Advanced in Earthquake Engineering Applications

First Edition 2008
MOHD. ZAMRI RAMLI, ROZAINA ISMAL & MELDI SUHATRIL 2008

All rights reserved. No part of this publication may be reproduced or transmitted
in any form or by any means, electronic or mechanical including photocopy,
recording, or any information storage and retrieval system, without permission
in writing from Universiti Teknologi Malaysia, Skudai, 81310 Johor Darul
Tak'zim, Malaysia.
Perpustakaan Negara Malaysia
Cataloguing-in-Publication Data
Advances in earthquake engineering applications / editor Mohd Zamri Ramli,

Rozaina Ismail, Meldi Suhatril.
ISBN 978-983-52-0571-2
1. Earthquake engineering. 2. Earthquake hazard analysis. I. Mohd Zamri
Ramli. II. Rozaina Ismail. III. Meldi Suhatril. IV. Universiti Teknologi
Malaysia. Fakulti Kejuruteraan Sivil.
624.1762
Pereka Kulit: MOHD. NAZIR MD. BASRI
Diatur huruf oleh / Typeset by
MOHD. ZAMRI RAMLI & RAKAN-RAKAN
Fakulti Kejuruteraan Awam

Universiti Teknologi Malaysia
81310 Skudai
Johor Darul Ta'zim, MALAYSIA
Diterbitkan di Malaysia oleh / Published in Malaysia by
PENERBIT
UNIVERSITI TEKNOLOGI MALAYSIA
34 38, Jalan Kebudayaan 1, Taman Universiti,

81300 Skudai,
Johor Darul Ta'zim, MALAYSIA.
(PENERBIT UTM anggota PERSATUAN PENERBIT BUKU MALAYSIA/
MALAYSIAN BOOK PUBLISHERS ASSOCIATION dengan no. keahlian 9101)
Dicetak di Malaysia oleh / Printed in Malaysia by
UNIVISION PRESS
Lot 47 & 48, Jalan SR 1/9, Seksyen 9

Jln. Serdang Raya, Tmn Serdang Raya
43300 Seri Kembangan, Selangor Darul Ehsan
MALAYSIA
Contents
CONTENTS
Preface
vii
Chapter 1
Development of Seismic Hazard Maps of 1

East Malaysia
Azlan Adnan, Hendriyawan, Aminaton Marto,
B.Selvanayagam P.N.
Chapter 2
Microzonation
Study
for
Putrajaya, 19
Malaysia
Azlan Adnan, Hendriyawan; Aminaton Marto,
Irsyam, M.
Chapter 3
Vulnerability Study of Public Buildings 35

Subjected to Earthquake by ATC-21, ATC22 And Finite Element Modeling
Mohd Zamri Ramli, Azlan Adnan, Suhana
Suradi.
Chapter 4
Buildings Classification using

Technology Council (ATC 21)
Mohd Zamri Ramli, Tay Tzer Yong
Chapter 5
Seismic Performance of Sultan Azlan Shah 61

Bridge under Low Earthquake Ground
Motion
Azlan Adnan, Meldi Suhatril, Ismail Mohd
Taib
Applied 49
Contents
Chapter 6
Seismic Performance of Rapid KL Elevated 77

Span Bridge under Low Earthquake Ground
Motion
Azlan Adnan, Meldi Suhatril, Ismail Mohd
Taib
Chapter 7
Analysis of Prestress Concrete Highway 93

Bridges with and Without Rubber Bearing
Mohd Zamri Ramli, Azlan Adnan
Chapter 8
Predicting of Bridge Condition Based on 111

Seismic Zonation by using Artificial Neural
Network
Azlan Adnan, Sophia C. Alih, Rozaina Ismail
Chapter 9
Database System and Digital Earthquake 127

Evaluation of Buildings
Azlan Adnan, Rozaina Ismail
Index
145
vii
Preface
PREFACE
This book talks about applications of advances earthquake

engineering. As we know, more than 6400 people died in the Kobe
Earthquake on January 17, 1995. Two to three million people died
in earthquakes during the 20th century. More than 240,000
perished in the Tangshan Earthquake in China and 20,000 in the
Izmit and Indian earthquakes. And every of us must still remember
with 2004 Acheh and 2008 Sinchuan earthquake.
This book is intended as a structural analyses guide for
practitioners and advanced students. Earthquake engineering is a
vast subject and the intention of this book is not to provide a fully
comprehensive treatment of all aspects. Rather, it is providing the
practicing engineer with an understanding of those aspects of the
subject that are important when analyzing structural. Although
earthquake does not respect national boundaries, the practice of
earthquake engineering does vary significantly between region,
and this is reflected in the differing formats and requirements of
national seismic codes.
The first chapter of this book reviews the process of development
of seismic hazard map. The second chapter gives briefed
introduction about the development of microzonation study. These
first two chapters are important to provide input for seismic design,
land use management, and estimation of the potential for
liquefaction and landslides. Chapter 3 and 4 discussed how the
vulnerability study of building is done using ATC-21, ATC 22 and
viii
Preface
finite element modeling. The vulnerability and analysis of

performance of the bridge are present in chapter 5 and 6. Chapter 7
introduced the method of modeling structures with and without
seismic hazard reduction material such as seismic rubber bearing.
Chapter 8 therefore discussed about the technology to do
prediction of structure under seismic using artificial neural
network. Lastly, chapter 9 introduces special tools to evaluate the
buildings using database system.
Mohd Zamri Ramli

Rozaina Ismail
Meldi Suhatril
Faculty of Civil Engineering,
2008
1
DEVELOPMENT OF SEISMIC HAZARD
MAPS OF EAST MALAYSIA
Azlan Adnan
Hendriyawan
Aminaton Marto
B. Selvanayagam, P.N.
INTRODUCTION
Earthquakes are natural phenomenon which can cause huge losses
of life and economy. In recent years, Malaysia is more aware to the
seismic effect on their buildings because the tremors were
repeatedly felt over the centuries from the earthquake events
around Malaysia (SEASEE, 1985). Peninsular Malaysia has felt
tremors several times from some of the large earthquakes
originating from the intersection areas of Eurasian plate and IndoAustralian plate near Sumatra, and some of the moderate to large
earthquakes originating from the Great Sumatran fault. On the
other hand, East Malaysia has experienced small to moderate
earthquakes from local origin and tremors originating from the
southern part of the intersection area of Eurasian and Philippines
plates as listed by Surat (2001) and Rosaidi (2001).
The 1976 earthquake of magnitude 5.8 in Lahad Datu caused
some houses and buildings to develop cracks in the walls. A four
storey police complex nearing completion suffered severe
structural damage. Several roads in the district were reported to
have cracked too, causing damage. Similarly, the 1991 Ranau
earthquake of magnitude 5.2 on Richter scale caused extensive
damages to a four-storey teachers quarters and were verified unfit
Advances in Earthquake Engineering Applications
for occupancy. The earthquake of magnitude 4.8 that occurred on

2 May 2004 near Miri, Sarawak likewise caused some damages to
the non-reinforced concrete buildings and developed cracks on the
ground (Bernama, 2004).
The frequent occurrence of tremors within the country and
nearby region seems to suggest that seismic risk in Malaysia is
evident. The question now is the level of risk and its regional
variation and whether it is necessary to consider seismic factors in
the planning and design of structures and/or infrastructures. These
questions have so far remained unanswered due to a lack of
understanding of seismicity and inadequate seismic data in
Malaysia. Hence, the level of seismic risk in Malaysia is still
barely known. It is not known if such risk should be considered in
future design of structures and/or infrastructures. This is further
compounded by the fact that Malaysia is rapidly developing and
major installations and high-rise structures are being constructed at
a rapid pace.
Based on the above facts, the earthquake engineering research is
urgently required in order to predict the possibility of earthquakes
in the future that can cause damages to the buildings and structures
in Malaysia and to find the solutions for mitigating the effects.
The engineers have a responsibility to quantify the earthquake
risks in Malaysia quantitatively and find the optimal solutions to
deal with those effects.
This paper presents the development of seismic hazard maps for
East Malaysia using total probability theorem. The analysis covers
the development of seismotectonic model, the determination of
seismic hazard parameters, and the selection of appropriate
attenuation relationships for East Malaysia.
SEISMOTECTONIC SETTING
East Malaysia is located at the triple junction of the Pacific
(through the Philippine plate), Indo-Australian and Eurasian Plates
(Figure 1.1). The interactions among the plates are very complex
and active. According to the Seismo-Tectonic Map published by

Jabatan Penyiasatan Kajibumi Malaysia (JPKM, 1994) the
seismicity around this location is affected by the low seismic active
level of stable Sunda tectonic plate and moderately active seismic
level of Sabah and East Kalimantan (Figure 1.2).
Figure 1.1 Plate tectonic setting of Southeast Asia (after McClay, 2000)
Generally tectonic features that affected East Malaysia can be

divided into three classifications, i.e. subduction zone, transform
zone and diffuse seismicity zone. The first classification is
subduction zone. All of those earthquakes that occurred near
convergent boundaries where an oceanic plate is being subducted
under an island arc or continent are classified into this zone. This
zone is formed due to the movement of Sangihe plate that is
subducting an island of Sulawesi on the north side as shown in

