Table 3.2). If the
cubic compressive strength of overall concrete data is converted to the standard
cylinder strength value, then f c and δfc are found to be 24.87 N/mm2 and 10.5 %,
respectively. Here, f c denotes the compressive strength and δfc stands for the c.o.v.
of fc, which is a measure of basic (inherent) variability in the compressive strength

of concrete. When the most common construction conditions in Turkey are
considered, these values can be taken directly as representative of whole Turkey.
On the other hand, in practice, the strength of concrete in a structure may differ
from its specified design strength. The major sources causing these differences are
variations in material properties, proportions of concrete mix, variations in mixing,
transporting, placing and curing methods and variations in testing procedures
(Mirza et al., 1979).
Some factors, which cause additional uncertainty besides inherent variability, will
be considered in the following paragraphs. Mostly, since there is no adequate local
information, these factors are quantified within the scope of the international
literature.
The strength of concrete in a structure is lower than the strength of control cylinder
molded from the same concrete. This difference results from the effects of different
curing and placing processes, the segregation of concrete in deep members, the
effects of difference in size and shape and the effects of different stress conditions
in the structure and in the specimens. Due to these factors, the actual core strength
of concrete is lower than the laboratory measured strength in most of the situations.
The average ratios of core strength to standard cylinder strength reported in
different studies varied within the range of 0.74 and 0.96 with an overall average
22
value of 0.87 (Mirza et al., 1979). Likewise, in another study conducted by
Ellingwood and Ang (1972), this ratio was stated to range from 0.83 to 0.92. In
order to compensate for the difference between the in-situ actual strength and the
strength obtained from control cylinders, a correction factor, N1, will be introduced.
Based on the reported ranges, the mean value of the corresponding correction
factor, N 1 , is taken as 0.86, which is the average of the above ranges. The c.o.v. of
N1 is denoted by ∆1. Ellingwood and Ang (1972) recommended a value of 0.16 for
∆1. In the study performed by Mirza et al. (1979), it was proposed that ∆1=0.1.
Here, the value of ∆1 will be taken to be 0.13 as the average of these two values.
The apparent strength of concrete rises as the rate of loading increases. The increase
in strain with time because of the micro cracking and creep causes the decrease of
observed strength of concrete. To account for this effect, the correction factor, N2,
with a prediction uncertainty, ∆2, is introduced. Mirza et al. (1979) used the
following empirical relationship to express the mean value of N2 in their study.
N 2 =0.89(1+0.08 log (R)) (3.1)
where R is rate of loading (psi/sec). Using the value of 1 psi (6.895 kN/m2 per
second) for R, N 2 is found to be 0.89. Kömürcü (1995) suggested the value of 0.88
for N 2 with no prediction uncertainty, i.e. ∆2=0. Based on other studies, Mirza et
al. (1979) stated that the prediction uncertainty associated with N2 can be ignored,
and if the rate of loading, which is approximately 234.43 kN/m2 per second for
standard cylinder test, is decreased to 6.895 kN/m2 per second, then apparent
strength of concrete decreases by a value of 12%. In this study, N 2 is taken as 0.88
which is a conservative estimate, and prediction uncertainty, ∆2 is neglected.
Thirdly, the specimens are obtained from a special concrete batch rather than
obtaining them randomly from the actual mix. Standard testing procedures may not
23
be applied; for example, testing on the specimens may not be carried out at the
correct time by properly calibrated machines. In other words, due to human factor, a
prediction error may occur. As a result, to be on the safe side, a correction factor, N3
is introduced with a mean of 0.95 and a prediction error, ∆3 of 0.05 (Kömürcü,
1995).
The in-situ value of compressive strength of concrete can be modeled according to

Eq. (2.20) as follows:
f c = N f .f̂ c (3.2)
c
where fc is the in-situ value of compressive strength of concrete, f̂ c is the

compressive strength of the cylindrical specimens tested at the laboratory, and
N fc is overall bias in fc.
The overall mean bias in fc is computed as N fc = N 1 .N 2 .N 3 =0.86x0.88x0.95=0.72

and the mean value of the in-situ compressive strength of concrete will be
f c =0.72x24.87=17.87 N/mm2. Here, C14, C16, C18, C20, C25 and C30 concrete
classes, which are given in TS 500 (2000), are taken into consideration. The
weighted average value of concrete class used in Turkey, which is obtained from
the third column of Table 3.2, is found to be 21.55 (in other words, the weighted
average concrete class is “C21.55”). Regarding this weighted average, the nominal
21.55
compressive strength is taken as =14.37 N/mm2. Hence, the ratio of mean to
1.5
f c 17.90
nominal compressive strength, i.e. = will be equal to 1.25.
f c 14.37
On the other hand, the total prediction uncertainty due to these three sources of
uncertainty will be ∆ f c = 0.13 2 + 0 2 + 0.05 2 =0.14. The inherent variability in
24
compressive strength based on laboratory test results and the total prediction error
are combined to find the total variability, which becomes
Ω f = 0.105 2 + 0.14 2 = 0.18 . In a similar study, the total uncertainty was taken as
c
0.21 by Kömürcü and Yücemen (1996). Likewise, Real et al. (2003) used three
different values for the overall variability of concrete in their study as 0.10, 0.15,
and 0.20. The value computed here is consistent with the values given in these
previous studies.
The normal distribution was used for concrete compressive strength by the majority
of researchers (Mirza et al., 1979). Yücemen and Gülkan (1989), Kömürcü (1995)
and Nowak and Szerszen (2003a) also used the normal distribution for the
compressive strength of concrete. In this study, a statistical analysis of collected
data based on the Chi-Square and Kolmogorov- Smirnov tests also supported the
normal distribution at a significance level α=0.05; consequently a normal
distribution for the compressive strength of concrete is used herein.
3.2 REINFORCING STEEL BARS
Turkish iron and steel sector, the base of which was established in the 1930’s, plays
an important role in the industrialization and development of Turkish economy. Iron
and steel production was first started in Kirikkale, which is now known as
Mechanical and Chemical Industry Corporation (MKEK). The first integrated plant,
Karabük Iron and Steel Works (KARDEMIR) began to operate in Karabük in 1937.
In order to meet the demand for flat products, the second integrated plant, Eregli
Iron and Steel Works (ERDEMIR) started production in 1965. In 1977, Iskenderun
Iron and Steel Works (ISDEMIR), Turkey’s third integrated steel mill, came on line
to meet the demand for long products and semi-finished products. The Turkish iron
and steel industry has shown great progress in the last fifteen years in terms of
quality and capacity and has been one of the growing iron and steel industries in the
25
world. Currently, Turkish iron and steel sector ranks 11th among the 64 steel-
making countries in the world and 3rd in Europe. In 2003 steel products were
exported to more than 130 countries in the world (Turkish Iron and Steel Producers
Association, 2004).
Figure 3.1 The map showing the locations of iron and steel plants founded in
Turkey (from http://www.dcud.org.tr/indextur.htm)
In Figure 3.1, the locations of iron and steel plants founded in Turkey are shown.
Since steel products are exported to more than 130 countries in the world, most of
these steel and iron plants prepare their products according to different standards as
specified by different countries, and some of these steel products are also sold in the
domestic market. Furthermore, Turkey also imports steel from different countries.
Therefore, the quality of steel used in construction sites is affected negatively in
terms of coefficient of variation, since there are a lot of different types of steel in
the domestic market. Some country standards concerning reinforcing steel bars are
given in Table 3.4.
26
Table 3.4 Different country standards related to reinforcing steel bars
(from Özsoy, 2001)
Minimum
Minimum
Ultimate Yield/ Minimum
Yield
Country Standard Class Limit Ultimate Elongation
Strength
Strength Strength (%)
( N/mm² )
( N/mm² )
12
TS 708 III a 420 500 1.1 (Ø8-Ø28)
Turkey
(1996) IV a 500 550 1.08 10
(Ø32-Ø50)
BS 4449
England Gr 460B 460 - 1.08 14
(1997)
BSt 420 S 420 500 1.05 10

Germany DIN 1986
BSt 500 S 500 550 1.05 10
NS 3576-2
(1997) B500B 500 - 1.08 -
Norway
NS 3576-3 B500C 500 - 1.15 -
(1997)
ASTM
Gr 40 300 500 - 11 ~ 12
A615 /
USA Gr 60 420 620 - 7~9
A616M
Gr 75 520 690 - 6~7
(1996)
Within the scope of this study, the local data on reinforcing steel bars are obtained
from the production reports of the following steel and iron plants:
-Habas iron and steel plant,

-Icdas iron and steel plant,
-Ekiciler iron and steel plant,
-Colakoglu iron and steel plant,
-Egecelik iron and steel plant,
-Kroman iron and steel plant,
-Yesilyurt iron and steel plant.
27
3.2.1 Evaluation of Data
In Turkey, BCIII(a) reinforcing steel bars are widely used. Therefore, the test
results on 24875 BCIII(a) reinforcing steel bar specimens with different bar sizes
ranging from 8 mm to 26 mm are obtained from the test reports. After the analysis
of data, the mean and c.o.v. of yield strength, ultimate strength and elongation
values are determined for each steel and iron plant. Additionally, the statistical
parameters of yield strength, ultimate strength and elongation based on civil
engineering material laboratories of ITU, METU and SU are investigated. The test
reports of university laboratories do not belong to a specific steel and iron plant, i.e.
the reinforcing steel bars tested in university laboratories are obtained from
different firms. The values based on civil engineering material laboratories of only
ITU and METU are arranged according to years. The results are presented in Tables
3.5 to 3.8, and the detailed results are given in Appendix B. In Table 3.8 and in
Appendix B, “number of the values under the limit” refers to the number of
reinforcing steel bars yield strength of which is less than “fy= 420 N/mm2”, as
specified in TS 708 (Steel Bars for Concrete).
Table 3.5 Mean value and c.o.v. of yield strength of BCIII(a) reinforcing steel bars
produced by different steel and iron plants in Turkey
Plant Habas Icdas Ekiciler Colakoglu Egecelik Yesilyurt Kroman Overall
Number of 19709
9619 1400 2390 530 1073 3024 1673
samples
Yield
strength 530.01 516.79 480.94 473.63 489.71 464.4 460.71 503.46
(N/mm2)
Coefficient
of 0.034 -- 0.047 0.048 0.052 0.036 0.035 0.038
variation
28
Table 3.6 Mean value and c.o.v. of ultimate strength of BCIII(a) reinforcing steel
bars produced by different steel and iron plants in Turkey
Plant Icdas Ekiciler Colakoglu Egecelik Kroman Yesilyurt Overall
Number of
samples 1400 2390 530 1073 1673 3024 10090
Ultimate strength
(N/mm2) 621 563.81 617.95 631.45 675.68 593.1 609.1
Coefficient
of variation -- 0.047 0.065 0.057 0.040 0.043 0.045
Table 3.7 Mean value and c.o.v. of elongation of BCIII(a) reinforcing steel bars
produced by different steel and iron plants in Turkey
Plant Icdas Ekiciler Colakoglu Egecelik Kroman Yesilyurt Overall
Number of
samples 1400 2390 530 1073 1673 3024 10090
Elongation
(%) 18.32 18.78 19.25 19.02 23.69 21.7 20.46
Coefficient
of -- 0.076 0.094 0.097 0.105 0.081 0.087
variation
Table 3.8 Statistical parameters of yield strength of BCIII(a) reinforcing steel bars
according to years
Years Number Required Mean Standard Coefficient Number Percentage

of samples yield value of deviation of of values of values
strength yield (N/mm2) variation under under the
(N/mm2) strength the limit limit (%)
(N/mm2)
1998-1999 846 420 491.89 72.31 0.147 122 14.4
2000 877 420 490.16 58.82 0.12 84 9.6
2001 911 420 502.16 62.26 0.124 64 7
2002 897 420 498.35 47.84 0.096 35 3.9
2003 770 420 492.78 59.13 0.12 49 6.4
2004 532 420 491.06 44.19 0.09 25 4.7
Overall 4833 420 494.76 58.38 0.118 379 7.8
29
3.2.2 Uncertainty Analysis of Reinforcing Steel Bars
The sources of variation in the yield strength and the ultimate strength of steel bars
are listed as follows:
-Variation in the strength of material itself,

-Variation in the cross-sectional area of the bar,
-Effect of rate of loading,
-Effect of bar diameter on mechanical properties of bars,
-Effect of the strain at which yield is defined.
The variability of yield strength of reinforcing steel bars depends on the nature and
source of the data. The variation in the yield strength within a single bar is
comparatively small, while the in-batch variation is insignificantly higher.
However, the variability of samples obtained from different batches and sources
may be high due to the fact that the quality measures change for different
manufactures and different bar sizes. When the data are taken from different
producers, an increase in coefficient of variation from 5% to 8% is observed for
individual bar sizes (Mirza and Macgregor, 1979). In the literature related to yield
strength of reinforcing bars, the basic variability in general ranges from 1% to 4%
for individual bar sizes supplied that the bars are produced by the same
manufacturer (Kömürcü, 1995). However, these values change from 2% to 7% for
individual bar sizes in Turkey. Here, the basic (inherent) variability in the
reinforcement yield strength for individual bar sizes is taken to be 3.8% as an
overall value of the data obtained from steel and iron plants, i.e. δfy=0.038 (Table
3.5).
The observed strength of a test specimen is affected considerably by the strain rate
and the rate of application of the load: the lower the rate of loading, the lower the
apparent strength. Since tests on reinforcing steel bar specimens are generally
performed at much greater strain rates than those encountered in structures
30
subjected to static loads, they are inclined to overestimate the yield strength (Mirza
and Macgregor, 1979). Ellingwood and Ang (1972) estimated the mean bias, N1 ,
accounting for this effect as 0.9. On the other hand, the prediction error, ∆1, due to
this discrepancy was assumed to be negligible (Kömürcü, 1995). In this study, the
mean bias is also taken as 0.9 and the prediction error is neglected.
Yield strength of reinforcing steel bars may be determined corresponding to upper

or lower yielding points or at a specific strain. These upper or lower yielding points
and specific strain values affect the yield strength. Mirza and Macgregor (1979) did
not include this effect in their study. Ellingwood and Ang (1972) used the value of
5% of prediction error in order to account for this effect. Kömürcü (1995) suggested
a value of 9% for the associated prediction error. Here, prediction error, ∆2, is taken
to be 5% as an average of these values, and a correction factor, N2, is introduced
with a mean value of 1.00.
On the other hand, the reality that most of the structures are built with reinforcing
steel bars produced in different iron and steel plants in Turkey has to be considered
as a prediction uncertainty. We mainly have two types of data related to reinforcing
steel bars, first of which are acquired from iron and steel plants separately. The
second type of data is obtained by test reports of reinforcing steel bars evaluated at
university laboratories. The average total inherent variability in the yield strength of
reinforcing steel bars of different steel and iron plants is computed to be 11.8 % and
the average inherent variability, i.e. within group variability of steel and iron plants
is found to be 3.8 % as an overall value. In connection with this study, the
maximum value of prediction uncertainty for structures which are built with
reinforcing steel bars of different iron and steel plants will be 0.112 based on Eq.
(2.21); that is, the maximum value of group variability between plants in addition to
within group variability is 11.2 %. If the structure is built with reinforcing steel bars
produced by only one plant, this value can be taken as 0. Here, the prediction error,
∆3, is taken to be 0.056 as an average value of 0 and 11.2%, and N 3 will be taken
31
as 1.00.
Mirza and Macgregor (1979) stated that the use of different bar diameters like Φ8,
Φ28 did not have a significant effect on the yield strength. At the end of data
analysis, it is seen that there is no significant difference among bar diameters from
the point of view the yield strength of reinforcing steel bars. Therefore, the effect of
bar diameter on yield strength is neglected.
After conducting a statistical analysis on the collected data, the yield strength is
computed to be 501.37 N/mm2 as a weighted average value obtained from the data
supplied by iron and steel plants and university laboratories. In connection with this
study, the overall mean bias is computed as N fy = N1 .N 2 .N 3 =0.9x1x1=0.90.
Using this value, the mean yield strength will be equal to 0.9x 501.37= 451.23
N/mm2. The nominal yield strength for BCIII(a) is 365 (420/1.15) N/mm2 as
given in TS 500 (2000). Consequently, the ratio of mean to nominal yield strength
is 1.24 (i.e. f y f f′ =451.23/365=1.24).
Kömürcü and Yücemen (1996) assigned a value of 0.14 for the total variability of
the yield strength of reinforcing steel bars. Real et al. (2003) stated that the total
variability in reinforcing steel bars change from 0.05 to 0.10 in their study. In this
study, the overall prediction error for reinforcing steel bars is computed to be
∆ f y = (0 2 + 0.05 2 + 0.056 2 ) = 0.08 . By combining the inherent variability and
overall prediction error, the total variability in the yield strength of reinforcing steel
bars is computed as Ω f y = (0.038 2 + 0.08 2 ) = 0.09 .
The sources of variation in the ultimate strength of reinforcing bars are the same as
those causing variation in the yield strength (Mirza and Macgregor, 1979). By using
the data values obtained from iron and steel plants and university laboratories, the
weighted average of the ultimate strength is computed as 615.96 N/mm2. The
32
overall mean bias of ultimate strength of reinforcing steel bars is computed as
N fu = N 1 .N 2 .N 3 = 0.9x1x1=0.90 and the mean ultimate strength will be equal to
0.9x 615.96= 554.36 N/mm2. The nominal ultimate strength for BCIII(a) is 435
(500/1.15) N/mm2 as given in TS 500 (2000). The ratio of mean to nominal ultimate
strength is 1.27 ( f u f u′ =554.36/435=1.27). For ultimate strength, the overall
prediction error is computed to be ∆ f u = (0 2 + 0.05 2 + 0.056 2 ) = 0.08 . Here, the

basic variability in the reinforcement ultimate yield strength for individual bar sizes
is taken as 4.5 %, which is a weighted average value obtained from steel and iron
plants (Table 3.6). Combining the inherent variability and overall prediction error of
ultimate strength, the total variability in the ultimate strength of reinforcing steel
bars is calculated to be Ω f u = (0.045 2 + 0.08 2 ) = 0.09 .
As for the distribution of the yield strength of reinforcing steel bars, Kömürcü
(1995) used the lognormal distribution for reinforcing steel bars. Nowak and
Szerszen (2003a) and Topcu and Karakurt (2006) recommended that normal
distribution is suitable for the reinforcing steel bars in their study. Here, the normal
distribution will be used for the yield strength of reinforcing steel bars. The
statistical analysis of the collected data also supports this selection.
3.3 DIMENSIONS
The dimensional characteristics of reinforced concrete members constructed at the

site sometimes do not match with those calculated in the design stage of the
structure. These geometrical discrepancies and imperfections are generally observed
in certain dimensional properties which are listed below:
-the cross-sectional shape and dimensions,

-the vertical and horizontal position of columns, beams, floors and
foundations,
33
-the levels of reinforcing bars, ties and stirrups.
Geometrical discrepancies may occur during different stages of the construction

process, which mainly depend on shape, size and quality of forms used. Concreting
and vibrating operations generally create the geometrical discrepancies.
Additionally, the strength of concrete, the dimension of orientation (depth, width)
and the position of the cross-section (mid-span, support) may affect geometrical
deviations. For these reasons, the degrees of geometrical discrepancies vary from
country to country, region to region and even from structure to structure, and these
differences are mainly dependent on the quality of construction techniques,
equipment and training of the site personnel.
Udoeyo and Ugbem (1995) used normal and lognormal distributions for the errors
on the dimensions in their study. On the other hand, a number of researchers have
recommended the normal distribution (Mirza and MacGregor, 1979). Yücemen and
Gülkan (1989) and Kömürcü (1995) also used the normal distribution in their
studies, which were conducted for Turkey. Since it is simple to use normal
distribution as the probability model, the normal distribution is used here to
represent the distribution of errors pertaining to dimensions of reinforced concrete
members, consisted with the previous studies mentioned above.
3.3.1 Data Collection and Analysis
Unfortunately, almost no local study concerning the geometrical discrepancies has

been performed in Turkey. Furthermore, the process of collecting and reporting data
for geometrical discrepancies has not been standardized. Therefore, the comparison
of the results of measurements is difficult in order to derive quantitative conclusions
for Turkey. As stated before, engineering judgment plays a critical role in
quantifying the epistemic uncertainties. Using the idea and inputs obtained from
experience-based professional engineering judgment, some additional prediction
34
uncertainties (epistemic uncertainty) related to the dimensions can be quantified.
Moreover, epistemic uncertainties in beam dimensions are compared to epistemic
uncertainties in dimensions of other structural members, within the scope of the
engineering judgment in cases where there is a lack of information.
Accordingly, the results of the uncertainty analysis mentioned above on

dimensional properties will be compared against the results reported in other
countries. In order to quantify the geometrical discrepancies and uncertainties as to
dimensions of reinforced concrete members, the data belonging to aleatory
uncertainty are obtained by measuring the dimensions of structural members after
the removal of forms. On the other hand, additional prediction uncertainties having
relatively small values (in terms of c.o.v) in the dimensions must be estimated. The
main characteristics of these additional prediction uncertainties can be defined as
follows:
-The increase in the variability of dimensions parallel to the decrease in the

design value (i.e. value of the dimension specified in the design),
-The complexity of direct measurement of effective depth,
-The unfixity of forms,
-Problems of vibration and concreting operations.
In spite of the fact that the variability in dimensions changes with different design
values, the same variability will be used for all of the design values considered in
this study. The prediction error associated with this effect will be assumed to be 2%.
The actual dimension may be affected due to unfixity of forms, problems of
vibration and concreting operations, i.e. the quality of workmanship and the
construction techniques affect the variability. In order to account for this effect, a
prediction error of 2% is introduced. The total prediction error due to these sources
of uncertainty will be ∆ bw = ((0.02)2 + (0.02)2 ) =0.03. On the other hand, considering
the effective depth of column and beam, it is very difficult to measure the effective
35
depth directly, and we used the depth of structural member and relevant steel cage.
A c.o.v. of 6% is assigned to this prediction error resulting from the indirect
measurement of effective depth. This value seems to be high; however, Ellingwood
et al. (1980) and Kömürcü (1995) took this prediction error also as 6% in their
studies due to the high probability of change in the position of bars during placing
of concrete. Accordingly, the total prediction error (i.e. epistemic uncertainty) in the
effective depth of beams and columns will be ∆ bw = ((0.02)2 + (0.02)2 + (0.06) 2 ) = 0.07
Considering the above characteristics of epistemic uncertainties related to

dimensions, Ellingwood and Ang (1972) quantified the total prediction uncertainties
as 0.02, 0.02 and 0.05 for width, depth and effective depth of beams, respectively.
On the other hand, Kömürcü (1995) quantified these uncertainties for width, depth
and effective depth of beams as 0.03, 0.03 and 0.07, respectively. In this study,
owing to similarities of quality of workmanship, construction systems and operating
techniques in beam, column and shear wall dimensions, the prediction uncertainty
will be taken conservatively as 0.03, 0.03 and 0.07, as computed above, for the
width, depth and effective depth of these structural members, respectively.
In the following sections, the basic variabilities and total uncertainties will be
evaluated separately in addition to prediction uncertainties.
3.3.2 Beam Width and Depth
Variations in beam width and depth dimensions influence the weight, strength and
deflections. Variations in beam width and depth generally depend on concreting and
vibration operations and unfixed forms. Furthermore, the quality of workmanship
and the construction techniques influence the beam width and depth dimensions.
In order to investigate the width and depth dimensional variations of beams, a total
of 3725 measurements were carried out in all of the building sites. Results are given
36
in Table 3.9. The nominal (specified) beam external dimensions were usually
between 250 to 500 mm for width and 300 to 1150 mm for depth. The mean to
nominal ratio changed within a range of 0.980 and 1.011 with an overall mean value
of 0.998 for beam width, and 0.967 and 1.012 with an overall mean value of 0.996
for beam depth. The basic variability changed between 3.1% and 8.3% and the
overall mean value was equal to 4.5% for the beam width; whereas for the beam
depth, the basic variability changed between 1.3% and 6.6% and the overall mean
value was equal to 2.5%.
Combination of the basic variability and the overall prediction error results in the
total variability on the beam width and depth, which are Ω bw = (0.045 2 + 0.03 2 ) =
0.054 and Ω bd = (0.025 2 + 0.032 ) = 0.04 , respectively.
The average of mean to nominal ratio is found to be 0.998 and 0.996 based on the
analysis of the collected data for beam width and depth, respectively. This ratio is
pointed out as 1.0 by Yücemen and Gülkan (1989) and Kömürcü (1995) for Turkey.
37
Table 3.9 Results of the statistical analysis of beam external dimensions for
different regions
Region Dimension Number of Design Mean of Mean of basic

observations value mean to variability
(mm) nominal
ratio
Width 515 200-600 1.007 0.046
Istanbul
Depth 580 300-850 0.998 0.022
Width 363 200-500 0.999 0.040

Ankara
Depth 345 500-1000 0.999 0.021
Width 247 200-600 0.996 0.042

Konya
Depth 224 300-900 0.998 0.021
Width 306 200-600 0.995 0.046

Adana
Depth 356 400-1150 0.991 0.030
Width 226 200-500 0.995 0.044

Malatya
Depth 215 350-800 0.994 0.027
Width 165 200-600 0.988 0.053

Ordu
Depth 183 350-800 0.994 0.031
Width 1822 250-600 0.999 0.045

Overall
Depth 1903 300-1150 0.996 0.025
3.3.3 Beam Effective Depth
Effective depths of reinforced concrete beams are directly dependent on the position
of the top and bottom reinforcing steel bars, which are influenced by several factors,
e.g. external dimensions, chair heights, concrete placing operations and lack of
transverse support. In addition to these, the position of beam reinforcement changes
as a result of conflicts with other exterior bars at beam to beam or beam to column
intersections. Besides, the exterior bars of beam may be formless due to unsuitable
stirrups or truss bars and also, the stirrups are opened as people walk on the beam
38
for concrete placing.
Since the direct measurement of the effective depth of a beam is rather difficult, the
depth of the steel cage and the beam depth are taken into consideration together.
Effective depth of a beam can be calculated by using Eq. (3.3). The dimensions of
the steel cage were measured before the placement of concrete. The specified d*
ranged from 300-600 mm and the basic variability was found to be 0.023, as shown
in Table 3.10.
1 *
d be = (d b + d b ) (3.3)
2
where:
dbe: beam effective depth
db: beam depth
db*: depth of steel cage
The distance between the outer sides of the steel cage and the external dimensions
of beam can be treated as perfectly correlated variables. Based on the collected data,
the mean of mean to nominal ratio of the beam effective depth is found to be 1.00,
and the average basic variability in the beam effective depth is computed as 0.024
by using Eq. (3.3).
Table 3.10 Results of the statistical analysis of beam internal dimensions before the
placement of concrete
Dimension Number of Design Mean of mean Mean of basic

description observations value to nominal variability
(mm) ratio
Steel cage depth (db*) 415 300-600 0.998 0.023
Depth (db) 1903 300-1150 0.996 0.025
39
The total prediction error and the basic variability in the effective depth are
combined to find the total variability; yielding Ω d be = (0.024 2 + 0.07 2 ) = 0.074 .

Ellingwood and Ang (1972) suggested the total prediction uncertainty as 0.086. In
similar studies conducted by Yücemen and Gülkan (1989) and Kömürcü (1995),
total variability in the beam effective depth was found to be 8%.
3.3.4 Column Width and Depth
Variations in reinforced concrete column external dimensions affect the load

carrying capacity of columns. These variations are results of concreting and
vibration operations and unfixed forms. The most important problem on column
concreting and vibration operations is the opening of molds during the casting of
fresh concrete. This causes an increase in external dimensions due to the weight of
concrete and the vibration effect. The pressure caused by the concrete pump
machine also expands the external dimensions of columns. In the light of this
explanation, the quality of workmanship and the construction techniques stands out
as a significant property.
A total of 4216 measurements were taken for columns in all building sites. Results
are given in Table 3.11. The nominal column external dimensions were usually
between 200 and 800 mm for width and between 350 and 1200 mm for depth. The
mean to nominal ratio ranged from 0.932 to 1.027 with an overall mean value of
1.007 for column width, and ranged from 0.922 to 1.033 with an overall mean value
of 1.013 for column depth. The basic variability was found to be 1.8%-9.3% and the
overall mean value was equal to 3.2% for column width whereas, for the column
depth, the basic variability changed between 1.2%- 7.2% and the overall mean
value was equal to 2.4%.
40
Since the actual external dimensions of columns are greater than nominal external
dimensions due to the opening of molds during the casting of fresh concrete, a
correction factor, N, is introduced to take into consideration this effect. In order to
find the value of N, 42 observations were carried out in the construction site. First,
the forms constructed with exact interior sizes were chosen and measured before the
placement of concrete. Then, after the removal of forms, measurements of columns
were again taken. At the end of observations and measurements, a mean value for N
is recommended as 1.02 whereas the prediction error associated with N is not
considered. The mean of mean to nominal ratio are corrected by these correction
factors yielding to: 1.02x1.007=1.02 and 1.02x1.013= 1.03 for width and depth,
respectively.
When the basic variability and the overall prediction error are combined, the total
variability in the column dimensions becomes Ω cw = (0.032 2 + 0.03 2 ) = 0.044 for
width and Ω cd = (0.024 2 + 0.03 2 ) = 0.038 for depth.
41
Table 3.11 Results of the statistical analysis of column external dimensions for
different regions
Region Dimension Number of Design Mean of mean Mean of

observations value (mm) to nominal basic
ratio variability
Width 580 350-800 1.018 0.027
Istanbul
Depth 596 400-1200 1.019 0.020
Width 431 300-600 1.009 0.029

Ankara
Depth 441 400-1200 1.020 0.024
Width 315 200-600 1.00 0.029

Konya
Depth 313 400-1200 0.999 0.029
Width 304 250-600 1.002 0.038

Adana
Depth 311 400-1000 1.016 0.019
Width 252 300-600 1.013 0.030

Malatya
Depth 237 400-1200 1.015 0.021
Width 215 350-600 0.984 0.048

Ordu
Depth 221 400-800 0.996 0.039
Width 2097 250-800 1.007 0.032

Overall
Depth 2119 400-1200 1.013 0.024
3.3.5 Column Effective Depth
Effective depth of a reinforced concrete column is directly related with the position
of the steel cage which is considerably affected by overall depths and widths of
columns. At the same time, the exterior bars of steel cage may be out of position
due to formless stirrups. In addition to these, the position of column reinforcement
can be affected by the conflicts with other exterior bars at column to column or
beam to column intersections.
Due to difficulties of direct effective depth measurement, the distance between the
42
outer sides of the steel cage was measured as well as column depth, but
measurements of steel cage were taken before concrete placing. The specified d*
ranged from 300 to 700 mm and the basic variability changed from 1% to 4.2%
with an average value of 2.7%. The results for steel cage depth and column depth
measurements are given in Table 3.12. Effective depth of columns can be calculated
by the following relationship with the help of these results:
1 *
d ce = (d c + d c ) (3.4)
2
where:
dce: effective depth of column
dc: column depth
dc*: depth of steel cage
The depth of steel cage and the external dimension of the reinforced concrete
columns can be treated as perfectly correlated. Based on the column external
dimension and the steel cage depth, the mean of mean to nominal ratio and the
average basic variability of the column effective depth are found to be 1.01 and
0.025, respectively.
It is necessary to combine the basic variability in the effective depth and the total
prediction error in order to find the total variability. Thus, the total variability in the
effective depth of a column becomes Ω d ce = (0.025 2 + 0.07 2 ) = 0.074 .
43
Table 3.12 Results of the statistical analysis for the column internal dimensions
Dimension Number of Design Mean of mean Mean of basic

description observations value to nominal variability
(mm) ratio
Steel cage depth (dc*) 555 300-700 1.005 0.027
Depth (dc) 2119 400-1200 1.013 0.024
3.3.6 Shear Walls
Concreting and vibration operations and unfixity of forms influence the variations
in shear wall external dimensions. One of the main problems on shear walls like
columns is the opening of molds during casting of fresh concrete which causes an
increase in dimensions due to the weight of concrete, vibration effect and
disturbance caused by concrete pump machine.
To examine the variations of external dimensions on shear walls, 2942 dimensions

were measured. Results are given in Table 3.13. As the nominal shear wall widths
were 200, 250 or 300 mm, the nominal shear wall depths usually varied between
1200 and 1500 mm. The mean to nominal ratio ranged from 0.932 to 1.027 with an
average value of 1.003 for width and 0.922 to 1.033 with an average value of 0.998
for shear wall depth. The basic variability was found to be 2%-7.6% with an
average value of 4.7% and 0.4%- 2.7% with an average value of 1.4% for shear
wall width and shear wall depth, respectively.
44
Table 3.13 Results of the statistical analysis of shear wall external dimensions for
different regions
Region Dimension Number of Design Mean of mean Mean of basic

observations value (mm) to nominal variability
ratio
Width 348 200-300 1.006 0.045
Istanbul
Depth 366 1200-4400 1.00 0.014
Width 257 200-300 1.004 0.043

Ankara
Depth 252 1350-5000 0.998 0.014
Width 235 200-300 1.004 0.054

Konya
Depth 238 1200-3000 0.999 0.014
Width 284 200-350 1.001 0.048

Adana
Depth 270 1200-4000 1.001 0.012
Width 195 200-300 0.998 0.051

Malatya
Depth 199 1450-2500 0.995 0.015
Width 141 200-300 1.006 0.044

Ordu
Depth 157 1200-1800 0.995 0.018
Width 1460 250-350 1.003 0.047

Overall
Depth 1482 1200-4400 0.998 0.014
As a result of the opening of molds during casting of fresh concrete, actual shear
wall width is greater than the nominal value. A correction factor, N, is introduced
to take this effect into consideration. Since the depth of shear wall is considerably
greater than other dimensions, the ratio of increase in this dimension is quite small;
therefore, the correction factor related to shear wall depth is ignored. In order to
find the value of N, a total of 36 observations were carried out. First, the forms
constructed with exact interior sizes were chosen and measured before the
placement of concrete. Then, after the removal of forms, measurements of shear
wall widths were again taken. At the end of the observations and the measurements
carried out only for shear wall width, a mean value for N is recommended as 1.05
and the prediction error associated with N is ignored.
45
If we correct the mean of mean to nominal ratio; 1.05x1.003=1.05 is found for the
shear wall width. On the other hand, as it is seen on Table 3.13, the mean of mean
to nominal ratio in shear wall depth is equal to 0.998 based on the analysis of the
collected data.
When the basic variability and the overall prediction error are combined, the total
variability in the shear wall width and depth, respectively, becomes:
Ω sw = (0.047 2 + 0.03 2 ) = 0.056 and Ω sd = (0.014 2 + 0.03 2 ) = 0.033
3.3.7 Reinforcement Area
The basic variability in the reinforcement area of any structural element can
be computed from the variability in the individual bar area assuming that there is a
perfect correlation among the bar areas. Except for the fabrication errors, the other
errors do not considerably affect the variability of reinforcement bar areas. In
Turkey, BCIII(a) type reinforcing bar is widely used. Thus, test results on 1587
BCIII(a) reinforcing bar specimens with different bar sizes from 8mm to 22 mm
were obtained. After the analyses of the data, the mean of mean to nominal ratio,
standard deviation and mean of basic variability of each bar size were determined.
The results are presented in Table 3.14.
46
Table 3.14 Results of the statistical analysis of reinforcement areas
according to bar sizes
Bar Number of Mean of mean to Standard Mean of basic

size observations nominal ratio deviation variability
8 185 1.01 0.16 0.02
10 172 0.99 0.10 0.01
12 256 1.00 0.15 0.013
14 126 1.01 0.15 0.011
16 185 1.00 0.13 0.008
18 106 1.00 0.19 0.011
20 284 0.99 0.26 0.013
22 112 0.99 0.26 0.012
25 96 1.03 0.31 0.012
Overall 1522 1.00 0.20 0.012
As it could be seen from Table 3.14, the ratio of mean to specified values of
reinforcing steel bar areas and the basic variability in reinforcement area is equal to
1.00 and 1.2%, respectively, based on the overall data. Prediction uncertainties arise
from fabrication errors, and the mean areas of smaller diameter bars are less
predictable (Ellingwood and Ang, 1972). To account for the additional sources of
variability, Ellingwood and Ang (1972) and Kömürcü (1995) used a total prediction
error of 3% in their studies. Here, the total prediction uncertainty will also be taken
as 3%. Thus, the total uncertainty of reinforcement area, quantified in terms of
c.o.v., equals to Ω As = (0.012 2 + 0.03 2 ) = 0.03 .
47
CHAPTER 4
MODELING OF CAPACITY OF REINFORCED

CONCRETE MEMBERS IN DIFFERENT FAILURE
MODES
In the previous chapter, various sources of uncertainties associated with concrete

compressive strength, yield strength of reinforcing steel bars and the dimensions of
beams, columns and shear walls are analyzed and quantified. Based on these
uncertainties, the resistance statistics for flexural and shear failure modes of
reinforced concrete beams, columns and shear walls as well as columns in
combined action of axial load and flexure failure modes are computed within the
framework of reliability analysis in the following sections.
4.1 CAPACITY OF REINFORCED CONCRETE BEAMS IN DIFFERENT

FAILURE MODES
Beams are structural members whose main function in a structure is to carry

transverse loads that create flexural moments and shear forces. Due to monolithic
nature of reinforced concrete structures, beams can be subjected to axial loads and
torsional moments as secondary loads in addition to transverse loads. In this study,
only flexural and shear capacities are treated since they are usually the most critical
parameters that govern the beam design. The statistics of basic variables concerning
the flexural and shear capacities of beams are summarized in Table 4.1.
48
Table 4.1 Statistics of the basic variables involved in the calculation of flexural and
shear capacities of reinforced concrete beams
Nominal
Parameter Mean to Inherent Prediction Total
(Specified)
nominal ratio variability error Uncertainty
value
Compressive strength
21.55 MPa 1.25 0.105 0.14 0.18
of concrete (fc)
Yield strength
365 MPa 1.24 0.038 0.08 0.09
of BC III (f y)
Beam width (bw) 250-600 mm 0.998 0.045 0.03 0.054
Beam depth (h) 300-1150 mm 0.996 0.025 0.03 0.04
Effective depth (dbe) 250-1100 mm 1.00 0.024 0.07 0.074
Reinforcement area
100-4000 1.00 0.012 0.03 0.03
(As)
mm2
4.1.1 Flexural Capacity of Beams
When beams are under the influence of flexure, bending strains are produced. It
should be noted that under positive bending moments, tensile strains are produced
in the bottom of the beam while compressive strains are produced in the top of the
beam. Accordingly, these strains produce tensile stresses in the bottom of the beam,
and compressive stresses in the top of the beam. Hence, both tensile stresses and
compressive stresses must be resisted by beams in a suitable manner.
The skill in determining the sense of bending is highly critical in the design of
reinforced concrete members. The concrete must be embedded with steel on the
tension side since the tensile strength of concrete is very low compared with its
compressive strength. The absence of properly placed steel will most likely cause
structural failure (Meyer, 1996).
49
By using the equivalent rectangular stress block approach, the nominal flexural
strength of a rectangular reinforced concrete beam can be determined. A typical
rectangular beam cross-section with strain, stress and force distribution diagrams
are shown in Figure 4.1. The equivalent rectangular block of intensity 0.85fc and
depth, a, as shown in this figure. The depth of the equivalent stress block “a” can be
determined by the following equation (TS 500, 2000):
a= k1c (4.1)
where:
k1: a factor of function of strength of concrete that takes a value of 0.85 for C16,
C20, C25, a value of 0.82 for C30, and a value of 0.79 for C35 (TS 500, 2000)
c: distance from the neutral axis to the outer compressive fiber.
h d
As
dı
bw
Cross-section
0.003 0.85fc
a
Fc
Neutral
axis
(d-a/2)
Fs
Strains Stresses Internal forces
Figure 4.1 Stresses and forces in reinforced concrete beams
50
In a reinforced concrete beam, flexural failure may occur in three different ways:
a) Balanced failure: This type of failure occurs when yielding of tension steel
and crushing of concrete in the extreme concrete fiber take place at the same
time; for this reason, the behavior is brittle. The amount of steel required to
produce this condition, which is a borderline case, is large.
b) Tension failure: This type of failure occurs when the balanced steel ratio is
higher than the steel ratio of the cross section. Such beams are said to be
underreinforced.
c) Compression failure: This type of failure occurs when the concrete in the
compression zone crushes before the steel in the tension zone yields; that is,
if a beam has a larger amount of steel than that required to create the
balanced condition, compression failure occurs. The beams showing this
type of failure are said to be overreinforced.
Since compression failure is dominated by concrete, the behavior of an

overreinforced beam is brittle and sudden. For this reason, overreinforced beams are
forbidden in design codes. In order to determine whether a beam is underreinforced
or overreinforced, balanced case should be investigated first. It is to be noted that
the equivalent rectangular stress block approach assumes a value of εcu=0.003 for
the strain in the extreme compression fiber at ultimate stage in the all types of
failure (TS 500, 2000). The balanced steel ratio, ρb, and the existing steel ratio, ρ,
can be computed from Eq. (4.2) and Eq. (4.3), respectively.
0.85f cd cb
ρb = k1 (4.2)
f yd d
As
ρ= (4.3)
bwd
where cb is the depth of neutral axis at the balanced case. In the case of double
51
reinforced beams, (ρ- ρ ′ ) should be checked against balanced steel ratio ρb to
control whether the beam is overreinforced or underreinforced. Here, ρ ′ is the
compression reinforcement ratio.
Reinforced concrete floor systems usually consist of beams and slabs that are placed
monolithically. Typically, a certain portion of a slab which is adjacent to a beam
becomes a part of the beam. This part of the slab that behaves together with the
beam against the flexural moment gives rise to a beam section of T or L shapes as
shown in Figure 4.2. In reinforced concrete structures, the most commonly observed
beams are this type of beams. Therefore, in this study, flanged beams are treated as
reinforced concrete beams.
bw
Figure 4.2 Beam sections of T and L shapes (Ersoy and Özcebe, 2004)
Since flanged beams ensure a large compression area in the compression zone of
the beam, they can carry high compression force with this portion of the section
without necessitating additional steel reinforcement in this zone. Flanged beams are
advantageous owing to that the shifting of the centroid of the compression zone
toward the compression face raises the moment arm, which leads to a great benefit
in terms of moment carrying capacity of the beam.
52
The neutral axis of a T beam may be either in its flange or in the web (i.e. the part
of the T beam below the slab) depending upon the strength of materials, the cross
sectional properties and the amount of reinforcing steel bars (Nilson et al., 2004). If
the depth of the flange thickness is greater than the compression block, k1c (i.e.
t≥k1c), then the analysis of a flanged beam is identical to the analysis of a
rectangular beam. The depth of the flange thickness is rarely less than the depth of
the compression block except in precast beams (Ersoy and Özcebe, 2004). For this
reason, in this study, the analysis of the flanged beam is carried out in the same way
as the analysis of a rectangular beam.
Using the equivalent rectangular stress block approach, the ultimate moment
capacity, Mr, of a flanged beam where “t≥k1c” can be computed by using the
following equations:
 k c
M r = A s f yd  d − 1  (4.4)
 2 
A s f yd
k1c = (4.5)
0.85f cd b
In order to compute the mean to nominal ratio and the total variability for the
flexural capacity of a beam, a code is written in MathCAD 12 in accordance with
FOSM method, and beams with various dimensions and with different
reinforcement areas are analyzed using this code. Besides, the beams whose depth is
much smaller than its width are considered. The results are given in Tables 4.2 and
4.3. It is to be noted that, in these tables, the mean to nominal ratios of the flexural
capacity of the beams are computed by substituting the in-situ and nominal values
of the basic variables into the following equation (a bar “ - ” over a basic variable
stands for the “true” mean and a prime “ ′ ”denotes the nominal value):
53
 A s f yd 
A s f yd  d − 
Mr  1.7f b 
=  cd 
(4.6)
M ′r  A′ f ′ 
′  d ′ − s yd 
A ′s f yd
 ′ b ′ 
1.7f cd

From Tables 4.2 and 4.3, it can be seen that M r / M ′r =1.24 as an average value for
the studied cases. The corresponding value was suggested by Ellingwood et al.
(1980) as 1.14 and 1.05 for Grade 40 and Grade 60 steels, respectively. In a similar
study, Kömürcü and Yücemen (1996) reported a value of 1.19 for the flexural
capacity of beams. In the following chapters, a value of 1.24 will be used in the
reliability calculations.
Considering the total uncertainties of the basic variables involved in the flexural
capacity of a beam, the overall standard deviation, σ Mr , is calculated by utilizing the
FOSM format. For the flexural capacity of a beam, based on Eq. (2.11), the
standard deviation equation can be rewritten as follows:
2 2 2 2 2
 ∂M   ∂M   ∂M   ∂M   ∂M 
σ Mr =  r  σ 2As +  r  σ fy2 +  r  σ d2 +  r  σ fc2 +  r  σ 2b (4.7)
 ∂As   ∂fy   ∂d   ∂fc   ∂b 
After the computation of standard deviation by using Eq. (4.7), the total variability
in terms of coefficient of variation is calculated by dividing the standard deviation
to the mean value. As given in Tables 4.2 and 4.3, the total variability varies from
0.119 to 0.141 with a weighted average value of 0.12. In addition, it is necessary
that the analysis uncertainty (i.e. epistemic uncertainty resulting from the difference
between actual behavior and estimated behavior predicted by analysis) should be
included into the overall uncertainty. For this reason, Ellingwood et al. (1980) and
Nowak and Szerszen (2003a) used the values of 0.046 and 0.06, respectively in
their studies. Here, the analysis uncertainty is taken as 0.06 in order to be on the
54
safe side. Thus, the total uncertainty in the flexural capacity of a reinforced concrete
beam becomes Ω M r = (0.12 2 + 0.06 2 ) = 0.13 .
Ellingwood et al. (1980) used the values of 0.14 and 0.11 for Grades 40 and Grades
60 steels, respectively. Yücemen and El-Etoom (1986) reported Ω Mr to be 0.16.
Kömürcü and Yücemen (1996) also calculated this value as 0.16 in their study.
Despite being slightly smaller, the value found in this study is consistent with the
values reported in the studies mentioned above.
55
Table 4.2 Mean to nominal ratios and coefficients of variation of the beam flexural
capacity (b=200 mm and b=300 mm)
b d As Mr
ΩM
(mm) (mm) (mm2) Mr ′
100 1.24 0.120
200 1.24 0.120
260 400 1.241 0.119

600 1.241 0.119
800 1.241 0.121
200 1.240 0.120
400 1.240 0.119
360 800 1.241 0.120

200
1200 1.241 0.119
2000 1.242 0.132
200 1.240 0.120
400 1.240 0.119
460 900 1.241 0.119

1200 1.241 0.120
2000 1.242 0.126
400 1.240 0.119
800 1.241 0.120
360 1600 1.242 0.128

2400 1.244 0.152
400 1.240 0.119
800 1.241 0.119
300
460 1600 1.242 0.123
2400 1.243 0.134
400 1.242 0.120
800 1.242 0.119
560 1600 1.242 0.121

2400 1.242 0.127
Average value 1.241 0.123
56
Table 4.3 Mean to nominal ratios and coefficients of variation of the beam flexural
capacity (b=400 mm and b=1000 mm)
b d As Mr
ΩM
(mm) (mm) (mm2) Mr ′
800 1.242 0.119
460 1600 1.242 0.119

2400 1.242 0.121
3200 1.242 0.123
800 1.242 0.120
560 1600 1.242 0.119

400
2400 1.242 0.120
3200 1.242 0.121
800 1.242 0.120
660 1600 1.242 0.119

2400 1.242 0.119
3200 1.242 0.120
800 1.242 0.119
210 1600 1.242 0.119
3200 1.242 0.122
4000 1.242 0.125
800 1.242 0.120
260 1600 1.242 0.119

