'I

,
UILU.;.ENG-84-2007
CIVIL ENGINEERING STUDIES
STRUCTURAL RESEARCH SERIES N.O. 516
"'::
"'-"
SEtS E
ISSN: 0069-4274
N L SIS N
N
F R.
ITI
BUlL
By
Y. J. PARK
A. H-S. ANG
and
Y. K. WEN
Technical Report of Research
Supported by the
BSTerence
.Un1vers1t.y o:f
BI06 NCEL
208 N. Romine Street
Urbana, Illinois
NATIONAL SCIENCE FOUNDATION
Division of Civil & Environmental Engineering
(Under Grants CEE 80-02584 and CEE 82-13729)
UNIVERSITY OF ILLINOIS
at URBANA-CHAMPAIGN
URBANA, ILLINOIS
OCTOBER 1984
50272 -101
i
REPORT DOCUMENTATION '1. I4
R
f.POflf NO.
PAGE UILU-ENG-84-2007
4. Title and Subtitle
SEISMIC DAMAGE ANALYSIS AND DAMAGE-LIMITING DESIGN OF R. C.
BUILDINGS
IlL Date
OCTOBER 1984
1--------------- -----------..--------------
7. Author(s)
Y. J. Park, A. H-S. Ang, K. Wen
9. Perform Ins O'llanlzlllltion Name and AddrHlIl
Department of Civil Engineering
University of Illinois
208 N. Romine Street
Urbana, IL 61801
12. Orwan1mlon Hllmll and Add,,"_
National Science Foundation
Washington, D.C.
15. Supplemllntary NotH
L PGtrforml.,. Or"llllnlxlltlon Rept. No.
SRS No. 516
10. Projed/Tnk/Woril Unit No.
U. ContntCt(C) or Gnant(G, No.
(C)
NSF CEE 80-02584
W>NSF CEE 82-13729

14.
.. -------------------------------i
16. Abllltnact (limit: 200 words)
A model for evaluating structural damage in reinforced concrete structures under
earthquake ground motions is proposed. Damage is expressed as a linear combination
of the maximum deformation and the effect of repeated cyclic loadings. Available static
(monotonic) and dynamic (cyclic) test data were analyzed to evaluate the statistics of
the appropriate parameters of the proposed model. Accordingly, a method of nonlinear
random vibration technique was developed for obtaining the response statistics necessary
for damage assessment.
From an extensive damage analysis of various reinforced concrete buildings, an
intensity scale was derived to describe the potential destructiveness of ground motion.
It is a simple function of the rms acceleration and strong-motion duration. The pro-
posed damage model is calibrated with observed building damages from past earthquakes.
,. On this basis, a design procedure is developed accordingly to limit the potential
structural damage to a tolerable level.

Destructiveness of Earthquake
Earthquake-Resistant Design
Random Vibration
b. Identlfierlll/()penEnded Terms
c. COSATI Field/Group
18. Availability Statement
(SIN ANSI-Z39.18)
Reinforced Concrete
Seismic Damage
Structural Reliability
21. No. of Pales !119. Security Ciauis (This Report)
__ ________
; 20. Security ChllllUI (This PallID)
! IJNCl.ASS I FI ED
SIN Instructions on RfJlVfJlrfut
22. Price
OPTIONAL FORM 212 (4-77)
(Formerly NTIS-35)
Department of Commerce
ii
ACKNOWLEDGMENTS
This report is based on the thesis of Dr. Young-Ji Park submitted in
partial fulfillment of the requirements for the Ph.D. in Civil
Engineering at the University .of Illinois at Urbana-Champaign.
The study is part of a program on the safety and reliability of
structures to seismic hazard, supported by the National Science
Foundation under grants CEE 80-02584 and CEE 82-13729. Supports for
this study are gratefully acknowledged. The authors wish to express
their appreciation to the' Architectural Institute of Japan for
permission to use the damage data and photos of the 1978 Miyagiken-Oki
earthquake.
CHAPTER
1
2
3
iii
TABLE OF CONTENTS
Page
INTRODUCTION 1
1.1 Object and Scope of Study 1
1 .. 2 Related Previous Studies ................ " .................. " .......... 2
1.2.1
1.2.2
Inelastic Behavior of Reinforced Concrete 2
Damage Assessment .............................. co ............... 4
1.3 Organization 6
1.4 Notation ... fi) CD .. ., .......... ., " ., .; ., ., ". 7
DAMAGE MODEL FOR REINFORCED CONCRETE e
2,,1
2 .. 2
2 .. 3
2 .. 4
General Remarks
P r o ~ o s e d Damage Model
Determination of Model Parameters
2.3.1 Determination of 8
., ., ., " (I ., Cit
u
2 .. 3.2 Determination of
S
o " ., 1/1 ...... CIt " .,
2.3.3 Determination of Q
and
Q
" ., e ., ., .,
E
Y
(8)
u
2.3 .. 4 Determination of CD ., It II
c
Damage Index Statistics e ., " G 0 0 e CIt e It e . e ........ ,.
DAMAGE ANALYSIS OF REINFORCED CONCRETE BUILDINGS
3 .. 1
3 .. 2
3.3
General Remarks
Methodology
3 .. 2 .. 1
3.2 .. 2
3.2 .. 3
3 .. 2.4
3 .. 2.5
Damage Index Statistics
Force-Deformation Relationship
Response of Hysteretic Systems
Maximum Response Statistics
Absorbed Hysteretic Energy
Damage Analysis of SDF Systems
10
10
11
12
12
21
22
22
23
25
25
26
26
27
28
31
33
34
4
3.4
3 .. 5
iv
Page
Modeling of Buildings ............................. e CD ....... "
3.4.1 General ............................................ " . ., ..... .
35
35
3.4 .. 2
3.4.3
A Hybrid Model .............................. " 36
Force-Deformation Relation and Damage Index
of Inelastic Springs ........................... .
3.4.4 Overall Damage Assessment
Examination of Building Types
39
40
40
CALIBRATION OF DAMAGE INDEX " .......................... . 44
44
44
45
45
46
46
46
47
47
47
47
4 .. 1
4 .. 2
4.3
4 .. 4
General Remarks ............. It ..... ., .......... G ., ...... ., .. 0
Description of Building Damage .......................... .
4 .. 2.1 Building A (Olive View Hospital) ............... .
4.2 .. 2
4.2 .. 3
4.2 .. 4
4 .. 2.5
4.2.6
4 .. 2.7
4.2 .. 8
4 .. 2 .. 9
Building B (Taiyo Gyogyo Building)
Building C (Tohoku Kogyo University)
Building D (Saigo School) ...................... .
Building E (Tonan High School)
Building F (Kinoshita Menko Building)
Building G (Obisan Office
Building H (Fukushi Kaikan Building)
Building I (Izumi High School) ..................
Estimation of Ground Excitation 48
Results of Damage Analyses 49
4 e4.1 Building A e. G .0. I) It G ......... CIt ............. ., .... lit. 49
4.4.2
4.4.3
4.4.4
4.4 .. 5
4.4.6
4.4.7
Building B
Building C
Building D
Building E
Building F
Building G
49
50
50
51
51
51
5
4.5
4 .. 4.8
4.4.9
Building H
Building I
v
0' .............. .; .................. .
CIt It CIt " ., I/) .,
Def ini tion of Damage
Page
52
52
52
DAMAGE-LIMITING ASEISMIC DESIGN ....................... . 53
53
54
54
5.1
5 .. 2
5.3
5.4
5.5
5.6
5.7
Objectives of Aseismic Design ., 0 ., ............... .
Strength Requirements . ., ....................... ., .. " . ., .. .
General 5.2.1
5.2.2
5.2 .. 3
Optimal Strength Distribution and Mode Shape 55
Application to Weak-Beam-Type Buildings ......... 58
Equivalent SDF Concept It " ..... It
5,,3.1 Modeling of Buildings
5.3.2 Equivalent Short Duration Excitation
5.3.3 Response of SDF Systems
Ductility Requirements
CD " " ....... .
5.4.1
5 .. 4.2
Ductility Requirement of Components
Ductility Requirement of Building
Design Procedure
5.5 .. 1
5.5 .. 2
5.5.3
5.5.4
Design Setup ...................... .
Strength Design
Structural Analysis
Ductility Check ......... .
Design Examples ........... "." ......... " ........... " ........ .
5.6.1
5.6.2
5.6.3
5.6.4
5.6.5
Design of Frame A
Design of Frame B
" " CD
ell" e
Design of Frame A' ,. 4\) " 0 CD ....
Design of Frame B' 00@& 00o
Examination of Designed Frames
Limitation
60
60
61
62
63
63
65
66
66
67
67
67
68
69
71
71
73
73
74
6
7
vi
Page
RELIABILITY OF PROPOSED ASEISMIC DESIGN .. It
6.1
6 .. 2
6.3
Introductory Remarks ................... " .... " .... e ....... " ........ " ...... ..
Uncertainty in Structural Response ...................... .
6 .. 2.1
6.2.2
Stiffness and Strength ......................................... .
Damping .. 0 " " lit ... " G' ., ..... 0
6.2.3 Predominant Frequency of Excitation w
. G
6.2.4 Hysteretic Energy per Cycle .................... .
........................................... 6.2 .. 5 Mass
Reliability of Design .. " ., CIt 0 ., 0 ., lit 4)
SUmiARY AND CONCLUSIONS ............................................... " ...... ., .... " .. "
75
75
75
77
77
78
78
78
79
82
TABLES .... ' .......................................................... . 84
FIGURES ee CIII 98
APPENDIX
A YIELD CURVATURE AND YIELD MOMENT ................................ 143
B DERIVATION OF EQUATION 3.17 149
C DETAILS OF EQUATION OF MOTION . 0 00.0 150
D THE EQUIVALENT SDF SYSTEM .. 154
vii
Page
LIST OF REFERENCES ., ............. ., lit .. lit ., " " ., lit ................. (I ., Cit 157
Table
3 .. 1
4.1
4 .. 2
4 .. 3
4.4
4.5
4.6
4 .. 7
4.8
4 .. 9
4 .. 10
5 .. 1
5 .. 2
5 .. 3
5.4
5 .. 5
5.6
5 .. 7
5 .. 8
5 .. 9
5 .. 10
5 .. 11
6 .. 1
viii
LIST OF tABLES
Page
Excitation Level and Damage Index
Data and Results of Damage Analyses
0 fill
Results of Building-A
Results of Building-B
Results of Building-C
Results of Building-D
Results of Building-E
Results of Building-F
Results of Building-G
Results of Building-H
Results of Building-I
....... ., ...... ..
" " Cit ........... 0 ....... ., . ....... .,
0 ., 0 .,
., 4) CD .......... .
CD " 8: " \9 ., ., /I
' ., " lit 0 .............. .
., ., " 0 " It ., ., .,
., ............ ., II " tit ., " " .... .
... ., " .......... III " .,
Lateral Forces and Interstory Shears of Frame A ........... ..
Strength Design of Frame A ................................. .
Ductility Design of Frame A ................. .
Lateral Forces and Interstory Shears of Frame B
Strength Design of Frame B
Ductility Design of Frame B
Strength Design of Frame A'
Ductility Design of Frame A'
Strength Design of Frame B'
Ductility Design of Frame B'
... ., ., " II ........... ., ,.
., ., ., fit ., fit CD
I/) GIl ., It " It " CD CD 0 " II " " 0
" " ............... ., ., 0 .. " .,
Overall Damage Index
Model Parameter Uncertainties ..........................
85
86
87
87
88
88
89
89
89
90
90
91
91
92
92
93
94
94
95
95
96
96
97
Figure
2 .. 1
2 .. 2
2 .. 3
2 .. 4
2.5
2.6
2.7
2.8
2 .. 9
2 .. 10
2 .. 11
ix
LIST OF FiGURES
Page
Three Components of Inelastic Deformation at Yielding ....... 99
Cross-Section of a Member .............................................. 99
Stress-Strain Relationship ......................................... 100
Bond Slippage at Anchorage .............................................. 100
Pullout Test by Ferguson
Pullout Tests by Several Researchers .............................. ..
101
102
Effect of Shear Span Ratio on Shear Deformation ............ 103
Bond Failure Mechanism ...................................... 104
Shear Cracking Model ............................................... 104
Effect of Bond Stress .on Shear Deformation ........................ 105
Yield Deformation ........................................... 00 .................... 105
2.12 Ultimate Failure Mechanism 106
2.13 Ultimate Ductility. Factor Failure Mode .. " .................... " 1 07
2 .. 14
2 .. 15
2 .. 16
2 .. 17
2.1B
Ultimate Ductility Factor Confinement Ratio ............. .
Prediction Error in Ultimate Ductility Factor .............. .
Prediction Error in B for Eq. 2.1 .......................... .
Prediction Error in B for Eq. 2 .. 2
Experimental Data of Hysteretic Energy (at Small
lOB
109
110
110
Deformation) ...................................... " .............. III
2 .. 19 ,Experimental Data of Hysteretic Energy (at Large
Deformation) ........... " .......................................................... 112
2.20 Idealized Normalized Hysteretic Energy ........................... 113
2 .. 21
2.22
Regression Analysis for f and f
1 2
113
Damage Index of Eq. 2 .. 1 Plotted on Lognormal
Probability Paper ................................................. 114
2 .. 23 Damage Index of Eq ... 2 .. 2 Plotted on Lognormal
Probability Paper 114
3.1 Force-Deformation Relationship of Members 115
3.2 Degrading Hysteretic Model .................................... 115
Hexerenoe Hoo.
University of Illinois
BI06 NCEL
208 N. Romine street
Urbana, Illinois