Figure 1.3.
Legend:
Strongly Active:
A. Outer Burmese Arc
B. Inner Burmese Arc
C. Indonesian Arc
Moderately Active:
D. Shan Plateau
E. Sabah and East
Kalimantan
Active:
F. Irrawaddy/Andaman
Trough
Stable:
G. Junction South China
Sea and Sunda Shelf
Figure 1.2 Seismotectonic setting around East Malaysia (JPKM, 1994)
Figure 1.3 Cross section of North Sulawesi Subduction Fault (Cardwell,

1980)
The Sangihe plate is subducting down to more than 300 km on

the north side of Sulawesi (Figure 1.3). There are two large
earthquakes recorded by National Oceanic and Atmospheric
Administration (NOAA), United Stated and International
Seismological Catalogue (ISC), United Kingdom occurred in this
region. The first one was occurred on 22nd January 1905. The
earthquake was located at the longitude of 123.0oE and latitude of
1.0oN. The depth of the earthquake was 90.0 km and magnitude,
MS, was 8.4. The second was occurred on 21st December 1939 at
the longitude of 123.0oE and latitude of 0oN and magnitude, MS, of
8.6. The depth of the earthquake was 150 km.
The second classification is transform zone. Transform zone is a
terminology for earthquakes that occurred on boundary between
two lithospheric plates that are sliding past one another (transform
plate boundary). The shallow crustal faults such as Palu-Koro,
Walanae or Paternoster (Sulawesi), and Melange (Java) faults can
be classified into this zone. Figure 1.4 depicts the location of the
shallow crustal faults.
The Palu-Koro fault has long been established and mapped as a
traversing fault zone across the Sulawesi Island. It roughly divided
into two halves, i.e. West and East Sulawesi (Bemmelen, 1949;
Brouwer, 1947). The largest earthquake around this region was
occurred on 19th May 1938 at the longitude of 123.0oE and latitude
of 1oS and magnitude, MS, of 7.9. The latest big earthquake was
occurred on 19th October 2001. The earthquake was located at the
longitude of 123.91oE and latitude of 4.1oS and magnitude, MS,
was 7.5.
The third classification is diffuse seismic zone. All earthquakes
that occur in areas where seismicity is not associated with a single
fault or fault type are classified into this zone. This zone is found
in Batui, Poso, Mamuju Sulu, Tarakan Basin, and Kutai-Mahakam
Basin source zones. The detail of these faults system is shown in
Figure 1.4.
Figure 1.4 Fault sources in Borneo Island (McClay, 2000)
REGIONAL SEISMICITY
The primary seismicity database used for this study was compiled
primarily from four sources:
1. Earthquake listings held by National Earthquake Information
Center (NEIC), U.S. Geological Survey (USGS) of the
United State.
2. International Seismological Center (ISC), United Kingdom.
3. Malaysian earthquake listing prepared by the Malaysian
Meteorological Service.
4. Pacheco and Sykes catalogue (1992).
The combined catalogues cover areas from 105oE to 125oE
longitude and from 10oS to 10oN latitude and include 7039
earthquake events occurred during 108 years period of observation
(1897-2004). The location of earthquake epicenter during that
period of observation is shown in Figure 1.5.
Figure 1.5 Distribution of epicenters around project Location
Some analyses have been performed to the earthquake catalogues

as to obtain the reliable earthquake data. The processes are as
follows:
1. Choosing a consistent magnitude for SHA, and then the other
magnitude scales are converted to this magnitude scale. In
this research, a moment magnitude, Mw, was chosen as a
measurement to quantify the size of the earthquake. Other
types of magnitude in the catalogues were then converted to
Mw by using an appropriate formula.
2. Declustering the earthquake data in the catalogues for
separating main shock and accessory shock events in order to
obtain independent earthquake data. The analysis was
performed using time and distance windows criteria
proposed by Gardner and Knopoff (1974). The algorithm
eliminated more than 50% of earthquake data.
3. Adjusting the length of the statistical time window in order
to improve estimates of parameters using completeness
analysis method proposed by Stepp (1973). Based on the
result, the earthquake events with magnitude less than 6.0 are
completely reported only during the most recent 41 year

interval or since 1964, and for events greater than or
equivalent 6.0 are completely reported for the whole of 104year sample interval.
SEISMIC SOURCES MODELS
Identification and characterization seismic sources are the first step
to analyze seismic hazard. In this stage, seismic source zones are
identified including all potential seismic sources capable of
generating significant ground motion at the site. A seismic source
represents a region of the earths crust where the characteristics of
earthquake activity are recognized to be relatively different than
those of the adjacent crust.
The source geometries were constructed based on earthquake
spatial distributions and regional tectonic setting of East Malaysia.
Generally, the seismic source model can be divided into three
classifications, i.e. North Sulawesi Subduction (NSS) zone,
shallow crustal (SC) zone, and background zone.
The seismic source models are divided into 15 fault sources and
one background source. The fault sources are North Sulawesi
Subduction zone (zone 1, 2), Palu-Koro (zone 3, 4, 5, 6), Sulu
(zone 7), Tarakan Basin (zone 8), Kutai-Mahakam Basin (zone 9),
Walanae (zone 10, 11, 12), Melange (zone 13, 24), and Mamuju
(zone 15). The background source was used in this analysis to
accommodate uncertainty associated with unknown faults around
site region.
The seismic source model used in this study can be seen in
Figure 1.6. The fault source zones within radius of 500 km from
East Malaysia are considered in the analysis. At farther radius,
amplitudes of incoming seismic shear waves can be considered
small in affecting the site location.
Figure 1.6 Seismic source models for East Malaysia.
The NSS was investigated by plotting the spatial distribution of

earthquakes around Sulawesi using the regional seismicity data
recorded since 1900. A plot of seismicity around NSS can be seen
in Figure 1.7. Due to the sparse of epicentral data, NSS is divided
into two sections i.e. zone 1 and zone 2.
The hypocentral profile of NSS is shown in Figures 1.8. The
profiles reveal downward dipping zones of seismicity that mark the
subsurface location of the Sangihe plate. Based on the profile,
earthquake sources for North Sulawesi can be modelled as an
interface and intraslab source zones with 7o and 61o of dip angles,
respectively. Based on these dip angles and the focal depth of
focus, NSS were divided into Megathrust zone (interface) and
Benioff zone (intraslab). This model is consistent with the model
of Sangihe Plate as presented by Cardwell et al (1980).
10
Figure 1.7 Earthquake spatial distributions
Figure 1.8: The hypocentral profiles of NSS
11
SEISMIC HAZARD PARAMETERS

Characterization of seismicity at particular site or region is
commonly expressed in seismic hazard parameters. Seismic hazard
parameters are needed for fully describing earthquake activity
within the earth crust in a certain region. There are three
parameters that are most commonly considered in seismic hazard
assessment, i.e. a-b parameter, recurrence rate, and maximum size
of future earthquakes for each source. Usually, temporal
distribution of earthquakes is assumed to follow frequencymagnitude relationship proposed by Gutenberg-Richter (G-R)
(1954). Three methods for assessing seismic hazard parameters
were used in this research; i.e., Least Square (LS), Weichert
(1980), and Kijko & Sellevoll (KS) (1989). The results show that
the seismicity b-values in this study are within the range 0.7 to 1.2
which can be considered in the normal range of b-values. The
results are shown in Figure 1.9.
ATTENUATION RELATIONSHIPS
One of the critical factors in seismic analysis is to obtain or to
select appropriate attenuation relationship. This formula, also
known as ground motion relation, is a simple mathematical model
that relates a ground motion parameter (i.e. spectral acceleration,
velocity and displacement) to earthquake source parameter (i.e.
magnitude, source to site distance, mechanism) and local site
condition (Campbell, 2003). Since there is no attenuation function
derived for Malaysia region, therefore several attenuation
relationships that consider appropriate according to mechanism
that likely to occur in Malaysia region from previous researcher
were selected. There has been a number of attenuation relations
derived in the last two decades since the record of ground motions
becomes more available. In general, they are categorized according
to tectonic environment (i.e. subduction zone and shallow crustal
earthquakes) and site condition.
12
1.00E+01
1.00E+01
Data
Least Square
1.00E+00
Weichert
1.00E+00
KS
1.00E-01
N(M > m o)
N (M > m o)
1.00E-01
1.00E-02
1.00E-02
1.00E-03
Data
1.00E-03
Least Square
1.00E-04
Weichert
KS
1.00E-04
1.00E-05
1.00E-06
1.00E-05
5.0
6.0
7.0
8.0
5.0
5.5
6.0
Magnitude
(a) Palu-Koro
1.00E+00
1.00E-01
1.00E-01
1.00E-02
1.00E-02
N(M > m o)
1.00E+00
N(M > m o)
1.00E+01
1.00E-04
Data
7.5
8.0
1.00E-03
1.00E-04
Least Square
1.00E-05
7.0
b) Other shallow crustal
1.00E+01
1.00E-03
6.5
Magnitude
Weichert
Data
1.00E-05
KS
Least Square
Weichert
1.00E-06
1.00E-06
1.00E-07
KS
1.00E-07
5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
Magnitude
(c) N-S Megathrust
5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
Magnitude
d) N-S Benioff
Figure 1.9 Recurrence relationship chart for East Malaysia
13
There are several attenuation relationships derived for

subduction zone earthquake, which are commonly used, such as
Crouse (1991), Youngs (1997), Atkinson and Boore (1997),
Petersen (2004), whereas attenuation relationships, which were
developed by Campbell (1997, 2003), Sadigh et al. (1997), Toro
(1997), are frequently used to estimate ground motion for shallow
crustal earthquakes.
Most of the attenuation functions were developed using
empirical method. Therefore, the limitation of the function will
depend on the quality of strong motion data such as quantity and
the distribution of parameters of attenuation function such as
magnitude, depth, distance and peak acceleration. Usually
attenuation relationships are derived for near source earthquakes;
consequently, most of the attenuation relationships have a distance
limit.
In this study, three attenuation functions are used in seismic
analysis. Campbell (2003) attenuation is used for calculating
distant earthquakes while Sadigh et al. (1997) and Boore et al.
(1997) for short distance earthquakes (less than 100 km).
LOGIC TREE
Logic trees (Power et al., 1981; Kulkarni et al., 1984; Coppersmith
and Youngs, 1986) are used in this study in order to allow
uncertainty in selection of models for attenuation, recurrence rate,
and maximum magnitude to be considered. In this study,
attenuation model of Campbell (2003), Sadigh et al (1997), and
Boore et al (1997) are assigned a relative likelihood of 0.33 each.
The recurrence rates calculated according to the method of Least
Square, Kijko & Sellevoll (1989) and that of Weichert (1980) are
considered equally likely to be correct. At final level, different
relative likelihoods are assigned to the maximum magnitude. The
logic tree model is illustrated in Figure 1.10.
14
Figure 1.10 Simple logic tree for incorporation of model uncertainty
THE RESULTS OF ANALYSIS