3200 1.242 0.120
1000
4000 1.242 0.122
800 1.242 0.120
360 1600 1.242 0.119
3200 1.242 0.119
4000 1.242 0.120
4.1.2 Shear Capacity of Beams
Reinforced concrete structural members are usually subjected to shear besides

flexural and axial loads. In the classical sense, not only shear but also torsion
produces shear flows in structural members. While shear strength of concrete is
57
lower than its compressive strength, it is considerably larger than its tensile
strength. In fact, shear by itself does not usually create any critical condition in
reinforced concrete members. Nevertheless, the principal tensile stresses caused by
shear can produce an important problem in reinforced concrete members, as the
tensile strength of concrete is very low (Ersoy and Özcebe, 2004).
If there is inadequate shear reinforcement in a structural member, structural

collapse, which is brittle, can occur owing to the principal tensile stresses. In order
to prevent this type of failure, structural members should be provided with adequate
shear reinforcement so that these members can reach their ultimate limit state by
flexure, but not by shear. Therefore, in the design of structural members, the
following equation should be satisfied:
Vr ≥ Vd (4.8)
where:
Vr: shear strength of the member,
Vd: maximum design shear force to be calculated based on convenient load factors
and combinations.
TS500 (2000) defines the shear strength of a structural member as the sum of the
resistance of concrete and the resistance of the shear reinforcement. The resistance
of concrete, Vc, which is commonly taken as 80% of the cracking shear strength of
concrete and the resistance of shear reinforcement, Vw, can respectively be
calculated by using the Eq. (4.9) and Eq. (4.10), as specified in TS500 (2000):
Vc = 0.80(0.65f ctd b w dΨ ) (4.9)
A sw
Vw = f ywd d (4.10)
s
where:
58
fctd: design tensile strength of concrete
fywd: design yield strength of shear reinforcement
Asw: cross-sectional area of stirrup
s: spacing of stirrups
If a reinforced concrete member is subjected to axial stresses in addition to shear

resulting from the applied axial forces or imposed deformations, such as shrinkage
and temperature change, the magnitude and direction of the principle tensile stresses
can change. A stress analysis indicates that axial tensile stresses lower the diagonal
cracking resistance whereas axial compressive stresses increase it. The value of Ψ
given in Eq. (4.9) can be found from the following formulas for beams subjected to
axial tension or axial compression besides shear, respectively:
Nd
Ψ = 1 + 0.3 (4.11)
Ac
Nd
Ψ = 1 + 0.007 (4.12)
Ac
It is worth noting that Nd has to be entered (-) in Eq. (4.11) and (+) in Eq. (4.12).
When Nd/Ac<0.5 N/mm2, one can take Ψ as 1.0.
Based on the discussion above, shear strength of a reinforced concrete beam can be
calculated by using the following relationship:
A sw
Vr = f ywd d + 0.52f ctd b w dΨ (4.13)
s
In order to find the mean to nominal ratio and the total variability of the shear
capacity of a beam, a code was developed in MathCAD 12 according to FOSM
format. Various cases with different reinforcement areas and dimensions are taken
into account by using this code, and the results are presented in Table 4.4. The mean
59
to nominal ratio of the shear capacity of a beam is obtained by substituting the in-
situ and nominal values of the basic variables into the following equation:
A sw
f ywd d + 0.52f ctd b w d Ψ
Vr s
= (4.14)
Vr′ A ′sw
f ywd
′ d ′ + 0.52f ctd
′ b ′w d ′Ψ ′
s′
From Table 4.4, one can infer that for the studied cases Vr / Vr′ has an average
value of 1.24. This value was indicated by Ellingwood et al. (1980) as 1.09. In a
similar study, Nowak and Szerszen (2003a) reported the value of 1.23 for the shear
capacity of the beam. In the following chapters, a value of 1.24 will be used in the
reliability calculations.
Considering the total uncertainties of the basic variables involved in the shear
capacity of a beam, the overall standard deviation, σ Vr , is calculated on the basis of
FOSM procedure. For the shear capacity of a beam, by using Eq. (2.11), the
standard deviation equation can be rewritten as follows:
 ∂V 2 2 2 2 
  r  σ 2 +  ∂V r 
  σ2 + 
 ∂V r 
 σ2 +
 ∂V r 
  σ2 
  ∂ As  As  ∂s  s  ∂ fyw  fyw  ∂d  d
        
σ
V
=
 2 2 2  (4.15)
r   ∂ V r   ∂V r   ∂V r  
+  σ2 +  σ2 +  σ2

 fct  ∂ bw  bw  ∂Ψ  Ψ
  ∂ fct      
If the standard deviation of the shear capacity is divided by its mean value, the total
variability in the shear capacity of the beam is obtained. From Table 4.4, it can be
seen that the total variability changes within a range of 0.116 to 0.157 with a
weighted average value of 0.14. On the other hand, the actual behavior of shear
capacity of a beam is different from the estimated behavior predicted by structural
analysis. Ellingwood et al. (1980) and Nowak and Szerszen (2003a) assigned 0.115
and 0.10 in their studies, respectively as the c.o.v. to account for this effect. In this
60
study, the epistemic uncertainty resulting from the effect mentioned above (analysis
uncertainty) will be taken as 0.10 as reported in the study of Nowak and Szerszen
(2003a) due to fact that it is a more recent study and contains up to date data and
information in comparison with the study of Ellingwood et al. (1980).
Consequently, the overall uncertainty in the shear capacity of a beam will be
Ω Vr = (0.14 2 + 0.10 2 ) = 0.17 .
Ellingwood et al. (1980) reported Ω Vr to be 0.17. Nowak and Szerszen (2003a)

used a value of 0.11. In spite of the fact that there is significant difference between
the results obtained from studies conducted by Ellingwood et al. (1980) and Nowak
and Szerszen (2003a), the value found by Ellingwood et al. (1980), which is the
same as that also computed in this study, is on the safe side. Therefore, a value of
0.17 will be used for the overall uncertainty of the shear capacity of a beam.
61
Table 4.4 Mean to nominal ratios and coefficients of variation of the beam shear
capacity
bw h d Asw s Vr
ΩV
(mm) (mm) (mm) (mm2) (mm) Vr′
50 180 1.244 0.132
200 400 370 78.5 180 1.243 0.122
100 280 1.243 0.126
50 145 1.244 0.132
250 500 460 100 145 1.243 0.119
100 230 1.244 0.127
100 330 1.244 0.136
250 700 660 100 400 1.244 0.138
200 200 1.242 0.116
50 280 1.245 0.157
300 600 570 78.5 280 1.245 0.144
100 280 1.244 0.136
78.5 220 1.244 0.137
300 500 470 100 280 1.244 0.136
200 400 1.244 0.147
100 330 1.245 0.146
350 700 660 78.5 320 1.246 0.157
200 320 1.243 0.125
78.5 200 1.245 0.142
400 1000 960 100 250 1.245 0.142
200 350 1.243 0.132
78.5 100 1.245 0.142
800 300 270 100 150 1.245 0.147
200 200 1.244 0.135
62
4.2 CAPACITY OF REINFORCED CONCRETE COLUMNS IN
DIFFERENT FAILURE MODES
Columns are vertical structural members which are designed to support load-
carrying beams. The height of a column is generally considerably larger than its
cross-sectional dimensions. These structural members transfer loads from the roof
or the upper floor to the lower levels, and then to the soil through foundations.
Since columns are compression members, failure of one column in a critical region
can lead to progressive collapse of the adjoining floors resulting in total collapse of
the entire structure. Structural column failure is of major importance in terms of not
only economic but also human loss. Hence, extreme care should be given to design
a column with a higher reserve strength compared to beams and other horizontal
structural elements, especially since compression failure provides little visual
warning (Nawy, 2005).
Columns can be classified into four general types as tied columns, spiral columns,
composite columns and concrete-filled pipe columns, generally according to the
forms used and the arrangement of reinforcing steel bars (Atimtay, 1998):
- In tied columns, which are usually in rectangular, square or L shape and rarely in
circular shape, the longitudinal reinforcement bars are held in position by
intermitted lateral ties,
-In spiral columns, which are usually in circular, and rarely in rectangular or square
shape, the longitudinal reinforcement bars are enveloped by a continuous helix of
steel bar or wire,
-In composite columns, which are usually in rectangular, square or circular shape,
steel profiles are thoroughly encased in concrete, which is also reinforced with both
longitudinal and spiral reinforcement,
63
-Concrete filled pipe columns are manufactured by filling steel pipes with concrete.
Macgregor and Wight (2005) indicated that over 95 percent of the columns in the
buildings located in nonseismic regions are tied columns. In spite of the fact that
spiral columns are also used where increased ductility is needed, tied columns are
the most commonly used type of columns due to their lower construction costs
(Nawy, 2005). In Turkey, the number of the rectangular tied columns is
considerably more than the number of the other column types. For this reason, in
this study, rectangular tied columns will be studied. The statistics of the basic
variables concerning the flexural and shear capacity of reinforced concrete columns
are summarized in Table 4.5.
Table 4.5 Statistics of the basic variables involved in the calculation of the
combined action and shear capacities of columns
Nominal
(Specified)
nominal ratio variability error Uncertainty
value
21.55 MPa 1.25 0.105 0.14 0.18
of concrete (fc)
Yield strength
365 MPa 1.24 0.038 0.08 0.09
of BC III (f y)
Column width (bw) 250-800 mm 1.02 0.032 0.03 0.044
Column depth (h) 400-1200 mm 1.03 0.024 0.03 0.038
Effective depth (dbe) 350-1150 mm 1.01 0.25 0.07 0.074
Reinforcement area
1400-6400 1.00 0.012 0.03 0.03
(As)
mm2
4.2.1 Combined Action of Flexure and Axial Load for Columns
A great majority of columns as a part of monolithic reinforced concrete structures

are not subjected to only uniaxial loading. Due to asymmetrically placed floor
64
loads, initial crookedness of the column, and the possibility of lateral loads, such as
wind load and earthquake load, flexural moment, shear and sometimes torsion take
place in addition to axial compression in structural columns. In fact,
nonhomogeneous nature of concrete creates an eccentricity even when only an axial
load is applied to the geometric centroid of the cross-section. This is why, most of
the design codes avoid column designs with no moment, and assign a minimum
eccentricity in order to stay on the safe side (Ersoy and Özcebe, 2004). Therefore,
not only axial forces but also moments, namely the combined action of flexure and
axial load, are taken into consideration for the columns considered in this study.
The design of an eccentrically loaded column necessitates the determination of two

basic quantities: (i) the moment that can be resisted for a specified axial load and
(ii) the axial load that can be carried for a specified moment.
The same basis concerning the stress distribution and the equivalent rectangular
stress block utilized for beams are equally applicable for columns (Nawy, 2005).
As stated above, most of the columns in structures are tied columns, which are of
rectangular cross-sections. A typical rectangular column cross-section with strain,
stress and force distribution diagrams are shown in Figure 4.3. For rectangular
sections, the equivalent rectangular stress block approach leads to the following
equations for the calculation of axial load, Nr, and moment capacity, Mr:
n
N r = 0.85f cd k 1c b + ∑ A si σ si (4.16)
i =1
h k c n
M r = 0.85f cd k 1c  − 1  + ∑ A si σ si x i (4.17)
2 2  i =1
 h
 xi − 
σ si = 0.003E s 1 + 2≤f (4.18)
yd
 c 
 
 
65
where:
Asi: steel area in the ith layer
σsi: steel stress in the corresponding steel area in ith layer
Es: modulus of elasticity of steel reinforcement
xi: distance between neutral axis and ith steel layer
dı
As3
d
h As2
As1
Cross-section
N
0.003 0.85fc
e Fs3
Neutral axis εs3 k1c
Fc
x Fs2
εs2
Fs1
εs3
Strains Stresses Internal forces
Figure 4.3 Stresses and forces in reinforced concrete columns
As mentioned above, all columns in a structure should be designed with respect to

the combined action of flexure and axial load. All possible load combinations
should be taken into consideration in the structural analysis. For each load
combination, a pair of N and M values is computed. Then, the plot of M versus N
obtained from these three equations for a given cross-section represents the strength
envelope called interaction diagrams or interaction curves. Most of the structural
codes limit the axial load on a column and also specify a minimum eccentricity. The
column design should be carried out for the most critical Nd and Md combinations
66
which are determined from the analysis.
In order to obtain the values of mean to nominal ratio and total variability for the
combined action of axial and flexural capacity of a column, a code was written in
MathCAD according to FOSM. With the help of this code, the mean to nominal
ratio and total variability are computed for different cases with various
characteristics by using both the nominal and in-situ values as discussed before. For
different design cases, the statistics of mean to nominal ratios associated with the
combined action of flexure and axial load capacity of reinforced concrete columns
are summarized in Tables 4.6 - 4.9. From these tables, the mean to nominal ratio of
axial capacity of a column is observed to vary from 1.212 to 1.263 with a weighted
average of 1.25 whereas the mean to nominal ratio of flexural capacity vary from
1.145 to 1.276 with a weighted average of 1.22. Accordingly, the mean to nominal
ratio of the combined action in columns will be taken as 1.24, which corresponds to
the average of these two values. In similar studies, Ellingwood et al. (1980)
recommended a value of 1.10 for axially loaded columns while Nowak and
Szerszen (2003a) reported the value of 1.26 for this ratio.
As shown in Tables 4.6-4.9, the total variability changes in a range of 0.058 and
0.162 with a weighted average value of 0.12 for the axial capacity of a column
whereas for the flexural capacity of a column, the total variability ranges from
0.059 to 0.107 with a weighted average value of 0.07. To be on the safe side, the
total variability in the combined action of the axial and the flexural capacity of
columns will be taken as 0.12, which corresponds to the total variability in the axial
capacity of the column. In addition, it is to be noted that, a value of 0.046 for the
analysis uncertainty is reported by Ellingwood et al. (1980) in their study whereas
this value is reported as 0.08 by Nowak and Szerszen (2003a). Here, the analysis
uncertainty will also be taken as 0.08. Thus, the overall uncertainty for the
combined action of axial and flexural capacity of a column will be
Ω ca = (0.12 2 + 0.08 2 ) = 0.14 .
67
flexural capacities for eccentrically loaded columns (b/h=0.6)
dı As1
d d ′′ As1 1
h
As2
dıı = 0.9 , λ = =
As1
h 2As1 + As 2 4
Design case Axial force Moment

Nr Mr
ρ c ΩN ΩM
N ′r M ′r
c=1.5h 1.262 0.162 1.147 0.066
c=h 1.258 0.163 1.272 0.106
c=0.75h 1.260 0.160 1.170 0.073
0.01
c=0.5h 1.262 0.142 1.240 0.067
c=0.25h 1.260 0.097 1.240 0.068
c=cb 1.262 0.14 1.240 0.067
c=1.5h 1.254 0.149 1.147 0.065
c=h 1.246 0.148 1.271 0.106
c=0.75h 1.235 0.136 1.173 0.073
0.02
c=0.5h 1.261 0.113 1.240 0.068
c=0.25h 1.252 0.074 1.240 0.068
c=cb 1.255 0.116 1.240 0.067
c=1.5h 1.246 0.136 1.148 0.065
c=h 1.237 0.134 1.271 0.106
c=0.75h 1.222 0.121 1.172 0.073
0.03 c=0.5h 1.257 0.097 1.240 0.068
c=0.25h 1.249 0.063 1.240 0.068
c=cb 1.250 0.100 1.240 0.067
c=1.5h 1.240 0.125 1.148 0.065
c=h 1.228 0.124 1.271 0.106
c=0.75h 1.212 0.109 1.172 0.073
0.04
c=0.5h 1.255 0.087 1.240 0.068
c=0.25h 1.247 0.058 1.240 0.068
c=cb 1.246 0.089 1.240 0.068
Average value 1.248 0.118 1.218 0.075
68
flexural capacities for eccentrically loaded columns (b/h=1)
dı As1
d
h As2 dıı d ′′ As1 1
= 0.8 , λ = =
As1 h 2As1 + As 2 4

Nr Mr
ρ c ΩN ΩM
N ′r M ′r
c=1.5h 1.260 1.160 1.148 0.065
c=h 1.258 0.164 1.274 0.107
c=0.75h 1.252 0.157 1.174 0.074
0.01
c=0.5h 1.265 0.135 1.240 0.068
c=0.25h 1.258 0.100 1.240 0.068
c=cb 1.262 0.142 1.240 0.068
c=1.5h 1.249 0.142 1.147 0.065
c=h 1.245 0.147 1.274 0.107
c=0.75h 1.235 0.137 1.174 0.074
0.02
c=0.5h 1.259 0.108 1.240 0.068
c=0.25h 1.252 0.074 1.240 0.068
c=cb 1.255 0.116 1.240 0.068
c=1.5h 1.241 0.127 1.147 0.065
c=h 1.235 0.134 1.274 0.107
c=0.75h 1.223 0.121 1.174 0.074
0.03 c=0.5h 1.257 0.097 1.240 0.068
c=0.25h 1.249 0.063 1.240 0.068
c=cb 1.25 0.100 1.240 0.068
c=1.5h 1.234 0.115 1.147 0.065
c=h 1.226 0.123 1.274 0.107
c=0.75h 1.213 0.110 1.174 0.074
0.04
c=0.5h 1.253 0.082 1.240 0.068
c=0.25h 1.247 0.058 1.240 0.068
c=cb 1.246 0.089 1.240 0.068
Average value 1.246 0.158 1.219 0.075
69
flexural capacities for eccentrically loaded columns (b/h=2)
dı
As1
As2
dıı d ′′ As1 1
h AAs1 d = 0.8 , λ = =
h 2As1 + As 2 4

Nr Mr
ρ c ΩN ΩM
N ′r M ′r
c=1.5h 1.260 0.160 1.145 0.063
c=h 1.257 0.113 1.207 0.084
c=0.75h 1.253 0.158 1.181 0.075
0.01
c=0.5h 1.266 0.139 1.240 0.067
c=0.25h 1.258 0.100 1.240 0.067
c=cb 1.263 0.141 1.240 0.067
c=1.5h 1.248 0.141 1.145 0.063
c=h 1.244 0.147 1.206 0.084
c=0.75h 1.237 0.137 1.181 0.076
0.02
c=0.5h 1.261 0.113 1.240 0.067
c=0.25h 1.259 0.080 1.240 0.068
c=cb 1.259 0.117 1.240 0.067
c=1.5h 1.240 0.126 1.145 0.063
c=h 1.233 0.133 1.206 0.084
c=0.75h 1.225 0.122 1.181 0.076
0.03 c=0.5h 1.257 0.097 1.240 0.068
c=0.25h 1.249 0.063 1.240 0.068
c=cb 1.253 0.100 1.240 0.068
c=1.5h 1.233 0.115 1.128 0.063
c=h 1.224 0.122 1.206 0.084
c=0.75h 1.216 0.111 1.181 0.076
0.04
c=0.5h 1.255 0.087 1.240 0.068
c=0.25h 1.247 0.058 1.240 0.068
c=cb 1.247 0.088 1.240 0.068
Average value 1.248 0.115 1.208 0.071
70
flexural capacities for eccentrically loaded columns (b/h=0.5)
dı
As1
d d ′′ As1 1
As2 = 0.9 , λ = =
h d ıı
h As1 + As 2 3
As1

Nr Mr
ρ c ΩN ΩM
N ′r M ′r
c=1.5h 1.262 0.160 1.158 0.059
c=h 1.264 0.162 1.276 0.099
c=0.75h 1.245 0.158 1.160 0.068
0.01
c=0.5h 1.253 0.138 1.231 0.065
c=0.25h 1.245 0.106 1.234 0.064
c=cb 1.257 0.144 1.233 0.066
c=1.5h 1.252 0.141 1.158 0.059
c=h 1.256 0.146 1.276 0.099
c=0.75h 1.231 0.138 1.159 0.068
0.02
c=0.5h 1.240 0.111 1.231 0.065
c=0.25h 1.233 0.078 1.234 0.064
c=cb 1.247 0.119 1.233 0.066
c=1.5h 1.244 0.127 1.158 0.059
c=h 1.25 0.133 1.276 0.099
c=0.75h 1.217 0.122 1.159 0.068
0.03 c=0.5h 1.232 0.094 1.231 0.065
c=0.25h 1.228 0.066 1.234 0.064
c=cb 1.239 0.162 1.233 0.066
c=1.5h 1.238 0.115 1.158 0.059
c=h 1.244 0.122 1.276 0.099
c=0.75h 1.207 0.111 1.159 0.068
0.04
c=0.5h 1.227 0.083 1.231 0.066
c=0.25h 1.224 0.059 1.234 0.064
c=cb 1.234 0.091 1.243 0.066
Average value 1.24 0.12 1.216 0.070
71
4.2.2 Shear Capacity of Columns
Shear strength of a column can be computed in the same way as done in the case of
beams. However, for the shear capacity of a column, the values of axial load, Nd,
which appears in Eq. (4.12) is considerably larger than that used for the shear
capacity of a beam.
In order to acquire the values of mean to nominal ratio and total variability for the
shear capacity of a column, a code was developed in MathCAD complying with
FOSM format. Using this code and the statistical parameters compiled in Chapter 3,
the mean to nominal ratios and total variabilities are computed for various cases
with different reinforcement areas, dimensions and axial loads. The mean to
nominal ratios and total uncertainties are almost the same with those presented for
beams in Table 4.4. Therefore they are not given here once again.
The mean to nominal ratio of the shear capacity of a column is found to be 1.24 by
using both nominal and in-situ values as stated in the case of the shear capacity of a
beam (Section 4.2.2). For the beams, this value was indicated as 1.09 by
Ellingwood et al. (1980) and as 1.23 by Nowak and Szerszen (2003a). In the
following chapters, a value of 1.24 will be used in the reliability calculations.
It is observed that the total variability in the shear capacity of a column varies from
0.116 to 0.157 with a weighted average of 0.14 (Table 4.4). Although, the values of
axial load, Nd, acting on a reinforced concrete column are considerably more than
those acting on a reinforced concrete beams, the value of total variability does not
change significantly. On the other hand, the actual behavior of shear capacity of a
column is different than the estimated behavior predicted by structural analysis. In
order to account for this effect, Ellingwood et al. (1980) and Nowak and Szerszen
(2003a) used a value of 0.115 and 0.10, respectively in their studies. Here, the
analysis uncertainty will be taken as 0.10. As a result, the overall uncertainty in the
72
shear capacity of a column will be Ω Vr = (0.14 2 + 0.10 2 ) = 0.17 .
4.3 CAPACITY OF REINFORCED CONCRETE SHEAR WALLS IN

DIFFERENT FAILURE MODES
Large lateral loads acting on a reinforced concrete structure, such as seismic loads
or very high wind loads are mostly resisted by shear walls, which may be
constructed between column lines or may be combined into stair wells and elevator
shafts. Shear walls are vertical structural members whose cross-sectional dimension
ratio is greater than 7. The additional function of shear walls is to decrease the
possibility of damage to the non-structural members that most building contain
(Nawy, 2005). As stated by Paulay and Priestley (1992), “individual shear walls
may be subjected to axial, translational and torsional forces; to the extent to which a
wall will contribute to the story torsion, story shear forces and resistance of
overturning moments depends on its geometric appearance, orientation and location
within the plane of building”.
When shear walls are placed in advantageous positions in a building, they can be
very efficient in resisting lateral loads (Park and Paulay, 1975). In a structural
system, shear walls can be used together with a frame system. There are also
structural systems in which vertical load carrying elements are only shear walls.
Shear walls are considerably stiffer than regular frame elements, and thus can
absorb or respond to greater lateral forces, while controlling interstory drift.
Therefore, structures constructed with shear walls are significantly effective
compared to rigid frames from structural integrity, damage control and structural
safety point of view. (Nawy, 2005)
The use of shear walls becomes imperative in certain types of high-rise buildings if
the interstory deflections caused by the lateral loading have to be controlled. Shear
walls can both provide adequate structural safety and give a great prevention against
73
costly non-structural damage during a moderate seismic excitation (Park and
Paulay, 1975). Shear walls act as cantilever beams fixed at the base of the structure
in order to transfer lateral loads, such as earthquake load and wind load, to the
foundation. In a shear wall, gravity loads produce uniform compression over the
cross-section while flexure creates compression on one face of the wall and tension
on its opposite face. Flexure and shear have maximum values at the base of the
shear wall which vary over its height as shown in the Figure 4.4. The statistics of
basic variables concerning the flexural and shear capacity of a reinforced concrete
shear wall are summarized in Table 4.10.
a) b) c) d) e)
Figure 4.4 a) Lateral loads b) Isolated wall c) Shear diagram d) Moment diagram
e) Axial loads
74
and shear capacities of shear walls
Nominal
(Specified)
nominal ratio variability error uncertainty
value
21.55 MPa 1.25 0.105 0.14 0.18
of concrete (fc)
Yield strength
365 MPa 1.24 0.038 0.08 0.09
of BC III (f y)
Shear wall width (bw) 250-350 mm 1.05 0.047 0.03 0.056
1200-4400 0.998 0.014 0.03 0.03

Shear wall depth ( l w ) mm
Reinforcement area 1500-7500
1.00 0.012 0.03 0.03
(As) mm2
4.3.1 Flexural Capacity of Shear Walls
The diagram shown in Figure 4.4.c. indicates that shear walls, which can be
subjected to very high moments, must be designed to resist the moment at their
base. Additionally, the axial compressive force caused by gravity loads act on the
base section. In order to take into account the effects of both forces in combination,
an interaction diagram can be developed. Because of its large cross-sectional area,
the axial compressive load on a shear wall is often considerably smaller than the
one which would cause a balanced failure condition, in other words, in low-to-
moderate rise structures, the axial load is generally quite low; consequently, the
moment capacity is usually increased by the gravity forces in shear walls (Park and
Paulay, 1975). Therefore, it is conservative to consider only the moment at the base
in comparison with the axial compressive force. The moment capacity of a shear
wall can be computed from the following equations (Ferguson et al., 1988):
  z 
M r = 0.9 0.5A st f yd l w 1 −  (4.19)
  lw 
75
z
= 1 / (2 + 0.85k1l w hf cd / (A st f yd )) (4.20)
lw
where:
l w : the horizontal length of the shear wall
h: the thickness of the wall
Ast: the horizontal reinforcement area
In order to calculate the mean to nominal ratio and total variability of the flexural
capacity of a shear wall, different cases with varying characteristics of shear walls
are taken into account through another code written in MathCAD 12 according to
the FOSM format. The results of mean to nominal ratios and total variabilities for
the shear wall flexural capacity are given in Table 4.11.
The mean to nominal ratio of the flexural capacity of a shear wall is found to be
1.24 by using both nominal and in-situ values as explained before; and also, the
total variability in the flexural capacity of a shear wall is found to vary from 0.092
to 0.097 (Table 4.11). Correspondingly, in this study, the total variability will be
taken as 0.09. Again, the total variability in shear wall flexural design should be
combined with the analysis uncertainty. Accordingly, the analysis uncertainty
should be estimated. Ellingwood et al. (1980) used a value of 0.046 for beams and
columns in order to include this type uncertainty, whereas Nowak and Szerszen
(2003a) recommended the values of 0.06 and 0.08 for beams and columns,
respectively. On the other hand, the analysis uncertainty in the shear capacity of a
beam is given as 0.10 by Nowak and Szerszen (2003a) in their study. In the analysis
of shear walls, the assumption of “plane section remains plane after deformation” is
considered similar to the beam analysis. The studies, however, have showed that
this assumption is not correct if the height to width ratio of the cross-section is
excessively high. Therefore, in this study, the analysis uncertainty of the flexural
equation for shear walls is taken larger than the values taken for the flexural
76
equation of beams. Accordingly, in this study, the analysis uncertainty is set equal
to 0.10.
As a result, the total variability in the flexural capacity of a shear wall and the
analysis uncertainty for the equation of flexural capacity of a shear wall are
combined to find the overall uncertainty; yielding Ω M r = (0.092 + 0.10 2 ) = 0.13 .
Table 4.11 Mean to nominal ratios and coefficients of variation for the flexural
capacity of shear-walls
bw lw As Mr
ΩM
(mm) (mm) (mm2) M ′r
200 1500 3500 1.241 0.092
200 2800 2500 1.240 0.094
200 2000 4500 1.241 0.092
200 3500 7500 1.242 0.092
250 1500 3500 1.241 0.093
250 1500 4500 1.242 0.092
250 1500 7500 1.242 0.092
250 2500 7500 1.242 0.092
250 4400 7500 1.241 0.093
300 1500 1000 1.240 0.096
300 1500 1500 1.241 0.095
300 1500 2500 1.241 0.094
300 1500 3500 1.241 0.093
300 2800 1500 1.240 0.097
300 1500 2500 1.241 0.094
300 1500 3500 1.241 0.093
300 2800 1500 1.240 0.097
77
4.3.2 Shear Capacity of Shear Walls
Shear walls behave like fixed supported beams, and the design of shear walls can be
carried out in the same way as beams (Park and Paulay, 1975; Celep and Kumbasar,
2005).
Shear capacity of wall cross sections, Vr, can be calculated by using the following
equation, as specified in TEC-2006 (Specification for Structures to be Built in
Earthquake Areas, 2006):
Vr = A ch (0.65f ctd + ρsh f yd ) (4.21)
where:
Ach: gross section area of the shear wall
ρsh: ratio of horizontal web reinforcement of shear wall to the gross area of shear
wall web [(ρsh)min=0.0025]
On the other hand, the shear force, Vd, should satisfy the conditions defined below:
Vd ≤ Vr (4.22)
Vd ≤ 0.22A ch f cd (4.23)
In order to calculate the mean to nominal ratio and the total variability of the shear
capacity of a shear wall, various dimensions with different reinforcement areas are
analyzed using a code written in MathCAD 12 according to the FOSM formulation.
The results are given in Table 4.12.
As observed in Table 4.12, the analyzed cases indicate a value of 1.24 for the mean
to nominal ratio of the shear capacity of shear walls by using both nominal and in-
situ values as stated before, and also the total variability in the shear capacity of
78
shear walls is found to vary from 0.085 to 0.103, with a weighted average value of
0.10.
Nowak and Szerszen (2003a) recommended a value of 0.10 for the analysis
uncertainty for the shear capacity of a beam in their study, based on the research
conducted by Ellingwood et al. (1980) and engineering judgment. Here, the analysis
uncertainty will also be taken as 0.10; in other words, in order to account for the
analysis uncertainty of the shear capacity of a shear wall, the same value that is used
for the shear capacity of a beam will be used. Consequently, the overall uncertainty
in the shear capacity of a shear wall will be Ω Vr = (0.10 2 + 0.10 2 ) = 0.14 .
Table 4.12 Mean to nominal ratios and coefficient of variations of resistances for
shear-walls in the shear failure mode
bw lw Mr
ρs ΩM
(mm) (mm) M ′r
200 1500 0.0025 1.244 0.102
200 2000 0.0040 1.243 0.094
200 3000 0.0055 1.243 0.091
200 4000 0.0040 1.243 0.094
200 5000 0.0055 1.243 0.091
250 1500 0.0030 1.243 0.098
250 2000 0.0040 1.243 0.094
250 3000 0.0048 1.243 0.092
250 4000 0.0032 1.244 0.097
250 5000 0.0025 1.244 0.102
300 2000 0.0026 1.244 0.101
300 3000 0.0033 1.244 0.096
300 4000 0.0040 1.243 0.094
300 5000 0.0050 1.243 0.091
300 6000 0.0025 1.244 0.102
79
CHAPTER 5
MODELING OF LOADS
5.1 INTRODUCTION
The information on the design loads is one of the main components in structural
engineering for the analysis of structures. The loads acting on a structure can take
on a wide variety of forms. They are usually considered as primary or secondary
loads. Secondary loads are due to temperature changes, construction eccentricities,
shrinkage of structural materials, settlement of foundations, or other such loads.
Primary loads include the own weight of the materials from which the structure was
built, the occupants and the furniture, various weather conditions, loading
conditions that the structure is experienced during construction and environmental
loads. In this study, only five major loads, namely dead, live, wind, snow and
earthquake will be considered as the primary loads.
Dead load can be defined as the weight of the permanent elements of a structure.
These elements are always present; therefore, the dead load will nearly remain
constant unless any major changes are carried out in the building.
Imposed or live loads include all the movable objects in a building, such as people,
furniture, machines and fixtures, partitions and other non-structural elements.
The degree of snow load effect depends on the climate of a region. Some countries
80
and regions have snowfall for six months of the year or more while some have little
or no snowfall during a year. Structures have to be designed to withstand the
appropriate snow load consistent with the weather conditions of the regions in
which they are located.
Wind load is the lateral pressure on the structure due to wind blowing in any
direction. Wind loads play a much more important role in modern construction than
they did in the past. In modern construction, where a steel framework is used, wind
loads primarily affect the strength and stability of the building.
The importance of earthquake loading is related to its likelihood of occurrence and

possible magnitude in any given region. Structural engineers mainly deal with the
local effects of large earthquakes in regions where the ground motions appear to be
intense enough to lead to the structural damage. Earthquake load is assumed to be
lateral load acting in any horizontal direction on the structural frame due to the
dynamic action of earthquakes. The structure can be designed considering the
maximum shear force at the base of structure. In the earthquake resistant design, the
reliability-based design criteria is required to include the characteristics of
earthquake load, structural capacity and structural response, which are all random in
nature and involve uncertainties.
Any type of structural load becomes meaningful if it leads to a load effect.

Assuming a linear relationship between a load and its effect, the following formula
gives the load effect, Si, on the basis of its corresponding structural load, Li,
(Ellingwood et al.1980):
Si = C i N i L i (5.1)
where:
Ci: influence coefficient
81
Ni: modeling parameter
Li: structural load
It should be noted that in this formula, the parameters on the right hand side of the
equations, i.e. Ci, Ni, and Li are assumed to be statistically independent. According
to the FOSM method, the mean and coefficient of variation for any load type can be
computed from the following formulas, respectively:
Si = Ci N i L i (5.2)
Ω Si = ∆2Ci + ∆2Ni + ∆2Li (5.3)
In the following sections, structural load effects resulting from dead, live, wind,
snow and earthquake loads will be evaluated. The published data in literature and
the local data compiled in Turkey constitute the main sources of information in the
evaluation of the load statistics. For calibration purposes, the ratio of mean to
nominal load values should be determined. The nominal values of loads will be
obtained from TS 498 (Design Loads for Buildings). It is worth mentioning here
that although the dead and live loads acting on a structure are independent of the
geographical location of the structure, environmental loads, such as snow, wind and
earthquake loads are highly dependent on the location of the structure. Accordingly,
for the assessment of statistics of environmental loads, different cities, which will
represent the whole Turkey, are chosen. In this selection, cities with the highest
critical environmental load are given priority. Also, cities with larger populations
are preferred. Another criterion in this selection is that these cities be located in
geographically different regions of Turkey. For the snow depth and wind speed, the
data are obtained from the meteorological stations that are in the centers of these
cities. The locations of these cities are shown in the following map:
82
Figure 5.1 The map showing the locations of selected cities
5.2 DEAD LOAD
Dead loads result from the weight of the structural elements, such as beams,
columns, slabs, and non-structural permanent fixtures like roofing, surfacing and
covering. The main characteristics of dead load can be described as follows (JCSS
Probabilistic Model Code, 2001):
-Throughout the life of the structure, dead load remains constant and the
probability of occurrence of dead load at an arbitrary point-in-time is one.
- The variability of dead load with time is generally negligible.
- The uncertainties related to dead load are normally small in comparison
with other types of loads.
Taking the above statements into consideration the following factors cause
variability in dead load:
83
- The errors and tolerances in the measurement of dimensions,
- The differences in the actual and nominal unit weights of the construction
materials,
- The tributary area method which represents a rough approximation of the
real situation.
In addition to the factors listed above, the variability in dead load also results from
the variability in the weight of the non-structural items, such as roofing, partitions.
The most important variability in dead load is due to weight of these non-structural
components. Therefore, Ellingwood et al. (1980) expressed that the material that
constitute the structural system has a very weak influence on variability of dead
load.
Concerning the uncertainties in dead load, the following sources of variability can
be taken into consideration: variability within a structural member, variability
between different structural members of the same structure and variability among
various structures (JCSS Probabilistic Model Code, 2001).
Most of the investigators indicated that the probability distribution of dead load
should be normal and assumed that the ratio of mean to nominal dead load, D D′ , is
unity. Some of the values on the mean to nominal ratio and the coefficient of
variation as reported by different reliability based design studies are listed in Table
5.1. It is observed in Table 5.1 that the variability in dead load ranges from 5% to
10%. In his study, Kömürcü (1995) assumed that normal distribution is suitable for
dead load. He took the mean to nominal ratio of dead load, and total variability as
1.05 and 0.10, respectively. It is assumed that D D ′ =1.05 and ΩD=0.10 as the
approximate average value of these studies, and normal distribution is taken for the
dead load. Here, D and D ′ denote the mean and nominal value of dead load,
respectively, and ΩD stands for the total uncertainty in dead load.
84
Table 5.1 Statistics of dead load (from Ellingwood et al., 1980)
References D D′ ΩD
Galambos and Ravindra, 1973 1.0 0.08

Allen, 1976 1.0 0.10
Elingwood, 1978 1.0 0.10
Lind, 1976 1.05 0.09
Lind et al., 1978 1.0 0.05
Ellingwood et al. 1980 1.03 0.10
5.3 LIVE LOAD
Most of the live load research works were conducted about two decades ago and the
development in the understanding of live loads has been hampered by the
insufficiency of live loads survey data (Kumar, 2002a). Live loads on floors of
buildings include the weight of furniture, equipment, stored objects, movable
partitions, and persons and their possessions. At the same time, rare events expected
in the lifetime of structure, such as gathering of persons or moving of objects which
may occur during reorganization or redecoration, cause to this type of load. Any
structural or non-structural elements like covering, extraordinary equipment or
partition walls are not included in live loads. The live load can be separated
according to categories of usage, such as domestic and residential activities, office
areas, shopping areas, hotels, hospitals, restaurants, dance halls where people may
concentrate in number.
A wide variety of loading possibilities, which the structural engineers may not
predict, can occur within the lifetime of a structure. In addition, live loads vary in
time and space in a random manner. On the other hand, owing to the correlation
between room use and the room area, unit floor loads are strongly correlated with
the room area. “The total live load on a floor area may be thought of conveniently
85
as consisting of a sustained component which remains relatively constant within a
particular occupancy, referred to as the arbitrary point-in-time live load, and an
extraordinary component which arises from infrequent clustering of people above
and beyond normal personal load, or from activities such as remodeling”
(Elingwood et al., 1980).
It is rational that unit floor loads are insignificantly affected by geographical

locations, height of storey, sector, age and traditional habits in countries. Although
there are differences among the results which are based on UK, USA, Australian
and Indian surveys, the overall characteristic and variation of live loads are very
similar. (Kumar, 2002b)
5.3.1 Arbitrary Point-in-Time Live Load (Lapt)
A number of load surveys which were analyzed by using probabilistic load models
have been conducted in different countries. Although most of the studies have
focused on office buildings, some data on residence, retail establishments and other
occupancies were also taken into consideration. A summary of results obtained
from the analyses of load survey data is shown in Table 5.2.
In addition to studies given Table 5.2, the arbitrary point-in-time live load
intensities for office buildings obtained from different studies are compared in
Table 5.3. The differences among the live load survey results of various studies are
due to a number of typical reasons, e.g. sample size of surveys, time interval
between different surveys, methodology of data collection, cultural backgrounds of
social groups using relevant buildings, such as occupancy duration and properties of
devices used in the buildings.
86
Table 5.2 Statistics of arbitrary point-in-time live loads for offices reported in
different studies (from Ellingwood et al., 1980)
Area
18.58 m2 92.9 m2 464.5 m2 929 m2

Reference
L apt L ′ ΩLapt L apt L ′ ΩLapt L apt L ′ ΩLapt L apt L ′ ΩLapt

McGuire and
0.24 0.89 Varies 0.52 Varies 0.41 Varies 0.40
Cornell (1973)
Ellingwood and
Culver (1977)
Chalk and
Corotis (1979)
Allen (1976) 0.16 0.70 Varies 0.48 Varies 0.38 Varies 0.36
Sentler (1975) 0.15 0.59 Varies 0.26 Varies 0.20 Varies 0.18
Mean of ΩLapt 0.80 0.50 0.45 0.40
According to Table 5.3, the average arbitrary point-in-time loads for office
buildings are 0.609 kN/m2, 0.58 kN/m2, 0.51 kN/m2, 0.488 kN/m2 for surveys
conducted by Mitchell and Woodgate (1971), Culver (1975), Choi (1992) and
Kumar (2002a), respectively. In this study, the arbitrary point-in-time live load will
be taken as 0.55 kN/m2 for office buildings, which is the average of the values
reported in the studies mentioned above. On the other hand, the average basic
variability of arbitrary point-in-time live load is taken as 0.78 based on Table 5.3.
The c.o.v. in live load effect should include the uncertainties in the analysis which
converts the equivalent uniformly distributed loads to a load effect and errors in the
load modeling. The uncertainty due to the transformation of the equivalent
uniformly distributed load to a load effect is assumed to be ∆C=0.05, and the
modeling uncertainty is taken as ∆N=0.10 (Ellingwood, et al., 1980). The total
prediction error due to these two sources of uncertainty will be
∆ L = 0.102 + 0.052 =0.11. As a result, the total uncertainty in arbitrary point-in-
87
time load becomes Ω L = 0.112 + 0.782 = 0.79 . As it is seen in Table 5.2, the mean
of total variability in arbitrary point-in-time live load, ΩLapt ranges from 0.8 to 0.4
depending on the area with an average value of 0.60. In this study, taking into
consideration the values associated with the arbitrary point-in-time load, the total
variability, ΩLapt, will be taken as 0.70 as an average value of total variabilities
obtained from Tables 5.2 and 5.3.
In the Turkish Code of design loads for buildings, i.e. TS 498 (1997), the nominal
live load, L ′ , is given as 2 kN/m2 for housing or terrace room and corridors, stores
up to 50 m2, hospital rooms and offices. Thus, the ratio of arbitrary point-in-time
L apt 0.55
load to nominal live load is = = 0.275
L′ 2
The normal, log-normal and gamma distributions were tested by different

researchers to describe the distribution of live load. The normal distribution did not
show a good fit. Among the three distributions studied, the use of gamma
distribution seems to provide a reasonable probabilistic model for live load (Corotis
and Doshi, 1977). Accordingly, the gamma distribution is taken as the probability
distribution for the arbitrary point-in-time live load in this study.
88
Table 5.3 Results of arbitrary point-in-time live load surveys for office buildings
which was reported in different studies (from Kumar, 2002a)
Survey Bounds of room Mean live Standard Coefficient of

floor area, A (m2) load (kN/m2) deviation variation
2.4 0.66 0.65 0.98

5.2 0.64 0.53 0.83
14.0 0.62 0.43 0.69
Mitchell and Woodgate 31.2 0.61 0.34 0.56
(1971) 58 0.59 0.30 0.51
111.3 0.58 0.26 0.45
192.4 0.56 0.21 0.38
A≤4.7 0.83 0.82 0.99

Culver (1975) 4.7<A≤9.3 0.63 0.60 0.95
9.3<A≤27.9 0.44 0.31 0.70
A>27.9 0.42 0.43 1.02
A≤5 0.50 0.66 1.32

5<A≤10 0.62 0.64 1.03
Choi (1992) 10<A≤20 0.55 0.47 0.85
20<A≤40 0.45 0.53 1.18
40<A≤80 0.43 0.45 1.05
A>80 0.51 0.41 0.80
A≤8 0.68 0.41 0.60

8<A≤16 0.60 0.32 0.54
16<A≤24 0.50 0.36 0.72
Kumar (2000) 24<A≤32 0.50 0.29 0.58
32<A≤40 0.47 0.26 0.55
40<A≤48 0.45 0.24 0.53
48<A≤56 0.45 0.25 0.56
64<A≤72 0.46 0.15 0.33
72<A≤80 0.46 0.19 0.41
A>80 0.31 0.20 0.65
5.3.2 Maximum Live Load (L)
Maximum live load which is used for design purposes is composed of lifetime
maximum sustained load and extraordinary load. The lifetime maximum sustained
load is the maximum of the various sustained loads which are normally present as a
89
consequence of everyday activity, and this type of load is dependent on the rate of
change of occupancy. Crowding of people or stacking of furniture causes additional
loads which act at a moment during the lifetime of building except for sustained
load. This type of loading is defined as extraordinary load.
Loads acting on a structure at any point in-time can be defined by load surveys, but
the determination of the maximum load which may be expected to act on the
structure at any time during its lifetime is impossible with these load surveys
performed during a limited duration of time. In addition to survey data of arbitrary
point-in-time loads, the frequency of occupant changes and extraordinary load
events must be predicted. Ellingwood et al. (1980), Kömürcü and Yücemen (1996)
and Kumar (2002a) used Type I extreme value distribution to describe the
distribution of maximum live load. Type I extreme value distribution is also
preferred in this study. Some of the results obtained from the studies related to
maximum live load for office buildings are shown in Table 5.4. In spite of the fact
that the results in Table 5.4 are derived for offices, examination of data for a
number of other occupancies covering multistory residences and retail
establishments shows similar properties (Ellingwood et al., 1980).
Table 5.4 Statistics of maximum live loads as reported in different studies

(from Ellingwood et al., 1980)
Area
Reference 18.58 m2 92.9 m2 464.5 m2 929 m2

L L′ δL L L′ δL L L′ δL L L′ δL
McGuire and
Cornell (1973)
Ellingwood and
Culver (1977)
Chalk and
Corotis (1979)
Sentler (1975) - 0.26 Varies 0.18 Varies 0.14 Varies 0.12
90
In the studies performed by Ellingwood et al. (1980) and Kömürcü (1995), the
uncertainty due to the transformation of the equivalent uniformly distributed load to
a load effect is assumed to be ∆C=0.05, which is the same value for Lapt, and
uncertainty in the load modeling is taken as ∆N=0.20. Here, the value of ∆C and ∆N
are also taken as 0.20 and 0.05, respectively. The basic variability, δL, is taken to be
0.17 as the average of the values reported in Table 5.4. Accordingly, in the case the
prediction uncertainties and inherent variability are combined, the total uncertainty
in maximum live load, L, will be equal to Ω L = 0.17 2 + 0.20 2 + 0.05 2 = 0.27 .
Kömürcü (1985) took mean to nominal ratio of maximum live load as 1.00. Also, in
another study, Nowak and Szerszen (2003b) proposed this ratio to be 1.00. In this
study, the mean to nominal maximum live load ratio is also taken as 1.00.
5.4 SNOW LOAD
Snow loading is one of the main reasons of structural performance failures which
can cause significant economic losses in certain regions; therefore, structural
engineers should anticipate snow action on structures (O’Rourke and Wreen, 2004).
In some cases, the drifted snow on roofs may cause severe damage and even
collapse of structures in Turkey if the design snow load is less than its real value
(Durmaz and Daloglu, 2005). Buildings, particularly roofs can be affected by snow
in numerous ways; the important ones can be listed as follows:
- Snow load can cause the collapse of roofs due to heavy accumulation.
- The drifting around buildings can obstruct people and vehicles to access
inside.
- Snow slides from sloped roofs and skylights can endanger pedestrians.
- Ice and ice dams cause water leakage under roofs.
- Snow melting causes gathering of water at low or irregular areas on a roof.
91
Although all snow problems should be paid attention, in this study, the snow load
on roofs, which is probably the most important problem in terms of human safety
and economic properties will be taken into account. Design snow loads for roof
structures are based on a number of parameters. These parameters include
geographical location, roof type and slope, building configuration, air temperature,
amount of sunshine, thermal building conditions and roof exposure to wind. In
addition to these parameters, amount and type of snowfall, sliding snow, snow
drifts, unbalanced snow loads, building characteristics and surrounding environment
must be considered (Jeff Quell, 1998; Schiever, 1978). To be aware of how they
affect snow loads is important both in evaluating the validity of calculated snow
loads for a special roof at a specific site and in determining the characteristic ground
snow load.
Among the parameters that influence design snow loads on roofs, the wind effect is
probably the most important one because the wind exposure usually causes the
snow drift. Winds incline to decrease snow loads on the windward side of a roof
and raise snow loads on the leeward side. In addition, terrain features, rooftop
projections, parapet walls and adjacent buildings can all cause snow drifts on roofs.
Extra snow can also be accumulated on roofs equipped with satellite dishes and
solar collectors. On the other hand, impact snow loads resulting from snow sliding
off or falling from a higher roof cause additional snow load. The roof surface
temperatures, the slipperiness of the roof surface and rooftop obstructions influence
the amount of sliding. Since most structural failures associated with snow loads
came into existence due to overloads, such as snow drifts and sliding snow, for the
convenient design of the roof, and in some cases, the removal of snow drifts is key
to the integrity of the roof structure( (Jeff Quell, 1998).
If a building is adjacent to a small building, it is important to understand that the

small building may be exposed to more severe snow action. On the other hand, the
design snow load on an existing roof structure may be changed by building
additions and alterations, or by the construction of a new building which is adjacent
92
to the old structure.
In the structural engineering problems, snow loads are generally an important

design factor if:
-The building is constructed in regions that are cold and snowy.

