b1461930.006592.001

DYNAMIC SHEAR AMPLIFICATION IN SEISMIC RESPONSE OF
STRUCTURAL WALL SYSTEMS
by
Umut Utku CELEP
B.S., Civil Engineering, Dokuz Eylül University, 1998
M.S., Earthquake Engineering, Boğaziçi University, 2001
Submitted to the Kandilli Observatory and

Earthquake Research Institute in partial fulfillment of
the requirements for the degree of
Doctor of Philosophy
Graduate Program in Earthquake Engineering

Boğaziçi University
2008
iii
To my precious wife Beyza CELEP

iv
ACKNOWLEDGEMENTS
I would like to express my deepest gratitude to Professor Nuray Aydınoğlu for his
invaluable guidance and help during the preparation of this dissertation, who had been a
father to me as well as a wise mentor when I need. I would like to thank for his patience
and understanding for the times I had to split focus to my job as a practicing engineer.
I would like to express my sincere gratefulness to my dear wife Beyza Celep for her
support and encouragement. This dissertation would not have been completed without her
presence.
I would also like to mention my special gratitude to my boss and mentor Osman
Müyesser for his support and understanding.
Finally, special thanks to my old friend and colleague Göktürk Önem for his valuable
discussions and feedback.
v
ABSTRACT
DYNAMIC SHEAR AMPLIFICATION IN SEISMIC RESPONSE OF

STRUCTURAL WALL SYSTEMS
Previous research indicates that, shear force demand in yielding walls are not
proportional to the design moments calculated by code procedures and higher shear force
demands develop along the wall with respect to code predictions as a consequence of the
higher mode effects after the plastic hinge formation at the base of the wall. 2007 version
of the Turkish Seismic Design Code takes the dynamic shear amplification phenomena into
account with a constant base shear amplification factor of 1.5 regardless of the first mode
period and ductility level of the wall. However, results obtained from extensive non-linear
time history analyses performed on generic walls in this study indicate that dynamic shear
amplifications increase with increasing first mode period, Strength Reduction Factor (R)
and ground motion intensity. A dynamic base shear amplification relationship as a function
of the first mode period and strength reduction factor has been proposed for the Turkish
Seismic Design Code (2007), based on the regression analysis of the non-linear time
history analysis. A story shear force profile has been suggested for the Turkish Seismic
Design Code (2007), which is intended for not only preventing shear failures at the base
but also along the height of the wall. As a side product of the non-linear time history
analyses, a moment profile has also been proposed for use in the Turkish Seismic Design
Code (2007). A modal decomposition technique is presented in this study for
demonstrating the effects of the higher modes on the dynamic shear amplification
phenomenon.
vi
ÖZET
PERDE SİSTEMLERDE DEPREM NEDENİYLE OLUŞAN DİNAMİK

KESME KUVVETİ BÜYÜTMESİ
Geçmiş çalışmalar, perde tabanında plastik kesit oluşması sonrasında ortaya çıkan
yüksek mod etkileri nedeniyle kesme kuvveti talebinin yönetmelik yöntemleri ile elde
edilen tasarım kesme kuvvetlerinden büyük olduğunu ve tasarım momentleri ile uyumsuz
olduğunu göstermiştir. 2007 Türk Deprem Yönetmeliği dinamik kesme kuvveti
büyütmesini periyot ve deprem yükü azaltma katsayısından (R) bağımsız olarak tasarım
taban kesme kuvvetlerine uygulanan 1.5 katsayısı ile dikkate almaktadır. Fakat temsili
perde duvarlar üzerinde yapılan zaman tanım alanında doğrusal olmayan analizlerinin
sonucunda, dinamik kesme kuvveti büyütmelerinin artan birinci mod periyodu ve deprem
yükü azaltma katsayısı ile arttığını göstermektedir. Zaman tanım alanında doğrusal
olmayan analizlerden elde edilen sonuçlara yapılan regresyon analizlerinden; 2007 Türk
Deprem Yönetmeliği’nde kullanılabilecek, periyot ve deprem yükü azaltma katsayısının
fonksiyonu olan bir dinamik taban kesme kuvveti büyütmesi ilişkisi önerilmiştir. Bunun
yanında, 2007 Türk Deprem Yönetmeliği’nde kullanılabilecek perde yüksekliği boyunca
kesme göçmelerini de engellemek amacıyla bir kat kesme kuvveti dağılımı önerilmiştir.
Yapılan zaman tanım alanında doğrusal olmayan analizlerin bir yan ürünü olarak, 2007
Türk Deprem Yönetmeliği için bir perde moment profili önerilmiştir. Çalışmada, yüksek
modların dinamik kesme kuvveti büyütmesi üzerindeki etkisi bir modal ayrıştırma tekniği
ile gösterilmiştir.
vii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS..................................................................................................iv
ABSTRACT...........................................................................................................................v
ÖZET ....................................................................................................................................vi
LIST OF FIGURES ..............................................................................................................ix
LIST OF TABLES..............................................................................................................xiii
LIST OF SYMBOLS ..........................................................................................................xiv
1. INTRODUCTION .............................................................................................................1
1.1. Motivation of the Study .............................................................................................2
1.2. Scope of Work ...........................................................................................................2
2. LITERATURE REVIEW ..................................................................................................4
3. DESIGN OF THE GENERIC STRUCTURAL WALLS................................................11
3.1. Pre-Design of Generic Structural Walls...................................................................12
3.2. Design of Generic Structural Walls .........................................................................14
4. A NOVEL MODAL DECOMPOSITION TECHNIQUE FOR IDENTIFYING
HIGHER MODE EFFECTS IN DYNAMIC SHEAR AMPLIFICATION....................17
4.1. Linear and Non-Linear Modes of Vibration ............................................................18
4.2. Modal Decomposition Technique ............................................................................23
4.2.1. Modal Expressions in The Linear Response ..................................................23
4.2.2. Incremental Form of Modal Expressions in The Non-Linear Response........24
5. NON-LINEAR TIME HISTORY ANALYSES OF GENERIC WALLS.......................37
5.1. Group 1 Analyses and Results .................................................................................39
5.5. Remarks on Results Obtained from Non-Linear Time History Analyses ...............57
6. PROPOSED DYNAMIC SHEAR AMPLIFICATION FACTOR
RELATIONSHIP AND STORY SHEAR PROFILE.....................................................58
6.1. Comparison of the Proposed Amplification Relationship with the Other
Relationships............................................................................................................60
6.2. Proposed Story Shear Force Profile Along the Wall Height ...................................65
viii
6.3. Implications on Shear Design of Structural Walls...................................................68

6.4. Proposed Moment Profile Along the Wall Height...................................................72
7. CONCLUSIONS .............................................................................................................77
APPENDIX A: DESIGN MOMENTS AND STORY SHEAR FORCES OF THE
GENERIC WALLS.........................................................................................................80
APPENDIX B: LINEAR AND NON-LINEAR MODE SHAPES OF GENERIC
WALL BUILDINGS.......................................................................................................85
APPENDIX C: GROUND MOTION RECORDS FOR TIME HISTORY
ANALYSES....................................................................................................................90
APPENDIX D: ILLUSTRATIVE WALL DESIGN BY USING THE PROPOSED
BASE SHEAR AMPLIFICATION RELATIONSHIP, SHEAR AND MOMENT
PROFILE ........................................................................................................................98
REFERENCES ..................................................................................................................103
REFERENCES NOT CITED ............................................................................................107
ix
LIST OF FIGURES
Figure 2.1. Inertia forces before and after yielding at the base of the wall ......................4
Figure 2.2. Base shear amplification factors suggested by Derecho and Corley
(1984) .............................................................................................................6
Figure 2.3. Base shear amplification factors suggested by Seneviratna and
Krawinkler (1994)..........................................................................................8
Figure 3.1. Illustration of generic walls considered in this study...................................11
Figure 3.2. Design acceleration response spectrum used in the study ...........................14
Figure 3.3. Modelling representation of structural walls................................................15
Figure 4.1. Linear and non-linear 1st mode shapes of the 16 story generic wall ...........19
Figure 4.2. Linear and non-linear 2nd mode shapes of the 16 story generic wall..........20
Figure 4.3. Linear and non-linear 3rd mode shapes of the 16 story generic wall ..........20
Figure 4.4. Identification of linear and non-linear phases of the wall through plastic
hinge rotations at the base ............................................................................25
Figure 4.5. Plastic rotation history at the base of 16 story generic wall (R=6)
obtained from non-linear time history analysis (Whittier Narrows, 1987-
Brea Dam Record)........................................................................................28
Figure 4.6. Modal and superposed base shear history of the 16 story generic wall
designed with R=6........................................................................................29
Figure 4.7. Agreement between the base shear of the 16 story generic wall obtained
from mode superposition and Group 2 non-linear time history analysis .....30
Figure 4.8. Agreement between the overturning moment of the 16 story generic wall
obtained from mode superposition and Group 2 non-linear time history
analysis.........................................................................................................30
Figure 4.9. Increment of modal seismic forces of the 16 story generic wall designed
with R=6 (Immediately before the peak response instant)...........................31
Figure 4.10. Modal seismic forces of the 16 story generic wall designed with R=6 (at
the peak response instant) ............................................................................32
Figure 4.11. Modal and superposed story shear forces of the 16 story generic wall
designed with R=6 (at the peak response instant)........................................33
x
Figure 4.12. Increment of modal moments of the 16 story generic wall designed with
R=6 (immediately before the peak response instant)...................................34
Figure 4.13. First ten seconds of the modal and superposed dynamic base shear
amplification factor history of the 16 story generic wall designed with
R=6...............................................................................................................35
Figure 5.1. Illustration of damping ratio assumption in time history analyses ..............38
Figure 5.2. Illustration of analytical model for Group 1 analyses..................................39
Figure 5.3. Story moment diagrams of 16 story wall with strength level R=6 (Group
1 time history analyses)................................................................................40
Figure 5.4. Story shear force diagrams of 16 story wall with strength level R=6
(Group 1 time history analyses) ...................................................................41
Figure 5.5. Mean story moment diagram of 16 story wall with strength level R=6
Figure 5.6. Mean story shear force diagram of 16 story wall with strength level R=6
Figure 5.7. Mean story shear force amplification factors computed for the 16 story
wall with strength level R=6 (Group 1 time history analyses).....................43
Figure 5.8. Mean base shear amplification factors obtained from group 1 time history
analyses ........................................................................................................44
Figure 5.15. Mean base shear amplification factors obtained from Group 3 time
history analyses ............................................................................................48
xi
Figure 6.1. Agreement between the proposed relationship and mean base shear
amplification factors.....................................................................................59
Figure 6.2. Base shear amplification factors computed by using Ghosh and
Markevicius (1990) relationship ..................................................................61
Figure 6.3. Ratio of the base shear amplification factors between the relationship
suggested by Ghosh and Markevicius (1990) and this study.......................61
Figure 6.4. Base shear amplification factors computed by using Eibl and Keintzel
(1988) relationship .......................................................................................62
xii
suggested by Eibl and Keintzel (1988) and this study .................................63
Figure 6.6. Base shear amplification factors computed by using Rutenberg and Nsieri
(2006) relationship .......................................................................................64
suggested by Rutenberg and Nsieri (2006) and this study...........................64
Figure 6.8. Observed trend of the shear force profile obtained from non-linear time
history analyses of generic walls..................................................................66
Figure 6.9. Proposed story shear force profile................................................................66
Figure 6.10. Eurocode 8.1 (CEN, 2004) story shear force profile ...................................67
Figure 6.11. Shear force profile proposed by Rutenberg and Nsieri (2006) ....................68
Figure 6.12. Proposed moment profile .............................................................................72
Figure 6.13. Agreement between proposed moment profile and non-linear time history
analysis of the generic walls designed with R=2 .........................................73
Figure 6.16. Agreement Between proposed moment profile and non-linear time
history analysis of the generic walls designed with R=8 .............................75
analysis of the generic walls designed with R=10 .......................................75
xiii
LIST OF TABLES
Table 3.1. Generic wall characteristics .........................................................................13

Table 4.1. Linear and non-linear higher mode periods of the 16 story generic wall ....22
xiv
LIST OF SYMBOLS
R Strength reduction factor

T1 First mode period of the wall corresponding to gross section
stiffness
T1−cr First mode period of the wall corresponding to cracked section
stiffness
Hw Total wall height
ω1 Natural vibration frequency of the wall
EI g Bending stiffness of the wall corresponding to gross section

properties
m( y ) Mass per unit length of the wall
βb Dynamic base shear amplification factor
lw Dimension of the wall perpendicular to major bending axis
Ψo Flexural overstrength factor

Veb Design base shear obtained by using code procedures
Vab Amplified design base shear
M eb Moment demand at the base of the wall obtained from code

procedures
M rb Provided moment strength of the section at the base of the wall
M Eb Elastic moment demand at the base of the wall obtained from

code procedures
M nb Nominal moment strength of the section at base of the wall
associated with characteristic material strengths.
Vfb Sliding shear strength at the base of the wall provided by dowel
action of the longitudinal reinforcement
Asw Total longitudinal reinforcement area in the web of the section
at the base of the wall
xv
f yk Characteristic yield strength of reinforcement.