The hazard calculations were performed for several combinations
of attenuation functions, seismic hazard parameters, and maximum
magnitudes using total probability theory as proposed by Cornell
(1968). The result of each analysis is then multiplied by the
relative likelihood of its combination of branches in the logic tree.
In order to develop macrozonation maps of East Malaysia, the
procedure is then applied for every grid (or site) spacing of 0.5
degrees in latitude and longitude around East Malaysia. The results
can be seen in Figures 1.11 to 1.12.
15
Figure 1.11 Peak ground acceleration (PGA) maps for 500 year
Figure 1.12 Peak ground acceleration (PGA) maps for 2,500 year
16
CONCLUSION
Two macrozonation maps representing 10% and 2% probability of
exceedance (PE) in 50 years ground motions for East Malaysia
have been developed in this study. The results show the ground
motions across the East Malaysia range between 60 and 120 gals
and between 160 and 220 gals for 10% and 2% PE in 50-year
hazard levels, respectively. The hazard levels show the peak
ground acceleration contours increase from the west to the east of
East Malaysia.
REFERENCES
Atkinson, G.M. and Boore, D.M. (1997). Some Comparisons
Between Recent Ground Motion Relations. Seismological
Research Letters. Vol. 68. No. 1.
Bernama. (2004). Earthquake Tremors Felt in Miri, Bintulu.
Malaysian News National Agency. 2 May 2004.
Boore, D.M., Joyner, W.B., and Fumal, T.E. (1997). Equation for
Estimating Horizontal Response Spectra and Peak Acceleration
from Western North America Earthquakes: A Summary of
Recent Work. Seismological Research Letters, Vol. 68, No. 1,
January/February 1997, pp. 128-153.
Brouwer, H.A. (1947). Geological explorations in Celebes
summary of results. p. 1-64. In Brouwer, H.A. (ed.),
Geological explorations in the island of Celebes. North
Holland Publishing Company, Amsterdam.
Campbell, K.W. (1997). Empirical Near-source Attenuation
Relationships for Horizontal and Vertical Components of Peak
Ground Acceleration, Peak Ground Velocity and Pseudoabsolute Acceleration Response Spectra. Seismological
Research Letters. Vol. 68.
Campbell, K.W. (2003). Prediction of strong ground motion using
the hybrid empirical method and its use in the development of
ground-motion (attenuation) relations in Eastern North
17
America. Bulletin of the Seismological Society of America.

Vol. 93, pp. 10121033.
Cardwell, R.K.I., Isacks, B.L., and Karig, D.E. (1980). The Spatial
Distribution of Earthquakes, Focal Mechanism Solutions and
Subducted Lithosphere in the Philippine and Northeastern
Indonesian Islands. In Hayes, D.E. (ed) The Tectonic and
Geology Evolution of South East Asian Seas and Islands.
American Geophysical Union Monograph, 23, 1-35.
Coppersmith, K.J., and Youngs, R.R. (1986). Capturing
Uncertainty in Probabilistic Seismic Hazard Assessments with
Intraplate Tectonic Environments. Proceeding, 3rd U.S.
National Conference on Earthquake Engineering. Charleston,
South Carolina. Vol. 1, pp. 301-312.
Cornel, C.A. (1968). Engineering Seismic Risk Analysis. Bulletin
of the Seismological Society of America. Vol 58, No. 5: 15831606.
Crouse, C.B. (1991). Ground Motion Attenuation Equation for
Earthquake on the Cascadia Subduction Zone. Earthquake
Spectra, Vol. 7, No. 2.
Gardner, J.K., and Knopoff, L. (1974). Is the Sequence of
Earthquakes in Southern California, with Aftershocks removed,
Poissonian? Bulletin of the Seismological Society of America,
Vol. 64, No. 5, 1974, pp. 1363-1367.
2
MICROZONATION STUDY FOR
PUTRAJAYA, MALAYSIA
Azlan Adnan
Hendriyawan
Aminaton Marto
Irsyam, M.
INTRODUCTION
Earthquake is one of the most devastating natural disasters on the
earth. Generally, the effects of strong earthquakes are caused by
ground shaking, surface faulting, liquefaction, and less commonly,
by tsunamis. Although it is impossible to prevent earthquakes from
happening, it is possible to mitigate the effects of strong
earthquake shaking and to reduce loss of life, injuries and
damages. The most effective way to reduce disasters caused by
earthquakes are to estimate the seismic hazard and to disseminate
this information for used in improved building design and
construction so that the structures posses adequate earthquake
resistant capacity.
Geotechnical factors often exert a major influence on damage
patterns and loss of life in earthquake events. For example, the
localized patterns of heavy damage during the 1985 Mexico City
and 1989 Loma Prieta earthquakes provide illustrations of the
importance of understanding the seismic response of deep clay
deposits and saturated sand deposits. The pronounced influence of
local soil conditions on the characteristics of the observed
earthquake ground motions also can be seen during 1957 San
Francisco Earthquake. Even in one city, however, building
20
response and damage were varying significantly due to variation of

soil profiles in the city. In other countries, several attempts have
been made to identify their effects on earthquake hazards related to
geotechnical factors in the form of maps or inventories. Mapping
of seismic hazard at local scales to incorporate the effects of local
geotechnical factors is called microzonation.
Microzonation for seismic hazard has many uses as mentioned
by Finn et al. It can provide input for seismic design, land use
management, and estimation of the potential for liquefaction and
landslides. It also provides the basis for estimating and mapping
the potential damage to buildings.
Putrajaya is a planned city and a federal territory that acts as a
federal government administration centre of Malaysia. Therefore,
there are a lot of investments and assets that should be protected
against earthquake hazard in Putrajaya such as high rise and
monumental buildings. This paper presents the results of
microzonation study for developing microzonation maps for
Putrajaya.
GROUND RESPONSE ANALYSIS
In this study, ground response analysis was performed using onedimensional shear wave propagation method (1-D analysis). 1-D
method is based on assumption that all boundaries are horizontal
and that the response of a soil deposit is predominantly caused by
shear wave propagating vertically from the underlying bedrock.
Although the soil layers are sometimes inclined or bent, they are
regarded as horizontal in most cases. Furthermore, the length of a
layer is infinite compared with its thickness. It is thus practical to
model them as 1-D horizontal layers. Analytical and numerical
procedures based on this concept, incorporating linear
approximation to nonlinear soil behavior, have shown reasonable
agreements with field observations in a number of cases.
The ground response analysis should consider the nonlinearity
of soil behavior to provide reasonable results. There are two
21
approaches to include the effect of nonlinearity of soil material into

the analysis: equivalent linear and nonlinear approaches.
Equivalent linear models imply that the strain will always return to
zero after cyclic loading, and since a linear material has no limiting
strength, failure cannot occur. The nonlinear of soil behaviors are
approximated by determining the values that consistent with the
level of strain induced in each layer.
The equivalent linear approach is incapable of representing the
changes in soil stiffness those actually occur during the earthquake.
It also means that it cannot be used directly for problems involving
permanent deformation or failure. An alternative approach is to
analyze the actual nonlinear response of a soil deposit using direct
numerical integration in the time domain. The advantages of
nonlinear method are: (1) the stiffness of an actual nonlinear soil
changes over the duration of large earthquake, such high
amplification levels that occur in equivalent linear approach, will
not develop in the field; and (2) nonlinear method can be
formulated in terms of effective stresses to allow modeling of the
generation, redistribution, and eventual dissipation of excess pore
pressure during and after earthquake shaking.
In this study, the ground response analyses were performed
using nonlinear approach. The analyses were carried out using
program NERA, which stands for Nonlinear Earthquake Response
Analysis. This program use soil model proposed by Iwan and Mroz
to model nonlinear stress-strain curves of soil.
DYNAMIC SOIL PROPERTIES
Ground response analysis requires profile of dynamic soil
parameters such as maximum shear modulus, Gmax or shear wave
velocity, VS and damping, . This parameter can be obtained from
field dynamic tests or by converting from static field tests using
empirical formula. Numerous researchers have investigated the
relationship between maximum shear modulus or shear wave
velocity and N-values of Standard Penetration Test (SPT). Most of
22
the studies were performed in the 1970s in Japan. Since then,