3.3
3.4
3.5
3.6
3 .. 7
3.8
3 .. 9
3,,10
3 .. 11
3.12
3.13
3 .. 14
3 .. 15
3.16
3.17
3.18
3.19
4.1
4 .. 2
4.3
4.4
4.5
5.1
5.2
\
5 .. 3
5 .. 4
5 .. 5
5.6
6 .. 1
6.2
x
Page
Apparent Yield Deformation 116
Ultimate Strength ...................................................... 116
Force-Deformation Relation and Damage Index of Columns 117
Comparison to Monte Carlo Simulations II II G
Contour Lines for Mean
Probability of D ~ 0.5
Damage Index .............. ' ............ .
118
119
119
Characteristic Intensity vs. Damage Index 120
Modeling of Buildings ........................................ 121
Moment and Rotation at Hinges ................................. 121
Detail of FraIIles ................................................ 122
Static Analysis Result " " ., Gil
Response and Damage of Building-SA " I) fit " " ...... 8
Response and Damage of Building-SB ., lilt " ., II " ., ....
Response and Damage of Building-SC e
Characteristic Intensity vs. Maximum Damage Index
Characteristic Intensity vs. Overall Damage Index
C. O. V. of Damage Index """"" .. "" ................................ " ...... ..
Plan and Elevation of Damaged Buildings ............. "
Photographs of Damaged Buildings ........................... .
Characteristic Intensity vs. Peak Acceleration "
Attenuation Character of Characteristic Intensity .......... .
Calculated Damage Index vs. Observed Seismic Damage
Typical Force-Deformation Relationship ........................ ..
Response of Equivalent SDF Systems ........................ .
Allowable Rotational Angle of Components .................. .
Des ign Procedure, .............................. " ........................... ..
Dimension of FraIIls " .... " ............................ " ........... " .. G " ............ ..
Results of Damage Analysis ......................... " ...... "" ......... " ........ ..
Modeling Error in Stiffness and Strength ....................... .
Damping Coefficient of R.C. Buildings .................... " ....... .
123
124
124
125
126
126
127
128
129
132
132
133
134
134
135
136
137
138
139
140
......
6.3
6.4
A.l
A.2
A.3
xi
Page
Variability in Response Due to Modeling Error 141
Reliability of Design Procedure ...
Coefficient of C vs. p
Coefficient of C VB. N
o
Coefficient of C for Columns with N = 0.3 ...
o
142
148
148
148
1
CHAPTER 1
INTRODUCTION
1.1 Object and Scope of Study
Experience from past earthquakes, such as the 1968 Tokachi-Oki
earthquake in Japan and the 1971 San Fernando earthquake in California,
have shown the vulnerability of reinforced concrete buildings to strong
ground shakingso For economic reasons, however, some level of damage
should be expected and permitted in the aseismic design of structures,
particularly for low-rise and buildings when subjected to
severe earthquake forces. In spite of this recognition, the potential
seismic damage of structures and the associated aseismic provisions are
based largely on linear theory and engineering judgment. Clearly, in
order to assure seismic safety and limit the damage' of reinforced
concrete buildings, the failure mechanism of structural systems under
dynamic earthquake loadings needs to be delineated and incorporated in
the development of aseismic design provisions. This may require, in
particular, the identification of the ultimate structural capacity of
reinforced concrete components under inelastic loading reversals.
Under earthquake loadings, reinforced concrete structures are
generally damaged by a combination of repeated stress reversals and high
stress excursions. Accordingly,a method for quantitative assessment of
building damage as a function of both the maximum response and the
effect of cyclic loadings should be developed.
The development of a method for damage analysis is of prime
importance for rational quantification of seismic damage; related design
criteria must also be developed and verified using actual seismic damage
data. A continuing effort is being made by engineering researchers and
practitioners in accumulating such data from major earthquakes, e.g.,
the 1968 Tokachi-Oki earthquake in Japan, the 1971 San Fernando
earthquake in California, the 1978 Miyagiken-Oki earthquake in Japan,
2
and the 1979 Imperial Va11y earthquake in California. Although reliable
damage data are limited, tentative conclusions can be developed
regarding tolerable design damage levels based on the analysis of such
available data.
The objectives of this research are, therefore; (i) to develop a
method for damage assessment, and (ii) to formulate a damage-limiting
procedure for earthquake-resistant design.
Past seismic damage of reinforced concrete structures has shown that
building collapse is caused mainly by damage concentration (especially
in the first story) and/or the lack of ductility in the constituent
members. In the 1971 San Fernando earthquake, the main unit of the
Olive View Hospital was severely damaged due to damage concentration 1n
the lower two stories; whereas, several weak-bearn-type buildings, such
as the Orion Avenue Holiday Inn, sustained only minor damage because of
a more uniform distribution of damage over the building (U.S. Department
of Commerce, 1973). This would imply that the seismic performance of
buildings can be improved without greatly increasing construction cost,
if uniform potential damage distribution can be assured in the design
stage.
A design method is
properly selecting both
members. The reliability
collapse is also examined.
developed for achieving uniform damage by
the strength and ductility of the structural
of the proposed aseismic design against
1.2 Related Previous Studies
1.2e1 Inelastic Behflvior of Reinforced Concrete
Since the early 1950's, there has been considerable research on the
inelastic behavior of reinforced concrete structures. The early
research concentrated primarily on the flexural behavior of components
under monotonic loadings (e.g., Gaston, Siess and Newmark, 1952;
3
McCollister, Siess and Newmark, 1954; Burns and Siess, 1962). Research
on the inelastic behavior of reinforced concrete components under
earthquake-like loading reversals was initiated in the 1960's (e.g.,
Naka, 1963; Aoyama, et al., 1966; Umemura and Aoyama, 1969; Ikeda,
1970). Under inelastic loading reversals, the ductility capacity of
components is generally decreased, mainly due to the increase of shear
deformation and bond deterioration (Naka, 1963; Ikeda, 1970).
By testing twelve reinforced concrete beams under various loading
reversals, Brown and Jirsa (1971) showed the close relationship between
the shear deformation and the amount of stirrups. Celebi and Penzien
(1973) conducted dynamic destructive tests on relatively ductile beams;
the results indicated a significant effect of the shear span ratio and
the amount of longitudinal reinforcement on the shear deformation under
repeated cyclic loadings.
Wight and Sozen (1973) conducted similar tests by changing both the
axial load and the transverse reinforcement. Their results showed a
significant decay of both the ~ t r e n g t h and stiffness in specimens
subjected to high axial load\ and/or containing poor transverse
reinforcement.
KHstH and Bouwkamp (1975) conducted cyclic loading tests on eight
beam-column subassemblages. Extremely brittle shear failure was
observed in all of the columns although every column was designed to
fail in flexure under monotonic loadings--revealing the inadequacy of
design provisions based entirely on monotonic test results.
A comprehensive research program was conducted by Japanese
researchers on about three hundred half-scale reinforced concrete
columns under reversed loadings (Short Column Committee, 1973-77). The
main factors considered included the loading history, axial load, shear
span ratio, stirrup ratio, axial steel ratio, and concrete strength.
The results revealed the complex failure mechanism of reinforced
concrete components under reversed loadings. In short, a member
designed to fail in flexure tends to fail through shear under repeated
loading reversals.
significantly, the
When the number of loading reversals is increased
member tends to fail through bond at a much lower
4
deformation than would be expected under a monotonic loading. Analyses
of these test results showed a relatively small error in predicting
strength; however, a significant scatter was observed in predicting the
ultimate deformation capacity and energy absorbing capacity.
1.2.2 Damage Assessment
Whitman, et al. (1972, 1973, 1974) developed a method for seismic
damage assessment of buildings. The severity of ground motions is
represented by the MMI scale, and seismic damage is expressed by the
ratio of the cost of repair to the replacement cost of a building
(damage ratio).
Blume, et al. (URS/Blume, 1975) proposed the spectral matrix method
for potential damage assessment of a building or a group of buildings.
Gound motion characteristics is represented by the velocity response
spectrum, and the structural capacity is expressed by the base shear at
yielding. The spectral velocity corresponding to the base shear is then
calculated. The overall damage is expressed by the ratio of cost of
repair to the total replacement cost, which is crudely related to the
ductility factor.
Similar attempts have also been made by Culver, et al. (1980). A
sophisticated computer program is used for obtaining the detailed
response under specified ground excitations. However in assessing the
safety of buildings, the relevant damage criteria are determined based
on engineering judgments.
Bertero and Bresler (1977) attempted to give a more complete
definition of damage, by defining local damageability, global
damageability, and cumulative damageability. Local damageability is a
measure of damage of the constituent components, expressed as a ratio of
the maximum response to the ultimate deformation capacity. Global
damageability is a measure of building damage defined as the sum of the
local damages, weighted by an appropriate importance factor. Cumulative
damageability is a measure of the overall damage as the result of
5
previously sustained damage.. Based on these definitions, Biejaws and
Bresler (1979) proposed a method of damageability evaluation using a
quasi-static structural analysis method. In applying this method to
actual buildings, two critical quantities, namely, the ultimate
deformation capacity of components and the appropriate importance
factors, should be specified. The use of relevant experimental data for
the former and the appropriate engineering judgment for the latter were
suggested ..
As suggested in many of the foregoing studies, .structural damage may
be defined as a ratio of "demand," i. e., the response under earthquakes,
to the ultimate structural Numerous studies have been made
for obtaining the "demand" using dynamic response analysis; an extensive
literature survey is available in Umemrua and Takizawa (1982). On the
other hand, the determination of the "capacity" is more limited in spite
of its critical importance in damage assessment.
Gosain, et ale (1977) proposed the "work index" as a measure of
energy absorbing capacity of reinforced concrete components subjected to
cyclic loadings. The incremental damage in each cycle is expressed as a
function of the ductility ratio, peak load, axial load, and the shear
span ratio. Recently, Banon, et ale (1980) proposed a more
sophisticated damage model in which damage is represented by a
two-dimensional failure surface of the total absorbed energy and the
damage ratio defined in Lybas and Sozen (1977). Because the model did
not incorporate the complicated behavior of reinforced concrete
components under inelastic loading reversals, such as shear and bond
failures, and also due to the relatively small sample size (e.g., only
29 test specimens), a poor correlation was observed between the
calculated structural capacity and the observed failures of the test
specimens.
6
1.3 Organization
In Chapter 2, a seismic damage model is developed for reinforced
concrete components. Structural damage is expressed as a linear
combination of the maximum deformation and the absorbed hysteretic
energy. A large set of U.S. and Japanese test data of reinforced
concrete beams and columns tested to' failure were analyzed to define the
ultimate structural capacity and variabilities. A hysteretic restoring
force model of reinforced concrete components is also developed based on
experimental data.
Based on the foregoing damage model, a method for stochastic
assessment of building damage is developed in Chapter 3. The random
vibration method developed by Wen (1980) is adopted and modified for
reinforced concrete structures. For analyzing both weak-column-type and
weak-bearn-type buildings, a hybrid model for structural frames is
introduced. Based on extensive damage analysis of both SDF and MDF
reinforced concrete buildings, the potential destructiveness of ground
motions is expressed as a simple function of the rIDS acceleration and
strong-phase duration of ground motions.
In Chapter 4, the damage calculated with the proposed model IS
calibrated with respect to observed damage of nine reinfoced concrete
buildings, that were moderately or severely damaged during the 1971
San Fernando earthquake and the 1978 Miyagiken-Oki earthquake in Japan.
Based on this calibration, a damage criterion is developed for
specifying the tolerable damage state as well as extreme damage state
that corresponds to collapse.
Based on the proposed damage criterion, an aseismic design method is
developed in Chapter 5. The proposed design insures that the potential
structural damage will be within a tolerable damage level by properly
selecting both the strength and ductility of the constituent components.
Applications to various reinforced concrete frames are also. illustrated.
In Chapter 6, the reliability of a structure designed according to
the proposed method is examined. The uncertainties, including those in
seismicity, modeling error in structural analysis, and in the structural
7
capacity, are systematically identified and incorporated in the
reliability analysis.
c
C
1 .. 4 Notation
A = area of longitudinal reinforcing bar
b = width of cross-section
c = viscous damping coefficient
=
B
base shear coefficient
= effective strength of SDF system
e
D = damage index, or diameter of main reinforcement
D = limitiag damage index
a
D = damage
i
index of i-th story
D = response D value
r
D = overall damage index
T
d = effective height of cross-section, or diameter
dE = incremental absorbed hysteretic energy
E (6 ) = hysteretic energy per cycle at deformation 6
c
F = lateral seismic force at i-th floor
i
f' = compr es s i ve strength of concrete
c
f = restoring force of hinged columns in
i,.;,th
story
l.
h = height of cross-section
h = interstory height of i-th story
l.
I = characteristic intensity
c
I = intensity of strong motion phase
0
K initial stiffness
K = elastic stiffness of member
e
K = secant stiffness at yielding
y
shear span length
~
=
M(E;;,t) = number of peaks exceeding t;. at time t
M = cracking moment
c
}! = yield moment
y
m = mass
of stirrup
8
m = mass of i-th floor
i
= number of peaks per unit time
n = restoring bending moment of rotational spring
i
n
o
p
t
P
w
= normalized axial stress
= tension steel ratio
= stirrup ratio
Q ,Q ,Q
c y u
= cracking, yield, maximum shear strength, respectively
Q. = story shear force at i-th floor
1
q. =
1
restoring force of elastic columns in i-th story
R
d
= required ductility index
R
i
= allowable rotational angle of member
R = ductility index
T
T = fundamental natural period
T = predominant period of excitation
G
t
o
u, u., u
u
i
w
=
=
=
=
strong-motion duration
displacement, velocity
story drift
total building weight
w = i-th story weight
'i
x = ground acceleration
and acceleration,
= distance from centroid of tension
compression reinforcement
S = constant value in damage index
respectively
reinforcement to
o ,0 ,0 ,0 = flexural, bond slippage, inelastic shear, elastic shear
f b s e
deformation at yielding, respectively
o = maximum
M
o = ultimate deformation under monotonic loading
u
o = yield deformation
y
= damping ratio of the Kanai-Tajimi spectrum
G
e = rotational deformation at i-th floor
i
A = energy absorbing contribution factor of inelastic spring
1
= ductility factor
= ultimate ductility factor
u
P = confinement ratio
w
9
a = rms of strong-motion acceleration
G
a ,a,a = rms of displacement, velocity
u u u
respectively
~ . = interstory mode shape function
l.
~ = yield curvature
and
~ = perimeter of longitudinal reinforcing bar
w = predominant frequency of excitation
G
acceleration,
10
CHAPTER 2
DAMAGE HODEL FOR REINFORCED CONCRETE
2.1 General Remarks
Under earthquake loadings, a reinforced concrete structure is
weakened or damaged by a combination of stress reversals and high stress
excursions; consequently, any damage criterion should include not only
the maximum response, but also the effect of repeated cyclic loadings
(Gosain, et al., 1977; Banon, et al., 1980). Current aseismic design
codes, such as the UBC and SEAOC codes, are based primarily on research
results of flexural behavior of reinforced concrete under monotonic
loadings (e.g., Gaston, et al., 1952; McCollister, et al., 1954; Burns
and Siess, 1962). Under monotonic loadings, brittle shear failure can
be avoided through careful detailing of the members, and the ultimate
flexural capacity can be accurately evaluated (e.g., Burns and Siess,
1962). However, under repeated loadings, it is difficult to insure that
such members will not fail in shear. Several researchers (e.g., KBstB
and Bouwkamp, 1975; Japanese Short CO,lumn Committee, 1973-77) have shown
that reinforced concrete components designed to fail in flexure under a
monotonic loading tend to fail in shear or through loss of bond under
repeated loading reversals.
The unpredictable nature of the ultimate deformation capacity or
energy absorbing capacity may be another significant aspect of the
inelastic behavior of reinforced concrete structures under repeated
loading reversals. Recent studies by Banon, et ale (1980) and Hwang
(1982) showed a high uncertainty in the prediction of the ultimate
structural capacity of reinforced concrete components subjected to
loading reversals. Structural damage was expressed as a function of the
maximum deformation and absorbed hysteretic energy by Banon, et al.,
whereas, a simple scaling, in terms of the "energy index" was used by
Hwang. The results were based on limited data, 29 test specimens in
11
, It!t:;) LZ .t!e.i.: 0 "
UlJ.iTl8:C '. ,,1 enoe J:t
" s ~ 'Cy Or. III ~
208 N B106 NCEL in0l$:
'(J:rb ROIlline
ana, 1111 St:ceet
. nots 6180l)
Banon, et ale (1980) and 34 specimens in Hwang (1982), and significant
scatters were observed.
In the present study, a damage model is developed for expressing the
potential damage of reinforced concrete components as a linear function
of the maximum deformation and the absorbed hysteretic energy. In
developing the necessary damage function, extensive test data were
examined. Monotonic and cyclic test data of reinforced concrete, beams
and columns reported in the U.S. (e.g., Atay, et a l ~ , 1975; Burns and
Siess, 1962; Celebi and Penzien, 1973; Ma, et al., 1976; McCollister, et
al., 1954; Newmark, et al., 1963; Wight and Sozen, 1973) and Japan
(e.g., Hirosawa, et al., 1973; Hirosawa, 1975; Naka, 1963; Short Column
Committee, 1973-77) were used in a systematic regression analysis. The
pertinent data were all for rectangular sections reinforced with
deformed bars (loaded under a single-axis bending).
2.2 Proposed Damage Model
Consistent with the behavior of reinforced concrete components
described above, seismic structural damage is expressed as a linear
combination of the damage caused by excessive deformation and that
contributed by repeated cyclic loading effecte This may be represented
in terms of a damage index,
OM 8
D = 8 + -Q 0 IdE
u y u
or
where,
dE
E (0)
c
o = maximum response deformation under an earthquake;
M
o = ultimate deformation capacity under monotonic loading;
u
(2.1)
(2.2)
12
Q = calculated yield strength ( if the maximum strength,
Q
y u
< Q ,
Q
is replaced by Q ) ;
y y u
E (0 ) = hysteretic energy per cycle at deformation 0;
c
a,S = non-negative parameters;
0 = amplitude of deformation in each cycle of oscillation; and
dE = incremental absorbed hysteretic energy.
Values of the damage index, D, are such that D 1.0 signifies
collapse or total damage. Structural damage, therefore, is a function
of the response 0 and dE, that are dependent on the loading history,
M
and the parameters a, S, 0 , Q and E (0), that specify the structural
capacity_ In Eq. 2.2, the is weighted by a quantity
proportional to the power (a) of deformation levels, whereas in Eq. 2.1,
no such weighting is assumed.
2.3 Determination of Model Parameters
The damage model proposed in Eq. 2.1 contains the three parameters
0, Q and S.
u Y.
also
In Eq .. 2.2, two additional parameters a" and E (0) are
c
These parameters can be evaluated on the basis of
deformation analysis and available experimental data as described below.
2.3.1 Determination of a
u
When brittle shear failure is of no concern, such as for very
slender beams and columns, the ultimate deformation can be evaluated
from the stress-strain relationship of the material. However, as
mentioned earlier, components may fail shear under repeated cyclic
loadings even though a flexural failure is expected under monotonic
loadings. All possible failure modes should be taken into account in
evaluating a when repeated cyclic loadings are involved.
u
13 .
To date, there appears to be no reliab.le method for determining the
ultimate deformation of reinforced concrete components, especially when
shear deformation and bond-slippage may be dominant. Even highly
sophisticated finite element analysis cannot reproduce the deformation
behavior up to the ultimate stage because of uncertainties in the dowel
action, shear cracking, bond deterioration, three-dimensional stress
concentration, etc. In this .light, a simple empirical relation is
developed for determining 0 using available monotonically loaded test
u
data ..
Because the yield deformation, 0 , can be predicted with good
accuracy, a practical means to the ultimate deformation may be
to amplify this with the ductility factor ; namely,
o
u
0
u y
u
(2.3)
The values 11 and 0 may be determined independently.
YieldingUof a concrete component may be defined as the
first yielding in the tension reinforcement, or when the extreme
compressive fiber strain in the concrete exceeds 1.5 times the crushing
strain E ..
o
The yield deformation can be regarded as composed of the flexural
component, 0 ,
f
the deformation due to bond-slippage of the reinforcing
bar from its anchorage, 0 , the inelastic shear deformation, 0 , and the
b s
elastic shear deformation, 0 ; i.e .. ,
o
y
e
(2.4)
o may be evaluated by the conventional elastic beam theory. Three of
e
the components in Eq. 2.4 can be evaluated with reference to Figure 2.1
as follows ..
Flexural Deformation -- It is well-known that the yield curvature
can be reasonably determined by the plane section assumption, and the
curvature distribution along a member is approximately linear (Burns and
Siess, 1962) .. By assuming that the concrete in compression remains
14
elastic up to yielding of the tension reinforcement, the yield curvature
for beams can be expressed as follows (Park and Pauley, 1974),
E
Y
<P '
y (l-k)d
where:
p
a cr
- ~
- bdf'
p'
c
a 0'
c y
= bdf'
c
The above parameters are
(2.5)
E d
= ~
and S
c
Ct ; =-
Y
E C d
0
defined further in Figures 2.2 and 2.1.
Because of the inelasticity of concrete and the effect of axial forces,
Eq. 2.5 tends to underestimate the actual curvature. Based on results
of iterative analysis (e"g., Aoyama, 1964), the following improvement is
proposed:
n
{C
I
+ (C
2
- C
I
) O ~ 3 } ;
where:
c = 1 .. 05, for
p'
1=
0
1
+ le9p
2

4
, = 1 for
p'
= 0;
c = 1 + 0.45/(0 .. 84 + 2P' - p);
2
n
=
N/ (f' bd) ;
0 c
N = axial force.
When the extreme fiber strain E exceeds 1.5
c
will occur. In such cases, equating E =
C
becomes
I.SE
o
xh
(2.6)
E, compressive yielding
o
1.5 E , the yield curvature
o
(2.,7)
15
in which, x is the solution of
0.778 S'x
2
- (n _ yp' _ 1.5 )
c 0 ~ p x
or
x
where,
8'
C
y
1.286 (n + p - yp')/a'
o c
hid; y 1
f'
C
cr
y
and y'
Details are described in Appendix-A.
1.5
a.
y
1.5
~ p
y c
o (y':s; 1.0)
(y' > 1.0)
lis' - x
__ c __ :s; 1.0
x
Deformation due to Bond-Slippage -- With the assumption that the
deformation of concrete in the anchorage is negligible, the following
relation can be obtained between the bond stress, L, and corresponding
bond slippage S,
where:
A = area of the reinforcing bars;
~ = p e ~ i m e t e r of the reinforcing bars; and
E = modulus of elasticity of steel ..
s
(2 .. 8)
Using the idealized bond-slippage relation shown in Figure 2.4, the
stress in the reinforcing bar at. the face of the anchorage, cr , may be
o
expressed as a simple function of the normalized. slippage,
a-
o
(in ksi) (2 .. 9)
where:
A = S ID :::: 1/40; and
o
16
S = SID, the normalized slippage.
n
The validity of Eq. 2.9 had been examined with available pullout test
data (Ferguson, et al., 1965), assuming LM = 1.5 ksi for the bottom
bars, and L = 0.9 ksi for the top bars. When the anchorage length is
M
longer than 10D, Eq. 2.9 agrees well with experimental results as shown
in Figure 2 .. 5 ..
Results of other pullout tests (Clark, 1946 ; Jirsa and Marques,
1972 ; Koike, et al., 1974; Ma, et al .. 1I 1976 ; Morita, et al .. , 1967 ) are
also shown in Figure 2 .. 6 e On the basis of the above experimental
results, it can be observed that the bond slippage is largely a function
of the degree of compactness of the concrete, and independent of the
concrete. strength.. For the purpose of determining the deformation due
to bond slippage, a mean value of L = 1.2 ksi may be assumed if the
M
degree of compactness is not specified.
Inelastic Shear Deformation -- The shear deformation, 0, may be
S
obtained by subtracting the calculated 0 ,0 and 0 from the measured
f b e
yield deformation 0.. The ratio of the shear deformation to the
flexural as a function of the shear span ratio is shown in
Figure 2.7 for 244 beams and columns in which yielding of the tension
reinforcement is clearly recorded. In slender beams and columns, the
shear deformation is not prominent; however, as the shear span ratio
approaches unity (i.e., + 1.0), the shear deformation becomes
dominant ..
It is well-known that the bond stress and stirrup ratio also affect
the shear deformation. Available experimental results (e.g .. , Japanese
Short Column Committee, 1973-77) indicated that as the stirrup ratio is
increased to about 1%, the shear deformation generally decreases;
however, higher stirrup ratios do not always insure decreasing shear
deformations. Figure 2.8 illustrates a well-known bond failure
mechanism .. As the ribs of the longitudinal reinforcements bear against
17
the surrounding concrete, conical micro cracks are formed around the bar
0
with an inclined angle of about 45 (Goto, 1971). The presence of a
stirrup will disturb this failure mechanism and reduce the bearing
stress transmitted to the surrounding concrete .. Assuming that the
shaded part of the longitudinal bar (see Figure 2.8) is not effective
for bond, the effective average bond stress Tis,
B
I::.T
(2.10)
'(l - 1.71 nd)
where:
I::.T = difference of the forces in the reinforcement between both
ends of a member calculated by flexural analysis;
= perimeter of the longitudinal reinforcement;
. = length of member;
n = number of pairs of stirrups; and
d = diameter of stirrups ..
.
Therefore, according to Eq. 2 .. 10, an excessive number of stirrups has
the effect of reducing the effective bond length, and thus increasing
the bond stress. Figure would appear to suggest a weak positive
correlation between the shear deformation and the normalized effective
average bond stress
T IR. As seen in both Figures 2 .. 7 and 2.10, the
B c
scatter in the shear deformation is high if 0 is expressed as a
s
function of a single parameter. This scatter can be reduced by
expressing 0 as a function of several parameters, ./d, T and
s B
p (stirrup ratio).
w
The shear cracking pattern is different from member to member,
however, an idealized shear cracking model may
crack in a member
Figure 2 .. 9 with
deformation of a
o
s
I.e.. 8
l S
is assumed to be inclined
an identical shear rotation
cantilever is
be assumed.
0
at 45 as
e . Then,
s
Every shear
shown in
the shear
(2 .. 11)
18
in which, = the arm. length measured from the end of the shear crack.
i
According to Olsen, et ale (1967), the shear cracking load Q is
c
Q = If' bd + M (. - d/2)
c c c
(2.12)
~ n which, M = the flexural cracking moment. By slightly modifying
c
Eq .. 2.12, the length of a "no shear. crack zone," ', at yielding is
.' M /(Q - If' bd) + Z
eye
(2.13)
where, Z = the effective depth of the member. The location and number
of shear cracks are obviously unpredictable. It is reasonable to assume
that cracks are equally likely within the "shear crack zone if ( - ') ,
with an occurrence rate of p = liz. Then the mean shear deformation is
(2.14)
Using Eq. 2.14, the shear rotations,
available test beams and columns. On
e , were evaluated for the 244
s
the basis of the calculated
results, the following observations were obtained:
(i) There is a high negative correlation between the shear span
( ii)
( iii)
ratio and e
s
However, for shear span ratios /d <
8 is approximately independent of /d.
s
For T /1' > 5.0 and /d ~ 4.0, a positive correlation
B c
between 8 and T /II' was observed.
s B c
For /d > 2.5, the stirrup ratio, p , does not affect e
w s
Also, when p > 1%, e is independent of p
wsw
Based on these observations, the following is proposed for the shear
rotation of beams and columns.
0.002
8S = il d-O. S
0.002 .
l/d-O.S {1+0.27(u-S)}
19
'for u<S or l/d>4
for u>S and 2.S<l/d<4
(2.15)
0.002 . u-S
l/d-O.S {1+0.18S }
vP -0.4
for u>S and l/d<2.5
w
in which, Q,/d = shear span ratio (replaced by 1 .. 5 if 9.Jd < 1.5) ;
p
= stirrup ratio, in percent (replaced by 0 .. 2% if p < 0.2%); and
w
Iff.
w
u = T
B c
The total yield deformation, therefore, can be calculated with
Eq .. 2.4 .. A comparison of the calculated and normalized experimental
yield. deformation is shown in Figure 2.11 ..
Ultimate Ductility Factor Using Eq. 2.3, the ultimate ductility
factor l.l were 'calculated for 142 monotonically loaded beams and
u
columns.
In order to clarify'the definition of the "ultimate stage," the test
specimens were divided into four categories according to their ultimate
failure mode, as defined in Figure 2.12.
For a sudden failure, in which a member loses its load-carrying
capacity suddenly, there is no difficulty in identifying the failure
point.. Buckling of the compression reinforcement or fracture of the
tension or web reinforcement usually show this type of failure. For a
gradual failure, in which a member loses its strength continuously, the
failure point is identified on the load-deformation curve at which the
strength drop exceeds a certain percentage of its maximum strength.
Analyses of the ultimate ductility factors for the test specimens
showed a strong correlation with the flexural and shear deformations,
whereas there is virtually no correlation with the bond slippage. After
examining several mathematical formations, the following deformation
parameter was found to correlate well with l.l :
e:
p
u
(2.16)
20
in which,
maximum principal strain;
flexural concrete strain at the location
bar calculated using Eq. 2.6; and
of the compression
e = shear rotation of Eq. 2.15.
s
Figure 2.13 shows the plot between the principal strain normalized
and the ultimate ductility factor. The correlation between ~ and by e:
o
e: Ie: is
p 0
members)
maximized when the strength drop (for gradually
is assumed to be. about 10% of the maximum strength.
u
failing
However,
the strength drop at the failure point has been defined as 20% of the
maximum strength in this study, because the majority of tests to
ultimate collapse indicate that total repair is generally needed beyond
this point (e.g., Japanese Short Column Committee, 1973-77).
From igure 2.13, it can be observed that the ultimate ductility
factor is not influenced by the failure mode, indicating the validity of
Eq. 2.16. Figure 2.14 shows the same plot for ~ versus e: Ie: with
u p.o.
different confinement ratio p , which is defined as the volumetrlc ratlo
w
of the stirrups to the core concrete. A positive correlation between
~ and p may be observed in Figure2.14. Based on these observations,
u w
the following is proposed for the ultimate ductility factor ~ :
u
e: 0.218p -2.15
(1) W exp (0.654 p + 0.38)
e: W
o
(2.17)
in which, p = confinement ratio in percent (replaced by 2% if p > 2%),
w w
and ~ is replaced by 1 .. 0 if ~ < 1.0.
u u
Figure 2.15 shows the correlation between the calculated and
experimental ultimate ductility factor. Observe that the c.o.v. of 38%
is more than twice the corresponding c.o.v. in the calculated yield
deformation ..
21
2.3.2 Determination of B
The effect of cyclic loadings on structural damage is represented by
the parameter S in Eqs. 2.1 and 2.2. The absorbed hysteretic energy
(excluding potential energy) is integrated up to the failure point for a
large set (261) of cyclic test data of beams and columns. These
experimental data were carefully selected from a larger set of test
specimens; only those in which an abrupt failure is clearly observed or
gradual failure can be identified on the envelope curve were included.
Relatively ductile columns sometimes become unstable under loading
reversals due to the p - ~ effect; in this case, the failure point is
identified
stiffness ..
at which the load-deformation curve shows a negative
. Using.Eq. 2.1 or 2 .. 2, the load-deformation curve for each test is
traced up to the failure point. Then, at the point of failure, with
D = 1.0, the corresponding value of S is evaluated. Based on the
calculated values, a negative correlation was observed between Sand
the confinement ratio P , and also weak positive correlations were
w
observed between S and the shear span ratio 9..,/d, longitudinal steel
ratio p , and axial stress n .
t 0
Through trial-and-error, the minimum-variance values of S for
Eq. 2.1 were determined in such a way that the standard deviation of D
is minimized and the mean value of D is close to unity. In a similar
manner, and with a = 0.6, the corresponding values of S for Eq. 2.2 were
also obtained. The results yielded the following:
for Eq .. 2.1;
for Eq II 2 .. 2 ; S
in which,
(-0.447 + 0.73 lId + 0.24
P
w
(-0.165 + 0.0315 lid + 0.131 p
t
)0.84
(2.18)
(2.19)
22
~ / d = shear span ratio (replaced by 1 .. 7 if. ~ / d < 1 .. 7);
n = normalized axial stress (replaced by 0 .. 2 if n < 0.2) ;
0 0
p = longitudinal steel ratio in percent (replaced by 0.75%
t
< 0 It 7 5%); and
p = confinement ratio in percent ..
w
if
Pt
A large scatter can be observed between the calculated and
experimental results of S as shown in Figures 2.16 and 2.17.
2.3 .. 3 Determination of Q .and Q
y u
The method for evaluating the yield strength Q is described in
y
Appendix-A.. For the maximum strength Q of reinforced concrete beams
u
and columns subjected to reversed loadings, the following is proposed
based on the analysis of experimental data ..
Q = (1.24 - 0.15 p - 0.5 n )Q
u 0 y
(2.20)
in which p and n are defined in Eqs. 2.5 and 2.6.. The scatter
o
associated with Eq. 2.20 is relatively small: c.oov. = 12% (see
Figure 3.4).
2.3.4 .Determination of E ~
c
E (0) is the hysteretic energy per cycle at deformation 0" About
c
800 hysteresis loops from 261 cyclic test data were evaluated to
determine this parameter. The hysteretic energy (normalized by Q 0 ) is
plotted against the ductility factor in Figures 2.18 and 2 . r 9 ~ On
this basis, the relationship between the hysteretic energy and ductility
factor may be approximated by a bilinear curve with a transition point
at around ~ . = 1.5 (see Figure 2.20).
23
Figure 2.18 shows the normalized hysteretic energy at relatively
small- deformations" As the axial stress increases, the normalized
hysteretic energy tends to increase slightly; however, considering that
there is a significant scatter in the plot, the effect of axial stress
may be Linear regression of the experimental data for
< 1.5 yields the following:
E = Q 8 (0.77 - 0.22)
c y y
(2.21)
Figure 2.19 shows the data of normalized hysteretic energy for
> 1 .. 0. In this range, the ratio of the flexural deformation to the
total yield deformation, 8 /8, controls the _ energy absorbing
. . f y / . .
characterlstlcs. As the ratlo 8 8 lncreases, the rate of lncrease of
. f (y . 2 19) A 1 f
the absorbed energy also lncreases see Flgure " t low va ues 0
8 /8 , the scatter of the data increases considerably as shown in
f y
Figures 2 .. 19a through 2.19c; in this range, 8 and 8 become dominant.
s b
Linear regressions were also performed using the data with 1.0 to
determine the constants f
1
summarized in Figure 2 .. 21.
and f of Figure 2.20. The results are
2
Therefore, the hysteretic energy per cycle
in the large deformation range > 1.0) is given by:
8
f
8
f
E = Q 8 {(O.S + 2.34 - 1) + (0.7 - 1.54
c y y y y
(2 .. 22)
In evaluating
Equations 2.21
the ratio, 0 /0 , Eqso 2.4 and 2.6 may be used.
and 2.22 mayf also be used for properly modeling the
hysteretic behavior of reinforced concrete components under dynamic
loadings. The average c.o.v.'s about Eqs. 2.21 and 2.22 are
approximately 25%.
2.4 Damage Index Statistics
With the appropriate model parameters evaluated
damage indices for each of the 142 monotonic
empirically above,
and 261 cyclic test
24
specimens were evaluated with Eqs. 2.1 and 2.2 at the respective failure
points. The results obtained with Eqsa 2.1 and 2.2 are plotted on
lognormal probability paper in Figures 2.22 and 2.23, respectively.
These show that the damage index is reasonably lognormally distributed
with the respective standard deviation indicated in Figures 2.22 and
2.23. The standard deviation associated with Eq. 2.2 is slightly lower
than that of Eq. 2.1. However, by virtue of its simplicity, Eq. 2.1 is
preferred for seismic damage assessment of reinforced concrete
structures.
The high degree of scatter (or uncertainty) in the damage capacity
of reinforced concrete components (c.o.v. ~ 0.5) indicated in
Figures 2.22 and 2.23 is in agreement with results of similar previous
studies, e.g., Banon, et al. (1980) and Hwang (1982). It may be
emphasized that such high uncertainty should be expected, as the
capacity under repeated cyclic loadings is much less predictable than
under monotonic loadings.
25
CHAPTER 3
DAMAGE ANALYSIS OF REINFORCED CONCRETE BUILDINGS
3 .. 1 General Remarks
Based on the damage model developed in the previous chapter, a
method for assessing building damage is developed herein. As shown
earlier, a high degree of scatter (or uncertainty) is observed in the
ultimate structural capacity of reinforced concrete components.
Random vibration technique is used to obtain the response statistics
necessary for damage assessment" The application of random vibration
techniques to nonlinear earthquake response problem has attracted the
attention of many researchers for the last two decades (e"g., Iwan and
Lutes, 1968; Minai, et al,,", 1970; Kobori, et al., 1973; Asano, 1976;
lemurs, 1977; Wen, 1980; and Grossmayer, 1981). In this study, the
equivalent linearization method (Atalik and Utku, 1976;" Wen, 1980) is
used. The versatility and efficiency of this method for MDF degrading
hysteretic systems have been demonstrated by Baber and Wen (1981) and
Sues, Wen, and Ang (1983).
In assessing damage of building structures, the dynamic failure
mechanism pertinent to reinforced concrete structures should be properly
reflected. A hysteretic model of reinforced concrete components and a
structural modeling technique are introduced for this purpose.
The general trend of the damaging process of reinforced concrete
structures, as well as the ground motion parameters
structural damage are examined through the damage analysis
SDF and MDF reinforced concrete structures.
pertinent to
of various
26
3.2 Methodology
3.2.1 Damage Index Statistics
The damage index defined in Eq. 2.1 ~ s used for assessing the
structural damage of reinforced concrete components under earthquake
loadings. In order to incorporate the uncertainties in the structural
capacity as well as in the random response, a damage index ratio may be
defined as follows:
D
r
Di = D
a
where:
D
=
the
i
damage index ratio for member-i;
D
= structural damage defined by
Eq"
2.1 ;
r
D
= ultimate structural capacity, with
a
In such a formation D. > 1 means D
~ r
using second-order approximation
> D , and
a
(Ang and
D
a
(3.1)
and
= 1.0, a
D
= 0 .. 54.
therefore collapse. By
Tang, 1975), the mean and
variance of D
1
is (the correlation between 0 and
M
dE is neglected),
D.
1
Var [D. ]
1
(3 .. 2)
(3.3)
Thus, the statistics of damage index ratio can be obtained from the mean
and variance of the responses 0 and dE.
M
27
3.2.2 Force-Deformation Relationship
The force deformation relationship of a structural component is the
basic and essential structural property necessary for damage analysis.
This must include, in particular, the envelope function of the
force-deformation curve and the hysteretic energy absorbed per cycle of
loading. A detailed and. complete description of the force-deformation
relationship for reinforced concrete would also require the shape of the
hysteresis loop (including "pinching effect") and the strength
degradation under a repeated cyclic loading. However, these latter two
characteristics have only minor influence on the. resulting response
(e.g., Umemura and Takizawa, 1982), but will add considerable complexity
to the damage analysis of MDF buildings.
Envelope Curve -- It is a common practice to express the envelope
curve for reinforced concrete as a multilinear function with three
turning points; namely, the cracking point, yield point, and the maximum
strength point, as shown in Figure 3.1.
The coordinate of the yield point was discussed in the previous
chapter .. The cracking point can be determined by the elastic stiffness
K and the cracking moment, M , which is usually defined as,
e c
M
c
., . r;:-t"f' '7 ...L. Tl.TL. I c.
I V r' I...J I lUll U
C e
in which, Z = elastic section modulus; and N = axial force.
e
On the
basis of a large number of force-deformation curves, the "apparent
cracking momentlYM' is proposed for the first turning point, as
M'
c
c
11 1fT Z + Nh/6
c e
(3.5)
Whereas, for the second turning point, the "apparent yield deformation"
0' is proposed as
y
o y
y
(0.485 + 3.3 K /K - 0.836 n )0
yeo y
(3.6)
28
in which;
K = elastic stiffness;
e
K = secant stiffness at yielding; and
y
n = normalized axial stress (replaced by 0.2 if n < 0.2).
o 0
The ratio
error in 0'
K /K should be also replaced by 0.2 if K /K > 0.2.
y eye
as modeled with Eq. 3.6 can be seen in Figure 3.3,
has
y
a c .. o.v .. of 21%.
The
which
For the maximum strength of reinforced concrete beams and columns
subjected to reversed loadings, Eq. 2.20 may be used. As shown in
Figure 3 .. 4, the scatter in predicting Q is relatively small; CeO.V. of
u
12%. The portion of the maximum strength point is affected largely by
the loading history.
Hysteretic Epergy Per Cycle -- The hysteretic energy per cycle
by Q 0) may be expressed as a function of the ductility
f
yy. 31 Th d ..
actor as shown ln Flgure e etermlnatlon of the necessary
parameters, and f , were explained in Chapter 2.
1 2
Degrading -- The degrading hysteretic model for reinforced
concrete structures is shown in Figure 3.2. The hysteretic restoring
force is assumed to deteriorate in both strength and stiffness as a
function of the mean maximum response 8 .
M
3.2 .. 3 Response of Hysteretic Systems
The equation of motion of a hysteretic system under
excitation may be given by,
,.
mu + eu + aku + (1 - a)kZ
and,
z = - +
-mx
g
ground
(3.7)
(3 .. 8)
29
Metz Beference
University Illinois
BI06 NeEL
208 N. Romine Street
Illinois 61802
in which, m, c, k are the mass, viscous .damping and initial stiffness,
respectively; x = ground acceleration; Z = hysteretic component of
g
restoring force; and a, y = parameters. The parameters a and yare time
dependent and are functions of the energy dissipation per cycle E (0)
c
and the peak response (8 ,p). The restoring force at the peak, P , is
M M M
determined on the envelope curve as illustrated in Figure 3.2.
Therefore, a and y degrade according to
/ 2 2
a = O.SeA + 3A - 2 + I(e /4 + 3e + 4)A - (8 + 3e)A + 4
1 - a
y
where:
A = P /K 6 , restoring force deteriorating ratio; and
M M
e = E (6 ) /8 P, energy ratio.
c M M M
(3 .. 9)
(3.10)
Following Wen (1980), through an equivalent linearization procedure
(Atalik and Utku, 1976), Eq. 3.8 is reduced to
o (3 .. 11)
in which,
C
21
= o. 5 II [E ( liZ) + 0 ] - 1. 0 an d
y 1T 0" Z
U
k22 = o.sy/i [oe +
1T U 0
Z
The earthquake excitation may be modeled as a filtered shot noise
with the Kanai-Tajimi spectrum (Tajimi, 1965),
30
4 2 2 2
wG + 4wG l;G
W
S (w) S
0
(2 2)2
2 2 2
W - wG + 4wG l;G
W
The equation of motion, expressed in matrix form, is
Gy + F
in which,
and,
T
Y = {u, z, x , u, x }
G g
0 0
0 -K
22
G = 0 0
-aK/m -(1-a)K/m
0 0
T
F = {O, 0, 0, 0, x }
g
0 1 0
0 -c 0
21
0 0 1
2
W
-c/m
2l; W
G G G
2
-w
0
-2l; W
G G G
(3.12)
(3.13)
This leads to the equation in terms of the covariance matrix S (Lin,
1967) ,
~ S
dt
where:
GS + SG + B
31
S = E[y(t) y(t)
T
] ;
B
0, except B = I( t) ; and