1
1. INTRODUCTION
Structural walls systems constitute a noteworthy share in high rise building stock
built in seismic zones. Structural walls offer advantages not only to structural engineers
such as their considerable lateral stiffness and strength and consequent reduction of seismic
deformation demands; but also to architects by whom the structural system can also be
used as partitioning elements with reduced disturbance.
On the other hand structural walls exhibit the disadvantage of redundancy problems
under seismic attack unless they are designed to respond fully elastic, which would be
impractical and uneconomical if not impossible. Thus, general design trend is to
incorporate Strength Reduction Factors (hereinafter will be called as R factors) to reduce
seismic demands on the walls by taking the risk of an acceptable damage that is deemed to
develop under design ground motion levels. Traditionally strength reduction factors
uniformly apply to all design seismic effects including wall moments and shear forces, that
is, wall design shear forces are assumed to be proportional to design moments. However,
previous nonlinear time history analyses by various researchers revealed that, this is not the
case for wall structures designed to respond the seismic attack in the nonlinear range and
shear force demands are greater than the demands expected in the usual design process.
This amplification with respect to design shear forces have been attributed to the higher
mode effects after yielding occurring at the base of the wall; however quantified
explanation of this phenomenon and the range of amplification is still an open field of
research.
Current rationale of most seismic design codes is to inhibit brittle shear failure modes
while attaining the flexural capacity at the base of the wall, which stems from capacity
design principles. However, amplified shear forces, particularly at the base of the wall,
may cause unexpected premature brittle shear failure modes of diagonal tension, diagonal
compression and sliding before the plastic hinge development if required measures are not
taken. This may cause destructive consequences for the structural wall systems designed
for high ductility demands since adequate reinforcement and detailing might not be
provided for an extra shear force demand.
2
1.1. Motivation of the Study
Despite the fact that the effect of higher modes are held responsible for shear force
amplification after the plastic hinge formation at the base of the wall, analytical
demonstration of the problem is rather limited in the literature.
Dynamic shear force amplification is generally taken into account in design by

means of base shear amplification factors applied to the design base shear derived from
code procedures. Most countries have incorporated base shear amplification factors in their
national seismic design codes based on their perspectives to the problem. These
amplification factors are presented principally as a function of first mode period and/or
strength reduction factors. 1997 version of the Turkish Seismic Design Code (1997) did
not incorporate such a factor in the shear design of structural walls; however the 2007
version takes the base shear amplification phenomena into account with a constant base
shear amplification factor of 1.5 regardless of the first mode period and ductility level of
the wall. It is believed that this must be revised with more rigorous base shear
amplification factors consistent with the expected behavior of structural walls in the non-
linear range. It is also believed that a shear force profile is also necessary in the Turkish
Seismic Design Code (2007) not only for preventing shear failures at the base but also
along the height of the structural wall.
1.2. Scope of Work
This work comprises a parametric study performed on generic walls for deriving a
dynamic base shear amplification factor relationship, which could be proposed for in lieu
of the constant base shear amplification in the Turkish Seismic Design Code (2007). A
modal decomposition technique has been presented to demonstrate the responsibility of
higher modes in dynamic shear amplifications.
Previous studies on dynamic shear amplification have been covered in Chapter 2,

with emphasis given to various country perspectives on the subject. Design and
characteristics of the generic walls that will be the basis of the numerical findings in this
study are presented in Chapter 3. Effect of higher modes on the dynamic shear
3
amplifications has been demonstrated in Chapter 4 by using a modal decomposition

technique. Subsequently, a relationship for dynamic base shear amplification factors and
shear force as well as bending moment profile along the wall height have been suggested
for the Turkish Seismic Design Code (2007) in Chapter 6 based on the results of extensive
non-linear time history analyses described in Chapter 5.
4
2. LITERATURE REVIEW
Dynamic shear amplification of structural wall systems designed to respond non-

linearly during seismic attack has been documented by limited number of researchers in
the past. Majority of the researchers had focused on the base shear amplification as the
maximum shear forces develop at the base of the wall. Inevitably, research objectives on
the subject have been founded on a particular code perspective since the motivation was
the code development for inhibiting premature shear failure of structural walls. Majority of
the shear amplification recommendations by researchers are based on parametric non-
linear time history analyses and presented in the form of base shear amplification factors.
However, descriptive research on the analytical demonstration of the effect of higher
modes on the base shear amplification is scarce. The dynamic shear amplification is
principally explained in the literature by the deviation of the centroid of inertia forces in
the non-linear phase with respect to the centroid of inertia forces prescribed by the
triangular code distribution as illustrated in Figure 2.1.
LOWERED CENTROID
VE1
VE2
heff
heff
My My
VE2 > VE1
INERTIA FORCE DISTRIBUTION AT INERTIA FORCE DISTRIBUTION
FIRST MODE RESPONSE AT THE ONSET AFTER YIELDING AT THE BASE
OF YIELDING AT THE BASE
Figure 2.1. Inertia forces before and after yielding at the base of the wall
Dynamic base shear amplification effect is generally presented by researchers in the

form of a factor (presented as β b in this study) that is multiplied with the base shear
5
derived from conventional code procedures ( Veb ) to obtain the expected base shear demand
( Vab )as given in Equation (2.1).
Vab = β bVeb (2.1)
Dynamic base shear amplification has not been recognized until the work of
Blakeley et al. (1975), who pioneered the idea that higher mode effects amplify the base
shear with respect to the base shear obtained from conventional code procedures after the
formation of a plastic hinge at the base of the wall. They stated that after a plastic hinge
formation at the base of the wall, the center of lateral inertial loading becomes lower or
higher with respect to the center predicted by the code distribution, effectively causing
higher shear forces than predicted by the code. Blakely et al. (1975) performed non-linear
time history analyses on various structural wall models and recommended base shear
amplification factors as a function of number of stories (n) of the wall structure. Based on
these recommendations the provisions of New Zealand seismic code NZS3101 (Standards
New Zealand 1982) have changed in 1982 and formed the basis of the New Zealand code
perspective for the base shear amplification factors as given in Equation (2.2) which is still
in use (Standards New Zealand 2006).
n
ωn = 0.9 + n≤6
10
(2.2)
n
ωn = 1.3 + ≤ 1.8 n>6
30
It is seen from Equation (2.2) that, New Zealand perspective on the base shear
amplification is dependent on the first mode period (fundamental period) of the wall,
which is implicitly related to the number of stories of the wall (n).
A parametric investigation study carried out by Derecho and Corley (1984) revealed
the increasing dependency of base shear amplification not only to the first mode period but
also to the ductility level of the wall. The authors presented their base shear amplification
results in the form of a chart format as shown in Figure 2.2. Derecho and Corley also stated
6
that there is a linear relationship between the base shear amplification factor and intensity
of the ground motion.
Figure 2.2. Base shear amplification factors suggested by Derecho and Corley (1984)
Kabeyasawa (1987) proposed base shear amplification factors, as given in Equation

(2.3), based on the analytical verification of observed data from full scale pseudo dynamic
test of a seven story model structure:
Vmax = Vn + c × W × PGA (2.3)
in which Vn is the base shear capacity for the structure calculated by ultimate limit state
analysis assuming a triangular force distribution, c represents a coefficient that stands for
the higher mode effects varying between 0.25 and 0.30, W is the seismic weight of the wall
and PGA is the design peak ground acceleration. Kabeyasawa concluded that maximum
dynamic shear in a wall under seismic attack is higher than the statically computed shear
7
due to the higher mode effects and the additional shear force demand in the non-linear
range is a function of the peak ground acceleration.
A study by Eibl and Keintzel (1988) has also shown that base shear amplification is
both functions of the first mode period and expected ductility level. The authors have
proposed a base shear amplification factor relationship as given in Equation (2.4), which
later formed the basis of the Eurocode 8 EN 1998-1 (CEN, 2004) provisions for base shear
amplifications of wall systems designed for high ductility. Eibl and Keintzel have stated in
their study that the first term under the square root denotes the contribution the first mode
that is reduced by the strength reduction factor and the second term implicitly denotes the
importance of the second mode in yielding systems.
2 2
⎛γ M ⎞ ⎛ S (T ) ⎞
ε = q ⎜ rd rd ⎟ + 0.1⎜ e C ⎟ ≤ q (2.4)
⎝ q M Ed ⎠ ⎝ Se (T1 ) ⎠
in which ε is the base shear amplification factor, q is the strength reduction factor, γRd the
overstrength factor, MRd the design flexural strength at the wall base, MEd the design
moment at the wall base, TC the upper limit period of the spectral acceleration plateau in
the design spectrum and Se(T) is the ordinate of the elastic response spectrum.
Ghosh and Markevicius (1990) has concluded in their parametric study that no
dependence with the first mode period could be established and claimed that dynamic
shear amplification is likely to be a function of ground motion intensity. The authors has
suggested a relationship for the maximum base shear demand after yielding at the base of
the wall, in which the first term is a function of the intensity of the ground motion and the
second term is the base shear corresponding to plastic hinge formation at the base wall as
presented in Equation (2.5).
PGA My
Vmax = 0.25 × W × + (2.5)
g 0.67 H
where W is the seismic weight of the wall, PGA the design peak ground acceleration, My
the yield moment capacity at the base of the wall and H is the total wall height.
8
Experimental investigations on nine small-scale reinforced concrete structural

models by Eberhard and Sozen (1993) revealed that base shear response of structural walls
are very sensitive to higher mode effects. The authors observed that story inertial-forces
differed increasingly from the triangular distribution assumed by the code provisions after
yielding initiates at the base of the wall. Eberhard and Sozen demonstrated that modal
stiffness of the first mode will practically be zero after yielding at the base and incremental
shear force demand in this stage will be due to higher mode effects.
Seneviratna and Krawinkler (1994) have emphasized the dependency of the shear
amplification of structural walls on both strength level and the first mode period of the
structure as shown in Figure 2.3. The authors stated that shear amplifications increase for
longer periods and lower strength levels, which can be attributed to the higher mode
vibrations, particularly the second mode, of the elastic portion of wall above the plastic
hinge at the base. They also compared their results with the relationship proposed by
Ghosh and Markevicius (1990) and concluded that Ghosh and Markevicius relationship
predicts the base shear amplifications in good agreement with their results for medium-
high walls with relatively low strength levels, however underestimates amplifications for
tall structures due to the lack of a term related to the period of the structure.
Figure 2.3. Base shear amplification factors suggested by Seneviratna and Krawinkler
(1994)
9
Filiatrault et al. (1994) have studied the problem from the Canadian code approach
and concluded that brittle shear failures are possible for ductile walls designed by the
Canadian seismic code provisions. Based on a parametric study of walls designed
according the Canadian code, they suggested seismic zone dependent strength reduction
factors of 1.0 or 1.5 for shear design to account for base shear amplifications.
Rutenberg and Nsieri (2006) have studied the problem through Eurocode 8
perspective and claimed that current provisions of Eurocode 8 are in need of revision since
they do not predict the shear amplifications realistically for the medium and high ductility
classes of structural walls. Based on a parametric study by using non-linear time history
analyses, the authors recommended the relationship given in Equation (2.6) in lieu of the
current Eurocode relationship for dynamic base shear amplifications.
ε = 0.75 + 0.22 × (T + q + Tq) (2.6)
in which q is the strength reduction factor and T is first mode period of the structure.
Rutenberg and Nsieri (2006) concluded that modification is necessary to Eurocode 8 base
shear amplification relationship, which should increase with increasing period and strength
reduction factors and should take into account the distribution of plasticity along the wall
height.
Krawinkler (2006) expressed that “Once a plastic hinge forms at the base of a wall
structure, the dynamic response characteristics of the structure above the base change
radically. Different ‘modes’ become predominant and the effective lever arm (M/V ratio)
becomes smaller, and for relatively tall and slender structures it becomes much smaller.
Correspondingly, the dynamic shear force that can be generated may become much larger
than that indicated by a static M/V concept in which the resultant lateral force is applied
approximately at 2/3 of the structure height (or higher).” The author emphasized that
despite the detrimental reality that base shear demand is higher than predicted by
traditional code provisions, higher actual shear strength of walls than recommended by
code equations may be compensating this negative effect. The author also pointed out a
very critical issue of wall behavior based on the recent test results, where wall response is
10
often controlled by sliding at the base and shear force distribution may be totally different
than predicted by analytical methods.
It is interesting to note that USA has been rather hesitant to apply dynamic shear
amplification factors in the national seismic design codes, despite several researchers in the
USA have stressed the fact on shear amplifications in their studies. As an exception,
Commentary to the Recommended Lateral Force Requirements of the Structural Engineers
Association of California (SEAOC, 1999) recommended adopting the amplification factors
advocated by the New Zealand code.
Turkish Seismic Design Code (1997) did not take into account the dynamic base
shear amplifications in structural walls at any level of ductility, and the design for shear
was based on the shear forces corresponding to the lateral force distribution used for
flexural design. Present version of the code (2007) takes the base shear amplification
phenomena into account with a constant base shear amplification factor of 1.5 regardless of
the first mode period and ductility level of the wall.
Seismic design codes could prefer utilizing base shear amplification factors
suggested by researchers, as long as they are practical to implement in design practice.
However they have a serious drawback of being related to the intensity of ground motion
records they are derived from. Research on shear amplification of structural walls is still
continuing and there is limited global consensus between the researchers (also between the
codes) on the degree of base shear amplifications, despite the fact that the source of this
amplification is recognized as the higher mode effects initiating after the plastic hinge
formation at the base of the wall.
11
3. DESIGN OF THE GENERIC STRUCTURAL WALLS
The numerical and statistical findings in this study are based on the results obtained
from the analysis of generic structural walls that are intended to represent the structural
wall systems in the building stock. Thus, generic structural walls with 8, 12, 16, 20 and 30
stories and typical story heights of 3.0m have been considered as illustrated in Figure 3.1,
in order to cover a broad range of wall characteristics in terms of flexibility and strength.
m
m
m
m
m
m
m
m
m
m
m m
m m
m m
m m
30 @ 3m = 90 m
m m m
m m m
m m m
m m m
m m m m
20 @ 3m = 60 m
m m m m
m m m m
16 @ 3m = 48 m
m m m m
m m m m m
12 @ 3m = 36 m
m m m m m
m m m m m
8 @ 3m = 24 m
m m m m m
m m m m m
m m m m m
m m m m m
m m m m m
Figure 3.1. Illustration of generic walls considered in this study
Firstly generic walls have been pre-designed to provide section dimensions.