-The building is adjacent to a higher existing building.
-The building has a relatively large roof.
-The building has an unusually shaped roof.
The principles and methods used to calculate the design snow loads are determined
in the model building codes, or can be clearly described by the local building codes.
The application of these building codes by a design engineer determines the design
snow load for each building separately (Jeff Quell, 1998).
5.4.1 Roof Snow Load
According to standards, roof snow load required for structural analysis is defined as
the product of a design ground snow load and different conversion factors that
transform the ground snow load to roof snow load. This transformation procedure
can be carried out by using the following formula (ASCE/SEI 7-05, 2006):
S = α.C e .C t .I.Pk0 (5.4)
where:
S: roof snow load
α: basic reduction factor
Ce: exposure factor
Ct: thermal factor
I: risk factor
93
Pko: ground snow load.
The basic reduction factor, α, takes into account the redistribution of the snow on
the roof caused by wind, in other words, α results from wind effect on snow
accumulation. It was proposed to be 0.47, 0.6 and 0.7 by Q’ Rourke et al. (1982),
Kömürcü (1995) and ASCE/SEI 7-05 (2006), respectively. In this study α will be
taken as 0.6, which is approximately the average of these three values.
The thermal coefficient, Ct, accounts for the roof’s thermal special features. If a
structure is heated, thermal energy moving smoothly through the roof should cause
a change in roof snow load. The structures can be mainly divided into heated
structures, heated structures with ventilated roofs in freezing and cold climates and
unheated structures. According to ASCE/SEI 7-05 (2006), the corresponding
thermal factors range from 0.85 to 1.2 for heated structures. Since the heated
structures are considered as a basis for design in this study, thermal factor will be
taken as 1.0.
The exposure factor, Ce, depends on the location of both structure and roof, as well
as theirs surrounding (O’Rourke et al., 1982). Ce in ASCE/SEI 7-05 (2006) ranges
from 0.7 to1.2 for the separate classifications with an approximate average value of
1.0, which will be also valid in this study.
The aim of the importance factor, I, used in snow load analysis is to increase design
loads for cases where the consequences of failure are greater than normal, and is to
decrease design loads for structures where consequences of failure are less than
normal. It ranges from 0.8 to 1.2 ASCE/SEI 7-05 (2006). In this study, importance
factor value will be taken as 1.0.
As a term, all of these conversion factors can be defined as snow load coefficient,
Cs, and, Eq. (5.4) can be rewritten as “ S = C s .Pk0 ”. This coefficient is consequently
94
calculated as 0.6. The total variability in snow load coefficient, ΩCs, which results
from the above conversion factors is assumed to be 0.23 (Ellingwood et al., 1980;
Kömürcü and Yücemen, 1996).
5.4.2 Ground Snow Load, Pko
In Turkey, snowfall generally increases from west to east and from south to north; it
also increases with altitude of the site from the sea level (Ayaroglu, 1991). The
ground snow load is determined from the basic meteorological data, and naturally
shows diversity according to the geographical location of the building sites.
Therefore, site specific case studies determine design ground snow load.
Accordingly, the annual maximum snow depth is obtained for Ankara, Izmir, Bursa,
Antalya, Gaziantep, Samsun, Malatya, Erzincan, Canakkale and Hakkari.
Considering the earthquake effect, Istanbul region has two different seismic zones,
seismic zone I and seismic zone II; some of the buildings in Istanbul are located in
seismic zone 1 whereas other buildings are located in seismic zone II. Therefore,
two centers for Istanbul are classified as Göztepe located in seismic zone I and Sile
located in seismic zone II. Another reason to consider this classification is that
Istanbul has the largest population among the cities in Turkey; naturally the number
of buildings is quite more than an ordinary city. Also snowfall is expected for only
three or four months in most of the cities in Turkey; we do not mention any snow
loads in the other months for these locations. Furthermore, snowfall may not be
seen in some years for specific locations. Therefore, years in which there is no
snowfall are not taken into account and for locations where there is regular
snowfall; 0.01 m snow depth values are ignored. On the other hand, 1.00 m snow
depth value in the Gaziantep region, which deviates from the mean by more than
3σ, is treated as an outlier, and is not taken into consideration. The modified data is
given in Appendix C.
The average values of the maximum snow depths for Ankara, Izmir, Bursa,
95
Antalya, Gaziantep, Samsun, Malatya, Erzincan, Canakkale, Hakkari,
Göztepe/Istanbul and Sile/Istanbul are found to be 0.127 m, 0.02 m, 0.188 m, 0.015
m, 0.16 m, 0.142 m, 0.250 m, 0.207m, 0.107 m, 1.057 m, 0.146 m, 0.147 m,
respectively. The matching variabilities in annual snow depth values for these cities
are computed as 0.54, 0.63, 0.71, 0.47, 0.66, 0.88, 0.51, 0.52, 0.85, 0.50, 0.66, and
0.76, respectively.
The ground snow load can be determined by using the following equation:
Pk0 = h.ρ (5.5)
where:
h: depth of snow on the ground (m)
ρ: average density of snow (kN/m3)
The value of snow density is determined by using its water equivalent (O’Rourke et
al., 1982, Rusten et al., 1980). Snow density is space and time dependent; while
snow depth grows, snow density, which also rises with time, will naturally increase.
TS 7046 (Bases for Design of Structures-Determination of Snow Loads of Roofs)
and Ghiocel and Lingu (1975) proposed the following empirical relationships
between snow depth, h (m) and snow density, ρ (kg/m3):
ρ = 300 − 200.e −1.5h (5.6)
Considering the mean snow depths, mean snow densities can be computed for
Ankara, Izmir, Bursa, Antalya, Gaziantep, Samsun, Malatya, Erzincan, Canakkale,
Hakkari, Göztepe/Istanbul and Sile/Istanbul by using Eq. (5.6). By putting in the
mean values of annual maximum snow depth, h an , and the mean values of snow
density, ρ , in Eq. (5.5), the mean value of the annual extreme ground snow load,
Pko, for different locations are computed. The total variability in ground snow loads
96
is calculated by using the variability in maximum snow depths and the variability in
snow density according to Eq. (5.7). The results are summarized in Table 5.5.
2 2 (5.7)
Ω Pko = Ω ρ + Ω h an
Table 5.5 Statistical parameters of the annual extreme ground snow load
Parameter
ρ Ωn han Ω han Pko ΩPko
Location (kN/m ) 3
(kN/m )2
(m)
Ankara 1.35 0.12 0.127 0.54 0.171 0.55
Izmir 1.06 0.03 0.02 0.63 0.021 0.63
Bursa 1.49 0.18 0.188 0.71 0.280 0.73
Antalya 1.04 0.02 0.015 0.47 0.016 0.47
Gaziantep 1.43 0.16 0.16 0.66 0.229 0.68
Samsun 1.38 0.20 0.142 0.88 0.196 0.90
Malatya 1.62 0.15 0.250 0.51 0.405 0.53
Erzincan 1.53 0.15 0.207 0.52 0.317 0.54
Canakkale 1.30 0.16 0.107 0.85 0.139 0.86
Hakkari 2.59 0.11 1.057 0.50 2.738 0.51
Göztepe/Istanbul 1.39 0.16 0.146 0.66 0.203 0.68
Sile/ Istanbul 1.40 0.18 0.147 0.76 0.206 0.78
5.4.3 Annual Extreme Roof Snow Load
The mean annual roof snow load, San , can be computed for Ankara, Izmir,
Gaziantep, Samsun, Malatya, Erzincan, Canakkale, Hakkari, Göztepe/Istanbul and
97
Sile/Istanbul by using the appropriate values of the parameters Cs and Pko .
an
Statistical parameters of the annual extreme roof snow load are shown in Table 5.6.
It is observed that the lognormal distribution shows a reasonable fit to the data
according to Minitab and BestFit computer programs for the annual maximum snow
depth of different locations. The Chi-Square and Kolmogorov- Smirnov tests results
for Lognormal, Normal, Extreme-value (Type I), Gamma, Weibull, Rayleigh
probability distributions are given in Appendix C. The distribution of Pko can also
an
be taken as lognormal and regarding the distribution of component variables Cs

and Pko , the distribution of San could be approximated by a lognormal distribution
an
like its component Pko (Ellingwood et al., 1980). The parameters of the
an
corresponding lognormal distribution are computed from the following general

equations (Ang and Tang, 1984):
ζ = ln(1 + Ω 2 ) (5.8)
1
λ = ln µ − ζ 2 (5.9)
2
98
Table 5.6 Statistical parameters of the annual extreme roof snow load
Parameter
San Ω San ζ S an λ San
Location
(kN/m2)
Ankara 0.103 0.60 0.555 -2.427
Izmir 0.013 0.67 0.609 -4.528
Bursa 0.168 0.77 0.682 -2.017
Antalya 0.009 0.52 0.489 -4.83
Gaziantep 0.137 0.72 0.646 -2.197
Samsun 0.118 0.93 0.789 -2.449
Malatya 0.243 0.58 0.538 -1.560
Erzincan 0.190 0.59 0.547 -1.810
Canakkale 0.083 0.89 0.764 -2.781
Hakkari 1.643 0.56 0.522 0.36
Göztepe/Istanbul 0.122 0.72 0.646 -2.313
Sile/ Istanbul 0.123 0.81 0.710 -2.348
5.4.4 Maximum Roof Snow Load (S)
Maximum roof snow load, S, can be defined as the 30-year maximum value for
Canada and Russia, and the 50-year maximum value for Europe and U.S.A.
(Durmaz and Daloglu, 2005). In this study, a 50-year maximum value will be used
for maximum roof snow load.
The probability distribution of maximum roof snow load is based on the

distributions of Cs and mainly on Pko , which is lognormally distributed. Type II
an
extreme value distribution for S is used on the basis of the fact that the limiting
distribution for a series of lognormally distributed variates is the Type II extreme
value distribution of largest values (Ellingwood et al., 1980). Considering that the
initial distribution of a random variable is lognormal with parameters, ζ and λ , the
99
distribution of the largest value of a random variable converges to the Type II
asymptotic form with following parameters (Ang and Tang, 1984):
 lnln(n)+ ln4π 
 ζ S 2ln(n) −ζ S + λS 
 2 2ln(n)  (5.10)
vS = e
2ln(n)
kS = (5.11)
ζS
The mean value and variability of S are obtained substituting the above parameters
in the following equations:
 1 
S = v S Γ1 −  (5.12)
 kS 
 2 
Γ1 − 
ΩS =  kS  −1 (5.13)
 1 
Γ 2 1 − 
 kS 
where Γ(.) stands for the gamma function.
100
Table 5.7 Statistical parameters of the 50-year maximum roof snow load
Parameter
vS kS S (kN/m2) ΩS
Location
Ankara 0.283 5.044 0.329 0.31
Izmir 0.039 4.593 0.046 0.35
Bursa 0.558 4.099 0.679 0.41
Antalya 0.022 5.718 0.025 0.27
Gaziantep 0.432 4.328 0.519 0.39
Samsun 0.454 3.543 0.577 0.50
Malatya 0.652 5.194 0.754 0.30
Erzincan 0.516 5.118 0.598 0.31
Canakkale 0.309 3.662 0.389 0.48
Hakkari 4.295 5.356 4.938 0.29
Göztepe/Istanbul 0.385 4.328 0.462 0.38
Sile/ Istanbul 0.425 3.938 0.523 0.43
5.4.5 Mean to Nominal Ratios of San and S
The statistical parameters of San / S′ and S / S′ can be found after determining the
nominal snow loads. The nominal snow load, S′ , can be obtained from TS 498
(1997). Results of mean to nominal ratios of San and S are shown in Table 5.8 for
the different locations considered.
Since the distribution of San is lognormal San/ S′ will have a lognormal distribution.
However, S/ S′ has a Type II distribution because Type II distribution is used for
maximum snow load, S.
101
Table 5.8 Statistical parameters of the mean to nominal ratio for annual extreme and
50-year maximum snow loads
Parameter
S′ San Ω San ζ San λ San S ΩS vS kS
S
kN/m2 S′ S′ S′
Location
Ankara 0.95 0.11 0.60 0.56 -2.36 0.35 0.31 0.30 5.04
Izmir 0.75 0.02 0.67 0.61 -4.10 0.06 0.35 0.05 4.59
Bursa 0.75 0.22 0.77 0.68 -1.75 0.91 0.41 0.75 4.10
Antalya 0.75 0.01 0.52 0.49 -4.73 0.03 0.27 0.03 5.72
Gaziantep 1.25 0.11 0.72 0.65 -2.42 0.42 0.39 0.35 4.33
Samsun 0.75 0.16 0.93 0.79 -2.14 0.77 0.50 0.62 3.54
Malatya 1.35 0.18 0.58 0.54 -1.86 0.56 0.30 0.48 5.19
Erzincan 1.49 0.13 0.59 0.55 -2.19 0.40 0.31 0.34 5.12
Canakkale 0.75 0.11 0.89 0.76 -2.50 0.52 0.48 0.42 3.66
Hakkari 1.84 0.89 0.56 0.52 -0.25 2.68 0.29 2.32 5.36
Göztepe/Istanbul 0.75 0.16 0.72 0.65 -2.04 0.62 0.38 0.52 4.33
Sile/ Istanbul 0.75 0.16 0.81 0.71 -2.09 0.70 0.43 0.57 3.94
5.5 WIND LOAD
The assessment of wind load and reliability of structures under the influence of
wind has been receiving growing interest during the last few decades. However, a
number of significant issues as to wind load still remain unsolved (Minciarelli et al.,
2001). Structural engineers should make certain that the structures subjected to
wind load will be sufficient during its expected life with regard to both
serviceability and structural safety. Accordingly, the information on the behavior of
the structure under the wind action is required in order to realize the relation
between the wind environment and the wind action.
102
Wind loads are derived by using the statistical data based on wind speed, mass
density of air, pressure coefficient, parameters related to wind speed and exposure,
and a gust factor that incorporates the effects of short gusts and the dynamic
response of the structure. The wind load acting on a structure can be determined
from the wind speed by using the standard hydrodynamic relationship, which can be
rewritten for particular structures or surfaces of structures as follows (Melchers,
2002):
W = c.C p .E z .G.V 2 (5.14)
where;
W: wind load
c: a constant related to mass density of air
Cp: pressure coefficient
Ez: exposure coefficient
G: gust factor
V: wind speed
The pressure factor, Cp, depends on the geometry and shape of the structure. It is the
ratio of the pressure at relevant surface of the structure to the dynamic pressure of
the wind (Simiu and Scalan, 1978). The exposure coefficient, Ez, depends on the
actual topographical conditions (e.g. urban area, enclosed valleys, slopes, hills, open
country and also the presence of constructions near the structure). The gust factor is
associated with the turbulence of the wind and the dynamic interaction between the
structure and wind.
Since the velocity parameter appears with its square in Eq. (5.14), it is a particularly
important parameter in comparison with the other parameters. The overall
uncertainty in wind load is also affected by uncertainties in the estimation of the
pressure coefficient, the exposure factor and the gust factor.
103
5.5.1 Analysis of Wind Speed
Most of the statistical data available related to wind load are obtained from the
monthly maximum or annual maximum wind speeds; the pressure coefficients and
gust factors are consistent with this maximum wind speeds. The wind speed
changes with latitude, longitude and altitude from the sea level and time (Sahin,
2001). In this study, since it is impractical to perform reliability analysis separately
for each and every location where wind speed data is available, it is decided to
collect this necessary data for twelve different locations in Turkey. Hence, the same
cities considered in the calculation of the snow load, namely, Ankara, Izmir, Bursa,
Göztepe/Istanbul and Sile/Istanbul are selected in order to carry out the wind load
analysis for Turkey. The required data on maximum yearly and maximum daily
wind speeds are taken from the Turkish Meteorological Department (DMI). The
wind speed is measured at 10 meters above the ground level in these locations
(Dündar et al., 2002). These data and the statistics computed are shown in Appendix
D and Appendix E for the locations above.
Type I and Type II extreme value distributions are the most common and
appropriate probability distributions for wind speeds (Simiu and Scalan, 1978 and
Ellingwood et al., 1980). Simiu and Scanlan (1978) indicated that if the wind speed
data are collected in locations where extraordinary wind speeds are very rare, the
use of Type I distribution is more suitable as a probabilistic model. Yücemen and
Gülkan (1989) and Kömürcü (1995) used the Type I distribution for yearly
maximum wind speed, daily maximum wind speed and lifetime maximum wind
speed that were measured in Turkey. Because unusual winds are not seen frequently
in Turkey, Type I distribution can be used to describe the wind speed data. Also the
data analysis with BestFit and Minitab computer programs showed that Type I
distribution fits data satisfactorily. For Lognormal, Normal, Extreme-value (Type
I), Gamma, Weibull, Rayleigh probability distributions, the Chi-Square and
Kolmogorov- Smirnov tests results, which were performed by using computer
104
programs mentioned above, are given in Appendix D and Appendix E.
In this study, daily maximum wind speed, Vapt, and annual maximum wind speed,
Van, will be used, which are respectively necessary for the calculation of the daily
maximum wind load and the yearly maximum wind load. In addition to these two
parameters, maximum wind speed, V50, and nominal wind speed, V ′ , are needed.
V50 is derived from the annual maximum wind speed, Van, and V ′ is obtained from
TS 498 (1987). Since the data for the annual maximum wind speed fits the Type I
distribution, maximum wind speed could be described by a Type I distribution, and
the mean value and the basic variability of maximum wind speed are computed
from the mean value and variability of Van by making use of the following
equations. Also, the prediction error due to insufficient sampling, ∆1, is defined by
using these computed values according to Eq. (5.17) (Ellingwood et al., 1980).
 6 
V50 = Van 1 + δ Van ln50  (5.15)
 π 
δ V50 = Van δ Van / V50 (5.16)
3.8Van δ V an
∆1 = (5.17)
( n V50 )
where n is the sample size. As n increases, ∆1 decreases.
Dündar et al. (2002) indicated that due to systematic errors and wrong calibration of
devices associated with measurement of wind speed, an uncertainty of 0.05 should
be included. In addition to the uncertainties involved in the recorded data,
conversions including modeling, features of climatic parameters and roughness
parameters of surface also create additional uncertainties. Therefore, an additional
uncertainty of 0.02 is assumed to account for these factors (Ellingwood et. al., 1980;
Kömürcü, 1995).
105
The mean values and variabilities of Vapt, Van, and V50 for Ankara, Izmir, Bursa,
Göztepe/Istanbul and Sile/Istanbul are presented in Table 5.9. Considering the TS
498 (1997), the nominal wind speeds, V ′ , for these locations are also shown in
Table 5.9.
Table 5.9 Mean values and coefficients of variation of wind speeds for different
locations
Parameter
Vapt Ω Vapt Van Ω Van V50 δ V50 ∆1 Ω V50 V′
Location (m/s) (m/s)
(m/s) (m/s)
Ankara 8.50 0.33 20.13 0.24 34.25 0.135 0.045 0.15 36
Izmir 12.45 0.29 26.19 0.14 36.58 0.093 0.036 0.11 36
Bursa 6.53 0.44 23.92 0.21 38.51 0.124 0.043 0.14 36
Antalya 10.23 0.47 26.62 0.21 42.83 0.124 0.042 0.14 36
Gaziantep 5.49 0.44 17.53 0.25 30.63 0.139 0.043 0.16 36
Samsun 8.06 0.53 23.11 0.24 39.23 0.135 0.040 0.15 36
Malatya 6.52 0.54 15.09 0.46 36.26 0.191 0.041 0.20 36
Erzincan 6.65 0.44 19.07 0.29 35.94 0.154 0.046 0.17 36
Canakkale 11.14 0.41 28.99 0.18 44.02 0.112 0.038 0.13 36
Hakkari 7.07 0.54 20.19 0.22 33.12 0.128 0.045 0.15 36
Göztepe/Istanbul 6.95 0.53 19.67 0.30 37.67 0.157 0.046 0.17 36
Sile/ Istanbul 8.72 0.46 25.21 0.29 46.74 0.151 0.043 0.17 36
5.5.2 Maximum, Yearly Maximum and Daily Maximum Wind Loads
As it is stated before, if the wind speed data are collected in locations where
extraordinary wind speeds are very rare, the use of Type I distribution is more
suitable as the probabilistic model; besides the data analysis with BestFit and
Minitab computer programs showed that Type I distribution fits the data
106
satisfactorily (Appendix D, Table D.2 and Appendix E, Table E.13). Accordingly,
Type I distribution is adopted as the probabilistic model for the wind speed data. On
the other hand, the wind load may not have the same distribution of the wind speed.
Ellingwood et al. (1980) used Monte Carlo techniques assuming that Cp, Ez and G
were described by normal distributions. As a result, it was found out that wind load,
W, could be described by a Type I distribution over the range of the distribution
above its 90th percentile. The parameters of the Type I largest extreme value
distribution can be calculated through the following relationships (Melchers, 2002):
π2
α= (5.18)
6σ
γ
u=µ− (5.19)
α
where, γ =0.577 is the Euler’s constant.
Kömürcü (1995) proposed the mean values of the parameters Cp, Ez and G to be
0.80, 1.61 and 0.45, respectively. Ellingwood et al. (1980) quantified the
variabilities of these parameters as 0.12, 0.11 and 0.16. These variability values are
suitable for use in the code calibration related to wind load (Melchers, 2002).
Ghiocel and Lungu (1975) proposed the constant c to be equal to 0.0625, and
Ellingwood et al. (1980) quantified the variability of this parameter as 0.05. These
values will also be used in this study for all locations considered. The mean and
total variability of the maximum wind load can be computed from the following
equations by utilizing FOSM method:
W = c.Cp .E z .G.V 2 (5.20)

2 2 2 2 2
Ω W = Ω c + Ω C p + Ω G + Ω E z + 2Ω V50 (5.21)
The statistics related to the yearly maximum wind load, Wan, and daily maximum
wind load, Wapt, can be calculated in a similar way. The mean values and total
107
uncertainties of the maximum, yearly maximum and daily maximum wind loads for
relevant locations are presented in Table 5.10. In addition, mean to nominal wind
load ratios are shown in Table 5.12. Type I distribution parameters for wind loads
mentioned above are also presented in Table 5.12 for different locations.
Table 5.10 Mean values and the total uncertainties of wind loads for different
locations
Parameter
W apt Wan W W′ Ω W apt Ω W an ΩW
(kN/m2) (kN/m ) 2 (kN/m2) (kN/m2)
Location
Ankara 0.056 0.147 0.425 0.96 0.52 0.41 0.32
Izmir 0.015 0.248 0.485 0.96 0.47 0.31 0.28
Bursa 0.038 0.207 0.537 0.96 0.67 0.38 0.31
Antalya 0.011 0.257 0.665 0.96 0.71 0.38 0.31
Gaziantep 0.011 0.111 0.34 0.96 0.66 0.42 0.33
Samsun 0.024 0.193 0.558 0.96 0.79 0.41 0.32
Malatya 0.015 0.082 0.476 0.96 0.80 0.69 0.37
Erzincan 0.016 0.132 0.468 0.96 0.67 0.47 0.34
Canakkale 0.045 0.304 0.702 0.96 0.63 0.35 0.30
Hakkari 0.018 0.148 0.397 0.96 0.80 0.39 0.32
Göztepe/Istanbul 0.017 0.14 0.514 0.96 0.40 0.48 0.34
Sile/ Istanbul 0.028 0.23 0.791 0.96 0.70 0.47 0.34
108
Table 5.11 Mean to nominal wind load ratios and associated total uncertainties for
different locations
Parameter
W apt / W ′ Ω W apt Wan / W ′ Ω W an W / W′ ΩW
2 2
Location (kN/m ) (kN/m ) (kN/m2)
Ankara 0.058 0.52 0.153 0.41 0.443 0.32

Izmir 0.016 0.47 0.258 0.31 0.505 0.28
Bursa 0.040 0.67 0.216 0.38 0.559 0.31
Antalya 0.011 0.71 0.268 0.38 0.693 0.31
Gaziantep 0.011 0.66 0.116 0.42 0.354 0.33
Samsun 0.025 0.79 0.201 0.41 0.581 0.32
Malatya 0.016 0.80 0.085 0.69 0.496 0.37
Erzincan 0.017 0.67 0.138 0.47 0.488 0.34
Canakkale 0.047 0.63 0.317 0.35 0.731 0.30
Hakkari 0.019 0.80 0.154 0.39 0.414 0.32
Sile/ Istanbul 0.029 0.70 0.240 0.47 0.824 0.34
Table 5.12 Type I distribution parameters of mean to nominal wind load ratios for
different locations
Parameter
W apt / W ′ Wan / W ′ W / W′
Location α µ α µ α µ
Ankara 0.044 42.53 0.125 20.44 0.379 9.05
Izmir 0.013 170.55 0.222 15.03 0.441 9.07
Bursa 0.028 47.85 0.179 15.63 0.481 7.401
Antalya 0.007 164.21 0.222 12.59 0.596 5.97
Gaziantep 0.007 176.66 0.094 25.33 0.301 10.98
Samsun 0.016 64.94 0.164 15.56 0.497 5.89
Malatya 0.010 100.20 0.059 21.86 0.413 5.99
Erzincan 0.012 112.63 0.109 19.77 0.413 7.73
Canakkale 0.034 43.32 0.267 11.56 0.632 5.84
Hakkari 0.012 84.38 0.127 21.35 0.354 9.68
Sile/ Istanbul 0.020 63.18 0.189 11.37 0.698 4.58
109
5.6 EARTHQUAKE LOAD
The properties of the ground motion at the site and the magnitude and location of
the next earthquake in a seismic source zone are inherently random. Inadequacy of
analytical models in evaluating nonlinear structural behavior leads to additional
uncertainties. On the other hand, the capacity of a structure cannot be determined
with reasonable accuracy due to various factors, such as material properties,
workmanship, and climatic nature. Naturally, seismic load, structural response and
structural capacity are all random and involve uncertainties, and a reliability-based
design criterion is required to include all these aspects in earthquake-resistant
design (Hwang and Hsu, 1993). Reliability analysis of structures under the
influence of earthquakes involves the identification of limit states associated with
the structural performance requirements, analysis of overall structural system
response and its components in addition to determination of the seismic hazard.
In this study, advanced computational methods for probabilistic analysis of

structural response to earthquakes are not addressed with all its specifics. A detailed
study of earthquake loading is beyond the scope of this study. However, regarding
the local conditions and data, probabilistic methods are used to determine the load
and resistance factors by taking the uncertainties into account, supported by the
results of studies which have been performed in other countries.
The peak ground acceleration is usually compatible with seismic specifications.

Accordingly, it is used as the seismic hazard parameter in this study. In other words,
the seismic hazard is described in terms of the peak ground acceleration, A. The
probability distribution of peak ground acceleration can be described by the Type II
extreme value distribution (Ellingwood, 1994). In a suitable manner, the cumulative
distribution function of A can be obtained from the following relationship:
110
−k
FA (a) = e −(a/v) a≥0 (5.22)
where:
v : the characteristic extreme value,
k : the shape parameter of the distribution.
Regarding the 100, 225, 475 and 1000 years return periods, peak ground
accelerations for Ankara, Izmir, Bursa, Antalya, Gaziantep, Samsun, Malatya,
Erzincan, Canakale, Hakkari, Göztepe/Istanbul Sile/Istanbul are obtained from the
study conducted by Gülkan et al. (1993) for the development of seismic zones map
of Turkey.
Table 5.13 Geographical coordinates and seismic zones of selected locations, and
corresponding peak ground acceleration values for different return periods
Peak ground acceleration, A (in g)

Longitude Latitude Seismic
Location (North) (East) Zone
100 225 475 1000
(in degrees) (in degrees)
years years years years
return return return return
period period period period
Ankara 32.853 39.929 4 0.13 0.16 0.19 0.21
Izmir 27.145 38.433 1 0.36 0.43 0.51 0.59
Bursa 29.075 40.196 1 0.35 0.42 0.50 0.58
Antalya 30.709 36.893 2 0.30 0.37 0.44 0.52
Gaziantep 37.389 37.069 3 0.14 0.17 0.20 0.24
Samsun 36.331 41.293 2 0.19 0.24 0.31 0.36
Malatya 38.309 38.355 1 0.29 0.35 0.41 0.48
Erzincan 39.504 39.740 1 0.40 0.49 0.59 0.70
Canakkale 26.414 40.155 1 0.40 0.48 0.57 0.66
Hakkari 43.751 37.568 1 0.38 0.47 0.56 0.65
Istanbul/Göztepe 29.082 40.980 1 0.29 0.35 0.42 0.50
Istanbul/Sile 29.628 41.175 2 0.21 0.25 0.32 0.38
111
The peak ground accelerations in these locations are set equal to the values shown
in Table 5.13 for different return periods. By using the acceleration values for 225
and 475 years return periods, the parameters of Type II distribution can be
computed. The probability that the peak ground acceleration will not be exceeded
over a period of 50 years is 0.8 and 0.9 for the return periods of 225 and 475 years,
respectively. It is to be noted that the economic life of a structure is mostly assumed
to be 50 years in codes in different countries. For the purpose of determining the
parameters of the Type II distribution, v and k, for each location, Eq. (5.22) is
solved for 225 years and 475 years return period, simultaneously. The calculated v
and k values for the locations mentioned above are given in Table 5.14.
Table 5.14 Parameters of Type II distribution for peak ground acceleration for
different locations
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
v 0.13 0.32 0.30 0.26 0.12 0.15 0.26 0.34 0.37 0.33 0.35 0.24
k 4.37 4.98 4.30 4.33 4.63 3.36 4.74 4.04 5.06 4.28 3.04 4.11
The values of the means and coefficients of variation of peak ground acceleration
corresponding to Ankara, Izmir, Bursa, Antalya, Gaziantep, Samsun, Malatya,
Erzincan, Canakkale, Hakkari, Göztepe/Istanbul and Sile/Istanbul can be found by
substituting the computed v and k values into Eqs. (5.23) and (5.24), respectively.
The computed values of δA indicate only the basic variability (aleatory uncertainty)
in peak ground acceleration. In addition, modeling error, which is quite high due to
various uncertainties associated with the earthquake process, should be taken into
consideration.
112
 1
A = vΓ1 -  (5.23)
 k
 2
Γ1 − 
 k  −1
δA = (5.24)
 1
Γ 2 1 − 
 k
where, Γ(.) is the gamma function.
In this study, the modeling error, ∆A, will be taken as 0.9 in view of the studies of
Ellingwood et al. (1980) and Kömürcü (1995) who have reported this uncertainty to
be equal to 0.9. Consequently, total variability, ΩA, is calculated using the basic
variability, δA and the modeling error, ∆A. Results are shown in Table 5.15.
Table 5.15 Mean value and total variability of peak ground acceleration for
different locations
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.15 0.37 0.36 0.32 0.15 0.20 0.30 0.41 0.43 0.40 0.48 0.29
δA 0.38 0.32 0.38 0.38 0.35 0.55 0.34 0.42 0.31 0.39 0.66 0.41
∆A 0.90 0.90 0.90 0.90 0.90 0.90 0.90 0.90 0.90 0.90 0.90 0.90
ΩA 0.98 0.95 0.98 0.98 0.97 1.06 0.96 0.99 0.95 0.98 1.12 0.99
5.6.1 Determination of Total Equivalent Lateral Earthquake Load
Building codes specify the seismic load with regards to the maximum shear force at
113
the base of the building. In other words, the seismic analysis of buildings can be
performed in accordance with the equivalent lateral force procedure. The formal
seismic code TEC-1998 (Specification for Structures to be Built in Disaster Areas)
is used in determining the nominal earthquake loads for Turkey. It is hard to say
which code gives the best estimate of the “true” value of the lateral base shear force
for buildings in Turkey, considering the mean value of earthquake load. TEC-2006
and IBC 2003 (International Building Code) seem to reflect the most up-to-date
version of the equivalent lateral force procedure in terms of the estimation of the
nominal earthquake load and the mean earthquake load for this study, respectively.
However, in IBC- 2003, the earthquake spectral acceleration at short periods (Ss)
and at 1-second period (S1) are taken from the maps which were not mapped for
Turkey. If the base shear is taken into consideration, UBC-1994 is compatible with
TEC-2006 in terms of design variables, such as response modification factor, site
coefficient for soil characteristics. On the other hand, Yüksel (1997) pointed out
that the analysis and design of reinforced concrete buildings and its calculation
principles in UBC-1994 and TEC-1996 (the draft of TEC-1998) are almost the
same. TEC-1996 is the basis of TEC-2006 and the calculation procedure of base
shear in TEC-2006 has not been changed since 1996. Therefore, for the
computation of nominal earthquake loads, TEC-2006 can be also taken into
consideration. In the following sections, brief information associated with these
codes, i.e. UBC-1994 and TEC-2006 is presented.
5.6.1.1 UBC-1994
In UBC-1994, the total design base shear, V, in a given direction is determined

from the following equations:
ZIC
V= W (5.25)
RW
114
1.25 S
C= < 2.75 (5.26)
T2/3
C
≥ 0.075 (5.27)
RW
where:
Z: seismic zone factor
I: importance factor
W: total building weight
C: numerical coefficient
Rw: response modification factor
S: site coefficient of soil properties
T: fundamental period of vibration of the building for the direction considered
Seismic zone factor, Z, is the ratio of the design ground acceleration to the
acceleration of gravity, g (9.81 m/s2). Zone 4 in UBC-1994 is the most critical one,
whereas Zone 1 in TEC-2006 is the most critical one; that is, Zone 4 in UBC-1994
corresponds to Zone 1 in TEC-2006. The values of the seismic zone factor, Z in
UBC-1994 are given in Table 5.16 for different seismic zones.
Table 5.16 Seismic zone factor in UBC-1994
ZONE 1 2A 2B 3 4
Z 0.075 0.15 0.20 0.30 0.40
The ductility of the structural system, the types of member and material which are
normally ignored in linear elastic calculations is quantified by the response
reduction factor, RW in connection with the energy dissipation capacity of the
structure. The response modification factor in UBC-1994 is given in Table 5.17.
115
The first natural vibration period of the building can be calculated by the following
expression that gives an approximate value:
T1 = C t H 3/4
N (5.28)
Ct value equals to 0.0731 in UBC-1994 for reinforced concrete moment resisting

frames and eccentrically braced frames.
Table 5.17 Response modification factor (RW) for reinforced concrete buildings in
UBC-1994
Basic structural
Lateral force- resisting- system description RW
system
Bearing wall Reinforced concrete shear walls 6

system
Building frame Reinforced concrete shear walls 8

system
Special moment-resisting frames(SMRF) 12

Moment
resisting frame Intermediate moment-resisting frames (IMRF) 8
system Ordinary moment-resisting frames (OMRF) 5
Dual system with SMRF capable of resisting at

Dual system 12
least 25%of prescribed seismic forces
Dual system with IMRF capable of resisting at

Dual system 8
least 25% of prescribed seismic forces
5.6.1.2 TEC-2006
The general principle of earthquake resistant design in TEC-2006 is to avoid any

damage in the structural and non-structural elements of buildings for low intensity
116
earthquakes, to limit the repairable damage levels in structural and non-structural
elements for medium intensity earthquakes, and to avoid the partial or overall
collapse of buildings for high intensity earthquakes in order to prevent the loss of
life.
In TEC-2006, the total design base shear, Vt, is determined by using the following
equation for a given direction:
A(T1 )
Vt = W ≥ 0.10A 0 IW (5.29)
R a (T1 )
where:
W :total building weight,
A(T1) : spectral acceleration coefficient,
Ra(T1) : seismic load reduction factor,
T1 :fundamental period of vibration of the building for translation motion in the
direction considered,
A0 :effective ground acceleration coefficient,
I :importance factor.
Total building weight is determined from the following equation:
N
W = ∑ wi (5.30)
i =1
In the above equation, wi is the individual storey weight.
The spectral acceleration coefficient corresponding to 5% damping is given by the

following equation:
117
A(T) = A 0 IS(T) (5.31)
Effective ground acceleration coefficient, A0, can be defined as the ratio of the
design ground acceleration to the acceleration due to gravity, g (9.81 m/s2). The
effective ground acceleration coefficients specified for different seismic zones in
TEC-2006 are shown in Table 5.18.
Table 5.18 Effective ground acceleration coefficients (A0) in TEC-2006
Seismic Zone A0
1 0.40
2 0.30
3 0.20
4 0.10
In TEC-2006, there is no equation in order to calculate the first natural period,

except for two limitations as the highest and lowest level. In TEC-1998 and UBC
1994, the first natural vibration period of a building can be calculated from the
following approximate expression:
T1 = C t H 3/4
N (5.32)
Ct value equals to 0.07 in TEC-1998 for buildings whose structural system is

composed of only reinforced concrete frames or structural steel eccentric braced
frames. Approximately, the same calculation procedure is given in both UBC-1994
and TEC-1998 for the value of Ct of buildings where seismic loads are fully resisted
by reinforced concrete structural walls.
The spectrum coefficient S(T), which appears in Eq. (5.31) is determined from the
118
following equations depending on the building’s natural period, T, and the local
site classes. Spectrum characteristic periods (TA, TB) in TEC-2006 are given in
Table 5.19.
S(T) = 1 + 1.5T/TA (0≤T≤TA) (5.33 a)

S(T) = 2.5 (TA≤T≤TB) (5.33.b)
S(T) = 2.5 + (TB /T) 0.8 (T>TB) (5.33.c)
where:
T: fundamental period of vibration of the building in the direction considered
TA, TB : spectrum characteristic periods which depend on the local soil class given
Table 5.19 Spectrum characteristic periods (TA, TB) in TEC-2006
Local site class TA(second) TB (second)

Z1 0.10 0.30
Z2 0.15 0.40
Z3 0.15 0.60
Z4 0.20 0.90
These local site classes are classified according to the thickness of the soil topmost
layer and the soil groups, such as massive volcanic rocks, soft deep alluvial layers
with high water table, and so on.
In order to account for the specific nonlinear behavior of the structural system,
seismic load reduction factor, Ra(T), which corresponds to response modification
factor, RW, in UBC-1994 is used. Regarding the various structural systems and
natural vibration periods, seismic load reduction factors are determined from the
119
following equations in terms of structural behavior factor, R, which is given in
Table 5.20 according to TEC-2006.
Ra(T)=1.5+(R-1.5)T/TA (0≤T≤TA) (5.34.a)

Ra(T)=R (T>TA) (4.34.b)
Table 5.20 Structural system behavior factor (R) in TEC-2006 for cast-in-situ
reinforced concrete buildings
System of System of
nominal high
Lateral force- resisting- system description
ductility ductility
level level
Buildings in which earthquake loads are fully resisted 4 8

by frames
by coupled structural walls
by solid structural walls
Buildings in which earthquake loads are jointly 4 7

resisted by frames and solid and/or coupled structural
walls
Here, in order to determine the mean to nominal ratios of earthquake loads, E / E ′ ,

different cases are taken into consideration with respect to importance factor,
seismic zone and seismic load reduction factor, Ra(T), (this term is given as
response modification factor, RW, in UBC-1994). It is to be noted that, in these
cases, the mean value of earthquake load, E , is computed from UBC-1994 and the
nominal value, E ′ , from TEC-2006 by using a code developed in MathCAD 12.
120
5.6.1.3 Case 1
Considering UBC-1994, importance factor, I, is taken as 1.0 for buildings which are
residential and office buildings, hotels, industrial structures, etc., and the response
modification factor, RW, is assumed to be 8 suitable for buildings in which seismic
loads are resisted by reinforced concrete shear walls and its basic structural system
is building frame system.
As for TEC-2006, the importance factor, I, is taken as 1.0 corresponding to

buildings in which small numbers of people live (houses, hotels, employment
buildings, restaurants, and industrial buildings) and seismic load reduction factor,
Ra(T), is assumed to be 7 in connection with buildings in which earthquake loads
are resisted by frames and solid and/or coupled structural walls.
In the light of above descriptions and values, the mean to nominal ratios of
earthquake load in terms of different local site classes and building heights are
computed by using a code written in MathCAD 12, and the results are summarized
in Tables 5.21 and 5.22.
121
Table 5.21 Mean to nominal ratios of earthquake load in terms of different local site
classes and building heights (RW=8, Ra(T)=7)
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.19 0.51 0.50 0.44 0.20 0.31 0.41 0.59 0.57 0.56 0.42 0.32
A0 0.10 0.40 0.40 0.30 0.20 0.30 0.40 0.40 0.40 0.40 0.40 0.30
Building height (HN) is equal to 8 m, and local site class is Z1

E / E′ 1.27 0.85 0.83 0.98 0.67 0.69 0.68 0.98 0.95 0.93 0.70 0.71

E / E′ 1.01 0.68 0.66 0.71 0.53 0.55 0.54 0.78 0.76 0.74 0.56 0.57

E / E′ 0.87 0.58 0.57 0.67 0.46 0.47 0.47 0.68 0.65 0.64 0.48 0.49

E / E′ 0.77 0.51 0.50 0.59 0.40 0.42 0.41 0.60 0.57 0.56 0.42 0.43

E / E′ 1.83 1.23 1.20 1.41 0.97 1.00 0.99 1.42 1.37 1.35 1.01 1.03

E / E′ 1.15 0.77 0.76 0.89 0.61 0.63 0.62 0.89 0.86 0.85 0.64 0.65

E / E′ 1.00 0.67 0.66 0.77 0.53 0.54 0.54 0.78 0.75 0.74 0.55 0.56

E / E′ 0.89 0.60 0.58 0.68 0.47 0.48 0.48 0.69 0.67 0.65 0.49 0.50
122
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.19 0.51 0.50 0.44 0.20 0.31 0.41 0.59 0.57 0.56 0.42 0.32
A0 0.10 0.40 0.40 0.30 0.20 0.30 0.40 0.40 0.40 0.40 0.40 0.30
E / E′ 1.83 1.23 1.20 1.41 0.97 1.00 0.99 1.42 1.37 1.35 1.01 1.03
E / E′ 1.83 1.23 1.20 1.41 0.97 1.00 0.99 1.42 1.37 1.35 1.01 1.03
E / E′ 1.16 0.78 0.76 0.90 0.61 0.63 0.65 0.90 0.87 0.86 0.64 0.65
E / E′ 1.04 0.70 0.68 0.80 0.55 0.57 0.56 0.81 0.78 0.77 0.57 0.58
E / E′ 1.83 1.23 1.20 1.41 0.97 1.00 0.99 1.42 1.37 1.35 1.01 1.03
E / E′ 1.83 1.23 1.20 1.41 0.97 1.00 0.99 1.42 1.37 1.35 1.01 1.03
E / E′ 1.83 1.23 1.20 1.41 0.97 1.00 0.99 1.42 1.37 1.35 1.01 1.03
E / E′ 1.79 1.20 1.18 1.38 0.94 0.97 0.96 1.39 1.34 1.32 0.99 1.00
123
5.6.1.4 Case 2
Considering UBC-1994, the importance factor, I, is taken as 1.0 for buildings which
are residential and office buildings, hotels, industrial structures, etc., and the
response modification factor, RW, is assumed to be 6 corresponding to buildings in
which seismic loads are resisted by reinforced concrete shear walls and its basic
structural system is bearing wall system.
Regarding TEC-2006, importance factor, I, is used as 1.0 corresponding to

buildings in which small number of people live (houses, hotels, employment
buildings, restaurants, industrial buildings) and seismic load reduction factor, Ra(T),
is assumed to be 6 corresponding to buildings in which earthquake loads are fully
resisted by solid structural walls
In view of above descriptions and values obtained from UBC-1994 and TEC-2006,
which are used for the calculation of the mean earthquake load, E , and the nominal
earthquake load, E ′ , respectively. The mean to nominal ratios of earthquake load
with corresponding to different local site classes and building heights are computed
and summarized in Tables 5.23 and 5.24.
124
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.19 0.51 0.50 0.44 0.20 0.31 0.41 0.59 0.57 0.56 0.42 0.32
A0 0.10 0.40 0.40 0.30 0.20 0.30 0.40 0.40 0.40 0.40 0.40 0.30

E / E′ 1.45 0.97 0.95 1.12 0.76 0.79 0.78 1.12 1.08 1.07 0.80 0.81

E / E′ 1.15 0.77 0.76 0.89 0.61 0.63 0.62 0.90 0.87 0.85 0.64 0.65

E / E′ 0.99 0.67 0.65 0.77 0.52 0.54 0.54 0.77 0.75 0.73 0.55 0.56

E / E′ 0.88 0.59 0.58 0.68 0.46 0.48 0.47 0.68 0.66 0.65 0.40 0.49

E / E′ 2.09 1.40 1.38 1.61 1.10 1.14 1.13 1.62 1.57 1.54 1.16 1.17

E / E′ 1.32 0.88 0.87 1.02 0.69 0.72 0.71 1.02 0.97 0.97 0.73 0.74

E / E′ 1.14 0.77 0.75 0.88 0.60 0.62 0.62 0.89 0.86 0.84 0.63 0.64

E / E′ 1.01 0.68 0.67 0.78 0.53 0.55 0.55 0.79 0.76 0.75 0.56 0.57
125
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.19 0.51 0.50 0.44 0.20 0.31 0.41 0.59 0.57 0.56 0.42 0.32
A0 0.10 0.40 0.40 0.30 0.20 0.30 0.40 0.40 0.40 0.40 0.40 0.30

E / E′ 2.09 1.40 1.38 1.61 1.10 1.14 1.13 1.62 1.57 1.54 1.16 1.17

E / E′ 2.09 1.40 1.38 1.61 1.10 1.14 1.13 1.62 1.57 1.54 1.16 1.17

E / E′ 1.33 0.89 0.87 1.02 0.70 0.72 0.72 1.03 0.99 0.98 0.73 0.74

E / E′ 1.09 0.80 0.78 0.92 0.63 0.65 0.64 0.92 0.89 0.88 0.66 0.67

E / E′ 2.09 1.40 1.38 1.61 1.10 1.14 1.13 1.62 1.57 1.54 1.16 1.17

E / E′ 2.09 1.40 1.38 1.61 1.10 1.14 1.13 1.62 1.57 1.54 1.16 1.17

E / E′ 2.09 1.40 1.38 1.61 1.10 1.14 1.13 1.62 1.57 1.54 1.16 1.17

E / E′ 2.04 1.37 1.34 1.58 1.07 1.11 1.10 1.59 1.53 1.50 1.13 1.15
126
5.6.1.5 Case 3
Considering the UBC-1994, the importance factor I is taken as 1.0 for buildings
which are residential and office buildings, hotels, industrial structures, etc., and the
response modification factor, RW, is assumed to be 6 suitable for buildings in which
seismic loads are resisted by reinforced concrete shear walls and its basic structural
system is bearing wall system.
In TEC-2006, however, the importance factor I is used as 1.0 corresponds to

buildings in which small number of people live (houses, hotels, employment
buildings, restaurants, industrial buildings) and the seismic load reduction factor,
Ra(T), is assumed to be 4 for systems with normal ductility level in connection with
buildings in which earthquake loads are fully resisted by solid structural walls
In the light of above descriptions based on UBC-1994 and TEC-2006, the mean to
nominal ratios of earthquake load according to different local site classes and
building heights are obtained, and results are presented in Tables 5.25 and 5.26.
127
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.19 0.51 0.50 0.44 0.20 0.31 0.41 0.59 0.57 0.56 0.42 0.32
A0 0.10 0.40 0.40 0.30 0.20 0.30 0.40 0.40 0.40 0.40 0.40 0.30

E / E′ 0.96 0.65 0.63 0.74 0.51 0.52 0.52 0.75 0.72 0.71 0.53 0.54

E / E′ 0.77 0.52 0.51 0.59 0.40 0.42 0.41 0.60 0.58 0.57 0.42 0.43

E / E′ 0.66 0.44 0.44 0.51 0.35 0.36 0.36 0.51 0.50 0.49 0.37 0.37

E / E′ 0.58 0.39 0.38 0.45 0.31 0.32 0.31 0.45 0.44 0.43 0.32 0.33

E / E′ 1.39 0.94 0.92 1.08 0.73 0.76 0.75 1.08 1.05 1.03 0.77 0.78

E / E′ 0.88 0.59 0.58 0.68 0.46 0.48 0.47 0.68 0.66 0.65 0.48 0.49

E / E′ 0.76 0.51 0.50 0.59 0.40 0.41 0.41 0.59 0.57 0.56 0.42 0.43

E / E′ 0.68 0.45 0.44 0.52 0.36 0.37 0.36 0.52 0.51 0.50 0.37 0.38
128
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.19 0.51 0.50 0.44 0.20 0.31 0.41 0.59 0.57 0.56 0.42 0.32
A0 0.10 0.40 0.40 0.30 0.20 0.30 0.40 0.40 0.40 0.40 0.40 0.30