55
= intensity function of excitation.
In proceeding response deterioration is included by
continuously updating the foregoing parameters A and e.
3.2.4 Maximum Response Statistics
In the stochastic assessment of seismic damage, a method for
evaluating the maximum response statistics is clearly important. Since
rigorous solution is not obtainable especially. for nonstationary
processes, an effort has been made for developing approximate methods
based on the Poisson assumption
. Designating the number of peaks by M and the number of peaks per
unit time by N, the mean value of M above the at time tis,
t)
t
f E [N (, t) ] d.t
o
(3.15)
The response may be approximated by a Gaussian stationary process during
each incremental time intervale Then, according to Middleton (1960),
the following is obtained;

where:
a
u
0
u
(3.16)
a = 0 ..
3 u
I s I
The following approximate solution to Eq. 3.16 is proposed (its
32
derivation is given in Ap lendix-B);
E [N ( ~ , t) ]
, for ~ <0 ( 3 .. 1 7 )
in which,
Using Eqs. 3.15 and 3.17, the mean peak number can be evaluated by a
simple numerical integration at several discrete points-" Based on the
assumption of Poisson occurrence of peaks, the probability distribution
function for the maximum response 0 may be expressed by-the peak
M
distribution function (Lin, 1967). Since the relevant distribution
function is for the absolute maximum value, both positive and negative
peaks should be taken into account; i"e.
where:
(
_
_ ~ J t {02 + 0 3 ( 2) } )
M = total number of peaks I-v dt ..
T 271" 0 01 O
2
This leads to the Gumble Type I distribution (Gumble, 1954),
in which:
-a. (y-u )
n n
exp [-e ]
u = characteristics extreme; and
n
a. = inverse measure of dispersion.
n
(3 .. 18)
(3.19)
33
By equating Eq .. 3 .. 19 and Eq. 3 .. 18, the following are obtained:
a.
n
where:
M'(x,t) =,jl M(x t)
dx '
The mean and standard deviation of
u (Gumble, 1954). The rms response
n
directly from Eq. 3.7, viz
CJ
u
E[ {- x
g
c
m
. k
u-a.-u
m
o are obtained
M
of acceleration
(3.20)
(3.21)
from a. and
u
is obtained
(3.22)
Eq" ,3 .. 22 can be evaluated using the response covariance matrix S ..
3.2 .. 5 Absorbed Hysteretic Energy
The mean of the absorbed hysteretic energy is determined as
t
f dE
o
t
J (1
o
(3 .. 23)
The variance of the absorbed hysteretic energy is obtained from the
two-time covariance matrix Set ,t ) (Pires, 1983), which is the solution
1 2
of the following differential equation,
(3.24)
The matrix. SCt
l
,t
2
) may
integration with the initial
Pi r e 8, 1 983) "
34
be . obtained. through
condition t = t =
1
a backward numerical
t (for details, see
2
3.3 Damage Analysis of SDFSystems
In order to examine the accuracy of the above-mentioned random
vibration method and to examine the basic characteristics of the damage
process of reinforced concrete members, two columns are analyzed as SDF
systems. A relatively ductile column-A (reported in Japanese Short
Column Committee, 1973-77) is modeledas,a SDF system with a natural
period of 0.6 Sec. and yield acceleration capacity of 0 .. 2g; a
relatively brittle column-B (reported in Naka, 1963) has a natural
period of 0.3 sec. and yield acceleration capacity of .0.4g (see
Figure 3.5).
Figure 3 .. 6 shows the comparison of the analytical results to
Monte-Carlo (100 samples) under white noise excitations.
The numerical error in this random vibration solution is negligibly
small for engineering purposes.
The damage index under earthquake excitations is also examined.. The
ground motion is modeled as a filtered shot noise with the Kanai-Tajimi
spectrum and. the Amin-Ang type intensi ty function (Amin and Ang, 1968) ..
For the filter parameters of the spectrum, the following were assumed;
w = 571' and
C;; = 0 .. 6
G G
The intensity function is ,as,