Subsequently, the traditional design procedure as applied in practice has been followed
12
where possible so as not to deviate from realistic results. Details of the pre-design and
design steps are presented in the following sections of this chapter.
Analysis and design effects obtained herein have been computed by using SAP 2000
v11 (CSI, 2006) structural analysis software licensed to the Department of Earthquake
Engineering of Boğaziçi University.
3.1. Pre-Design of Generic Structural Walls
A simple approach has been utilized to determine the section dimensions of the
generic walls. It has been assumed that the first mode period corresponding to gross
(uncracked) stiffness shall approximately satisfy an old empirical relation correlated with
the total building height (H) as given in Equation (3.1) (Turkish Seismic Design Code,
1997).
3
T1 = 0.05 H w 4 (3.1)
Thus, by assigning building heights to Equation (3.1), first mode periods have been
determined for each generic wall in concern. Assuming that the wall mass and stiffness are
uniformly distributed along the wall height, the natural vibration frequency of the system
can be estimated by Equation (3.2) (Chopra 2006);
3.516 EI g
ω1 = (3.2)
H w2 m( y )
where Ig is the moment of inertia of the gross section properties and m(y) is the mass per
unit length of the wall.
Thus the first mode period of the system will be;
2π
T1 = (3.3)
ω1
13
Substituting Equation (3.1) in Equation (3.3) will result in Equation (3.4) to be

solved for Ig/m(y) , provided that modulus of elasticity for concrete (E) is assumed as 3x107
kN/m2.
3 2π
0.05 H w 4 =
3.516 EI g (3.4)
H w2 m( y )
Solution of Equation (3.4) for the generic walls resulted in the following first mode
period, section dimensions and typical story masses, as given in Table 3.1.
Table 3.1. Generic wall characteristics
WALL Hw (m) T1 (sec) b(m) d(m) mtypical (t)

8 STORY 24 0.54 0.30 4.00 30.95
12 STORY 36 0.73 0.30 5.75 30.12
16 STORY 48 0.91 0.30 7.50 29.68
20 STORY 60 1.08 0.30 9.30 29.88
30 STORY 90 1.46 0.30 13.85 29.73
14
3.2. Design of Generic Structural Walls
Subsequent to the determination of characteristics of generic structural walls as

above, structural design procedure based on response spectrum method has been utilized.
Design basis acceleration response spectrum given in Turkish Seismic Design Code
(2007) corresponding to 10 percent probability of exceedence in 50 years has been used for
the representation of ground motion (Figure 3.2). The design acceleration response
spectrum has been assumed to be based on local site conditions with site class Z3 (corner
periods of TA = 0.15sec and TB = 0.60 sec) and Seismic Zone 1 (Effective peak ground
acceleration of 0.40g), respectively.
12
10
8
A (m/s )
2
0
0.00 0.50 1.00 1.50 2.00 2.50 3.00
Period (sec)
Figure 3.2. Design acceleration response spectrum used in the study
Generic structural walls have been modeled with frame (beam) elements in terms of
stiffness and lumped masses at each story level as shown in Figure 3.3. Flexural stiffness
of the walls has been reduced by a factor 0.50 to obtain cracked section stiffness,
approximating the cracking and softening of linear stiffness under seismic attack.
The boundary conditions at the base of the walls have been assumed to be fixed, thus
rotations arising from the foundation or soil stiffness have been neglected.
15
Figure 3.3. Modelling representation of structural walls
Response spectrum analyses have been performed for each generic wall in order to
obtain the design story moments and shear forces along the wall height. All modes of the
structure exhibiting modal amplitudes in the lateral direction, that is equal to the number of
stories of the wall in concern, have been taken into account during response spectrum
analysis, and critical damping ratio has been assumed as 5 percent.
Each generic wall has been analyzed by taking into account Strength Reduction
Factors R = 2, 4, 6, 8 and 10 in order to provide a strength range. These Strength Reduction
Factors have been incorporated into response spectrum analyses by reducing the design
spectrum as per the requirements of the Turkish Seismic Design Code 2007. It can be
argued that realistic wall buildings in practice are not designed for strength reduction
factors of 8 or 10, which would result in drastically under-designed structures. However, it
is intended that high values of R may have some merit in cases of seismic assessment of
old structural walls designed with lesser seismic demands and weaker capacities.
Modal response quantities derived from response spectrum analyses have been
combined with the CQC modal combination method to obtain the design effects, namely
the story moments and shear forces, which are presented in Appendix A for each generic
wall with their corresponding strength reduction factors.
16
Design moments obtained at the base of the wall from response spectrum analyses
with the corresponding strength reduction factor have been hypothetically assumed as
equal to the design moment capacity or yield moment strength of the section at the base of
the generic walls. Thus, story moment and shear force profiles along the wall height have
also been supposed to be equal to the profiles obtained from the response spectrum
analyses as the yield moment is attained at the base.
17
4. A NOVEL MODAL DECOMPOSITION TECHNIQUE FOR

IDENTIFYING HIGHER MODE EFFECTS IN DYNAMIC SHEAR
AMPLIFICATION
As with systems exhibiting cantilever like behavior, structural wall systems are
statically unstable but dynamically stable structures when flexural yielding occurs at their
bases. During seismic excitation, as the system alters from linear to nonlinear behavior
mode, a plastic hinge or a series of hinges develop at the base of the wall and the structural
wall loses its redundancy for a time interval until it is stabilized back by inertia forces.
During this time interval, a new system with a completely unique dynamic characteristics
prevail until the system velocity becomes zero, and the linear phase initiates with
unloading. Status of the structural wall before and after the plastic development at the base
of the wall shall be called hereinafter as “the linear phase” and “the non-linear phase”,
respectively, for the ease of understanding.
It has been observed in this study and by various researchers described in Chapter 2
that greater shear forces are obtained from nonlinear time history analyses than the shear
forces expected from design. It is a general consensus that higher mode effects in the non-
linear phases causes this amplification of shear forces with respect to the ones obtained
from linear design, however studies demonstrating individual contribution of higher mode
effects to dynamic amplification are very limited in the literature. Kabeyasawa (1987) has
called the response of second mode effects as “residual fluctuating forces”. Based on a
triangular first mode shape, he predicted the second mode shape from orthogonality
condition with the first mode and extracted the second mode response by eliminating the
triangular first mode effects from the non-linear time history analyses.
A modal decomposition technique is presented in this study for demonstrating the

individual effects of the higher modes on the dynamic shear amplification phenomenon.
The technique is based on inverse extraction of modal responses from non-linear time
history analyses results by using the mode superposition method. A computer program has
been coded in MATLAB (The MathWorks Inc., 2004) for the numerical presentation of
18
the technique. The technique has been demonstrated on the non-linear time history results
of 16 story generic wall designed with a Strength Reduction Factor of R=6 which have
been obtained from the Whittier Narrows 1987 - Brea Dam (Downstream) record in Group
2 analysis as explained in Chapter 5. (single plastic hinge at the base of the wall with
elasto-plastic moment-curvature hysteretic characteristics-see Section 5.2).
4.1. Linear and Non-Linear Modes of Vibration
Modal characteristics of the wall in the linear and non–linear phases should be
determined prior to the implementation of the modal decomposition technique. Vibration
modes before the plastic hinge development will be denoted as the linear modes, whereas
vibration modes in the non-linear phases of the plastic hinge will be referred to as the non-
linear modes. Eigenvalues and mode shapes both for linear and non-linear phases of the
walls have been obtained by using a built-in feature of MATLAB. This feature utilizes the
Jacobi Method (Bathe, 1996) that converge accurate eigenvalues regardless of the stiffness
matrix conditions, as is the case for the wall systems in the non-linear phase.
General expression for the dynamic eigenvalue problem can be derived from the
equation of motion of an undamped multi degree of freedom system having a mass matrix
of M and a stiffness matrix of K as:
KΦ n = ωn2 MΦ n (4.1)
where, Φn and ωn2 denote the eigenvector and eigenvalue of the n’th mode of the system,
respectively.
Solution of Equation (4.1) is straightforward for the linear phase of the wall, however
the system is theoretically unstable in the nonlinear phase due the plastic hinge formation
and thus eigenvalues can not be obtained through classical methods.
It is interesting to note that the Jacobi Method is able to provide the eigenvalues of
the system even in the non-linear phase since the solution of the eigenvalue problem by the
Jacobi Method does not contain any division by any number. Hence, the method always
19
converges and yields an accurate solution for positive, zero or negative eigenvalues.
(Bathe, 1996; Wilson, 2006)
All linear and non-linear eigenvalues and corresponding eigenvectors having

components in the direction of earthquake excitation are necessary to be computed in this
technique for accurate calculation of modal accelerations. Thus, for an n story wall
subjected to unidirectional horizontal ground motion n eigenvalues and associated
eigenvectors in the lateral direction should be computed.
Figure 4.1 to Figure 4.3 shows the first three of the computed 16 linear and non-
linear mode shapes (eigenvectors) of the 16 story generic wall, respectively.
1st MODE SHAPE

16
LIN. PHASE
14 NONLIN. PHASE
12
10
STORY
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
MODAL AMPLITUDE
Figure 4.1. Linear and non-Linear 1st mode shapes of the 16 story generic wall
20
2nd MODE SHAPE

16
LIN. PHASE
14 NONLIN. PHASE
12
10
STORY
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
MODAL AMPLITUDE
Figure 4.2. Linear and non-linear 2nd mode shapes of the 16 story generic wall
3rd MODE SHAPE

16
14
12
10
STORY
2 LIN. PHASE
NONLIN. PHASE
0
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
MODAL AMPLITUDE
Figure 4.3. Linear and non-linear 3rd mode shapes of the 16 story generic wall
21
Figure 4.1 reveals a transformation of the first mode shape pattern from parabolic to
straight line as a consequence of the plastic hinge development at the base of the wall in
the non-linear phase. Period of vibration in the first mode has been calculated as 1.30 sec
in the linear phase, whereas in the non-linear phase it is infinite (ω1=0). The zero natural
frequency indicates that modal stiffness for the first mode becomes effectively zero after
yielding takes place provided that no hardening is assumed for the moment-curvature
relationship at plastic hinge forming at the base. Thus, non-linear first mode can be called
as the unstable mode of the system, which would be stabilized by the inertia forces during
seismic excitation.
However, the non-linear first mode of the system has a definite Modal Participation
Factor, since eigenvectors are definite
Φ1*MTx
Γ1* = (4.2)
Φ1*T MΦ1*
where the superscript asterisk indicate the non-linear mode.
It is also interesting to note that, non-linear first mode of the system satisfy the
orthogonality conditions with the non-linear second mode as reported by Kabeyasawa
(1987):
Φ1*T MΦ*2 = 0 (4.3)
Modification of the second and third mode shapes in the non-linear phase has been
observed as a shift of the bulk of the mode shape downwards as shown in Figure 4.2 and
Figure 4.3, due to the hinge development at the base of the wall. Similar behavior has been
observed for higher modes of the wall above the third non-linear mode. In order to
demonstrate the eigenvectors of the other generic walls buildings with 8, 12, 16, 20 and 30
stories, first two mode shapes, both in the linear and non-linear phases, are presented in
Appendix B.
22
Non-zero periods of vibration have been computed for the higher modes as shown in
Table 4.1. Those modes can be called as stable modes regardless of the plastic hinge
formation at the base of the wall.
Table 4.1. Linear and non-linear higher mode periods of the 16 story generic wall
LINEAR NON-LINEAR
PHASE PHASE
T (sec) T (sec)
MODE 2 0.2145 0.3064
MODE 3 0.0767 0.0947
MODE 4 0.0392 0.0454
MODE 5 0.0237 0.0266
MODE 6 0.0159 0.0174
MODE 7 0.0114 0.0123
MODE 8 0.0086 0.0092
MODE 9 0.0067 0.0071
MODE 10 0.0054 0.0057
MODE 11 0.0045 0.0047
MODE 12 0.0038 0.0039
MODE 13 0.0033 0.0034
MODE 14 0.0030 0.0030
MODE 15 0.0027 0.0028
MODE 16 0.0026 0.0026
23
4.2. Modal Decomposition Technique
4.2.1. Modal Expressions in The Linear Response
Static-equivalent m’th mode seismic force at time t can be expressed as:
f m (t ) = M Φ m Γ m am (t ) (4.4)
where am (t ) refers to m’th mode pseudo-acceleration response history. Φ m represents

m’th mode shape vector and Γ m is the corresponding modal participation factor:
ΦTm M 1
Γm = (4.5)
ΦTm M Φ m
in which 1 refers to the vector of ones.
Define scaled mode shape vectors as:
Φm = Φm Γm
(4.6)
Φ = [Φ1 Φ 2 ..... Φ m ..... Φ n ]
where n refers to total number of degrees of freedom in the direction of seismic excitation.
Effective mass matrix can be defined as:
M = MΦ (4.7)
Thus, mode-superposition can be applied to the seismic force vector given in

Equation (4.4) for the m’th mode:
f (t ) = M a(t ) (4.8)
where pseudo-acceleration vector is defined as:

24
T
a(t ) = a1 (t ) a2 (t ) .... am (t ) .... an (t ) (4.9)
4.2.2. Incremental Form of Modal Expressions in The Non-Linear Response
Equation (4.4) can be written in incremental form at the (i)’th step representing the
response increment in between the time stations ti and ti+1 :
m Γ m ∆am
∆f m(i) = M Φ(i) (i) (i)
(4.10)
In this case Φ(i)

m represents either linear or non-linear m’th mode shape vector
prevailing in the (i)’th step, depending on whether the base plastic hinge response is along
loading/unloading lines or on the yielding plateau of the response hysteresis. Now ∆am (t )
refers to m’th mode pseudo-acceleration response increment in the (i)’th step.
Thus Equation (4.8) can be written in incremental form as,
∆f (i) = M (i) ∆a(i) (4.11)
where,
M (i) = M Φ(i) (4.12)

Φ(i) (i) (i)
m = Φm Γm
(4.13)
Φ(i) = [Φ1(i) Φ(i)
2 ..... Φ(i) (i)
m ..... Φ n ]
Thus, in nonlinear response, two different n by n effective mass matrices, M (i) , are
defined, one each corresponding to linear and non-linear modes, respectively.
In the implementation of modal decomposition technique, seismic force increment

∆f k(i) acting on any story k is obtained as the difference of the story shear demand
increments at consecutive stories k and k–1 as follows:
∆f k(i) = ∆Vk(i) − ∆Vk(i)−1 (4.14)

25
From plastic rotation time-history response at the base of the wall, the time steps
associated with loading/unloading and the non-linear phases of the hysteretic response of
the base plastic hinge can be detected as illustrated in Figure 4.4.
ZERO ROTATIONAL
MOMENT VELOCITY
(LINEAR MODES)
LINEAR PHASE
PLASTIC
ROTATION (rad)
ZERO ROTATIONAL
VELOCITY NON-LINEAR PHASE
(NON-LINEAR MODES)
Figure 4.4. Identification of linear and non-linear phases of the wall through plastic hinge
rotations at the base
Utilizing the appropriate M (i) matrix (corresponding to linear or non-linear modes),

increment of pseudo-acceleration vector at any (i)’th step can be obtained from Equation
(4.11) as:
∆a(i) = M (i) −1 ∆f (i) (4.15)

T
∆a(i) = ∆a1(i) ∆a1(i) .... ∆am(i) .... ∆an(i) (4.16)
Thus modal increments of seismic forces, ∆f m(i) , can be readily obtained from
Equation (4.10). Further, base shear and base bending moment modal increments can be
obtained as:
∆Vmb(i) = 1 ∆f m(i) (4.17)

∆M mb(i) = y ∆f m(i) (4.18)
26
Adding to those at the previous time station ti, m’th mode base shear and base
bending moment at time station ti+1 can be obtained as:
Vmb (ti+1 ) = Vmb (ti ) + ∆Vmb(i) (4.19)