some similar studies have been reported in the United States. Some
of the correlations were compiled by Barros.
In this research, the static parameters from SPT test were
converted into VS by using formula proposed by Ohta & Goto and
Imai & Tonouchi. In order to verify the formulas, the seismic
down-hole tests were performed on several locations and the
results were then compared to the results from empirical
correlations. A figure 2.1 shows the comparison of shear wave
velocity, VS, obtained from empirical correlation and seismic
down-hole tests. It can be seen in the figure, the empirical
formulas are relatively reliable to predict VS from NSPT values.
SITE CLASSIFICATION
Site classification analyses were performed by using 14 soil data in
Putrajaya. For each data, the soil dynamic properties are calculated
by using formulas proposed by Ohta & Goto and Imai &
Tonouchi. The results were summarized in Figure 2.2. The
classification of a particular site was determined by referring three
specifications: 1997 UBC/2000 IBC, Eurocode 8, and Bray and
Rodriguez-Marek. Based on the existing data, the soil in Putrajaya
can be classified as SD and SE in accordance with 2000 IBC as
shown in Table 2.1.
23
Figure 2.1 The result of seismic down-hole test
100
VS (m/sec.)
200
VS (m/sec.)
300
400
0.0
0.0
10.0
10.0
Depth (m)
Depth (m)
20.0
200
300
20.0
30.0
30.0
Site Class SD
40.0
100
Site Class SE
40.0
Figure 2.2 Soil dynamic properties for Putrajaya
400
24
RESULTS OF 1-D ANALYSIS

Shear wave propagation analyses were performed for all existing
soil data to obtain peak acceleration and amplification factor at the
surface. Two hazard levels were used in the analysis to represent
10% and 2% Probability Exceedance (PE) in design time period of
50 year or correspond to return period of approximately 500 and
2,500 years, respectively. These hazard levels were calculated
using total probability theorem as proposed by Cornel. Based on
our previous study, the peak ground accelerations for Putrajaya are
0.073g (73.4 gal) and 0.149g (149 gal) for 500 and 2,500 year
return periods of ground motions, respectively. The seismic hazard
map of Peninsular Malaysia for those two hazard levels can be
seen in Figure 2.3.
Table 2.1 Soil Classification of Putrajaya
No.
Location
VS (m/s)
Tn
(sec.)
Soil Classification
2000
IBC
[15]
EC8
[16]
BR 1997
[17]
PJ-1
213.40
0.47
C-2
PJ-2
167.12
0.73
C-3/E-1
PJ-3
278.68
0.35
C-1
PJ-4
267.36
0.49
C-2
PJ-5
345.75
0.29
C-1
PJ-6
278.20
0.48
C-2
PJ-7
211.00
0.58
C-3/E-1
PJ-8
193.78
0.60
C-3/E-1
PJ-9
195.54
0.49
C-2
10
PJ-10
11
PJ-11
214.75
304
0.77
0.39
D
D
C
C
C-3/E-1
C-2
12
PJ-12
205
0.63
C-3/E-1
13
PJ-13
322
0.36
C-2
14
PJ-14
341
0.59
C-3/E-1
25
TR=500 yr
TR=2500 yr
Figure 2.3 Seismic hazard maps of Peninsular Malaysia (site class SB)
Four time histories were used in the analysis: Synth-1, Synth-2,

Synth-3, and Synth-4. Synth-1 and Synth-2 represent ground
motion for 500 years return period, while Synth-3 and Synth-4
26
Acceleration (g
represent for 2,500 years return period. The time histories used in
the analysis can be seen in Figure 2.4.
0.08
0.06
0.04
0.02
0
-0.02
-0.04
-0.06
-0.08
a) synth-1
TR=500 yr
50
100
150
200
Time (sec)
0.1
Acceleration (g
b) synth-2
TR=500 yr
0.05
0
-0.05
-0.1
0
10
20
30
40
50
Time (sec)
0.2
Acceleration (g
c) synth-3
0.1
TR=2500 yr
0
-0.1
-0.2
0
25
50
75
100
125
150
Acceleration (g
Time (sec)
0.2
0.15
0.1
0.05
0
-0.05
-0.1
-0.15
d) synth-4
20
TR=2500 yr
40
60
Time (sec)
80
100
Figure 2.4 Time histories used in ground response analysis
27
The results of acceleration and amplification factors at surface

of Putrajaya were summarized in Tables 2.2 and 2.3, respectively.
The amplification factors show the ratio between acceleration at
bedrock and at surface. Generally, the amplification factors for 500
years return period are higher than 2,500 years return period. The
effects of using different time histories can be seen in Figures 2.5
and 2.6 for 500 and 2,500 years return periods of ground motions,
respectively. The results indicate that the selection of appropriate
time histories is one of the most critical in ground response
analysis.
Table 2.2 Results of 1-D analyses for Putrajaya

No.
Location
PJ-1
PSA (g's)
Soil
Type
Synth-1
Synth-2
Synth-3
Synth-4
SD
0.168
0.143
0.236
0.270
PJ-2
SE
0.144
0.128
0.297
0.260
PJ-3
SE
0.201
0.185
0.348
0.337
PJ-4
SD
0.197
0.162
0.366
0.337
PJ-5
SD
0.174
0.168
0.319
0.315
PJ-6
SD
0.134
0.118
0.182
0.193
PJ-7
SD
0.160
0.141
0.250
0.315
PJ-8
SD
0.173
0.153
0.225
0.285
PJ-9
SD
0.157
0.149
0.213
0.258
10
PJ-10
SD
0.137
0.129
0.234
0.225
11
PJ-11
SD
0.173
0.119
0.271
0.297
12
PJ-12
SD
0.163
0.147
0.281
0.288
13
PJ-13
SD
0.175
0.129
0.270
0.307
14
PJ-14
SD
0.155
0.146
0.186
0.196
Response spectra at the surface for Putrajaya can be seen in

Figures 2.7 to 2.8. The predominant periods of the spectra
generally occur in the range 0.2 - 0.8 second. It also can be seen in
the figures, the frequency content of the spectrum is relatively not
28
much different. In other hand, the figures show that the frequency
content of the spectrum is more affected by the stiffness of the soil.
According to the figures, soft soil deposits produce greater
proportions of long period (low frequency) motions than stiff soil.
Table 2.3 Results of 1-D analyses for Putrajaya (cont.)
No.
Location
Amplification factor
Soil
Type
Synth-1
Synth-2
Synth-3
Synth-4
SD
2.30
1.95
1.58
1.81
PJ-1
PJ-2
SE
1.98
1.75
1.99
1.75
PJ-3
SE
2.76
2.54
2.33
2.26
PJ-4
SD
2.69
2.22
2.46
2.26
PJ-5
SD
2.39
2.29
2.14
2.11
PJ-6
SD
1.84
1.61
1.22
1.29
PJ-7
SD
2.20
1.93
1.68
2.12
PJ-8
SD
2.36
2.10
1.51
1.91
PJ-9
SD
2.15
2.04
1.43
1.73
10
PJ-10
SD
1.87
1.77
1.57
1.51
11
PJ-11
SD
2.37
1.63
1.82
2.00
12
PJ-12
SD
2.24
2.01
1.89
1.93
13
PJ-13
SD
2.40
1.77
1.81
2.06
14
PJ-14
SD
2.12
2.00
1.25
1.32
Acceleration (g's)
0.05
0.08
0.10
0.13
0.15
Acceleration (g's)
0.18
0.20
0.05
0.0
35.0
Synth-1
Synth-2
Depth (km)
Depth (km)
0.15
0.18
10.0
Site Class SD
20.0
30.0
0.13
5.0
10.0
25.0
0.10
0.0
5.0
15.0
0.08
15.0
Site Class SE
20.0
25.0
30.0
35.0
Figure 2.5 1-D analysis for 500 years return period
Synth-1
Synth-2
0.20
29
Acceleration (g's)
0.10
0.15
0.20
Acceleration (g's)
0.25
0.30
0.10
0.0
0.20
0.25
0.30
0.0
Site Class SD
5.0
5.0
10.0
Site Class SE
10.0
Depth (km)
Depth (km)
0.15
15.0
20.0
25.0
15.0
20.0
25.0
Synth-3
Synth-4
30.0
Synth-3
Synth-4
30.0
35.0
35.0
Figure 2.6 1-D analysis for 2500 years return period
The results of site response analysis at several points were used

to develop contour map of surface acceleration and amplification
factor for 500-years and 2,500-years return periods. The isoacceleration contour maps for Putrajaya are shown in Figures 2.9
to 2.10, while the contour of amplification factors can be seen in
Figures 2.11 to 2.12.
According to the figures, the accelerations at the surface of
Putrajaya range between 13% g (130 gal) and 19% g (190 gal) for
10% PE in 50-year hazard levels and between 22% g (220 gal) and
34% g (340 gal) for 2% PE in 50-year hazard levels. The
amplification factors for those two hazard levels range between 1.5
and 2.6. Generally, the peak accelerations and amplifications
factors contours occur around Precinct 4.
1.000
Spectral Acceleration (g)
1.00
0.10
Synth-1
0.100
Synth-1
Synth-2
Synth-2
0.010
0.01
0.0
1.0
2.0
Period (second)
a) Site Class SD
3.0
4.0
0.0
1.0
2.0
3.0
4.0
Period (second)
b) Site Class SE
Figure 2.7 Response spectra at surface for 500 years return period
30
10.000
10.00
1.00
0.10
Synth-3
Synth-4
1.000
0.100
Synth-3
Synth-4
0.010
0.001
0.01
0.0
1.0
2.0
3.0
4.0
0.0
1.0
2.0
Period (second)
Period (second)
a) Site Class SD
b) Site Class SE
3.0
4.0
Figure 2.8 Response spectra at surface for 2500 years return period
a) Synth-1
(b) Synth-2
Figure 2.9 Contour of acceleration at surface for 500 years return period
(PGA=0.073 g)
31
a) Synth-3
(b) Synth-4
Figure 2.10 Contour of acceleration at surface for 2500

years return period (PGA=0.149 g)
a) Synth-1
(b) Synth-2
Figure 2.11 Contour of amplification factor for 500 years return period
(PGA=0.073 g)
32
a) Synth-3
(b) Synth-4
Figure 2.12 Contour of amplification factor for 2500 years return period
(PGA=0.149 g)
SUMMARY AND CONCLUSION