2
l(t) = I for 0 < t < t
o tl 1
= I for t < t < t
0 1 2
(3 .. 25)
= I {exp - 2( t - t )/t
}
for t < t
0 2 0 2
35
in which:
t = 0 .. 15 t and
t2
= 1 .. 15 t ;
1
,
0 0
t
= the strong-motion duration; and
0
I = intensity of strong motion.
0
Figure 3 .. 7 shows the mean damage index for various strong-motion
duration t and rms of ground motion a (in terms of g). Assuming D to
o G
be lognormally distributed, the probability of exceeding a specified
damage level is illustrated in Figure 3.8. Although the strength of
column-B is twice that of column-A, the damage is considerably higher in
column-B; indicating the importance of ductility relative to
earthquake-resistance .. On the basis of these response results, it was
found that for given ground motion parameters, i.e., w and ~ , the mean
G G 1.5 0.5
value of D is approximately proportional to the quantity a t
G 0
as shown in Figure 3.9. Considering that there is no significant
scatter in the plots over a wide range of strong-motion durations (from
1 sec. to 36 sec.), the "characteristic intensity" may be a reasonable
representation of the potential. destructiveness of ground motions, i.e.,
I
c
This simple relationship is
analysis of MDF buildings. It
(3.26)
examined further below for the damage
is expected, however, that when the
variability of w and ~ are introduced, the scatter would be higher ..
G G
3.4 Modeling of Buildings
3.4.1 G e n e r ~ J
In modeling buildings for earthquake response analysis, the
so-called "shear beam model" has been used extensively. Among the
36
drawbacks in this simple model is the incapability of reflecting the
coupling between stories for weak-beam-type buildings. Takizawa (1975)
proposed one solution to this problem; based on a member-by-member
analysis of frames with various failure mode_B, -the failure-coupling was
accounted for by assigning an equivalent SDF to the coupled stories,
resulting in an equivalent-degree-of-freedom system
simple model is useful for an approximate estimation
damage; however, for nlOre quantitative assessment
detailed modeling would be necessary.
3.4.2 A Hybrid Model
model.. Such a
of the overall
of damage a more
The proposed structural model for multistory, and multi-frame
reinforced, concrete building is illustrated in Figure 3.10. It is
essentially an extension of the conventional shear-beam model by adding
elastic bending columns at each floor and connected to the floor by
inelastic rotational springs.. The deformations of the constituent
members in' terms of the hinge rotation are explicitly related to the
story-level response through simple geometric relations (see
Figure 3.11); thus, the story-level damage index is obtained through a
combination of member-level damages. As shown in Figure 3.10, each
column is classified into three categories depending on the failure mode
at the two joints. A column weaker than the beams at both joints is
classified as a "hinged column," and modeled with the conventional shear
beam spring by assuming rigid floors. ,When a weak beam-strong column
condition prevails at both joints, the columns are regarded as elastic
springs; in this case, the inelastic behavior is represented by the
hinge rotations of the beams. Where a mixed failure mode is involved, a
column is classified as a "half-hinged column;" the half of the column
length is assumed to' remain elastic, whereas the other half is allowed a
hinge rotation ..
37
. The equilibrium at each floor yields the following:
n
m.( I u. + i ) +q; - qi+l + f. - f
i
+
1
= 0 ,
1. j=l J g. 1.
i=l,n (3 .. 27)
in which:
6EI. Zu.
1. ( ___ _ e e )
q; = -Z- h . - . 'I
h. i 1. 1.-
(3.28)
1.
and
ZEI. 3u. 2EI'+
1
3u'
1
1. (Z e 1.) 1. (Z e e _ 1.+) = 0 0
n
t
e
i
+ i-l-, hi + h
i
+
1
i+ i+l h
i
+
1
,i= ,n
(3 .. 29)
where:
- the mass of each floor; m
i
EI = the bending stiffness of an elastic column (or walls); and
i
h = the height of each story.
i
The nonlinear restoring force of the 'translational
spring n are, respectively,
i
f.
1.
n.
1.
c. u. + u. k. u. + (1 - a.) k. Z.
C 1. 1. C 1. C 1. 1. C 1. C 1. 1.
b
a
. bk. e. + (1 - ba')b
k
. Y.
1. 1. 1. 1. 1. 1.
spring f and
i
(3 .. 30)
(3.31 )
Equations 3.27 and 3 .. 29 may be expressed in matrix form as'follows:
{ ii }, = [ c ] { + [ k] { u } + [ B ] { Z} + [A] { e }
(3.32)
[F]{e} + [D]{Y} + [E]{u} o
Eliminating the rotational deformation {e} yields,
38
{ti} [ c ] { ~ } + {[k]-[A] [F)-l[E] }{u} +[B]{Z} -:- [A] [F]-l[D]{Y}
(3.34)
For the hysteretic components Z. and Y., the following equations are
1. 1
obtained after linearization (see Eq. 3.11); namely,
..
Z. + C'll u. +
1. C "-. 1.
1.
k"'2
c "-
.I
Z.
].
o
(3.35)
(3 .. 36)
Introducing the following relationship between the rotational and the
translational deformations;
{e} = [H]{u}
(3.37)
where the matrix [H] is determined from the compatibility condition of
the continuous elastic columns; Eq. 3.36 becomes
o (3.38).
Equations 3.34, 3.35 and 3.38 form a set of first-order differential
equations identical to Eq. 3.13.
response vector y is
In this case, the corresponding
y
.. .. T
{ u
l
' ... ,u ,u
l
'..., U , Z 1 ' ... ,Z ,Y , ... , y }
n n non
(3.39)
It may be observed that the hysteretic components Y
e
are added to the
l.
response vector of the corresponding shear-beam model. The details of
the matrices in Eqs. 3.32 and 3.33 are illustrated for a three-story
building in Appendix-C.
39
3.4.3 Force-Deformation Relation and Damage Index_of Inelastic Springs
The translational (shear-beam) springs and rotational springs
illustrated in Figure 3 .. 10a may be simply called "inelastic springs" and
regarded as a parallel system consisting of the constituent elements,
i.e., hinged columns and beams. Therefore, the deformation response,
after adjustment for the rigid zone as shown in Figure 3.11, is the same
in every constituent element within each inelastic spring; the restoring
force and absorbed energy of each spring is obtained by superposition.
The envelope curve and the hysteretic energy per cycle of each spring
may then be simplified by a multilinear function as shown in Figure 3.1.
The ultimate deformation 0 for each spring is determined such that the
u
potential energy up to the ultimate deformation is equivalent to the sum
of those of the constituent members.
The coefficient S for each spring is obtained from those of the
constituent members as follows:
(Q 0 ).
\' Y U 1
LS.
i 1 LQ.o(:8
1 U
(3.40)
l.n which,
S = the
S
for each spr ing;
s
S. =
the
S
for a constituent member; and
1-
s
0 = the ultimate deformation for each spring.
u
Considering that the constituent members in each spring are
constructed under identical condition, it is reasonable to assume that
the ultimate damage capacities of the members are highly correlated.
Therefore, the standard deviation of D of each spring is assumed to be
the same as that of the individual elements.
40
3.4.4 Overall Damage Ass2ssment
In addition to the nember and story-level damage measures, an
indicator for assessing tne overall damage sustained by a building would
also be useful. Such an indicator should reflect the damage
concentration in the weakest part of a building, e.g., the first story
or top story as frequently observed, as well as the distributed damage
throughout the building, which is likely in weak-beam-type buildings.
It is well recognized that- the damage distribution is closely
correlated with the absorbed energy.distribution (e.g., Akiyama, 1980).
Therefore, the overall damage may be expressed as the sum of the damage
indicies of each spring D., weighted by the corresponding energy
].
contribution factor A.; namely,'
].
D = EA D
T i i
(3.41)
where, A. = E./EE., in which E. is the total absorbed energy (including
]. ]. ]. ].
the potential energy) of spring-i. D may be called the "overall damage
T
index .. "
3.5 Examination of Building Types
Three different reinforced concrete frames (three-story and
three-bay) are designed according to the SEAOC Code (1975) and the ACI
Code (1979). Figure 3.12 shows the configuration and detail of the
frames. The basic design parameters are as follows: concrete strength
is 3 ksi;yield strength of steel is 50 ksi; combined dead and live load
is 0.18 kips/sq. ft for -every floor; span length is 20 ft in both
directions; lateral design seismic force is defined by a seismic
coefficient of CS = 0.14 (SEAOC, 1975).
Frame-A is a typical weak-column-type frame, whereas Frame-B is
designed according to the SEAOC Code and the ACI Code for ductile
moment-resisting space frame. Frame-C is designed as a
41
"rocking-waIL-type" frame; the shear walls are designed to be stronger
than the sum of the strength of the connected beams and the rocking
resistance at the base .. Figure 3 .. 13 shows the load-deformation curves
in terms of'the base shear and the deformation at the top of the frames.
The failure modes under lateral static loads are also shown. Combining
these three types of frames, a variety of buildings may be obtained ..
The following three buildings, designated as SA, SB, and SC were
examined ..
SA--only Frarne-A is used (weak-column-type building). The uniformly
distributed weight of a story is determined so that the yield base
shear is 0.26g. This satisfies the minimum requirement of the SEAOC
Code when K = 1.0. The calculated fundamental natural period is
0 .. 592 sec.
SB--only Frame-B is used (weak-bearn-type building). The yield base
shear is 0.15g, which satisfies the minimum requirement of the SEAOC
Code when K = 0.67 (moment ductile space frame). The fundamental
natural period is 0.955 sec.
SC--two Frame-A and one Frame-C are connected in parallel. The yield
base shear 'is the same as SA. The fundamental natural period is
0.68 sec. For the rocking spring at the base of the shear wall,
hysteresis loop is determined so that hysteretic energy dissipation
through rocking motion is equivalent to 10% of critical damping at
all deformation levels ..
The ground motion is represented as a filtered shot noise described
before. The viscous damping is assumed to be 3% of critical. Durations
of 5 sec. and 15 sec. were considered for the strong-motion phase. Two
to four values of the rms excitation IT (in terms' of g) were used. The
, G
final results in terms of the maximum damage index and the overall
damage index are summarized in Table 3.1 for different excitations
(Le. , different IT and t ).'
G 0
42
The mean value of the story drift, absorbed. hysteretic energy, and
damage index for building SA are shown in Figure 3.14. The damage
concentration in the first story is clearly observed both in the stqry
drift and in the absorbed hysteretic energy. For the shorter excitation
(t = 5 sec.), the damage is caused primarily by excessive deformation;
o
whereas for the longer excitation (t = 15 sec.), the effect of repeated
o
cyclic loadings becomes dominant.
The corresponding results for building SB are shown in Figure 3.15.
The effect of the interstory coupling is reflected in the almost uniform
distribution of the story drifts. Although the sustained deformations
at the hinges are almost the same for all members, the damage is
concentrated in the bottom of the columns in the first story, indicating
the poor ductility of the columns (because of relatively high axial
stress). The dominant effect of repeated cyclic loading on structural
damage under the longer excitation duration is also observed in this
building. By comparing the results for buildings SA and
SB, the advantage of the weak-beam-type building is apparent (see also
Table 3.1). In the weak-beam-type building, SB, the input energy is
distributed more uniformly throughout the building; whereas, the input
energy tends to concentrate in the weakest story in the case of the
weak-column-type building, SA. However, the increased energy absorbing
capacity is made possible at the expense of larger inelastic deformation
throughout the building (see Figure 3.15), which may lead to more
extensive nonstructural damages.
The results for building SC are shown in Figure 3.16; damages in the
columns and beams are shown separately. The interstory coupling due to
the rocking-type shear walls is reflected in Figure 3.16. Because of
the uniform distribution of the story drift, an almost uniform
distribution of energy absorption is achieved in the weak-column-type
frames .(Frame-A). The damage in the lower story columns (see
Figure 3.16a) tends to increase because of increased axial stresses.
In the damage analysis of SDF systems, under specified ground motion
parameters, a linear relationship was observed between the
"characteristic intensity," I , and the damage index, D. Similar linear
c
43
plots are also obtained for the story-level damage indices as shown in
Figure 3.17; the corresponding relationship for the overall damage index
of the three buildings is also shown in Figure 3.18. These results
would suggest that the potential destructiveness of ground motions can
be appropriately represented by the characteristic intensity, I
c
The overall seismic performance of buildings may be expressed by the
"damage ratio,"
R
(3.42)
In this term (see Figure 3.18), the building SC is almost equivalent to
building SB, whereas SC or SB is two times more damage-resistant than
building SA (i. e., R = l.R ).
SC 2 SA
The overall uncertainty in the estimated damage is measured by the
c.o.v. of the damage index; such c.o.v.'s are shown in Figure 3.19 as a
function of the mean damage index D ~ It may be observed that the degree
of uncertainty in the estimated damage is largely independent of the
building type, excitation duration, and level of damage.
44
CHAPTER 4
CALIBRATION OF DAMAGE INDEX
4.1 General Remarks
As mentioned earlier, some level of damage should be anticipated and
permitted in the earthquake-resistant design of buildings, especially
lowrise and medium-rise reinforced concrete buildings. Considering the
infrequent occurrence of damaging earthquakes, only damage of critical
consequence may be relevant in the design of most buildings; namely,
excessive structural damage or collapse that may cause loss of building
function and more importantly loss of human lives.
For practical meaning and usefulness, the proposed damage index may
be defined through calibration with observed seismic damages from past
earthquakes. For this purpose, the actual damage data of nine
reinforced concrete buildings, that were moderately or severely damaged
during the 1971 San Fernando earthquake and the 1978 Miyagiken-Oki
earthquake in Japan, were used.
4.2 Description of Building Damage
Building details and results of extensive damage inspection of the
above-mentioned buildings are available in the literature. The typical
plan and elevation for each building are shown 1n Figure 4.1. For
building A, i.e., Olive View Hospital, damaged during the 1971 San
Fernando earthquake, the primary source of information is the U.S.
Department of Commerce (1973); for buildings B through I, damaged during
the 1978 Miyagiken-Oki earthquake, the information from the
Architectural Institute of Japan (1980) is used. Photographs for each
building (reproduced from the U.S. Department of Commerce, 1973 and the
Architectural Institute of Japan, 1980) are shown in Figure 4.2. The
45
directions in which the photographs were taken are also indicated in
Figure 4.1. In almost all the buildings, the damage was prominent in
the longitudinal direction caused mainly by the lack of shear walls in
this direction; accordingly, the damage of the buildings was evaluated
in the corresponding directions indicated in Figure 4.1. The structural
damage for each of the buildings may be described briefly as follows.
4.2.1 Building A (Olive View Hospital)
This six-story building is the main unit of the Olive View Hospital
which was heavily damaged during the 1971 San Fernando earthquake. The
building was located 3 miles from Pacoima Dam where a maximum horizontal
ground acceleration of 1.25g was recorded. Because of the symmetric
plan, only one-quarter of the building was analyzed as shown in
Figure 4.1A. The damage was concentrated in the lower two stories.
Almost all of the tied columns in the base story collapsed as shown in
Figure 4.2A-b. The first story suffered a lateral displacement ranging
from 15 to 28 inches as shown in Figure 4.2A-a. Most of the columns in
the first story were reinforced with closely spaced spirals; however,
these spirals were not e x t e n d ~ d into the joints. Since large-size
deformed bars (#18) were closely reinforced in the longitudinal
direction, conspicuous bond failure was observed in these columns. The
upper stories suffered slight or no damage.
4.2.2 Building B (Taiyo Gyogyo Building)
The first story of the three-story office building collapsed as
shown in Figures 4.2B-a and 4.2B-b. The maximum acceleration at the
building was believed to be more than 0.45g according to the
Architectural Institute of Japan (1980). There were no shear walls in
the longitudinal direction in the first story.
46
4.2.3 Building C (Tohoku Kogyo University)
This four-story building consists of two weak-column-type frames at
both ends, with a weak-beam-type frame in the middle. All of the
columns in the north side frame are shortened by hanging walls and
standing walls as shown it. Figure 4.2C-a. Figure 4.2C-b shows the shear
failure in these columns. Because, of the weak-beam-type frame, the
structural damage was distributed throughout the building. The building
was demolished although there were no permanent horizontal distortions.
4.2.4 Building D (Saigo School)
One-third of 'this two-story school building was buil t in 1962.
Eleven years later, i ~ 1973, the extension was added in the longitudinal
direction. During this period, the Japanese
was changed primarily to require ductility.
seismic design provision
As shown in Figure 4.2D-b,
major structural damage, such as shear cracking and crushing of concrete
were concentrated in the columns of the older part. This part was
demolished, whereas the newer part (with much less damage) was repaired
for continued use.
4.2.5 Building E (Tonan High School)
This three-story building is a typical school building in
The damage was concentrated in the first and second stories.
slight damage was observed in the third story. Conspicuous
failure of short columns is shown in Figure 4.2E-a.
Japan.
Only
shear
47
4.2.6 Building F (Kinoshita Menko Building)
This three-story office building suffered major damage. Almost all
of the columns in the first story collapsed through shear failure as
shown in Figure 4.2F-b. A slight sag was also observed in the second
floor. On the second and third stories, extensive shear and flexural
cracks were observed in the columns and shear walls.
4.2.7 Building G (Obisan Office Building)
This single-bay office building collapsed during the earthquake as
shown in Figure 4.2G-a. The columns on the west side collapsed
completely as shown in Figure 4.2G-b, whereas the columns on the east
side continued to support one half of the building. The upper two
stories, supported by shear walls, suffered no damage.
4.2.8 Building H (Fukushi Kaikan Building)
This two-story building suffered only slight damage. After the
earthquake, sporadic occurrence of crackings was observed in the short
columns as shown in Figure 4.2H-b.
4.2.9 Building I (Izumi High School)
This is a typical school building in Japan similar to building E.
Damage was concentrated in the short columns of the north frame as shown
in Figure 4.2I-a. The relatively long columns of the south frame
suffered only slight or minor damage as shown in Figure 4.2I-b. This
building, with relatively less damage than that of building E, barely
escaped demolition.
48
4.3 Estimation of Ground Excitation
As mentioned in Chapter 3, the potential destructiveness of the
ground motion may be represented by the characteristic intensity I. In
c
order to further examine and establish the usefulness of the
characteristic intensity, a data set of 140 horizontal components
recorded in California were analyzed. This is the same data set
selected by McGuire and Barnard (1977), and used by Vanmarcke and Lai
(1980). The strong-motion duration, t , was determined using the method
o
proposed by Vanmarcke and Lai (1980). Figure 4.3 shows the relation
between the peak acceleration and the characteristic intensity. The
attenuation of I in the 1971 San Fernando earthquake is shown in
c
Figure 4.4.
In analyzing the above buildings, the characteristic intensity 1S
used for estimating the severity of the ground motion. Because no
accelerograms were recorded for any of these buildings, both the
intensity and frequency content of the input excitations were estimated
based on the available
For building A, which was damaged
earthquake, the attenuation relationship
in the 1971 San Fernando
given in Figure 4.4 is used
directly for determining the ground motion intensity.
is described by the Kanai-Tajimi spectrum with
The ground motion
filter parameters
w . = 16e5 and = Oe8e
G G
In the 1978 Miyagiken-Oki earthquake, accelerograms were recorded at
several locations; moreover, an extensive investigation was conducted on
the damage to tombstones for estimating the peak acceleration at and
around the damage sites (Architectural Institute of Japan, 1980). Based
on these data, the characteristic intensity of the excitation at each
building site was estimated as summarized in Table 4.1; the filter
parameters of the Kanai-Tajimi spectrum were assumed to be w = 6.3 and
G
= 0.9 based on the calculated spectra
G
(Architectural Institute of Japan, 1980).
of recorded accelerograms
49
4.4 Results of DamageAnalyses
The buildings were modeled based on the available design plans and
elevations, and dynamic as well as static analyses were performed using
the method described in the previous two chapters. The results of the
respective damage analyses are summarized in Table 4.1. The basic
characteristics of the buildings are also listed; namely; the
fundamental natural period, yield acceleration capacity of the weakest
story (in terms of g), and the ultimate ductility factor capacity of the
most brittle story. The results of the damage analysis for each
building are described below.
4.4.1 Building A
The results of static analysis (story strength and ultimate
ductility factor) and dynamic analysis (the energy contribution factor
A and the damage index of each story) are summarized in Table 4.2.
i
The calculated ultimate ductility factor ~ of the tied columns in the
u
ground story ranges from 1.8 (for columns with high axial stress) to 6.5
(for columns with low axial stress). That of the spiral columns in the
first story ranges from 2.7 to 4.8; whereas for the corner columns, ~
u
1S no greater than 2.0. Under a static loading (linearly distributed
lateral forces), the lower two stories yielded at a base shear
coefficient of 0.22, whereas the upper stories remain in the elastic
range. Consequently, damage is concentrated in the lower two stories as
indicated by the values of D in Table 4.2.
1
4.4.2 B u i l d i ~
The story-strength of this.weak-column-type three-story building is
relatively well-proportioned. Under a static loading, all of the
stories would yield at a base shear coefficient of 0.48. However, the
50
columns in the first story is relatively brittle due to the high axial
stress and poor confinement (confinement ratio p = 0.15%).
w
Consequently, damage was concentrated in the first story; the calculated
damage index exceeds 1.0 in the first story, whereas the corresponding
damage indices in the second and third stories were only 0.11 and 0.05,
respectively.
4.4.3 Building C
As mentioned earlier, this building consists of two weak-column-type
frames and a weak-beam-type frarne. The columns in the first story is
relatively brittle ( ~ = 5.5), whereas the beams in the middle frame are
u
very ductile ( ~ > 20). Due to this weak-bearn-type frame, an
u
approximately uniform story drift is achieved beyond yielding under a
static loading. Consequently, the calculated damage is distributed more
uniformly compared to weak-column-type buildings (eGg., bu'i1dings A and
B) as indicated in Table 4.4.
4.4.4 Building D
As explained earlier, two parts of this two-story school building
were designed under different seismic codes. Because of the change in
the building code requirement for ductility, the confinement ratio of
the columns in the older part was only 0.15%, whereas that of the newer
part was raised to 0.6%. The calculated ultimate ductility factor of
the first story is 4.0 for the older part and about 10 for the newer
,part. As indicated in Table 4.5, t ~ e calculated damage index of the
first story (for the two parts) is 0.24. However, if the index was
evaluated separately for each part, the damage index for the older part
would be 0.45, whereas fer the newer part it is only 0.14.
51
4.4.5 Building E
The story-strength of this three-story school building is weakest in
the second story; under a static loading, the second story yields first
at a base shear coefficient of 0.50. The ductility, however, is the
lowest in the first story; i.e., ~ = 4.7 for the first story and
u
~ = 6.1 and 6.9 for the second and third stories, respectively.
u
Consequently, the damage was concentrated in the lower two stories as
indicated in Table 4.6.
4.4.6 Building F
The first story of this three-story building is weaker and more
brittle than the upper two stories as shown in Table 4.7. Consequently,
the damage was concentrated in the first story. The building nearly
collapsed during the earthquake.
4.4.7 Building G
The columns in the first story
building are extremely brittle ( ~
of . this single-bay three-story
= 1 .. 7) mainly because of poor
u
confinement (p = 0.22%). Under a static loading, the first story
w
yields at a base shear coefficient of 0.67, whereas the upper two
stories, supported by shear walls, remain elastic. The calculated
damage index exceeds 1.0 for the first story. For the upper stories,
the calculated values were negligibly low.
52
4.4.8 Building H
This two-story building has high strength and ductility as indicated
in Table 4.9. Accordir.g1y, the calculated damage index is very low
(D ~ 0.02).
T
4.4.9 Building I
The strength and ductility of this three-story school building, as
indicated in Table 4.10, appear to be reasonably high. However, the
columns with a shear span ratio of 1.5 ln the north side frame are
extremely brittle ( ~ ~ 1.0). Although the calculated story damage
u
indices are not high, as indicated. in Table 4.10, the member-level
damage indices of these columns were geater than 1.0. This may
correspond to the shear failure sustained by these columns as shown in
Figure 4.2I-a.
4.5 Definition of Damage
The calculated damage indices of the above buildings and the
corresponding
in Figure 4.5.
results of post-earthquake damage surveys are summarized
The degree of structural damage may be classified into
five categories; namely, slight, minor, moderate, severe, and collapse.
The results shown in Figure 4.5 serve to calibrate the damage measure
and significance of the damage index, D. In light of the results of
this calibration, an overall damage index of D < 0.4 may be considered
T-
to be reparable, whereas buildings with D > 0.4 represents damage
T
beyond repair, and D > 1.0 represents total collapse.
T-
53
CHAPTER 5
DAMAGE-LIMITING ASEISMIC DESIGN
5.1 Objectives of Aseismic Design
The philosophy of earthquake-resistant design generally recognizes
that some level of damage is acceptable against a moderate earthquake
that can be expected over the life of a structure, whereas structural
collapse should be prevented under a severe earthquake defined by a
relatively long recurrence period (e.g., 500 years according to
ATC 3-06, 1978).
Using the results of calibration of the damage index presented in
the previous chapter, the overall damage level of D = 0.4 is a
T
reasonable level for limiting the potential damage 1n the design of
reinforced concrete buildings .. Based on this criterion, a design method
is developed herein to insure a limiting damage of D = 0.4 against a
T
specified load in terms of the characteristic intensity I
c
In aseismic designs of buildings, economical considerations are also
necessary. Past seismic damage of reinforced concrete structures has
shown that building collapse is caused mainly by damage concentration
(especially in the first story) and/or the lack of ductility in the
constituent members. In the 1971 San Fernando earthquake, the main unit
of the Olive View Hospital was severely damaged due to damage
concentration in the lower two stories; whereas, several weak-bearn-type
buildings, designed with the same code provisions, such as the Orion
Avenue Holiday Inn Hotel, sustained only minor damage owing to a more
uniform distribution of damage throughout the building (UoS .. Department
of Commerce, 1973). This would imply that the seismic performance of
buildings can be improved without increasing the strength requirement
and, therefore, without increasing construction cost, if uniform
potential damage distribution is achieved in design.
BOOR
U;n1versity of Il11no16
BI06 NeEL
, 208 N. Romine Street
'U.rba.na.
e
. Illinois 618Gl.
54
Proposed herein is a procedure for design to achieve a uniform
potential damage distribution at a limiting damage level of D = 0.4;
this can be accomplished by properly selecting both the strength and
ductility of the constituent members. The proposed design procedure is
equally as simple as current design methods and in which both economical
efficiency and reliability of design have been taken into consideration.
The proposed design procedure is illustrated for both
weak-column-type and weak-beam-type reinforced concrete buildings of
three to five stories.
5.2 Strength Requirements.
5.2 .. 1 General
The requirement for strength in design is to provide enough lateral
strength and to proportion the strength of each structural component in
order to avoid damage concentration. There are two practical ways for
preventing damage concentration by controlling component strength;
namely,
( i) pinpoint the ideal story strength distribution 1n a
weak-column-type building; or
(ii) insure a weak beam-strong column failure mode at the majority
of the joints ..
The first method is appropriate when the story strength can be made
proportional to the lateral force; whereas in the second method, the
strength of the constituent components should be proportioned so that
the weak-beam-type failure mode is dominant at every story under the
"design lateral forces." Therefore, adequately determining the design
force distribution is essential in either method.
55
5.2.2 Optimal Strength Distribution and Mode Shape
When the response of a multistory structure remains within the
elastic range, the story drift distribution is approximately represented
by the fundamental mode shape. According to the basic theory of
kinetics, this may correspond to the minimization of the following ratio
minimize
where,
E = potential energy; and
E
P
= kinetic energy.
k
(5.1)
When inelastic response is involved, the application of the above
formulation is not immediately obvious. However, according to Akiyama
(1980) and several other researchers, the total
structural system subjected to a specified
sensitive to small changes in the structural
regarded as a constant value. Therefore, the
converted to the following problem,
subject to E = constant
p
potential enel?gy of a
ground motion is not
parameters and can be
above formulation can be
( 5 .. 2)
Introducing the interstory modal vector {.}, the
1.
inters tory
deflection {u } may be expressed as
1.
{u. }
1.
in which,
{. }q
1.
q = response of the equivalent SDF system.
(5.3)
56
Then, both the kinetic energy and potential energy may be expressed as
follows;
E
P
where;
[B]
1 0:- _ ...
1
o!
i
0;
:-1 :
(5.4)
(5.5 )
, a lower triangular matrix
[M] a diagonal mass matrix with m = mass of
i
i-th floor
E.(u.) = potential energy of i-th floor
1 1.
Using the Lagrange's multiplier, Eq. 5.2 can be expressed as,
a
\E ] = 0 (i I,n)
~ [ E k
1.
P
The solution of which is,
aE
[A .. ]{<p.}
. --E
\"{Q.}
),'{a<p.}
1.J 1. 1.
1.
(5.6 )
(5.7)
57
where:
n
A =
L m
ij k=max(i, j) k
Q.
= shear force of i-th story
1-
A' ,:\ " =
arbitrary constants
The above Eq. 5.7 gives a one-to-one relationship between the interstory
mode shape and the applied inters tory force distribution.
For the case of a linear mode shape, the interstory mode { } may be
i
expressed as
{h
Y
h
Y
.}
i+l - ].
1-n which,
h' = the height of the i-th floor from the base.
i
(5.8 )
By substituting Eq. 5.8 into Eq. 5.7, the interstory shear forces Q. and
1-
the lateral forces F become
i.
n
I
w . h ~
j=i
J J
Q.
CBW
]. n
I
w . h ~
j=l
J J
w h
Y
F.
cBw
i i
]. n
L
w . h ~
j=l
J J
where:
W = weight of the i-th floor;
i
C = base shear coefficient; and
B
W = total weight of a building.
(5.9)
(5.10)
58
The above Eq. 5.10 gives the linear lateral force distribution
currently used in seiEmic code provisions. Therefore, the linear
lateral force distributicn is the optimal strength distribution when a
linear mode shape is prescribed. In order to achieve the linear mode
shape near the elastic range as well, the story stiffness should be
proportional to the story shear force; i.e., the same distribution
should be adopted for story strength and story stiffness.
5.2.3 Application to Weak-Beam-Type Buildings
As illustrated in Chapter 3, the near linear mode shape is
automatically achieved in weak-bearn-type buildings because of interstory
couplings. If every frame is designed with weak beam-strong column
joints, the strength requirement may be applied to the whole building
rather than to each story. With the assumption of a linear mode shape,
the required strength is then
IM. = h CBW
1 e
h
e
where:
(5.11)
(5.12)
M = yield moment of beams or half-hinged columns (adjusted for
1.
rigid zone); and
h = equivalent height of building.
e
An additional consideration is needed when and
weak-column-type frames are mixed within a building. The strength of
the constituent members should be proportioned more carefully so that
the weak-beam-type failure mode is dominant at every story.
59
Accordingly, the following restriction may be imposed on every story.
and
Q. < 0.5 Q.
ell
in which;
Q. = applied shear force at the i-th story;
l.
Q. =
sum of strength of weak-column-type frames;
c l.
bQi
=
sum of strength of weak-beam-type frames ..
( 5.13)
(5 .. 14)
and
Equation 5.13 indicates that at least 50% of the applied shear force
1S resisted by the weak-beam-type frames, whereas Eq. 5.14 requires that
the stress in the columns of the weak-column-type frames are within 50%
of their strength .. The restriction of Eq. 5.14 is based on the
observation of laboratory data of reinforced concrete columns, which
generally show that inelastic deformation becomes prominent when the
applied load exceeds 50% of the yield strength (e.g., Short Column
Committee, 1973-77).
When the above restrictions are satisfied, the overall strength
requirement may be expressed as,
I Q. h. + LM.
ell l
(5 .. 15)
60
5.3 Equivalent SDF Concept
5.3.1 Modeling of Buildings
As indicated in Eq. 5.3, the response of MDF buildings may be
represented by that of Hn equivalent SDF system. Since the linear mode
shape is used for both weak-column-type and weak-beam-type buildings,
the derivation of the relevant SDF system is straightforward. Using the
method described in Appendix-D, the equivalent building weight W,
e
equivalent strength Q , and the equivalent response u are expressed as
e e
follows (see Appendix-D for details):.
W
e
Q
e