M mb (ti+1 ) = M mb (ti ) + ∆M mb(i) (4.20)
Final quantities at time station ti+1 are obtained by mode superposition:
V b (ti+1 ) = V1b (ti+1 ) + V2b (ti+1 ) + .... + Vmb (ti+1 ) + .... + Vnb (ti+1 ) (4.21)
M b (ti+1 ) = M 1b (ti+1 ) + M 2b (ti+1 ) + .... + M mb (ti+1 ) + .... + M nb (ti+1 ) (4.22)
which should be exactly equal to those directly obtained from the nonlinear response
history analysis of the wall.
Equation (4.10) can also written in terms of incremental modal displacements as

shown in Equation (4.23):
( )
2
m Γ m ∆d m ω m
∆f m(i) = M Φ(i) (i) (i) (i)
(4.23)
Increments of modal seismic forces for the first non-linear mode should be obtained
as zero from Equation (4.23) since the frequency associated with the first non-linear mode
is obtained as zero provided that there is no strain hardening at the base plastic hinge.
Thus, increments of modal base shear and modal overturning moment associated with the
first non-linear mode should also be zero from Equations (4.17) and (4.18). This finding
indicates that first mode of the wall in the non-linear phase do not contribute to any force
related response. However, non-linear first mode contribute to increments of modal
deformations as the associated incremental displacements will be non-zero as shown in
Equation (4.24).
∆u1(i) = Φ1(i) Γ1(i) ∆d1(i) (4.24)
On the contrary, higher modes in the non-linear phases will maintain their
contribution to base shear and deformation response as their frequencies and incremental
27
displacements are non-zero. On the other hand, it is interesting to note that increments of
modal overturning moments associated with non-linear higher modes shall also be zero at
the base since the product
( )
2
m Γ m ∆d m ωm
y Φ(i) (i) (i) (i)
(4.25)
is calculated as zero for these modes. Thus, it can be stated that non-linear higher modes
initiates non-zero incremental shear forces, which also satisfy zero increments of modal
overturning moments at the base of the wall. Hence, during time intervals of non-linear
phase, incremental shear forces shall continue to develop on the structural wall until the
commencement of unloading in the plastic hinge, which modify the proportionality
assumed between the base shear forces and overturning moments by code procedures.
Figure 4.5 shows the plastic rotation history of the 16 story generic wall designed
with R=6, which has been arranged to demonstrate only the non-linear phases of the wall
with non-zero plastic rotations (linear phases of the wall are shown with zero rotations).
The wall has exhibited the first non-linear response at time t = 4.62 seconds of the record
and sustained this phase for 0.04 seconds with plastic rotations in the positive direction. A
total of 40 non-linear phases has been noted throughout the response history of the generic
wall with 38 of them are associated with plastic rotations in the negative direction which
indicates a response around a residual deformation developed after t = 5.0 seconds.
28
-3
x 10
1
-1
PLASTIC ROTATION (rad)
-2
-3
-4
-5
-6
-7
0 5 10 15 20 25 30
TIME (sec)
Figure 4.5. Plastic rotation history at the base of 16 story generic wall (R=6) obtained from
non-linear time history analysis (Whittier Narrows, 1987-Brea Dam Record)
Modal base shear history of the 16 story generic wall, decomposed with the
technique explained above, is given in Figure 4.6 for the first five modes of the wall. Base
shear response combined with mode superposition method is also provided in last pane of
Figure 4.6. With reference to Figure 4.5, maximum base shear response of the 16 story
generic wall has been observed in the first non-linear phase of the wall which occurred at t
= 4.65 seconds (hereinafter called “the peak response instant”) of the response history as
indicated in Figure 4.6 with a dark circle. This non-linear phase has been observed to
initiate at t = 4.62 seconds and continued through 0.04 seconds to switch to a linear phase
with unloading in the plastic hinge. It is evident from Figure 4.6 that, modal base shear of
the second mode also reaches its first peak value at the peak response instant, where the
contribution of the first mode is comparatively lower. Effect of third and higher modes has
been observed to be negligible at peak response instant as shown in Figure 4.6, indicating
the significant role of the second mode on the base shear response.
Base shear and overturning moment response histories of the 16 story generic wall
computed by the superposition of decomposed modal responses have been compared with
the actual base shear and overturning moment response histories obtained from the Group
29
2 non-linear time history analysis in order to verify the accuracy of the technique as shown
in Figure 4.7. and Figure 4.8, respectively. Both Figure 4.7 and Figure 4.8 demonstrate an
almost perfect match between the responses obtained from mode superposition and non-
linear time history analysis, which indicates that non-linear response can be decomposed
into modal responses by using modal characteristics of the structural walls in an
incremental fashion.
BASE SHEAR (kN)
2000
MODE 1
1000
0
-1000
-2000
0 5 10 15 20 25
BASE SHEAR (kN)
2000
MODE 2
1000
0
-1000
-2000
0 5 10 15 20 25
BASE SHEAR (kN) BASE SHEAR (kN)
2000
MODE 3
1000
0
-1000
-2000
0 5 10 15 20 25
2000
MODE 4
1000
0
-1000
-2000
0 5 10 15 20 25
BASE SHEAR (kN)
2000
MODE 5
1000
0
-1000
-2000
0 5 10 15 20 25
BASE SHEAR (kN)
2000
MODAL SUP.
1000
0
-1000
-2000
0 5 10 15 20 25
TIME (sec)
Figure 4.6. Modal and superposed base shear history of the 16 story generic wall designed
with R=6
30
2000
1500
1000
500
BASE SHEAR (kN)
-500
-1000
-1500
MODAL SUP.
-2000 NL. TIME HIST.
-2500
0 5 10 15 20 25 30
TIME (sec)
Figure 4.7. Agreement between the base shear of the 16 story generic wall obtained from
mode superposition and Group 2 non-linear time history analysis
4
x 10
2.5
2
MODAL SUP.
1.5 NL. TIME HIST.
OVERTUNING MOMENT(kNm)
0.5
-0.5
-1
-1.5
-2
-2.5
0 5 10 15 20 25 30
TIME (sec)
Figure 4.8. Agreement between the overturning moment of the 16 story generic wall
obtained from mode superposition and Group 2 non-linear time history analysis
31
Modal increment of seismic forces associated with the first three modes of the 16
story generic wall obtained immediately before the peak response instant are shown in
Figure 4.9.
RESPONSE IMMEDIATELY BEFORE t=4.65 sec

16
MODE 1
MODE 2
14
MODE 3
12
10
STORY
0
-8 -6 -4 -2 0 2 4 6
MODAL SEISMIC FORCE INCREMENT (kN)
Figure 4.9. Increment of modal seismic forces of the 16 story generic wall designed with
R=6 (Immediately before the peak response instant)
Figure 4.9 demonstrates zero increment of seismic forces corresponding to non-linear

first mode of the structural wall, where modal increments of seismic forces associated with
the higher modes are non-zero regardless of the plastic hinge formation at the base of the
wall.
Modal seismic force pattern of the wall at the peak response instant is shown in
Figure 4.10, which emphasizes considerably higher modal seismic forces provided by the
non-linear second mode with respect to other modal contributions.
32
RESPONSE AT t=4.65 sec

16
14
12
10
STORY
MODE 1
4
MODE 2
MODE 3
2 MODE 4
MODE 5
0
-300 -200 -100 0 100 200 300
MODAL SEISMIC FORCE (kN)
Figure 4.10. Modal seismic forces of the 16 story generic wall designed with R=6 (at the
peak response instant)
Figure 4.11 demonstrates the significant effect of the non-linear second mode to
story shear forces combined with mode superposition at the peak response instant. Story
shear force quantities are given as absolute values in Figure 4.11 for the ease of
interpretation. It is evident from Figure 4.11 that majority of story shear forces in the non-
linear phases of the wall are provided by the non-linear second mode. Story shear forces
associated with the first mode at the peak response instant shown in Figure 4.11 are those
retained from the prior linear phase, since increments of seismic forces associated with the
first mode are zero in this phase as demonstrated in Figure 4.9. Contribution of the third
and higher modes to story shear force response are virtually negligible based on the
observation that modal increments of seismic forces associated with these modes tend to
reduce the modal shear response in the non-linear phase contrary to the case for the second
non-linear mode.
33
RESPONSE AT t=4.65 sec

16
MODE 1
14 MODE 2
MODE 3
MODE 4
12
MODE 5
MODAL SUP.
10
STORY
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
MODAL STORY SHEAR FORCE (kN)
Figure 4.11. Modal and superposed story shear forces of the 16 story generic wall designed
with R=6 (at the peak response instant)
Effect of the non-linear second mode has also been observed on the moment
response of the 16 story generic wall as shown in Figure 4.12 in terms of modal increment
of moments obtained immediately before the peak response instant. Increment of modal
moments associated with the first non-linear mode has been obtained as zero due to zero
increment of seismic forces associated with this mode. However, second non-linear mode,
with zero modal moment increment at the base, has been observed to induce moments at
upper story levels. This effect may contribute to the spread of plasticity at the upper levels
of the wall in the non-linear phases as observed in Group 1 and Group 3 non-linear time
history analysis (See Chapter 5).
34
RESPONSE IMMEDIATELY BEFORE t=4.65 sec

16
MODE 1
14 MODE 2
MODE 3
12
10
STORY
0
-350 -300 -250 -200 -150 -100 -50 0 50 100
MODAL MOMENT INCREMENT (kNm)
Figure 4.12. Increment of modal moments of the 16 story generic wall designed with R=6
(Immediately before the peak response instant)
Dynamic base shear amplification factors can also be presented in terms of modal
contributions. Figure 4.13 shows first ten seconds of modal dynamic amplification factor
history of the 16 story generic wall, obtained by dividing each modal base shear to the
design base shear of the wall. Dynamic base shear amplification factor combined with
mode superposition method is also provided in last pane of Figure 4.13 and dynamic
amplification factor at the peak response instant has been marked with a dark dot for the
ease of understanding.
35
DYN. AMP. FACT.
3
2 MODE 1
1
0
-1
-2
-3
0 1 2 3 4 5 6 7 8 9 10
DYN. AMP. FACT.
3
2 MODE 2
1
0
-1
-2
-3
0 1 2 3 4 5 6 7 8 9 10
DYN. AMP. FACT.
3
2 MODE 3
1
0
-1
-2
-3
0 1 2 3 4 5 6 7 8 9 10
DYN. AMP. FACT.
3
2 MODE 4
1
0
-1
-2
-3
0 1 2 3 4 5 6 7 8 9 10
DYN. AMP. FACT.N
3
2 MODE 5
1
0
-1
-2
-3
0 1 2 3 4 5 6 7 8 9 10
DYN. AMP. FACT.
3
2 MODAL SUP.
1
0
-1
-2
-3
0 1 2 3 4 5 6 7 8 9 10
TIME (sec)
Figure 4.13. First ten seconds of the modal and superposed dynamic base shear
amplification factor history of the 16 story generic wall designed with R=6
Figure 4.13 indicates that, non-linear second mode is chiefly responsible for the
dynamic base shear amplifications, where effect of higher modes above the second mode
has been found to be negligible at the peak response instant. Dynamic base shear
amplification factor associated with the first mode has been computed greater than unity
between 6.30 seconds and 6.40 seconds of the response history (Figure 4.13), due to slight
numerical errors in the decomposition process.
Based on the results of this study, it is believed that dynamic amplification

phenomenon should be related to second mode effect rather than higher mode effects.
36
Significant role of the second mode on dynamic shear amplifications has been
demonstrated in this study by utilizing a modal decomposition technique. It has been
shown that, response of yielding structural walls can be decomposed into modal response
contributions by using the modal decomposition technique explained above.
37
5. NON-LINEAR TIME HISTORY ANALYSES OF GENERIC

WALLS
Owing to the cantilever-like characteristics of structural wall systems, maximum

shear force effects occur at the base of the wall. Thus, focusing on the amplification factors
of the base shear forces has a merit for inhibiting brittle modes of failure associated with
shear, such as the diagonal tension failure, diagonal compression failure and sliding at the
cold joints.
In this chapter, base shear amplification factors and their relation with the first mode
period of walls and the strength reduction factors have been investigated through a
parametric study based on non-linear time history analyses. Base shear amplification
factors are proposed based on regression analyses of the parametric time history analysis
results, which are indented to be used in Turkish Seismic Design Code (2007) for shear
design of walls. Furthermore, a story shear force envelope along the wall height is
recommended and interrelated with the amplified base shear.
The generic walls with geometric characteristics given in Chapter 3 have been
analyzed with non-linear time history analyses by using step-by-step integration procedure
in the time domain. Each wall has been consecutively analyzed for the strength levels
corresponding to strength reduction factors of R=2, 4, 6, 8 and 10 with 20 strong ground
motion acceleration records scaled to fit the Turkish Seismic Design Code (2007)
Acceleration Response Spectrum as explained in Appendix C.
Newmark’s average acceleration method (γ=1/2 and β=1/4) has been used as an
unconditionally stable algorithm in step-by-step integration of non-linear time history
analyses. In order to avoid the accuracy problems, time steps has been chosen as 0.005 sec,
which is below 1/10 of the third mode period of all generic walls.
Damping matrix of the system has been constructed with Rayleigh method that is,
the damping matrix has been assumed to be proportional to the stiffness and mass matrices
of the walls. Damping ratios for the first and the third modes have been chosen to be 1 per
38
cent, so as not to over emphasize the damping effects in the second mode as illustrated in
Figure 5.1
ξ 1= ξ3= 0.01
ξ2
ω1 ω2 ω3 ω
Figure 5.1. Illustration of damping ratio assumption in time history analyses
Non-linear time history analyses has been executed under four groups (hereinafter
called Group 1, Group 2, Group 3 and Group 4) in order to investigate the sensitivity of the
base shear amplification factors to the distribution of plastic hinges along the wall height
and the hysteresis characteristics of the plastic hinges. Modeling details and base shear
amplification results for each group are presented in the subsequent sections. Due to vast
number of analyses performed in this study, only the moment and story shear force
diagrams along the wall height, that have been obtained from the generic 16 story wall
with strength reduction factor of R=6 has been presented.
Group 1 and Group 2 non-linear time history analyses have been performed by using
SAP 2000 v11 (CSI, 2006) structural analysis software, whereas Group 3 and Group 4
non-linear time history analyses have been performed by using RUAUMOKO (Carr, 2000)
39
structural analysis software, which are both licensed to the Department of Earthquake
Engineering of Boğaziçi University.
5.1. Group 1 Analyses and Results
Group1 analyses involve the generic wall analytical models of multiple plastic hinge
distribution along the wall height with elastoplastic moment-curvature hysteresis relations
as shown in Figure 5.2. Other structural modeling assumptions described in Chapter 3 are
also valid for the Group 1 models.
STORY j MASS
PLASTIC HINGE
STORY j-1 MASS MOMENT
PLASTIC HINGE
MASS
PLASTIC HINGE
MASS
CURVATURE
PLASTIC HINGE
STORY 3 MASS
PLASTIC HINGE
STORY 2 MASS
ELASTOPLASTIC HYSTERESIS
PLASTIC HINGE RELATION FOR PLASTIC HINGES
STORY 1 MASS
PLASTIC HINGE
Figure 5.2. Illustration of analytical model for Group 1 analyses
It has been assumed that yield moment capacity is uniform along the wall, being
equal to base design moment corresponding to the strength reduction factors as described
in Chapter 3. Actually, reinforced concrete wall sections do not exhibit perfectly
elastoplastic hysteresis characteristics during flexure as reversals due to seismic attack
become significant, and some degree of stiffness degradation should be expected. Yet,
elastoplastic plastic hinges are customarily used in most non-linear time history analyses
and it is believed that reliable results can be obtained for the benchmarking of studies with
more complex hysteresis relationships.
Non-linear time history analyses have been performed for generic wall models of 8,
12, 16, 20 and 30 stories with strength levels corresponding to R = 2, 4, 6, 8 and 10. For
40
each generic wall model with the aforementioned strength level, maximum story moments
and shear forces have been obtained for each of 20 ground motion record as shown in
Figure 5.3 and Figure 5.4 for 16 story wall with strength level of R = 6.
16
14
12
10
STORY
0
0 0.5 1 1.5 2 2.5
MOMENT (kNm) 4
x 10
Figure 5.3. Story moment diagrams of 16 story wall with strength level R=6 (Group 1 time
history analyses)
41
16
14
12
10
STORY
0
0 500 1000 1500 2000 2500 3000
STORY SHEAR FORCE (kN)
Figure 5.4. Story shear force diagrams of 16 story wall with strength level R=6 (Group 1
time history analyses)
Subsequently story moment and shear forces have been averaged to obtain mean
story moment and story shear force diagrams, as shown in Figure 5.5 and Figure 5.6 for 16
story wall with R = 6 strength level with comparison to the design moment and shear force
diagrams.
42
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 0.5 1 1.5 2 2.5
MOMENT (kNm) 4
x 10
Figure 5.5. Mean story moment diagram of 16 story wall with strength level R=6 (Group 1
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 500 1000 1500 2000 2500
(Group 1 time history analyses)
Consequently mean story shear forces obtained from time history analyses have been
divided to design story shear forces in order to obtain story shear force amplification
factors as shown in Figure 5.7 for the 16 story wall with R = 6 strength level.
43
16
14
12
10
STORY
0
1.5 2 2.5 3 3.5 4
STORY SHEAR FORCE AMPLIFICATION
Figure 5.7. Mean story shear force amplification factors computed for the 16 story wall
with strength level R=6 (Group 1 time history analyses)
As shown in Figure 5.4, there is a scatter of story shear forces due to the different
characteristics of the ground motion in spite of scaling has been applied to fit a target
response spectrum. This scatter has been observed at all generic walls and corresponding
strength levels and is obviously reflected to the base shear amplification factors; however it
is believed that averaging time history results from 20 ground motion records is justifiable
due to the sufficient number of samples in the group.
Mean base shear amplification factors derived from Group 1 non-linear time history
analyses have been collected in a single diagram (βb – T – R diagram) related with the first
mode periods of the wall corresponding the cracked section stiffness and strength reduction
factors as shown in Figure 5.8
44
GROUP 1 ANALYSES
4
R=2
R=4
MEAN BASE SHEAR AMPLIFICATION FACTOR