E / E′ 1.39 0.94 0.92 1.08 0.73 0.76 0.75 1.08 1.05 1.03 0.77 0.78

E / E′ 1.39 0.94 0.92 1.08 0.73 0.76 0.75 1.08 1.05 1.03 0.77 0.78

E / E′ 0.88 0.59 0.58 0.68 0.47 0.48 0.48 0.69 0.66 0.65 0.49 0.50

E / E′ 0.79 0.53 0.52 0.61 0.42 0.43 0.43 0.61 0.59 0.58 0.44 0.44

E / E′ 1.39 0.94 0.92 1.08 0.73 0.76 0.75 1.08 1.05 1.03 0.77 0.78

E / E′ 1.39 0.94 0.92 1.08 0.73 0.76 0.75 1.08 1.05 1.03 0.77 0.78

E / E′ 1.39 0.94 0.92 1.08 0.73 0.76 0.75 1.08 1.05 1.03 0.77 0.78

E / E′ 1.36 0.91 0.90 1.05 0.72 0.74 0.73 1.06 1.02 1.00 0.75 0.76
129
The average values of mean to nominal ratios of the earthquake load displayed in
Table 5.21 to Table 5.26 are summarized in Table 5.27.
Table 5.27 The average mean to nominal ratios of earthquake load obtained from
UBC 1994 and TEC-2006 for different locations
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Situation
RW=8, 1.37 0.92 0.90 1.05 0.72 0.75 0.74 1.06 1.03 1.01 0.76 0.77
Ra(T)=7
RW=6 1.56 1.05 1.03 1.21 0.82 0.85 0.85 1.21 1.17 1.15 0.86 0.88
Ra(T)=6
RW=6 1.04 0.70 0.69 0.81 0.55 0.57 0.56 0.81 0.78 0.77 0.58 0.58
Ra(T)=4
Average 1.32 0.89 0.87 1.02 0.70 0.72 0.72 1.03 0.99 0.98 0.73 0.74
On the other hand, assuming that TEC-2006 gives the best estimate of the “true”
value of lateral base shear force for buildings in Turkey, this code can also be used
to compute “true” value of the earthquake load. However, since the nominal
earthquake load is also computed based on TEC-2006, the ratio of E / E ′ will be
equal to A/A0. The resulting mean to nominal ratios of earthquake load are shown in
Table 5.28.
Table 5.28 Mean to nominal ratios of earthquake load for different locations where
both values are computed based on TEC-2006
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Samsun
Antalya
Ankara
Bursa
Izmir
Parameter
E / E′ 1.90 1.28 1.25 1.47 1.00 1.03 1.03 1.48 1.43 1.40 1.05 1.07
130
In his study, Ellingwood (1994) stated that the uncertainty in earthquake load is
dominated by the uncertainty involved in the seismic hazard analysis; the c.o.v. in A
is typically around 0.80 or more while the c.o.v. due to other structural response
parameters is 0.30 or less. Kömürcü and Yücemen (1996) assumed that the basic
variability due to other factors different than those involved in the estimation of A
in the earthquake load is equal to 0.6. In the light of these studies, we can take the
c.o.v. to be 0.45 as the average of the two values given above for the modeling error
associated with the seismic load in terms of maximum shear forces at the base of
buildings.
It is assumed that mean to nominal ratio of earthquake load exhibits a Type II

extreme value distribution like the peak ground acceleration. The results of the
average mean to nominal ratio of earthquake load and total variability are shown in
Table 5.29 together with the parameters of Type II extreme value distribution. It is
to be noted that in this table, the mean to nominal ratio of earthquake load is the
average value obtained from the mean to nominal ratios given in Tables 5.27 and
5.28.
Table 5.29 Statistical parameters of the mean to nominal ratio for earthquake load
Location
Sile/Istanbul
Canakkale
Gaziantep
Erzincan
Göztepe/
Hakkari
Malatya
Istanbul
Antalya
Samsun
Ankara
Bursa
Izmir
Parameter
A 0.15 0.37 0.36 0.32 015 020 0.30 041 0.43 0.40 0.48 0.29
ΩA 0.98 0.95 0.98 0.98 0.97 1.06 0.96 0.99 0.95 0.98 1.12 0.99
E / E′ 1.61 1.08 1.06 1.24 0.85 0.88 0.87 1.25 1.21 1.19 0.89 0.91
∆ 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45
Ω E / E′ 1.08 1.05 1.08 1.08 1.07 1.15 1.06 1.09 1.05 1.08 1.21 1.09
v 1.10 0.74 0.73 0.85 0.58 0.59 0.54 0.86 0.83 0.82 0.60 0.62
k 2.46 2.49 2.46 2.46 2.47 2.42 2.48 2.46 2.49 2.46 2.38 2.46
131
CHAPTER 6
ASSESSMENT OF THE SAFETY LEVELS INHERENT

IN THE CURRENT DESIGN PRACTICE
6.1 INTRODUCTION
In developing, a probability based load and resistance criterion for reinforced

concrete beam, column and shear wall design in flexure and shear as well as column
design in combined action of flexure and axial load, it is first necessary to evaluate
the current level of safety inherent in the design practice in Turkey and the relevant
“Turkish Codes”. Here, the term of “Turkish Codes” stands for a group of codes
including “Requirements for Design and Construction of Reinforced Concrete
Structures” (i.e. TS 500 (2000)), “Design Loads for Buildings” (i.e. TS 498 (1997))
and “Specification for Structures to be Built in Disaster Areas (1998)”. The required
information and data which will be used in the assessment of the levels of safety
inherent in the Turkish Codes are given in the previous chapters of this dissertation.
Probability based load and resistance design traditionally employs resistances which
are multiplied by resistance factors and loads or load effects which are multiplied
by load factors in a set of safety checking equations (Ellingwood et al., 1980). In its
most general form, probability based load and resistance design criterion can be
represented as follows (Israel et al., 1987):
Factored resistance ≥ Effects of factored load (6.1)
132
In the above equation, resistance factors are generally smaller than one while load
factors are greater than one. While developing a probability based load and
resistance criterion, these factors should be determined in such a way that the
probability of an unfavorable combination of resistance and load variables is
acceptably small for all likely combinations of the design variables which are
involved in the safety checking equation (Ellingwood and Galambos, 1982).
6.2 LOAD COMBINATION
Different types of loads act on a structure during its lifetime, and most of these
loads, such as wind, snow and earthquake exhibit a large degree of variation from
one location to another, with both time and their randomness in nature. In a load
combination case, if a structural member is subjected to only one time-varying load
together with dead load, it is essential to make use of the lifetime maximum load for
the time-varying load. On the other hand, when more than one time-varying load act
on a structure, it is not suitable to consider their lifetime maximum all together; in
this respect, the load combination problem becomes rather complex. In such a case,
the total maximum load effects of these time-varying loads are not the sum of their
maximums since the simultaneous occurrence of the maximum load effects of these
time-varying loads are rarely observed. For example, there is a very low chance of a
strong earthquake load taking place along with an extreme high live load.
Consequently, a structural component should rationally be designed for a total load
which is less than the sum of the time-varying loads at their maximum (Galambos et
al., 1982).
Three methods are generally used in load combination problems, namely Ferry-
Borges Model, Wen’s Load Coincide Method and Turkstra’s rule (Aktas et al.,
2001). Since Turkstra’s rule has been widely used in code calibration processes in
most of today’s codes due to its practicality in application, this method is adopted in
this study. Another reason for the selection of this load combination rule is that it is
133
consisted with the observations which failures frequently occur resulting from only
one load reaching its maximum value.
6.2.1 Turkstra’s Rule
As the probability of simultaneous occurrence of time-varying loads at their

maximum simultaneously is so small, this rule assumes that the maximum of
combined loads will occur when one of the loads is at its maximum while the rest
are at their arbitrary point-in-time values. If the total load effect U(t) is assumed as
given in Eq. (6.2), then a group of sub-combinations is achieved as given in Eq.
(6.3), and the maximum value obtained from these groups is maximum of the
combined loads.
U(t)= X1(t)+X2(t)+…+Xn(t) (6.2)
X 1 max + X 2 + X 3 + ... + X n 
X + X 
 1 2 max + X 3 + ... + X n 
max U= max   (6.3)
M 
X 1 + X 2 + X 3 + ... + X n max 
where Xi represents the arbitrary point-in-time load and Ximax stands for the
maximum value of ith load.
If the combination of dead, live and snow loads are considered, dead load can be
assumed to be time independent owing to the fact that variation in dead load
throughout the lifetime of a structure is significantly low in comparison with live
and snow loads. Hence, the combined effect of dead load, live load and snow load
would be the maximum of dead load plus the maximum combination of maximum
live load with arbitrary point-in-time value of snow load or maximum snow load
with arbitrary point-in-time value of live load, which can be represented as follows:
134
U(t)= D+L(t)+…+S(t)
L max + Sapt  (6.4)
max U= D + max  
L apt + S max 
6.2.2 Safety Criterion
The safety criterion for a reinforced concrete structural element can be written as
follows:
φR ≥ γDD + γYQ + ∑γJ Q apt j (6.5)

where:
γ : load factor
D : dead load
Q : principle variable load
Q apt j : arbitrary point-in-time values of the other variable loads.
The expression on the right hand side of the inequality consists of nominal loads
multiplied by load factors, and is consistent with Turkstra’s rule for load
combination (Kömürcü and Yücemen, 1996). As it is indicated before, R is the
nominal capacity, and φ is the resistance factor. One expects that the capacity R and
the different loads considered in this study to be correlated to a certain extent since
the basic variables related to dimensions appear both in the resistance and the load
effect equations. This is observed especially between the dead load, D, and the
capacity, R. However, since the variabilities in the dimensions are quite small (see
Section 3.3) compared to the other basic variables, the contribution of the
correlation to the total uncertainty will be of negligible magnitude. Accordingly, the
possible correlations between capacity and different load effects are not considered.
135
The fundamental formulation related to the load combinations in TS 500 (2000) is
given by Eq. (6.6), and safety criterion given in Eq. (6.5) is based on this formula
(Ersoy and Özcebe, 2004).
   
Factored Load Effects = U  γ d D ' + γ Q  Q i + ∑ α oj Q j   (6.6)
  i≠ j  
where:
Qi : characteristics value of principle variable load
Qj : characteristics value of other less important variable load
αoj : ratio of arbitrary point-in-time value of the ith load to its nominal value.
In view of the above equations, only dead, live, wind, snow and earthquake loads
will be taken into consideration in this study. Furthermore, instead of all possible
combinations, only five load combinations, which are also considered in TS 500,
will be analyzed. These combinations are as follows:
a) U = γ D D′ + γ L L′
b) U = γ D D′ + γ L S′
c) U = γ D D′ + γ L apt L′apt + γ W W ′ (6.7)

d) U = γ D D′ + γ L L′t + γ W apt Wapt
′
e) U = γ D D′ + γ L apt L′apt + γ E E′
6.2.3 Load Statistics
As stated before, it is necessary to deal with the effects of loads rather than the
loads themselves. In Chapter 5, structural load effects resulting from dead, live,
wind, snow and earthquake loads are evaluated. The data published in literature and
136
the local data collected in Turkey constitute the main sources of the information
used in the evaluation of the load statistics. For calibration purposes, the ratios of
mean to nominal load values are determined. The nominal values of loads are
obtained from TS 498 (1997) and Specification for Structures to be Built in Disaster
Areas (1998)
Although dead and live loads acting on a structure are independent of the
geographical location of the structure, environmental loads, such as snow, wind and
earthquake loads are highly dependent. Therefore, in order to compute the
environmental loads, some representative cities (see Section 5.1), namely Ankara,
Izmir, Antalya, Bursa, Gaziantep, Samsun, Malatya, Erzincan, Canakkale, Hakkari,
Göztepe/Istanbul and Sile/Istanbul are considered. The overall load effects of snow,
wind and earthquake loads for Turkey include the mean values of arbitrary point-in-
time and maximum lifetime loads in these locations.
On the other hand, it is impractical to compute the values of the safety level
inherent in the Turkish Codes for each location separately. For this reason,
considering each environmental load individually, the average mean to nominal
load ratio and total uncertainty for the corresponding load for Turkey are
determined from the twelve locations mentioned above. Furthermore, for a specific
load, it is necessary to compare these average values computed for the whole
Turkey with the highest of the 12 pairs (mean to nominal ratio and total uncertainty
for each location). It should be noted that the location having the highest mean to
nominal ratio and total uncertainty will be selected as the critical location.
According to relevant standards, these critical locations are taken as Bursa,
Canakkale and Ankara for snow, wind and earthquake loads, respectively. In fact,
for the snow load, the highest mean to nominal ratio among these 12 locations is
observed for Hakkari with respect to the present standard (i.e. TS 498 (1997)).
However, the mean to nominal maximum snow load ratio computed for Hakkari
( S ′ = 2.68 ) is not in accordance with the ratios of maximum snow loads proposed
S
137
in the foreign load standards. Furthermore, the nominal snow load values in TS 498
(1997) are expected to be revised in the near future by the authorities, especially in
locations where snowfall is excessive. Accordingly, since Bursa has the second
highest mean to nominal ratio among the above-mentioned 12 locations, based on
the Turkish Standards, this location is taken as the critical one for snow load.
The statistical parameters of dead, live, snow, wind and earthquake loads for critical
locations and for the whole Turkey, which were assessed in Chapter 5, are
summarized in Table 6.1. Note that the values computed for Turkey are the
averages of these twelve locations. In Table 6.1, APT stands for arbitrary point-in-
time.
Table 6.1 Statistics of dead, live, wind and earthquake loads
Type of Load Mean to Nominal Ratio Total Variability
Canak- Canak-
Ankara Bursa Turkey Ankara Bursa Turkey
kale kale
Dead Load
1.05 1.05 1.05 1.05 0.10 0.10 0.10 0.10
(D)
Live Load
1.00 1.00 1.00 1.00 0.27 0.27 0.27 0.27
(L)
APT Live Load
0.28 0.28 0.28 0.28 0.70 0.70 0.70 0.70
(Lapt)
Wind Load
0.44 0.55 0.73 0.55 0.32 0.31 0.30 0.32
W
APT Wind Load
0.15 0.22 0.32 0.19 0.41 0.38 0.35 0.43
Wapt
Snow Load
0.35 0.91 0.52 0.58 0.31 0.41 0.48 0.39
S
Earthquake Load
1.61 1.06 1.21 1.09 1.08 1.08 1.05 1.09
E
138
6.3 RESISTANCE STATISTICS
The published data in literature and the local data collected in Turkey constitute the
main sources of information in evaluating the resistance statistics. For calibration
purposes, the ratios of mean to nominal values are determined. The nominal values
of resistance parameters are obtained from TS 500 (2000), TS EN 206-1 (2002) and
TS 708 (1996). In this study, C14, C16, C18, C20, C25 and C30 concrete classes,
which are given in TS 500 (2000), are taken into consideration. The nominal
average value of concrete (i.e. weighted average concrete class) and in-situ average
value (i.e. real weighted average compressive strength) are determined from the
statistical analysis of 28 day compressive strength data for Turkey. BCIII(a)
reinforcing steel bars with diameters of 8, 10, 12, 14, 16, 18, 20, 22, 24, 26 and 32
mm are examined. The average yield strength of BCIII(a) reinforcing steel bars
used in reliability analysis include above diameters. In addition, the dimensional
characteristics of reinforced concrete members are also taken into account. The
mean to nominal ratios of beam, column and shear wall dimensions like width,
depth and effective depth and total variabilities in these dimensions are used to
obtain the resistance statistics.
In Chapter 3, the mean to nominal ratios and the total variabilities of basic
resistance variables, such as compressive strength of concrete, yield strength of
reinforcing steel bars and dimensions of three types of reinforced concrete elements,
beams columns and shear walls are quantified. In Chapter 4, using these values, the
mean to nominal ratios and the total uncertainties of resistances for different failure
modes of these structural members are computed within the framework of reliability
analysis. In this chapter, probability based resistance criterion for the design of
reinforced concrete beams, columns and shear walls in the flexure and shear failure
modes as well as the design of column in the combined action of flexure and axial
load failure mode will be examined within the scope of the information provided in
previous chapters. The resistance statistics for different failure modes of structural
members, which are determined in Chapter 4, are given in Table 6.2.
139
Table 6.2 Resistance statistics for different reinforced concrete members in different
failure modes
Structural R
Failure mode R′ Ω
member
Flexure 1.24 0.13

Beam
Shear 1.24 0.17
Combined
1.24 0.14
Column action
Shear 1.24 0.17
Flexure 1.24 0.13

Shear wall
Shear 1.24 0.14
When the mean to nominal ratios of the capacities of a reinforced concrete member
are computed by using the mean to nominal ratios of all basic variables, it is
observed that while the computed values may vary for different cases, their average
values are equal to approximately 1.24 (see Tables 4.6 to 4.9). In fact, this is not an
unexpected result considering that the mean to nominal ratios of the capacities are
governed by two “primary” basic variables, namely, compresive strength of
concrete and yield strength of reinforcing steel bars, whose mean to nominal ratios
are very close to each other, respectively 1.24 and 1.25, which are significantly
greater than that of the other parameters, such as the member dimensions for which
this ratio is 1.0. On the other hand, since the coefficients of variation of these
“primary” variables, 0.18 for concrete compressive strength and 0.09 for yield
strength of reinforcing steel bars, are much different compared to their mean to
nominal ratios; such a consistency could not be achieved in the coefficients of
varition for the capacities of structural members in different failure modes; which
are observed to range from 0.13 to 0.17.
140
6.4 COMPUTATION OF RELIABILITY INDEXES
Utilizing the statistical information presented in the previous chapters of this

dissertation, one can compute reliability index, β, which provides a basis for the
reliability based design criteria. Reliability indexes give guidance for the
determination of the target reliabilities to be used in the establishment of load and
resistance factor design criteria. However, since they include the shortcomings of
existing standards, strict reliance on these computed values is not recommended by
researchers. In this section, the β values corresponding to the safety level inherent
in the current Turkish design practice and the Turkish Codes will be computed for
the following failure mode and structural member combinations:
- reinforced concrete beams in the flexural failure mode,

- reinforced concrete beams in the shear failure mode,
- reinforced concrete columns in the combined action of flexure and axial
load failure mode,
- reinforced concrete columns in the shear failure mode,
- reinforced concrete shear walls in the flexural failure mode,
- reinforced concrete shear walls in the shear failure mode.
The weights for the relative frequency of live, wind, snow and earthquake loads in
proportion to dead load, which are given by Kömürcü and Yücemen (1996)
according to the conditions specific to Turkey, are presented in Table 6.3. These
values also agree with the values proposed by Ellingwood et al. (1980), with minor
differences. The most likely value of reliability index, β, can be found by computing
the value of β for each design situation, which is defined by a set of nominal load
and resistance variables in proportion to dead load, and then by taking its
expectation over the relative frequency distribution given in Table 6.3.
141
Table 6.3 Relative frequency distribution of the ratio of a given load to dead load
(from Kömürcü and Yücemen, 1996)
Y ′ / D′ 0.25 0.50 1.00 2.00 3.00

Y (Load Type)
Live Load 0.10 0.45 0.35 0.10 0.00
Wind Load 0.40 0.30 0.20 0.10 0.00
Snow Load 0.00 0.40 0.30 0.20 0.10
Earthquake Load 0.00 0.25 0.25 0.25 0.25
6.4.1 Reliability Indexes for Reinforced Concrete Beams in the Flexural

Failure Mode
6.4.1.1 Gravity Loads
In this study, the load combination alternatives involving gravity loads are taken as
the sum of dead load and maximum live load (D+L), and dead load and maximum
snow load on roofs (D+S). In TS 500 (2000), the following load combination is
considered:
U = 1.4D′ + 1.6L ′ (6.8)
For the snow load on the roofs, by replacing the live load, L′, with snow load, S',
Eq. (6.8) can be rewritten as follows:
U = 1.4D′ + 1.6S′ (6.9)
Based on the load combination format in TS 500 (2000) and the loads in TS 498
(1997), the reliability indexes, β, for the flexural failure mode of reinforced concrete
beams subjected to D+L combination are given in Table 6.4. In this table, each
142
design situation, which is defined by a set of nominal load and resistance variables
consists of three main design variables: R ′ / D ′ , D ′ / D ′ and L ′ / D′ . In computing
R ′ / D ′ values, the safety criterion (R= 1.4D+1.6L) specified in TS 500 (2000) is
used. In the computations of β values, the relative frequency distribution of L ′ / D′
is taken from Table 6.3.
Table 6.4 Reliability indexes and design situations for the flexural failure mode of
reinforced concrete beams subjected to D+L combination
Expected
R′ D′ L′ value of
β
D′ D′ D′ reliability
index
1.8 1 0.25 2.98
Design situations
2.2 1 0.50 2.99

2.90
3.0 1 1.00 2.84
4.6 1 2.00 2.64
As it can be observed from Table 6.4, the reliability index changes within a range of
2.64 and 2.99 depending on the design situations, i.e. the L ′ / D′ ratios. For this load
combination, the expected value of the reliability index is equal to 2.90. Ellingwood
et al. (1980) obtained a range from 2.6 to 3.2 for the corresponding β with an
expected value of 2.9 based on the design practice in the USA (ACI Code 1977,
1983, 1989). Likewise, Yücemen and El-Etoom (1986) reported that the β values
for this load combination change within a range of 2.6 and 3.09 with an expected
value of 2.7 based on the design practice in Jordan (Arabic Code 1977 was used). In
a similar study, Kömürcü (1995) reported a range of 2.30 and 2.40 for the β values
with an expected value of 2.38 with respect to TS 500 (1984). The value computed
here, 2.90, is consistent with the values reported in these studies.
Snow load is an environmental load which changes from one location to another.
143
For this reason, reliability index, β, is calculated mainly for three different critical
locations, namely, Ankara, Bursa and Canakkale (in Table 6.5, the reliability index
values corresponding to Hakkari are given for the comparison of the values
obtained from these three locations). Besides, while computing the average snow
load reliability indexes for the whole Turkey, Izmir and Antalya are not taken into
consideration since there is no snowfall in Izmir and Antalya. Here, it is important
to note that the nominal snow load values in TS 498 (1997) will be revised in the
near future by the authorities, especially in locations where snowfall is excessive
(personal communication by TSE Standard Preparation Group, 2006). Therefore,
the reliability indexes pertaining to Hakkari are not used in the determination of the
load and resistance partial factors (in Table 6.5, the reliability index values
corresponding to Hakkari are given for the comparison of the values obtained from
three locations). Based on the load combination format given in TS 500 (2000) and
the loads in TS 498 (1997), the reliability indexes for the flexural failure mode of
reinforced concrete beams subjected to dead load and snow load combination are
given in Table 6.5.
reinforced concrete beams subjected to D+S combination
Reliability indexes
R′ D′ L′
Canakkale
Hakkari
Ankara
Turkey
D′ D′
Bursa
D′
2.20 1.00 0.50 4.07 2.76 3.28 1.27 3.39

Design situations
3.00 1.00 1.00 4.08 2.50 3.07 0.68 3.16
4.60 1.00 2.00 3.95 2.32 2.93 0.24 3.01
6.20 1.00 3.00 3.90 2.25 2.88 0.05 2.96
Expected value of reliability 4.03 2.54 3.11 0.76 3.20

indexes
144
As shown in Table 6.5, depending on the design situations, the reliability indexes
are found to range from 3.90 to 4.08; 2.25 to 2.76; 2.88 to 3.28 and 0.05 to 1.27 for
Ankara, Bursa, Canakkale and Hakkari, respectively. Accordingly, for these four
locations, the most likely reliability index values become 4.03, 2.54, 3.11 and 0.76,
respectively. For the whole Turkey, it is observed that the reliability index varies
between 2.96 and 3.39 with a most likely value of 3.20.
From Tables 6.4 and 6.5, it can be seen that while the value of L ′ / D′ or S′ / D ′
increases, the value of reliability index decreases. The relatively higher value of β
corresponding to small ratios of L ′ / D′ or S′ / D′ is because the dead load is the
dominant component. It should be noted that dead load has smaller variability and
higher mean to nominal ratio compared to live and snow loads. For these reason, β
decreases with the increasing values of L ′ / D′ or S′ / D ′ . In Appendix F, the
variations of β values for load combinations considered in this study are displayed
in a graphical form.
6.4.1.2 Gravity and Wind Loads
The load combination alternatives involving gravity and wind loads will be dead
load plus maximum live load plus arbitrary point-in-time wind load (D+L+Wapt),
and dead load plus arbitrary point-in-time live load plus maximum wind load
(D+Lapt+W). In TS 500 (2000), the corresponding load combination is :
U = 1.0D′ + 1.3L ′ + 1.3W ′ (6.10)
Based on the load combination format in TS 500 (2000) and the nominal loads in
TS 498 (1997), the reliability indexes, β, for flexural failure mode of reinforced
concrete beams subjected to D+L+Wapt and D+Lapt+W combinations are given in
Table 6.6 and Table 6.7, respectively.
145
reinforced concrete beams subjected to D+Lapt+W combination
Reliability indexes
R′ D′ L ′apt W′
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
1.98 1.00 0.50 0.25 3.36 3.25 3.18 3.29
2.30 1.00 0.50 0.50 3.63 3.43 3.24 3.48
Design situations
2.95 1.00 0.50 1.00 3.97 3.60 3.17 3.64

4.25 1.00 0.50 2.00 4.05 3.54 2.96 3.55
2.62 1.00 1.00 0.25 3.27 3.47 3.20 3.24
2.95 1.00 1.00 0.50 3.48 3.67 3.34 3.43
3.60 1.00 1.00 1.00 3.81 3.91 3.54 3.72
4.90 1.00 1.00 2.00 4.34 3.87 3.34 3.90
Expected value of reliability indexes 3.59 3.51 3.24 3.45
As shown in Table 6.6, depending on the design situations, the reliability indexes, β,
for the first load combination, i.e., for D+L+Wapt range from 3.27 to 4.34; 3.25 to
3.91; 2.96 to 3.54, with expected values of 3.59, 3.51 and 3.24 for Ankara, Bursa
and Canakkale, respectively. For the whole Turkey, this range is 3.24 to 3.90 with a
most likely value of 3.45.
146
reinforced concrete beams subjected to D+L+Wapt combination
Reliability indexes
R′ D′ L′ Wapt
′
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
1.98 1.00 0.50 0.25 2.42 2.38 2.32 2.40
2.30 1.00 0.50 0.50 3.01 2.94 2.81 2.97
Design situations
2.95 1.00 0.50 1.00 3.85 3.72 3.50 3.78

4.25 1.00 0.50 2.00 4.77 4.55 4.12 4.65
2.62 1.00 1.00 0.25 2.32 2.21 2.17 2.22
2.95 1.00 1.00 0.50 2.66 2.62 2.55 2.64
3.60 1.00 1.00 1.00 3.34 3.26 3.12 3.29
4.90 1.00 1.00 2.00 4.26 4.12 3.86 4.18
From Table 6.7, it can be seen that the reliability indexes, β, for D+L+Wapt
combination, range from 2.32 to 4.77; 2.21 to 4.55 and 2.17 to 4.12 and expected
values are 2.99, 2.90 and 2.78 for Ankara, Bursa and Canakkale, respectively. For
the whole Turkey, reliability index changes within a range of 2.22 and 4.65 with a
6.4.1.3 Gravity and Earthquake Loads
The load combination alternative involving gravity loads and earthquake load will
be D+Lapt+E, which is the most critical load combination in terms of three design
variables: D, L (S or W) and E. According to TS 500 (2000), this load combination
is as follows:
147
U = 1.0D′ + 1.0L ′ + 1.0E ′ (6.11)
While the live load to be used in load combinations is taken from TS 498 (1997) as
stated before, the design earthquake load is obtained from the calculation methods
proposed in Specification for Structures to be Built in Disaster Areas (1998). The
average mean to nominal ratios and total uncertainties associated with earthquake
loads in twelve locations mentioned before are used in the determination of
reliability index for Turkey. In addition, Ankara, Bursa and Canakkale are
considered as critical locations. Based on the load combination format given in TS
500 (2000) and the loads specified in TS 498 (1997) and Specification for
Structures to be Built in Disaster Areas (1998), the reliability indexes for flexural
failure mode of reinforced concrete beams subjected to D+Lapt+E combination are
computed and are given in Table 6.8.
As shown in Table 6.8, the reliability index inherent in Turkish design practice
varies in the range of 0.27 and 1.55; 0.91 and 1.96; 0.76 and 1.88 with a
corresponding most likely values: 0.76, 1.31 and 1.19 for Ankara, Bursa and
Canakkale, respectively for the flexural failure mode of beams. For the whole
Turkey, it is observed that the reliability index varies between 0.93 and 1.98 with a
most likely value of 1.33. From Table 6.8, the variation of β values with L'/D' and
E'/D' ratios can be seen; as observed there is a tendency for the values of reliability
index to decrease, as the values of L'/D' and E'/D' increase. The corresponding plots
are given in Appendix F.
148
reinforced concrete beams subjected to D+Lapt+E combination
Reliability indexes
R′ D′ L′apt E′
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
2.00 1.00 0.50 0.50 1.13 1.60 1.50 1.62
2.50 1.00 0.50 1.00 0.72 1.26 1.14 1.28
Design situations
3.50 1.00 0.50 2.00 0.40 1.01 0.87 1.03

4.50 1.00 0.50 3.00 0.27 0.91 0.76 0.93
2.50 1.00 1.00 0.50 1.55 1.96 1.88 1.98
3.00 1.00 1.00 1.00 1.06 1.55 1.44 1.56
4.00 1.00 1.00 2.00 0.65 1.21 1.09 1.23
5.00 1.00 1.00 3.00 0.46 1.06 0.93 1.09
6.4.2 Reliability Indexes for Reinforced Concrete Beams in the Shear

Failure Mode
As stated before, the load combination alternative involving gravity loads is the sum
of dead load and maximum live load (D+L), and dead load and maximum snow
load on roofs (D+S). According to TS 500 (2000), these load combinations are as
follows:
U = 1.4D′ + 1.6L ′ (6.12)
U = 1.4D′ + 1.6S′ (6.13)
149
Based on the load combination format in TS 500 (2000) and loads in TS 498
(1997), the reliability indexes, β, for the shear failure mode of reinforced concrete
beams subjected to dead load plus live load combination are given in Table 6.9. In
this table, each design situation, which is defined by a set of nominal load and
resistance variables consists of three main design variables: R ′ / D ′ , D ′ / D ′ and
L ′ / D′ . In computing R ′ / D ′ values, the safety criterion (R= 1.4D+1.6L) specified
in TS 500 (2000) is used. In the computations of β values, the relative frequency
distribution of L ′ / D′ is taken from Table 6.3.
reinforced concrete beams subjected to D+L combination
Expected
β
index
1.8 1 0.25 2.36
situations
2.2 1 0.50 2.42

Design
2.41
3.0 1 1.00 2.42
4.6 1 2.00 2.35
As it can be observed in Table 6.9, the reliability index changes within a narrow
range of 2.36 and 2.42 depending on the design situations, i.e. L ′ / D′ ratios. For
this load combination, the most likely reliability index value is equal to 2.41.
Based on the load combination format in TS 500 (2000) and loads in TS 498
(1997), the reliability indexes, β, for the shear failure mode of reinforced concrete
beams subjected to dead load and snow load combination are given in Table 6.10.
150
reinforced concrete beams subjected to D+S combination
Reliability indexes
R′ D′ L′
Canakkale
Hakkari
Ankara
Turkey
D′ D′
Bursa
D′
2.20 1.00 0.50 3.18 2.51 2.98 1.05 2.95

Design situations
3.00 1.00 1.00 3.63 2.35 2.91 0.57 3.05
4.60 1.00 2.00 3.82 2.20 2.79 0.19 2.94
6.20 1.00 3.00 3.79 2.14 2.75 0.03 2.89
Expected value of reliability 3.50 2.36 2.90 0.63 2.97

indexes
As shown in Table 6.10, depending on the design situations, the reliability indexes
are found to range from 3.18 to 3.82; 2.14 to 2.51; 2.75 to 2.98 and 0.03 to 1.05 for
Ankara, Bursa, Canakkale and Hakkari, respectively. Accordingly, for these four
locations, the most likely reliability index values are 3.50, 2.36, 2.90 and 0.63,
respectively. For the whole Turkey, it is observed that the reliability index varies
between 2.89 and 3.05 with a most likely value of 2.97.
From Tables 6.9 and 6.10, it can be seen that while the value of L ′ / D′ or S′ / D ′
increases, the value of reliability index decreases. As it is stated before the
relatively higher value of β corresponding to small ratios of L ′ / D′ or S′ / D′ is
because the dead load which has smaller variability and higher mean to nominal
ratio in comparison with live load and snow load is the dominant component. For
these reason, β, decreases with the increase of the values of L ′ / D′ or S′ / D ′ . The
related figures are given Appendix F.
151
Here, the load combination alternatives involving gravity loads and wind loads will
be “D+L+Wapt” and “D+Lapt+W”. In TS 500 (2000), the following load
combination is provided:
U = 1.0D′ + 1.3L ′ + 1.3W ′ (6.14)
Based on the load combination format in TS 500 (2000) and snow loads in TS 498
(1997), the values of the reliability index for the shear failure mode of reinforced
concrete beams subjected to D+Lapt+W and D+L+Wapt are given in Table 6.11 and
6.12, respectively
reinforced concrete beams subjected to D+Lapt+W combination
Reliability indexes
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
1.98 1.00 0.50 0.25 2.68 2.62 2.52 2.63

2.30 1.00 0.50 0.50 2.87 2.76 2.57 2.77
Design situations
2.95 1.00 0.50 1.00 3.15 2.94 2.63 2.96

4.25 1.00 0.50 2.00 3.38 3.06 2.60 3.08
2.62 1.00 1.00 0.25 3.03 2.99 2.94 3.00
2.95 1.00 1.00 0.50 3.20 3.13 3.01 3.14
3.60 1.00 1.00 1.00 3.43 3.27 3.03 3.29
4.90 1.00 1.00 2.00 3.60 3.32 2.91 3.34
152
As shown in Table 6.11, depending on the design situations, the reliability indexes,
β, range from 2.68 to 3.60; 2.62 to 3.32; 2.52 to 3.03, and expected values are 3.04,
2.93 and 2.75 for Ankara, Bursa and Canakkale, respectively. For the whole
Turkey, it is observed that the reliability index varies between 2.63 and 3.34 with a
reinforced concrete beams subjected to D+L+Wapt combination
Reliability indexes
R′ D′ L′ Wapt
′
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
1.98 1.00 0.50 0.25 1.97 1.94 1.89 1.95
2.30 1.00 0.50 0.50 2.42 2.37 2.27 2.39
Design situations
2.95 1.00 0.50 1.00 3.06 2.96 2.79 3.00

4.25 1.00 0.50 2.00 3.74 3.60 3.33 3.66
2.62 1.00 1.00 0.25 1.91 1.89 1.85 1.90
2.95 1.00 1.00 0.50 2.27 2.22 2.16 2.24
3.60 1.00 1.00 1.00 2.80 2.73 2.61 2.76
4.90 1.00 1.00 2.00 3.48 3.36 3.15 3.41
In Table 6.11, for D+Lapt+W combination, β increases as W ' / D' increases up to
153
W ' / D' =1 and after this ratio β values do not change significantly. In Table 6.12, for
D+L+Wapt combination, β values decrease while W ' / D' and L ′ / D′ increase. In
Appendix F, β values for D+Lapt+W and D+L+Wapt load combinations are displayed
in a graphical form.
Here, the load combination alternative involving gravity loads and earthquake loads
will be D+Lapt+E, which is the most critical load combination in view of three load
variables: D, L (S or W) and E. According to TS 500 (2000), this load combination
is as follows:
U = 1.0D′ + 1.0L ′ + 1.0E ′ (6.15)
While the live load to be used in load combinations is taken from TS 498 (1997) as
stated before, the design earthquake load is obtained utilizing the calculation
methods proposed in “Specification for Structures to be Built in Disaster Areas
(1998)”. The average mean to nominal ratios and total uncertainties associated with
earthquake load in these 12 locations mentioned before are used in the
determination of reliability index for Turkey. Additionally, Ankara, Bursa and
Canakkale are considered as critical locations.
The reliability indexes and design situations for the shear failure mode of reinforced
concrete beams subjected to D+Lapt+E combination are given in Table 6.13. In this
table, each design situation, which is defined by a set of nominal load and resistance
variables consists of four main design variables: R ′ / D ′ , D ′ / D ′ , L ′ / D′ and E ′ / D ′ .
In computing R ′ / D ′ values, the safety criterion (R= 1.0D+1.0L+1.0W) specified in
TS 500 (2000) is used. In the computations of β values, the relative frequency
distribution of L ′ / D′ is taken from Table 6.3.
154
From Table 6.13, the reliability index inherent in Turkish design practice varies
between 0.26 and 1.49; 0.93 and 1.93; 0.74 and 1.82 with the corresponding most
likely values: 0.73, 1.30 and 1.15 for Ankara, Bursa and Canakkale in connection
with the shear failure mode of beams. For Turkey, reliability index ranges from 0.91
to 1.92 with an expected value of 1.29.
reinforced concrete beams subjected to D+Lapt+E combination
Reliability indexes
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
2.00 1.00 0.50 0.50 1.07 1.56 1.43 1.55
2.50 1.00 0.50 1.00 0.69 1.26 1.10 1.24
Design situations
3.50 1.00 0.50 2.00 0.39 1.02 0.85 1.01

4.50 1.00 0.50 3.00 0.26 0.93 0.74 0.91
2.50 1.00 1.00 0.50 1.49 1.93 1.82 1.92
3.00 1.00 1.00 1.00 1.02 1.54 1.40 1.53
4.00 1.00 1.00 2.00 0.63 1.22 1.06 1.20
5.00 1.00 1.00 3.00 0.45 1.08 0.91 1.06

Combined Action of Flexure and Axial Load Failure Mode
Based on the load combination format in TS 500 (2000) and loads specified in TS
498 (1997), the reliability indexes, β, for reinforced concrete columns in combined
155
action of flexure and axial load subjected to D+L and D+S load combinations are
given in Tables 6.14 and 6.15, respectively.
Table 6.14 Reliability indexes and design situations for reinforced concrete columns
in the combined action of flexure and axial load subjected to D+L
Expected
β
index
1.8 1 0.25 2.80
Design situations
2.2 1 0.50 2.84

2.78
3.0 1 1.00 2.74
4.6 1 2.00 2.58
in the combined action of flexure and axial load subjected to D+S
Reliability indexes
R′ D′ L′
Canakkale
Hakkari
Ankara
Turkey
D′ D′
Bursa
D′
2.20 1.00 0.50 3.81 2.67 3.12 1.17 3.35

Design situations
3.00 1.00 1.00 4.01 2.42 2.96 0.63 3.14
4.60 1.00 2.00 3.89 2.25 2.83 0.21 2.99
6.20 1.00 3.00 3.84 2.18 2.78 0.03 2.94
Expected value of reliability

3.89 2.46 2.98 0.67 3.17
indexes
As shown in Table 6.15, the reliability indexes range from 3.81 to 4.01; 2.18 to
2.67; 2.78 to 3.12 and 0.03 to 1.17 for Ankara, Bursa, Canakkale and Hakkari,
156
respectively. Accordingly, for these four locations, the most likely reliability index
values are 3.89, 2.46, 2.98 and 0.67, respectively. For the whole Turkey, it is
observed that the reliability index varies between 2.94 and 3.35 with a most likely
value of 3.17.
The reliability indexes and design situations for the flexural failure mode of
reinforced concrete beams subjected to D+Lapt+W combination and D+L+Wapt are
given in Table 6.16 and Table 6.17, respectively. As it can be shown in Table 6.16,
depending on the design situations, the reliability index, β, changes within a range
of 3.12 and 4.11; 3.03 and 3.73; 2.92 and 3.42, and expected values are 3.50, 3.38
and 3.13 for Ankara, Bursa and Canakkale, respectively. For the whole Turkey, the
reliability index varies between 3.04 and 3.76 with an expected value of 3.36.
in the combined action of flexure and axial load (for D+Lapt+W)
Reliability indexes
Canakkale
Ankara
D′ D′ D′ Turkey
Bursa
D′
1.98 1.00 0.50 0.25 3.12 3.03 2.92 3.04

2.30 1.00 0.50 0.50 3.39 3.24 3.02 3.25
Design situations
2.95 1.00 0.50 1.00 3.71 3.43 3.02 3.45

4.25 1.00 0.50 2.00 3.87 3.43 2.86 3.46
2.62 1.00 1.00 0.25 3.40 3.44 3.29 3.36
2.95 1.00 1.00 0.50 3.62 3.53 3.39 3.54
3.60 1.00 1.00 1.00 3.92 3.73 3.42 3.75
4.90 1.00 1.00 2.00 4.11 3.73 3.21 3.76
157
in the combined action of flexure and axial load (for D+L+Wapt)
Reliability indexes
R′ D′ L′ Wapt
′
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
1.98 1.00 0.50 0.25 2.28 3.26 2.20 2.28
2.30 1.00 0.50 0.50 2.81 2.78 2.66 2.81
Design situations
2.95 1.00 0.50 1.00 3.56 3.51 3.30 3.56

4.25 1.00 0.50 2.00 4.36 4.28 3.91 4.36
2.62 1.00 1.00 0.25 2.14 2.13 2.09 2.14
2.95 1.00 1.00 0.50 2.54 2.52 2.45 2.54
3.60 1.00 1.00 1.00 3.16 3.13 2.99 3.16
4.90 1.00 1.00 2.00 3.98 3.92 3.67 3.98
For reinforced concrete columns in the combined action of flexure and axial load,
subjected to D+Lapt+E load combination, the reliability index inherent in the
Turkish design practice varies between 0.27 and 1.53; 0.94 and 1.98; 0.76 and 1.86
with the corresponding most likely values: 0.76, 1.33 and 1.18 for Ankara, Bursa
and Canakkale, respectively. Reliability index changes within a range of 0.92 and
158
1.97 with a most likely value of 1.32 for the whole Turkey. The design situations
and corresponding reliability index values belonging to these locations are
summarized in Table 6.18. In this table, each design situation, which is defined by a
set of nominal load and resistance variables consists of four main design variables:
R ′ / D ′ , D ′ / D ′ , L ′ / D′ and E ′ / D ′ . In computing R ′ / D ′ values, the safety criterion
(R= 1.0D+1.0L+1.0E) specified in TS 500 (2000) is used. In the computations of β
values, the relative frequency distribution of L ′ / D′ is taken from Table 6.3.
in the combined action of flexure and axial load (for D+Lapt+E)
Reliability indexes
Canakkale
Ankara
Turkey
D′ D′ D′ D′
Bursa
2.00 1.00 0.50 0.50 1.12 1.62 1.49 1.60
2.50 1.00 0.50 1.00 0.71 1.29 1.13 1.27
Design situations
3.50 1.00 0.50 2.00 0.40 1.05 0.87 1.03

4.50 1.00 0.50 3.00 0.27 0.94 0.76 0.92
2.50 1.00 1.00 0.50 1.53 1.98 1.86 1.97
3.00 1.00 1.00 1.00 1.05 1.57 1.43 1.56
4.00 1.00 1.00 2.00 0.65 1.24 1.08 1.23
5.00 1.00 1.00 3.00 0.46 1.10 0.92 1.08
6.4.4 Reliability Indexes for Reinforced Concrete Columns in the Shear

Failure Mode
The mean to nominal ratio and the total uncertainty of the shear failure mode of
columns are the same as those computed for the shear failure mode of beams.
159
Furthermore, the load statistics shown in Table 6.1 will be used for all failure modes
and all structural members. As a result, the reliability indexes to be computed for
the shear failure mode of columns will be the same as those found for the shear
failure mode of beams. Accordingly, the reliability indexes reported in Section 6.4.2
are exactly the same in this section. However, for the sake of completeness, the
results are duplicated in a summary form in Table 6.19.
Table 6.19 Reliability indexes for reinforced concrete columns in the shear failure
mode
Ankara Bursa Canakale Turkey

Locations
Most likely
Most likely
Most likely
Most likely
Range
Range
Range
Range
value
value
value
value
Load
combinations
D+L 2.36-2.42 2.41
D+S 3.18-3.82 3.50 2.14-2.51 2.36 2.75-2.98 2.90 2.89-3.05 2.97
D+Lapt+W 2.68-3.60 3.04 2.62-3.32 2.93 2.52-3.03 2.75 2.03-3.34 2.94
D+L+Wapt 1.91-3.74 2.44 1.89-3.60 2.38 1.85-3.33 2.28 1.90-3.66 2.40
D+Lapt+E 0.26-1.49 0.73 0.93-1.93 1.30 0.74-1.82 1.15 0.91-1.92 1.29

Flexural Failure Mode
The mean to nominal ratio and the total uncertainty of shear walls in the flexural
failure mode are the same as those computed for the flexural failure mode of beams.
Furthermore, the load statistics shown in Table 6.1 will be used for all failure modes
and all structural members. For these reasons, the reliability indexes to be computed
for the flexural failure mode of shear walls will be the same as those found for the
flexural failure mode of beams. Accordingly, the reliability indexes reported in
Section 6.4.1 are exactly the same in this section, too. A summary of results of
160
reliability indexes is given in Table 6.20.
Table 6.20 Reliability indexes of reinforced concrete shear walls in the flexural
failure mode
Location
Most likely
Most likely
Most likely
Most likely
Range
Range
Range
Range
value
value
value
value
Load
combination
D+L 2.64-2.99 2.90
D+S 3.90-4.08 4.03 2.25-2.76 2.54 2.88-3.28 3.11 2.96-3.99 3.20
D+Lapt+W 3.27-4.34 3.59 3.25-3.91 3.51 2.96-3.54 3.24 3.24-3.90 3.45
D+L+Wapt 2.52-4.77 2.99 2.21-4.55 2.90 2.17-4.12 2.78 2.22-4.05 2.73
D+Lapt+E 0.27-1.55 0.76 0.91-1.96 1.31 0.76-1.88 1.19 0.93-1.98 1.33
6.4.6 Reliability Indexes for Reinforced Concrete Shear Walls in the Shear
Failure Mode
The values to be computed in this section are the same as those reported in Section
6.4.3. In other words, the reliability indexes computed for reinforced concrete
columns in combined action and flexure are also the same for the reinforced
concrete shear walls failing in the shear failure mode. However, for the sake of
completeness, the previous results are repeated in a summary form in Table 6.21.
161
Table 6.21 Reliability indexes of reinforced concrete shear walls in the shear failure
mode
Locations
Most likely
Most likely
Most likely
Most likely
Range
Range
Range
Range
value
value
value
value
Load
combinations
D+L 2.58-2.84 2.78
D+S 3.81-4.01 3.89 2.18-2.67 2.46 2.78-3.12 2.98 2.94-3.35 3.17
D+Lapt+W 3.12-4.11 3.50 3.03-3.73 3.38 2.52-3.42 3.13 3.04-3.76 3.36
D+L+Wapt 2.14-4.36 2.79 2.13-4.28 2.76 2.09-3.91 2.64 2.14-4.36 2.79
D+Lapt+E 0.27-1.53 0.76 0.94-1.98 1.33 0.76-1.86 1.18 0.92-1.97 1.32
6.4.7 Assessment of Target Reliability Indexes
In general, basic requirements for the safety of a structure can be expressed in terms
of either the accepted minimum reliability index or accepted maximum failure
probability. From a reliability based safety assessment point of view, the acceptable
risk level can be quantified by specifying the value of the reliability index known as
target reliability index, βT.
After the determination of the safety levels inherent in the existing standards that
have resulted in acceptable performance earlier, the optimum value of target
reliability index can be assessed based on two factors: cost of safety and
consequences of failure (Kömürcü and Yücemen, 1996; Melchers, 2002 and Nowak
and Szerszen 2003b). Ellingwood et al. (1980), Kömürcü (1995), and Nowak and
Szerszen (2003b) first calculated the safety levels and then proposed a set of target
reliability indexes for both different failure modes of structural members and
different load combinations. These safety levels and selected target reliabilities are
summarized in Table 6.22. In the following paragraphs, considering also these
162
earlier studies, the target reliability indexes for the studied failure modes of
reinforced concrete beams, columns and shear walls will be determined consistent
with the safety level inherent in the current standards.
Table 6.22 Reliability indexes corresponding to the current design practice and the
target reliability indexes for different load combinations and different structural
members according to different studies
Failure Load
Reference βcurrent βT
mode Combination
D′ + L′ 2.62-3.58 3.0
D′ + S ′ 3.33/3.08 3.0
D ′ + Lapt
′ +W ′ 2.74/2.50/3.90/3.24/2.90 2.5
Ellingwood et al.
Beam flexural capacity
(1980) D ′ + L ′ + Wapt
′ 2.98/3.34/4.55/2.78/3.76 2.5
D′ + L′ + E ′ 1.75
D′ + L′ 2.38 2.7
D′ + S ′ 2.97/2.96/2.30 2.7
D ′ + Lapt
′ +W ′ 3.19/3.17/2.93 2.5
Kömürcü(1995)
D ′ + L ′ + Wapt
′ 3.17/3.14/3.16 2.5
D′ + L′ + E ′ 0.98/1.22/1.04 1.75
Nowak and
D′ + L′ 3.83 3.5
Szerszen (2003b)
D′ + L′ 1.99-2.45 3.0
Beam shear capacity
D′ + S ′ 3.0
Ellingwood et al. D ′ + Lapt
′ +W ′ 2.5
(1980) D ′ + L ′ + Wapt
′ 1.97/2.32/2.38 2.5
D′ + L′ + E ′ 1.75
Nowak and
D′ + L′ 3.78 3.5
Szerszen (2003b)
D′ + L′ 2.98-3.49 3.0
D′ + S ′ 3.0
Column axial
D ′ + Lapt
′ +W ′ 2.5
capacity
Ellingwood et al.
(1980) D ′ + L ′ + Wapt
′ 2.74/2.96 2.5
D′ + L′ + E ′ 1.75
Nowak and
D′ + L′ 4.68 4.0
Szerszen (2003b)
163
As listed in Table 6.22, when the flexural failure mode of a beam is taken into
consideration, Ellingwood et al. (1980) proposed the target reliability index to be
3.0 for gravity loads, 2.5 for gravity loads combined with wind load and 1.75 for
gravity loads combined with earthquake load. For the same load combinations,
Kömürcü (1985) selected the target reliability index as 2.7, 2.5, 1.75. In another
study, Nowak and Szerszen (2003b) proposed this index to be 3.5 for gravity loads.
While the reliability indexes which are computed in this study based on Turkish
Codes are higher than those reported by Ellingwood et al. (1980) and Kömürcü
(1995), especially for the D+L+W combination, one can see that they are smaller
than the values computed by Nowak and Szerszen (2003b). However, it should also
be noted that the study of Nowak and Szerszen (2003b) was conducted only for the
D+L combination (1.2D+1.6L). On the other hand, the coefficient of variations of
the variables used in the study of Nowak and Szerszen (2003b) are usually small in
comparison with those used in this study. For example, the c.o.v.’s of concrete
compressive strength and maximum live load are equal to 0.10 and 0.18,
respectively, while these values are taken as 0.18 and 0.27, respectively in this
study.
Considering the studies mentioned above (Table 6.22) and the safety levels inherent
in the Turkish current design practice, a set of target reliabilities are selected as
given in Table 6.23 for beams in the flexural failure mode.
164
Table 6.23 Current and the target reliability indexes for different load combinations
considering beams in the flexural failure mode
Reliability Indexes Values