This paper has described the microzonation study for Putrajaya in
Peninsular Malaysia. Ground response analyses were performed
using 1-D shear wave propagation analysis. The analysis was
performed for two hazard levels that represent 500 and 2,500 years
return periods of earthquake. Four time histories were used in the
analysis to represent ground motion for 500 years (Synth-1 and
Synth-2) and 2,500 years (Synth-3 and Synth-4) return periods. In
this study, the analysis was performed using nonlinear approach in
order to consider the actual nonlinear response of a soil deposit.
The results of site response analysis at several points were used to
develop microzonation maps of Putrajaya for 500 and 2500-years
return periods. Four microzonation maps were produced in this
research that can be used as input for seismic design, land use
management, and estimation of the potential for liquefaction and
landslides.
33
The results of ground response analysis also show that both of

time histories and local soil conditions (soil properties and
stratigraphy) are critical to the results of ground response analysis.
Generally, time histories affect the amplitude of spectral
acceleration, whilst the soil conditions (stiffness, stratigraphy,
ground water level) influence the frequency content of the
spectrum. Therefore, these two subjects should be considered and
determined carefully in ground response analyses.
REFERENCES
Hu, Y.X. (1996). Earthquake Engineering. London: E & FN Spon.
Bray, J. D., Seed, R. B., Cluff, L. S., Seed, H. B. (1994).
Earthquake Fault Rupture Propagation through Soil. Journal
of Geotechnical Engineering, American Society of Civil
Engineers. Vol. 120, No. 3: 543-561.
Seed, H.B., Chaney, R.C., and Pamucku, S. (1991). Foundation
Engineering Handbook. New York: Van Nostrand Reinhold.
Finn, W.D.L., Onur, T., Ventura, C.E. (2004). Microzonation:
Developments and Applications. In: Ansal, A ed. Recent
Advances in Earthquake Geotechnical Engineering and
Microzonation. Netherlands: Kluwer Academic Publisher. 326.
Kramer, S. L. (1996). Geotechnical Earthquake Engineering. New
Jersey: Prentice-Hall.
Bardet, J.P. and Tobita, T. (2001). NERA-A Computer Program
for Nonlinear Earthquake site Response Analyses of Layered
Soil Deposits. Department of Civil Engineering University of
Southern California.
Iwan, W. D. (1967). On A Class of Models for The Yielding
Behavior Of Continuous And Composite Systems. Journal of
Applied Mechanics, ASME. Vol. 34: 612-617.
Mrz, Z. (1967). On The 'Description of Anisotropic Work
hardening. Journal of Mechanics and Physics of Solids. Vol.
15: 163-175.
34
Barros, J.M.C. (1994). Factor Affecting Dynamic Properties of

Soil. Report, University of Michigan.
Ohta, Y. and Goto, N. (1978). Empirical shear wave velocity
equations in terms of characteristic soil indexes. Earthquake
Engineering Structural Dynamic. Vol. 6:167-187.
Imai, T. and Tonouchi, K. (1982). Correlation of N-value with SWave Velocity and Shear Modulus. Proceeding, 2nd
European Symposium on Penetration Testing. Amsterdam.
Pp. 57-72.
Ohsaki, Y. and R. Iwasaki. (1973).On dynamic shear moduli and
Poisson's ratio of soil deposits. Soils and Foundations. Vol.
13, No.:61-73.
Seed, H.B., I.M. Idriss, and I. Arango. (1983). Evaluation of
Liquefaction Potential Using Field Performance Data. J.
Geotech. Eng., ASCE. Vol. 109, No. 3: 458-482.
Building Seismic Safety Council. (1998). 1997 Edition NEHRP
Recommended Provisions for the Development of Seismic
Regulations for New Buildings, FEMA 302/303. Part 1
(Provisions) and Part 2 (Commentary). Developed for the
Federal Emergency Management Agency. Washington, DC.
International Code Council. (2000). International Building Code
2000. International Code Council. International Conference of
Building Officials. Whittier, CA, and others.
3
VULNERABILITY STUDY OF PUBLIC
BUILDINGS SUBJECTED TO
EARTHQUAKE BY ATC-21, ATC-22 AND
FINITE ELEMENT MODELING
Mohd Zamri Ramli
Azlan Adnan
Suhana Suradi
INTRODUCTION
ATC21 and ATC22 are handbooks in the FEMA series: (Rapid
Visual Screening of Buildings for Potential Seismic Hazard: A
Handbook, (ATC, 1988). The handbook has descriptions and
illustrations of building types that will be helpful to the engineer in
determining the age, building type, and the need for evaluation of
the building. The purpose of evaluation is to know how the
building performed. In ATC21, the evaluations are performed fast
and relatively inexpensive (without detailed analysis for potential
hazardous buildings). There is no structural analysis calculations
are performed in this evaluation (Figure 3.1). The evaluation only
takes 15-30 minutes per building includes inspection, data
collection, and decision-making process.
In ATC22, the methodology is centred on a set of question one
set for each of fifteen model building types that is designed to
uncover the flaws and weaknesses of the building. The engineer
addresses each statement and determines whether it is true or false.
True statements identify conditions that are acceptable. False
Statements identify issues or concerns that need further
investigation. In dealing with the statements that have been found
36
to be false, the further analyses are required to identify the

problem.
Figure 3.1 ATC-21 form for Mahkamah Labuan
37
MODELING AND ANALYSIS

In order to analyze the seismic performance of the buildings, a
single main frame was chosen for modelling in finite element
analysis. The finite element software, SAP2000 was used to
analyse the structures through a linear dynamic analysis.
The analyses were carried for four intensities of seismic load,
i.e. 0.05g, 0.10g, 0.15g, and 0.20g. Three analysis methods were
applied for the model, i.e. Free Vibration Analysis, Response
Spectra Analysis and Time Histories Analysis. The response
spectra analyses used UBC94 code for S2 soil. Data recorded from
El Centro Earthquake in 1940 was used in Time histories analysis.
From the Free Vibration Analysis, the natural period (s),
frequency (Hz), angular frequency, and mode shapes were
determined. For each building, four mode shapes were analyzed.
From Response Spectra Analysis, maximum axial force, shear and
maximum moment were determined. The results were compared to
static analysis to define the differences between static and dynamic
analyses. The maximum allowable axial force, shear and moment
for each element of the buildings were compared with the
maximum design capacity to determine the capability of each
building to retain different capacity of earthquake load.
LOAD CASES
i) Static Linear Analysis
Load Case
1. LC1 WL1,2,3,4 = Dead + Live + Wind 1,2,3,4
2. LC2 = Static EQ
3. LC1 LC2 WL1,2,3,4 = Dead + Live + Static EQ + Wind
1,2,3,4
Wind load (CP3: Chapter V: Part 2)
Basic Wind Speed
38
V1 = WL1 = 20 m/s
V2 = WL2 = 30 m/s
V3 = WL3 = 40 m/s
V4 = WL4 = 50 m/s
Design Wind Speed, Vs = V. S1. S2. S3
Where :
V = Basic wind speed
S1 = Topography factor
S3 = Factor from ground roughness, building size,
and height above ground
S3 = Statistical factor
Dynamic pressure, q = 0.613 Vs 2
(N/m2 and m/s)
Static Earthquake (UBC 97)

Seismic Dead Load, W
W = Total structure dead load + partition + 25% floor live load
Building Period, T
3
Ta = C t hn 4
where:
0.035 for steel moment frames
Ct = 0.030 for concrete moment frames
0.030 for eccentric braced frames
0.020 for all other buildings
hn = the height of the building in feet
Design Base Shear, V
V = max { min [ 2.5 Ca.I.W / R , Cv.I.W / R.T ], 0.11Ca.I.W,
0.82 Z.Nv.I.W / R }
Where : W = Seismic dead load
T = Building period
I = Seismic Importance Factor (UBC Tab 5-1)
R = Response Modification Coef. (UBC Tab 5-2)
Cv = Numerical coefficient dependent on the soil
conditions at the site and regional seismicity
39
(UBC Tab 5-3)

Ca = Another seismic coefficient dependent on the
soil conditions at the site and regional
seismicity (UBC Tab 5-4)
Soil Profile Types , S1, S2 and S3, (UBC Tab 5-5)
Seismic sources type , (UBC Tab 5-6)
Z = Seismic zone factor (Figure - Map Malaysia)
Na = Near sources factor (UBC Tab 5-7)
Nv = Near sources factor zone 4 (UBC Tab 5-8)
Distribution of Lateral Force, Fx
(V Ft )( wx hx )
Fx =
n
wi hi
i =1
If T > 0.7 sec. : Ft = 0.07TV 0.25V