R=6
3.5
R=8
R=10
2.5
1.5
1
0.5 1 1.5 2 2.5
CRACKED SECTION ELASTIC FIRST MODE PERIOD (sec)
Figure 5.8. Mean base shear amplification factors obtained from Group 1 time history
analyses
45
Group 2 analyses include the generic wall analytical models with a single plastic
hinge at the base of the wall with elastoplastic moment-curvature hysteresis relation as
shown in Figure 5.9. Other structural modeling assumptions described in Chapter 3 are
also valid for the Group 2 models.
STORY j MASS
MASS
MASS
CURVATURE
STORY 3 MASS
STORY 2 MASS
ELASTOPLASTIC HYSTERESIS
RELATION FOR PLASTIC HINGE
STORY 1 MASS
PLASTIC HINGE
It is expected in real wall structures that a single plastic hinge formation at the base
of the wall is not possible and the plasticity spreads upwards from the base with increasing
seismic demand. However, it is intended that Group 2 analyses shall provide insight on the
effect of spread of plasticity on structural response with respect to Group 1 analyses.
Non-linear time history analyses have been performed for Group 2 generic wall
models similar to the Group 1 analyses. Similarly, for each generic wall model with the
corresponding strength level, maximum story moments and shear forces have been
obtained for each of 20 ground motion record. Subsequently mean story moments, story
shear forces and base shear amplification factors have been computed as shown in Figure
5.9 to Figure 5.14 for the 16 story wall with R = 6 strength level.
46
16
14
12
10
STORY
0
0 0.5 1 1.5 2 2.5 3
MOMENT (kNm) 4
x 10
Figure 5.10. Story moment diagrams of 16 story wall with strength level R=6 (Group 2
16
14
12
10
STORY
0
0 500 1000 1500 2000 2500
47
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 0.5 1 1.5 2 2.5
MOMENT (kNm) 4
x 10
Figure 5.12. Mean story moment diagram of 16 story wall with strength level R=6 (Group
2 time history analyses)
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 500 1000 1500 2000 2500
48
16
14
12
10
STORY
0
1.5 2 2.5 3 3.5 4
βb – T – R diagram derived from Group 2 non-linear time history analyses are

presented in Figure 5.15
GROUP 2 ANALYSES
5
R=2
4.5 R=4
R=6
4 R=8
R=10
3.5
2.5
1.5
1
0.5 1 1.5 2 2.5
analyses
49
Group 3 analyses involve the generic wall analytical models with multiple plastic
hinge distribution along the wall height with Takeda (1970) degrading stiffness moment-
curvature hysteretic relationships as shown in Figure 5.16. Other structural modeling
assumptions depicted in Chapter 3 are also valid for the Group 3 models.
STORY j MASS
PLASTIC HINGE
PLASTIC HINGE
MASS
PLASTIC HINGE
MASS
CURVATURE
PLASTIC HINGE
STORY 3 MASS
PLASTIC HINGE
STORY 2 MASS
TAKEDA HYSTERESIS RELATION
PLASTIC HINGE FOR PLASTIC HINGES
STORY 1 MASS
PLASTIC HINGE
It has been intended that Group 3 analyses shall reflect the effect of degrading
stiffness moment-curvature hysteretic relationship with respect to the Group 1 analyses
that were performed with elastoplastic hysteretic relationship.
Non-linear time history analyses have been performed for the generic wall models in
the Group 3 analyses similar to Group 1 and Group 2 analyses. For each generic wall
model with the corresponding strength level, maximum story moments and shear forces
have been obtained for each of 20 ground motion records. Subsequently mean story
moments, story shear forces and base shear amplification factors have been computed as
shown in Figure 5.17 to Figure 5.21 for the 16 story wall with R = 6 strength level.
50
16
14
12
10
STORY
0
0 0.5 1 1.5 2 2.5
MOMENT (kNm) 4
x 10
16
14
12
10
STORY
0
0 500 1000 1500 2000 2500 3000
51
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 0.5 1 1.5 2 2.5
MOMENT (kNm) 4
x 10
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
52
16
14
12
10
STORY
0
1.5 2 2.5 3 3.5 4 4.5

GROUP 3 ANALYSES
4
R=2
R=4
R=6
3.5
R=8
R=10
2.5
1.5
1
0.5 1 1.5 2 2.5
analyses
53
Group 4 analyses constitute the generic wall analytical models with a single plastic
hinge at the base of the wall height Takeda (1970) degrading stiffness moment-curvature
hysteresis relations as shown in Figure 5.23. Other structural modeling assumptions
described in Chapter 3 are also valid for the Group 4 models.
STORY j MASS
MASS
MASS
CURVATURE
STORY 3 MASS
STORY 2 MASS TAKEDA HYSTERESIS RELATION

FOR PLASTIC HINGES
STORY 1 MASS
PLASTIC HINGE
In spite of an unrealistic behavior expected with a single plastic hinge at the base of
the wall, it has been anticipated that Group 4 analyses shall provide insight on the effect of
spread of plasticity on structural response with respect to Group 3 analyses.
Similar to the other analysis groups, non-linear time history analyses have been
performed for the generic wall models in the Group 4 analyses. For each generic wall
model with the corresponding strength level, maximum story moments and shear forces
have been obtained for each of 20 ground motion record. Subsequently mean story
moments, story shear forces and base shear amplification factors have been computed as
shown in Figure 5.24 to Figure 5.28 for the 16 story wall with R = 6 strength level.
54
16
14
12
10
STORY
0
0 0.5 1 1.5 2 2.5 3
MOMENT (kNm) 4
x 10
16
14
12
10
STORY
0
0 500 1000 1500 2000 2500
55
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 0.5 1 1.5 2 2.5
MOMENT (kNm) 4
x 10
16
MEAN (T.HIST)
14 DESIGN
12
10
STORY
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
56
16
14
12
10
STORY
0
1.5 2 2.5 3 3.5 4

GROUP 4 ANALYSES
5
R=2
4.5 R=4
R=6
4 R=8
R=10
3.5
2.5
1.5
1
0.5 1 1.5 2 2.5
analyses
57
5.5. Remarks on Results Obtained from Non-Linear Time History Analyses
Based on the information gathered from the Group 1, Group 2, Group 3 and Group 4
non-linear time history analyses results, the following points have been observed;
• Maximum shear forces develop at the base of the wall as expected.

• Story shear forces are noticeably amplified with respect to the design values
regardless of the distribution of the plastic hinges in the model or the hysteretic
characteristics of the plastic hinges.
• Single hinge models (Group 2 and Group 4) resulted in comparatively higher
story shear force amplification factors with respect to the distributed hinge models
(Group 1 and Group 3), particularly for the models with elastoplastic plastic hinge
properties. It is believed that due to the multiple plastic hinge formations along the
wall, complex mode shapes occur, which reduce seismic forces.
• Single hinge models (Group 2 and Group 4) provide incorrect moment quantities
and patterns above the base of the wall after the plastic hinge formation, where
these quantities are higher than the yield moment capacity of the plastic hinge.
• Difference between the story shear force amplification factors obtained from
Group 1 and Group 3 are barely noticeable and it is believed that analyses using
elastoplastic moment-curvature hysteresis properties can provide reasonable
estimations with respect to analyses using degrading stiffness hysteresis
characteristics in cases the analyst lacks software with such complex hysteresis
libraries.
• Both Group 1 and Group 3 analyses results indicate that, there is a tendency of
reduction in base shear amplification factors for walls with low strength levels
(R=8 and R=10) and with high first mode periods, which has not been observed
in the single hinge model analyses (Group 2 and Group 4). This can be attributed
to the multiple plastic hinge formations along the wall and associated complex
mode shapes.
• Base shear amplification factors tend to increase with increasing Strength
Reduction Factors (or decreasing strength levels) and with increasing first mode
period, except for the wall with high first mode periods and considerably low
strength levels.
58
6. PROPOSED DYNAMIC SHEAR AMPLIFICATION FACTOR

RELATIONSHIP AND STORY SHEAR PROFILE
Based on the remarks stated in Chapter 5, it has been concluded that multiple hinge
distribution along the wall better present the expected wall behavior during seismic attack.
Moreover, it is believed that Takeda (1970) degrading stiffness hysteresis model is more
suitable to represent realistic reinforced concrete flexural behavior. Thus, regression
analysis have been performed for the base shear amplification factors derived from Group
3 non-linear time history analyses in order to propose a relationship between the base shear
dynamic amplification factor, first mode period of the wall corresponding to cracked
section stiffness and the strength reduction factor. It is suggested that the relationship
proposed herein can be used for the design and evaluation of structural wall systems in
accordance with Turkish Seismic Design Code (2007) provisions.
Prior to regression analyses, an expression as given in Equation (6.1) has been

suggested as the initial structure of the proposed relationship.
β b = 1 + (aT1-cr + b)( R − 1.5)c (6.1)
Consequent to regression analyses, regression variables a, b and c has been

calculated as:
a = 0.281 b = 0.394 c = 0.553
which formed the proposed expression given in Equation (6.1) for the dynamic base shear
amplification factors as presented in Equation (6.2);
β b = 1.0 + (0.281× T1-cr + 0.394) × ( R − 1.5)0.553 (6.2)

1.0 ≤ β b ≤ R (6.3)
The coefficient of multiple determination of the regression (R2) has been computed
as 0.9743 over 1.0000, which suggested a successful fit of the proposed relationship to the
samples. The agreement between the obtained mean dynamic base shear amplification
59
factors and the proposed dynamic base shear amplification relationship is shown in Figure
6.1.
4.5
R=2
R=4
4
BASE SHEAR AMPLIFICATION FACTOR
R=6
R=8
3.5 R=10
2.5
1.5
1
0.5 1 1.5 2 2.5
Figure 6.1. Agreement between the proposed relationship and mean base shear
amplification factors
Figure 6.1 suggests that the proposed relationship is in good agreement with the
mean amplification factors for strength reduction factors between 2 and 6, and
overestimates the amplification factors for strength reduction factors 8 and 10 in the high
period range that is believed to be acceptable in terms of a conservative approach.
60
6.1. Comparison of the Proposed Amplification Relationship with the Other

Relationships
Proposed dynamic base shear amplification relationship given in Equation (6.2) have
been compared with the relationships proposed by other researchers mentioned in Chapter
2. It should be noted that such a comparison may not be appropriate since the base shear
amplification factors are highly affected by the ground motion intensity considered,
number of ground motion records considered and modeling assumptions. However, it is
found useful to compare the trends between the proposed relationship and those proposed
by other researchers.
Figure 6.2 shows the base shear amplification factors computed for the generic walls
considered in this study by using the relationship proposed by Ghosh and Markevicius
(1990). In order to illustrate the deviation between the two relationships base shear
amplification factors computed with Ghosh and Markevicius relationship is divided by the
amplification factors obtained from the proposed relationship as shown in Figure 6.3
Figure 6.3 reveals that proposed relationship in this study provides greater base shear
amplification ratios particularly in the high period range with respect to the relationship
proposed by Ghosh and Markevicius, which may be attributed to lack of a period term in
the Ghosh and Markevicius relationship.
61
3.5
R=2
R=4