Load Combination Location
β β βT
(According to TS 500)
1.4 D ′ + 1.6 L ′ Turkey 2.64-2.99 2.90 3.0
Ankara 3.90-4.07 4.03
Bursa 2.25-2.76 2.54
1.4 D ′ + 1.6 S ′
Canakkale 2.88-3.28 3.11
Turkey 2.96-3.39 3.20 3.0
Ankara 3.27-4.34 3.59
Bursa 3.25-3.91 3.51
1.0D′ + 1.3L ′apt + 1.3W ′
Canakkale 3.17-3.54 3.24
Turkey 3.24-3.90 3.45 3.0
Ankara 2.32-4.26 2.99
Bursa 2.21-4.55 2.90
1.0D′ + 1.3L ′ + 1.3Wapt
′ Canakkale 2.17-4.12 2.78
Turkey 2.22-4.65 2.93 2.7
Ankara 0.27-1.55 0.76
Bursa 0.91-1.96 1.31
1.0D′ + 1.0L′apt + 1.0E ′ Canakkale 0.76-1.88 1.19
Turkey 0.93-1.98 1.33 1.75
165
As far as the shear failure mode is considered for beams, Ellingwood et al. (1980)
proposed the target reliability index to be 3.0 for D+L and D+S combination, 2.5 for
D+L+W combination and 1.75 for D+L+E combination whereas Nowak and
Szerszen (2003b) proposed the target reliability index to be 3.5 for D+L
combination. The target reliabilities selected by considering these studies and the
safety levels inherent in the Turkish design practice are tabulated in Table 6.24
considering beams in the shear failure mode
Reliability Index Values

Load Combination
Location
(According to TS 500) β β βT
1.4 D ′ + 1.6 L ′ Turkey 2.36-2.42 2.41 3.0

Ankara 3.18-3.82 3.50
Bursa 2.14-2.51 2.36
1.4 D ′ + 1.6 S ′
Canakkale 2.75-2.98 2.90
Turkey 2.89-3.05 2.97 3.0
Ankara 2.68-3.60 3.04
Bursa 2.62-3.32 2.93
1.0D ′ + 1.3L ′apt + 1.3W ′
Canakkale 2.52-3.03 2.75
Turkey 2.63-3.34 2.94 3.0
Ankara 1.91-3.74 2.44
Bursa 1.89-3.60 2.38
1.0D′ + 1.3L′ + 1.3Wapt
′ Canakkale 1.85-3.33 2.28
Turkey 1.90-3.66 2.40 2.7
Ankara 0.26-1.49 0.73
Bursa 0.93-1.93 1.30
1.0D′ + 1.0L ′apt + 1.0E ′ Canakkale 0.74-1.82 1.15
Turkey 0.91-1.92 1.29 1.75
166
For reinforced concrete columns, the target reliability values proposed by
Ellingwood et al. (1980) are the same with those proposed for the flexural and shear
failure modes of beams. Nowak and Szerszen (2003b) proposed the target
reliability index as 4.0 for D+L combination. Failure of one column in a critical
region can lead to the subsequent collapse of the adjoining floors yielding to total
collapse of the entire structure. Accordingly, the target reliabilities to be used for
columns should be higher than those selected for beams. Taking the studies
mentioned above and the safety levels inherent in the Turkish design practice into
account, the proposed target reliabilities are listed in Table 6.25.
considering columns in the combined action failure mode
Load Combination Location Reliability Index Values

β β βT
1.4 D ′ + 1.6 L ′ Turkey 2.58-2.84 2.78 3.2

Ankara 3.81-4.01 3.89
1.4 D ′ + 1.6 S ′ Bursa 2.18-2.67 2.46
Canakkale 2.78-3.12 2.98
Turkey 2.94-3.35 3.17 3.2
Ankara 3.12-4.11 3.50
1.0D′ + 1.3L ′apt + 1.3W ′ Bursa 3.03-3.73 3.38
Canakkale 2.92-3.42 3.13
Turkey 3.04-3.76 3.36 3.2
Ankara 2.14-4.36 2.79
Bursa 2.13-4.28 2.76
1.0D′ + 1.3L′ + 1.3Wapt
′ Canakkale 2.09-3.91 2.64
Turkey 2.14-4.36 2.79 3.0
Ankara 0.27-1.53 0.76
Bursa 0.94-1.98 1.33
Turkey 0.92-1.97 1.32 1.75
167
Structural column failure is of main concern with respect to not only economic but
also human losses. Hence, great care should be given during the design of a column
with a higher reserve strength compared to beams and other horizontal structural
elements, especially since failure provides little visual warning (Nawy, 2005).
Accordingly, the target reliabilities to be used for columns should be higher than
those selected for beams and other horizontal structural elements. Considering the
safety levels inherent in the Turkish design practice, the selected target reliabilities
are listed in Table 6.26
considering columns in the shear failure mode

Load Combination
Location
1.4 D ′ + 1.6 L ′ Turkey 2.36-2.42 2.41 3.0

Ankara 3.18-3.82 3.50
Bursa 2.14-2.51 2.36
1.4 D ′ + 1.6 S ′
Canakkale 2.75-2.98 2.90
Turkey 2.89-3.05 2.97 3.2
Ankara 2.68-3.60 3.04
Bursa 2.62-3.32 2.93
1.0D ′ + 1.3L′apt + 1.3W ′
Canakkale 2.52-3.03 2.75
Turkey 2.63-3.34 2.94 3.2
Ankara 1.91-3.74 2.44
Bursa 1.89-3.60 2.38
1.0D′ + 1.3L′ + 1.3Wapt
′ Canakkale 1.85-3.33 2.28
Turkey 1.90-3.66 2.40 3.0
Ankara 0.26-1.49 0.73
Bursa 0.93-1.93 1.30
Turkey 0.91-1.92 1.29 1.75
168
As it is already mentioned, shear walls behave like fixed supported beams, and the
design of shear walls can be carried out in the same way as beams (Park and Paulay,
1975; Celep and Kumbasar, 2005). On the other hand, the values of safety levels in
shear walls (Table 6.26 and 6.27) are higher than those computed for beams (Table
6.22 and Table 6.23). In view of these considerations, the target reliabilities are
selected considering the existing safety levels in the Turkish design practice and the
selected target reliabilities are listed in Tables 6.27 and 6.28.
considering shear walls in the flexural failure mode

Load Combination
Location
1.4 D ′ + 1.6 L ′ Turkey 2.64-2.99 2.90 3.0

Ankara 3.90-4.07 4.03
Bursa 2.25-2.76 2.54
1.4 D ′ + 1.6 S ′
Canakkale 2.88-3.28 3.11
Turkey 2.96-3.39 3.20 3.2
Ankara 3.27-4.34 3.59
Bursa 3.25-3.91 3.51
1.0D′ + 1.3L ′apt + 1.3W ′
Canakkale 3.17-3.54 3.24
Turkey 3.24-3.90 3.45 3.2
Ankara 2.32-4.26 2.99
Bursa 2.21-4.55 2.90
1.0D′ + 1.3L ′ + 1.3Wapt
′ Canakkale 2.17-4.12 2.78
Turkey 2.22-4.65 2.93 2.7
Ankara 0.27-1.55 0.76
Bursa 0.91-1.96 1.31
Turkey 0.93-1.98 1.33 1.75
169
considering shear walls in the shear failure mode
Load Combination Location Reliability Index Values

β β βT
1.4 D ′ + 1.6 L ′ Turkey 2.58-2.84 2.78 3.2

Ankara 3.81-4.01 3.89
Bursa 2.18-2.67 2.46
1.4 D ′ + 1.6 S ′
Canakkale 2.78-3.12 2.98
Turkey 2.94-3.35 3.17 3.2
Ankara 3.12-4.11 3.50
Bursa 3.03-3.73 3.38
1.0D′ + 1.3L ′apt + 1.3W ′
Canakkale 2.92-3.42 3.13
Turkey 3.04-3.76 3.36 3.2
Ankara 2.14-4.36 2.79
Bursa 2.13-4.28 2.76
1.0D′ + 1.3L ′ + 1.3Wapt
′ Canakkale 2.09-3.91 2.64
Turkey 2.14-4.36 2.79 3.0
Ankara 0.27-1.53 0.76
Bursa 0.94-1.98 1.33
Turkey 0.92-1.97 1.32 1.75
170
CHAPTER 7
SELECTION OF LOAD AND RESISTANCE FACTORS
7.1 INTRODUCTION
In this chapter, based on the statistical analysis carried out in the previous chapters,
the resistance and load factors for different load combinations will be developed by
using the Advanced First Order Second Moment (AFOSM) method for reinforced
concrete beams, columns and shear walls in the flexure and shear failure modes and
also for columns subjected to the combined action of flexure and axial load.
The required information and data needed for the computations of the resistance and
load factors are presented in the previous chapters of this dissertation. In Chapter 3,
the parameters which influence the capacity of structural members are described in
a detailed way and the various sources of uncertainties are quantified using the local
and international data and information. In Chapter 4, using these values, the mean to
nominal ratios and the total uncertainties of resistances for different failure modes
of three basic types of reinforced concrete structural members, namely beams
columns and shear walls, are computed. In Chapter 5, structural load effects
resulting from dead, live, wind, snow and earthquake loads are evaluated. In
Chapter 6, reliability based resistance criterion for the failure modes of reinforced
concrete structural members are examined within the scope of the previous chapters
and the target reliability indexes are selected. These target reliability indexes will be
used in the AFOSM computations associated with the new load and resistance
171
factors, which will be proposed at the end of this chapter.
7.2 LOAD AND RESISTANCE FACTORS CORRESPONDING TO THE

SELECTED TARGET RELIABILITIES
In this section, the load and resistance factors corresponding to target reliabilities
are determined for different design situations, i.e., the situations for L'/D', S'/D',
W'/D' and E'/D' values according to D+L, D+S, D+Lapt+W, D+L+Wapt and
D+Lapt+E load combinations. In Figures 7.1 to 7.36, the values of the load and
resistance factors computed for different load combinations considered in this study
are plotted for different failure modes of beams, columns and shear walls.
As stated before, snow, wind and earthquake loads for the whole Turkey include the
mean values of loads for some representative locations, namely Ankara, Izmir,
Antalya, Bursa, Gaziantep, Samsun, Malatya, Erzincan, Canakkale, Hakkari,
Göztepe/Istanbul and Sile/Istanbul. The critical locations are taken as Bursa,
Canakkale and Ankara for snow, wind and earthquake loads, respectively.
Correspondingly, the figures are plotted for both each critical location and for the
whole Turkey for each load combination.
These figures show that the resistance factors are rather insensitive to the time-
varying loads in the load combinations, when these loads are small. For all of the
studied failure modes of the structural members, the dead load factor, γD, is
observed to be about 1.1. The reason why these values are the same is due to the
fact that the variability in dead load is small in comparison to the other time-varying
loads. These figures also indicate that the live load factor in D+L+W and D+L+E
load combinations is quite small. This is because the value of Lapt is much less than
the nominal live load, L'. If the different resistance statistics of failure modes of
structural members are considered, load factors do not appear to be sensitive to the
resistance statistics. It should be noted that while the resistance factors depend on
172
the material and the limit state of interest, the load factors are independent of these
considerations. On the other hand, the different load and resistance factors are
observed for each failure mode of each structural member due to the fact that the
target reliability as well as resistance statistics are different. Besides, the
observations indicate that the load factors for the time-varying loads in the load
combination increases as the proportion of that load to dead load, such as L'/D',
S'/D' and E'/D', increases. The computed load and resistance factors are given in a
graphical form in Sections 7.2.1 to 7.2.6.
7.2.1 Load and Resistance Factors for Reinforced Concrete Beams in the
Flexural Failure Mode
Based on the load combination format mentioned in Section 6.2 and the nominal
loads in TS 498 (1997) and “Specification for Structures to be Built in Earthquake
Areas (2006)”, the load and resistance factors for reinforced concrete (RC) beams in
the flexural failure mode are computed considering different load ratios. The
variations are shown in a graphical form in Figures 7.1 to 7.9.
Load and Resistance Factors
1.9
1.7
Resistance
1.5 factor
1.3 Dead load
factor
1.1 Live load
0.9 factor
0.7
0.25 0.5 1 2
L'/D'
Figure 7.1 Variation of the load and resistance factors for RC beams in the flexural
failure mode (Turkey; D+L; βT=3.0)
173
3.4
3.2

3
2.8
2.6
Resistance
2.4
factor
2.2
Dead load
2 factor
1.8 Snow load
1.6 factor
1.4
1.2
1
0.8
0.5 1 2 3
S'/D'
failure mode (Bursa; D+S; βT=3.0)
2.2
2
1.8
1.6 Resistance
factor
1.4 Dead load
factor
1.2 Snow load
factor
1
0.8
0.5 1 2 3
S'/D'
failure mode (Turkey; D+S; βT =3.0)
174
1.6
1.5
1.4
1.3
1.2 Resistance factor

(L/D=0.5)
Resistance factor
1.1 (L/D=1)
Dead load factor
(L/D=0.5)
1
Dead load factor

(L/D=1)
0.9 Live load factor
(L/D=0.5)
Live load factor
0.8 (L/D=1)
Wind load factor
(L/D=0.5)
0.7
Wind load factor
(L/D=1)
0.6
0.5
0.4
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
failure mode (Canakkale; D+Lapt+W; βT =3.0)
175
1.2
1.1

(L/D=0.5)
Resistance factor
(L/D=1)
0.8
Dead load factor
(L/D=0.5)
Dead load factor

0.7 (L/D=1)
Live load factor
(L/D=0.5)

(L/D=1)
Wind load factor
(L/D=0.5)
0.5 Wind load factor
(L/D=1)
0.4
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
failure mode (Turkey; D+Lapt+W; βT =3.0)
176
1.8
1.7
(L/D=0.5)
1.5

1.3 (L/D=1)
1.2 Dead load f actor
1.1 (L/D=0.5)
1 Dead load f actor
0.9 (L/D=1)
0.8 Live load f actor
0.7 (L/D=0.5)
0.6
Live load factor
0.5 (L/D=1)
0.4
Wind load factor
0.3
(L/D=0.5)
0.2
0.1 Wind load f actor
(L/D=1)
0
0.25 0.5 1 2
W'/D'
failure mode (Canakkale; D+L+Wapt; βT =2.7)
1.8
1.7
1.5 (L/D=0.5)

1.3 (L/D=1)
1.2 Dead load factor
1.1 (L/D=0.5)
1
Dead load factor
0.9 (L/D=1)
0.8
Live load
0.7
f actor(L/D=0.5)
0.6
0.5 Live load
f actor(L/D=1)
0.4
0.2 (L/D=0.5)
0 (L/D=1)
0.25 0.5 1 2
W'/D'
failure mode (Turkey; D+L+Wapt; βT =2.7)
177
3.8
3.6
3.2 (L/D=0.5)

3
Resistance factor
2.8
(L/D=1)
2.6
2.2 (L/D=0.5)
2 Dead load factor
1.8 (L/D=1)
1.4 (L/D=0.5)
1 (L/D=1)
0.8
Earthquake load
0.6
f actor (L/D=0.5)
0.4
0.2 Earthquake load
f actor (L/D=1)
0
0.5 1 1.5 2 2.5 3
E'/D'
failure mode (Ankara; D+Lapt+E; βT =1.75)

(L/D=0.5)
2.4
Resistance factor
2.2 (L/D=1)
2 Dead load factor

(L/D=0.5)
1.8
Dead load factor
1.6 (L/D=1)

(L/D=0.5)
1.2
Live load factor
1 (L/D=1)
0.8 Earthquake load
factor (L/D=0.5)
0.6
Earthquake load
0.4 factor (L/D=1)
0.2
0
0.5 1 1.5 2 2.5 3
E'/D'
failure mode (Turkey; D+Lapt+E; βT =1.75)
178
7.2.2 Load and Resistance Factors for Reinforced Concrete Beams in the
Shear Failure Mode
In this section, the load and resistance factors are computed for reinforced concrete
beams in the shear failure mode, on the basis of load combination format mentioned
in Section 6.2 and the nominal loads in TS 498 (1997) and Specification for
Structures to be Built in Earthquake Areas (2006). The corresponding plots are
given in Figures 7.10 to 7.18.
1.8
1.6
1.4 Resistance
factor
Dead load
1.2 factor
Live load
factor
1
0.8
0.6
0.25 0.5 1 2
L'/D'
Figure 7.10 Variation of the load and resistance factors for RC beams in the shear
failure mode (Turkey; D+L; βT =3.0)
179
3.4
3.2

3
2.8
2.6
2.4 Resistance
2.2 factor
2 Dead load
factor
1.8
1.6 Snow load
factor
1.4
1.2
1
0.8
0.5 1 2 3
S'/D'
failure mode (Bursa; D+S; βT =3.0)
2.2
2
1.8
Resistance
1.6 factor
Dead load
factor
1.4
Snow load
factor
1.2
0.8
0.5 1 2 3
S'/D'
180
1.3
1.2
1.1

(L/D=0.5)
Resistance factor
0.8 (L/D=1)
Dead load factor
(L/D=0.5)
(L/D=1)
Live load factor
0.6 (L/D=0.5)
Live load factor
(L/D=1)
(L/D=0.5)
Wind load factor
0.4 (L/D=1)
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
181
1.2
1.1
0.9
Resistance factor
(L/D=0.5)
0.8
Resistance factor
(L/D=1)
Dead load factor
0.7
(L/D=0.5)
Dead load factor
(L/D=1)
(L/D=0.5)
Live load factor
0.5 (L/D=1)
Wind load factor
(L/D=0.5)
(L/D=1)
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
182
1.5
(L/D=0.5)
1.3
Resistance factor
1.2

(L/D=1)
1.1
Dead load factor
1 (L/D=0.5)
0.8 (L/D=1)
0.6 (L/D=0.5)

(L/D=1)
0.4
Wind load factor
0.3 (L/D=0.5)
0.1 (L/D=1)
0
0.25 0.5 1 2
W'/D'
1.5
1.4
Resistance factor
1.3 (L/D=0.5)

1.1 (L/D=1)
1 Dead load f actor
0.9 (L/D=0.5)
(L/D=1)
0.7
(L/D=0.5)
0.5
L ive load factor
0.4
(L/D=1)
0.3
Wind load factor
0.2 (L/D=0.5)
0 (L/D=1)
0.25 0.5 1 2
W'/D'
183
3.8
3.6
3.2 (L/D=0.5)
3 Resistance factor

2.8 (L/D=1)
2.4 (L/D=0.5)
2 (L/D=1)
1.6 (L/D=0.5)
1.2 (L/D=1)
1 Earthquake load
0.8 f actor (L/D=0.5)
0.6 Earthquake load
0.4 f actor (L/D=1)
0.2
0
0.5 1 2 3
E'/D'
2.6
Resistance f actor
2.4
(L/D=0.5)
2.2 Resistance f actor
2 (L/D=1)
Dead load factor
1.8
(L/D=0.5)
(L/D=1)
1.4
Live load f actor
1.2 (L/D=0.5)
1 Live load f actor
(L/D=1)
0.8
Earthquake load
0.6 f actor (L/D=0.5)
0.4 Earthquake load
f actor (L/D=1)
0.2
0
0.5 1 2 3
E'/D'
184
7.2.3 Load and Resistance Factors for Reinforced Concrete Columns in the
Combined Action Failure Mode
Established upon the load combination format mentioned in Section 6.2 and the
nominal loads in TS 498 (1997) and “Specification for Structures to be Built in
Earthquake Areas (2006)”, the load and resistance factors for reinforced concrete
columns in the combined action failure mode are computed. In Figures 7.19 to 7.27,
the values of load and resistance factors computed for different load combinations
considered in the study are plotted.
2.2
2
1.8
1.6 Resistance
factor
1.4 Dead load
factor
1.2
Live load
factor
1
0.8
0.6
0.25 0.5 1 2
L'/D'
combined action failure mode (Turkey; D+L; βT =3.2)
185
4
3.8
3.6

3.4
3.2
3
2.8 Resistance
2.6 factor
2.4 Dead load
2.2 factor
2
Snow load
1.8 factor
1.6
1.4
1.2
1
0.8
0.5 1 2 3
S'/D'
combined action failure mode (Bursa; D+S; βT =3.2)
2.4
2.2
1.8
Resistance
1.6 factor
Dead load
1.4 factor
Snow load
1.2 factor
0.8
0.5 1 2 3
S'/D'
combined action failure mode (Turkey; D+S; βT =3.2)
186
1.6
1.5
1.4
1.3
1.2
Resistance factor
(L/D=0.5)
1.1
Resistance factor
(L/D=1)
1
Dead load factor
(L/D=0.5)
(L/D=1)
(L/D=0.5)

(L/D=1)
Wind load factor
0.6
(L/D=0.5)
Wind load factor
0.5 (L/D=1)
0.4
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
combined action failure mode (Canakkale; D+Lapt+W; βT =3.2)
187
1.4
1.3
1.2
1.1
1 Resistance factor
(L/D=0.5)
Resistance factor
0.9
(L/D=1)
Dead load factor
0.8 (L/D=0.5)
Dead load factor
(L/D=1)
0.7
Live load factor
(L/D=0.5)
(L/D=1)
Wind load factor
0.5 (L/D=0.5)
Wind load factor
(L/D=1)
0.4
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
combined action failure mode (Turkey; D+Lapt+W; βT =3.2)
188
1.8
1.7
1.6
Resistance f actor
1.5
(L/D=0.5)

1.4
(L/D=1)
1.2
1 (L/D=0.5)
0.8 (L/D=1)
0.6 (L/D=0.5)
0.4 (L/D=1)
0.2 (L/D=0.5)
0 (L/D=1)
0.25 0.5 1 2
W'/D'
combined action failure mode (Canakkale; D+L+Wapt; βT =3.0)
1.8
1.7
1.6
(L/D=0.5)
1.4
(L/D=1)
1.2
1 (L/D=0.5)
0.8 (L/D=1)
0.6 (L/D=0.5)
0.4 (L/D=1)
0.2 (L/D=0.5)
0 (L/D=1)
0.25 0.5 1 2
W'/D'
combined action failure mode (Turkey; D+L+Wapt; βT =3.0)
189
3.8
3.6
3.4
3.2

3 Resistance factor
2.8 (L/D=0.5)
2.6
Resistance factor
2.4 (L/D=1)
2.2
Dead load f actor
2
(L/D=0.5)
1.8
1.4 (L/D=1)
1 (L/D=0.5)
0.6 (L/D=1)
0.4 Earthquake load
0.2 factor (L/D=0.5)
0
Earthquake load
0.5 1 2 3 factor (L/D=1)
E'/D'
combined action failure mode (Ankara; D+Lapt+E; βT =1.75)
2.6
2.4
2.2
Resistance factor
2
(L/D=0.5)
1.6 (L/D=1)
Dead load factor
1.4
(L/D=0.5)
1 (L/D=1)
Live load f actor
0.8
(L/D=0.5)
0.6
Live load f actor
0.4 (L/D=1)
0.2 Earthquake load

f actor (L/D=0.5)
0
Earthquake load
0.5 1 2 3 f actor (L/D=1)
E'/D'
combined action failure mode (Turkey; D+Lapt+E; βT =1.75)
190
7.2.4 Load and Resistance Factors for Reinforced Concrete Columns in the
Shear Failure Mode
Based on the load combination format mentioned in Section 6.2 and the nominal
loads in TS 498 (1997) and “Specification for Structures to be Built in Earthquake
Areas (2006)”, the load and resistance factors for reinforced concrete columns in
the shear failure mode are computed. In Figures 7.28 to 7.36, the load and resistance
factors for different load combinations considered in this study are displayed in a
graphical form.
1.6
1.5
1.4
1.3
Resistance
1.2 factor
1.1 Dead load
factor
1
Live load
0.9 factor
0.8
0.7
0.6
0.25 0.5 1 2
L'/D'
Figure 7.28 Variation of the load and resistance factors for RC columns in the shear
failure mode (Turkey; D+L; βT =3.0)
191
3.8
3.6
3.4

3.2
3
2.8
2.6 Resistance
2.4 factor
2.2 Dead load
2 factor
1.8
Snow load
1.6 factor
1.4
1.2
1
0.8
0.5 1 2 3
S'/D'
failure mode (Bursa; D+S; βT =3.2)
2.3
2.1
1.9
1.7
1.5 Resistance
factor
1.3 Dead load
factor
1.1
Snow load
0.9 factor
0.7
0.5
0.5 1 2 3
S'/D'
192
1.3
1.2
1.1
Resistance factor
(L/D=0.5)
0.9
Resistance factor
(L/D=1)
Dead load factor
0.8 (L/D=0.5)
Dead load factor
(L/D=1)
(L/D=0.5)
Live load factor
0.6 (L/D=1)
Wind load factor
(L/D=0.5)
Wind load factor
0.5
(L/D=1)
0.4
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
193
1.2
1.1
0.9
Resistance factor
(L/D=0.5)
Resistance factor
0.8
(L/D=1)
Dead load factor
(L/D=0.5)
(L/D=1)
Live load factor
(L/D=0.5)
0.6
Live load factor
(L/D=1)
Wind load factor
0.5 (L/D=0.5)
Wind load factor
(L/D=1)
0.4
0.3
0.2
0.1
0
0.25 0.5 1 2
W'/D'
194
1.5
1.4
Resistance f actor
1.3 (L/D=0.5)

1.2
Resistance f actor
1.1 (L/D=1)
1 Dead load f actor
0.9 (L/D=0.5)
0.7 (L/D=1)
0.5 (L/D=0.5)
0.3 (L/D=1)
0.1 (L/D=0.5)
0 Wind load factor
0.25 0.5 1 2 (L/D=1)
W'/D'
1.5
1.4

(L/D=0.5)
1.2
Resistance factor
1.1
(L/D=1)
1
Dead load factor
0.9 (L/D=0.5)
(L/D=1)
0.7
Live load
0.6
f actor(L/D=0.5)
0.5
Live load
0.4 f actor(L/D=1)
0.2
(L/D=0.5)
Wind load f actor
0.1
(L/D=1)
0
0.25 0.5 1 2
W'/D'
195
3.8
3.6
3.2 (L/D=0.5)
3 Resistance factor

2.8 (L/D=1)
2.4 (L/D=0.5)
2 (L/D=1)
1.6 (L/D=0.5)
1.4
Live load f actor
1.2
(L/D=1)
1
0.8 Earthquake load
factor (L/D=0.5)
0.6
0.4 Earthquake load
0.2 factor (L/D=1)
0
0.5 1 2 3
E'/D'
2.6
2.4
Resistance f actor
2.2 (L/D=0.5)
2 Resistance f actor
(L/D=1)
1.8
Dead load factor
1.6 (L/D=0.5)
(L/D=1)
1.2
Live load f actor
1
(L/D=0.5)
0.6 (L/D=1)
0.4 Earthquake load

f actor (L/D=0.5)
0.2
Earthquake load
0 f actor (L/D=1)
0.5 1 2 3
E'/D'
196
7.2.5 Load and Resistance Factors for Reinforced Concrete Shear Walls in
the Flexural Failure Mode
It is observed that the load and resistance factors computed for reinforced concrete
shear walls in the flexural failure mode are almost identical to those computed for
beams in the flexural failure mode (Section 7.2.1) For this reason, the graphs for the
load and resistance factors are not replotted in this section; the figures given in
Section 7.2.1 are also applicable for this section.
7.2.6 Load and Resistance Factors for Reinforced Concrete Shear Walls in
the Shear Failure Mode
Similar to the previous section, the resistance statistics and target reliabilities for
shear walls in the shear failure mode are observed to be of the same values as those
computed in Section 7.2.3. Hence, the results related to load and resistance factors
to be computed for the shear failure mode of shear walls will be the same as those
found for columns in the combined action failure mode. Accordingly, the figures
drawn for the resistance and load factors displayed in Section 7.2.3 are exactly the
same for this section.
197
7.3 OPTIMAL LOAD AND RESISTANCE FACTORS
Load and resistance factors depend on the mean, variance and probability
distribution of basic variables involved in the limit state functions as well as type of
the load combinations considered. As observed in Section 7.2, load and resistance
factors corresponding to a specified target reliability show quite a high variability
depending on the ratio of the selected load to the dead load. If a constant set of load
and resistance factors are recommended, the corresponding reliabilities will deviate
from the target reliabilities for most of the design situations. This is due to the fact
that when the magnitudes of the time-varying loads increase, the related load factors
also increase. Nevertheless, it is possible to select one set of load and resistance
factors that minimizes the extent of the deviation from the target reliability when all
likely combinations of a particular load, which is given in Table 6.1, are taken into
consideration.
In fact, as emphasized by Ellingwood et al. (1980), “in general, an optimal set of

load factors can be selected by first defining some function which measures the
“closeness” between the target reliability and the reliability associated with the
proposed load and resistance factors set, and then selecting the load factors so as to
minimize this function”.
By using the load and resistance factors given in Section 7.2, a nominal resistance,
R IIn , corresponding to a given set of nominal loads and a target reliability, can be
obtained. In addition, a design equation, which prescribes a set of load factors that
are constant for all load ratios will also result in a nominal resistance, R In . For
example, in the case of D+L+E load combination, if the factored resistance and the
D, L and E loads are linearly related, R In will be equal to:
R In = ( γ D D ′ + γ L L′ + γ E E ′) / ϕ (7.1)
198
Considering these two nominal resistance values, a set of load and resistance factors
can be selected in order to minimize the following difference equation:
( )2
D(ϕ, γ i ) = ∑i R IIni − R Ini p i (7.2)
where pi is the relative weight assigned to the ith load situation for the relative
frequencies shown in Table 6.3. Ellingwood et al. (1980) stated that optimal load
and resistance factors are significantly more sensitive to the range of the ratio of
time-varying load to dead load than to the distribution of pi within that range,
implying that any deviation from the actual values of pi does not considerably
influence the values of optimal load and resistance factors.
Here, the optimal load and resistance factors are computed by utilizing a computer
program (MINMUM) in which Fletcher-Powell numerical search method is used
for the minimization procedure. The results corresponding to different failure
modes of beams, columns and shear walls are given in Tables 7.1 through 7.6.
Table 7.1 Optimal load and resistance factors for reinforced concrete beams in the
flexural failure mode
Optimal Load and Resistance Factors

γ φ
D+L Turkey 1.11D+1.80L 0.92
Bursa 1.06D+3.38S 1.14
D+S
Turkey 1.07D+2.00S 1.11
Canakkale 1.10D+0.49L+1.23W 0.88
D+Lapt+W
Turkey 1.10D+0.55L+0.89W 0.89
Canakkale 1.10D+1.40L+0.43W 0.90
D+L+Wapt
Turkey 1.10D+1.44L+0.21W 0.89
Ankara 1.02D+0.26L+3.75E 1.19
D+Lapt+E
Turkey 1.03D+0.23L+2.51E 1.19
199
Table 7.2 Optimal load and resistance factors for reinforced concrete beams in the
shear failure mode

γ φ
D+L Turkey 1.10D+1.26L 0.65
Bursa 1.06D+3.30S 1.03
D+S
Turkey 1.08D+1.55S 0.87
Canakkale 1.10D+0.41L+1.00W 0.68
D+Lapt+W
Turkey 1.10D+0.44L+0.74W 0.68
Canakkale 1.09D+1.24L+0.38W 0.67
D+L+Wapt
Turkey 1.09D+1.26L+0.21W 0.67
Ankara 1.06D+0.25L+3.62E 1.13
D+Lapt+E
Turkey 1.06D+0.20L+2.43E 1.12
Table 7.3 Optimal load and resistance factors for reinforced concrete columns in the
combined action failure mode

γ φ
D+L Turkey 1.10D+1.92L 0.83
Bursa 1.06D+3.85S 1.10
D+S
Turkey 1.07D+2.29S 1.08
Canakkale 1.10D+0.48L+1.17W 0.80
D+Lapt+W
Turkey 1.10D+0.56L+0.88W 0.80
Canakkale 1.10D+1.41L+0.43W 0.81
D+L+Wapt
Turkey 1.10D+1.46L+0.25W 0.81
Ankara 1.06D+0.27L+3.71E 1.13
D+Lapt+E
Turkey 1.06D+0.28L+2.45E 1.11
200
Table 7.4 Optimal load and resistance factors for reinforced concrete columns in the
shear failure mode

γ φ
D+L Turkey 1.10D+1.41L 0.71
Bursa 1.06D+3.65S 1.02
D+S
Turkey 1.07D+0.94S 0.54
Canakkale 1.10D+0.40L+1.05W 0.65
D+Lapt+W
Turkey 1.10D+0.44L+0.76W 0.66
Canakkale 1.09D+1.23L+0.38W 0.66
D+L+Wapt
Turkey 1.09D+1.23L+0.22W 0.66
Ankara 1.06D+0.25L+3.62E 1.13
D+Lapt+E
Turkey 1.06D+0.20L+2.43E 1.12
Table 7.5 Optimal load and resistance factors for reinforced concrete shear walls in
the flexural failure mode

γ φ
D+L Turkey 1.11D+1.80L 0.92
Bursa 1.06D+3.94S 1.13
D+S
Turkey 1.07D+2.33S 1.10
Canakkale 1.10D+0.49L+1.20W 0.89
D+Lapt+W
Turkey 1.11D+0.54L+0.88W 0.89
Canakkale 1.10D+1.40L+0.43W 0.90
D+L+Wapt
Turkey 1.10D+1.44L+0.21W 0.89
Ankara 1.02D+0.26L+3.75E 1.19
D+Lapt+E
Turkey 1.03D+0.23L+2.51E 1.19
201
Table 7.6 Optimal load and resistance factors for reinforced concrete shear walls in
the shear failure mode

γ φ
D+L Turkey 1.10D+1.92L 0.83
Bursa 1.06D+3.85S 1.10
D+S
Turkey 1.07D+2.29S 1.08
Canakkale 1.10D+0.48L+1.17W 0.80
D+Lapt+W
Turkey 1.10D+0.56L+0.88W 0.80
Canakkale 1.10D+1.41L+0.43W 0.81
D+L+Wapt
Turkey 1.10D+1.46L+0.25W 0.81
Ankara 1.06D+0.27L+3.71E 1.13
D+Lapt+E
Turkey 1.06D+0.28L+2.45E 1.11
7.4 MODIFIED OPTIMAL LOAD AND RESISTANCE FACTORS
As shown in Section 7.3, the resistance factors range from 0.65 to 0.92 for D+L and
from 0.65 to 0.90 for D+L+W load combinations. On the other hand, for the D+S
and D+L+E load combinations, these ranges are respectively 0.54 to 1.14 and 1.11
to 1.19. However, a resistance factor greater than 1.0 is not accustomed
(Ellingwood et al, 1980; Kömürcü and Yücemen, 1996). In order to reduce the
resistance factors which are greater than 1.0, the dead load factor is increased to 1.2,
which has been computed approximately to be 1.1 for all load combinations and for
all failure modes (Tables 7.1 to 7.6). It should be noted that in similar studies,
Ellingwood et al. (1980) and Kömürcü (1995) also increased the value of the dead
load factor from the optimal value of 1.1 to 1.2. In this study, to decrease the
resistance factor, the live load factors are also slightly increased and rounded to 0.4
in D+L+E load combination for all failure modes and for all structural members.
Also the resistance factors are set equal to the same value for all load combinations
202
for each structural member failing in a specific failure mode. The resistance factor
values are selected close to the most frequently observed value for the case under
consideration (i.e. the case corresponding to a certain structural member in a
specified failure mode), and are rounded. After these modifications, a new set of
load and resistance factors are obtained for each failure mode and for each
structural member. These load and resistance factors, which are called as modified
optimal load and resistance factors, are given in Tables 7.7-7.12.
beams in the flexural failure mode
Load Load and Resistance Factors

Location
Combination
γ φ
D+L Turkey 1.2D+1.74L 0.90
Bursa 1.2D+2.56S 0.90
D+S
Turkey 1.2D+1.44S 0.90
Canakkale 1.2D+0.54L+1.24W 0.90
D+Lapt+W
Turkey 1.2D+0.59L+0.91W 0.90
Canakkale 1.2D+1.40L+0.43W 0.90
D+L+Wapt
Turkey 1.2D+1.44L+0.22W 0.90
Ankara 1.2D+0.40L+2.36E 0.90
D+Lapt+E
Turkey 1.2D+0.40L+1.37E 0.90
203
beams in the shear failure mode

Location
Combination
γ φ
D+L Turkey 1.2D+1.38L 0.70
Bursa 1.2D+2.01S 0.70
D+S
Turkey 1.2D+1.03S 0.70
Canakkale 1.2D+0.41L+1.00W 0.70
D+Lapt+W
Turkey 1.2D+0.44L+0.74W 0.70
Canakkale 1.2D+1.39L+0.40W 0.70
D+L+Wapt
Turkey 1.2D+1.39L+0.32W 0.70
Ankara 1.2D+0.40L+1.68E 0.70
D+Lapt+E
Turkey 1.2D+0.40L+0.91E 0.70
columns in the combined action failure mode

Location
Combination
γ φ
D+L Turkey 1.2D+1.80L 0.80
Bursa 1.2D+2.65S 0.80
D+S
Turkey 1.2D+1.45S 0.80
Canakkale 1.2D+0.49L+1.18W 0.80
D+Lapt+W
Turkey 1.2D+0.56L+0.87W 0.80
Canakkale 1.2D+1.38L+0.42W 0.80
D+L+Wapt
Turkey 1.2D+1.40L+0.24W 0.80
Ankara 1.2D+0.40L+2.10E 0.80
D+Lapt+E
Turkey 1.2D+0.40L+1.27E 0.80
204
columns in the shear failure mode

Location
Combination
γ φ
D+L Turkey 1.2D+1.39L 0.70
Bursa 1.2D+2.22S 0.70
D+S
Turkey 1.2D+1.20S 0.70
Canakkale 1.2D+0.64L+1.12W 0.70
D+Lapt+W
Turkey 1.2D+0.59L+0.80W 0.70
Canakkale 1.2D+1.47L+0.40W 0.70
D+L+Wapt
Turkey 1.2D+1.27L+0.25W 0.70
Ankara 1.2D+0.40L+1.67E 0.70
D+Lapt+E
Turkey 1.2D+0.40L+0.91E 0.70
shear walls in the flexural failure mode

Location
Combination
γ φ
D+L Turkey 1.2D+1.74L 0.90
Bursa 1.2D+3.15S 0.90
D+S
Turkey 1.2D+1.69S 0.90
Canakkale 1.2D+0.49L+1.24W 0.90
D+Lapt+W
Turkey 1.2D+0.54L+0.91W 0.90
Canakkale 1.2D+1.40L+0.43W 0.90
D+L+Wapt
Turkey 1.2D+1.44L+0.22W 0.90
Ankara 1.2D+0.40L+2.36E 0.90
D+Lapt+E
Turkey 1.2D+0.40L+1.37E 0.90
205
shear walls in the shear failure mode

Location
Combination
γ φ
D+L Turkey 1.2D+1.80L 0.80
Bursa 1.2D+2.65S 0.80
D+S
Turkey 1.2D+1.45S 0.80
Canakkale 1.2D+0.49L+1.18W 0.80
D+Lapt+W
Turkey 1.2D+0.56L+0.87W 0.80
Canakkale 1.2D+1.38L+0.42W 0.80
D+L+Wapt
Turkey 1.2D+1.40L+0.24W 0.80
Ankara 1.2D+0.40L+2.10E 0.80
D+Lapt+E
Turkey 1.2D+0.40L+1.27E 0.80
For the beam flexure, Ellingwood et al. (1980), El-Etoom (1985) and Kömürcü and
Yücemen (1996) also suggested a different set of load and resistance factors. Their
results are given in Tables 7.13 and 7.14.
Table 7.13 Load and resistance factors in USA and Jordan for beams in the flexural
failure mode
Load Combination Country Load and Resistance Factors
γ φ
USA 1.2D+1.6L 0.81
D+L (or D+S)
Jordan 1.2D+1.7L 0.77
1.2D+0.5L+1.3W 0.84
USA
1.2D+1.6L+0.1W -
D+L+W
1.2D+0.4L+1.6W 0.78
Jordan
1.2D+1.7L 0.72
USA 1.2D+0.2L+1.5E 0.82
D+L+E
Jordan 1.2D+0.4L+2.0E 0.80
206
Table 7.14 Load and resistance factors recommended by Kömürcü and Yücemen
(1996) for beams in the flexural failure mode
Load and Resistance

Factors
γ φ
D+L Turkey 1.3D+1.67L 0.90
Erzincan 1.3D+1.80S 0.90
D+S
Turkey 1.3D+1.20S 0.90
Erzincan 1.3D+0.49L+0.96W 0.90
D+L+W
Turkey 1.3D+0.5L+0.8W 0.90
Ankara 1.3D+0.30L+2.21E 0.90
D+L+E
Turkey 1.3D+0.30L+2.10E 0.90
In order to come up with a single set of load factors applicable to all members and
failure modes, the load factors given in Tables 7.7 to 7.12 are averaged. The
resulting load factors are tabulated in Table 7.15. It is to be noted that these load
factors should be used together with the appropriate resistance factors, which will
be computed in Section 7.5 for rounded load factors.
Table 7.15 Average load factors for all structural members in different failure
modes
Load Average Load Factors

Location
Combination
D+L Turkey 1.2D+1.64L
Bursa 1.2D+2.52S
D+S
Turkey 1.2D+1.38S
Canakkale 1.2D+0.53L+1.15W
D+Lapt+W
Turkey 1.2D+0.57L+0.89W
Canakkale 1.2D+1.34L+0.38W
D+L+Wapt
Turkey 1.2D+1.37L+0.23W
Ankara 1.2D+0.40L+2.05E
D+Lapt+E
Turkey 1.2D+0.40L+1.18E
207
In TS 500 (2000), material safety factors are used but there is no resistance factor;
that is, the resistance factor, φ, is equal to 1.0. Furthermore, the specified load
factors have to be used for all structural members and failure modes. Here, a set of
load factors are developed just for the sake of establishing a parallelism with the
provisions in TS 500 (2000), based on the load factors modified for achieving a
resistance factor, φ=1.0. The resulting load factors, which will be used for all
members and for all failure modes, are given in Table 7.16. However, the reliability
index values corresponding to these load factors will vary according to different
structural members and failure modes. To examine the extent of this variability, the
reliability index values corresponding to this case are computed and shown in Table
7.16. Although the table is constituted only for beams in the flexural and shear
failure modes, and columns in the combined action failure mode, they also represent
the reliability index values for the other failure modes. The reliability index values
computed for columns in the shear failure mode are identical to those of beams.
Similarly, the reliability indexes pertaining to the flexural and the shear failure
mode of shear walls are the same as those for the flexural failure mode of beams
and the combined action failure mode of columns, respectively.
Table 7.16 Load and resistance factors and corresponding reliability index values
for different structural members and failure modes for φ=1.0
Beams in the Columns in the

Beams in the shear
Load Factors flexural failure combined action
failure mode
mode failure mode
β β β β β β
1.3D+1.9L 2.93-3.09 3.07 2.32-2.67 2.54 2.78-2.98 2.94
1.3D+1.6S 2.92-3.27 3.12 2.60-2.98 2.88 2.91-3.23 3.09
1.3D+0.80L+1.0W 3.02-3.28 3.18 2.54-2.80 2.65 2.98-3.16 3.06
1.3D+1.6L+0.4W 2.77-3.10 2.93 2.33-2.62 2.41 2.67-3.07 2.79
1.3D+0.6L+1.6E 1.51-2.12 1.78 1.51-2.00 1.68 1.50-2.11 1.74
208
As observed in Table 7.16, for example, for D+L load combination, while the
reliability index value for beams in the flexural failure mode is observed to be about
3.07 (βT=3.0), this value is 2.54 (βT=3.0) and 2.94 (βT=3.2) for beams in the shear
failure mode and for columns in the combined action, respectively. In other words,
for the same resistance factor, φ=1, the reliability indexes deviate significantly from
the selected target reliability index value, βT, according to the type of structural
member and failure mode. This shows the deficiency of TSE 500 (2000).
Accordingly, different resistance factors should be used for different failure modes
and different structural member types if the same load factors are used.
7.5 RECOMMENDED LOAD AND RESISTANCE FACTORS
In order to simplify the works of structural engineers, it is desirable to propose only

one set of load factors. For this purpose, the average load factors computed for
Turkey (see Table 7.15) are slightly rounded towards larger values and are
presented in Table 7.17. These are the final load factors that are recommended to be
used in design practice for all structural members in different failure modes.
Consistent with these proposed load factors, new resistance factors are computed
for each structural member failing in a specific failure mode, based on the target
reliability indexes given in Section 6.4.7. The reliability index values corresponding
to the recommended load and resistance factors are computed in order to compare
them with the selected target reliabilities. The results are tabulated in Tables 7.18-
7.23.
209
Table 7.17 Recommended load factors for all structural members in different failure
modes for Turkey
Load Recommended
Combination Load Factors
D+L 1.2D+1.7L
D+S 1.2D+1.4S
D+Lapt+W 1.2D+0.6L+0.9W
D+L+Wapt 1.2D+1.4L+0.3W
D+Lapt+E 1.2D+0.40L+1.3E
reliability index values for beams in the flexural failure mode
Load Resistance Recommended Reliability index values

Combination Factor Load Factors ΒT β β
D+L 1.2D+1.7L 3.0 3.0-3.14 3.10
D+S 1.2D+1.4S 3.0 2.89-3.28 3.04
D+Lapt+W 0.90 1.2D+0.6L+0.9W 3.0 2.92-3.25 3.11
D+L+Wapt 1.2D+1.4L+0.3W 2.7 2.70-2.95 2.84
D+Lapt+E 1.2D+0.4L+1.3E 1.75 1.39-1.98 1.62
reliability index values for beams in the shear failure mode