L, 1 1

1 1
c W
B
h R
e
(5.16)
( 5.17)
(5 .. 18)
in which, R is the rotational deformation of the building corresponding
to the linear mode shape ..
Combining Eqs. 5.16 and 5.17, the effective strength of the SDF
system (in terms of g) is
C
e
W x
L 1 1
2
C
B
(5.19)
Observe that the overall strength is generally increased from that of
the static strength C
B
(e.g., when both the floor weight and interstory
height are uniform, C = 1.167 C
e B
for a 3-story building, whereas
C = 1.222 C for a 5-story building).
B e
The corresponding elastic stiffness of the SDF system may be
determined from the fundamental natural period T. Among a number of
61
approximate methods for estimating fundamental natural periods, the
so-called "gravitational equation" may be one of the simplest and
practical (Muto, 1963); i.e.,
T = iu/3.68
p
(5.20)
in which, u = top lc.teral deformation of a building (in inches) under
uniform l a t ~ r a l static load o ~ 1.0g.
5.3.2 Equivalent Sh2rt Duration Excitation
According to the definition of the damage index, structural damage
is expressed as a combination of the maximum deformation and hysteretic
energy dissipated in a structure. However, the damage analysis 1n
Chapter 3 has shown that the damage caused by cyclic loading effect is
prominent only when the duration of the ground motion is relatively
long; under a very short-duration excitation, the damage would be
primarily due to excessive deformation. This may suggest that the
structural damage is expressed in terms of the maximum deformation under
a short-duration excitation with an equivalent intensity.
In this study, a strong motion duration of t = 2T 1S selected to
o
represent a "short duration." By using the definition of the
characteristic intensity,
acce I eration a is
G
the corresponding rIDS strong-phase
(5.21)
Therefore, the overall structural damage of a building can be
approximately represented by the equivalent SDF system subjected to an
"equivalent short-duration excitation."
62
5.3.3 Response of SDF Systems
In order to specify the force-deformation relationship of the
equivalent SDF system, an appropriate multilinear envelope curve and
normalized hysteretic energy per cycle are needed. A study of several
reinforced concrete frames has shown that when the entire structure is
modeled as a SDF system, the resultant deformation shape does not vary
much from building to building. In Figure 5.1 ~ s shown a typical
force-deformation relationship that has been obtained by averaging those
of several reinforced concrete buildings analyzed in Chapters 3 and 4.
By assigning several discrete values to the basic parameters T,
T (predominant period of excitation), I and C , the responses of 180
G c e
SDF systems were calculated. The ranges of the parameter values are as
follows:
I
=
0.1
-
0.2 ;
c
T
= 0.2 - 5 .. 0 (sec .. );
T
=
0.4 - 1 .. 0 (sec.);
G
C
=
0 .. 1 - 0 .. 6 ; and
e
l;G
=
0 .. 9.
Multiple regression analysis of the results yielded the. following:
u (inch)
e
40 I T
1
.
5
c
0.67-0.42 In(T/T
G
)
C
e
(5.22)
A comparison of the simulated maximum response values and those obtained
with Eq. 5.22 is shown in Figure 5.2. The relevant modeling error in
terms of the c.o.v. is 0.19. The proposed approximation tends to
underestimate the elastic response and overestimate the inelastic
response. However, for moderate response values, say 2 to 10 inches,
which are the allowable response of 'low-rise and medium-rise reinforced
concrete buildings, the prediction error is relatively small
(c.o.v. :::: 0.1).
63
5.4 Ductility Requirements
5.4.1 Ductility Requirement of Components
Combining Eqs. 5.18 and 5.22, the overall rotational deformation R
of a building may be expressed as follows:
The value
40 I T
I
.
5
c
0.67-0.42 In(T/T
G
)
h C
e e
of R may be called
d
constituent components are required
the above rotational deformation.
(5.23)
the "required ductility index.
"
The
to have enough duct i I ity Ito sustain
Since the target is to achieve a
uniform damage distribution at a limiting damage level of D = 0.4, the
relevant potential damage of the constituent components should also
satisfy this limit of D = 0.4. By eliminating the hysteretic energy
term (under equivalent short duration excitation), the damage index for
each component should be
D 0.4 (5 .. 24)
This may be expressed in terms of the allowable rotational angle of a
member as,
R.
1.
0.40
u
(5.25)
l.n which, ~ = the shear span length. Therefore, the potential ductility
of the constituent member is expressed by the value of R
i
The study on the inelastic behavior of reinforced concrete
components in Chapter 2 has shown that the ductility is a function of
the following structural parameters;
64
i/d = shear span ratio;
p
= confinement ratio in percent;
w
p = normalized steel content
= a cr /bdf';
t y c
n
= normalized axial stress
=
N/bdf' ; and
0 c
f'
= concrete strength, in ksi.
c
Larger values of i/d, p , and f' will increase ductility, whereas higher
values of p and n
o
w c
will decrease ductility.
A parametric study was conducted using the results of about
different combination of parameter values for reinforced
45,000
concrete
components. A mUltiple regression analysis of the results yielded the
following;
R. (%)
l
0.4 (l/d)n
1n which,
n = 1.2 + O.lf' + 0.43p
c w
3.7:n
o
(1.5 + 0 .. 35p
w
4.2n )p
o
(5.26)
When the value of the confinement ratio p exceeds 2%, p should be
w w
replaced by 2%. The applicable range of the above relationship is
1.0 < i/d < 7.0;
0.1 ~ p < 2.0;
w
0.05 ~ p ~ 0.5;
0.0 ~ n < 0.55; and,
o
2 .. 5 < f' < 5,,0.
c
Comparison with experimental data is shown in Figure 5.3 for 261 cyclic
test specimens. In evaluating the values of R for these specimens,
1
the ultimate deformation under static loading was calculated using the
definition of the damage index. The relevant scatter gives a c.o.v. of
0.54 (assuming a lognormal distribution), which is almost the same as
the CeO.V. of the damage index.
65
The application of Eq. 5.26 to design is straightforward since four
parameters 'lId, p'" nand f' have already been determined in the
o c
strength design. The confinement ratio P is the only parameter that
w
is governed by ductility requirements. However, there is a saturation
point for the confinement ratio at P = 2%; beyond this value, the
w
increase of P
may not improve ductility (see the definition of the
w
ultimate deformation Q in Chapter 2). This imposes a restriction that
u
must be considered in designing the constituent members for strength
requirement.
5.4.2 Ductility Requirement of Building
Completely uniform potential damage distribution
perfectly uniform ductility distribution) among
(and,
the
therefore,
constituent
components is difficult to achieve in design; however, uniform potential
damage distribution among the stories should be achieved for
weak-column-type buildings. Using the same formulation as the overall
damage index D (see Eq. 3.41), the overall potential ductility of each
T
story (assuming a linear mode shape) may be expressed as a "ductility
index" as follows:
LQ.R.
1 1
LQ.
1
Accordingly, the requirement for ductility of a building is
R > R
T - d
where:
R = minimum ductility as defined in Eq. 5.23;
d
Q. = strength of the constituent columns; and
1-
(5.27)
R = allowable rotation angle of the columns (adjusted for rigid
zone)
66
For the case of weak-bearn-type buildings, the ductility requirement may
be applied to the whole building as,
(5.28)
5.5 Design Procedure
Earthquake-resistant design is usually initiated by the selection of
a design seismic load. The regional potential seismic activity, the
desired aseismic protection level, as well as
underlying the design process should be duely
the uncertainties
reflected in such a
selection. In Chapter 6, some of these issues are addressed; the design
seismic load (in terms of the characteristic intensity) is determined as
a function of the acceptable risk. In general, a tr'ade-off between
economy and safety is required in the selection of the design seismic
load.
A damage-limiting
procedure is divided
design
into
procedure is suggested herein. The
four consecutive steps each consisting of
three items as shown in Figure 5.4.
follows ..
Each step may be described as
5.5.1 Design Setup
A significant feature of the proposed design method is that the base
shear coefficient C is regarded as a design parameter rather than a
B
design constraint. This does not contradict present code provisions in
which only the minimum lateral strength requirement is specified (e .. g. ,
the SEAOC Code) Thus, the designer may select a
"strong-but-not-so-ductile design" or a "ductile-but-not-so-strong
design .. " In reality, however, the designer may be obliged to adopt the
67
former design with a weak-column-type building, and the latter design
with a weak-beam-type building.
5.5.2 Strength Design
The dimension and axial are determined at this stage
according to the design base shear coefficient and the intended failure
mechanism. Due consideration for the ductility of the constituent
components is necessary in order to avoid repeated modifications of
strength design. As mentioned earlier, there is a limitation to the
range of ductility design. Beyond the
improvement in ductility is achieved
confinement ratio of 2%, no
by increasing the web
reinforcement. By sUbstituting p = 2% in Eq. 5.25, the maximum value
w
of the allowable rotational angle of a member is,
R. (%)
1
2.06+0.1f'-3.7n -(2.2-4.2n )p
O.4(./d) coo
(5 .. 29)
A tentative evaluation of the ductility requirement may be carried out
using Eq. 5.29.
5 .. 5 .. 3 Structural Analysis.
The detailed design is made under service loads at this step. In
particular, the elastic lateral deformation is calculated for estimating
the fundamental natural period using a conventional method of structural
analysis.
5.5.4 Ductility
Some amount of web reinforcement may have already been provided to
resist the shear stress under service loads. Additional web
68
reinforcement is provided here to meet the ductility requirement given
in Eqs. 5.27 and 5.28. As there is a limit to the control of ductility
by means of confinement, the relevant ductility requirement mayor may
not have been satisfied.
If the ductility is only slightly insufficient, the strength design
may be modified (feedback A in Figure 5.4) since the potential ductility
of the constituent members is a function of the structural parameters
tid, p and n. However, if the insufficiency of ductility is large, the
o
designer has to increase the base shear coefficient C (feedback B in
B
Figure 5.4), thus decreasing the required ductility index R
d
5.6 Design Examples
Four 3-bay reinforced concrete frames are designed to illustrate the
proposed design procedure. The dimensions of the frames are shown in
Figure 5.5. Frames A and Bare weak-column-type frames, whereas
Frames" A' and B' are intended to be weak-beam-type structures. All
frames have a uniform story weight of 420 kips and uniform inters tory
height of 12 feet.
A design seismic load of characteristic intensity I = Oe1s 18
C
assumed for all the frames. According to the relationship between the
characteristic intensity and the peak acceleration evaluated from 140
acceleration records in the California area, a characteristic intensity
of 0.15 corresponds to a peak acceleration of about 0.4g (see
Chapter 4). The predominant period of ground motion is assumed (firm
soil condition) to be T" = 0.4 sec. Nominal concrete strength of
G
f' = 3 ksi and reinforcing steel yield strength of a = 50 ksi are
c y
prescribed.
71
T = 11.69/3.68 = 0 .. 35 sec.
The corresponding required ductility
selected value of confinement ratio p ,
R
w
was calculated for each column
i
Eq .. 5 .. 27 , the ductility requirement
yielding:
index is R = 0.77%. With a
d
the allowable rotational angle
as shown in Table 5.3 .. Using
was checked for each story,
Story 1 R = (2x84xl.057+2xl05xO.559)/(2x84+2xl05) 0.778 > 0.77
Story 2 R = (2x71 .. 4xO.985+2x86.5xO.606)/(2x71.4+2x86.5) = 0.78 > 0.77
Story 3 R = (2x44xl.04+2x51xO.75)/(2x44+2x51) = 0.883 > 0.77
The results are also summarized in Table 5.3.
5.6.2 Design of Frame B
Frame B is a weak-column-type 5-story 3-bay f r a m e ~ For this frame,
a base shear coefficient of C = 0.4 is selected. The lateral force F
B i
and story shear Q are summarized in Table 5.4. The results of the
i
strength design is shown in Table 5.5. The calculated fundamental
natural period is
ductility index is R
d
shown in Table 5.6.
0.486 seconds, and the corresponding required
= 0.572%. The results of the ductility design are
5.6.3 Design of Frame A'
Frame A' 18 a weak-beam-type 3-story 3-bay frame. The frame is
designed so that failures in the beams are dominant at every joint
except the flexural failure of the column bases in the first story.
Therefore, the components expected to be subjected to inelastic
deformations are beams Bl and B2, and columns Cl and C2 of the first
story (see Figure 5.5). Because of the relatively ductile behavior of
70
The normalized axial stress n = N/bdf' are shown in Table 5 .. 2. The
o c
tension reinforcement ratio is assumed to be 1%, 1.15%, and 1.1% for the
first, second, and third stories, respectively.
normalized steel ratios pare:
Story 1
Story 2
Story 3
p = p cr If' = 0.0Ix50/3 = 0.167
t Y c
= 0.0115x50/3 = 0.192
= 0.011x50/3 = 0.183
Therefore, the
Yield moments, M , were calculated for the columns using the method
described in yielding the following:
Exteria Columns
Interia Columns
Story 1
4370
5470
eM
y
Story 2
3720 2290
4500 2650
in kips-inch)
The story strength are calculated as
Story 1
Story 2
Story
v = l:2M /Q,
Y
= 2x2x547 0/ 1 04+2x2x43 70/104 378 kips
= 2x2x4500/104+2x2x3720/104 = 315.8 kips
= 2x2x2650/104+2x2x2290/104 = 190 kips
The results are summarized in Table 5.2. The next step is to calculate
the natural peoriod T. The top lateral deformation, u , of the building
under a uniform lateral load of Ig is calculated as follows:
Story 1 1260 kips
/
1820 kips/in = 0.69 in
Story 2 840 /
1483 = 0 .. 56
STory 3 420
/ 950 = 0.44
Total (u )
:::
1.69 in
p
Then, the natural period is
73
R = (2234.5x12x4.13+1308.x6x6.45+3422.x2x3.80+4756.x2x1.52)/51018.
T
= 3.95 > 3.53 - ~ O.K.
5.6.4 Design of Frarne B'
Frame B' is also a weak-bearn-type 5-story 3-bay frame. The intended
failure mode is the same as that of Frame A'. A base shear coefficient
of C = 0.1 is selected for this case. The required strength is,
B
~ M > hew = 524xO.1x2100 = 110880 (kips-in)
i - e B
The results of the strength design are shown in Table 5.9.
~ M = 2886x24+1776x6+6458x2+9033.2x2 = 110902.4 > hew - ~ O.K.
i e B
The calculated fundamental natural period is 1.93 seconds, and the
corresponding required ductility index is R = 3.03%. The results of
d
the ductility design are summarized in Table 5.10. In this case, the
overall ductility index of the designed frame is
R = (2886x24x3.36+1776x6x5.11+6458x2x3.61+9033.2x2xl.29
T
= 3.22 > 3.03 - ~ O.K.
5.6.5 Examination of Designed Frames
Damage analyses were performed on the designed frames. The
strong-motion durations t 5 sec. and t = 15 sec. were used. The
o 0
corresponding values of rIDS acceleration are O.165g and O.1145g,
respectively. The calculated story drifts (as a ratio to the interstory
height) and damage indices are shown in Figure 5.6. A nearly uniform
distribution of the story drifts as well as damage indices can be
72
the beams, the "ductil e-but-not-so-strong design" is adopted. In this
regard, a base shear coefficient of C = 0.12 is selected. The required
B
strength is
~ M > h C W = 336x(:.12x1260 = 50803.2 (kips-in)
1. - e B
The dimensions of the com1'onents w e ~ e determined as shown in Table 5.7 ..
The tension reinforcemeltt ratio is assumed to be 0.5% for the columns
and 1.2% for the beams. The normalized steel ratio p is determined as
follows:
Columns
Beams
0.005x50/3 = p :: p cr / f' =
t Y c
= 0.012x50/3 =
0 .. 083
0.2
The yield moments were calculated using the method described in
Appendix-A; yielding,
Beam B1 M = 2050 kips-in
Beam B2
MY
= 1200
Column C1
MY
= 4100
Column C2
MY
= 2950
Y
A moment diagram was drawn to assure the weak beam-strong column
condition at the joints and to determine the point of inflection for the
columns. Accordingly, the shear span ratio was calculated and the yield
moment was adjusted for rigid zone as summarized in Table 5.7. The
strength capacity is, therefore,
~ M = 2234.5x12+1308.x6+3422.x2+4756.x2 = 51018.0 > heW - ~ O.K.
i e B
The calculated natural period is T = 1.24 seconds, and the corresponding
required ductility index is R = 3.53%.
d
design are shown in Table 5.8.
The results of the ductility
74
observed for the weak-column-type frames (i.e., Frames A and B). The
calculated overall damage' indices D are shown in Table 5.11.
T
5.7 Limitation
The proposed design pl'ocedure is limited to buildings up to seven
stories. For such build:ngs, a near uniform damage distribution within
a limiting value of D = 0.4 can be obtained for both.weak-beam-type and
weak-column-type buildings. However, for highrise buildings, e.g.,
higher than ten stories, the deviation of the story damage from the
tolerable level of D = 0.4 could be unacceptably large; this may be
attributed to the contribution of the higher modes to the building
response. For highrise buildings, a more complete analysis of the
structural response is needed. and should be incorporated in the strength
and ductility requirements.
75
Metz Ratera
YnlverSlt neg
y o:f Il
208 BI06 NeEL 11.no1.s
U N. RO
m
l
Illi ne Street
1J.o 18. .618QJ:iJ
CHAPTER 6
RELIABILITY OF PROPOSED ASEISMIC DESIGN
6.1 Introductory Remarks
Occurrence of earthquakes .is unpredictable both in time and space;
moreover, the structural response under a specified intensity cannot be
predicted precisely. Structural engineers, therefore, have to make
critical decisions, as the selection of design earthquake and
aseismic protection. level, under condition of uncertainty.
The objective of this chapter is to examine the reliability of the
earthquake-resistant design method developed in the previous chapters.
In the proposed procedure, the potential damage is limited to a
tolerable damage level of D = 0.4 under a specified design seismic load.
The same structure, however, may collapse under a more severe load. In
light of the unpredictable nature of earthquakes and the uncertainties
in the design process, the structure may collapse even under the design
earthquake. Uncertainties may arise from three sources; namely,
uncertainty in the ground excitation (amplitude, duration, frequency
content, and inherent randomness in the time history), variability of
structural capacity (e.g., energy absorbing capacity), and modeling
error in structural analysis. Based on a systematic evaluation of these
uncertainties, the collapse probability associated with a structure
designed according to the proposed met,hod is examined as a function of
the design load (in terms of characteristic intensity I ).
c
6.2 Uncertainty in Structural Response
The variability in the damage index is given in Eq. 3.3 in terms of
the variance of D. This consists of the uncertainty in the structural
capacity and the randomness in the calculated maximum deformation and
76
absorbed hysteretic energy. In addition to this variability, the
response variability or uncertainty associated with modeling errors is
examined here ..
In the damage analysis, structures are reduced to lumped
mass-discrete spring systems having a simplified load-deformation
relationship; moreover, several parameters necessary for response
calculations, such as the viscous damping and filter parameters of the
input excitations, cannot be determined precisely. The uncertainty in
the response associated with these imperfections is not negligible and
should be included in the overall uncertainty (Sues, Wen, and Ang,
1983) ..
The variance of the response value of D due to the model parameter
uncertainties may be obtained by the following approximation (Ang and
Tang, 1984),
'\ '\ aD alD I
L L (-,,-I )(-,,- )a.a.p ..
.. op. II op. 11 1 J 1J
1J 1 J
where:
P. = model parameter;
1
aD I
~ l l
1
a
i
p
ij
= partial derivative of response value of D evaluated at
mean values of the model parameters;
=
=
standard deviation of parameter p.; and
1
coefficient of correlation between p. and p ..
1 j
(6 .. 1)
the
The model parameter uncertainties are evaluated below based on available
data and information.
77
6.2.1 Stiffness and Strength
The uncertainties in the stiffness, K, and strength, Q, consist of
the inherent material variability expressed in terms of the c.o.v. 0,
and the modeling error, expressed in terms of c.o.v. 1:..
A comprehensive study was conducted by Tichy and Vorlicek (1972) for
evaluating the inherent variability in stiffness and ultimate strength
of reinforced concrete components. The main findings are: (i) the
degree of scatter in the component strength is largely a function of the
axial stress, and insensitive to other factors such as
reinforcement ratio, (ii) the value of a(c.o.v.) of the component
strength ranges from 0.1 to 0.15 for beams, whereas it ranges from 0.1
to 0.3 for columns, (iii) the variability of stiffness due to material
uncertainty is relatively small (a 0.1). Based on this study,
o = 0.20 and 0.10 are assumed for strength and stiffness, respectively
. To evaluate the modeling errors in both stiffness and strength, 260
reinforced concrete component test data were examined. The results are
shown in Figure 6.1. The secant stiffness at yielding, K, and the
ultimate shear strength
described in Chapter 2. As
are evalua ted using h
. X 1 h d
t e met 0
shown in Figure 6.1, the stiffness and
strength are partially correlated with a correlation coefficient of
0.42. Based on these results, the modeling errors are, respectively,
the total uncertainty in the stiffnes.s
I:.
=
0 .. 29 and I:. = 0 .. 12, giving
K
Q
of n =
Ic 2 2
= 0.31, 0 .. 1 +0.29
K
whereas for strength,
10 .2
2
+0 .. 12
2
n = = 0.23 ..
Q
6.2.2 Damping
The viscous damping of structures under earthquake motions is
believed to be mainly contributed by the energy dissipation of the soil
and the friction between the foundation and the soil (e.g., Kanai, et
al., 1962). Figure 6.2 shows observed damping coefficients of the
fundmnental mode in actual reinforced concrete buildings (Tanaka, et
78
ale , 1968) These 'data indicate that the damping coefficient is
approximately inversely proportional to the fundamental natural period
of buildings. The large scatter observed in the damping coefficient may
be attributed to the variability of the soil property (Kanai, et al.,
1962). Based on the data in Figure 6.2, a c.o.v. of n = 0.52 ~ s
h
assumed in the damping.
6.2.3 Predominant Frequency of Excitation w
G
The frequency contpnt of the ground motion may be estimated based on
the soil property of surface layer.Lai (1982) examined the frequency
content for "soft soil" and "hard soil" conditions by fitting 70
earthquake records to the Kanai-Tajimi spectrum; the coefficient of
variation for the predominant frequency,
information is also incorporated in
damage analysis.
6.2.4 Hysteretic Energy per Cycle
W
G
' was about 0.44. This
the overall uncertainty in the
-The uncertainty in the hysteretic energy per cycle has already been
examJned in Chapter 2; a c.o.v. of 0 .. 25 was obtained through the
analysis of about 800 h y s ~ e r e s i s loops.
6.2 .. 5 Mass
Based on the study by Portillo and Ang (1976), the uncertainty in
estimating the mass of buildings is assumed to be c.o.v. = 0.12.
The above results are sUmmarized in Table 6.1. For estimating the
overall variabil ity of the damage index D due to mode 1 ing error as well
as the contribution of each model parameter, three SDF systems were
79
analyzed. Only the natural period is chosen as a variable structural
parameter, i. eo , T
= 0 .. 4, 0 .. 8 and 1 .. 2 seconds for SDF systems A, B and
. C,
respectively .. Other parameters are: <5
=
5.0 <5 , S = 0 .. 2, the
u y
strong-motion duration is 10 sec .. , the filter parameters of the ground
motions are w
= 51T (T
= 0.4 sec .. ) and C;
= 0.9
G G G
Using Eq. 6.1,the variability of D at various damage levels were
evaluated in terms of the c.o.v. as shown in Figure 6.3 .. In this
figure, the contribution of each model parameter is also expressed in
terms of the c.o.v .. ; i.e.,
1 aD I
I dp. l.l
1
Qi
x -=-
D
in which,.Q is the coo.v. of the model parameter p .
i 1
(6 .. 2)
As observed in the above results, the contribution of the frequency
of excitation, w, is dominant over the other parameters. When the
G
natural frequency of a structure is close to w (system A), the c.o.v.
G
of D tends to fluctuate at different damage levels. However, as the
natural frequency deviates from w , the overall variability of D (in
G
terms of the coo.v.) tends to converge to a constant value of about 0.6
(see Figure 6 .. 3).
6.3 Reliability of Design
In the proposed design scheme, the potential building damage is
limited to D ~ 0.4 against a specified design load, in terms of I
c,des
Correspondingly, the risk associated with such a design may be expressed
in terms of the probability of potential building collapse. p CD > 1.0),
F -
under different mUltiples of I "
c,des
In the ordinary scheme of evaluating the fragility of structures,
the inherent variability and the variability due to modeling error are
simply combined to obtain the overall uncertainty in structural
responses. However, the modeling error, being largely dependent on the
designer's skill as well as the quality. of available information, should
80
be treated differently from the inherent variability in evaluating the
potential collapse probability. In this light, the effect of modeling
error may be represented with the range of calculation error in the
probability PF' whereas the inherent variability is. represented in the
calculated probability PF; this may be achieved by regarding the
modeling error as the representation of the randomness of the mean value
of the estimated damage response, n. In this scheme, the relevant
probability of building collapse PF is also random variable since the
PF is implicitly a function of D. For a practical use, such a
variability in the estimated probability may be expressed in terms of
the q-percentile value of the probability (Ang and Tang, 1984); namely,
q
(6.3a)
or
(PF) q
(6.3b)
in which:
F (p) = the CDF of the estimated probability of collapse; and
p F
F
(p) = the probability of collapse at the q-percentile level.
F q
As shown in Figure 6.3, the degree of variability due to modeling
error (in terms of the c.o.v.) can be regarded as a constant value
except when the fundamental natural period of the structural system, T,
is very close to the predominant period of excitation, T 0 Based on
G
this observation, a c.o.v. of 0.60 is assumed for the variability of D,
whereas a c.o.v. of 0.50 is assumed for the inherent variability in D
(see Figure 3.19). The probability of building collapse may be
-
evaluated for a given load, I, assuming that D and D are both
c
lognormally distributed. Figure 6.4 gives the probability of collapse,
p{D > 1.0), as a function of the characteristic intensity of the ground
81
motion (given as multiples of the design intensity I ). At q = 50%,
c,des
it can be observed that at the design earthquake intensity I , the
c,des
collapse is less than 1%.
82
CHAPTER 7
SUMMARY AND CONCLUSIONS
A seismic damage model for reinforced concrete structure is
proposed. The model is based on the premise that the total structural
damage from earthquake motions is composed of the damage caused by the
maximum structural and that associated with the absorbed
hysteretic energy. Damage is expressed in terms of a damage index, D,
with D 1.0 representing total collapse. Available monotonic and
cyclic test data of reinforced concrete beams and columns were analyzed
to evaluate the statistics of the appropriate parameters of the proposed
damage model. The lognormal distribution was found to be appropriate
for describing the capacity (in terms of D) of reinforced concrete
componen t s ..
Based on the component-level damage model, a method for stochastic
assessment of building damage was developed. In the proposed method,
the uncertainty in the vltimate structural capacity and the randomness
in structural responses are systematically incorporated. In modeling
IMDF buildings, the conventional shear beam model is modified to include
the coupling between stories such as in weak-bearn-type buildings.
A comprehensive damage analysis was performed on various SDF MDF
reinforced concrete buildings for examining the ground motion parameters
pertinent to structural damage. On the basis of this parametric study,
the destructiveness of ground motions is expressed as a function of the
rms acceleration and the strong-motion duration; accordingly, the
"characteristic intensity" is proposed as a damage-consistent intensity
scale of earthquake ground motions.
Damage indices were calculated for nine reinforced concrete
buildings, that were moderately or severely damaged during the 1971 San
Fernando earthquake and the 1978 Miyagiken-Oki earthquake in Japan. The
results were then correlated with the corresponding observed damages;
this correlation (or calibration) suggested that the overall damage
83
index DT ~ 0.4 represents reparable damage., whereas DT > 0.4 represents
damage beyond repair, and D ~ 1.0 represents total collapse.
T
Based on this cal ibration, a criterion for "damage-limiting design"
is proposed; namely, the potential structural damage may be tolerated
within the reparable range (i.e., D ~ 0.4) against an expected maximum
ground motion during the life of a building. Accordingly, a detailed
aseismic design procedure is developed to insure D ~ 0.4. The proposed
procedure consists of the strength design and ductility design. In the
strength design, the dimensions and axial reinforcement of reinforced
concrete components are determined based on a selected base shear
coefficient and an intended failure mechanism, i.e., weak-beam-type or
weak-column-type failure mode. Following the strength design, the
ductility of components is checked and the amount of web reinforcement
is determined. The structural deformation, expected under the design
ground motion, is expressed as a function of design parameters such as
the .ground motion intensity and the base shear coefficient. In the
ductility design, a simple scheme is provided to limit the potential
damage at D = 0.4.
The applicability of the proposed design method to buildings up to
seven stories was illustrated and confirmed. For such buildings, a near
uniform damage distribution within a limiting value of D = 0.4 is
obtained for both weak-beam-type and weak-column-type buildings.
However
j
for highrise buildings, e.g., higher than ten stories, the
deviation of the story damage from the limiting damage level of D = 0.4
may be unacceptably large, caused mainly by the contribution of the
higher modes to the structural response.
Based on a systematic evaluation of the major uncertainties, the
collapse probability associated with a structure designed according to
the proposed procedure was evaluated as a function of the characteristic
intensity of the ground motion. It was observed that at the design
intensity of the ground motion, the collapse probability is of the order
-3
of 10
84
TABLES
85
Table 3.1, Excitation Level and Damage Index
t
=
5 sec. t
=
15 sec.
0 0
G(g)
0.079 0.126 0.165 0.200 0.055 0.087 0.115 0.126
SA
D 0.266 ' 0.540 0.870 1.250 0.266 0.489 0.874 1.050
max
DT
0.209 0.416 0.655 0.941 0.208 0.375 0.668 0.794
G (g)
0.109 0.172 0.226 0.075 0.120 0.157
SB
D 0.366 0.626 0.961 0.326 0.576 0.827
max
D
T
, 0.181 0.310 0.469 0.178 0.324 0.473
G(g)
0.122 0.193 0.253 0.111 0.176
SC
D 0.316 0.594 0.937 0.423 0.814
max
DT
0.196 0.379 0.601 0.275 0.570
Table 4.1 Data and Results of Damage Analyses
.. _- .
*
Building A B C D E F G
Number of Stories 6 3 4 2 3 3 3
Natural Period (sec.) 0.56 0.23 0.34 0.15 0.20 0.30 0.15
Strength (g) 0.22 0.48 0,.33 0.59 0.50 0.46 0.67
llu
4.0 5.5 5.5 9.7 4.7 7.5 1.7
Characteristic Intensity 0.31 0.23 0.12 0.19 0.19 0.23 0.23
Overall Damage Index 1.47 1 .. 05 0.48 0.22 0.39 0.85 1.25
* Building C is a weak beam-strong column type structure
H I
2 3
0 .. 13 0.18
0.71 0.63
7.9 7.4
0.12 0.23
0.02 0.27 .
00
(j\
Story
G
1
2
3
4
5
Story
1
2
3
87
Table 4.2 Results of Building-A
Strength (g)
llu
A.
1-
0.22 4.06 0.70
0.28 3.96 0.29
0.86 4.14 0.01
0.95 8.50 0.00
1.43 10.0 0.00
2.06 10.0 0.00
Table 4.3 Results of Building-B
Strength (g)
0.48
0.81
1.19
5.51
6.81
7.50
A.
1.
0.91
0.07
0.02
D.
1-
1 .69
0.95
0.03
0.01
0.00
0.00
D.
1.
1.14
0.11
0.05
Story
1
2
3
4
p
Story
1
2
88
Table 4.4 Results of Building-C
(Weak-Column-Type Frames) (Weak-Beam-Type
A. D. A.
l l l
0.24 0.89 0.04
0.22 0 ~ 5 2 0.06
0.18 0.41 0.08
0.11 0 .. 24 0.04
0.03
Tab]e 4.5 Results of Building-D
A.
D.
l
Frames)
D.
l
0.39
0.17
0.21
0.13
0.10
Strength (g)
l
Entire Bldg. Old Part New Part
0.59 9 .. 72 0.90 0.24 0 .. 45 0.14
0.91 11.0 0.10 0.05 0.09 0 .. 02
Story
1
2
3
Story
1
2
3
Story
1
2
3
89
Table 4.6 Results of Building-E
Strength (g)
0.51 4.72
0.50 6.11
0.81 6 .91
A.
l
0.52
0.43
0.05
Table 4.7 Results of Building-F
Strength (g) A.
1
0 ~ 4 6 7 .5 0.76
0.61 7 .7 0.20
0.81 8.6 0 .. 04
Table 4.8 Results of Building-G
Strength (g)
0.67 1.7
1.20 2.6
1.65 3.5
A.
1
1.00
0.00
0 .. 00
D.
~
0.45
0.35
0.09
D.
~
0.97
0.51
0.10
D.
1
1.25
0.00
0.00
Story
1
2
Story
1
2
3
90
Table 4.9 Results of Building-H
Strength (g)
0.71 7 .. 92
0.81 14.4
A.
1
0.67
0.33
Table 4.10 Results of Building-I
Strength (g) A.
~ u
1
0.63 7.41 0.57
0.67 9 .. 48 0 .. 39
1.14 8.30 0 .. 04
D.
1
0 .. 03
0.01
D.
1
(Story) (Member)
0.33 1.08
0.23 0.95
0.01 0.29
91
Table 5.1 Lateral Forces and Interstory Shears of Frame A
Story Q.
.1
3 189 189
2 126 315 (in kips)
1 63 378
Table 5.2 Strength Design of Frame A
Component Dimension 9-/d
* "iV.
n p
V.
0 1..
1..
Cl 0.9 44.0
Story 3 17"x17
H
3.45 0.183 190 .. 0
C2 0.18 51.0
C1 0.14 71.4
Story 2 19"x19" 3 .06 0.192 315.8
C2 0.29 86.5
C1 0.19 84.0
Story 1 20
iii
x20
ii
2.89 0.167 378 .. 0
C2 0.39 105.0
* V. = yield strength in kips (adjusted for rigid zone of columns)
1..
92
Table 5.3 Ductility Design of Frame A
Component
P (%)
.*
Check Rr > Rd
w
Ri . (%)
..
C1 1.04
Story 3 0.20 0.883 > 0 .. 77 - ~ O.K ..
C2
0" 7 5
C1 0 .. 985
Story 2 0.80 0" 7 80 > 0 .. 77 - ~ O .. K.
C2 0.606
C1 1.057
Story 1 1.46 O. ii 8 >O .. ii - ~ O.K.
C2 0.559
* adjusted for rigid zone of columns
Table 5.4 Lateral Forces and Interstory Shears of Frame B
Story F. Q.
1 . 1
5 280 280
4 224 504
3 168 672 (in kips)
2 112 784
1 56 840
93
Table 5.5 Strength Design of Frame 13
Component Dimension
tid n p v. r.v .
0 1 1
C1 0.09 65.9
Story 5 17 "x17 " 3.47 0.308 280.8
C2 0.18 74.5
C1 0.14 118.7
Story 4 19"x19" 3.06 0 .. 366 505.4
C2 0.29 134.0
C1 0 .. 19
154.8 ..
Story 3 20 .. 5"x20.5" 2.81 0.362 672.4
C2 0.37 181.4
C1 0.22 181.4
. Story
2 21.5"x21.5" 2.67 0.342 785.8
C2 0.44 211.5
C1 0.26 193.4
Story 1 22"x22" 2 .. 60 0.325 842.8
C2 0.53 228.0
94
Table 5.6 Ductility Design of Frame B
Component
P (%)
w
R. (%)
l
Check Rr > Rd
Cl 0.869
Story 5 0.20 0.760 > 0.572 -;> O .. K.
C2 0 .. 664
C1 .0.685
Story 4 0.41 0.574 > 0.572 -;> O.K.
C2 0.476
C1 0.693
Story 3 1.00 0.573 > 0.572 -;> O.K.
C2 0.463
C1 0.731
Story 2 1.47 0.578 > 0 .. 57 2 -;> 0 .. K ..
C2 0.448
C1 0.762
Story 1 1.93 0.576 > 0.572 -;> O.K.
C2 0 .. 417
Table 5.7 Strength Design of Frame A'
Dimension 2/d * Component n p M.
0 l
B1 20"x12" 6.11 0.0 0.2 2234.5
B2 16"x12" 7.86 0.0 0.2 1308 .. 0
C1 20"x20" 6.11 0.19 0.083 3422.0
C2 20"x20" 6.11 0.39 0.083 4756 .0
* M. is the yield moment in kips in
l
(adjusted for rigid zone)
95
Table 5.8 Ductility Design of Frame A'
*
Component
p (%) R. (%)
w 1
Bl 0.38 4.13
B2 0.53 6.45
Cl 1.50 3.80
C2 2.00 1.52
* adjusted for rigid zone
Table 5 .. 9 Strength Design of Frame B'
Component Dimension tid n p M.
0 1
Bl 20"x1211 6.11 0 .. 0 0.255 2886.0
B2 16"x12" 7.86 0 .. 0 0.255 1776.0
Cl 24"x24n 6 .68 0.22 0 .. 083 6458.0
C2 24"x24" 6 .. 68 0 .. 44 0 .. 083 9033 .. 2
96
Table 5.10 Ductility Design of Frame B'
Component p (%)
w
R. (%)
1
B1 0.38 3.36
B2 0.53 5.11
C1 1.50 3.61
C2 2.00 1.29
Table 5.11 Overall Damage Index
G
t Frame A Frame B Frame A'. Frame B'
0
0 .. 165g 5 sec. 0 .. 354 0 .. 408 0.382 0.427
0 .. 115g 15 sec. 0.347 0.411 0.398 0.416
97
. Table 6 .. 1 Model Parameter Uncertainties
Model Parameter Parameter Uncertainty (c.o.v.)
Stiffness, K 0.31
. Strength, Q 0.23
Damping, h 0.52
Predominant Frequency, wG
Hysteretic Energy per Cycle, E
Mass, M
0.44
0.25
0.12
98
FIGURES
99
Fig. 2.1 Three Components of Inelastic Deformation at Yielding
de
e
0[
o 0 0 0 CTS CTs
~
d
ae
h
z
c f ~
at
~ CTy CTy
0 0 0
dt
I.
b
~ I
Cross-Section Strain Stress Equivalent Stress
Fig .. 2 .. 2 Cross-Section of a Member
100
O " / f ~
1.0
0" ~ ~ 2
-. ::: 2(-)-(-)
fc EO EO
1.0
Concrete Steel
Fig. 2.3 Stress-Strain Relationship
Infinite
U)
~
U)
<V
boo
+-
(f)
'M
.......... 0"0
. "C
c:
::: (...... )iJ
3
0
OJ
So
....
So
S, Slippage
Fige 2.4 Bond Slippage at Anchorage
90
80
70
60
-
en
~
50
~
c
m
't-
40
0
en
en
Q)
30 ~
+-
(f)
...
b
20
10
0
0 0.01
#'14-18.3'
#14-27.5
#18-18.3
Casting
) ... I Top Bar
--Cut Into Halves
1 ~ I Bottom Bar
:IF (Bar No.) - (Anchorage Length)
0.02
0.03 .
Sn \I Normalized Slippage
Fig. 2.5 Pullout Test by Ferguson
o u::: 225.5 X 1.5
112
X Sn 2/3
u ::: 225.5 X 0.91J2 x $n
2
/
3
0.04
.......
o
.......
90
80
70
-
fI)
60 ~
-
....
0
co
59
I\+-
0
fI)
40
fI)
w
....
0+-
en
30
b
20
10
Ferguson
(Average of Bottom Bars)'
Koike
D16-25
Koike
0.01
Ferguson
(Average of Top Bars)
0.02 0.03
Sn , Norma Hzed Slippage
0.04
Fig. 2.6 Pullout Tests by Several Researchers
Ma (Member Test)
#6-3A
~
o
N
103
11.0
'+-
to
9.0 I-
7.0 t- .
e
'u, 5.0 I-m
to . !
@
e
3.0 - e
e
$
1.0-
o
o
-
-
-
-
-
-1.0 L-______ ____ __ ________ ______ _________ ______
1.0 6.0 7.0
Fig. 2.7 Effect of Span Ratio on Shear Deformation
104
......... , III #!
Concrete
Fig. 2.8 Bond Failure Mechanism
Shear Cracking
Zone
L
No Shear
Cracking Zone
j,'
Fig. 2.9 Shear Cracking Model
105
11.0 ~ - - ~ - - ~ I - - - - - - - - ~ I ~ - - - - - - ~ I - - - - - - - - ~ I ~ - - - - - - - ~ I - - - - - - ~
9.0 f-
7.0 -
Cot-
<0
"- 5.0 f-
U)
<0
3.0 f-
1.0 f-
o
00
o
o
o
o
o
o
o
o 0
o
-
-
-
o
-
-
-1.0 ~ ______ ~ I _________ l ~ ______ ~ I ________ ~ L ~ _______ ~ I ______ ~
0.0 5.0 10.0 15.0 20.0 25.0 30.0
Bond Stress, T s/-fF c
Fig. 2.10 Effect of Bond Stress on Shear Deformation
2.0 ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - . - - - - - - - - - - - - - - - - - - - - - - - - - ~
N=244
COV=16.5% .
o
o
+- 1.0
c
(J.)
E
L
(J.)
0..
X
W
o
0.0 ~ __________________________ ~ ________________________ ~
0.0 1.0
Calculated (100xo
y
/l)
Fig. 2.11 Yield Deformation
2.0
(Failure Mode) (Illustration)
(Typical Load - Def. Curve)