R=6
3
R=8
R=10
2.5
1.5
1
0.5 1 1.5 2 2.5
Figure 6.2. Base shear amplification factors computed by using Ghosh and Markevicius
(1990) relationship
0.95 R=2
R=4
0.9 R=6
R=8
RATIO GHOSH / SUGGESTED
0.85 R=10
0.8
0.75
0.7
0.65
0.6
0.55
0.5
0.5 1 1.5 2 2.5
Figure 6.3. Ratio of the base shear amplification factors between the relationship suggested
by Ghosh and Markevicius (1990) and this study
62
Figure 6.4 shows the base shear amplification factors by using the relationship
proposed by Eibl and Keintzel (1988). Similarly base shear amplification factors computed
with Eibl and Keintzel relationship is divided by the amplification factors obtained from
the proposed relationship as shown in Figure 6.5 for comparison.
5
R=2
4.5 R=4
R=6
R=8
4
R=10
3.5
2.5
1.5
1
0.5 1 1.5 2 2.5
Figure 6.4. Base shear amplification factors computed by using Eibl and Keintzel (1988)
relationship
63
2
R=2
R=4
R=6
R=8
R=10
RATIO EIBL / SUGGESTED
1.5
0.5
0.5 1 1.5 2 2.5
by Eibl and Keintzel (1988) and this study
Figure 6.5 indicates the lower amplification factors computed by Eibl and Keintzel
relationship where parallel-like trends of the ratios with respect to strength reduction
factors also suggests the idea that amplification factors have been affected by the different
intensities of ground motion taken into account between the two studies.
Base shear amplification factors by using the relationship proposed by Rutenberg and
Nsieri (2006) is illustrated in Figure 6.6, whereas the ratio between the base shear
amplification factors computed with Rutenberg and Nsieri relationship and the proposed
relationship is shown in Figure 6.7 for comparison.
64
6
R=2
5.5 R=4

R=6
5
R=8
R=10
4.5
3.5
2.5
1.5
1
0.5 1 1.5 2 2.5
Figure 6.6. Base shear amplification factors computed by using Rutenberg and Nsieri
(2006) relationship
2
R=2
R=4
R=6
RATIO RUTENBERG / SUGGESTED
R=8
R=10
1.5
0.5
0.5 1 1.5 2 2.5
by Rutenberg and Nsieri (2006) and this study
65
Figure 6.7 depicts slightly higher amplification factors computed by Rutenberg and
Nsieri in which the ratio trends are almost in a linear fashion with increasing strength
reduction factor and period. This result is believed to be related to the higher intensity SAC
ground motion suite (Somerville et al. 1997) used by Rutenberg and Nsieri in their studies.
It should be emphasized that despite the lack of consensus between the researchers
on the magnitude of base shear amplification factors, the above-given results address the
fact that base shear amplification factors increase with increasing strength reduction factor
and increasing first mode period of structural walls. It is believed that suggested base shear
amplification relationship in this study is suitable for the design of structural walls against
shear failures according to Turkish Seismic Design Code (2007).
6.2. Proposed Story Shear Force Profile Along the Wall Height
Mean shear force distribution patterns derived from the non-linear time history
analyses reveal that amplified shear force magnitudes drop gradually to half value at about
40 per cent of the total wall height and follow a practically uniform path up to the tip of the
wall excluding the top one or two stories as shown in Figure 6.8. It is interesting to note
that, normalized story shear force profiles shown in Figure 6.8 encompass all generic walls
designed with the strength reduction factors considered in this study (R=2, 4, 6, 8 and 10),
which indicates that amplified story shear force pattern is neither a function of first mode
period nor the strength reduction factor.
It is believed that a constant shear force zone just above the base of the wall, equal to
the critical wall height as per Turkish Seismic Design Code (2007) will be rational
considering the tension shift effects (Paulay and Priestley, 1992) in the plastic hinge zone.
Above this zone, it is suggested that the story shear forces shall linearly descend to half
magnitude at the 40 per cent of the total wall height and remain constant along the rest of
the wall as shown in Figure 6.9. In cases where 40 per cent of the wall height do not
coincide with a story level, upper end of the linearly descending zone point may be
rounded up to the nearest upper story level.
66
1
OBSERVED TREND
0.9 NL. TIME HIST.
0.8
0.7
NORMALIZED HEIGHT
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
NORMALIZED SHEAR FORCE
Figure 6.8. Observed trend of the shear force profile obtained from non-linear time history
analyses of generic walls
60% Hw
40% Hw
max(l w , Hw /6)
0.5 1.0
Figure 6.9. Proposed story shear force profile
Shear force profile proposed in this study is similar to that given in Eurocode 8 EN
1998-1 (CEN, 2004) in most aspects, as shown in Figure 6.10. Major notable differences
67
between the two profiles are the level where the shear forces drop approximately to 50 per
cent of the base shear is defined as 1/3 (33 per cent) of the wall height and no constant
shear in the tension shift zone at the base of the wall is taken into account.
67% H w
33% Hw
0.5 1.0
Figure 6.10. Eurocode 8.1 (CEN, 2004) story shear force profile
Rutenberg and Nsieri (2006) also recommended a similar story shear force profile
based on the non-linear time history results as illustrated in Figure 6.11. The plastic hinge
zone above the base has been recommended as 10 per cent of the total height by Rutenberg
and Nsieri (2006) and the level where the shear forces drop 50 per cent of the base shear
has been related with first mode period of the wall with the expression given in Equation
(6.4)
ξ = 1.0 − 0.3T1 ≥ 0.50 (6.4)
It is believed that shear force profile proposed by Rutenberg and Nsieri over punishes
the structures in the medium period range, since ξ corresponds to values in the order of 70
per cent and 85 per cent of the total wall height for a structural wall with 1.0 sec and 0.50
sec period, respectively.
68
ξH w
0.10H w
0.5 1.0
Figure 6.11. Shear force profile proposed by Rutenberg and Nsieri (2006)
6.3. Implications on Shear Design of Structural Walls
As expressed in Chapter 2, base shear amplification factors are very practical to

implement in design, however they have the drawback of being related to the intensity of
ground motion records they are derived from. Correspondingly, the base shear
amplification relationship proposed in this study is based on the base shear amplification
factors derived from ground motion records that have been scaled to fit acceleration
response spectrum intensity associated with the Seismic Zone 1 of Turkey, with 10 per
cent probability of exceedence in 50 years. It should be noted that the ground motion
records selected in this study are recorded in the mid-range distances from the faulting
plane and the designer should use the proposed relationship with caution when designing
the wall systems for near field earthquakes. The proposed relationship is free of the
earthquake intensity; nevertheless, it has the characteristics of the design intensity. It
should be noted that for ground motion intensities higher or lower than the ground motion
intensity corresponding to 10 per cent in 50 years not only the amplitudes of the ground
motion may change but also the frequency content varies. Thus, it is recommended at least
69
few non-linear time history analyses should be performed to support the proposed
relationship if the structure is complex in geometry and/or configuration or the importance
level of the structure is high. It should also be noted that the proposed relationship has been
derived from non-linear time history analyses of generic structural walls covering a limited
period range. Thus, it is recommended that the proposed relationship should not be used
for structural wall systems above 30 stories.
In majority of the cases of design practice, even flexural strength of the wall
computed with minimum reinforcement ratios can well exceed the flexural strength
required by analysis. This provided flexural overstrength effectively reduce the strength
reduction factor used in design since the plastic hinge at the base of the wall will tend to
yield at a higher moment demand than required by analysis. Thus, Equation (6.2) can be
modified to take into account the aforementioned flexural overstrength as shown in
Equation (6.5)
R
β b = 1.0 + (0.281× T1-cr + 0.394) × ( − 1.5)0.553 (6.5)
Ψ o
where, Ψo is the flexural overstrength factor as given Equation (6.6)
M rb
Ψo = (6.6)
M eb
In Equation (6.6), M eb denotes the moment demand at the base of the wall obtained
from code procedures and M rb denotes the provided moment strength of the section at the
base of the wall computed with code requirements.
Application procedure of the proposed dynamic base shear amplification relationship

and story shear force profile to shear design of structural walls is summarized below;
1. Obtain design base shear ( Veb ) by using the code procedures such as the equivalent
seismic load method or the response spectrum method with the appropriate strength
reduction factor (R)
70
2. Multiply the design base shear ( Veb ) with base shear amplification factor ( β b )
computed by using Equation (6.5) to obtain amplified design base shear ( Vab ) as
shown in Equation (6.7).
Vab = Veb × β b (T1-cr , R ) (6.7)
3. Multiply the amplified design base shear ( Vab ) with the unitless story shear profile
proposed in Section 6.2 to obtain the amplified story shear forces along the wall.
An illustrative example covering the procedure explained above is given in Appendix

D.
Even though the proposed dynamic base shear amplification relationship is intended
for use in design, it can be employed in the seismic assessment of structural wall systems
by using linear and non-linear analysis procedures as well. Strength reduction factor
obtained by using single mode pushover analysis procedures can be implemented in
Equation (6.5) in order to compute the amplified base shear forces. Implementation of a
dynamic base shear amplification factor is not necessary in multi-mode pushover analysis
(Aydınoğlu, 2003) since the analysis procedure takes the effects of higher modes into
account. For the seismic assessment of structural wall systems by using linear analysis
procedures, third term in Equation (6.5) may be modified as shown in Equation (6.8).
β b = 1.0 + (0.281× T1-cr + 0.394) × ( Ry − 1.5)0.553 (6.8)
where, Ry can be computed with:
M Eb
Ry = (6.9)
M nb
In Equation (6.9), M Eb denotes the elastic moment demand at the base of the wall
obtained from code procedures and M nb denotes the nominal moment strength of the
section at base of the wall associated with characteristic material strengths.
71
Some degree of sliding shear in the plastic hinge zone of walls is inevitable
particularly at high ductility demands. Recent tests of walls performed on E-Defense
shaking table in Japan (NIED, 2006) and UC San Diego shaking table in the US (UCSD,
2005), indicate the dominance of sliding shear in the plastic hinge zone on the wall
behavior (Krawinkler, 2006). The severity of sliding shear increase with increasing
ductility demand and this phenomenon causes strength deterioration as the number of
cycles experienced by the plastic hinge. Thus, most of the sliding shear resistance will be
provided by the dowel action of the longitudinal reinforcement as a consequence of open
cracks and deteriorated concrete strength in the plastic hinge. Hence, it may be rational to
define an upper bound for base shear amplifications where the behavior will be primarily
governed by sliding at the base of the wall after attaining sliding shear strength.
Based on the recommendations of Paulay and Priestley (1992) for the sliding shear
resistance of longitudinal reinforcement in squat walls, sliding shear resistance at the base
of a of cantilever wall can be predicted by Equation (6.10);
Vfb = 0.25 Asw f yk (6.10)
in which, Vfb is the sliding resistance provided by the longitudinal reinforcement, Asw is the
total longitudinal reinforcement area in the web of the section at the base of the wall and fyk
is the characteristic yield strength of reinforcement. Thus, upper bound of the base shear
amplification factor can be expressed as given in Equation (6.11)
0.25 Asw f yk
β max
b
= (6.11)
Veb
It is believed that the upper bound limitation of base shear amplifications as given in
Equation (6.11) will provide rational results for design and prevent unnecessary overdesign
in practice.
72
6.4. Proposed Moment Profile Along the Wall Height
As a side product of this study, a unitless moment profile along the wall height is
suggested for the Turkish Seismic Design Code (2007), based on the regression of
normalized moments obtained from non-linear time history analyses of the generic walls. It
has been observed in this study that moment profile is strongly dependent on the strength
reduction factor utilized, due to the increasing spread of plasticity above the base of the
wall with increasing strength reduction factor. However, such a dependency was not
observed for the first mode period. A tri-linear moment profile is suggested as shown in
Figure 6.12.
Hw
ηH w
δHw
λ 1.0
Figure 6.12. Proposed moment profile
Intermediate points on the proposed moment profile are suggested as a function of

strength reduction factor as given Equation (6.12), Equation (6.13) and Equation (6.14). It
should be noted that minimum value of the strength reduction factor should be taken as 2
in the equations since it has been assumed that strength reduction factors below 2 will
correspond to linear response.
73
λ = 1.05 − 0.025 × R (6.12)

δ = 0.003 × ( R − 2) 2 + 0.025 × ( R − 2) (6.13)
η = 0.125 × ( R − 2) − 0.005 × ( R − 2) 2 (6.14)
Coefficients of multiple determination of variables λ, δ and η have been calculated

as 0.989, 0.999 and 0.987, respectively.
Agreement of the proposed moment profile with those obtained from non-linear time
history analyses are shown in Figure 6.13 to Figure 6.17 with respect to the strength
reduction factor utilized.
1
PROPOSED PROFILE
0.9 NL. TIME HIST.
0.8
0.7
NORMALIZED HEIGHT
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
NORMALIZED MOMENT
analysis of the generic walls designed with R=2
74
1
PROPOSED PROFILE
0.9 NL. TIME HIST.
0.8
0.7
NORMALIZED HEIGHT
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
NORMALIZED MOMENT
1
PROPOSED PROFILE
0.9 NL. TIME HIST.
0.8
0.7
NORMALIZED HEIGHT
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
NORMALIZED MOMENT
75
1
PROPOSED PROFILE
0.9 NL. TIME HIST.
0.8
0.7
NORMALIZED HEIGHT
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
NORMALIZED MOMENT
1
PROPOSED PROFILE
0.9 NL. TIME HIST.
0.8
0.7
NORMALIZED HEIGHT
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1
NORMALIZED MOMENT
76
It should be noted that tension shift effect (Paulay and Priestley 1992) on the
proposed moment profile has not been taken into account. Equation (6.13) and Equation
(6.14) should be modified as shown below to incorporate the tension shift effect in
accordance with the Turkish Seismic Design Code (2007):
lw
δ = 0.003 × ( R − 2) 2 + 0.025 × ( R − 2) + max( , 0.167) (6.15)
Hw
l
η = 0.125 × ( R − 2) − 0.005 × ( R − 2) 2 + max( w , 0.167) (6.16)
Hw
77
7. CONCLUSIONS
Dynamic shear amplification phenomenon in structural walls has been studied by