If T 0.7 sec. : Ft = 0.0
ii. Dynamic Linear Analysis
Load Case
LC3 = Dead Load (slab) + Spec. 0.05g, 0.10g, 0.15g, 0.20g
Figure 3.2 Response Spectrum Analyses (RSA) UBC 94
40
iii. Dynamic Time History Analysis:

Load Case
LC4 = Dead Load + Live Load + TH 0.05g, 0.10g, 0.15g,
0.20g Time History Analysis (THA) (May 18,1940
El Centro-Figure 3.3)
Figure 3.3 Time History Analysis (THA) (May 18,1940 El Centro)
STUDY CASE - MAHKAMAH, LABUAN

Mahkamah Labuan is an 11 story concrete building was located at
Labuan city. Based on built drawing document, the building is a
concrete moment resisting frame (Type 8). The building consists of
government offices. The elevation of the frame can be seen in
Figure 3.4.
FREE VIBRATION ANALYSIS (FVA)
The mode shapes and time period of the building come from the
Free Vibration Analysis (FVA). Table 3.1 shows the buildings
natural period and frequency for the free vibration analysis. The
first 4 modes are selected in this analysis.
41
Figure 3.4 Frame model of Mahkamah Labuan building
Table 3.1 Natural period and frequency of Mahkamah Labuan

Mode
Shape
1
2
3
4
Time Period,
T (sec.)
0.52516
0.15919
0.09709
0.06958
Frequency,
f (Hz)
1.90418
6.28180
10.29972
14.97195
Angular
Frequency,
11.96433
39.46972
64.71506
90.3016
Referring to Table 3.1, the natural period for mode shape 1 (0.
52516s) is the highest among the other mode shapes and will be
used for further analysis. Mode shape 2 indicates 0. 15919s, while
mode shape 3 and 4 indicate 0.09709s and 0.06958s respectively.
According to UBC formula (T=Cthn3/4), where Ct = 0.030 for
concrete moment frames, and hn = the height of the building in
feet, natural period of the Mahkamah Labuan building is equal to
0.76s by calculation.
The time period for this building is 0.52516 sec (Mode Shape 1).
The maximum spectral acceleration curve suitable on that site
comes from Palu Koru Fault, Philippine. Figure 3.5 shows that, the
design acceleration for this building is equivalent to be 0.09g.
42
Figure 3.5 Spectral Acceleration for building Mode Shape1 Time Period
(Palu Koro Fault)
Table 3.2 Dynamic amplification factor
Note :
D < 1 No response
D = 1 Static
D > 1 Dynamic
Table 3.2 shows the dynamic amplification factor for Mahkamah

Labuan based on previous Tawau and Bintulu earthquake. It
clearly shows that the analysis of the building needs further
analysis on dynamic. This is because the dynamic amplification
factor exceeds the static value (D=1), which is 4.46 for Tawau and
1.73 for Bintulu respectively
43
STATIC AND DYNAMIC LINEAR ANALYSIS

Shear and moment capacity for beams in Mahkamah Labuan is
equal to 182 kN and 238 kNm respectively. Referring to the output
analysis as been summarized in Figure 3.6, the maximum shear
force, Vmax in beam under static analysis had exceeded the beam
capacity itself. Under load case 1 (LC1) and wind load (20-50
km/hr), Vmax ranges from 847-875 kN. These values increased
when LC1 and load case 2 (LC2) being considered together with
the same wind load. For these cases Vmax ranges from 1,1031,130 kN. Besides, for the all load case, maximum moment, Mmax
has also exceeded the moment capacity. Mmax ranges from 1,2152,219 kNm. It shows that under all cases, the beam is not sufficient
to resist the maximum shear force and moment under static
loading.
(a)
Static Linear Analysis on Beam
600
Shear, V(kN)
500
400
300
200
100
0
0
20
40
60
Basic Wind Speed, v (m/s)

DL+LL+WL
(b)
DL+LL+EQ+WL
Shear Capacity
Static Linear Analysis on Beam
Moment, M (kNm)
2500
2000
1500
1000
500
0
0
20
40
60
Basic Wind Speed, v (m /s)

DL+LL+WL
DL+LL+EQ+WL
Moment Capacity
Figure 3.6 Shear (a) and Moment (b) at Beam Element (Static Load)
44
(a)
Response Spectrum Analysis of Beam

300.00
Shear, V (kN)
250.00
200.00
150.00
100.00
0.14
50.00
0.00
0.00
0.05
0.10
0.15
0.20
0.25
Intensity (g)
Max. Shear
(b)
Shear Design Capacity
Response Spectrum Analysis of Beam
Momen (kN-mm)
500.00
400.00
300.00
200.00
100.00
0.07
0.00
0.00
0.05
0.10
0.15
0.20
0.25
Intensity (g)
Max. Momen
Momen Design Capacity
Figure 3.7 Shear (a) and moment (b) at beam element
Design capacity can be made to find the level of safety of the

element due to the various loading. Figure 3.7 shows the level of
safety for beam in term of shear and moment due to the dynamic
linear analysis (RSA). It shows that the element can sustain shear
up to 0.14g and 0.07g for moment.
For columns in Mahkamah Labuan, axial force, shear and
moment capacity is equal to 5,712 kN, 314 kN and 860 kNm
respectively. Referring to the output analysis as been summarized
in Figure 3.8, it can be seen that the maximum axial force (Amax),
45
shear force (Vmax), and moment (Mmax) for load case 3 (LC3)
under static analysis had exceeded the column capacity. But for
load case 1 (LC1), Vmax and Mmax are still lower compare than
design capacity itself except for Amax.
(a)
Static Linear Analysis on Column
12,000
Axial, A(kN)
10,000
8,000
6,000
4,000
2,000
0
0
20
40
60

DL+LL+WL
(b)
DL+LL+EQ+WL
Axial Capacity
Static Linear Analysis on Column
800
Shear, V (kN)
700
600
500
400
300
200
100
0
0
20
40
60

DL+LL+WL
(c)
DL+LL+EQ+WL
Shear Capacity
Static Line ar Analysis on Column
1,800
Moment, M(kNm)
1,600
1,400
1,200
1,000
800
600
400
200
0
0
20
40
60
Basic W ind Spee d, v (m/s)

DL+LL+WL
DL+LL+EQ+WL
Moment Capacity
Figure 3.8 Axial Force (a), Shear (b) and Moment (c) on Column (under
Static Load)
46
Figure 3.9 shows the Amax, Vmax, and Mmax curve for column
element due to the dynamic linear analysis (RSA). It clearly
presents that; all internal forces have the lower value compare that
design capacity (for Intensity less than 0.20g).
(a)
Response Spectrum Analysis of Column

6,000.00
Axial, A (kN)
5,000.00
4,000.00
3,000.00
2,000.00
1,000.00
0.00
0.00
0.05
0.10
0.15
0.20
0.25
Intensity (g)
Axial Design Capacity
(b)
Max. Momen
Response Spectrum Analysis of Column

350.00
Shear, V (kN)
300.00
250.00
200.00
150.00
100.00
50.00
0.00
0.00
0.05
0.10
0.15
0.20
0.25
Intensity (g)
Shear Design Capacity
(c)
Max. Shear
Response Spectrum Analaysis of Column