Q
r
Buckling
Compression of Bar
Flexural
Failure

8
Tension
V
Fracture of
Flexural
Failure . Tension Bar
.....
I-'
b
0"1
Compress ion

Shear Failure
Failure

L __
j
Tension
Shear
Failure
1---
4 V

. Fig" 2 .. 12 Ultimate Failure Hechanism
501-
lOt-
fL
5t-
2t-
1.01-
0.51-
I
00
X
I
0.2
I r I I
6
x
f SA _
o - Compo Flexural Failure
X - Tens. Flexural Failure
S - Compo Shear Failure
~ ~ (\0 >< 0
T - Tens. Shear Failure
SXo 0 0 8
o 0 0 0
X 0 00 0 0 ! e
o 0
0
T
J
~ 0
--rr 0
SJ S o
s OS /)
<',{\ ~ S
v ~ r (' 3 S
S.:> &0
~ , ':'c,s
t.':t-*) ',-I S ~ J S
.:.D ( : ~ .', C'
s. '( ...~ ~ v S
T
S c, J.... .J 'r)
w'-)" ).
S T S . oS
TT c:
J .... 1
o S
T
I
0.5 1.0
Ep lEo
s S
S
T
Sf
S
T
S
T
S T
ST.
1
2
Fig. 2.13 Ultimate Ductility Factor - Failure Mode
I
5
-
-
-
-
-
-
......
a
-...,J

--
Confinement Rat io, Pw (0;'0)
(f) 0-0 8'V 0.5
501-
2
a
1
I - 0.5 8'V 1.0
J
2 - 1.0 '" 1.5
@L:.&a2 1
23 .3
3 - 1.5 ""2.0
04 2 1
4 - 2.0 8'V
2 2 2
20 21 2 21 4
t
o 1
22
J.L

00
41 3
3e

t-'
<D
2 J
0
1
co
121-
321 2 2 4
0
2J: 12 i j 4
2 2 1 0 1 1 2
4
o 1 1 0(J 1
3


0 3 0 0 4
o 0 2 2
0
0.5r-
20
I I I I I
0.2 0.5 1.0 2 5
Ep lEo
Fig .. 2.14
Ultimate Ductility Factor -,Confinement Ratio
fLEXP
a

t:r 0 -4.

Z m t.:e
- \:):j)-l- III
{-lc+H..
0<:.4 ('0
\--100) \;
{-l\3 o('()
j-I'j-I- ZHltj
tj t::1 (') 0
b a> t:1 ';j (v
j-I. t""' r-
m(/) r-a::
x
c+ \-', 0
\; ';j (
Q) ('() 0 ,. 1.-1 _
50

5
N :: 142
C.O.v. :: 38 <"9'0
0
D'B s
's' S
.
Sf r
S
T 0
0.5r-- ff
I

to 5
r
10
Xo
1) 00
Co d)
0


fLCAL
X
S
()
o x
50
Fig. 2.15 Prediction Error in Ultimate Ductility Factor
I-'
o
\D
110
1.5
N=261 e
COV=60%
0
f!Jb0
(j)
1.0
0
co..
e
+-
C
<D
E 0.5
L
<D
n..
X
W
0.0
0
-.5
0.0 0.5 1.0
Ca I cu loted
f3
Fig. 2.16 Prediction Error in S for Eq. 2.1
0.6
fifo
$
0El <:) ~
0.4
0 0
co.. 0 0
0
+-
@ @
c:
I 0
0
@
a:>
., 0
0
r
t: 0.2
0@ @
0
L 0
a:> ~
0..
El
X
El
W
0.0
00
El
N=261
I 0
COV=55%
0
0
-.2
0.0 0.2 0.4
Calculated f3
Fig. 2.17 Prediction Error in S for Eq. 2.2
111
2.0 r-I ----.,------.,.------.-. l
no<O.05 e.
;...1.5 I'

;... 0 0 j
1.0 0 0 $
;.uU % 0 00
o
0.5 0 0 I
1IIJ
0
0 0
00 & e 0
____ ______ __
0.0 0.5 1.0 1.5 2.0
Ductility Fa :tor
(a)

e ii
0 i
0
1
;... 0 <l!> &> I
? 1. 0 0 00 0 . . 'i
...... 0
0
u @'iJ1S @ <:) ;
W Gl 0000, 01D 0 i
o "til '1
o Cl.
00
0 ]
o 0 0
o 0
0
0.0 L--__
0
__
0
_i:>eL.._0 ___ ---! _---', ___
0.5
0.0 0.5 1.0 1.5 2.0
2.0
1.5
>-

;...
CJ 1.0
'-..
U
W
0.5
Ducti!lty Factor
(b)
(c)
j
2.0
0 ,..------ :-------,------r-
0
------
1.5
>-

>-
CJ 1.0
'-..
U
W
0.5

o
o
o
o
o
o
o
o
o
0$ 0
0.0 '--____ --<-0_. __ , L. ____
0.0 0.5 1.0 1.5 2.0
Ductility Factor
(d)

1.5
>-

>-
01.0
"
(..)
w
0.5
O.45<n
o
00
o 0
o
o
o 00
o 0 0
o 0
. fl00

o 0
o
o
o
o
o
-

o. 0 ___ -'--___ -'--__ --'
0.0 0.5 1.0 1.5 2.0
Factor
(e)
Fig. 2.18 Experimental Data of Hysteretic Energy
(at Small Deformation)
112
5.0,------.------.------.------.------.
4.0
5.0 f
4.0
>-
to 3.0
>-
<:::1
'-.
WJUZ.O
1.0
o
(;)
o
2.0 3.0 4.0
Ductility Factor
(a)
o
4)
'"
o <:>
<:> 0
<? 0
(;) <it>
Jl00
0
0 (;) ..
o @ (;) .,
0
0
:8 0
o
5.0
o
6.0
4)1
!