limited number of researchers in the last 30 years and most of these research findings have
formed the basis of code provisions in seismic prone countries. Previous research revealed
that, shear forces in yielding walls are not proportional to the design moments and shear
force demands are higher than predicted by the code procedures due to higher mode effects
after plastic hinge formation at the base of the wall.
Amplified shear forces, particularly at the base of the wall, are probable to cause
unexpected premature brittle shear failure modes of diagonal tension, diagonal
compression and sliding before the intended plastic hinge development if the magnified
shear forces are not taken into account. Practically dynamic shear amplification is taken
into account by means of base shear amplification factors applied to the design base shear
calculated with the code procedures. These amplification factors are presented principally
as a function of first mode period and/or strength reduction factors. 2007 version of the
Turkish Seismic Design Code considers the base shear amplification phenomena with a
constant base shear amplification factor of 1.5 regardless of the first mode period and
ductility level of the wall. It is believed that this must be revised with base shear
amplification factors as functions of strength reduction factor and first mode period of the
wall. It is also believed that a shear force profile along the height of the wall is also
necessary in the Turkish Seismic Design Code (2007) not only for preventing shear
failures at the base but also along the height of the structural wall.
Extensive parametric non-linear time history analyses have been performed for code
designed generic structural walls, in order to suggest a base shear amplification
relationship and a shear force profile for the Turkish Seismic Design Code (2007).
Distribution and hysteretic characteristics of the plastic hinges have been considered in the
analyses models of the parametric study, in order to investigate the effect of modeling
assumptions to dynamic shear amplifications. It has been observed that dynamic shear
amplifications are not dependent on the hysteretic characteristics of the plastic hinge.
However, models with single plastic hinge at the base of the wall resulted in comparatively
78
higher story shear force amplification factors and incorrect moment patterns with respect to
distributed plastic hinge models. Results of the non-linear time history analysis revealed
that dynamic base shear amplification factors increase with increasing first mode period as
well as with strength reduction factor. A base shear amplification relationship as functions
of first mode period and strength reduction factor has been proposed based on a regression
analysis of the non-linear time history analysis results. It was suggested that dynamic base
shear amplifications at the base of the wall should be limited with the amplifications
corresponding to the sliding shear strength during load reversals for a better
characterization of the wall behavior under seismic attack. Moreover, a practical story
shear force profile has been suggested, based on the non-linear time history analysis
results, to be used in tandem with the Turkish Seismic Design Code (2007). It has been
observed that story shear force profile is neither a function of first mode period nor the
strength reduction factor. A moment profile has also been proposed for use in the Turkish
Seismic Design Code (2007) as a side product of the parametric non-linear time history
analyses performed in this study. Moment profiles obtained from non-linear time history
analyses indicated a strong dependency on the strength reduction factor utilized but not the
first mode period of the wall. A novel modal decomposition technique is presented in this
study for demonstrating the individual effects of the higher modes on the dynamic shear
amplification phenomenon. The technique is based on inverse extraction of modal
responses from non-linear time history analyses results by using the mode superposition
method. It was shown that non-linear second mode is chiefly responsible for the dynamic
base shear amplifications and the effect of higher modes above the second mode has been
found to be negligible. It is believed that the modal decomposition technique presented in
this study demonstrates the individual contribution of higher mode effects to dynamic
amplification where related studies on this subject are very limited in the literature.
Majority of the findings on dynamic shear amplification phenomenon are based on

analytical studies and limited experimental research performed by Kabeyasawa (1987) and
Eberhard and Sozen (1993) confirms this reality. However, recent full scale tests of walls
performed on E-Defense shaking table in Japan (NIED, 2006) and UC San Diego shaking
table in the US (UCSD, 2005), indicate the dominance of the sliding shear in the plastic
hinge zone on the wall behavior (Krawinkler, 2006), which might be reducing the rate of
amplifications as a shear fuse. It is believed that more information will be available on the
79
real dynamic behavior of yielding structural walls as more full-scale tests and supporting
numerical research simulating the complex reinforced concrete behavior are carried out.
80
APPENDIX A: DESIGN MOMENTS AND STORY SHEAR FORCES

OF THE GENERIC WALLS
Design Moments and Story Shear Forces of the 8 Story Wall
8
R=2
7 R=4
R=6
R=8
6
R=10
5
STORY
0
0 0.5 1 1.5 2 2.5
DESIGN MOMENT (kNm) 4
x 10
Figure A.1. Design moment profile of the 8 story wall
8
R=2
7 R=4
R=6
R=8
6
R=10
5
STORY
0
0 200 400 600 800 1000 1200 1400
DESIGN STORY SHEAR FORCE (kN)
Figure A.2. Design shear force profile of the 8 story wall

81
12
R=2
R=4
10 R=6
R=8
R=10
8
STORY
0
0 0.5 1 1.5 2 2.5 3 3.5 4
x 10
12
R=2
R=4
10 R=6
R=8
R=10
8
STORY
0
0 200 400 600 800 1000 1200 1400 1600 1800

82
16
R=2
14 R=4
R=6
R=8
12
R=10
10
STORY
0
0 1 2 3 4 5 6 7
x 10
16
R=2
14 R=4
R=6
R=8
12
R=10
10
STORY
0
0 500 1000 1500 2000 2500

83
20
R=2
18 R=4
R=6
16
R=8
R=10
14
12
STORY
10
0
0 1 2 3 4 5 6 7 8 9 10
x 10
20
R=2
18 R=4
R=6
16
R=8
R=10
14
12
STORY
10
0
0 500 1000 1500 2000 2500 3000

84
30
R=2
R=4
25 R=6
R=8
R=10
20
STORY
15
10
0
0 0.5 1 1.5 2 2.5
x 10
30
R=2
R=4
25 R=6
R=8
R=10
20
STORY
15
10
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500

85
APPENDIX B: LINEAR AND NON-LINEAR MODE SHAPES OF

GENERIC WALL BUILDINGS
Linear and Non-Linear Modes Shapes for 8 Story Wall Building
1st MODE SHAPE
8
LIN. PHASE
7 NONLIN. PHASE
5
STORY
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
MODAL AMPLITUDE
Figure B.1. Linear and non-linear 1st mode shapes of the 8 story wall building
2nd MODE SHAPE

8
LIN. PHASE
7 NONLIN. PHASE
5
STORY
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
MODAL AMPLITUDE
Figure B.2. Linear and non-linear 2nd mode shapes of the 8 story wall building
86

1st MODE SHAPE
12
LIN. PHASE
NONLIN. PHASE
10
8
STORY
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
MODAL AMPLITUDE
2nd MODE SHAPE

12
LIN. PHASE
NONLIN. PHASE
10
8
STORY
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
MODAL AMPLITUDE
87

1st MODE SHAPE
16
LIN. PHASE
14 NONLIN. PHASE
12
10
STORY
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
MODAL AMPLITUDE
2nd MODE SHAPE

16
LIN. PHASE
14 NONLIN. PHASE
12
10
STORY
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
MODAL AMPLITUDE
88

1st MODE SHAPE
20
LIN. PHASE
18 NONLIN. PHASE
16
14
12
STORY
10
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
MODAL AMPLITUDE
2nd MODE SHAPE

20
LIN. PHASE
18 NONLIN. PHASE
16
14
12
STORY
10
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
MODAL AMPLITUDE
89

1st MODE SHAPE
30
LIN. PHASE
NONLIN. PHASE
25
20
STORY
15
10
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
MODAL AMPLITUDE
2nd MODE SHAPE

30
LIN. PHASE
NONLIN. PHASE
25
20
STORY
15
10
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
MODAL AMPLITUDE
90
APPENDIX C: GROUND MOTION RECORDS FOR TIME HISTORY

ANALYSES
Non-linear time history analysis is considered as the most advanced and most
complicated analysis method with respect to the simplified analysis methods used in
engineering practice. Simplified analysis methods have the advantage of specified ground
motion characteristics mostly in the form of response spectra either in the codified or site
specific forms. This eliminates the effort required by the designer to deal with the
variations arising from the ground motion side of the problem and modeling assumptions
and representations becomes the major concerns to deal with. However, reliable results
expected from a time history analysis are also dependent on appropriately selected strong
ground motion records since the ground motion characteristics will also affect the
response.
Two major issues are confronted with the strong ground motion records to be
incorporated in the non-linear time history analyses. First issue particularly comprises the
composition of a record suite with sufficient number of ground motions records having
comparable characteristics. It is evident that ground motion characteristics are primarily
affected by distance to fault, magnitude of the earthquake and local site conditions. Thus,
selected ground motion records should be representative of these three criteria where the
structure shall be designed. Theoretically, the more the number of records incorporated in
analyses, the more reliable results are obtained and an increase of number of samples allow
for working with average analysis results. It is obvious that finding sufficient number of
locally recorded accelerograms satisfying the three criteria is not an easy task, thus either
accelerograms with similar conditions can be selected or synthetically simulated
accelerograms can be developed.
Second issue relates to the standardization of accelerograms; since the frequency

content and associated spectral accelerations of an accelerogram vary from record to record
naturally which will eventually result in high scatters in terms of analysis results.
Standardization is generally achieved by scaling the records to fit their response spectra to
a target response spectrum, in most cases a code spectrum.
91
Selection criteria and standardization procedure of the selected accelerograms are

explained in the following sections.
Selection of Strong Ground Motion Records
20 recorded accelerograms corresponding to magnitudes between M6.0 and M7.0,

distance to epicenter between 15 km and 50 km and the site class similar to Z3 defined in
Turkish Seismic Design Code (2007) have been selected as given in Table C.1
Table C.1. Selected ground motion records in the study

Distance Site PGA PGV PGD
No Earthquake Magnitude Station (km ) Condition (g) (cm/s) (cm)
1 Chalfant Valley 6.2 54428 Zack Brothers Ranch 18.7 D 0.45 36.90 7.01
2 Chalfant Valley 6.2 54429 Zack Brothers Ranch 18.7 D 0.40 44.50 8.56
3 Loma Prieta 1989 6.9 APEEL 2 - Redwood City 47.9 D 0.22 34.30 6.87
4 Loma Prieta 1989 6.9 1686 Fremont - Emerson Court 43.4 B 0.19 12.70 5.50
5 Mammoth Lakes 1980 6.0 54214 Long Valley dam 19.7 A 0.48 14.20 1.77
6 Mammoth Lakes 1980 5.7 54214 Long Valley dam 14.4 A 0.25 18.50 1.56
7 Mammoth Lakes 1980 6.0 54301 Mammoth Lakes H. S. 14.2 D 0.39 23.90 2.72
8 Morgan Hill 1984 6.2 47380 Gilroy Array #2 15.1 C 0.21 12.60 2.10
9 Morgan Hill 1984 6.2 57382 Gilroy Array #4 12.8 C 0.35 17.40 3.11
10 Northridge 1994 6.7 90074 La Habra - Briarcliff 61.6 C 0.21 12.30 1.23
11 Northridge 1994 6.7 24575 Elizabeth Lake 37.2 C 0.16 7.30 2.70
12 Northridge 1994 6.7 24611 LA - Temple &amp 32.3 B 0.18 20.00 2.74
13 Northridge 1994 6.7 90061 Big Tujunga, Angeles Nat F 24.0 B 0.25 12.70 1.12
14 Northridge 1994 6.7 90021 LA - N Westmoreland 29.0 B 0.40 20.90 2.29
15 Whittier Narrows 1987 6.0 Brea Dam (Downstream) 23.3 D 0.31 14.50 0.77
16 Whittier Narrows 1987 6.0 108 Carbon Canyon Dam 26.8 A 0.22 8.70 0.64
17 Whittier Narrows 1987 6.0 90034 LA - Fletcher Dr 14.4 C 0.21 12.60 1.45
18 Whittier Narrows 1987 6.0 90063 Glendale - Las Palmas 19.0 C 0.30 17.10 1.82
19 Whittier Narrows 1987 6.0 90021 LA - N Westmoreland 16.6 B 0.21 9.70 0.98
20 Whittier Narrows 1987 6.0 24461 Alhambra, Fremont Sch 13.2 B 0.33 22.00 2.42
It is intended that selected ground motion records suitably represent the expected
ground motion characteristics for Seismic Zone 1 defined in Turkish Seismic Design Code
(2007) in terms distance and magnitude. Pseudo-acceleration response spectra of the
selected records are given in Figure C.1 with respect to the design response spectrum
constructed for respective Site Class Z3 and the Seismic Zone 1.
92
16
DESIGN
PSEUDO SPECTRAL ACCELERATION (m/sec2)

14 SELECTED REC.
12
10
0
0 0.5 1 1.5 2 2.5 3
PERIOD (sec)
Figure C.1. Pseudo acceleration response spectra of the selected records
Standardization of Selected Ground Motion Records
Standardization of a suite of records is one of the key points for non-linear time
history analyses. Standardization is principally based on scaling of real accelerograms to fit
a target design ground motion level, which can be defined through design peak ground
acceleration, intensity or the response spectrum itself. Although details of this subject are
out of the scope of the present study, it should be reminded that any manipulation for the
sake of standardization has the risk of losing the natural characteristics of the real records.
Standardization of the selected records in this study has been achieved by scaling the
records to fit the target design response spectrum in all period ranges. An iterative
procedure has been applied by using a program coded in MATLAB with steps given
below:
1. Obtain target pseudo velocity response spectrum, for 5 per cent damping
ratio( S tpv ( f , ξ = 0.05) )
2. Obtain pseudo-velocity response spectrum of the original record, for 5 per cent
damping ratio ( S pv ( f , ξ = 0.05) ).
93
3. Divide the velocity response spectrum of the original record by the target pseudo-
velocity response spectrum to obtain the scale factors ( α ( f ) ) corresponding to
each frequency.
S pv ( f , ξ = 0.05)
α( f ) =
S tpv ( f , ξ = 0.05)
4. Compute Fourier Transform of the original record. FFT (a (t ))

5. Multiply the Fourier Transform of the original record with the scale factors obtained
in Step 3 in the frequency domain, in order to obtain a Scaled Fourier Transform
______
( FFT (a(t )) ).
______
FFT (a(t )) = α ( f ) FFT (a(t ))
6. Compute Inverse Fourier Transform of the Scaled Fourier Transform obtained in

Step 5 to get the Scaled Record.
__ ______
a (t ) = inv[ FFT (a(t ))]
7. Repeat steps between 2 and 6 by replacing the Scaled Record in Step 6 with the
original record in step 2 until a satisfactory convergence is achieved between the
pseudo acceleration response spectrum of the record and the target pseudo
acceleration response spectrum.
In order to illustrate the above scaling procedure, scaling applied to 90021 LA - N

Westmoreland accelerogram given in Table B.1 is presented below.
Figure C.2 shows the pseudo-acceleration response spectrum of the original record
with respect to target acceleration response spectrum before iteration. Figure C.3, Figure
C.4 and Figure C.5 show the spectra at iteration steps 5, 10 and 20, respectively.
94
15
RECORD

TARGET
10
0
0 0.5 1 1.5 2 2.5 3
PERIOD (sec)
Figure C.2 . Comparison of the record spectrum with target spectrum before iteration
15
RECORD
TARGET
10
0
0 0.5 1 1.5 2 2.5 3
PERIOD (sec)
Figure C.3. Comparison of the record spectrum with target spectrum at iteration step 5
95
15
RECORD

TARGET
10
0
0 0.5 1 1.5 2 2.5 3
PERIOD (sec)
15
RECORD
TARGET
10
0
0 0.5 1 1.5 2 2.5 3
PERIOD (sec)
Original and scaled accelerogram of the 90021 LA - N Westmoreland station are

shown in Figure C.6 for comparison.
96
ORIGINAL RECORD
0.4
ACCELERATION (g)
0.2
-0.2
-0.4
0 5 10 15 20 25 30 35 40
time (sec)
SCALED RECORD
0.4
ACCELERATION (g)
0.2
-0.2
-0.4
0 5 10 15 20 25 30 35 40
time (sec)
Figure C.6. Comparison of the original and scaled accelerogram of 90021 LA - N