D+L 1.2D+1.7L 3.0 2.83-3.06 3.00
D+S 1.2D+1.4S 3.0 3.08-3.25 3.19
D+Lapt+W 0.78 1.2D+0.6L+0.9W 3.0 3.04-3.25 3.12
D+L+Wapt 1.2D+1.4L+0.3W 2.7 2.77-2.93 2.82
D+Lapt+E 1.2D+0.4L+1.3E 1.75 1.57-2.18 1.90
210
reliability index values for columns in the combined action failure mode

D+L 1.2D+1.7L 3.2 3.15-3.27 3.24
D+S 1.2D+1.4S 3.2 3.02-3.44 3.26
D+Lapt+W 0.83 1.2D+0.6L+0.9W 3.2 3.16-3.43 3.30
D+L+Wapt 1.2D+1.4L+0.3W 3.0 2.91-3.20 3.03
D+Lapt+E 1.2D+0.4L+1.3E 1.75 1.50-2.11 1.74
reliability index values for columns in the shear failure mode

D+L 1.2D+1.7L 3.0 2.96-3.15 3.10
D+S 1.2D+1.4S 3.2 3.12-3.33 3.28
D+Lapt+W 0.75 1.2D+0.6L+0.9W 3.2 3.06-3.27 3.14
D+L+Wapt 1.2D+1.4L+0.3W 3.0 2.90-3.03 2.93
D+Lapt+E 1.2D+0.4L+1.3E 1.75 1.63-2.25 1.87
reliability index values for shear walls in the flexural failure mode

D+L 1.2D+1.7L 3.0 3.08-3.23 3.19
D+S 1.2D+1.4S 3.2 2.94-3.34 3.16
D+Lapt+W 0.88 1.2D+0.6L+0.9W 3.2 3.06-3.37 3.23
D+L+Wapt 1.2D+1.4L+0.3W 2.7 2.79-3.14 2.93
D+Lapt+E 1.2D+0.4L+1.3E 1.75 1.42-2.02 1.65
211
reliability index values for shear walls in the shear failure mode

D+L 1.2D+1.7L 3.2 3.15-3.27 3.24
D+S 1.2D+1.4S 3.2 3.02-3.44 3.26
D+Lapt+W 0.83 1.2D+0.6L+0.9W 3.2 3.16-3.43 3.30
D+L+Wapt 1.2D+1.4L+0.3W 3.0 2.91-3.20 3.03
D+Lapt+E 1.2D+0.4L+1.3E 1.75 1.50-2.11 1.74
As observed in Tables 7.18 to 7.23, the resulting mean reliability index values are
very close to the selected target reliability indexes, and in the majority of cases
these reliability index values are slightly greater than the target reliability values;
accordingly this creates a safer condition and it does not necessitate further
consideration. In Appendix G, the safety levels (in terms of reliability index) related
to the recommended load and resistance factors are plotted as functions of L'/D',
S'/D', W'/D' and E'/D' for the load combinations considered. It is to be noted that the
reliability index values computed for the combined action failure mode of columns
are identical to those of the shear failure mode of shear walls owing to the same
target reliability indexes and resistance statistics. Therefore, the graphs are not
plotted for the shear failure mode of shear walls. In these figures, the target
reliability indexes are also marked so that the deviations of reliability indexes from
target reliabilities, which is known to depend highly on the load and resistance
statistics and on the ratio of given loads to dead load, can also be evaluated.
212
CHAPTER 8
SUMMARY AND CONCLUSIONS
This thesis work is carried out to establish a reliability based design criterion for
different reinforced concrete structural members under different failure modes
considering the local conditions and the design practice in Turkey. The probabilistic
method of analysis utilized is the Advanced First Order Second Moment (AFOSM)
procedure. Factored resistance and factored loads form the basis for the limit state
function under consideration. The load combination model used is Turkstra’s rule.
Various sources of uncertainties associated with the load and resistance parameters
are analyzed and quantified based on the data available in Turkey. The published
data in the international literature are used to supplement the data compiled within
the framework of this study whenever local data are insufficient. In addition, related
results reported by other researchers are examined for the comparison of the results
found in this study.
The reliability indexes corresponding to the current design practice in Turkey for
reinforced concrete beam, column and shear wall design in flexure and shear as well
as column design in combined action of flexure and axial load subjected to different
load combinations are computed. Considering these values and those reported from
other studies, target reliabilities are selected for different failure modes of structural
members under different load combinations. Finally, a new set of load and
resistance factors corresponding to selected target reliabilities and levels of
213
uncertainties are proposed. The computer program A58LF (Ellingwood et al., 1980)
is utilized for the reliability calculations. A58LF computes the reliability index
values for a given design situation established by a set of nominal load and
resistance values (analysis problem) and computes the needed resistance and load
factors for a prescribed a set of nominal loads (design problem). The optimal load
and resistance factors are obtained by utilizing a code written in Fortran 77
(MINMUM). In addition to the programs mentioned above, the resistance statistics
of reinforced concrete members in different failure modes are obtained by using
five different codes, which are developed in MathCAD 12 for this study.
The main conclusions and recommendations derived from this study are listed
below:
1- Based on the evaluation of the 28 day concrete compressive strength test results
of approximately 11000 specimens collected in Turkey, the mean value of cubic
compressive strength is found as 29.87 N/mm2, and c.o.v is computed as 0.105. If
the cubic compressive strength of concrete data is converted to the standard cylinder
strength value, then f c and δfc are found to be 24.87 N/mm2 and 0.105, respectively.
It should be noted that the values of c.o.v. (i.e. δfc) are obtained for each
construction site individually in this study. The mean to nominal concrete
compressive strength ratio is computed as 1.25. In other words, the “true”
compressive strength obtained from the field data (i.e. in-situ data) is 25% higher
than the nominal compressive strength specified in the relevant standards. The
uncertainty analysis showed that the total uncertainty for concrete compressive
strength equals to 0.18 if the prediction uncertainties (epistemic uncertainties) are
combined with the basic variability (aleatory uncertainty). When these results are
compared with the ones obtained in the earlier years, it can be concluded that the
quality of concrete has been improved over the years from the compressive strength
and basic variability points of view.
214
2- The statistical parameters of yield strength, ultimate strength and elongation of
reinforcing steel bars are obtained from the production reports of the seven steel and
iron plants. In addition, the data based on the civil engineering materials
laboratories of Istanbul Technical University, Middle East Technical University and
Selcuk University are investigated. After conducting a set of statistical analyses on
the collected data, the mean yield strength is computed to be 501.37 N/mm2 as the
weighted average value, obtained from the data supplied by iron and steel plants
and university laboratories. The basic variability in yield strength is found to range
from 0.02 to 0.07 with a weighted average value of 0.038 for individual bar sizes
based on the data obtained from the iron and steel plants. In this study, the mean to
nominal ratio of yield strength of reinforcing steel bars and total uncertainty in them
are calculated as 1.24 and 0.09, respectively.
3- Depending on the analysis of local data obtained from different construction

sites, the average basic variabilities in the various dimensions, such as width, depth,
effective depth for beams, columns and shear walls range from 0.03 to 0.07. The
total variabilities in these dimensions change between 0.038 and 0.074. For all
dimensions, the mean to nominal ratios are observed to be about 1.00.
4- The mean to nominal ratios and the total uncertainties of resistances in different
failure modes for different reinforced structural members are computed by using the
compressive strength of concrete, the yield strength of reinforcing steel bars and the
dimensions of structural members within the framework of the reliability model.
While the mean to nominal ratio (the ratio of “true” mean value to nominal value
specified in relevant standards) is found as 1.24 for all failure modes and for all
reinforced concrete structural members considered in this study, the total
uncertainties vary within a range of 0.13 and 0.17 depending on the failure mode
and the member type.
5- Loads considered in this study are dead, live, snow, wind, and earthquake. Due to
lack of local data, the statistical parameters of dead and live loads are calculated
215
based on the information assessed from the foreign literature. This can be accepted
as a reasonable assumption owing to the fact that the characteristics of dead and live
loads are believed not to change noticeably from one country to another. It is
assumed that the mean to nominal ratio and total variability of dead load are 1.05
and 0.10, respectively, as the approximate average value of the studies reported in
the international literature. As for live load, the mean to nominal ratios are taken as
1.00 and 0.28 for maximum live load and arbitrary point-in-time live load, while
total variabilities are obtained as 0.28 and 0.70, respectively.
6- Although dead and live loads acting on a structure are independent of the
geographical location of the structure; snow, wind and earthquake loads, which can
be termed as environmental loads, are quite dependent. For this reason, in order to
compute the parameters of environmental loads for the whole Turkey, some
representative cities, namely Ankara, Izmir, Antalya, Bursa, Gaziantep, Samsun,
Malatya, Erzincan, Canakkale, Hakkari, Göztepe/Istanbul and Sile/Istanbul are
considered. In order to compute snow, wind and earthquake loads, the mean values
of arbitrary point-in-time and maximum lifetime loads in these locations are
considered separately.
7- Snow depth is the main parameter affecting the snow load. The data on snow
depth are obtained from the records of Turkish Meteorological Department. The
nominal roof snow load, S′ , can be obtained from TS 498 (1997). The mean to
nominal ratios range 0.01 to 0.89 and 0.06 to 2.68 for annual extreme roof snow
load and maximum roof snow load in the geographical locations listed above. As
indicated by these values, the differences between mean and nominal values are
excessive for some of these locations. Therefore, it is recommended that the
nominal snow load values in TS 498 (1997) should be revised. The total
uncertainties in annual extreme roof snow load and maximum roof snow load range
from 0.60 to 0.93 and 0.27 to 0.48, respectively.
8- The daily and annual wind speed data, which are necessary for the wind load
216
statistics, are obtained from the meteorological stations of Turkish Meteorological
Department located in twelve different cities. For these cities, the mean to nominal
ratios range from 0.09 to 0.32 and from 0.35 to 0.82 for arbitrary point-in-time and
maximum wind loads, and also the total uncertainties in these loads range from 0.60
to 0.93 and from 0.27 to 0.48, respectively. As it is seen from these values, the
nominal wind load values in TS 498 (1997) stay on the safe side for all locations.
9- Earthquake load is assumed to depend mainly on the peak ground acceleration.

For the twelve locations, the outcomes of the seismic hazard analysis conducted by
Gülkan et al. (1993) are used for estimating the statistical parameters of peak
ground accelerations. The mean to nominal ratios change between 0.85 and 1.61 for
earthquake loads, and also the total uncertainties in this load, which is quantified in
terms of coefficient of variation, change within a range of 1.05 and 1.21. It is to be
noted that earthquake load shows the most variation among all of the loads.
10- The current design practice is assumed to depend on TS 500 (2000), TS 498
(1997), and Specifications for the Structures to be Built in Disaster Areas (1998).
By utilizing the AFOSM method, the safety levels inherent in the current design
practice are examined for reinforced concrete beams, columns and shear walls in
the flexural and shear failure modes, and also columns in the combined action
depending on the load and resistance statistics compiled in this study. Considering
the failure modes of reinforced concrete structural members studied, the average
reliability index values for the whole Turkey (i.e. the average of values obtained
from twelve location) range from 2.41 to 2.90 for D+L; from 2.97 to 3.20 for D+S;
from 2.94 to 3.36 for D+Lapt+W; from 2.40 to 2.79 for D+L+Wapt and from 1.29 to
1.32 for D+Lapt+E. The target reliabilities are selected according to these safety
level values, and also by taking into consideration the target reliabilities reported
from the international literature. In this study, the selected target reliabilities vary
within a range of 1.75 and 3.2 depending on the load combination, the failure mode
and the member type.
217
11- A new set of load and resistance factors corresponding to the selected target
reliabilities are determined for different design situations, i.e. for L'/D', S'/D', W'/D'
and E'/D' values according to D+L, D+S, D+Lapt+W, D+L+Wapt and D+Lapt+E load
combinations. The load and resistance factors show variability based on the ratio of
the selected load to the dead load. It is possible to select one set of optimal load and
resistance factors that minimizes the extent of the deviation from the selected target
reliability. After computing these optimal load and resistance factors for each
failure mode of reinforced concrete structural members considered in this
dissertation, some modifications are made in order to obtain more accustomed
resistance factors (i.e. less than 1.0). These factors which are termed as modified
optimal load and resistance factors are as follows:
1.2D + 1.74L
1.2D + 1.44S

U= and φ= 0.90 for beams in the flexural failure mode
 1.2 D + 0.59 L + 0.91W
1.2D + 0.4L + 1.37 E
1.2D + 1.38L
1.2D + 1.03S

U= and φ= 0.70 for beams in the shear failure mode
1.2D + 0.44L + 0.74 W
1.2D + 0.4L + 0.91E
1.2D + 1.8L
1.2D + 1.45S

U= and φ= 0.80 for columns in the combined action
1.2D + 0.56L + 0.87 W failure mode
1.2D + 0.4L + 1.27 E
1.2D + 1.39L
1.2D + 1.20S

U= and φ= 0.70 for columns in the shear failure mode
1.2D + 0.6L + 0.8W
1.2D + 0.4L + 0.91E
218
1.2D + 1.74L
1.2D + 1.69S

U= and φ= 0.9 for shear walls in the flexural failure mode
1.2D + 0.54L + 0.91W
1.2D + 0.4L + 1.37 E
1.2D + 1.8L
1.2D + 1.45S

U= and φ= 0.8 for shear walls in the shear failure mode
 1.2 D + 0. 6 L + 0.87 W
1.2D + 0.4L + 1.27 E
12- In TS 500 (2000), material safety factors are used but resistance factors are not
considered; in other words, the resistance factor, φ, equals 1.0. Furthermore, the
specified load factors have to be used for all structural members and failure modes.
Therefore, a set of load factors are computed corresponding to φ=1 in order to have
a parallelism with the provisions in TS 500 (2000) as given below:
1.3D + 1.9L
1.3D + 1.6S

U= φ= 1.0 for all failure modes and all structural members
1.3D + 0.8L + 1.0 W
1.3D + 0.6L + 1.6E
When the reliability index values corresponding to this case are computed, it is seen
that the reliability indexes exhibit high variability according to structural members
and failure modes, and deviate significantly from the selected target reliability index
βT. Therefore, different resistance factors should be used for each failure mode and
each structural member in such a case where the same load factors are used.
13- In order to simplify the works of structural engineers, only one set of load
factors are proposed by averaging the load factors computed for reinforced concrete
structural members failing in different failure modes. However, the use of the same
resistance factors is not rational due to the fact that the related reliability index
219
values vary according to different failure modes and member types. Therefore, the
following one set of load factors and six different resistance factors are proposed for
each failure mode of the reinforced concrete structural members considered:
1.2D + 1.7 L
1.2D + 1.4S

U=
1.2D + 0.6L + 0.9 W
1.2D + 0.4L + 1.3E
0.90 for beams in the flexural failure mode

0.78 for beams in the shear failure mode

0.83 for columns in the combined action failure mode
ϕ=
0.75 for columns in the shear failure mode
0.88 for shear walls in the flexural failure mode

0.83 for shear walls in the shear failure mode
It is to be noted that these load and resistance factors are to be used in connection
with TS 500 (2000), TS 498 (1997) and Specification for Structures to be Built in
Earthquake Areas (2006).
14- In addition to load combinations consisting of primary loads, different load

combinations including secondary loads, (e.g. temperature and soil load) can be
considered, within the same framework.
15- The statistical data on loads and the probabilistic methodology provided in this
dissertation can be extended to the limit state design of different construction
materials, such as metal structures, engineering masonry and prestressed concrete.
16- The load and resistance factors proposed in this study are open to future
adjustments if new local and relevant international data as well as information
become available.
220
REFERENCES
American Concrete Institute, Building Code Requirements for Reinforced Concrete,

ACI 318/05, ACI, Detroit, Michigan, USA, 2005.
American Society of Civil Engineers, Minimum Design Loads for Buildings and
Other Structures, ASCE Standard, ASCE/SEI 7-05, USA, 2006.
Aktas, E., Structural Design Code Calibration Using Reliability Based Cost
Optimization, Ph. D. Thesis, The University of Pittsburgh, Pittsburgh, 2001.
Aktas E., Moses, F. and Ghosn, M., Cost and Safety Optimization of Structural
Design Specifications, Reliability Engineering and System Safety 73, pp. 205-212,
2001.
Ang, A. H.S and Tang, W.H., Probability Concepts in Engineering Planning and
Design, Volume II, John Wiley and Sons Inc., New York, 1984.
Atimtay, E., Reinforced Concrete Fundamentals, Bizim Büro Basimevi, Ankara,

Turkey, 1998.
Ayaroglu F., Türkiye’de Kar Yagislari, Uzmanlik Tezi, T.C. Basbakanlik Devlet
Meteoroloji Genel Müdürlügü, Ankara, 1991.
Celep, Z. and Kumbasar, N., Betonarme Yapilar, Beta Dagitim, Istanbul, 2005.
221
Dislitas, S., Ahiska R. and Yanmaz, H., Beton Karakteristik Basinc Dayanimi
Testinin Bilgisayar Kontrollü Yapilmasi, Bilgi Teknolojileri Kongresi, Akademik
Bilisim 2006, Pamukkale Üniversitesi, Denizli, 2006.
Durmaz, M. and Daloglu A., Kar Verilerinin İstatistiksel Analizi ve Dogu

Karedeniz Bölgesinin Zemin Kar Haritasinin Olusturulmasi, İMO Teknik Dergi,
pp 3619-3642, 2005.
Dündar C., Türkiye Rüzgar Atlasi, T.C. Basbakanlik Devlet Meteoroloji Genel
Müdürlügü ve Enerji Tabii Kaynaklar Bakanligi Elektrik İsleri Etüt İdaresi Genel
Müdürlügü, Ankara, 2002.
Ellingwood, B.R., Galambos, T.V., MacGregor, J.G. and Cornell, C.A.,

Development of a Probability Based Load Criterion for American National
Standards A58, NPS Special Publication 577, 1980.
Ellingwood, B. and Galambos, T.V., Probability-Based Criteria for Structural

Design, Structural Safety, No.1, pp. 15-26, 1982.
Ellingwood, B., Ang. A. H. S., Probabilistic Study of Safety Criteria for Design,
Structural Research Series No. 387, Department of Civil Engineering, University of
Illinois, Urbana, 1972.
Ellingwood, B.R., Probability-based codified design for earthquakes, Engineering

Structure, Vol. 16, Number 7, pp. 498-506, 1994.
Ellingwood, B.R., Reliability-based condition assessment and LRFD for existing

structures, Structural Safety, Vol. 18, pp 67-80, 1996.
Ersoy, U. and Özcebe, G., Reinforced Concrete, Evrim Yayinevi, Ankara, 2004.
222
Faber, M.H., Kübler, O. and Köhler, J. Tutorial for the JCSS code calibration
program CodeCal 03, 2003.
Ferguson, P.M., Breen, E.B. and James, O.J., Reinforced Concrete Fundamentals,
John Wiley & Sons, Inc., Singapore, 1988.
Galambos, T. V., Ellingwood, B., MacGregor, J.G. and Cornell C.A., Probability
Based Load Criteria: Load Factors and Load Combinations, Journal of The
Structural Division, ASCE, Vol 108, No. ST5, pp. 978-997, 1982.
Ghiocel, D. and Lungu, D., Wind, Snow and Temperature Effects on Structures
Based On Probability, Abacus Press, Turbridge Wells, Kent, 1974.
Gülkan, P., Kocyigit A., Yücemen, M.S., Doyuran, V. and Basöz, N., En Son
Verilere Göre Hazirlanan Türkiye Deprem Bölgeleri Haritasi, Rapor No. 93-01,
Deprem Mühendisligi Arastirma Merkezi, ODTÜ, Türkiye, 1993.
Gündüz, A., Beton Mukavemetinin Betonarme Yapilarin Göcme Riski Üzerine

Etkisi, Ülkemizin Kalkinmasinda Mühendisligin Rolü Sempozyumu, Istanbul, 1988
Hasofer, A. and Lind, N.C., Exact and Invariant Second-Moment Code Format,
Journal of Engineering Mechanics Division, ASCE, Vol. 100, No. EM. 1, pp. 111-
121, 1974
Hwang, H.H.M. and Hsu, H.M., Seismic LRFD Criteria for RC Moment-Resisting
Frame Buildings, Journal of Structural Engineering, Vol. 119, Number 2921, pp.
1807-1824, 1993.
International Code Council, International Building Code 2003, USA, 2003.
Israel, B. M., Ellingwood, B. and Corotis, R., Reliability Based Code Formulations
223
for Reinforced Concrete Buildings, Journal of the Structural Engineering, Vol 113,
No. 10, October, 1987.
Turkish Iron and Steel Producers Association, Iron & Steel Sector in Turkey,
http://www.dcud.org.tr/indextur.htm, last accessed date 2004.
Jeff Quell, P.E, Determining Snow Loads, A Quarterly Information Source from
Benchmark, Inc., Volume 34, 1998.
Kömürcü, A.M., A Probabilistic Assessment of Load and Resistance Factors for

Reinforced Concrete Structures Considering the Design Practice in Turkey, M. Sc.
Thesis, Department of Civil Engineering, METU, Ankara, 1995.
Kömürcü, A.M. and Yücemen, M.S., Load and Resistance Factors for Reinforced
Concrete Beams Considering the Design Practice in Turkey, Concrete Technology
for Developing countries, Fourth International Conference, Eastern Mediterranean
University, Gazi Magusa, North Cyprus, 1996.
Kumar, S., Live Loads in Office Buildings: Point in Time Load Intensity, Building
and Environment, Vol. 37, pp- 79-89, 2002a.
Kumar, S., Live Loads in Office Buildings: Lifetime Maximum Load, Building and
Environment, Vol. 37, pp- 91-99, 2002b.
Melchers R.E., Structural Reliability Analysis and Prediction, John Wiley & Sons,
Inc., Chichester, England, 2002.
Meyer, C., Design of Concrete Structures, Prentice-Hall, Inc., New Jersey, USA,
1996.
Minciarelli, F., Gioffre, M., Grigoriu, M. and Simiu, E., Estimates of Extreme Wind
224
Effects and Wind Load Factors: Influence of Knowledge Uncertainties,
Probabilistic Engineering Mechanics, Vol. 16, pp 331-340, 2001.
Mirza, S. A. and MacGregor, J.G., Variations in Dimensions of Reinforced

Concrete Member, Journal Of The Structural Division, ASCE, Vol. 105, No. ST4,
pp. 921-937, 1979.
Mirza, S.A. and MacGregor, J.G., Variability of Mechanical Properties of

Reinforced Bars, Journal of the Structural Division, ASCE, Vol. 105, No. ST4, pp.
751-766, 1979.
Mirza, S.A., Hatzinikolas, M. and MacGregor, J.G., Statistical Description of the

Strength of Concrete, Journal of the Structural Division, ASCE, Vol. 105, No. ST6,
pp. 1021-1037, 1979.
Mirza, S.A. and MacGregor, J.G., Probabilistic study of strength of Reinforced

concrete members, Can. J. Civ. Engrg., Ottawa 9, pp. 431-438, 1982.
Nawy, E.G., Reinforced Concrete A Fundamental Approach, Pearson Education,

Inc., New Jersey, USA, 2005.
Nilson, A.H., Darwin, D. and Dolan, C.W., Design of Concrete Structures,

McGraw-Hill Companies, Inc. Singapore, 2004.
Nowak, A.S. and Szerszen, M.M., Calibration of Design Code for Buildings (ACI
318): Part 1- Statistical Models for Resistance, ACI Structural Journal, May-June,
pp. 377-382, 2003a.
Nowak, A.S. and Szerszen, M.M., Calibration of Design Code for Buildings (ACI
318): Part 2- Reliability Analysis and Resistance Factors, ACI Structural Journal,
May-June, pp. 383-391, 2003b.
225
Park R. and Paulay, T., Reinforced Concrete Structures, John Wiley & Sons, Inc.,
Singapore, 1975.
Paulay, T. and Priestley, M.J.N., Seismic Design of Reinforced Concrete and

Masonry Buildings, John Wiley & Sons, Inc., USA, 1992.
Q’ Rourke M.J., Redfield, R. and Bradsky, P.V., Uniform Snow Loads on

Structures, Journal of the Structural Division, ASCE, Vol. 108, No ST12, pp. 2781-
2798, 1982.
O’Rourke, M. and Wrenn, P.D., A Guide to the Use and Understanding of the Snow
Load Provisions of ASCE 7-02, ASCE, 2004.
Özsoy, D., İnsaat Celigi ve Kontrollü Su Sogutma Yöntemi”, http://www.mmo. org.

tr/ mühendis makine/arsiv/2001/martinsaatt.htm, last accessed date, 2007
Joint Committee on Structural Safety, Probabilistic Model Code, 2001.
Rackwitz, R., Optimization- the Basis of Code-Making and Reliability Verification,

Structural Safety, Vol. 22, pp. 27-60, 2000.
Real, M.V., Filho, A.C. and Sergio, R.M., Response Variability in Reinforced
Concrete Structures with Uncertain Geometrical and Material Properties, Nuclear
Engineering and Design, pp. 205-220, 2003.
Rusten, A., Sack, R. L. and Malnav, M., Snow Load Analysis for Structures,
Journal of the Structural Division, ASCE, Vol. 108, No ST1, pp.11-21, 1980.
Schiever, W.R., Estimating Snow Loads, CBD-193, Canadian Building Digest,

1978, http://irc.nrc-cnrc.gc.ca/pubs/cbd/cbd193_e.htm , last accessed date 2005.
226
Simiu E. and Scalan, R.H., Wind Effects on Structures, 1st ed., John Wiley, New
York, 1978.
Spiegel, L. and Limbrunner, G.F., Reinforced Concrete Design, Pearson Education,

Inc., New Jersey, USA, 2003.
Sahin, A.D., Türkiye Rüzgarlarinin Alan Zaman Modellemesi, Ph. D. Thesis, İTÜ,
Istanbul, 2001.
T.C. Bayindirlik ve İskan Bakanligi, Deprem Surasi-2004. Yapi Malzemeleri

Komisyonu Raporu, Ankara, Temmuz 2004.
T.C. Bayindirlik ve İskân Bakanligi, Afet Bölgelerinde Yapilacak Yapilar Hakkinda

Yönetmelik, Nihai Deprem Yönetmeligi Taslagi, Türkiye, 1996.

Yönetmelik, Türkiye, 1998.

Yönetmelik, Nihai Deprem Yönetmeligi Taslagi, Türkiye, 2006.
Türk Standartlari, Yapi Elemanlarinin Boyutlandirilmasinda Alinacak Yüklerin

Hesap Degerleri (TS 498 ), Türk Standartlari Enstitüsü , Ankara, 1997.
Türk Standartlari, Betonarme Yapilarin Hesap ve Yapim Kurallari (TS 500), Türk
Standartlari Enstitüsü, Ankara, 2000.
Türk Standartlari, Beton- Bölüm 1: Özellik, Performans, İmalat ve Uygunluk (TS

EN 206-1), Türk Standartlari Enstitüsü, Ankara, 2002.
International Conference of Buildings Officials, Uniform Building Code 1994,
227
Whitter, California, USA., 1994.
Toft-Christensen, P. and Baker, M. J., Structural Reliability Theory and Its

Applications, Springer Verlag, Berlin, 1982.
Topcu, İ.B. and Karakurt, C., Eskisehir’de Kullanilan Yapi Celiklerinin Istatistiksel
Olarak Incelenmesi, Seventh International Congress on Advances in Civil
Engineering, Yildiz Technical University, Istanbul, Turkey, 2006.
Udoeyo, F. F. and Ugbem, P.I, Dimensional Variations in Reinforced Concrete

Members, Journal of the Structural Engineering, December, pp. 1865-1868, 1995.
Vrouwenvelder, A.C.W.M., Developments towards Full Probabilistic Design

codes. Structural Safety. Vol.24, 417-432, 2002.
Yücemen, M.S. and Gülkan, P., Betonarme Yapilar İcin Yük ve Dayanim
Katsayilarinin Belirlenmesi, 10. Teknik Kongre, Cilt 2, İnsaat Mühendisleri Odasi,
s. 637-651, 1989.
Yücemen, M.S. and El-Etoom, A., Reliability Based Load and Resistance Factors
for Reinforced Concrete Beams in Jordan, Proceedings, The Second International
Conference on Concrete Tecnology for Developing Countries, Vol. I, Tripoli,
Libya, pp. 4.1-4.17, 1986.
Yüksel, S.B., A Comparative Study on Seismic Codes, M. Sc. Thesis, Department of

Civil Engineering, METU, Ankara, 1997.
228
APPENDIX A
STATISTICAL PARAMETERS OF 7 AND 28 DAY COMPRESSIVE

STRENGTH DATA ACCORDING TO CONCRETE CLASS AND REGION
Table A.1.a Statistical parameters of 7 day compressive strength data according to

concrete class and region
Concrete Number fck Mean Standard Coefficient

Class of samples (N/mm2) (N/mm2) deviation of variation
Corlu Region
C20 577 25 24.01 3.28 0.137
C25 405 30 26.18 4.11 0.157
C30 33 37 30.07 3.67 0.122
Overall 1015 25.07 3.61 0.144
Trabzon Region
C14 6 18 10.18 0.71 0.07
C18 139 22 13.87 2.54 0.183
C20 162 25 17.24 2.84 0.165
Overall 307 15.58 2.66 0.171
Erzurum Region
C14 12 18 12.53 2.07 0.165
C18 6 22 17.24 2.47 0.143
C20 135 25 15.14 0.908 0.06
Overall 153 15.02 1.08 0.072
Samsun Region
C14 6 18 22.02 5.15 0.234
C16 45 20 21.01 3.74 0.178
C18 3 22 17.86 6.97 0.39
C20 231 25 22.84 3.81 0.167
C25 3 30 30.33 0.52 0.017
C30 21 37 36.17 4.56 0.126
Overall 309 23.49 3.95 0.168
Bursa Region
C16 17 20 15.73 1.37 0.087
C18 10 22 19.94 0.89 0.045
C20 272 25 20.15 2.18 0.108
C25 138 30 22.57 2.39 0.106
Overall 437 20.74 2.18 0.105
229
Table A.1.b Statistical parameters of 7 day compressive strength data according to
Concrete Number of fck Mean Standard Coefficient of

Class samples (N/mm2) (N/mm2) deviation variation
(N/mm2)
Ankara Region
C16 9 20 19.47 3.12 0.160
C18 8 22 16.63 4.99 0.30
C20 193 25 23.13 2.94 0.127
C25 175 30 27.43 3.46 0.126
C30 143 37 32.71 3.57 0.109
Overall 528 26.99 3.37 0.125
Izmir Region
C20 603 25 26.82 - -
C25 627 30 28.61 - -
C30 159 37 35.39 - -
Overall 1389 28.60 - -
Gaziantep Region
C16 17 20 17.09 0.87 0.051
C20 181 25 26.34 3.08 0.117
C25 3 30 30.33 4.55 0.15
C30 5 37 37.76 261 0.069
Overall 206 - 25.91 2.88 0.111
Denizli Region
C18 41 22 21.18 3.35 0.158
C20 176 25 24.37 3.99 0.164
C25 31 30 27.26 2.67 0.098
Overall 248 24.20 3.75 0.155
Antalya Region
C16 137 20 20.79 2.39 0.115
C18 169 22 19.65 2.69 0.137
C20 345 25 20.92 2.85 0.136
C25 267 30 25.12 2.56 0.102
C30 102 37 28.08 2.13 0.076
Overall 1020 22.51 2.66 0.118
Konya Region
C16 193 20 20.42 3.88 0.190
C18 159 22 21.32 2.96 0.139
C20 216 25 22.53 2.63 0.117
C25 52 30 28.58 3.97 0.139
Overall 620 22.07 3.24 0.147
230
Table A.2.a Statistical parameters of 28 day compressive strength data according to
Concrete Number of fck Mean Standard Coefficient Number of Percentage of

Class samples (N/mm2) (N/mm2) deviation of variation values values under
(N/mm2) under the the limit (%)
limit
Corlu Region
C20 580 25 31.09 3.52 0.113 9 1.5
C25 392 30 34.04 4.02 0.118 18 4.6
C30 21 37 38.73 2.43 0.063 0 0
Overall 993 32.42 3.69 0.114 27 2.7
Trabzon Region
C14 6 18 17.95 1.87 0.104 0 0
C18 43 22 24.74 2.97 0.12 3 6.9
C20 163 25 25.98 1.64 0.063 2 1.2
Overall 212 25.5 1.94 0.076 5 2.4
Erzurum Region
C14 24 18 17.31 1.64 0.095 0 0
C18 37 22 19.21 1.36 0.071 2 5.4
C20 228 25 25.25 2.20 0.087 27 11.8
Overall 289 23.82 2.07 0.087 29 0.100
Kayseri Region
C18 179 22 24.96 2.21 0.10 3 1.7
C20 123 25 27.24 1.77 0.069 1 0.8
Overall 302 25.89 2.25 0.087 4 0.013
Samsun Region
C14 6 18 29.3 8.06 0.275 0 0
C16 45 20 28.67 5.05 0.176 2 4.4
C18 18 22 29.08 6.28 0.216 1 5.5
C20 229 25 30.51 3.94 0.129 10 4.4
C25 6 30 35.93 1.69 0.047 0 0
C30 20 37 39.04 4.02 0.103 2
Overall 324 30.78 0.140 15 4.6
Bursa Region
C16 17 20 24.04 2.57 0.107 0 0
C18 12 22 30.24 2.72 0.09 2 16.7
C20 235 25 28.33 2.21 0.078 3 1.3
C25 156 30 30.42 3.44 0.113 13 8.3
Overall 420 28.99 2.70 0.093 18 0.043
231
Table A.2.b Statistical parameters of 28 day compressive strength data according to
Concrete Number of fck Mean Standard Coefficient Number of Percentag

Class samples (N/mm2) (N/mm2) deviation of variation values e of values
(N/mm2) under the under the
limit limit (%)
Izmir Region
C20 603 25 33.34 3.29 0.099 - -
C25 627 30 35.92 2.72 0.076 - -
C30 159 37 43.33 3.11 0.072 - -
Overall 1389 - 35.64 3.03 0.085 - -
Gaziantep Region
C16 13 20 28.07 4.01 0.143 0 0
C20 170 22 34.2 3.76 0.11 2 1.2
C25 6 25 38.17 8.24 0.216 0 0
C30 10 30 48.12 4.96 0.103 0 0
Overall 199 34.62 3.98 0.115 2 1.0
Denizli Region
C14 9 18 24.89 5.25 0.211 0 0
C18 44 22 28.54 3.97 0.139 0 0
C20 200 25 30.91 4.27 0.138 2 1.0
C25 40 30 32.18 4.11 0.128 2 5.0
C30 6 37 31.78 1.21 0.038 0 0
Overall 299 30.57 4.19 0.137 4 1.3
Antalya Region
C16 137 20 25.22 2.32 0.092 2 1.5
C18 182 22 25.70 2.80 0.109 2 1.1
C20 353 25 27.79 2.58 0.093 6 1.7
C25 264 30 34.31 3.02 0.088 3 1.4
C30 99 37 39.39 1.69 0.043 2 2.0
Overall 1035 29.85 26.87 0.90 15 0.014
232
Table A.2.c Statistical parameters of 28 day compressive strength data according to
Concrete Number fck Mean Standard Coefficient Number of Percentage

Class of (N/mm2) (N/mm2) deviation of values of values
samples (N/mm2) variation under the under the
limit limit (%)
Ankara Region
C16 21 20 23.66 3.12 0.132 0 0
C18 16 22 24.13 4.89 0.203 5 3.1
C20 660 25 29.33 3.20 0.109 35 5.3
C25 217 30 33.05 4.00 0.121 9 4.1
C30 227 37 40.41 3.96 0.098 10 4.4
Overall 1141 32.06 3.53 0.11 59 5.2
Konya Region
C16 432 20 25.24 4.04 0.160 9 2.1
C18 390 22 26.44 3.28 0.124 7 1.8
C20 549 25 26.19 3.06 0.117 18 3.3
C25 126 30 33.98 3.77 0.111 6 4.8
C30 12 37 37.32 4.25 0.114 0 0
Overall 1509 26.72 3.5 0.131 40 2.7
Malatya Region
C14 92 18 19.80 2.34 0.115 1 1.1
C16 90 20 22.68 1.86 0.080 0 0
C20 281 25 31.96 4.34 0.135 3 1.1
C25 55 30 34.58 3.11 0.090 2 3.6
C30 12 37 42.63 3.26 0.065 0 0
Overall 530 28.79 0.116 6 1.1
Istanbul Region
C20 1443 25 24.90 6.63 0.263 - -
C25 878 30 28.63 6.40 0.219 - -
C30 304 37 38.40 6.62 0.174 - -
C35 252 45 38.21 6.85 0.18 - -
Overall 2877 --- 28.63 6.58 0.23 - -
233
APPENDIX B
STATISTICAL PAREMETERS OF BCIII(A) REINFORCING STEEL BARS

USED IN TURKEY
Table B.1. Statistical parameters of BCIII(a) reinforcing steel bars produced by

Habas steel plant
Diameter Number Required Mean value Standard Coefficient
(mm) of yield (N/mm2) deviation of
samples strength (N/mm2) variation
(N/mm2)
10 835 420 546.45 25.65 0.047
12 300 420 511.28 19.21 0.038
16 2174 420 521.42 20.55 0.039
18 30 420 574.83 20.39 0.036
20 1695 420 529.93 19.49 0.037
25 1830 420 531.24 14.24 0.027
32 2329 420 537.75 13.22 0.024
40 406 420 529.18 28.04 0.053
Overall 9619 420 530.01 18.06 0.034