Momen, M (kN-mm)
1,000.00
800.00
600.00
400.00
200.00
0.00
0.00
0.05
0.10
0.15
0.20
0.25
Intensity (g)
Momen Design Capacity
Max. Moment
Figure 3.9 Axial (a), shear (b) and moment (c) at column element
47
CONCLUSIONS
Based on ATC21, the study findings show that the selected
buildings on the inspection sites has Structural Score less then 2,
which indicates critical score and need further investigation. This
building are also classify by soft story building but not for highrise
building. Due to that result, the building is evaluated in detail using
ATC22.
Through ATC22, the building system and components are
investigated thoroughly. Building components are organized into
the following subsystem: vertical elements resisting horizontal
loads (i.e., moment resisting frames, shear walls and braced
frames), horizontal elements resisting lateral loads (i.e.,
diaphragms), foundations, and the connections between
subsystems. The study shows that, the buildings system is not
critical to earthquake load. The buildings show concrete
deterioration and have insufficient condition in vertical elements.
These include insufficient column tie spacing, stirrups spacing,
beam bar splices and joint eccentricities, which could be risky
during earthquake. The connections between subsystems are
adequate.
In detail, some procedures indicate that need further
investigation using computer model. The Linear static and
dynamic analysis is performed to investigate more about the
building characteristics and performance. The result shows that,
almost all cases, the beam and column is not sufficient to resist the
maximum shear force and moment under static and dynamic
loading. It is suggested that the beam element need further
investigation using non linear response spectrum analysis.
48
REFERENCES
Azlan Adnan (1998). Low Intensity Earthquake Effects on Steel
Girder Bridge. Universiti Teknologi Malaysia : Tesis Ph.D.
Balkema, A.A, Rotterdam and Brookfield (1992). Proceedings of
the tenth World Conference on Earthquake Engineering.
Madrid, Spain Volume 4.
British Standards Institusion (1982). Guide To Selection And Use
Of Elastomeric bearings For Vibration Isolation Of
Buildings~BS6177 :1982. BSI Ltd.: London
Derham, C.J. (1983) Proceedings, Natural Rubber For Earthquake
Protection Of Buildings And Vibration Isolation, Proc.,
International Conference, Kuala Lumpur. Direct Art Co: Kuala
Lumpur.
4
BUILDINGS CLASSIFICATION USING
APPLIED TECHNOLOGY COUNCIL
(ATC 21)
Mohd Zamri Ramli
Tay Tzer Yong
INRODUCTION
Every earthquake provides new lessons for the earthquake
engineering profession. Most naturally occurring earthquakes are
related to the tectonic nature of the earth. Such earthquakes are
called tectonic earthquakes. Earthquake is a shake in earth crust
that causes of a sudden high energy that release from the earth.
This energy maybe causes of a tectonic force in the earth crust.
The earth crust is made from a cold and fragile rock and it is
different from the layer under it where it is hotter than earth crust.
The crack earth crust will wreck and form to piece of rock. This
piece of rock will collide each other and cause a pressure that
produce a shake and known as earthquake. This shake will produce
a seismic wave that can be transforming to kinetic force and next
will cause damage to structure when it cant absorb or control this
kinetic energy. The method to absorb and control the kinetic force
before it damage the structure is equipped the structure with
seismic isolator.
This paper describes the ATC 21. The method whereby
buildings can be rapidly identified via a sidewalk survey from the
exterior as seismically acceptable or potentially seismically
hazardous. In generally, a Structural Score which consists of a
series of scores and modifiers based on building attributes that
50
can be seen from the street. The Structural Score S is related to the
probability of the building sustaining life-threatening damage
should a severe earthquake in the region occur. A low S score
suggests that the building requires additional study by a
professional engineer experienced in seismic design (further
investigation) and a high S score indicates that the building is
probably adequate. Finally, a location map of Putrajaya Precint 1 is
developed to show all buildings weather its need ATC 21 or
ATC 22.
PROBLEM STATEMENT
Earthquake is one of the disasters that occur in many countries and
make a serious damage especially to the structure, but also to loss
of life and destruction of property because of the collapse structure
of high-density people in it. It does happen maybe cause of the
structure didnt have any preparation and protection from
earthquake. So we have to make an preliminary analysis using
Applied Technology Council (ATC 21) to minimize the
structures damages in building.
OBJECTIVES
a. Classification for buildings uses Applied Technology
Council (ATC 21) at Putrajaya Precint 1.
b. To determine how effectiveness Applied Technology
Council (ATC 21) during earthquake against the structure
of a building.
c. To provide location map of Putrajaya Precint 1 to shows
the classifications of various types of buildings were needs
ATC 21 and ATC 22
51
METHODOLOGY
The analysis has been made using ATC 21 in the FEMA series:
(Rapid Visual Screening of Buildings for Potential Seismic
Hazard: A Handbook, (ATC, 1988). The handbook has
descriptions and illustrations of building types that will be helpful
to the engineer in determining the age, building types and the need
for evaluation of the building. In evaluation we want to know how
the building performed. In ATC 21, the evaluations are
performed fast and relatively inexpensive without detailed analysis
for potential hazardous buildings. In this evaluation is without
performing structural analysis calculations. The inspection, data
collection and decision-making process typically will occur at the
building site only takes an average of 15 to 30 minutes per building
(30 minutes to one hour if access to the interior is available).
Figure 4.1 shows the several steps in collection data, planning
and performing a rapid screening of potentially seismically
hazardous buildings for classification of building using with
Applied Technology Council (ATC 21).
RESULT AND DISCUSSION
A copy of the Data Collection Form for NEHRP Map Areas 1,2
Low is shown in Figure 4.2. The form has been designed to be
filled out in a smooth progressive manner with a minimum of
writing. A discussion of how structure hazard scores and modifiers
of the Rapid Screening Procedure method were developed is
presented in the Data Collection Form. For figure 4.3 is shows a
example of Data Collection Form for Building Block E3
Kementerian Pengajian Tinggi Malaysia. The Structural Score is
got only 1.9 marks. From the result is shown buildings Structural
Score, S is less than 2. Then the seismic performance of that
building may not be meet modern seismic criteria and the building
should be investigated further.
52
Start
Choosing A Location For Classification
Example : Precint 1, Wilayah Persekutuan Putrajaya
Building Identification Information
Sidewalk Survey From The Exterior
Selection And Review Of Data Collection Forms
Walking around and sketching buildings (Elevations and Plan)
Determining Structural Score, S
ATC 21
ATC 22
Prepare the Location Plan of Presint 1, Wilayah Persekutuan

Putrajaya with ATC 21 and ATC 22
Conclusion
End
Figure 4.1 Methodology of the study
53
Figure 4.2 Example of data collection form (NEHRP Map Areas 1,2 Low)
There have so many types of buildings in this analysis Applied

Technology Council (ATC 21). The types of buildings have
government building, school, quarter, commercial building, public
assembly, telecom and hotel. Table 4.1 is shows the 37 number of
building use analysis ATC 21 at Precint 1, Putrajaya. Form the
table, the government buildings were built up on 1997 years and
floor high is between 3 floors to 13 floors.
54
Table 4.1Buildings information use for ATC 21

Nos
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Building
Code
C1
C2
C3
C4
C5
C6
C7
C10
D1
D2
D3
D4
D5
D6
D7
D8
E1
E2
E3
E4 & 5
E6
E7
E8
E9
E10
E11
E12
E13
E14
E15
S1
S2
S3
P1
T1
M1
H1
Building Name
Jabatan Perkhidmatan Awam
Jabatan Perkhidmatan Awam
Jabatan Peguam Negara
Kem. Sains Teknologi & Alam Sekitar
Kem. Sains, Teknologi Dan Invorsi
Jabatan Perangkaan Malaysia
JKR Wilayah Persekutuan Putrajaya
Dewam Serbaguna
Kem. Keselamatan Dalam Negeri
Kementerian Dalam Negeri
Kementerian Sumber Manusia
Kementerian Sumber Manusia
Kementerian Pengangkutan Malaysia
Badan Pencegah Rasuah
Jab Kemajuan Islam Msia (JAKIM)
Audiotorium
Kementerian Kesihatan Malaysia
Kementerian Pelajaran Malaysia
Kementerian Pengajian Tinggi Msia
Kem.Tenaga, Air Dan Komunikasi
Pejabat Sekolah Sultan Alam Shah
Kelas Bijak Sek. Sultan Alam Shah
Kuarters Sekolah Sultan Alam Shah
Alamanda Parkson
Pejabat Telekom
Masjid Putrajaya
Shangri La Hotel
Nos Of
Stories
10
8
8
8
8
8
8
8
10
7
7
7
6
6
7
3
6
9
8
6
6
10
8
8
6
13
9
7
7
6
3
4
4
1
4
3
5
Year of
Build
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1997
1999
1999
1999
1999
1999
1999
1999
55
Table 4.2 Results from analysis ATC 21

Nos
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Notes :
Building Code
Modifiers
Structural Score Types Of ATC
C1
HR, VI
1.9
ATC 22
C2
HR, PI
2.4
ATC 21
C3
HR, PI
2.4
ATC 21
C4
HR, PI
2.4
ATC 21
C5
HR, PI
2.4
ATC 21
C6
HR
2.9
ATC 21
C7
HR
2.9
ATC 21
C10
SC
2.4
ATC 21
D1
HR
2.9
ATC 21
D2
3.4
ATC 21
D3
3.4
ATC 21
D4
3.4
ATC 21
D5
3.4
ATC 21
D6
3.4
ATC 21
D7
3.4
ATC 21
D8
3.4
ATC 21
E1
VI
2.4
ATC 21
E2
HR, VI, PI
1.4
ATC 22
E3
HR, VI
1.9
ATC 22
E4 & 5
VI
2.4
ATC 21
E6
VI
2.4
ATC 21
E7
HR, VI
1.9
ATC 22
E8
HR, VI
1.9
ATC 22
E9
HR, VI
1.9
ATC 22
E10
VI
2.4
ATC 21
E11
HR, VI, PI
1.4
ATC 22
E12
HR, VI, PI
1.4
ATC 22
E13
VI
2.4
ATC 21
E14
VI
2.4
ATC 21
E15
VI
2.4
ATC 21
S1
PI
2.9
ATC 21
S2
3.4
ATC 21
S3
3.4
ATC 21
P1
SC
2.4
ATC 21
T1
LHC
2.4
ATC 21
M1
SC
2.4
ATC 21
H1
VI
2.4
ATC 21
HR - High Rise
SC-Short Column VI-Vertical Irregularity
LHC- Large Heavy Cladding
PI-Plan Irregularity
56
There have 8 number of building from total buildings analysis

get the low value or less than 2 marks for Structural Score. So, this
buildings need to use ATC 22 for further investigation.
Figure 4.4 is shows the Percentage of Total 37 number
Buildings Analysis in ATC 21 And ATC 22. From the result,
we can say that 72.5 % of buildings are in seismically good
condition building at Precint 1. There are only 27.5 % of buildings
are considering as hazardous building.
80.0%
72.5%
70.0%
P ercen tage
60.0%
50.0%
40.0%
27.5%
30.0%
20.0%
10.0%
0.0%
ATC - 21
ATC - 22
ATC
Figure 4.4 Percentage from total buildings analysis In ATC 21 and