I

I
I

I


O.2<Of /Oy<O.3 i
o.olL ______ L' ______ ______ ______ ____
1.0 2.0 3.0 4.0 5.0 6.0
20.0 [
15.0
>-.
to
>-.
CJJ 10.0
'-.
C,)
W
5.0
Ductility Factor
(b)
o
<i.:J
.L{)!J%<:>


o
<l>
o 00
0<l>0 '"
0.0 I I ! I
i
o
j
1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Ductility Factor
(c)

15.0
o
>-
to
(:t
10
.
0
o
'u 0 0
W 5. (, l' f. :0 '!1 '"

0.0 _...l.- I I I-----L. . .-.L--. _L ___ -L __
1.0 2.0 3.0 4.0 5.0 6.0 7.0 B.O 9.0 10.0
15.0
'20.0 r
>.
to
>-
OJ 10.0
.......
C,)
w
Ductility 'Factor
(d)
0 $
o J 9
0
0
o .
5.0 0


__ __ __ __ __ __ __
1.0 2.0 3.0 4.0 5.0 6.0 7.0 B.O
Ductility Factor
(e)
9.0 10.0
Fig. 2.19 Experimental Data of Hysteretic Energy
(at Large Deformation)
f I
113
Normalized Hysteretic
Energy per Cycle
Ductility Factor 1 p.
Fig. 2.20 Idealized Normalized Hysteretic Energy
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.1 0.2 0.3
f I = 0.5 + 2.34 !!.
8y
0.4 0.5
Sf/Sy
0.6 0.7
Fig. 2.21 Regression Analysis for fl and f2
1.0
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
0.8
f2
x
(l)
"'0
C
(l)
Cl
o
E
2.0
1.0
8 0.5
o
114
o = at + Q ~ 8 u f dE
n = 403
fLo = 1.008
(j = 0.535
o QOO
10 50
Pro ba b iii t Y (CVa )
00
o
99
o
Fig. 2.22 Damage Index of Eq. 2.1 Plotted on Lognormal Probability Paper
Cl)
0'
o
E
2.0
1.0
o
::) 0.5
8 aJ dE 8 0.6
D=-+f--J (-)
, 8
u
Ec(8) 8
u
n =403
fLo = 1.008
0-=0.501
115
Normalized Hysteretic
Energy Per Cycle
Ducti lity Factor. fL
( Enve lope Curve)
(a)
(b)
Fige 3 .. 1 Force-Deformation Relationship of Members
&1 K::: Ke X By IBM
Deformation
8M
Fig. 3.2 Degrading Hysteretic Model
116
1.5
+-
c
<l) .
E
L

X
W
N=261
COV=21%
o
0
9

o
o 0
0.0
0.0 0.5
Calculated Oy/Oy
Fig. 3.3 Apparent Yield Deformation
1.0
2.0
.-l

u
o
o
o
i 0
o
N=261
COV=12%
o
a . *

x
w
a
o
i
o
0.0 ________ -L ________ __________ ________ __________ ________
1.0 2.0 3.0 4.0
Shear Span
5.0
Ratio, lid
Fig. 3.4 Ultimate Strength
6.0 7.0
c:
m
a.n
Q (KPS) E/Q
y
o8
y
Q (KIP)
E/Qm'8y
15
10
5
1
_l


I /1' :
" .I , I
I : I '---/--,
'-Test ,/
15
10
5
8 (in.)
g
I
P
t
= 1.11
Fe = 2.75 ksi
Pw= 0.51
N = 36kps
- 8M . It
o - 3.93 + 0.00202 0 dE
Column-A

o
V
" , .. ,
\
\
\
\
\
1.0
..... 8 .. 80n,)
g
IZ9in./
P
t
=
Fe = 3.2 ksi
Pw= 0.33
N = 88 kps
8M ,t
0= 0.651 +0.0216)0 dE
Column-B
Fig. 3.5 Force-Deformation Relation and Damage Index of Columns
I-'
I-'
'-l
118
.-
8..., (inch)
0
0 0.4 8...,{inchl
2.0
1.0
0.1
00
Maximum Displacement
Time

t at.
q) 5 10 15 20 0 0 5 10 15 T' 20
.me
Absorbed Hysteretic Energy
1.0 0
o
0,5
Damage Index, D
(Column - A) (Column - B)
o Monte - Carlo Simulation (n = 100)
Fig .. 3 .. 6 Comparison to Monte Carlo Simulations
119 '
D =1.0
15 = 1.0
0.6
0.20
0.5 0.75
Column-A
0.75
0.15
Column-B
0.50
t:n
en
...
0.10
E
...
en
E
c
...
c
0.05
0
4 9 16. 25 36 4 9 16 25 36
to (sec) to (sec)
Fig. 3.7 Contour Lines for Mean Damage Index
P =90OJo P = 90OJo
0.6 0.20
5CYlo
0.5
Coiumn-A
0.15
Column-B
t:n
0'1
en
E
en
... E
c
...
c
0
4 9 16 25 36
0
4 9 16 25 36
to (sec) to(sec)
Fig. 3.8 Probability of D ~ 0.5
x
Q)
"0
c:
Q)
0"
o
E
o
o
1.0
0.5
lLO
I
, at!
ytf
D
j
'
II cOlu7m-A JJ.(:i - '0 = I sec.
II - '0 =4sec.
o - to =9 sec.
o - to = 16 sec.
.i b. - to = 25sec.
r "V -to =36sec.
"
. 1.5 0.5
Characteristic IntenSity, Ie = erG 411 to
Fig. 3.9 Characteristic Intensity vs. Damage Index
Elast ic
Columns
121
p.
I
(Hinged
Column)
(b) Failure Patterns
(0) Hybr id Mode I
Moment
at Hinge
Rotation
at Hinge
Effective
Stiffness
Fig. 3.10 Modeling of Buildings
Beam
. '-Colu'n;'n--
C'o'lur;,n
II-Inlf i-iinnpril Hinml!d
(I-) .)Mj (I-2).)Mj
O-AL-).u
1h
f'
2 I
~
8j Uj
1-), 1-2), (I-).L - ) . u ~ h
3EI GEI 12EI
(1_).)3 (1- 2).)3 h ( 1-).L-Au)3h
3
1= H "I (I-ALi :1
~ ~
(Beams)
.s:::.
.:<.
.s:::. -i-
l!') .s:::.
o :<
I
l!')
o
8i
.1-2),
I
(Half -Hinged Column)
Fig. 3.11 Moment and Rotation at Hinges
82
82
12' 82
(Frame - A)
Member Columns
Cross-
D
Section
<I ..
......
Size (inch) 16xl6
Reinforcement 8-#8
122
84
83 J
83 I
81
I
I
>Ill 9-i
20'
I
(Frame- 8)
Walls 81 82
0 8
~
...
. ..
... . ..
72x6 15x30 IOx42
-
3-#8 2-#\0
84
83
I
' ..... .1
83
I I
I
81
I
J
odII
20'
9-i
(Frame-C)
83 84
D
o ..
[J
.. ..
12x20 12xl6
2-#8 2-#6
Web-Rein.
#3-at8::(For F-A) #3-at65/1 #3-atld'#3-at5
1i
#3-at 4 (For F-8,C) .
#3-at4
11
Fig. 3.12 Detail of Frames
L-
a
QJ
..c
(f)
QJ
fI)
0.2
~ 0.1
Frame-A
Frame - C
------
,,,----- Frame - B
'--
0.4 0.8 12
Top Floor Defor'mation ( ~ o )
('Failure Mode)
Fig. 3.13 Static Analysis Result
I--'
N
LV
124
a b c d e f g h
to (sec.) 5 5 5 5 15 15 15 15
CTG(g) 0079 0.126 0.165 Q20 0.055 0.087 0.115 0.126
(Story) (Story) (Story)
3 3 3
2 2 2
d d
b
g c d
I ---4 I I
5 2000 1.0
Story Drift Hysteretic Energy Damage Index
(inch) (kips-in.)
Fig. 3.14 Response and Damage of Building-SA
a b c d e f
to (sec.) 5 5 5 15 15 15
CTG(g) 0.109 0.172 0.226 0.075 0.120 0.157
(Story) (Story) (Story)
'r 1
)
3
\ , I
\ , I
, , \
2
, , ,
I , ,
I ,
I Ie If
dab c
c
I I
5 1500 1.0
Story Dr i ft Hysteret ic Energy Damage Index
(inch) (kips- in.)
Fig. 3.15 Response and Damage of Building-SB
(Story)
3
2
(Story)
3
2
125
a b c d
to (sec.)
'(TG (g)
5 5 5 15
0.122 0.193 0.253 0.111
(Story)
: l il
a b
d
c e
(Story)
3
2
~ ~ ~ - - + - ~ ~ I I I
Story Drift
(inch)
I
, I
I
I
, \
, \
\
,
,
I
,
ld
i
e
I I
a I b I c
I
0.02
Rotation of
Joints
5 4000'
Hysteretic Energy
, (kips-in.)
(a) Results for Columns
(Story) (Story)
,
\
3
\
3
\
\
I
2
I
I
2
I
.........
.........
........... e
c ........
I I
6000
Hysteret ic Energy
(k ips-in.)
(b) Results for Beams
e
15
0.176
~ \
a d b e c
Damage Index
1.0
Damage Index
Fig. 3.16 Response and Damage of Building-SC
)(
o
E
10
(1)
0\
o
E
o
c
E
:3
E
-;c
o
:E
\'0
0.5
b.-SA
O-SB
o-SC
0,1
126
11/
/
0 //0

Beams
.-t
o
= 5 sec.
0- t . = 15 sec.
o
0.2 0.3
1.5 0.5
Characteristic Intensity, Ie (g x sec. )
Fig. 3.17 Characteristic Intensity vs. Maximum Damage Index
....
0.7
;/
10 b.-SA
..
O-SB
,
)(
/
(1)
o-SC
"C
c
0,5
"
Q)
A' 0\
0 , b.
E
/
0
0.3
0
0
/
.-t
o
= 5 sec. '-
(1)
>
0,1
O-to = 15 sec.
0
0
0,1 0.2 0.3
Characteristic Intensity,
Ie (gl.5
x sec.
0.5 )
Fig. 3.18 Characteristic vs. Overall Damage Index
;>
o
0.6
0.4
u 0.2
b. -5A
o -58
o -5C
127

.-t
g
=5sec.
o -to = 15sec.
Mean of 0
Fig. 3.19 C. o. v. of Damage Index
~ .
128
I 68 M +1----4
11
.. I
Direction of Analysis
a,b ---
Direction of Photo, a (or b)
a
b 0",
( .. ----. -,,-: -':;'fu (C l....--w-,. -"-. -r. --",,--'.-\ :"--. ...- .. -r- ...,...--.,.
m
(D)
(E)
II--' --25M---I 1...- .. --54M----t

II T e, T e
. . .
J------.--:'T-\ b----99.5 M
'0
1-0>---71.8 M-----I
(H)
0\
(F) -
ffi

b
I I- '--45 M -----.j .. I
t 0
0\
Fig. 4.1 Plan and Elevation of Damaged Buildings
12.9.
(A-a)
(A-b)
(B-a)
(B-b)
(C-a)
(C-b)
Fig. 4.2 Photographs of Damaged Buildings
1.30 ..
CD-a)
(D-b)
(E-a)
(E-b)
(F-a)
(F-b)
Fig. 4.2 (continued)
131
(G-a)
(G-b)
(H-a)
(H-b)
(I-a)
(I-b)
Fig. 4.2 <continued)
132
0.12---....-,.---....,----...,....---,.......-0--,
u
H
>. 0.09
en
c::
Q.)
-
c::
u
-
en

Q.)
u
c

C
.t:
U
0.06
0.03
o
o
o
o
8 0
o
0
0
o
Pea k Aeee leration (G)
0.5
Fig. 4.3 Characteristic Intensity vs. Peak Acceleration
10
(,)
H
...

+-
(/)
10-
1
c
QJ
+-
f'f"\\/-f"\A,)
\".v. V. - v."'Tc..
C

+-
(/)
),..,
QJ
+-
U
C
),..,
c
.s:::.
U

o 20 40 60 80 100
Epicentral Distance (km)
Fig. 4.4 Attenuation Character of Characteristic
Intensity
Degree of
Damage Physical Appearance
Total or Partial Coilapse
COLLAPSE
of Building.
Extensive Crashing of
SEVERE
Concrete. Disclosure of
Buckled Reinforcements.
MODERATE
Extensive Large Cracks.
Spoiling of Concrete in
Weaker Elements.
Minor Cracks Throughout
MINOR Building. Partial Crashing
of Concrete in Columns.
SLIGHT
Sporadic Occurrence
of Cracking.
(Damage Inspection)
0
0'1
....
c: ....
c
-0
OJ
"-
::J
.s::::.
m
I-
'6-
0
(/)
(/)
0
....J
(/)
OJ
>
.-
....J
c:
c
E
::::::s
:c
C
0 0
B G
0--0 0
F
0---0
~
I E ~ '-Maximum D.L ..c
c 0-0 0-0
"-
Overall D. I. ,
c
Q.
OJ
0::::
000
oH
.
.
o 0.2 0.4 0.6 0.8 1.0
(Simulated Damage Index)'
D
Fig. 4.5 'Calculated Damage Index vs. Observed Seismic Damag'e
~
......
LV
LV
<1>
(,)
"-
o
1..1..
/
By
Deformation
(Envelope Curve)
134
>-
~
"0<1>
W c:
NW
-U
0._
En;
"- "-
Ow
Zu:;
>-
:x:
fLc
Ductility Factor, fL
(Hysteretic Energy)
Fig. 5.1 Typical Force-Deformation Relationship
.c
(,)
c:
ClJ
~
"0
W
-0
::J
E
en
i I j I I I
lOa::
t:
L
,
t
i
1
i-
i
1
!
1 0 ~
~
;-
f-
!
f-
;
1.0:
n :: 180
C.O.V. =0.19
1.0 10
Approximated U
e
(inch)
100
Fig. 5.2 Response of Equivalent SDF Systems
! I
135
n = 261
C.O. v. = 0.54
10.0-
10'70 Limit
-c
Q)
E
o
'1.0
10.0
Calculated R i ('Vo)
Fig. 5.3 Allowable Rotational Angle of Components
u ......
. ...
.n.".
136
1 0 Design Setup -------,
o Base Shear Coefficient, C
B
o Failure Mechanism
o Material
Seismicity, I
c

/ Requirements
.------ 2. Strength Des ign ------V\.--------------I
o Lateral Seismic Load
o Determine Strength and Stiffness
o Dimension and Axial Rein.

. Size of Building
\11 I / (m
i
, hi)
3. S t ruc tu r al Analy sis --------V'--------------l
o Load Combination
o Design of Joints and Footings
o Natural Period, T
\/
Service Load
Non-Structural
Components
Ductility
o Required Ductility Index, Rd .
o Confinement Ratio
....
o Check Rr > Rd
Fig. 5.4 Design Procedure
Seismicity
Ic' TG
(Controlling Factors)
Frome AS AI
82
81
-
C\J
~ II; )0> I
20
T
Frome S S SI
a.
>- .
t-
-
N
Cl
SI
81
SI
81
BI
r
.
T
C2 T
T
T
T
Common Features
Beams of Frame A and 8
0
11 II
Ore 4 x 12
" . I I
Columns of Frame A and 8
~ J r e 20" x 20" (Frame A' )
:24
11
x 2411 (Frame 8' )
Fig .. 5.5 Dimension of Frames
I-'
W
"-J
Story
:3
2
R (OJo)
I Qft
o
=15 sec
-'--.L.-...I-----L.'_,t
o
~
5
sec.
o 1.0 0 0.2 0.4
(Story Drift) (Damage)
Frame A
Story
5
4
I I
I
I
,. to= 5 sec.
-'-
;
I
~

I R ( ~ o
~
,
. I
. _ .... p eo- 0
o 1.0 2.0 3.0 4.0 0 0.2 0.4
(Story Drift) (Damage)
Frame AI
to:: 5 sec. I ' i
3
2
R ( ~ o )
--+--
o 1.0
(Story Drift)
o 0.2 0.4
(Damage)
Frome 8
o
R ( ~ o )
o 1.0 2.0 3.0 4.0 0
to:: 5 sec.
o
~
0.2 0.4
(Story Drift) (Damage)
Frame 8
1
Fig. 5.6 Results of Damage Analysis
I-'
W
00
139
(Stiffness, K)
2.0 ~ - - ~ - - - - ~ - - ~ - - - - ~ - - - - ~ - - ~ - - ~
...-l
<{
u
~
0
0
0
0
0 ~
0
00
~ Q ~
8
0
: .
o .
~
~ 0
0
I
0
0,
:B$ 0 0
0",
.l0.
0
"1.0 ~ - - ~ ~ ~ ~ ~ ~ ~ ~ - - - - - - - - - - - - - - - - 4
0...
x
W
~
e
0
a
e
0
0
0
0
0
0
0
~
\:I
0
0
~
0.5 ~ __ ~ ____ ~ __ ~ ____ ~ ____ ~ __ ~ __ ~
0.0 1.0 2.0 3.0 4.0 5.0
Shear Span, Railo:
6.0
a/d
7.0
""0
" 6.0
(Correlation)
o ~ 0
'n = 260
c.o.v. of K = 0.39
c.o.v. of Q 0.12
Coefficient Correlation = 0.42
0
5
.
0
+-
o
cr: 4.0
c
03.0
0..
(J)
L 2.0
o
<D
..c 1 0
(J)'
1.0
QEXp/QCAL
(Strength, Q)
Fig. 6.1 Modeling Error in Stiffness and Strength
2.0
140
o C.O.V. =0.52
.s:::. 0
.....
C
QJ 0.1
O.C?IOo I 0.5 1.0
Fundamental Period, T (sec.)
Fig. 6.2 Damping Coefficient of R.C. Buildings
.......
" ~ ..........; -, ' ..~ :.
d d
N t:S k

0) ('Oc:-4
liN
:e e '-1
m
V-'\-htr
I-Ipjl--lc+
CD
...... 0 0<:..::1 H
......
13
0'l CD


...,.. H Q
.m (f.lt:4l--1<D
C":.
f-i \-h
m (J t::i 0
t-A 0 Co

CD
c
o
..-
o
I-
o
>
'+-
o
c
Q)
u
'+-
'+-
Q)
o
U
Damage Index. 5
A (T= 0.4 sec.)
1.0
0.8
0.6
0.2
0.8
- 0.6


0.2 jK
y
A .....
A At

h W E
E J)
<)=:-:=Q="'::9 a' 1.0
1.0 0 Damage Index. D
Damage Index. D
B (T =0.8 sec.) C (T= 1.2 sec.)
fig. 6.3 Variability in Response Due to Modeling Error
I-'
+:---
I-'
142
-
q
10-
1
1\1
C
a..
...
~
-
..c
0
..c
0
.....
a..
Q)
en
c.
0
"0
u
Ratio lelle, des
Fig. 6.4 Reliability of Design Procedure
143
APPENDIX A
YIELD CURVATURE AND YI ELD MOMENT
A.I Tensip.!L Jielding .
For the tension yielding, the yield curvature may be expressed by
amplifying ' (in Eq. 2.5) with a coefficient C, as
y
. = C4l'
Y Y
(A. I )
For beams (i.e., with no axial load), the coefficient C is primarily a
of the normalized longitudinal steel ratio p and p'
Figure A.I shows the resul t of flexural of beams with various
steel ratios. Based on this result, the followlng was obtained for C;
C'
(1.9 - p')
p .
forp' :;2! 0
(A.2)
9.4p' + 0.Z3
For symmetrically reinforced beams (i.e., p p'), the value of C does
not vary and may be regarded as being constant,
C 1.05
for p (A. 3)
For singly reinforced beams, the following relationship was obtained;
C
1 + 1.
9p
Z.4
for p' o (A.4)
ii
For columns, the value of C tends to increase with the increase of the
normalized axial stress, n , as shown in Figure A.2. When the axial
o
stress is not excessively high, e.g., n < 0.4, the rate of increase of
o
C with n may be regarded as being constant. The calculated values of
o
C of columns with a normalized axial stress n = 0.3 and various steel
o
ratios are shown in Figure A.3. Based on these results, the following
144
is obtained for the value of C for columns with the axial stress of
n = 0 .. 3;
o
C 1
0.45
+ 0.84 + 2p' - p
(A.5)
Therefore, by assuming a linear relationship between C and n, the
o
following may be obtained for the y.ield curvature of beams and colttm.ns;
n
{C
1
+ (C
2
-
(A.6)
where:
for p' ;z! 0
1 + 1. 9
p
2. 4
for p' o
C
2
= 1 + 0.45/(0.84 + 2p' - p)
Based on the stress-strain relationship of concrete as shown in
stress block, a (see Figure 2.3, the height
Figure
a
2.2) is,
2Ec 1 -

=

y
E
= - 1/2)
- 1/3
Y
in which,
E IE
C 0
2
= -(
3
of

+ E
This can be approximated as,
the
IE
y
a = n.d
E
( 0.75 7 d
1 + E IE E
Y 0 0
compression
)ct =