Icdas steel plant
Diameter Number Required Mean Required Mean Required Mean
(mm) of yield value of ultimate value of minimum value of
samples strength yield strength ultimate elongation elongation
(N/mm2) strength (N/mm2) strength (%) (%)
(N/mm2) (N/mm2)
8 100 420 495 500 603 12 21.2
10 100 420 512 500 617 12 18
12 100 420 508 500 601 12 18.3
14 100 420 507 500 616 12 19.2
16 100 420 531 500 640 12 20
18 100 420 523 500 621 12 19.4
20 100 420 518 500 641 12 18
22 100 420 506 500 597 12 18.6
24 100 420 510 500 625 12 18.7
25 100 420 535 500 624 12 17.2
26 100 420 514 500 608 12 17.3
28 100 420 542 500 638 12 17.5
30 100 420 509 500 621 12 16.6
32 100 420 525 500 642 12 16.5
Overall 1400 420 516.8 500 621 12 18.32
234
Colakoglu steel plant
a. Yield strength
Diameter Number of Required Mean value Standard Coefficient of
(mm) samples strength (N/mm2) deviation variation
(N/mm2) (N/mm2)
8 57 420 487.00 31.59 0.065
10 70 420 457.60 14.91 0.033
12 62 420 481.42 24.30 0.050
14 72 420 469.64 20.46 0.044
16 77 420 465.81 19.93 0.043
18 58 420 477.08 20.58 0.043
20 51 420 470.41 23.94 0.051
22 9 420 497.44 18.39 0.040
25 56 420 478.00 30.73 0.064
32 18 420 489.00 23.33 0.047
Overall 530 420 473.63 22.79 0.048
b. Ultimate strength
(N/mm2) (N/mm2)
8 57 500 682.44 62.24 0.091
10 70 500 703.89 30.63 0.044
12 62 500 582.46 46.38 0.079
14 72 500 575.58 38.15 0.066
16 77 500 581.99 26.74 0.046
18 58 500 613.83 48.83 0.071
20 51 500 608.02 44.76 0.074
22 9 500 638.22 19.08 0.030
25 56 500 598.03 45.84 0.076
32 18 500 618.23 26.84 0.043
Overall 530 500 617.95 40.96 0.065
c. Minimum elongation
Diameter Number of Required Mean Standard Coefficient of
(mm) samples elongation value deviation variation
(%) (%) (%)
8 57 12 17.82 1.89 0.106
10 70 12 18.05 1.72 0.095
12 62 12 20.07 1.57 0.078
14 72 12 18.46 1.91 0.104
16 77 12 20.16 2.01 0.100
18 58 12 19.04 1.91 0.100
20 51 12 18.67 2.34 0.125
22 9 12 23.29 0.75 0.032
25 56 12 20.88 1.21 0.058
32 18 12 20.15 1.42 0.071
Overall 530 12 19.25 1.99 0.094
235
Egecelik steel plant
a. Yield strength
(N/mm2) (N/mm2)
12 454 420 489.98 23.85 0.049
14 395 420 485.58 25.24 0.052
16 146 420 501.35 34.33 0.068
18 54 420 487.83 14.81 0.039
20 24 420 485.96 28.92 0.059
Overall 1073 420 489.71 25.45 0.052
(N/mm2) (N/mm2)
12 454 500 632.68 33.93 0.053
14 395 500 626.39 37.92 0.061
16 146 500 640.12 44.18 0.069
18 54 500 625.61 24.11 0.039
20 24 500 651.75 29.24 0.045
Overall 1073 500 631.45 36.19 0.057
(%) (%) (%)
12 454 12 19.00 1.88 0.099
14 395 12 19.26 1.91 0.099
16 146 12 18.84 1.85 0.098
18 54 12 18.69 1.58 0.084
20 24 12 17.38 1.44 0.083
Overall 1073 12 19.02 1.86 0.097
236
Kroman steel plant
a. Yield strength
(N/mm2) (N/mm2)
8 166 420 463 19.52 0.042
10 349 420 463 16.42 0.035
12 478 420 462 16.02 0.035
14 350 420 464 16.33 0.035
16 226 420 453 14.54 0.032
18 57 420 452 14.36 0.032
20 40 420 447 9.73 0.022
22 7 420 439 7.86 0.018
Overall 1673 420 460.71 1.07 0.035
(N/mm2) (N/mm2)
8 166 500 667 33.82 0.051
10 349 500 677 27.30 0.040
12 478 500 680 26.24 0.039
14 350 500 681 26.82 0.039
16 226 500 667 24.44 0.037
18 57 500 666 25.02 0.038
20 40 500 668 18.72 0.028
22 7 500 657 21.39 0.033
Overall 1673 500 675.68 26.85 0.040
(%) (%) (%)
8 166 12 25 3.58 0.141
10 349 12 24 2.52 0.105
12 478 12 24 2.58 0.108
14 350 12 23 2.26 0.097
16 226 12 23 2.13 0.091
18 57 12 23 1.95 0.084
20 40 12 23 2.29 0.102
22 7 12 22 0.88 0.040
Overall 1673 12 23.69 2.50 0.105
237
Ekiciler steel plant
a. Yield strength
(N/mm2) (N/mm2)
8 198 420 485.5 24.25 0.050
10 459 420 480.2 21.98 0.046
12 580 420 478.6 25.34 0.053
14 573 420 476.8 17.03 0.036
16 430 420 485.4 24.88 0.051
18 31 420 493.5 36.85 0.075
20 74 420 490.5 25.1 0.051
22 8 420 483.1 20.67 0.043
24 10 420 505.5 26.87 0.053
25 21 420 469.9 24.9 0.053
32 6 420 501 44.79 0.089
Overall 2390 420 480.94 22.71 0.047
(N/mm2) (N/mm2)
8 198 500 595.0 20.63 0.035
10 459 500 587.7 18.16 0.031
12 580 500 593.2 38.42 0.065
14 573 500 585.9 26.07 0.044
16 430 500 593.4 25.88 0.044
18 31 500 617.7 46.20 0.075
20 74 500 611.4 46.38 0.076
22 8 500 589.0 18.91 0.032
24 10 500 613.6 24.10 0.039
25 21 500 600.0 24.06 0.040
32 6 500 604.0 43.00 0.071
Overall 2390 500 563.81 27.95 0.047
(%) (%) (%)
8 198 12 19.80 1.18 0.059
10 459 12 18.80 1.49 0.079
12 580 12 19.61 1.48 0.075
14 573 12 18.16 1.30 0.072
16 430 12 18.46 1.35 0.073
18 31 12 17.83 2.35 0.132
20 74 12 17.52 2.18 0.124
22 8 12 17.63 0.74 0.042
24 10 12 16.20 4.62 0.285
25 21 12 17.41 0.97 0.056
32 6 12 17.00 1.55 0.091
Overall 2390 12 18.78 1.43 0.076
238
Yesilyurt steel plant
a. Yield strength
(N/mm2) (N/mm2)
8 950 420 480 18.5 0.039
10 270 420 465 17 0.037
12 524 420 461 14.4 0.031
14 551 420 458 15.9 0.035
16 311 420 468 15.7 0.034
18 100 420 455 16.9 0.037
20 72 420 462 15.9 0.034
22 58 420 457 24.2 0.053
24 46 420 456 26.9 0.06
26 142 420 444 16.9 0.038
Overall 3024 420 466.4 16.9 0.036
(N/mm2) (N/mm2)
8 950 500 587 21.5 0.037
10 270 500 579 21.3 0.037
12 524 500 580 21.5 0.037
14 551 500 578 20.6 0.036
16 311 500 588 36.9 0.063
18 100 500 581 33.8 0.058
20 72 500 582 28.4 0.049
22 58 500 679 78.1 0.11
24 46 500 731 42.8 0.058
26 142 500 714 36.3 0.051
Overall 3024 500 593.1 25.6 0.043
(%) (%) (%)
8 950 12 23 1.8 0.078
10 270 12 22.9 1.8 0.079
12 524 12 22.7 1.3 0.057
14 551 12 22.1 1.8 0.081
16 311 12 21.2 2 0.094
18 100 12 20.7 1.9 0.092
20 72 12 21 2 0.095
22 58 12 15.3 3.3 0.22
24 46 12 12.9 1.9 0.147
26 142 12 13.2 1.7 0.129
Overall 3024 12 21.7 1.8 0.081
239
Table B.8. Statistical parameters of BCIII(a) reinforcing steel bars based on data
obtained from the Civil Engineering Materials laboratory of METU (1999)
a. Yield strength
Diameter Number Required Mean Standard Coefficient Number Percentage
(mm) of strength value deviation of of values of values
samples (N/mm2) (N/mm2) (N/mm2) variation under the under the
limit limit (%)
8 166 420 491.43 6758 0.14 0 0
10 349 420 452.29 9231 0.21 21 42.5
12 478 420 486.69 59.62 0.12 5 10.9
14 350 420 477.18 43.72 0.10 3 8.8
16 226 420 481.19 94.07 0.20 9 22.5
18 57 420 443.93 33.96 0.08 7 11.7
20 40 420 495.20 54.52 0.11 0 0
Overall 1673 420 469.34 63.13 0.14 45 17.1
limit limit (%)
8 166 500 698.9 46.73 0.07 0 0
10 349 500 631.72 120.36 0.19 9 18.4
12 478 500 700.46 69.41 0.10 0 0
14 350 500 694.76 59.47 0.09 0 0
16 226 500 648.21 92.40 0.14 6 15
18 57 500 700.34 64.07 0.09 0 0
20 40 500 689.33 96.22 0.14 0 0
Overall 1673 500 678.05 80.83 0.12 15 5.7
(mm) of elongation value deviation of of values of values
samples (%) (%) (%) variation under the under the
limit limit (%)
8 166 12 17.89 2.37 0.13 0 0
10 349 12 20.23 4.56 0.22 1 2
12 478 12 17.61 3.83 0.22 4 8.7
14 350 12 19.15 3.18 0.17 0 0
16 226 12 20.66 4.58 0.22 0 0
18 57 12 18.76 2.78 0.15 0 0
20 40 12 17.61 1.77 0.10 1 5.3
Overall 1673 12 19.04 3.52 0.18 6 2.3
240
a. Yield strength
limit limit (%)
8 40 420 444.35 72.03 0.16 13 33
10 58 420 469.26 52.76 0.11 6 10
12 52 420 499.54 57.49 0.12 1 2
14 43 420 463.84 35.00 0.08 1 2
16 49 420 460.36 60.68 0.13 1 2
18 25 420 487.35 60.79 0.12 3 12
20 16 420 444.33 51.18 0.12 6 37
Overall 283 420 469.13 55.64 0.12 31 11
limit limit (%)
8 40 500 629.93 128.32 0.20 10 25
10 58 500 691.05 85.23 0.12 4 7
12 52 500 721.45 103.53 0.14 0 0
14 43 500 684.71 37.97 0.06 0 0
16 49 500 697.29 102.89 0.14 0 0
18 25 500 695.19 65.77 0.09 0 0
20 16 500 670.85 56.80 0.08 0 0
Overall 283 500 687.34 87.23 0.12 14 5
limit limit (%)
8 40 12 21.09 5.02 0.24 0 0
10 58 12 19.18 4.16 0.22 0 0
12 52 12 16.94 3 0.18 4 8
14 43 12 18.63 1.43 0.08 0 0
16 49 12 19.10 2.76 0.14 3 8
18 25 12 18.45 1.69 0.09 0 0
20 16 12 22.13 3.18 0.14 0 0
Overall 283 12 19.04 3.14 0.16 7 2.5
241
a. Yield strength
limit limit (%)
8 31 420 482.81 23.06 0.05 0 0
10 49 420 478.66 51.02 0.11 4 8.2
12 91 420 504.79 93.30 0.18 10 11
14 42 420 475.17 35.45 0.07 0 0
16 55 420 458.09 73.12 0.16 7 13
18 13 420 444.30 46.97 0.11 3 23.1
20 70 420 505.75 59.51 0.12 0 0
Overall 351 420 486.29 62.65 0.13 24 6.8
limit limit (%)
8 31 500 718.17 28.82 0.04 0 0
10 49 500 689.69 60.28 0.09 0 0
12 91 500 682.12 94.02 0.14 8 8.8
14 42 500 708.78 38.44 0.05 0 0
16 55 500 671.37 94.78 0.14 7 12.7
18 13 500 711.4 70.73 0.10 0 0
20 70 500 682.02 46.79 0.07 0 0
Overall 351 500 688.94 66.74 0.10 15 4.3
limit limit (%)
8 31 12 19.12 4.05 0.21 2 6.5
10 49 12 17.5 3.81 0.22 2 4.1
12 91 12 16.56 2.7 0.16 3 3.3
14 42 12 18.51 2.01 0.11 0 0
16 55 12 18.76 2.14 0.11 1 1.8
18 13 12 18.01 1.52 0.08 0 0
20 70 12 19.1 2.24 0.12 0 0
Overall 351 12 18.06 2.67 0.15 8 2.2
242
a. Yield strength
limit limit (%)
8 52 420 508.69 51.54 0.10 1 1.9
10 50 420 512.08 59.48 0.12 1 2
12 61 420 513.86 55.78 0.11 0 0
14 37 420 507.23 47.98 0.09 3 8.1
16 67 420 499.05 36.91 0.07 0 0
18 31 420 521.07 31.41 0.06 0 0
20 41 420 527.61 55.09 0.10 0 0
Overall 339 420 511.48 48.78 0.09 5 1.5
limit limit (%)
8 52 500 674.13 58.19 0.09 1 1.9
10 50 500 677.12 32.13 0.05 0 0
12 61 500 683.05 54.69 0.08 0 0
14 37 500 659.03 48.48 0.07 0 0
16 67 500 657.57 57.57 0.09 0 0
18 31 500 667.58 51.57 0.08 0 0
20 41 500 684.90 53.22 0.08 0 0
Overall 339 500 671.99 51.33 0.08 1 0.3
limit limit (%)
8 52 12 18.31 3.59 0.2 3 5.9
10 50 12 16.24 3.22 0.2 4 8
12 61 12 16.46 2.33 0.14 2 3.3
14 37 12 18.48 2.37 0.13 1 2.7
16 67 12 18.51 1.7 0.09 0 0
18 31 12 17.51 1.4 0.08 0 0
20 41 12 16.83 2.09 0.12 1 2.4
Overall 339 12 17.48 2.42 0.14 11 3.2
243
a. Yield strength
limit limit (%)
8 30 420 479.65 66.88 0.14 4 13
10 54 420 484.72 57.04 0.12 4 7
12 72 420 499.92 60.65 0.12 2 3
14 51 420 458.38 50.88 0.11 6 12
16 75 420 502.38 68.09 0.14 0 0
18 14 420 505.25 74.81 0.15 0 0
20 68 420 491.70 55.28 0.11 3 4
22 14 420 473.79 33.09 0.07 1 7
Overall 378 420 488.77 59.32 0.12 20 5
limit limit (%)
8 30 500 680.77 99.38 0.15 2 7
10 54 500 694.51 59.23 0.09 0 0
12 72 500 700.79 47.58 0.07 0 0
14 51 500 661.94 89.58 0.14 3 6
16 75 500 702.81 94.89 0.14 0 0
18 14 500 722.88 100.03 0.14 0 0
20 68 500 714.72 87.39 0.12 0 0
22 14 500 655.18 64.33 0.10 0 0
Overall 378 500 695.10 78.13 0.12 5 1
limit limit (%)
8 30 12 18.35 4.84 0.26 3 10
10 54 12 16.87 2.98 0.18 2 4
12 72 12 16.72 3.22 0.19 1 1
14 51 12 18.41 3.69 0.20 1 2
16 75 12 16.77 2.82 0.17 5 7
18 14 12 19.65 2.49 0.13 0 0
20 68 12 19.84 2.32 0.14 3 4
22 14 12 18.76 1.72 0.09 0 0
Overall 378 12 17.85 3.06 0.18 15 4
244
a. Yield strength
limit limit (%)
8 27 420 510.01 58.34 0.11 0 0
10 28 420 495.35 61.81 0.12 4 14
12 53 420 520.58 60.08 0.12 0 0
14 16 420 459.14 20.38 0.05 0 0
16 30 420 493.62 33.18 0.07 0 0
20 39 420 514.09 46.01 0.09 0 0
22 5 420 483.40 78.78 0.16 0 0
Overall 198 420 504.30 50.50 0.10 4 2
limit limit (%)
8 27 500 695.31 45.88 0.07 0 0
10 28 500 707.47 73.65 0.10 0 0
12 53 500 683.30 50.97 0.07 0 0
14 16 500 663.09 87.21 0.13 0 0
16 30 500 692.94 38.89 0.06 0 0
20 39 500 708.11 60.27 0.09 0 0
22 5 500 690.3 6.61 0.01 0 0
Overall 198 500 693.24 55.29 0.08 0 0
limit limit (%)
8 27 12 14.97 4.12 0.28 6 22
10 28 12 17.23 3.29 0.19 1 4
12 53 12 16.72 2.33 0.14 0 0
14 16 12 19.71 2.93 0.15 0 0
16 30 12 17.86 2.97 0.17 0 0
20 39 12 16.38 1.85 0.11 0 0
22 5 12 16.48 1.04 0.06 0 0
Overall 198 12 16.89 2.72 0.16 7 4
245
obtained from the Civil Engineering Materials laboratory of ITU (1999)
a. Yield strength
limit limit (%)
8 59 420 434.89 85.66 0.17 23 32.2
10 76 420 519.87 84.42 0.16 8 9.2
12 78 420 500.24 85.48 0.17 12 15.4
14 83 420 493.34 71.20 0.14 11 14.5
16 78 420 494.65 70.95 0.14 9 10.3
18 43 420 518.29 76.38 0.15 5 11.6
20 96 420 523.72 74.61 0.14 8 6.3
22 70 420 520.36 52.61 0.10 1 1.4
Overall 583 420 502.07 74.95 0.15 77 12
limit limit (%)
8 59 500 596.29 99.4 0.17 12 20.3
10 76 500 658.10 80.80 0.12 5 6.6
12 78 500 631.78 88.86 0.14 8 10.3
14 83 500 606.28 78.68 0.13 7 8.4
16 78 500 621.97 67.00 0.11 2 2.6
18 43 500 641.12 59.08 0.09 2 4.7
20 96 500 652.92 68.18 0.10 4 4.2
22 70 500 677.55 59.00 0.09 0 0
Overall 583 500 636.34 75.31 0.12 40 6.9
limit limit (%)
8 59 12 22.86 5.80 0.25 1 1.7
10 76 12 18.50 3.81 0.2 1 1.3
12 78 12 19.47 4.33 0.22 0 0
14 83 12 19.21 3.84 0.2 1 1.2
16 78 12 18.96 3.80 0.2 1 1.3
18 43 12 17.60 3.55 0.2 1 2.3
20 96 12 18.01 2.79 0.15 1 1
22 70 12 17.17 1.75 0.1 0 0
Overall 583 12 18.93 3.65 0.19 6 1
246
a. Yield strength
limit limit (%)
8 82 420 460.12 77.45 0.17 24 29.3
10 78 420 506.26 77.17 0.15 9 11.5
12 79 420 509.34 68.7 0.13 5 6.3
14 77 420 495.13 43.21 0.09 1 1.3
16 76 420 493.43 56.74 0.11 7 9.2
18 63 420 491.90 65.78 0.14 5 7.9
20 78 420 524.03 46.81 0.09 2 2.6
22 61 420 527.2 41.41 0.08 0 0
Overall 594 420 500.18 60.19 0.12 53 8.9
limit limit (%)
8 82 500 614.25 101.48 0.17 11 13.4
10 78 500 654.68 69.34 0.11 4 5.1
12 79 500 631.53 61.62 0.10 1 1.3
14 77 500 625.71 53.84 0.09 2 2.6
16 76 500 620.65 60.83 0.10 1 1.3
18 63 500 621.09 67.84 0.11 3 4.8
20 78 500 638.34 64.17 0.10 2 2.6
22 61 500 634.72 52.15 0.08 0 0
Overall 594 500 630.15 67.05 0.11 24 4.04
limit limit (%)
8 82 12 22.63 5.35 0.24 2 2.4
10 78 12 18.85 4.55 0.23 0 0
12 79 12 19.91 3.74 0.19 0 0
14 77 12 19.41 1.95 0.10 0 0
16 76 12 18.81 2.75 0.15 0 0
18 63 12 19.11 3.16 0.17 0 0
20 78 12 18.05 2.57 0.14 0 0
22 61 12 18.22 1.85 0.10 0 0
Overall 594 12 19.44 3.3 0.17 2 0.3
247
a. Yield strength
limit limit (%)
8 86 420 488 76.89 0.16 16 18.6
10 85 420 510.06 73.52 0.14 9 10.6
12 96 420 524.26 51.51 0.10 0 0
14 77 420 505.83 57.24 0.11 5 6.5
16 88 420 502.98 74.29 0.15 7 8
18 49 420 508.61 60.42 0.12 2 4.1
20 64 420 554.5 55.58 0.10 0 0
22 56 420 508.93 43.06 0.08 1 1.9
Overall 601 420 512.11 62.70 0.12 40 6.7
limit limit (%)
8 86 500 631.41 81.29 0.13 5 5.81
10 85 500 649.73 60.68 0.09 2 2.4
12 96 500 640.57 61.40 0.08 0 0
14 77 500 630.73 50.77 0.08 1 1.3
16 88 500 649.54 65.69 0.10 3 3.4
18 49 500 649.41 52.85 0.08 0 0
20 64 500 676.23 53.82 0.08 0 0
22 56 500 647.89 42.08 0.08 0 0
Overall 601 500 645.81 60.11 0.09 11 1.8
limit limit (%)
8 86 12 21.37 5.31 0.25 1 1.2
10 85 12 19.04 4.33 0.23 1 1.2
12 96 12 18.41 2.59 0.14 0 0
14 77 12 19.13 2.33 0.12 0 0
16 88 12 19.07 3.78 0.2 0 0
18 49 12 18.53 2.77 0.15 0 0
20 64 12 16.59 2.87 0.17 0 0
22 56 12 18.32 2.45 0.14 0 0
Overall 601 12 18.92 3.4 0.18 2 0.3
248
a. Yield strength
limit limit (%)
8 42 420 468.71 74.21 0.16 9 21.4
10 87 420 494.74 59.72 0.12 3 3.4
12 81 420 500.93 56.73 0.11 2 2.5
14 80 420 480.64 42.81 0.09 2 2.5
16 80 420 474.41 58.79 0.12 12 15
18 51 420 492.56 41.22 0.08 2 3.9
20 81 420 510.61 36.37 0.07 0 0
22 56 420 492.09 34.05 0.07 0 0
Overall 558 420 490.58 50.16 0.10 30 5.4
limit limit (%)
8 42 500 614.67 87.73 0.14 6 14.3
10 87 500 689.84 67.52 0.10 0 0
12 81 500 651.32 65.51 0.10 2 2.5
14 80 500 633.56 56.71 0.09 2 2.5
16 80 500 635.71 57.38 0.09 1 1.3
18 51 500 637.9 62.08 0.10 2 3.9
20 81 500 644.25 45.34 0.07 0 0
22 56 500 631.07 51.28 0.08 0 0
Overall 558 500 645.49 60.40 0.09 13 2.3
limit limit (%)
8 42 12 21.26 3.91 0.18 0 0
10 87 12 18.29 2.49 0.14 0 0
12 81 12 19.34 3.14 0.16 0 0
14 80 12 19.85 2.16 0.11 1 1.3
16 80 12 19.76 2.42 0.12 0 0
18 51 12 19.26 2.40 0.12 1 2
20 81 12 18.89 2.89 0.15 1 1.2
22 56 12 18.98 1.81 0.10 0 0
Overall 558 12 19.34 2.62 0.13 3 0.5
249
a. Yield strength
limit limit (%)
8 55 420 461.24 72.18 0.16 15 27.3
10 57 420 515.91 52.44 0.10 0 0
12 60 420 500.65 85.21 0.17 0 0
14 54 420 512.96 51.59 0.10 0 0
16 55 420 494.57 43.86 0.09 0 0
18 31 420 482.87 56.82 0.12 1 3.2
20 54 420 501.39 52.57 0.11 2 3.7
22 26 420 497.16 62.35 0.13 2 7.7
Overall 392 420 496.65 59.93 0.12 20 5.1
limit limit (%)
8 55 500 640.31 87.86 0.14 6 10.9
10 57 500 688.17 16.85 0.10 0 0
12 60 500 666.57 61.52 0.09 0 0
14 54 500 668.04 59.85 0.09 0 0
16 55 500 667.31 47.38 0.07 0 0
18 31 500 638.10 52.18 0.08 0 0
20 54 500 648.26 54.45 0.08 0 0
22 26 500 629.00 62.63 0.10 0 0
Overall 392 500 659.07 54.87 0.09 6 1.5
limit limit (%)
8 55 12 21.85 4.49 0.21 0 0
10 57 12 18.81 2.39 0.13 0 0
12 60 12 20.08 2.32 0.12 0 0
14 54 12 18.89 2.38 0.13 0 0
16 55 12 19.27 1.88 0.10 0 0
18 31 12 19.93 2.21 0.11 0 0
20 54 12 19.17 2.42 0.13 0 0
22 26 12 20.44 2.47 0.12 0 0
Overall 392 12 19.75 2.6 0.13 0 0
250
a. Yield strength
limit limit (%)
8 29 420 480.21 40.94 0.085 1 3.5
10 57 420 479.52 40.97 0.085 2 3.5
12 50 420 491.24 47.54 0.097 1 2
14 54 420 458.35 41.41 0.09 8 14.8
16 42 420 495.6 49.52 0.10 0 0
18 40 420 467.28 35.56 0.08 6 15
20 47 420 501.21 39.93 0.08 3 6.4
22 35 420 502.38 28.98 0.058 0 0
Overall 354 420 483.66 41.04 0.085 21 5.9
limit limit (%)
8 29 500 657.96 78.83 0.12 1 3.4
10 57 500 655.18 45.04 0.07 0 0
12 50 500 681.03 65.55 0.096 1 2
14 54 500 644.4 58.95 0.09 0 0
16 42 500 663.59 60.88 0.09 0 0
18 40 500 619.08 44.88 0.07 0 0
20 47 500 629.65 52.32 0.08 3 6.4
22 35 500 631.50 26.93 0.04 0 0
Overall 354 500 648.6 53.86 0.08 5 1.4
limit limit (%)
8 29 12 21.21 2.96 0.14 0 0
10 57 12 20.44 2.25 0.11 0 0
12 50 12 19.10 2.58 0.14 0 0
14 54 12 20.5 2.08 0.10 0 0
16 42 12 18.86 2.30 0.12 0 0
18 40 12 20.13 2.72 0.14 0 0
20 47 12 19.77 2.67 0.14 0 0
22 35 12 19.64 1.52 0.08 0 0
Overall 354 12 19.93 2.37 0.12 0 0
251
obtained from different materials laboratories in Konya
a. Yield strength
limit limit (%)
8 68 420 468.27 58.13 0.124 10 14.71
10 47 420 478.42 62.50 0.131 3 6.38
12 64 420 466.31 69.57 0.149 10 15.63
14 72 420 470.00 64.74 0.138 6 8.33
16 45 420 479.86 55.31 0.115 6 13.33
18 23 420 473.52 48.28 0.101 2 8.69
20 14 420 514.14 108.31 0.211 1 7.14
Overall 333 420 473.56 63.43 0.134 38 11.41
limit limit (%)
8 68 420 693.72 73.96 0.107 1 1.47
10 47 420 710.36 72.75 0.102 2 4.26
12 64 420 677.61 86.35 0.127 3 4.69
14 72 420 701.18 100.44 0.143 3 4.17
16 45 420 691.93 48.21 0.070 0 0
18 23 420 686.61 51.79 0.075 0 0
20 14 420 714.00 113.21 0.158 0 0
Overall 333 420 694.71 78.54 0.113 9 2.7
limit limit (%)
8 68 420 27.26 12.51 0.459 1 1.47
10 47 420 25.24 8.67 0.343 0 0
12 64 420 24.67 7.96 0.323 0 0
14 72 420 23.39 5.86 0.250 0 0
16 45 420 22.17 3.92 0.177 0 0
18 23 420 21.36 3.64 0.170 0 0
20 14 420 22.21 4.64 0.209 0 0
Overall 333 420 24.33 7.55 0.303 1 0.3
252
APPENDIX C
ANNUAL MAXIMUM SNOW DEPTHS FOR DIFFERENT LOCATIONS
The data given in Table C.1 are obtained from Turkish Meteorological Department
Table C.1.a. Annual maximum snow depths for different locations (cm)
CANAKKALE
GAZIANTEP
ERZINCAN
MALATYA
/GÖZTEPE
ISTANBUL
ISTANBUL
ANTALYA
HAKKARI
ANKARA
SAMSUN
YEARS
BURSA
IZMIR
/SİLE
1947 18 * * * * 8 * 13 * * 12 *
1948 20 * * * 5 * * 14 * * 17 *
1949 20 * * * 16 20 * 17 * * 15 *
1950 * * 16 37 15 * * 25 17
1951 7 * * 22 23 24 14 * 5 2
1952 8 * * 8 35 40 7 * 6
1953 25 * * 35 26 10 23 2 * 14 20
1954 19 4 * * 16 30 36 46 5 * 9 28
1955 5 * * 7 * 13 8 5 *
1956 10 * * 11 25 29 25 24 * 18 13
1957 8 * 11 50 29 16 * 13
1958 * * 23 10 15 44 3 *
1959 8 * 2 25 21 33 29 10 * 3 2
1960 11 1 8 8 * 17 11 * 3
1961 15 35 16 49 54 23 97 8 32
1962 7 12 16 24 15 74 7
1963 9 26 4 6 13 7 10 52 15 22
1964 12 14 47 47 26 27 5 136 12 6
1965 15 22 20 5 15 26 3 167 11 18
1966 18 14 3 9 5 57 3
1967 10 22 18 32 37 26 12 204 15 22
1968 12 23 3 50 29 6 87 15 19
1969 12 14 11 5 26 24 3 216 19 16
1970 14 2 3 7 89 3 4
1971 9 8 4 30 20 12 90 7 5
1972 12 20 40 5 17 12 6 190 24
1973 14 11 21 4 25 11 6 89 21 20
1974 7 15 14 31 18 5 * 7 10
253
Table C.1.b. Annual maximum snow depths for different locations
CANAKKALE
GAZIANTEP
ERZINCAN
MALATYA
ISTANBUL
/GÖZTEPE
ISTANBUL
ANTALYA
HAKKARI
ANKARA
SAMSUN
BURSA
IZMIR
YEAR
/SİLE
1975 10 13 30 13 14 19 81 11
1976 24 11 15 36 25 30 2 114 25 8
1977 10 * 22 3 17 16 8 98 20
1978 6 10 14 21 30 98
1979 13 23 24 3 21 25 79 10 4
1980 22 23 15 16 19 17 100 18 13
1981 15 2 10 6 4 38 90 6 15
1982 4 50 9 23 7 11 80 19 6
1983 25 1 45 10 18 12 10 6 123 26 14
1984 4 * 9 2 35 9 51
1985 30 20 12 14 15 55 6 80 25 55
1986 12 1 12 2 9 41 33 10 70 4 14
1987 20 21 1 12 6 13 19 24 139 44 40
1988 3 9 6 15 28 5 160 3
1989 8 2 25 4 61 26 71 5 3
1990 6 8 22 22 14 62 8
1991 14 3 10 12 22 16 25 10 62 12 10
1992 16 41 28 24 48 35 10 260 21 18
1993 16 4 20 8 32 20 10 246 11 6
1994 13 13 4 19 29 30 4 79 17 9
1995 3 9 4 23 31 84 14
1996 10 3 21 4 16 6 7 86 4 7
1997 7 24 10 4 15 8 70 10 13
1998 7 12 4 2 21 16 9 50 4 2
1999 6 34 2 21 9 4 22 2
2000 28 9 26 10 38 14 108 12 19
2001 13 13 15 4 18 8 32 94 40 6
2002 30 23 22 20 33 17 3 71 22 12
2003 8 27 47 5 42 30 7 184 33 34
2004 7 66 12 11 22 17 36 111 32 28
2005 6 18 4 12 6 8 80 23 8
Aver 12.7 2 18.8 1.5 16.0 14.1 24.9 20.7 10.6 105.7 14.6 14.6
Std. 6.86 1.26 13.3 0.71 10.4 12.4 12.6 10.8 9.10 53.37 9.58 11.2
C.o. 0.54 0.63 0.71 0.47 0.66 0.88 0.52 0.51 0.85 0.50 0.66 0.76
Max 30 4 66 2 47 49 61 55 38 260 44 55
Note: In this table, “*” shows that no data that belongs to that year has been obtained.
254
Table C.2. The Chi-Square (χ2) and Kolmogorov- Smirnov (K-S) tests results for
annual snow depths
Annual snow depth