Westmoreland station
Pseudo acceleration response spectra of the selected records after scaling are given in
Figure C.7 with respect to design acceleration response spectra. Accelerograms scaled by
using the methodology above has been used as the ground motion data in the non-linear
time history analyses of the generic walls as explained in Chapter 5.
97
15
PSEUDO SPECTRAL ACCELERATION (m/sec2) TARGET
SCALED REC.
10
0
0 0.5 1 1.5 2 2.5 3
PERIOD (sec)
Figure C.7. Comparison of the acceleration response spectrum of the records with the
target spectrum after scaling
98
APPENDIX D: ILLUSTRATIVE WALL DESIGN BY USING THE

PROPOSED BASE SHEAR AMPLIFICATION RELATIONSHIP,
SHEAR AND MOMENT PROFILE
Shear Design of the 16 Story Generic Wall Designed With a Strength Reduction
Factor of R=6
The proposed shear design procedure is presented below for a 16 story generic wall
designed with strength reduction factor of R = 6.
Key characteristics of the wall are summarized below;
N= 16 stories
b = 0.30 m (section width in flexure)
lw = 7.50 m (section depth in flexure)
h = 3.00 m (Typical story height)
Hw = 48.00 m (Total wall height measured from the base)
T1-cr = 1.30 sec (First mode period of the wall based on cracked section stiffness
properties)
Design shear forces have been obtained in Chapter 3 by using the response spectrum
method as shown in Figure D.1 below;
16
14
12
10
STORY
0
0 100 200 300 400 500 600 700 800
Figure D.1. Design Shear Force Diagram of the 16 Story Wall (R = 6)

99
Thus the design base shear obtained is;
Veb = 741 kN
The base shear amplification factor is calculated by using the proposed relation given
in Equation 6.5 by assuming flexural overstrength of 1.0
M eb
= 1.0
M rb
β b = 1.0 + (0.281× 1.30 + 0.394) × (1.0 × 6 − 1.5)0.553

β b = 2.744
The amplified design base shear is obtained from Equation 6.7 will give;
Vab = Veb × β b = 741× 2.744 = 2033.3 kN
The story shear force diagram is obtained by using the suggested profile given
Section 6.2
• Determine the height of the critical wall height of constant shear right above the base
of wall
max(lw , H w / 6) = 8.0m → App. 3 stories above the base
• Determine the level where the story shear forces drop to 50 per cent of the base shear
0.40 × 48.00 = 19.2m → App. 7 stories above the base
Amplified story design shear force diagram obtained as above is given in Figure D.2
where it is compared with the quantities obtained from design and Group 3 non-linear time
history analyses.
100
16
AMPLIFIED
14 DESIGN
T. HISTORY
12
10
STORY
0
0 500 1000 1500 2000 2500
Figure D.2. Amplified design shear forces for the 16 story wall (R = 6)
101
Proposed Moment Diagram of the 16 Story Generic Wall Designed With a Strength
Reduction Factor of R=6
Design moment at the base of the 16 story generic wall has been obtained in Chapter
3 by using the response spectrum method.
M eb = 14677.34 kNm
Intermediate points on the proposed moment profile are obtained as per Section 6.4
by considering the tension shift effects at the base of the wall.
λ = 1.05 − 0.025 × 6 (Equation 6.12)

λ = 0.90
λ M eb = 13209.61 kNm
7.50
δ = 0.003 × (6 − 2) 2 + 0.025 × (6 − 2) + max( , 0.167) (Equation 6.15)
48.0
δ = 0.315
δ H w = 15.12 m (Just above 5th story level)
7.50
η = 0.125 × (6 − 2) − 0.005 × (6 − 2) 2 + max( , 0.167) (Equation 6.16)
48.0
η = 0.587
η H w = 28.18 m (Just below 10th story level)
Proposed moment diagram with intermediate points obtained as above is given in

Figure D.3 where it is compared with the quantities obtained from design and Group 3
non-linear time history analyses.
102
16
PROPOSED PROFILE
14 NL. TIME HIST.
DESIGN
12
10
STORY
0
0 5000 10000 15000
MOMENT
Figure D.3. Proposed moment diagram for the 16 story wall (R = 6)

103
REFERENCES
Aydınoğlu, M.N., 2003, “An Incremental Response Spectrum Analysis Procedure Based
on Inelastic Spectral Displacements for Multi-Mode Performance Evaluation”,
Bulletin of Earthquake Engineering, Vol. 1, pp. 3-36.
Blakeley, G., C. Cooney and M. Megget, 1975, “Seismic Shear Loading at Flexural
Capacity in Cantilever Wall Structures”, Bulletin of the New Zealand National
Society for Earthquake Engineering, Vol. 8, No.4, pp. 278-290.
Bathe, K-J., 1996, Finite Element Procedures, Prentice Hall, Upper Saddle River, NJ,
USA.
Chopra, A.K., 2007, Dynamics of Structures, Prentice Hall, Upper Saddle River, NJ, USA.
Carr, A.J., 2000, “RUAUMOKO, Program for Inelastic Dynamic Analysis”, Department
of Civil Engineering, University of Canterbury, Christchurch, New Zealand.
CEN, 2004, Eurocode EC8, “Eurocode (EC) 8 Design of Structures for Earthquake
Resistance-Part I General Rules, Seismic Actions and Rules For Buildings (EN
1998-1)”, Comité Européen de Normalisation, Brussels.
CSI, 2006, “SAP 2000, Static and Dynamic Finite Element of Structures”, Computers and
Structures Inc., Berkeley, California, USA.
Derecho, A. and W. Corley, 1984, “Design Requirements for Structural Walls in

Multistory Buildings”, Proceedings of Eighth World Conference on Earthquake
Engineering, San Francisco, California, July 1984, Vol. 5, pp. 541-548.
Eberhard, M. and M. Sözen, 1993, “Behavior Based Method to Determine Design Shear in
Earthquake-Resistant Walls”, Journal of the Structural Division, American Society of
Civil Engineers (ASCE), Vol. 119(2), pp. 619-640.
104
Eibl, J. and E. Keintzel, 1988, “Seismic Shear Forces in RC Cantilever Shear Walls”,
Proceedings of Ninth World Conference on Earthquake Engineering, Kyoto, Japan, 2
August-9 August 1991,Vol. VI, pp. VI/5-VI/10.
Filiatrault, A., D. D’Aronco and R. Tinawi, 1994, “Seismic Shear of Ductile Cantilever
Walls: A Canadian Code Perspective”, Canadian Journal of Civil Engineering, Vol.
21, pp. 363-376.
Gosh, S. and P. Markevicius, 1990, “Design of Earthquake Resistant Shearwalls to Prevent

Shear Walls”, Proceedings of Fourth U.S. National Conference on Earthquake
Engineering, Palm Springs, California, 20 May-24 May 1990, Vol. 2, pp. 905-913.
Kabeyasawa, T., 1987, “Ultimate-state Design of Reinforced Concrete Wall-Frame

Structures”, Proceedings of Pacific Conference on Earthquake Engineering,
Auckland, New Zealand, Vol. 4, pp. 431-440.
Krawinkler, H., 2006, “Importance of Good Nonlinear Analysis”, The Structural Design of
Tall and Special Buildings, Vol. 15, pp. 515-531.
NIED, 2006, “Test of a Six Story R/C Building”, Hyogo Earthquake Research Center, E-
Defense Facility, National Research Institute for Earth Science and Disaster
Prevention, Japan, www.bosai.go.jp.
Paulay, T. and M.J.N. Priestley, 1992, “Seismic Design of Reinforced Concrete and
Masonry Buildings”, John Wiley & Sons, New York, USA.
Rutenberg, A. and E. Nsieri, 2006, “The Seismic Shear of Ductile Cantilever Wall Systems
and the EC8 Provisions”, Bulletin of Earthquake Engineering, Vol. 4, pp.1-21.
SEAOC, 1999, “Recommended Lateral Force Requirements and Commentary”, Structural

Engineers Association of California, Sacramento, CA, USA.
105
Senaviratna, K. and H. Krawinkler, 1994, “Strength and Displacement Demands for

Seismic Design of Structural Walls”, Proceedings of Fifth U.S. National Conference
on Earthquake Engineering, Chicago, Illinois, 10 July-14 July 1994, Vol. II, pp.181-
190.
Somerville, P., N. Smith, S. Puntamurthula and J. Sun, 1997, “Development of Ground

Motion Time Histories for Phase-2 of the FEMA/SAC Steel Project”, SAC
Background Document, SAC/BD-97/04, SAC Joint Venture, Richmond, CA, USA.
Standards New Zealand, 1982, “NZS 4203 Code of Practice for Structural Design and
Design Loadings for Buildings”, Wellington, New Zealand.
Standards New Zealand, 2006, “Concrete Structures Standard: Part I-The Design of
Concrete Structures, Part II-Commentary on the Design of Concrete Structures”,
Wellington, New Zealand.
Takeda, T., M. Sözen and N. Nielsen, 1970, “Reinforced Concrete Response to Simulated
Earthquakes”, Journal of the Structural Division, American Society of Civil
Engineers (ASCE), Vol. 96(12), pp 101-113.
The MathWorks Inc., 2004, MATLAB, The Language of Technical Computing, USA.
Turkish Seismic Design Code, 2007, “Specification for Buildings to be Built on Seismic
Areas”, in Turkish, Ministry of Public Works and Settlement, Ankara, Turkey.
Turkish Seismic Design Code, 1997, “Specification for Structures to be Built on Disaster
Areas”, in English and Turkish, Ministry of Public Works and Settlement, Ankara,
Turkey.
UCSD, 2005, “Shake Table Test of a Full-Scale Seven Story Structural Wall”, Englekirk
Center, Powell Structural Research Laboratories, University of California, San
Diego, USA, www.jacobsschool.ucsd.edu/Englekirk.
106
Wilson, E., 2006, “The Eigenvalue Problem”, Technical Papers, www.csiberkeley.com.

107
REFERENCES NOT CITED
Clough , R.W. and J. Penzien, 1993, Dynamics of Structures, 2nd International Edition,
McGraw Hill Inc., Singapore.
Kappos, A.J. and P. Antoniadis, 2007, “A Contribution to Seismic Shear Design of R/C
Walls in Dual Structures”, Bulletin of Earthquake Engineering, Vol. 5, pp. 443-466.
Linde, P., 1998, “Evaluation of Structural Walls Designed According to Eurocode 8 and
SIA 160”, Earthquake Engineering and Structural Dynamics, Vol. 27, pp. 793-809.
Priestley, M.J.N., 2003, “Myths and Fallacies in Earthquake Engineering, Revisited”, The
Mallet Milne Lecture, IUSS Press, Pavia, ITALY.
Riva, P., A. Meda and E. Giuriani, 2003, “Cyclic Behavior of a Full Scale RC Structural
Wall”, Engineering Structures, Vol. 25, pp. 835-845.
Senaviratna, K. and H. Krawinkler, 1997, “Evaluation of Inelastic MDOF Effects for

Seismic Design”, The John A. Blume Earthquake Engineering Center, Report No.
120, June 1997, Department of Civil and Environmental Engineering, Stanford
University, CA, USA.

b1461930.006592.001

Uploaded by

Copyright:

Available Formats

b1461930.006592.001

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

b1461930.006592.001

Uploaded by

Copyright:

Available Formats

DYNAMIC SHEAR AMPLIFICATION IN SEISMIC RESPONSE OF

STRUCTURAL WALL SYSTEMS

Submitted to the Kandilli Observatory and

Graduate Program in Earthquake Engineering

To my precious wife Beyza CELEP

DYNAMIC SHEAR AMPLIFICATION IN SEISMIC RESPONSE OF

PERDE SİSTEMLERDE DEPREM NEDENİYLE OLUŞAN DİNAMİK

6.3. Implications on Shear Design of Structural Walls...................................................68

Table 3.1. Generic wall characteristics .........................................................................13

R Strength reduction factor

ω1 Natural vibration frequency of the wall

EI g Bending stiffness of the wall corresponding to gross section

βb Dynamic base shear amplification factor

lw Dimension of the wall perpendicular to major bending axis

Ψo Flexural overstrength factor

Vab Amplified design base shear

M eb Moment demand at the base of the wall obtained from code

M Eb Elastic moment demand at the base of the wall obtained from

f yk Characteristic yield strength of reinforcement.

1.1. Motivation of the Study

Dynamic shear force amplification is generally taken into account in design by

1.2. Scope of Work

Previous studies on dynamic shear amplification have been covered in Chapter 2,

amplifications has been demonstrated in Chapter 4 by using a modal decomposition

Dynamic shear amplification of structural wall systems designed to respond non-

Dynamic base shear amplification effect is generally presented by researchers in the

( Vab )as given in Equation (2.1).

Vab = β bVeb (2.1)

Kabeyasawa (1987) proposed base shear amplification factors, as given in Equation

Vmax = Vn + c × W × PGA (2.3)

Experimental investigations on nine small-scale reinforced concrete structural

ε = 0.75 + 0.22 × (T + q + Tq) (2.6)

3. DESIGN OF THE GENERIC STRUCTURAL WALLS

Figure 3.1. Illustration of generic walls considered in this study

Firstly generic walls have been pre-designed to provide section dimensions.

3.1. Pre-Design of Generic Structural Walls

Thus the first mode period of the system will be;

Substituting Equation (3.1) in Equation (3.3) will result in Equation (3.4) to be

Table 3.1. Generic wall characteristics

WALL Hw (m) T1 (sec) b(m) d(m) mtypical (t)

3.2. Design of Generic Structural Walls

Subsequent to the determination of characteristics of generic structural walls as

Figure 3.2. Design acceleration response spectrum used in the study

Figure 3.3. Modelling representation of structural walls

4. A NOVEL MODAL DECOMPOSITION TECHNIQUE FOR

A modal decomposition technique is presented in this study for demonstrating the

4.1. Linear and Non-Linear Modes of Vibration

All linear and non-linear eigenvalues and corresponding eigenvectors having

1st MODE SHAPE

2nd MODE SHAPE

3rd MODE SHAPE

where the superscript asterisk indicate the non-linear mode.

Φ1*T MΦ*2 = 0 (4.3)

4.2. Modal Decomposition Technique

4.2.1. Modal Expressions in The Linear Response

Static-equivalent m’th mode seismic force at time t can be expressed as:

where am (t ) refers to m’th mode pseudo-acceleration response history. Φ m represents

in which 1 refers to the vector of ones.

Define scaled mode shape vectors as:

Φ1T MΦ2 = 0 (4.3)