0

(A .. 7)
(A.8)
145
Therefore, the yield moment,
Figure 2.2),
M,
Y
and the axial force, N, are (see
M
Y
N
where:
ex.
c
S
c
cf'ab(d - d' - O.Sa) + a a cr (d
c c c y
d - d') + a cr d'
c t y
cf'ab + a Q cr - a cr
c c c Y t Y
s
s /s
(1 -
c
(a 1.0); 8 )--
8
:s;
s y c s c c
y
d /d
d' (1 - 8 ) d/ 2
c c
Combining Eqs. A.9 and Ae10, the yield moment is,
M
Y
0.5 f
1
bd
2
{(1 + 8
c c
n)n + (2- n)p + (n - 28 )a p'}
o c c
(A.9a)
(A.9b)
(A.10)
When the extreme fiber strain s exceeds 1.Ss, compressive
c 0
yielding may occur. In this case, the yield curvature may be expressed
as,
1.SE
o
= ---
y xh
(A.II)
in which, xh = the distance from the neutral axis to the extreme
compression fiber. Therefore, the strain of the tension reinforcing bar
is,
E
S
in which,
a s
c y
1.S(1/S' - x)
c
-----:...--- S
a x . y
y
(A .. 12)
146
s' h/d
c
The equilibrium condition of longitudinal stresses may yield,
1.5
N = bhf' 1
x
S f fCy)dy + ybdf'p'- a f'bdp
c . 0 C C C
where:
fCy)
y
2
2y - y, y 1.0
= 1.0
1 - f Y /a
C y
y > 1.0; and .
(A.13)
Hence, the value of x is obtained by solving the following equation;
0.778 S'x
2
c
(
,1.5) 1.5
no - yp - --;- p x - as' p
y y C
o (A.14)
In the case that the tension bar is also yielding (i.e., a = 1.0), x is
c
obtained simply as follows;
x = 1.286(n + p - yp')/S'
o C
The corresponding yield moment is,
M
Y
2 1.5
bh l Y + 0.5 - x)dy
+ (O.Sh - d )ya a + (O.Sh - d )a a a
esc . - t cst
This may lead to the following expression;
(A.16)
147
M f'bd
2
{6,2(0.389x - 0.3l5x
2
) + (1.0 - 0.56')a p
y c c c c
+ (0.56' - 6 )yp'}
c c
CA.17 )
in which, when "a > 1.0, a 'should be replaced by 1.0.
c c
1.4
1.3
1.2
c
o---op=pl
pi
0.1
0.2
0.3
0.4
I.
0.7
\.00 0 02
.I . 0.3 0.4 0.5 0.6 0.7
p
Fig. A.I Coefficient of C
. vs. p
c
pi =0.1
pl=constant
2.2
2p I_p = constant
2.0
1.8
0.2
\.6 0.3 .
1.4
1,2
0
0.1 0.2 0.3 0.4 0.5
p
148
C
1.6
1.5
P =0.3
p'=0.3
1.4
1.3
1.2
1.1
\.00
0.1 0.2 0.3
no
Fig. A.Z Coefficient of C
vs. N
o
c
2.2
2.0
C=I+ 0.45
0.84+(2p'-p}
1.8
1.6
1.4
.
-0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
(2pl_p)
Fig. A.3
Coefficient of C for Columns with N
o
0.3
149
APPENDIX B
DERIVATION OF EQUATION 3.17
The following approximation is assumed in Eq. 3.16;
2
x
2xU(x) exp (--2)
20"1
(B.1)
in wnich, U(x) = the unit step function. The numerical error 1n the
above approximation is negligibly small especially for large values of
x. Therefore, Eq. 3.16 is reduced to the following;
[integrand of Eq. 3.16] =
x 2: 0
x < 0
(B.2)
The integration yields Eq. 3 ~ 1 7 .
150
APPENDIX C
DETAILS OF EQUATION OF MOTION
The matrices in Eqs. 3 .. 32 and 3.33 are shown below for the case of a
3-story building.
1
cC
2
1
0 - C -
c 1 IDI IDl
[C]
C ~
c 1 IDI
C ( ~ + ~ )
c 2 IDl ID2
C ~
c 3 ID2
(C.l)
0
C ~
c 2 ID2
- C ( ~ + ~ )
c 3 m
2
ID3
k ~
12EIl
- C'L
2 c 1
c 1 IDl
ih1e1ID
1
[K] k ~ +
12EIl
cal
c 1 1lJ.
2
lh1elm
1
0
o
(C.2)
[B]
[A]
6EI I
1
o
151
o
6EI,.,
.:>
o
(1-c
a
3)c
k
3!m
Z
I
-(1- a ) k ( ~ + ~ J I
I C 3 c 3 m
2
m3
I .
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I -
I
I
I
I
(C.3)
(C.4)
152
4EI
I
2EIl
I I
I I I
I I I
0
I
bal bkl + -2-
I
[2h
I I
I I I
[ hI
I I I
I 1 I I
I I I
I I I
I I
2EI2
,
2EII
I
4EIl 4EI21
,
I
b
k
2
I
I b 0.
2
+--+--1 1
[2h [2h
l
[2h21
[2h
1
I 1
I
1 2 r
[F]
I I 1
1 1
,
I
2EI2
1
4EI2 4EI3
I
1 1 1
1 1 I
,
[2h
Ib
a
3
k+--+--
I
0
I
b 3 [2h2 [
2h
3
I
I I I
I 2 I I
I I I
I I I
I 1
2EI3
I
I I I
0
I
0
I I
I I
[2h
I
I I I
I I I
I I 3 I
o
o
(C.S)
153
o
[D] = (C.6)
o
6EI
I
0 0
2
. hIe
l
6EI
I
,6EI
2
O
2 2
. hIe
l
. h
2
e
2
[E] ,= (C.7)
0
6EI
2
.
.
6EI
3
2
1-
2
. h
2
e
2
I
. h3
e
3
I
I
I
I
I
6EI3 I
0 0
1
1-
2
1
1
. h3
e
3
1
in which, following transformations are made for the rotational spring
parameters using the average span length, i;
In this scheme, the rotational deformation e is
equivalent translational deformation w , as
i
e. w.lf..
1 1
i
expressed by the
(C .. 8)
154
APPENDIX D
THE EQUIVALENT SDF SYSTEM
With the prescription of a linear response shape, the floor
response, u., may be expressed as
~
u. h ~ R (D.l)
1 1
where:
h' = height of the i-th floor from the base; and
i
R = the deformation angle of the entire building corresponding to
the linear mode shape (as function of time).
The equation of motion for a MDF building (neglecting the damping term)
is
[M]{u,} + [A]{f.}
1 1
- [H] {l}x
G
(D .. 2)
where:
1 -1 0- - - .. 0
'. I
\. ,
, ,
[A]
\. I I
, I ;
'. . ',I I
I ~ . _. ___ ~ ~ ~ <: j
f = i-th story restoring force; and
.. 1
X = base shear ..
G
155
By combining Eqs. D.] and D.2, the following is obtained;
(D.3)
or
o (D.4)
The equation of motion for the equivalent SDF system may be expressed
as,
M (u .. ) f
e e + xG + e
o
where:
M
=
equivalent mass;
e
u
=
equivalent response; and
e
f
=
e
equivalent restoring force.
Compararing Eqs. D.4 and D.S, u may be obtained as,
e
u
e

1 1
}T[M]{l}
1

'-' 1 1
R=---R

1 1
The kinetic energy for the MDF and SDF systems may be written as,
(D.S)
(D.6)
(D.7)
By substituting Eq. D.6 into Eq. D.7, the equivalent mass is obtained
as,
M
e
156
Therefore, Eq. D.4 can be rewritten as,
x
L r n . h ~
~ ~
This leads to the following relationship between f
forces f.;
1
f
e
(D.S)
o (D .. 9)
and the restoring
e
(D.lO)
By substituting the design story shear forces Q given in Eq. 5.9 into
i
the above relationship, the equivalent strength, Q , is obtained as
e
where:
C = design base shear coefficient; and
B
W = total weight of building.
(D.ll)
1.
157
LIST OF REFERENCES
Akiyama, H., "Aseismic Ultimate Design of Building Structures,"
Tokyo Univ. Press, Tokyo, Japan, 1980.
2. American Concrete Institute, "Building Code Requirement for
Reinforced Concrete (ACI 318-77),n 1979.
3. Am in , M. and Ang, A. H-S., "Nonstationary Stochastic Model of
Earthquake Motions,1I Journal of the Engineering Mechanics Division,
ASCE, Vol. 94, EM2, pp. 559-583, April 1968.
4. Ang, A. H-S., "Structural Risk Analysis and Reliability-Based
Design," Journal of the Structural Division, ASCE, Vol. 99, ST9,
pp .. 1891-1910, 1973.
5. Ang, A. H-S. and Cornell, C. A., "Reliability Bases of Structural
Safety and Design," Journal of the Structural Division, ASCE, Vol.
100, ST9, pp. 1755-1769, September 1974.
6. Ang, A. H-S. and Tang, W. H., Probability Concepts in Engineering
Planning and Design, Votumne I-Basic Principles, John Wiley and
Sons, Inc., New York, NY, 1975.
7. Ang, A. H-S. and Tang, W. H., Probability Concepts in Engineering
Planning and Volume II: Decision. Risk, and Reliability,
John Wiley and Sons, Inc., New York, NY, 1984.
8. Aoyama, H., Characteristics of Reinforced
Concrete Members Subject to Axial Load and Reversal of Bending,"
Proceed"ings, International Symposium on the Flexural Mechanics of
Reinforced Concrete, ASCE-ACI, Miami, pp .. 183-212, November 1964.
9. Aoyama, H .. , Endo, T .. , and Minami, To, "Behavior of Reinforced
Concrete Frames Subject to Reversal of Horizontal Forces,"
Proceedings of dapan Earthquake Engineering Symposium 1966, Tokyo,
Japan, pp. 315-320.
IOe Architectural Institute of Japan, 1978 Miyagiken-Oki Earthquake
Damage Report, Tokyo, Japan, 1980.
11. Asano, K., "Stochastic Earthquake Response Analysis of the Lumped
Mass Structural System with Elasto-Plastic Characteristics,"
Transactions. Japans, pp. 75-82, 1976.
12. Ata1ey, M. B. and Penzien, J., Seismic Behavior of Critical
Regions of Reinforced Concrete Components as Influenced by Moment,
Shear and Axial Forces,u University of California at Berkeley,
EERC, Report No. 75-19, December 1975.
13.
158
Atalik, T .. S. and Utku, S., "Stochastic
degree-of-Freedom Nonlinear Systems, "
Engineering and Structural Dynamics,
April 1976.
Linearization of Multi-
Journal of Earthquake
4, No.4, pp. 411-420,
14. ATC (Applied Technology Council, associated with the SEADC) ,
Tentative Provisions for the Development of Seismic Regulations for
Buildings, ATC Publication, ATC 3-06, June 1978 .
15. Baber, T. T. and Wen, Y. K., "Stochastic Equivalent Linearization
for Hysteretic, Degrading, Multistory Structures," Civil
Engineering Studies, SRS 471, University of Illinois, Urbana,
Illinois, December 1979.
16. Baber, T. T. and Wen, Y. K., Vibration of Hysteretic
Degrading Systems," Journal of the Engineering Mechanics Division,
ASCE, Vol. 107, No. EM6, pp. 1069-1087, No. EM6, December 1981.
17. Banon, H., Biggs, J. Mo, and Irvine, H. M., "Prediction of Seismic
Damage in Reinforced Concrete Frames," Seismic Behavior and Design
of Buildings, Report No.3, MIT, May 1980.
18. Berlero, V. and Bresler, B. "Panel on Design and Engineering
Decisions Failure Criteria (Limit States)," Developing
for Evaluating Safety of Existing Buildings,
University of California, Berkeley, Report UCB/EERC 77-06, 1977 "
19. Blejwas, T. and Bresler, B., in inxisting
Buildings," University of California, Berkeley, EERC 78-12, 1979.
20., URS/John A. Blume and Associates, "Effects Prediction Guidelines
for Structures Subjected to Ground Motion," San Francisco,_ CA.,
1975 ..
21. Brown, R. H .. and Jirsa, J. D., "Reinforced Concrete Beams under
Load Reversals," ACI Journal, Vol .. 68, pp. 380-390, May 1971.
22. Burns, N .. H. and Siess, C. P., Characteristics of
Beam-Column Connections in Reinforced Concrete," Civil Engineering
Studies, SRS No. 234, University of Illinois, Urbana, Illinois,
January 1962.
23 e Celebi, M. and Penzien, J", "Experimental
Seismic Behavior of Critical Regions
Components as Influenced by Moment and
California at Berkeley, EERC 73-4, January
Investigation into the
of Reinforced Concrete
Shear," University of
1973.
24. Clark, A .. P., "Comparative Bond Efficiency of Deformed Concrete
Reinforcing Bars,"ACI Journal, Vol. 18, pp. 381-400, December
1946 ..
25 ..
159
Clough, R. W., "Effect of Stiffness Degradation
Ductility Requirements," Technical Report No. 66-16,
Materials Research, Department of Civil Engineering,
California, Berkeley, CA., October 1966.
on Earthquake
Struct ur es and
University of
26.. Culver, C. G .. , Lew, H. S., Hart, G.. C., and Pinkham, C. W.,
"Natural Hazard Evaluation of Existing Buildings," U .. S. Department
of Commerce, National Bureau of Standards Building, Science Series,
61, Washington, D.C., 1975.
27. Ferguson, P. M. and Thompson, J. N., Length of High
Strength Reinforced Bars in Bond," ACI Journal, Vol. 59, pp.
887-922, August 1962.
28. Ferguson, P. M., Breen, J. E., and Thompson, J. No, Test
on High Strength Reinforcing Bars," ACI Journal, Vol. 62, pp.
933-950, August 1965.
29. Fukada, Y., on the Restoring Force Characteristics of
Reinforced Concrete Buildings (Part'"I: Formation and Response
Evaluation of Degrading Tri-linear Model)," Proceedings of the 40th
Kanto District Symposium of Tokyo, Japan, pp. 121-124,
November 1969.
30. Gaston, J. R., Siess, C .. P., and Newmark, N. M., "An Investigation
of the Load-Deformation Characteristics of Reinforced Concrete
,Beams Up to the Point of Failure, fU Civil Engineering Studies, SRS
No. 40, University of Illinois, Urbana, Illinois, December 1952.
31. Gosain, N. K., Brown, R. H., and Jirsa, J. 0", "Shear Requirements
for Load Reversals on RC Members," Journal of the Structural,
Division, ASCE, Vol. 103, ST7, pp. 1461-1476, July 1977.
32. Goto, Y., "Cracks Found in Concrete Around Deformed Bars," ACI
Journal, Vol. 68, pp. 224-251, April 1971.
33. Grossmayer, R. L., "Stochastic Analysis of Elasto-Plastic Systems,U
Journal Qf the Engineering Mechanics Division, ASCE, Vol. 107, EMl,
pp. 97-117, February 1981.
34. Gumble, E. J., "Statistical Theory of Extreme Values and Some
Practical Applications," Applied Mathematics Series 33, National
Bureau of Standards, Washington D.C .. , February 1954.
35. Hasselman, T. K., Eguchi, R. T., and Wiggins, J. H., of
Damageability for Existing Buildings in a Natural Hazard
Development," J. He Wiggins Company, Redondo Beach, CA., 1980 ..
36. Hirosawa, M., Experimental Results on Reinforced Concrete
Shear Walls and Analysis on Them," Building Research Institute,
Ministry of Construction of Japan, March 1975.
160
37. Hirosawa, M., Goto, T., Fujiki, Y., '''Past Experimental Results on
Reinforced Concrete Columns ,." Building Research Institute, Ministry
of Construction of"Japan, February 1973.
38. Hwang, T-H., of Variation in Load History on Cyclic
Response of Concrete F1p.xural Members," Ph .. D. Thesis, Department of
Civil Engineering, University of Illinois, Urbana, IL .. , 1982.
39. Iemura, H., Response of Stationary and Deteriorating
Hysteretic Structures," Kyoto University, Japan, May 1977.
40.. Ikeda, A." "Experiment of Reinforced Concrete Columns Subject to
the 'Reversal Loading in Inelastic Region," Concrete Journal. Japan
National Council on Concrete, Vol. 8, No. 12, pp. 1-13, December
1970 ..
41. Iwan, W. D. and Lutes, L. D., "Response of Bilinear Hysteretic
Systems to Stationary Random Excitations," Journal of the
Acoustical Society of America, Vol. 43, No.3, pp. 545-552, March
1968.
42. Jirsa, J. 0 .. and Marques, J.L.G .. , "A Study of Hooked Bar Anchorage
in Beam-Column Joints," Report to Reinforced Concrete Research
Council (Project 33), University of Texas at Austin, July 1972.
43 .. Kanai, K. and Yoshizawa, S., nOn the Period and the Damping of
Vibration in Actual Buildings," Bulletin of the Earthquake Research
Institute, University of Tokyo, Vol. 39, pp. 477-489, 1961.
44. Kani, G.N.J., "The Riddle of Shear Failure and Its Solution," ACI
Journal, Vol. 61, pp. 441-467, April 1964
45. Kent, D. C. and Park, R., Members with Confined
Concrete," Journal of the Structural Division, ASCE, Vol .. 97, ST7,
pp. 1969-1990, July 1971.
46. Kobori, T., Minai, R., and Suzuki, Y., Linearization
Method for Nonlinear System Considering the Plastic Drift Effect,"
Annual Conference of Alj 1n Tohoku, Japan, pp. 777-778, 1973.
47. Koike, K., Takeda, J., Yoshioka, K., and Nakayama, T .. , nOn the
Yield Deformation of Reinforced Concrete Columns," Annual
Conference of AIJ, Japan, pp. 1297-1298, October 1974.
48 .. KilstH, O. and Bouwkamp, J. G., "Behavior of Reinforced Concrete
Deep Beam-Column Subassemblages Under Cyclic Loads," University of
California, Berkeley, EERC, Report No. 73-8, May 1975.
49. Lai, P. S-S., "Statistical Characterization of Strong Ground
Motions Using Power Spectral Density Functions," Bull. Seism. Soc.
Am., Vol. 72, No.1, pp. 259-274, February 1982 .
.:; .:....
161
50. Lin, Y. K., Probabilistic 'Theory of Structural, Dynamics,
McGraw-Hill, 1967.
51. Lybas, J. and Sozen, M. A., "Effect of Beam Strength and Stiffness
on Dynamic Behavior of Reinforced Concrete Coupled Walls," Civil
Engineering Studies, SRS No. 444, University of Illinois, Urbana,
Illinois, July 1977.
52. Ma, S-Y. M., Bertero, V. V., and Popov, E. P., "Experimental and
Analytical Studies on the Hysteretic Behavior of Reinforced
Concrete Rectangular and T-Beams," University of
Berkeley, EERC, Report No. 76-2, May 1976.
53. McCollister, H. Me, Siess, C. P., and Newmark, N. M.,
"Load-Deformation Characteristics of Simulated Beam-Column
Connections in Reinforced Concrete," Civil Engineering Studies, SRS
No. 76, University of Illinois, Urbana, Illinois, June 1954.
54. McGuire, R. K. and Barnard, J. A., Distance and
Intensity Data forC.I.T. Strong Motion Records," U.S. Geological
Survey Journal of Research, Vol. 5, No.4, pp. 437-443, July 1977.
55. Middleton, D., An Introduction !Q Statistical Communication Theory,
McGraw-Hill, 1960.
56. Minai, R. and Suzuki, Yo, "Reliability Analysis of Aseismic Safety
of Elasto-Plastic Structures Considering Random Fatigue,"
Proceedings of the Third Japan Earthquake Engineering Symposium,
pp. 723-730, 1970.
57.. Morita, Se and Tomita, K., "Basic Study on Bond Between Steel and
Concrete,iI Transactions. AIJ, Japan, March 1967---September 1967.
58. Morita, S., Sumi, T .. , and Kiguchi, S., "Bond Behavior of
Reinforcing Bars in Anchorage Under Cyclic Loadings," Annual
Conference of AIJ, Japan, pp. 1015-1015, October 1975.
59. Muto, K., Dynamic Analysis of Structures, Maruzen, Tokyo, Japan,
1963.
6 o. Naka, M., "Study on the Shear Failure of Reinforced Concrete
Columns," University of Tokyo, Japan, 1963.
61. Newmark, N. M., Siess, C. P., and Sozen, M. A.,
Characteristics of Reinforced Concrete and Ductility Requirement
for Earthquake Resistance," Proceedings, 30th Annual Convention of
SEAOC, pp. 54-66, 1963.
62.. Olsen, S. 0., Sozen, M. A., and Siess, C. P., "Investigation of
Prestressed Reinforced Concrete for Highway Bridge s; Part IV:
Strength in Shear of Beams with Web Reinforcement," Engineering
162
Engineering Station, Bulletin 493, University of Illinois,
1967 ..
63. Park, R. and Paulay, T., Reinforced Concrete Structures, John Wiley
and Sons, Inc .. , New York, NY, 1975.
64. Pires, J .. E.A., Wen, Yo K., and Ang, A. H-S., "Stochastic Analysis
of Liquefaction under Earthquake Loadings," Civil Engineering
Studies, SRS No. 504, University of Illinois, Urbana, Illinois,
April 1983.
65. Portillo, M. and Ang, A. H-S., "Evaluation of Safety of Reinforced
Concrete Buildings to Earthquakes," Ph.D. Thesis, Department of
Civil Engineering, University of Illinois, July 1976.
66. Short Column Committee, Commprehensive Research for the
Prevention of Failure in Short Reinforced Concrete Columns," Series
1 to 62, Annual Conference of AIJ, Japan, 1973-77.
67. SEAOC (Structural Engineers Association of California),
"Recommended Lateral Force Requirements and Commentary," 1975.
68. Sues, R. He, Wen, Yo K., and Ang, A. H-S., "Stochastic Seismic
Performance Evaluation of Buildings," Civil Engineering Studies,
SRS No. 506, University of Illinois, Urbana, Illinois, May 1983.
69. Tajimi, H., "Dynamics for Building Structures," Korona-sha, Tokyo,
Japan, 1965.
70. Takeda, T., Sozen, M. AD, and Nielsen, N. N., Concrete
Response to Simulated Earthquakes," Journal of the Structural
Division, ASCE, Vol. 96, ST12, pp. 2557-2573, December 1970.
71. Takizawa, He, Model for Simulating the Dynamic Damaging
Process of Lowrise Reinforced Concrete Buildings during Severe
Earthquakes," Earthquake Engineering and Structural Dynamics, Vol.
4, pp. 73-94, 1975.
72. Tanaka, Te , Yoshizawa, S., Osawa, Y., and Morishita, To, "Period
and Damping of Vibration in Actual Buildings during Earthquakes,"
the Earthquake Research Institute, University of Tokyo,
Vol. 47, pp. 1073-1092, 1969 ..
73. Tichy, Me and Vorlicek, M., Statistical Theory of Concrete
Irish University Press, Prague, 1972.
74. Umemura, H. and Aoyama, H., "Evaluation of Inelastic Seismic
Deflection of Reinforced Concrete Frames Based on the Tests of
Members, VI 4WCEE, Chile, pp. (B-2) 92-107, January 1969.
75. Umemura, H. and Takizawa, H., Response of Reinforced
Concrete Buildings," Structural Engineering Documents, IABSE, 1982.
163
76.. U. S. Department of Commerce, San 'Fernando, Earthquake .Q.f
February 1971, Washington, D.C., 1973.
77 .. Vanmarcke, E. H .. and Lai, P. S-S., "Strong-Motion Duration and RMS
Amplitude of Earthquake Records," Bull. Seism. Soc. Am., Vol. 70,
No.4, pp.. 1293 -13 07, Augu s t 1980 ..
78. Wen, Y. K., Linearization for Hysteretic Systems under
Random Excitations," Journal of Applied Mechanics, Transactions of
AS ME , Vol. 47, No .. 1, pp. 150-154, March 1980.
79. Whitman, R. V., Cornell, C. A., Vanmarcke, E. H., and Reed, J. W.,
''Methodology and Initial Damage Statistics," Department of Civil
Engineering, MIT, Research Report R73-57, No. ST380, 1972.
80. Whitman, R. V., Hong, J. T., and Reed, J. W., Statistics
for High Rise Buildings in the Vicinity of, the San Fernando
Earthquake," ,Seismic Design Analysis Report No.7, Department of
Civil Engineering, MIT, Research Report R73-24, 1973.
81. Whitman, R. V. and Biggs, J .. M., "Seismic Design Decision Analysis
- Methodology and Pilot Application," Report No. 10, Structures
Publication No. 385, MIT, 1974.
82. 'Wight, J. K. and Sozen, Me A., "Shear Strength Decay in Reinforced
Concrete Columns SUbjected to Large Deformation Reversals," Civil
Engineering Studies, SRS No. 403, University of Illinois, Urbana,
Illinois, August