Tests Tests
Prob. χ2 P value K-S Prob. χ2 P value K-S
Dist. Dist.
Ankara Erzincan
Lognormal 0.667 0.9996 0.0909 Rayleigh 2.842 0.9439 0.0649
Weibull 4.667 0.7295 0.0765 Normal 4.737 0.7855 0.0736
Ext. value 9.00 0.3426 0.1145 Lognormal 5.053 0.7519 0.0844
Rayleigh 12.00 0.1512 0.1357 Weibull 5.368 0.7196 0.0779
Normal 14.00 0.0815 0.1208
Bursa Canakkale
Lognormal 11.33 0.1247 0.0927 Lognormal 11.14 0.1326 0.0940
Gamma 11.33 0.1247 0.1053 Ext. value 7.791 0.3514 0.1570
Ext. value 7.143 0.4142 0.1258 Rayleigh 15.60 0.029 0.2613
Weibull 10.57 0.1584 0.1204 Hakkari
Rayleigh 15.14 0.0342 0.173 Ext. value 6.909 0.4384 0.1489
Gaziantep Lognormal 10.91 0.1426 0.1418
Ext. value 2.667 0.9535 0.08373 Gamma 13.09 0.0699 0.1588
Weibull 6.00 0.6472 0.07035 Weibull 18.55 0.0097 0.1669
Rayleigh 8.667 0.3712 0.1259 Göztepe
Normal 11.00 0.2017 0.1299 Rayleigh 2.868 0.9424 0.1025
Samsun Ext. value 4.226 0.8326 0.0766
Rayleigh 18.82 0.0088 0.2369 Weibull 6.264 0.6177 0.0741
Normal 34.82 1.2E-4 0.1888 Normal 17.81 0.0227 0.1143
Ext. value 38.38 2.6E-6 0.1992 Sile
Malatya Lognormal 4.095 0.7687 0.1199
Lognormal 11.36 0.1820 0.1124 Ext. value 5.238 0.6309 0.1007
Ext. value 7.962 0.4370 0.1385 Rayleigh 6.742 0.4541 0.1153
Weibull 13.06 0.1099 0.1081 Normal 22.76 0.0019 0.1047
Normal 16.45 0.0363 0.1501
Rayleigh 17.47 0.0256 0.1678
Note 1: At the significance level α=5%, 1-α percentile value of the Chi-Square test, χ.95,5=11.1 for Bursa,
Samsun, Canakkale, Hakkari and Sile; and , χ.95,6=12.6 for Ankara, Gaziantep, Malatya, Erzincan and
Göztepe.
.05
Note 2: The critical value of K-S test at the 5% significance level, D 50 =0.19.
255
APPENDIX D
ANNUAL MAXIMUM WIND SPEEDS FOR DIFFERENT LOCATIONS
The data given in Table D.1 are obtained from Turkish Meteorological Department
Table D.1.a. Annual maximum wind speeds for different locations (m/sn)
CANAKKALE
GAZİANTEP
/ISTANBUL
/ISTANBUL
ERZİNCAN
MALATYA
ANTALYA
HAKKARİ
GÖZTEPE
ANKARA
SAMSUN
YEARS
BURSA
IZMIR
SİLE
1949 12.9
1950 12.1 12 19.7
1951 11.7 11.4 15.6 19.2
1952 15.2 15 18 18.8
1953 15 8.4 21
1954 15 20.4 22.8
1955 18.7 15.2 27.6 24.6
1956 28.9 21.2 18.6 15.6 23 28
1957 28.9 20.8 20.4 20.1 13 21.1 17.2
1958 31.2 21 19.6 20.7 8.5 23.3 14.5
1959 23.8 21.2 19.2 20.4 12.1 21.6 18.5
1960 22.7 29.3 23 27.3 10.5 21.5 18.5
1961 24.8 28.9 23.8 28.5 20.5 21.5 16 15.2 23.2 13.9 15.6 26.6
1962 25.4 27 24.2 25.5 22.5 25.8 26.8 14.3 28.5 20.7 24 20.2
1963 28.3 27.5 25.2 25.5 19.6 18.3 23.2 15.6 28.2 13.3 17.2
1964 19.1 27.5 28.6 20.7 16.2 20.2 14.9 24.2 29.5 11.7 18.4
1965 29.6 27.4 31.1 30.6 24.1 28.8 17.2 30.4 30.3 30.2 13.3
1966 23 24.8 32.2 35.6 23 24 24.7 28 29 25.3 12.2 9.3
1967 23.2 21.2 28.2 29.5 24.9 25.5 21 27.9 31 22.7 10 14.4
1968 32.1 27.8 30 36.9 31 31.2 21.1 30.9 33.7 30.2 0
1969 24.4 28.8 29.6 38.7 22.1 23.9 25 26.2 34.1 27.4 10
1970 28.5 33.9 29 34.6 22.2 20.2 21.5 39.5 35.4 25 11.1
1971 29.2 32.5 31.2 36 24.4 19.1 18.9 25.8 33.2 26 10 19.4
1972 25.1 31.7 25.6 28.7 23 19.2 6.4 21 28 21 24.5 36
1973 23.2 28.9 29.1 32.4 17.9 27.8 7.8 23 31.8 22.3 27.7 42.4
1974 23.3 30 26 31.2 19.9 27 17.7 19.6 32 24.1 25.5 32.9
1975 17.8 26.8 26 29.8 23.4 30 22.6 22.8 25.9 25.6 27.4
1976 22 26 29.5 28.5 20.2 31.1 5 17.6 32.7 20 23.9 32.9
256
Table D.1.b. Annual maximum wind speeds for different locations (m/sn)
CANAKKALE
GAZİANTEP
/ISTANBUL
/ISTANBUL
ERZİNCAN
MALATYA
ANTALYA
HAKKARİ
GÖZTEPE
ANKARA
SAMSUN
YEARS
BURSA
IZMIR
SİLE
1977 16.9 27.3 26.7 26.5 17.7 28.2 5.6 16.5 32 21 24 35.2
1978 17.6 25.4 27.7 25.8 18 34.5 10 20.6 31.8 24.2 20 35
1979 16.5 25.9 29.9 25.5 16.7 31.2 21 17.6 34.5 25.2 21 39.5
1980 14.7 27 26 28.6 16.1 29.6 8.7 16.9 35.2 21.6 20.1 34.1
1981 17.7 27.8 30.7 22.4 10.6 34.2 4.8 17.3 28.5 21 21.7 36.3
1982 17 25 29.2 24 13.8 25.7 6.2 16.2 29 18.8 18.9 36.2
1983 17.2 21.5 23.3 22 13.6 27 4.6 15 31.8 16 19.6 27.4
1984 24.1 26.9 18.5 23.8 15.4 24 11.6 14.6 26.7 15.8 15.8 24.7
1985 22.2 24.2 20.6 23.6 16.1 26.9 8 13.8 29.4 17.1 25.8 29.6
1986 19.2 25.3 18.6 27.6 12.3 23.8 5.6 16.9 24 20 24 28.5
1987 18.2 29.1 22.9 22 13.3 23.1 3.8 19 35 20.7 22.7 28
1988 16.7 24.4 21.6 22.3 14.8 24.8 26.8 15.6 26.8 17.9 21.7 28.6
1989 15.8 25 22 19.8 13.7 26.9 3 16.2 32.6 18.6 25.1 28
1990 14.4 20 18.5 22.7 13.7 20 3.4 15 26.6 16.7 25.2 25.4
1991 15.3 22.3 22.7 20.6 13.6 15.2 21.3 17 38.7 13.4 27 28.3
1992 13.5 23.4 21.6 20.3 16.3 19.6 17.5 19.1 33.9 13.5 26.3 25.9
1993 13.5 22.1 19.8 22.1 14.6 23.2 15.2 14.2 31.8 15 21.2 23.3
1994 14.5 19 16 21.1 16.7 22 26.2 16.1 24.2 16.4 20.2 25.8
1995 18.1 26.8 19.1 21.9 13.7 25.2 17 16.5 32.7 20.5 21.5 28.4
1996 17.4 24.3 20.5 28.6 15.1 22.4 15.9 17.1 31.5 19.8 23.8 24
1997 17.7 23.5 17.2 25 15 26.3 19.6 15.7 27.6 18.3 19.2 27.5
1998 16.3 21.8 15.2 43.2 14.1 23.4 19.1 18.9 31 16.3 18.7 22.8
1999 20.2 23.8 21.3 22.6 16.6 27.1 17.4 16.6 36.2 14.5 20.4 26.2
2000 17.7 27.2 18.8 23.1 14.9 24.9 21.4 16.9 34.5 13.6 21.1 24.9
2001 19.4 29.2 22.2 25.5 15.6 20.3 20 20 29.9 18.5 20.5 22.2
2002 19.4 19.6 17.1 27.8 13.7 21.6 20.6 14.8 28 18.3 18.8 17.1
2003 19.3 29.8 23.1 30.8 11.3 22.4 19.5 17.2 28.7 19.5 25.3 23.1
2004 16.8 25.4 17.9 28.4 13.3 24.5 19.8 15.7 28.2 20.6 16.9 17.9
2005 17 23.6 19.2 23.8 13.4 22.4 20 17.9 25.8 20.9 18.6 19.2
Average 20.13 26.19 23.92 26.62 17.53 23.11 15.09 19.07 28.99 20.19 19.67 25.21
Std. dev 4.65 3.33 4.74 5.36 4.28 5.26 6.87 5.50 4.88 4.31 5.86 7.03
C.o.v. 0.23 0.13 0.20 0.20 0.24 0.23 0.46 0.29 0.17 0.21 0.30 0.28
Max. 32.10 33.90 32.20 43.20 31.00 34.50 26.80 39.50 38.70 30.20 27.70 42.40
257
Table D.2. The Chi-Square (χ2) and Kolmogorov- Smirnov (K-S) tests results for
annual wind speeds
Annual wind speed
Tests Tests
Dist. Dist.
Ankara Malatya
Ext. value 8.957 0.2558 0.5953 Normal 9.333 0.3150 0.1013
Normal 9.304 0.2315 0.1244 Weibull 9.333 0.3150 0.1074
Rayleigh 10.00 0.1886 0.8595 Rayleigh 13.67 0.0909 0.1541
Weibull 11.74 0.1095 0.4555 Ext. value 17.00 0.0301 0.1587
Lognormal 12.93 0.0871 0.4513 Erzincan
Izmir Lognormal 12.43 0.0871 0.1496
Normal 4.080 0.7705 0.0929 Ext. value 13.48 0.0613 0.1617
Weibull 4.080 0.7705 0.0884 Gamma 12.43 0.0871 0.1641
Ext value 7.600 0.3692 0.1296 Weibull 16.26 0.0228 0.1607
Bursa Rayleigh 23.91 0.0012 0.1999
Ext value 5.478 0.6018 0.0935 Normal 32.96 2.7x10-5 0.2149
Rayleigh 2.696 0.9117 0.0988 Canakkale
Weibull 2.696 0.9117 0.1014 Normal 4.667 0.7925 0.0919
Lognormal 4.087 0.7697 0.1030 Weibull 7.00 0.5366 0.0682
Normal 5.478 0.6018 0.0997 Ext. value 19.33 0.0132 0.1512
Antalya Hakkari
Weibull 3.408 0.8449 0.0727 Lognormal 4.442 0.7277 0.1220
Rayleigh 4.714 0.6948 0.1273 Normal 4.442 0.7277 0.1229
Ext value 6.020 0.5374 0.1010 Rayleigh 8.163 0.3185 0.0853
Gaziantep Göztepe
Ext. value 9.612 0.2116 0.0869 Normal 15.62 0.0288 0.1607
Rayleigh 8.633 0.2801 0.1160 Sile
Normal 10.92 0.1422 0.1484 Weibull 11.65 0.16.77 0.0969
Samsun Normal 14.47 0.0703 0.0990
Normal 9.333 0.3150 0.1074 Gamma 15.18 0.0558 0.0815
Rayleigh 13.67 0.0909 0.1541 Lognormal 15.18 0.0558 0.0808
Ext. value 17.00 0.0301 0.1587
Note 1: At the significance level α=5%, 1-α percentile value of the Chi-Square test, χ.95,5=11.1 for
Ankara, Izmir, Bursa, Antalya, Gaziantep, Erzincan, Hakkari and Göztepe; and , χ.95,6=12.6 for Samsun,
Malatya, Sile and Bursa.
.05
Note 2: The critical value of K-S test at the 5% significance level, D 50 =0.19.
258
APPENDIX E
DAILY WIND SPEEDS OBSERVED IN 2004 IN DIFFERENT LOCATIONS
The data in Tables E.1-E12 are obtained from Turkish Meteorological Department
Table E.1. Daily Maximum Wind Speeds observed during the year 2004 in Ankara
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JULY
JUNE
MAY
Days
1 6.7 10.8 7.2 6.7 4.6 9.8 10.3 10.3 10.3 7.7 4.4 4.9
2 8.2 6.7 7.2 11.3 8.8 4.6 7.5 11.3 14.4 8.0 8.8 5.2
3 7.2 5.2 6.7 11.3 9.8 9.0 8.8 9.3 9.8 7.0 10.3 4.1
4 10.3 10.3 6.2 5.7 7.2 6.2 12.4 7.2 10.8 10.0 9.3 4.6
5 11.8 5.7 10.8 5.7 10.3 11.8 10.3 7.7 11.6 9.8 7.2 7.7
6 8.8 6.2 9.3 7.2 8.8 13.9 8.8 6.7 11.3 10.8 5.9 5.2
7 8.8 10.3 6.7 10.3 12.9 7.2 10.8 12.4 11.8 9.0 5.4 9.3
8 6.2 13.9 8.2 8.8 6.7 10.8 13.9 10.8 7.2 7.2 6.2 5.9
9 6.2 10.8 7.0 6.2 7.2 13.9 11.8 9.3 11.8 6.2 6.2 6.4
10 9.8 11.3 13.4 5.7 11.6 7.2 8.2 11.8 9.8 6.7 5.7 4.6
11 12.9 7.7 8.8 5.7 11.8 5.2 7.7 10.3 8.2 6.4 4.1 5.4
12 8.8 16.5 7.2 6.7 9.3 4.1 10.3 10.8 6.7 14.4 3.6 4.1
13 5.7 18.5 9.8 6.2 9.0 6.2 10.3 5.4 5.7 11.8 4.1 5.9
14 5.2 10.3 10.8 19.1 11.8 11.8 9.0 7.2 5.7 8.2 5.2 10.8
15 11.8 10.0 6.2 10.8 9.8 10.3 11.3 13.4 5.2 4.6 7.7 7.7
16 13.4 10.3 9.8 14.9 7.7 6.7 11.8 15.2 5.7 7.7 15.5 5.4
17 7.2 7.2 8.2 3.6 11.6 8.8 11.8 9.5 10.3 8.2 9.8 4.1
18 6.7 9.3 10.8 7.7 11.3 11.6 9.3 10.3 8.2 9.0 4.1 8.2
19 6.7 7.2 5.2 5.2 10.3 11.8 10.8 10.3 5.2 8.2 5.7 6.2
20 6.2 7.7 6.2 7.0 5.7 6.2 13.4 8.2 5.7 7.2 14.9 6.4
21 11.8 12.4 9.3 9.8 5.2 11.3 10.8 7.2 5.2 5.2 5.9 7.0
22 19.1 11.6 10.3 8.2 6.2 7.2 9.3 5.2 5.7 5.2 8.2 6.7
23 9.8 4.6 9.8 6.2 7.2 10.3 8.2 8.2 6.2 7.2 6.7 8.2
24 6.7 5.7 12.4 7.7 9.5 6.7 6.7 7.2 7.2 5.7 7.7 6.2
25 12.4 9.8 11.8 4.1 9.5 9.8 13.4 9.3 6.2 4.6 6.4 4.4
26 7.7 9.8 10.0 9.3 11.3 7.7 8.8 6.2 5.2 5.4 11.3 7.7
27 5.2 8.2 11.8 15.5 9.8 12.9 9.3 6.7 4.6 8.8 5.2 8.2
28 5.2 13.1 7.2 10.8 9.8 8.8 8.2 15.5 3.6 5.7 3.6 5.7
29 8.8 10.3 10.8 9.0 11.8 6.2 7.7 7.7 7.7 6.2 4.1 9.8
30 10.3 9.3 6.7 11.8 12.4 9.8 6.2 7.7 2.8 5.7 7.7
31 13.4 6.2 7.2 8.2 7.2 5.2 5.7
Average: 8.50 Std. Dev: 2.83 C.o.v.: 0.33 Count: 366
259
Table E.2. Daily Maximum Wind Speeds observed during the year 2004 in Izmir
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 10.9 10.9 7.2 10.8 9.4 9.0 12.7 13.6 13.9 13.3 5.8 8.5
2 11.3 8.6 10.0 13.6 8.6 10.2 14.4 14.7 13.2 13.3 6.0 8.7
3 9.6 9.2 10.1 15.8 9.8 9.6 12.7 13.8 12.4 13.3 17.3 11.0
4 8.6 12.0 13.0 19.1 10.9 8.8 11.5 11.1 13.6 15.0 17.4 11.4
5 8.6 9.0 12.6 7.0 16.4 22.5 14.3 11.6 14.6 16.2 9.3 7.0
6 14.6 7.5 11.5 12.7 29.9 10.6 13.7 10.7 14.0 15.0 3.0 8.0
7 13.3 17.0 9.9 15.5 16.5 11.3 15.1 11.3 13.9 10.9 4.4 15.4
8 10.7 14.7 12.6 13.0 10.0 15.3 14.8 12.8 12.8 11.0 15.0 12.5
9 4.9 13.4 22.8 10.0 12.0 12.5 14.7 14.1 13.8 8.0 12.6 7.3
10 15.0 13.2 21.0 10.8 12.8 14.0 12.9 13.8 15.7 8.3 8.3 10.3
11 19.0 12.6 11.3 11.3 13.7 13.8 12.6 14.6 15.3 13.1 9.6 13.5
12 9.8 16.0 11.8 10.4 11.7 10.4 10.1 12.8 13.3 11.2 5.3 11.8
13 9.4 15.0 14.2 19.8 24.7 8.1 12.8 10.6 10.5 11.2 14.0 7.6
14 11.3 11.6 14.8 12.1 12.7 13.4 14.5 10.3 9.6 7.7 20.3 15.3
15 19.8 17.0 10.1 17.5 11.8 12.0 16.5 13.0 11.0 10.6 19.5 16.8
16 15.8 11.2 12.2 15.0 14.8 14.1 15.3 12.8 12.6 12.6 13.6 11.7
17 6.0 15.1 12.0 12.8 9.0 11.8 14.0 11.6 16.1 11.8 12.4 10.0
18 10.6 12.0 13.2 14.2 9.9 16.0 12.5 13.6 13.0 9.0 8.8 17.8
19 10.3 4.6 10.0 8.1 14.9 9.8 11.0 14.2 11.4 10.1 10.2 16.5
20 15.6 11.3 6.2 10.1 12.2 12.6 12.0 14.2 10.0 8.7 18.8 9.2
21 15.1 21.0 11.6 12.2 10.3 10.4 12.7 11.4 9.4 8.1 13.8 13.0
22 23.1 14.4 14.0 12.8 9.6 14.1 14.7 9.8 9.8 11.3 10.3 12.3
23 17.0 12.2 12.3 12.2 10.0 15.0 14.0 13.5 8.6 11.0 8.4 8.5
24 9.3 17.4 16.7 8.7 10.8 14.5 13.0 15.2 14.3 10.0 13.1 10.1
25 12.0 18.7 13.0 7.0 10.5 12.0 13.5 14.4 17.2 6.9 14.8 5.7
26 10.0 21.2 14.1 9.3 15.5 13.2 9.5 7.5 13.3 7.0 13.6 10.1
27 14.3 23.9 22.9 13.8 16.3 13.6 11.4 12.3 10.8 7.0 6.6 13.7
28 18.9 24.2 9.4 9.8 15.0 16.0 10.6 14.3 14.5 7.5 8.2 14.1
29 19.5 9.6 14.8 12.8 11.9 11.0 10.4 13.3 11.0 6.9 12.8 20.7
30 11.3 15.1 7.1 14.2 13.5 12.2 13.0 11.5 5.7 11.0 5.9
31 13.4 13.0 13.1 14.0 13.0 8.7 9.7
260
Table E.3. Daily Maximum Wind Speeds observed during the year 2004 in Bursa
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JULY
JUNE
MAY
Days
1 6.4 3.7 5.2 6.6 8.7 5.8 8.6 5 8.3 5.6 2.5 2.8
2 6.5 5 7.5 11.5 4.7 8 7 5.7 7.2 3.3 4.2 3.8
3 3.5 5.8 7.3 8.2 2.2 4.9 5.6 5.9 10.2 4.3 9.1 3.8
4 4.2 7.4 7.8 8.6 3.7 5 9.8 5.7 8.3 4.3 6.2 2.9
5 4.4 3.7 8 4.5 9.5 6 9.4 5.1 8.7 4.6 5.5 4.8
6 5.6 3.9 8.8 7.7 2.1 11.1 9.4 6 11 5.4 2.1 3.8
7 4.7 13 4.5 11.8 2.2 6.2 9.7 6 9.5 3.3 1.6 4.7
8 4.4 12 8.3 12.4 8.4 6.9 9.4 8.9 6.7 6.1 3.4 3.5
9 3.5 11.2 8.7 6.8 5.4 6.5 7.2 6.6 8.4 1.7 2.1 3.7
10 7 7.3 17.4 6.5 7.2 6.9 7.2 6.6 7.7 1.4 2 6.5
11 8.7 7.5 5.6 4.4 7.4 6.3 5.6 6 8 5.2 3.4 4.2
12 4.4 9.1 7.3 6.4 5.5 3.8 6.3 6.4 7.2 6.8 3.6 2.3
13 2.6 8.9 7.9 6.4 7 6 17.9 5.4 4.7 6.5 1.7 2.5
14 6.8 6.2 6.9 12.8 6.8 7.7 5 6.7 4.9 4 6 7
15 10 6.4 3.4 7.2 7.4 9 10.4 8.8 4.7 2.5 14.2 6.1
16 12 4 6.5 6.4 5.7 4.9 8.6 10.5 5 6 3.9 4
17 2.5 5.3 4.7 5.9 6.4 6.3 9.4 5.4 9.1 10.1 5.1 5.6
18 7 4.2 6.9 5.2 6.5 9.5 8.4 5 9.7 8.4 1.1 7
19 5.8 1.5 3.5 4.4 8.5 6.6 9.6 4.2 5.7 4.5 16 6.1
20 8.1 8.4 3.9 3.9 4.7 5.2 10.2 6.6 3.8 6.8 14 3.4
21 15 6.9 10.2 6.6 5.4 5.9 8.8 5 6 5 7.7 6.9
22 15 3.8 9.9 5.4 7 7.4 7.8 6.5 5.1 4 5.8 4.9
23 5 6.9 3.7 5 5.8 5.7 7.7 7.5 4.2 4.3 5.7 3.8
24 4.7 6.4 12 4.6 7.2 6.2 7.7 8 8.3 3.9 6.7 3.4
25 5.9 11.3 13 3.6 7.3 5.7 8.4 8.1 7.8 3 7.6 4.3
26 3 12.2 9.2 10 9.7 6.5 5.6 4 2.4 2.7 6.9 9.4
27 4.5 19 14.2 5.4 8.3 7.6 5.4 5.7 3.5 1.3 4.2 6.3
28 16 8.9 4.7 8.9 8 7.4 8.4 2.9 3 6.8 6.3
29 16.8 11.1 5.5 6.8 6.6 6.1 6.4 4.5 2 6.4 11.7
30 7.7 7.8 3.7 7.8 7.8 6.4 5 3.7 2 3.4 6.7
31 5.3 7.3 8.9 7.2 5.6 2.5 3.8
261
Table E.4. Daily Maximum Wind Speeds observed during the year 2004 in Antalya
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 7.0 12.9 6.3 8.5 6.6 15.3 8.1 7.3 8.0 5.7 8.2
2 9.5 9.3 4.3 13.7 6.8 9.5 9.2 7.7 7.4 6.8 4.3 5.6
3 6.3 8.0 7.6 16.0 7.3 10.3 7.6 7.8 8.4 12.9 5.4 3.6
4 24.9 9.6 12.2 15.3 8.8 7.0 11.2 9.0 9.8 10.0 7.3 3.6
5 5.0 11.0 17.7 10.2 5.9 9.4 8.8 12.8 6.5 9.0 8.2
6 16.7 11.6 25.0 8.3 6.2 9.1 8.2 10.0 13.1 5.4 9.0
7 20.9 8.0 16.7 9.0 6.2 8.1 12.5 8.2 13.0 8.7 13.9 16.9
8 15.5 6.7 7.2 8.0 8.2 11.5 10.0 8.0 8.5 6.6 9.4 19.4
9 9.8 9.0 9.0 6.0 11.4 11.6 14.1 7.2 8.0 7.6 6.2 6.9
10 8.1 8.7 10.8 7.3 9.2 11.0 12.8 7.2 20.0 6.7 5.6 8.0
11 26.9 22.0 14.0 6.5 11.8 10.9 10.0 13.2 14.6 6.9 6.0 18.3
12 9.6 20.7 8.7 8.2 15.1 7.7 11.6 12.9 11.1 7.1 6.1 11.8
13 6.9 23.4 16.8 6.2 7.6 6.9 9.3 12.7 7.8 7.2 5.4 5.6
14 8.1 21.2 15.8 8.8 7.8 11.6 12.2 9.2 6.4 7.7 4.9 17.7
15 7.6 12.0 15.1 26.0 16.3 9.8 9.2 7.0 7.4 7.6 18.0 20.1
16 13.5 7.2 7.6 25.6 15.0 11.5 20.1 8.4 6.9 9.2 11.1 11.3
17 14.9 6.5 11.4 8.1 9.5 10.0 9.8 8.6 7.0 7.1 17.3 5.9
18 6.3 7.6 14.6 10.2 16.8 17.1 10.8 9.4 8.0 5.0 12.6 18.8
19 5.0 7.6 12.4 8.4 12.0 11.6 8.9 13.5 7.3 5.7 6.9 21.8
20 4.5 5.3 10.0 8.2 10.0 8.2 10.0 10.2 7.0 5.0 9.9 20.3
21 19.5 24.4 7.0 12.5 8.3 7.5 12.2 8.3 5.5 12.6 25.0 5.8
22 32.3 22.1 7.0 8.8 8.7 7.5 14.3 11.4 7.6 7.0 25.9 5.7
23 18.2 6.1 6.3 8.0 6.2 18.6 8.9 11.6 9.4 6.1 17.9 3.5
24 13.8 6.3 12.7 7.9 6.7 14.0 12.9 9.9 8.5 5.4 5.9 9.4
25 23.0 7.0 9.7 7.8 6.3 8.6 9.2 9.2 8.2 5.0 20.7 8.3
26 7.8 5.2 6.2 7.1 6.5 9.4 7.9 9.9 9.2 5.9 23.2 7.4
27 14.6 3.9 4.8 18.0 10.8 10.0 8.2 9.2 7.5 6.6 17.0 3.7
28 7.3 6.9 6.3 8.6 10.7 11.8 7.3 11.5 10.1 5.2 7.5 4.0
29 19.0 5.5 7.9 8.2 9.1 12.0 7.0 8.6 8.5 7.3 4.6 15.0
30 22.5 16.4 9.6 10.2 8.2 8.0 8.0 7.0 6.0 5.0 15.0
31 10.8 10.0 11.0 7.1 8.2 5.3 5.5
262
Table E.5. Daily Maximum Wind Speeds observed during the year 2004 in
Gaziantep
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 3.9 8.6 3.3 8.5 4.0 7.8 8.3 5.8 3.7 6.8 4.2 4.0
2 8.1 6.7 3.8 9.5 13.9 7.3 8.9 5.4 3.3 7.2 2.9 3.4
3 3.6 3.6 6.2 9.4 3.9 4.4 5.6 6.7 2.7 7.8 3.6 2.1
4 4.2 7.9 5.7 9.5 6.5 4.0 8.2 5.1 4.6 3.3 3.4 2.0
5 6.5 3.6 13.8 5.1 6.9 3.7 7.8 4.1 10.6 3.8 4.3 3.2
6 10.0 7.8 7.8 4.3 4.3 6.4 3.8 6.0 8.2 8.7 4.8 3.8
7 11.1 7.2 6.8 5.2 5.0 10.0 3.9 5.0 4.9 4.1 3.5 2.0
8 6.7 4.7 5.0 6.9 5.8 8.0 4.0 4.9 3.8 3.5 3.0 7.4
9 3.5 3.8 3.6 3.5 6.4 8.1 11.8 3.7 6.5 5.0 3.6 3.5
10 6.6 4.5 6.2 5.7 7.8 7.5 7.9 4.0 8.7 3.8 3.5 3.5
11 4.6 6.0 7.0 6.5 8.2 5.5 7.0 6.9 8.9 3.7 3.8 2.7
12 4.6 4.3 4.0 7.0 10.8 5.9 5.7 6.4 11.6 3.6 3.0 3.4
13 5.9 10.6 4.0 3.4 9.8 3.4 5.6 5.5 3.4 3.0 2.7 2.1
14 3.8 11.9 6.3 7.3 6.7 7.4 5.6 3.9 2.0 3.2 2.3 3.0
15 4.2 9.2 9.1 11.0 7.3 8.1 8.1 7.4 2.2 3.2 2.1 4.4
16 4.5 4.9 4.9 13.5 8.1 9.7 6.9 8.8 1.6 3.4 3.4 5.9
17 5.6 4.1 5.6 6.7 5.1 6.8 5.8 3.8 3.6 2.4 5.9 2.3
18 4.5 5.3 6.2 3.2 8.3 5.6 9.2 4.9 4.7 1.9 3.8 5.4
19 3.0 2.0 5.4 8.2 9.0 7.0 8.7 7.8 4.3 4.9 2.2 5.2
20 3.8 3.4 5.3 4.5 8.6 5.8 3.0 8.4 3.3 3.9 2.9 4.3
21 3.2 9.0 4.8 5.8 3.2 7.4 4.7 3.3 5.8 4.5 11.7 1.9
22 13.6 12.0 6.9 6.2 5.0 8.1 8.3 3.6 4.4 3.0 7.7 2.3
23 10.3 7.2 3.5 5.8 4.4 6.6 8.3 4.9 6.2 3.2 8.4 2.5
24 10.0 3.9 8.2 3.8 4.5 6.3 7.0 4.0 3.7 4.1 3.9 8.0
25 8.5 5.0 5.8 4.6 7.0 5.0 4.9 3.3 5.3 5.4 7.1 6.0
26 4.2 3.3 3.4 6.1 5.4 7.8 5.4 6.8 3.6 5.0 5.1 2.9
27 4.4 3.3 2.8 10.8 5.0 6.2 5.1 3.1 3.2 2.1 8.0 1.8
28 3.1 2.2 3.8 5.4 7.2 8.0 4.0 5.0 3.2 3.3 2.0 2.0
29 4.0 3.2 5.1 6.0 6.0 9.5 5.0 6.4 3.6 3.7 2.1 4.0
30 4.0 8.5 5.0 5.1 5.4 9.0 3.2 3.4 3.2 1.4 3.7
31 6.4 3.1 6.8 3.6 3.3 6.0 4.7
263
Table E.6. Daily Maximum Wind Speeds observed during the year 2004 in Samsun
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 11.2 9.0 7.3 6.5 3.2 4.1 6.2 3.0 4.7 6.7 3.0 7.0
2 7.0 5.2 5.1 11.0 6.5 3.4 7.5 10.5 3.8 6.5 7.0 14.5
3 8.2 8.2 6.8 12.2 4.7 5.0 7.4 8.5 6.6 7.4 6.7 18.2
4 10.7 7.0 9.0 7.8 3.9 3.0 4.0 4.3 8.5 7.1 6.3 16.2
5 13.9 6.2 10.0 6.1 7.9 7.5 4.3 5.8 10.4 7.0 4.8 7.8
6 16.0 16.2 11.6 5.9 6.2 9.0 4.5 3.6 10.1 5.9 5.5 9.8
7 12.0 11.1 6.8 6.3 8.0 5.9 4.3 5.0 8.6 2.1 3.0 15.0
8 12.6 18.8 7.6 4.8 3.8 5.9 7.1 3.9 5.6 4.4 2.4 6.6
9 13.0 9.0 8.0 2.7 5.5 6.3 7.6 3.7 12.8 5.7 5.7 11.2
10 11.8 12.2 14.6 5.8 11.9 7.2 5.0 6.1 10.0 6.1 2.3 7.0
11 10.8 7.8 7.2 4.3 7.8 5.6 5.4 7.3 6.3 14.5 3.3 8.8
12 6.6 12.4 5.0 10.0 7.7 5.7 5.2 8.5 4.0 11.6 3.2 7.3
13 4.6 17.4 8.0 5.9 4.3 2.4 4.8 5.9 5.3 6.3 6.2 10.4
14 11.2 9.8 8.0 7.2 8.9 3.0 14.6 6.4 5.4 5.3 8.8 11.4
15 14.0 20.0 6.1 7.5 17.4 8.5 7.0 11.3 4.2 5.3 11.2 9.2
16 13.0 21.8 4.8 13.6 11.7 4.7 6.1 5.0 3.4 4.4 6.8 5.9
17 7.1 10.3 5.0 6.9 3.8 3.1 7.2 7.5 7.0 9.4 10.6 12.3
18 15.3 5.5 3.3 5.2 7.2 4.6 6.4 9.7 5.8 10.0 7.5 17.4
19 14.5 5.9 6.3 2.5 7.5 9.3 6.6 5.0 4.3 4.7 12.4 20.7
20 17.2 13.1 8.0 4.5 5.2 8.1 5.5 5.4 3.1 8.6 18.6 18.4
21 21.0 14.1 13.2 3.1 7.9 6.0 6.7 5.2 4.4 9.0 8.9 7.0
22 19.3 13.6 9.0 9.0 4.0 4.1 6.4 4.9 5.7 3.8 15.3 4.0
23 26.1 9.2 4.3 9.7 5.9 4.0 8.6 8.5 8.2 7.5 10.0 7.8
24 5.8 16.4 7.6 6.3 4.3 7.7 3.2 7.4 6.3 5.4 21.1 6.0
25 8.9 14.0 11.2 4.2 12.2 6.0 4.6 7.3 4.0 4.6 8.0 5.0
26 5.3 7.7 7.2 6.0 4.0 3.7 6.6 5.2 2.5 5.0 10.2 14.4
27 6.6 21.9 12.7 6.3 8.8 10.5 7.0 9.4 3.1 7.2 14.0 18.2
28 19.2 20.0 14.6 10.0 7.8 8.0 5.7 9.0 3.2 6.2 17.2 11.3
29 19.1 16.8 8.2 3.9 5.3 4.8 4.0 8.2 4.3 2.8 13.7 12.4
30 6.6 7.6 3.9 3.3 13.3 4.6 3.8 8.7 5.4 18.0 9.0
31 7.5 5.6 5.0 5.2 4.9 2.2 2.9
264
Table E.7 Daily Maximum Wind Speeds observed during the year 2004 in Malatya
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 2.2 2.2 3.8 3.8 6.7 8.7 11.8 3.6 3.3 4.3 2.1 8.0
2 3.0 4.4 2.3 10.0 7.8 9.3 9.4 8.7 5.0 6.4 3.0 4.8
3 4.0 4.3 3.6 12.6 3.1 13.0 4.7 8.8 5.1 8.3 4.3 2.9
4 4.0 2.9 3.2 13.1 8.0 5.8 8.9 6.4 4.1 3.0 6.7 2.0
5 4.8 5.0 18.3 5.8 5.0 4.5 5.7 5.9 9.6 3.2 3.2 3.4
6 4.4 10.6 16.0 4.2 5.0 11.5 5.2 9.8 7.5 3.8 3.1 4.2
7 13.5 11.0 9.0 4.0 4.2 5.9 5.6 5.4 3.0 9.7 3.2 3.0
8 8.6 8.6 6.0 6.9 4.9 10.0 14.5 5.2 8.7 4.6 3.0 2.8
9 3.8 8.6 3.8 4.8 10.1 8.4 15.1 7.7 8.4 4.5 2.5 5.0
10 3.9 10.8 8.4 4.9 8.2 6.3 7.8 4.7 9.9 3.4 2.9 8.4
11 2.3 7.5 9.3 4.0 10.7 8.3 8.2 13.1 8.0 3.3 2.8 2.7
12 1.7 4.0 6.8 3.6 13.3 8.7 9.4 9.7 6.8 6.0 2.8 10.0
13 4.6 12.6 7.1 5.2 11.0 4.2 9.8 4.7 2.9 8.0 2.5 4.2
14 2.9 12.8 9.2 19.8 9.9 9.8 9.0 4.3 3.0 6.0 2.0 9.8
15 2.7 10.0 8.6 10.0 10.0 9.6 11.2 4.0 3.2 3.7 3.9 13.0
16 4.0 3.0 11.2 10.0 7.6 9.8 13.3 5.4 3.1 9.0 3.6 10.8
17 13.6 3.6 8.7 5.1 9.7 9.8 10.0 8.6 7.1 4.1 8.3 5.0
18 4.4 3.0 4.6 9.8 16.0 4.2 10.8 10.3 2.5 2.4 7.4 3.0
19 2.9 4.2 7.1 13.0 16.6 14.6 6.3 9.3 2.9 2.3 2.6 3.0
20 2.5 4.0 4.4 7.7 8.7 9.3 4.3 6.3 3.3 6.1 7.2 2.4
21 2.2 10.4 4.0 8.5 4.4 7.2 9.9 5.6 3.5 10.8 10.5 2.2
22 6.4 11.0 6.2 7.1 5.0 9.6 11.4 4.1 4.5 6.3 16.2 2.3
23 16.1 10.4 8.3 10.8 4.4 8.5 6.2 3.6 8.8 3.0 11.5 2.5
24 10.5 1.8 6.9 2.8 4.3 9.8 4.7 7.0 5.0 2.6 5.2 2.2
25 3.9 1.5 9.3 7.5 11.0 9.8 9.6 6.8 4.4 3.1 8.3 3.2
26 5.0 2.2 4.3 6.1 13.1 8.0 3.8 2.8 3.3 2.6 5.0 3.8
27 2.1 2.9 3.7 13.1 9.1 4.8 8.5 4.4 4.0 3.1 7.8 2.8
28 3.1 3.4 4.4 11.0 6.6 12.6 9.8 7.0 3.7 2.4 5.0 3.0
29 3.7 3.3 8.8 17.0 8.1 10.4 4.7 10.0 4.2 5.2 3.2 3.7
30 2.7 12.4 9.6 9.0 4.7 4.6 3.2 3.4 2.1 3.0 2.7
31 2.8 6.4 12.3 4.9 3.6 3.0
265
Erzincan
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 2.3 3.7 2.1 4.5 5.6 7.9 9.5 3.8 6.2 8.3 3.5 5.0
2 4.2 3.4 3.6 10.0 3.8 9.0 7.5 6.8 8.7 6.5 3.0 8.1
3 4.1 3.3 2.2 11.0 7.5 12.7 4.2 9.6 8.5 6.3 6.4 2.5
4 5.9 3.0 9.0 12.8 9.7 3.9 8.3 6.7 6.8 6.8 9.0 2.8
5 6.5 3.4 10.2 5.7 7.3 3.2 8.0 5.6 10.0 4.1 3.2 2.8
6 9.1 7.1 9.3 4.1 5.6 9.7 4.3 9.6 9.4 4.8 4.5 4.0
7 9.5 7.2 10.9 4.9 6.7 7.5 5.7 7.7 3.5 6.7 4.9 4.1
8 7.0 7.0 4.0 6.3 9.1 10.0 7.5 8.8 10.0 9.2 3.3 3.1
9 1.1 5.0 3.2 6.0 12.6 7.5 8.3 7.8 8.4 8.0 2.8 2.6
10 3.0 7.1 3.6 4.2 10.9 9.2 15.7 5.9 9.1 5.6 3.0 3.1
11 2.1 7.6 9.7 6.5 9.9 6.0 7.6 14.7 9.8 3.0 2.0 2.7
12 3.0 6.6 7.3 3.9 13.0 6.6 8.5 10.9 10.2 4.2 3.0 10.6
13 4.8 10.8 9.6 8.2 7.9 5.0 8.1 6.4 5.7 5.9 2.3 4.4
14 5.2 8.2 9.5 9.8 11.7 10.1 4.9 3.8 5.1 3.6 2.8 7.8
15 2.5 9.0 10.6 12.1 9.9 9.3 7.8 9.4 6.9 4.9 3.3 9.3
16 9.3 4.1 8.4 12.4 7.5 7.3 8.8 3.8 4.6 4.20 6.7 7.9
17 7.5 3.2 7.2 12.0 8.2 7.5 10.9 6.0 7.9 3.3 7.8 3.7
18 2.0 2.6 4.8 8.5 9.8 6.6 10.3 11.0 5.2 3.6 9.8 3.1
19 3.1 2.0 9.5 10.0 8.2 12.3 11.3 10.0 3.1 4.8 3.1 3.4
20 2.2 3.2 7.3 7.6 9.7 10.9 8.2 8.6 3.3 6.9 9.1 3.9
21 2.2 9.0 3.5 8.9 7.2 6.1 10.0 7.6 5.8 8.2 9.1 3.9
22 13.1 10.0 5.3 7.2 4.3 9.4 9.9 5.5 3.5 4.7 11.9 4.9
23 11.9 10.4 5.6 9.2 9.4 7.1 7.7 7.2 5.3 2.9 9.2 4.0
24 9.6 1.8 7.9 9.9 4.1 7.2 6.2 7.9 3.4 2.9 4.1 3.7
25 5.8 2.2 8.7 11.1 9.1 8.1 6.6 8.8 8.8 2.7 9.9 3.9
26 6.0 2.0 7.6 4.1 9.9 7.4 9.2 6.5 5.7 2.5 9.7 3.1
27 3.5 2.6 3.3 10.2 11.4 7.0 12.5 4.9 2.8 2.8 9.9 3.0
28 3.9 3.7 5.6 11.4 7.6 10.3 6.1 4.2 7.1 3.3 6.4 3.1
29 2.9 3.7 8.8 11.1 6.4 10.1 6.9 7.0 4.0 4.7 3.9 4.0
30 5.0 10.0 6.1 8.2 8.4 5.1 7.2 3.4 6.0 3.1 5.0
31 3.0 12.4 12.3 6.2 10.0 2.8 3.7
Average: 6.65 Std. Dev:2.92 C.o.v.: 0.44 Count:366
266
Canakkale
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 4.1 5.5 6.7 9.1 10.9 5.4 10.3 10.8 9.9 7.0 3.4 6.1
2 6.8 3.6 11.1 15.0 10.4 4.4 9.6 8.1 12.1 6.1 13.5 12.0
3 9.4 12.9 14.7 13.9 5.4 8.0 9.1 7.5 16.5 10.5 17.5 11.2
4 17.0 11.1 17.8 11.3 5.9 10.4 15.3 4.0 15.3 14.0 21.0 11.7
5 9.5 5.0 15.4 5.5 18.5 7.0 17.2 5.3 16.7 16.5 16.5 9.0
6 11.7 8.8 7.4 12.2 17.8 9.0 17.5 8.1 22.0 18.4 11.5 5.6
7 11.2 10.1 8.2 14.6 15.0 6.7 18.2 8.8 19.4 13.8 5.2 13.9
8 5.8 15.1 3.0 15.0 11.0 11.0 19.0 9.5 10.9 5.7 10.0 6.7
9 4.0 17.0 14.0 10.8 5.0 7.7 13.8 9.1 11.8 4.3 12.0 10.5
10 8.4 9.8 13.9 8.7 6.4 10.2 12.8 11.2 14.0 8.0 3.6 14.3
11 22.9 11.1 8.1 8.5 8.0 10.7 8.9 7.1 14.0 15.2 7.3 12.3
12 11.0 18.8 12.3 8.9 6.0 5.0 7.1 6.6 13.0 17.4 8.1 6.8
13 6.0 14.7 12.1 11.1 7.5 4.5 10.3 6.6 12.0 17.0 4.4 7.9
14 12.4 4.2 14.1 10.7 6.8 17.1 9.7 13.4 14.0 12.9 24.6 12.6
15 18.4 10.7 8.5 10.6 9.2 12.7 13.8 13.7 13.3 4.4 20.5 12.8
16 20.1 7.5 10.9 9.9 11.7 6.7 14.3 13.9 8.6 13.1 15.8 7.1
17 8.2 12.7 7.1 5.0 14.1 13.0 13.3 10.8 16.3 17.6 13.5 10.2
18 13.7 9.4 9.1 4.5 10.6 11.0 16.1 5.7 15.8 13.9 6.3 20.4
19 10.0 5.5 3.9 10.1 6.7 7.6 15.5 5.7 15.2 10.9 19.3 22.8
20 22.3 11.8 7.8 5.6 3.5 8.1 15.2 6.2 12.9 8.0 19.7 8.9
21 20.7 14.2 12.2 11.0 5.0 7.1 17.6 5.2 7.0 9.0 10.1 15.2
22 20.0 5.4 16.8 8.1 10.7 12.4 14.1 11.8 7.5 12.5 11.5 16.3
23 9.4 6.5 13.3 7.5 12.1 16.0 11.4 9.3 8.6 11.2 13.6 9.9
24 4.0 10.4 22.0 5.4 13.4 8.7 10.9 12.4 15.0 5.9 12.1 3.7
25 13.8 17.7 14.9 9.0 11.0 8.1 11.0 11.3 16.3 9.1 12.4 3.8
26 5.6 18.9 15.4 10.2 13.9 8.0 12.1 6.0 11.9 4.9 7.6 15.0
27 15.8 31.0 16.5 13.9 14.3 12.8 12.3 10.1 9.6 8.4 12.7 16.1
28 22.8 14.1 7.3 6.1 15.2 9.4 12.4 10.0 6.7 3.5 8.6 17.1
29 22.9 13.9 11.3 7.0 12.0 8.9 7.5 6.3 7.0 4.4 8.2 23.0
30 14.4 15.3 9.3 11.7 12.3 11.8 7.4 6.3 3.8 9.2 12.6
31 20.2 13.6 7.5 9.0 8.0 3.2 15.0
267
Hakkari
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 2.5 5.3 3.0 5.4 8.8 11.3 9.2 4.7 7.9 6.7 2.2 1.4
2 2.0 8.3 4.7 11.3 5.2 8.7 8.4 5.7 5.5 6.7 2.1 1.7
3 3.1 2.5 2.2 14.3 9.2 8.1 7.6 9.9 8.3 12.2 8.6 1.4
4 1.4 3.2 4.6 10.2 7.3 8.1 11.2 6.7 6.8 10.3 3.9 1.1
5 8.6 3.1 17.9 5.1 9.8 6.2 9.7 6.1 14.1 7.1 4.0 1.2
6 2.2 5.7 15.3 2.9 8.7 6.9 9.7 5.8 11.3 11.7 4.6 4.6
7 18.1 9.2 10.2 5.9 6.5 9.1 10.0 7.0 11.2 5.5 2.2 3.4
8 15.9 10.0 3.2 5.0 6.8 9.4 12.1 6.1 11.4 20.6 2.0 3.4
9 8.2 8.5 2.8 5.5 10.2 13.3 13.3 4.8 13.1 10.5 2.0 1.4
10 10.0 8.6 5.1 7.1 14.9 10.0 9.8 8.7 11.8 12.2 5.7 7.5
11 1.2 10.3 4.4 6.6 11.5 9.5 8.8 9.3 8.4 6.8 5.1 6.5
12 2.2 8.1 4.9 6.5 12.0 6.2 7.6 10.2 9.4 11.9 3.4 3.3
13 3.9 15.6 7.0 2.8 11.2 9.8 6.9 5.6 7.0 9.1 3.1 1.8
14 6.7 17.1 6.0 3.6 11.4 11.0 7.3 4.1 5.9 10.4 3.0 6.2
15 3.1 4.9 9.2 12.8 14.0 11.0 6.8 3.8 7.8 6.3 3.6 9.9
16 2.8 12.2 6.4 14.1 9.7 18.5 11.4 4.6 8.7 4.0 7.9 3.0
17 3.0 7.5 5.3 7.9 15.9 11.0 5.8 5.4 6.0 4.6 6.9 6.3
18 3.7 1.3 6.5 10.4 11.0 9.9 8.1 8.8 6.5 2.6 8.7 1.5
19 3.6 2.0 6.7 10.0 13.0 5.7 8.3 8.8 7.5 5.0 7.0 1.0
20 2.0 7.3 6.5 6.0 12.8 9.0 8.7 7.5 9.0 4.4 6.8 0.9
21 2.2 8.6 3.9 8.1 6.0 10.2 8.1 7.4 4.0 12.8 9.0 1.4
22 9.0 5.2 4.5 17.1 9.6 8.0 10.0 6.5 6.2 4.9 12.0 1.1
23 14.2 5.6 6.6 13.7 6.9 8.1 8.9 4.4 11.7 5.2 4.8 1.3
24 10.5 1.8 4.7 7.9 10.1 7.4 10.0 6.5 8.5 3.7 5.3 1.0
25 3.2 1.4 4.9 8.7 7.0 10.2 10.0 7.4 3.4 7.5 1.6 6.2
26 1.4 1.6 5.8 4.4 10.6 9.4 8.1 8.8 3.0 4.2 11.2 13.2
27 4.9 1.5 5.0 3.8 14.0 8.8 5.6 4.4 5.1 6.5 7.9 4.9
28 3.9 1.5 4.8 6.5 11.9 6.0 10.4 9.2 8.2 11.1 1.5 1.4
29 3.5 1.5 5.3 2.1 12.5 14.0 9.6 3.8 11.3 4.2 1.4 1.1
30 2.0 16.2 9.2 11.8 5.7 4.0 6.8 3.2 11.5 1.2 1.1
31 2.3 6.6 7.9 7.9 8.7 1.5 1.1
268
Istanbul/ Göztepe
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JULY
JUNE
MAY
Days
1 1.7 2.6 3.5 4.6 7.0 2.1 5.0 6.2 7.8 4.1 1.5 2.3
2 2.7 3.6 5.3 10.0 9.8 4.2 7.8 4.9 8.0 5.8 7.6 7.8
3 8.3 8.0 9.1 7.9 2.9 9.8 5.0 5.0 10.4 7.5 10.1 5.7
4 11.0 9.2 9.1 4.8 2.2 8.4 9.0 3.8 10.9 9.9 15.1 4.2
5 9.0 4.1 11.9 2.3 8.0 6.8 14.9 6.3 12.5 15.5 11.2 4.8
6 11.5 3.1 9.2 3.2 2.3 3.1 13.5 7.1 14.7 14.5 5.9 3.9
7 8.7 7.8 2.7 6.1 10.7 5.7 12.7 8.9 13.3 9.6 6.9 8.7
8 4.7 4.0 3.1 7.3 2.7 2.6 10.3 6.9 8.1 5.6 6.2 2.5
9 1.8 9.6 4.0 2.3 2.9 4.4 10.2 6.3 12.3 4.5 3.7 7.1
10 2.5 6.0 11.0 4.7 8.3 5.0 7.8 5.9 11.4 6.4 4.2 7.7
11 11.0 4.2 3.3 1.6 7.5 6.0 4.1 4.8 11.1 8.3 8.4 3.5
12 6.8 13.4 8.5 2.8 4.4 2.6 5.3 4.9 7.7 15.6 5.1 1.9
13 4.4 9.9 7.0 2.3 3.9 3.7 5.8 3.1 6.2 13.4 1.9 3.3
14 3.8 4.9 6.4 11.0 4.0 10.9 5.3 7.2 8.3 12.2 6.0 9.6
15 12.8 15.9 3.7 10.3 10.1 7.0 5.8 9.1 7.0 1.8 9.7 7.1
16 15.9 2.2 4.2 4.2 5.8 2.6 8.2 15.0 5.9 6.3 11.0 2.2
17 1.9 5.5 2.5 1.6 9.9 7.2 8.3 8.0 12.0 8.5 16.4 3.9
18 8.3 3.2 4.9 4.2 10.0 4.4 9.8 4.0 10.7 6.5 6.0 2.6
19 3.5 3.6 2.6 3.9 2.7 4.4 12.3 4.2 8.8 4.0 15.6 3.1
20 11.7 10.1 1.9 4.7 3.6 6.3 10.9 2.7 7.2 5.0 16.9 5.1
21 13.1 9.9 11.4 10.0 3.4 4.3 11.9 2.8 4.8 4.8 7.5 14.0
22 16.8 3.3 8.1 4.0 3.7 3.9 9.7 14.8 4.4 8.8 11.0 14.8
23 5.9 2.1 2.5 2.7 3.6 4.1 6.1 15.0 8.8 7.0 13.8 3.8
24 2.9 2.2 11.8 4.2 12.9 5.1 5.9 9.2 5.0 1.9 14.8 2.7
25 3.0 7.4 12.8 8.0 6.0 4.1 6.3 6.4 10.4 4.6 13.3 2.7
26 9.3 3.3 9.3 11.0 11.4 4.6 8.5 5.4 2.0 5.2 10.0 5.9
27 4.6 14.2 9.4 12.1 9.2 5.5 5.4 6.5 5.0 3.2 9.8 5.0
28 11.8 9.9 10.0 3.3 12.4 5.2 4.0 10.0 8.2 1.7 6.9 5.0
29 13.4 6.8 9.1 3.8 6.5 2.5 9.5 5.0 7.1 6.0 2.0 7.3
30 9.0 8.1 4.7 4.7 6.5 9.9 2.6 5.5 2.1 9.2 8.4
31 13.6 10.5 7.2 11.0 4.4 1.6 7.7
269
Istanbul/ Sile
SEPTEMBER
NOVEMBER
DECEMBER
FEBRUARY
OCTOBER
JANUARY
AUGUST
MARCH
APRIL
JUNE
JULY
MAY
Days
1 3.3 8.0 9.7 13.2 4.8 5.7 8.0 6.0 5.5 4.1 5.3 7.0
2 4.4 11.5 6.2 10.8 6.2 6.8 9.2 4.0 6.1 5.6 8.5 10.0
3 8.3 8.0 7.1 8.9 3.5 8.3 5.8 4.5 12.1 7.9 14.3 17.1
4 12.6 7.1 12.0 8.0 9.1 8.0 6.7 4.1 9.0 14.2 14.8 10.4
5 11.0 15.9 14.5 6.8 13.7 9.0 10.2 6.0 13.9 15.1 9.8 9.3
6 11.3 14.7 8.9 10.4 13.4 8.0 9.4 5.9 16.5 16.3 3.0 6.8
7 9.0 14.8 7.4 14.8 14.8 5.0 7.0 8.3 15.0 13.8 3.7 7.5
8 6.4 15.6 11.3 19.5 9.3 5.0 7.2 6.1 9.0 8.7 7.9 5.0
9 6.0 12.9 4.9 10.0 8.9 5.0 7.5 6.0 10.5 7.0 5.2 7.0
10 5.0 8.6 15.9 13.5 9.4 10.8 8.0 4.0 10.7 8.1 5.2 8.0
11 11.9 8.4 10.0 6.0 6.1 7.0 4.1 5.0 8.0 16.6 4.9 7.7
12 5.2 13.2 5.9 13.6 7.2 5.7 10.5 6.0 7.8 19.7 5.0 3.8
13 8.0 16.2 6.1 9.8 13.7 9.4 14.5 5.4 4.4 19.4 7.0 6.7
14 11.3 14.0 6.6 5.6 8.0 7.0 5.4 6.4 5.7 14.4 10.4 7.8
15 17.8 16.6 9.0 13.8 6.9 5.0 6.8 12.0 5.5 4.0 10.8 8.3
16 11.0 6.4 5.0 6.2 5.9 4.5 10.6 21.5 5.4 8.1 9.1 4.6
17 6.0 6.2 6.1 8.5 7.1 5.7 14.6 8.2 8.7 9.5 13.9 7.4
18 12.3 5.0 8.4 4.8 5.3 11.5 8.3 4.4 9.7 6.9 8.9 7.2
19 11.4 5.6 5.5 3.8 5.0 5.5 8.3 6.6 6.5 4.5 12.7 7.5
20 21.0 11.2 10.1 5.0 6.8 5.0 7.0 3.4 5.6 8.2 17.6 5.9
21 14.2 12.0 13.6 5.8 6.4 9.8 7.9 4.3 6.0 9.6 11.2 18.1
22 21.4 10.0 15.4 4.8 6.1 4.0 6.0 10.2 4.4 8.0 16.1 15.0
23 13.7 5.7 8.9 5.0 6.9 5.7 4.4 13.4 7.8 6.1 11.2 7.2
24 3.4 13.4 9.8 5.0 8.7 4.1 5.0 7.7 10.1 3.3 16.8 4.5
25 5.2 16.8 17.2 7.8 9.8 4.3 5.3 5.2 8.0 6.2 15.5 5.9
26 7.7 13.0 15.1 5.6 8.7 5.0 6.7 10.0 7.3 3.9 13.5 6.2
27 5.0 29.1 7.7 6.1 8.0 7.9 6.9 8.4 7.8 4.3 13.9 9.9
28 19.4 14.0 8.7 5.4 9.3 5.3 6.3 10.6 5.0 5.0 9.9 10.0
29 15.4 9.3 8.6 5.1 7.7 5.3 5.3 5.8 7.0 3.8 8.0 11.7
30 10.3 7.3 6.5 5.4 7.9 8.2 6.0 5.0 6.0 10.8 7.1
31 15.5 8.0 8.3 5.7 4.0 4.0 9.0
270
Table E.13. The Chi-Square (χ2) and Kolmogorov- Smirnov (K-S) tests results for
daily maximum wind speeds recorded in the year 2004
Daily wind speed
Tests Tests
Dist. Dist.
Ankara Malatya
Weibull 56.36 7.8E-6 0.0655 Weibull 75.12 6.3E-9 0.0820
Ext. value 75.88 4.5E-9 0.0780 Lognormal 86.16 0.00 0.1036
Lognormal 83.25 2.3E-10 0.0736 Ext. value 109.8 0.00 0.1035
Normal 125.8 0.00 0.0999 Rayleigh 93.03 0.00 0.1512
Izmir Normal 170.4 0.00 0.1506
Gamma 31.75 0.0235 0.0648 Erzincan
Lognormal 32.38 0.0198 0.0624 Ext. value 86.85 3.4E-6 0.0878
Normal 37.36 0.0047 0.0783 Rayleigh 58.64 2.0E-10 0.092
Ext. value 50.03 7.5E-5 0.0781 Weibull 58.85 0.00 0.089
Weibull 52.93 2.7E-5 0.0869 Lognormal 83.56 0.00 0.0958
Bursa Normal 120.2 3.13E-7 0.1048
Gamma 25.40 0.1144 0.0397 Canakkale
Ext. value 31.97 0.0222 0.0480 Weibull 14.42 0.7016 0.038
Rayleigh 30.41 0.0337 0.0547 Lognormal 17.84 0.4661 0.038
Normal 43.46 6.7E-4 0.0866 Normal 30.20 0.0356 0.048
Antalya Hakkari
Gamma 41.23 0.0014 0.0914 Weibull 27.29 0.0737 0.05577
Ext. value 70.23 4.1E-3 0.0915 Rayleigh 30.40 0.0337 0.05468
Weibull 77.45 2.4E-9 0.1031 Ext. value 31.13 0.0278 0.06340
Gaziantep Göztepe
Ext. value 51.69 4.2E-5 0.0768 Weibull 22.31 0.2187 0.047
Lognormal 65.60 2.5E-7 0.0850 Rayleigh 29.99 0.0376 0.090
Rayleigh 59.79 2.2E-6 0.0910 Lognormal 33.93 0.0128 0.060
Normal 94.98 0.000 0.1005 Ext. value 35.08 0.0092 0.059
Samsun Sile
Ext. value 30.92 0.0294 0.0637 Ext. value 57.63 4.9E-6 0.0773
Weibull 36.01 0.0070 0.0708 Weibull 61.90 1E-8 0.055
Rayleigh 47.85 1.6E-4 0.133 Rayleigh 76.78 3E-8 0.116
Note 1: At the significance level α=5%, 1-α percentile value of the Chi-Square test, χ.95,16=26.3
Note 2: The critical value of K-S test at the 5% significance level, D .366
05
=0.071.
271
APPENDIX F
FIGURES SHOWING THE VARIATION OF RELIABILITY INDEX

OBTAINED ACCORDING TO THE TURKISH DESIGN PROVISIONS
3.2
3
Reliability index
2.8 Turkey
2.6
2.4
0.25 0.5 1 2
L'/D'
Figure F.1 Variation of reliability index for D+L combination

(RC beams in flexural failure mode)
4.5
4
Reliability index
Ankara
3.5
Bursa
Çanakkale
3
Turkey
2.5
2
0.5 1 2 3
S'/D'
Figure F.2 Variation of reliability index for D+S combination

272
4.5 Ankara
(L'/D'=0.5)
4.3
Bursa
4.1 (L'/D'=0.5)
Çanakkale
3.9 (L'/D'=0.5)
Turkey
Reliability index
3.7
(L'/D'=0.5)
3.5 Ankara
(L'/D'=1)
3.3 Bursa
(L'/D'=1)
3.1
Çanakkale
2.9 (L'/D'=1)
Turkey
2.7 (L'/D'=1)
2.5
0.25 0.5 1 2
W'/D'
Figure F.3 Variation of reliability index for D+Lapt+W combination

5
Ankara
(L'/D'=0.5)
4.5 Bursa
(L'/D'=0.5)
Çanakkale
Reliability index
4 (L'/D'=0.5
Turkey
(L'/D'=0.5)
3.5 Ankara
(L'/D'=1)
Bursa
3 (L'/D'=1)
Çanakkale
(L'/D'=1)
Turkey
2.5
(L'/D'=1)
2
0.25 0.5 1 2
W'/D'
Figure F.4 Variation of reliability index for D+L+Wapt combination

273
2.5
Ankara
(L'/D'=0.5)
Bursa
2
(L'/D'=0.5)
Çanakkale
Reliability index (L'/D'=0.5)
1.5 Turkey
(L'/D'=0.5)
Ankara
(L'/D'=1)
1 Bursa
(L'/D'=1)
Çanakkale
0.5 (L'/D'=1)
Turkey
(L'/D'=1)
0
0.50 1.00 2.00 3.00
E'/D'
Figure F.5 Variation of reliability index for D+Lapt+E combination

2.6
Reliability index
2.4 Turkey
2.2
0.25 0.5 1 2
L'/D'

(RC beams in shear failure mode)
274
4
3.8
3.6
3.4
Reliability index
Ankara
3.2
Bursa
3
Çanakkale
2.8
Turkey
2.6
2.4
2.2
2
0.5 1 2 3
S'/D'

Ankara
3.7 (L'/D'=0.5)
Bursa
3.5 (L'/D'=0.5)
Çanakkale
(L'/D'=0.5)
3.3
Reliability index
Turkey
(L'/D'=0.5)
3.1 Ankara
(L'/D'=1)
2.9 Bursa
(L'/D'=1)
Çanakkale
2.7 (L'/D'=1)
Turkey
2.5 (L'/D'=1)
0.25 0.5 1 2
W'/D'

275
4.3
Ankara
(L'/D'=0.5)
Bursa
3.8
(L'/D'=0.5)
Çanakkale
Reliability index
(L'/D'=0.5
3.3 Turkey
(L'/D'=0.5)
Ankara
(L'/D'=1)
2.8 Bursa
(L'/D'=1)
Çanakkale
(L'/D'=1)
2.3 Turkey
(L'/D'=1)
1.8
0.25 0.5 1 2
W'/D'

2.5
Ankara
(L'/D'=0.5)
Bursa
2 (L'/D'=0.5)
Çanakkale
(L'/D'=0.5)
Reliability index
1.5 Turkey
(L'/D'=0.5)
Ankara
(L'/D'=1)
1 Bursa
(L'/D'=1)
Çanakkale
0.5 (L'/D'=1)
Turkey
(L'/D'=1)
0
0.50 1.00 2.00 3.00
E'/D'

276
3.2
Reliability index
2.8 Turkey
2.6
2.4
0.25 0.5 1 2
L'/D'

(RC columns in combined action)
4.5
4
Reliability index
Ankara
3.5
Bursa
Çanakkale
3
Turkey
2.5
2
0.5 1 2 3
S'/D'

(RC columns in combined action failure mode)
277
4.3 Ankara
(L'/D'=0.5)
4.1 Bursa
(L'/D'=0.5)
3.9
Çanakkale
(L'/D'=0.5)
3.7
Turkey
3.5
Ankara
3.3 (L'/D'=1)
Bursa
3.1 (L'/D'=1)
Çanakkale
2.9 (L'/D'=1)
Turkey
2.7
(L'/D'=1)
2.5
0.25 0.5 1 2
W'/D'

4.5
Ankara
(L'/D'=0.5)
Bursa
4
(L'/D'=0.5)
Çanakkale
Reliability index
(L'/D'=0.5
3.5 Turkey
(L'/D'=0.5)
Ankara
(L'/D'=1)
3 Bursa
(L'/D'=1)
Çanakkale
(L'/D'=1)
2.5 Turkey
(L'/D'=1)
2
0.25 0.5 1 2
W'/D'

278
2.5
Ankara
(L'/D'=0.5)
Bursa
2 (L'/D'=0.5)
Çanakkale
(L'/D'=0.5)
Reliability index
1.5 Turkey
(L'/D'=0.5)
Ankara
(L'/D'=1)
1 Bursa
(L'/D'=1)
Çanakkale
0.5 (L'/D'=1)
Turkey
(L'/D'=1)
0
0.50 1.00 2.00 3.00
E'/D'

279
APPENDIX G
FIGURES SHOWING THE VARIATION OF SAFETY LEVEL

CORRESPONDING TO THE RECOMMENDED LOAD AND
RESISTANCE FACTORS
3.8
3.6
Reliability index
3.4
Safety level
Target reliability
3.2
2.8
0.25 0.5 1 2
L'/D'
Figure G.1 Variation of safety level for D+L combination

3.4
3.3
3.2
3.1
Reliability index
3
2.9
2.8 Safety level
2.7 Target reliability
2.6
2.5
2.4
0.5 1 2 3
S'/D'
Figure G.2 Variation of safety level for D+S combination

280
3.3
3.2
3.1
3 Safety level
Reliability index
(L'/D'=0.5)
2.9 Safety level
(L'/D'=1)
2.8
Target reliability
2.7
2.6
2.5
2.4
0.25 0.5 1 2
W'/D'
Figure G.3 Variation of safety level for D+Lapt+W combination

3.1
2.9
Safety level
Reliability index
(L'/D'=0.5)
2.8
Safety level
(L'/D'=1)
2.7
Target reliability
2.6
2.5
2.4
0.25 0.5 1 2
W'/D'
Figure G.4 Variation of safety level for D+L+Wapt combination

281
2.2
1.8
Safety level
Reliability index
1.6
(L'/D'=0.5)
1.4 Safety level
(L'/D'=1)
0.8
0.6
0.5 1 2 3
E'/D'
Figure G.5 Variation of safety level for D+Lapt+E combination

3.2
3
Reliability index
2.8
Safety level
2.6
Target reliability
2.4
2.2
2
0.25 0.5 1 2
L'/D'

282
3.3
3.2
3.1
Reliability index
2.9 Safety level
2.8 Target
2.7
2.6
2.5
2.4
0.5 1 2 3
S'/D'

3.3
3.2
3.1
3 Safety level
Reliability index
(L'/D'=0.5)
2.9 Safety level
(L'/D'=1)
2.8
Target reliability
2.7
2.6
2.5
2.4
0.25 0.5 1 2
W'/D'

283
3
2.9
Safety level
2.8
Reliability index
(L'/D'=0.5)
Safety level
2.7 (L'/D'=1)
Target reliability
2.6
2.5
2.4
0.25 0.5 1 2
W'/D'

2.2
1.8
Safety level
Reliability index
1.6
(L'/D'=0.5)
1.4 Safety level
(L'/D'=1)
0.8
0.6
0.5 1 2 3
E'/D'

284
3.8
3.6
3.4
Reliability index
3.2 Safety level
3 Target reliability
2.8
2.6
2.4
0.25 0.5 1 2
L'/D'

3.6
3.4
3.2
Reliability index
Safety level
3
Target reliability
2.8
2.6
2.4
0.5 1 2 3
S'/D'

285
3.6
3.4
3.2 Saf ety level

Reliability index
(L'/D'=0.5)
Saf ety level
3 (L'/D'=1)
Target reliability
2.8
2.6
2.4
0.25 0.5 1 2
W'/D'

3.6
3.4
3.2 Safety level

Reliability index
(L'/D'=0.5)
Safety level
3 (L'/D'=1)
Target reliability
2.8
2.6
2.4
0.25 0.5 1 2
W'/D'

286
2.2
1.8
1.6 Safety level
Reliability index
(L'/D'=0.5)
1.4 Safety level
(L'/D'=1)
0.8
0.6
0.5 1 2 3
E'/D'

3.4
3.3
3.2
3.1
Reliability index
3
Safety level
2.9
Target reliability
2.8
2.7
2.6
2.5
2.4
0.5 1 2 3
S'/D'

(RC columns in shear failure mode)
287
3.4
3.3
3.2
3.1
Reliability index
3
Safety level
2.9
Target reliability
2.8
2.7
2.6
2.5
2.4
0.5 1 2 3
S'/D'

3.4
3.3
3.2
3.1
Safety level
Reliability index
3 (L'/D'=0.5)
Safety level
2.9 (L'/D'=1)
2.7
2.6
2.5
2.4
0.25 0.5 1 2
W'/D'

288
3.6
3.4
3.2 Safety level

Safety level
3 (L'/D'=1)
Target reliability
2.8
2.6
2.4
0.25 0.5 1 2
W'/D'

2.2
1.8
Saf ety level
Reliability index
1.6 (L'/D'=0.5)
Saf ety level
1.4 (L'/D'=1)
Target reliability
1.2
0.8
0.6
0.5 1 1.5 2 2.5 3
E'/D'

289
3.8
3.6
3.4
Reliability index
3.2 Safety level
3 Target reliability
2.8
2.6
2.4
0.25 0.5 1 2
L'/D'

(RC shear walls in flexural failure mode)
3.4
3.3
3.2
3.1
Reliability index
3
Safety level
2.9
Target reliability
2.8
2.7
2.6
2.5
2.4
0.5 1 2 3
S'/D'

290
3.6
3.4
3.2 Safety level
Reliability index
(L'/D'=0.5)
Safety level
3 (L'/D'=1)
Target reliability
2.8
2.6
2.4
0.25 0.5 1 2
W'/D'

3.6
3.4
3.2 Safety level

Reliability index
(L'/D'=0.5)
Safety level
3 (L'/D'=1)
Target reliability
2.8
2.6
2.4
0.25 0.5 1 2
W'/D'

291
2.2
1.8
1.6 Safety level

1.4 Safety level
(L'/D'=1)
0.8
0.6
0.5 1 2 3
E'/D'

292
CURRICULUM VITAE
PERSONAL INFORMATION
Surname, Name: Fırat, Fatih Kürsat

Nationality: Turkish (TC)
Date and Place of Birth: 14 August 1973, Malatya
email: fkfirat@gmail.com.tr
EDUCATION
Degree Institution Year of Graduation

PhD METU, Department of Civil Engineering 2001-2007
Selcuk University, Department of Civil
MS 1997-1999
Engineering
Fırat University, Department of Civil
BS 1990-1995
Engineering
ACADEMIC EXPERIENCE
Year Place Rank

2000-2007 METU Department of Civil Engineering Research Assistant
Selcuk University, Department of Civil
1998-2000 Research Assistant
Engineering
Nigde University, Department of Civil
1996-1998 Research Assistant
Engineering
WORK EXPERIENCE
Year Place Rank

1995-1996 Sirkeciogulları Construction Company Civil Engineering
1994-1995 Gücar Construction Company Civil Engineering
293

Turkey RCC Load Combi

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Turkey RCC Load Combi

Uploaded by

Copyright:

Available Formats

DEVELOPMENT OF LOAD AND RESISTANCE FACTORS FOR

REINFORCED CONCRETE STRUCTURES IN TURKEY

FATİH KÜRŞAT FIRAT

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

DEVELOPMENT OF LOAD AND RESISTANCE FACTORS FOR

submitted by FATİH KÜRŞAT FIRAT in partial fulfillment of the requirements

Prof. Dr. Canan Özgen __________________

Prof. Dr. Güney Özcebe __________________

Prof. Dr. M. Semih Yücemen __________________

Examining Committee Members:

Prof. Dr. M. Yaşar Kaltakcı __________________

Prof. Dr. M. Semih Yücemen __________________

Assoc. Prof. Dr. H. Şebnem Düzgün __________________

Assist. Prof. Dr. Erdem Canbay __________________

Assist. Prof. Dr. Murat Güler __________________

Name, Last Name: Fatih Kürşat Fırat

DEVELOPMENT OF LOAD AND RESISTANCE

Firat, Fatih Kürsat

September 2007, 293 pages

In this dissertation, a study is conducted to develop a probability based load and

Keywords: Uncertainty, Load and Resistance Factors, Reliability, Reliability Index,

YÜK VE DAYANIM KATSAYILARININ

Fırat, Fatih Kürşat

Eylül 2007, 293 sayfa

Bu tez çalışmasında, Türkiye koşullarına göre istatistiksel yöntemler kullanılarak

Türkiye’deki geçerli tasarım kriterine uygun olarak değişik yük kombinasyonları

Anahtar Kelimeler: Belirsizlik, Yük ve Dayanım Katsayıları, Güvenilirlik,

I would like to express my deepest appreciation to my thesis supervisor Prof. Dr. M.

I also would like to send my thanks to Turkish Meteorological Department,

A. STATISTICAL PARAMETERS OF 7 AND 28 DAY COMPRESSIVE

B. STATISTICAL PARAMETERS OF BCIII(A) REINFORCING

C. ANNUAL MAXIMUM SNOW DEPTHS FOR DIFFERENT

D. ANNUAL MAXIMUM WIND SPEEDS FOR DIFFERENT

E. DAILY MAXIMUM WIND SPEEDS OBSERVED DURING THE YEAR

F. FIGURES SHOWING THE VARIATION OF RELIABILITY INDEX

G. FIGURES SHOWING THE VARIATION OF SAFETY LEVEL

Table 3.1 Statistical parameters of compressive strength data according to years

A Cross sectional area

A Peak ground acceleration

A0 Effective ground acceleration coefficient

Asw Cross sectional area of stirrups

A(T1) Spectral acceleration coefficient

C Coefficient related to the fundamental period of vibration of the structure

C1,C2 Tabulated coefficients of Type I distribution

Ce Exposure factor for snow load

Ci Influence coefficient of load “i”

Cs Snow load coefficient

Ct Thermal factor of snow load

c Air density constant

c Distance from the neutral axis to outer compressive fiber in a T cross-section

D Dead load effect

d Depth of the member

de Effective depth of the member

E Earthquake load effect (the lateral seismic base shear)

Es Modulus of elasticity of steel reinforcement

FX,fx Cumulative distribution function (CDF) and probability density function of

fc Concrete compressive strength

fct Tensile strength of concrete

fy Yield strength of steel bars

fyw Yield strength of shear reinforcement

G Gust factor of wind load

g(x) Limit state function

h Ground snow depth