ATC 22
Figure 4.5 is shows the Percentage Of Government Buildings

Analysis in ATC 21 And ATC 22. There have 30 number of
government building for this classification analysis. From the
result, 73.3 % of government buildings are in seismically good
condition building. There are only 26.7 % of buildings are
considering as hazardous building.
57
80.0%
73.3%
70.0%
Percentage
60.0%
50.0%
40.0%
26.7%
30.0%
20.0%
10.0%
0.0%
ATC - 21
ATC - 22
ATC
Figure 4.5 Percentage of government buildings analysis in ATC 21

and ATC 22
Figure 4.6 is shows the Percentage of High Rise Buildings

Analysis in ATC 21 And ATC 22. There have 15 number of
buildings is 8 stories and taller. From the result, 46.7 % of high rise
buildings are in seismically good condition building and 53.3 % of
high rise buildings are consider as hazardous building.
53.3%
54.0%
Percentage
52.0%
50.0%
48.0%
46.7%
46.0%
44.0%
42.0%
ATC - 21
ATC - 22
ATC
Figure 4.6 Percentage Of High Rise Buildings Analysis In ATC 21

And ATC 22
58
Figure 4.7 is shows the Location map indicated the buildings cover
by ATC 21 and ATC 22. The location map shown 29
buildings are under ATC 21 and 8 buildings are under ATC 22.
Figure 4.7 Location map indicated the buildings cover by ATC 21 and
ATC 22
CONCLUSIONS
From the study in this project, the conclusions that can be drawn
from study are as follows:
1. The building with calcify under Applied Technology
Council (ATC 21) have a less risk of damage when the
earthquake occurs compare than the building were calcify
under ATC-22.
59
2. Analysis Applied Technology Council (ATC 21) as an

effectives tool to classify the building, the fastest tool to
analyse the building, cheaper in cost and easy to use.
3. From the results, 72.5 % of buildings in analysis ATC
21and only have 27.5 % of buildings need to use ATC 22
for further investigation.
This study can be applied to whole part of Putrajaya in the
future. The local authority can used this classification map to make
decisions related to the maintenance job for the building and also
for prevention actions.
REFERENCE
Federal Emergency Management Agency, Rapid Visual
Screening Of Buildings For Potential Seismic Hazards : A
Handbook, April 1988
Supporting Documentation (second edition) (FEMA 155), Rapid
Visual Screening of Buildings for Potential Seismic Hazards,
September 2000.
ATC-21-T, Rapid Visual Screening of Buildings for Potential
Seismic Hazards Training Manual, April 30, 1996
5
SEISMIC PERFORMANCE OF SULTAN
AZLAN SHAH BRIDGE UNDER LOW
EARTHQUAKE GROUND MOTION
Azlan Adnan
Meldi Suhatril
Ismail Mohd Taib
INTRODUCTION
In recent years, Malaysia is more aware of the seismic effect on
their structures because the tremors were repeatedly felt over the
centuries from the earthquake events around Malaysia. Most
bridges in Malaysia do not take earthquake loadings into structural
design consideration. Therefore the seismic structural vulnerability
is very important in order to recognize the performance of the
bridges. The seismic analysis for bridges will be conducted as
linear and nonlinear problems. In this analysis, Sultan Azlan Shah
Bridge in Perak which consists of 5 spans (360 meters) was
modeled using two dimensional and three dimensional concepts.
The site specific analysis will be performed to determine the
earthquake loading (e.g. surface time history and design response
spectrum) using borehole data. Design response spectrum will be
constructed using IBC2000 modification. The seismic analyses
conducted were vibration analysis, time history analysis, response
spectrum analysis and damage inelastic analysis. Free vibration
analysis presented the periods and mode shapes of the structure
while time history and response spectrum analyses considered the
applied forces on the deck and piers. Damage inelastic analysis
62
showed the critical part of the bridge structural failure under

several peak ground accelerations (PGA). It can be concluded that
Sultan Azlan Shah Bridge is safe under earthquake loading when
subjected to local site effect of surface acceleration at 0.161g. The
bridge started to show initial cracking at 0.25g and collapsed at
0.32g.
RESEARCH OBJECTIVES
The overall objective of this phase of the study was to evaluate the
seismic response of Sultan Azlan Shah Bridge with the emphasis
on the two and three-dimensional effects of ground excitation.
Among the objectives are;
(i) To perform 2D and 3D modelling analysis to investigate the
seismic response of Sultan Azlan Shah Bridge in the
longitudinal direction;
(ii) To determine the time history at surface and construct design
response spectrum for Sultan Azlan Shah Bridge; and
(iii) To determine the seismic response of the bridge under
earthquake ground motion at different intensity levels, from
the initial failure stage up to the collapse level. At this stage,
the results would be able to show the critical portions of the
bridge under different earthquake ground motions (PGA).
BACKGROUND
Most bridges in Malaysia do not take into account earthquake
loadings in the structural design consideration, even though the
effects of earthquake are often felt in peninsular Malaysia. Thus,
the effects cannot be completely ignored, especially for critical
structures such as bridges. In this paper, the seismic performance
of Sultan Azlan Shah Bridge in Perak was investigated.
63
Sultan Azlan Shah Bridge is the main bridge in Perak which

crosses the Perak river. The bridge is a segmental prestressed box
girder concrete bridge which has 5 spans bridge and is
approximately 360 meters long and 41.5 feet (12.5 meters) wide
and will be analyzed to withstand earthquake and all other
anticipated loads. It consist of five spans supported by 4
intermediate piers and two abutments (Figure 5.1)
Figure 5.1 Elevation of Sultan Azlan Shah Bridge
Sultan Azlan Shah Bridge is a concrete segmental box girder

which has different depth for every segment. Figure 5.2 shows the
cross section of Sultan Azlan Shah Bridge.
Figure 5.2 Deck cross section of Sultan Azlan Shah Bridge
64
LOCAL SITE EFFECT ANALYSIS

Soil data were collected from existing soil investigation (SI) of the
bridge. The shear wave velocity (VS) were obtained by converting
the N-SPT value from Standard Penetration Test to shear wave
velocity using empirical formula proposed by Ohsaki and Iwasaki
(1973), Imai and Tonouchi (1982) and by averaging those two
formulas. Based on analysis, the VS-30 of BH-7 is 262 m/s,
respectively. Based on these results, generally the site can be
classified as stiff soil or site class D (SD) in accordance with 1997
UBC. Based on macrozonation study (Figure 5.3), the peak ground
acceleration (PGA) at the bedrock for 500-year return period of
earthquake is 0.073g (73 gal). The results of analysis can be seen
in Figures 5.4. Generally, the accelerations at the bedrock were
amplified on the surface in the range of 2.2-2.4. The predominant
periods of the spectra generally occur in the range of 0.3 0.8
second. Time history at the surface can be seen at the Figure 5.5.
Figure 5.3 Macrozonation map for 500 years return period (Azlan et al,
2006)
65
Figure 5.4 Result of ground response analysis
Figure 5.5 Time history at surface (0.161g) for Sultan Azlan Shah
Bridge
Response spectrum analysis used IBC2000 to construct the

design response spectrum. The design response spectrum can be
seen in Figure 5.6.
66
Figure 5.6 Design response spectrum using IBC2000 (modified)
SEISMIC ANALYSES
TWO DIMENSIONAL ANALYSIS OF AZLAN SHAH
BRIDGE
The seismic analysis of Sultan Azlan Shah Bridge was carried out
using SAP2000 for two-dimensional modelling. Figure 5.7 shows
the computer model simulation for Sultan Azlan Shah Bridge.
Three types of seismic analyses were implemented in this study;
the free vibration, Time History and Response Spectrum analyses
respectively.
Figure 5.6 2D modeling by SAP2000
67
FREE VIBRATION ANALYSIS

The free vibration analysis will consider five modes for twodimensional modeling (Figure 5.8). The periods of structure are
shown in Table 5.1. Figure 8 shows the mode shape 1 of the bridge
structure.
Figure 5.7 2D modeling by SAP2000
Figure 5.8 Mode shape 1 of bridge structure
Table 5.1 Natural periods of bridge

No
Period(s)
0.12995
0.08647
0.06065
0.04653
0.04396
TIME HISTORY ANALYSIS

Time history analysis of Sultan Azlan Shah Bridge model was
performed using two-dimensional models. The time history
analysis results with PGA of 0.161g are shown in Figure 5.5.The
maximum forces of the structure can be seen in Table 5.2.
68
Table 5.2 Maximum applied force for 2D time history analysis

Component
P(kN)
V(kN)
BM(kNM)
Deck
794
12985
Pier
2049
987
RESPONSE SPECTRUM ANALYSIS

IBC2000 was used to construct the design response spectrum
shown in Figure 5.6. The maximum responses for deck and pier
can be seen in Table 5.3.
Table 5.3 Maximum applied force for 2D response spectrum analysis
Component
P(kN)
V(kN)
BM(kNM)
Deck
Pier
2113
852
-
13728
1465
THREE DIMENSIONAL ANALYSIS OF AZLAN SHAH

BRIDGE
The seismic analysis of Sultan Azlan Shah Bridge used SAP2000
for its 3 dimensional modeling. The deck and pier of the bridge
were modelled using shell and beam elements. There are three
types of analysis implemented in this research; the free vibration,
time history and response spectrum. Figure 5.9 and 5.10 shows the
elevation and side view of bridge modeling.
69
Figure 5.9 The elevation of three dimensional modeling
Figure 5.10 Side view of three dimensional modeling

The free vibration analysis considered five modes. The periods of
structure are shown in Table 5.4. Figure 5.11 shows the mode
shape 1.
Figure 5.11 Mode shapes 1
Table 5.4 Five natural periods of bridge

No
Period(s)
0.19148
0.17500
0.15701
0.14494
0.12181
70

Time history analysis of Sultan Azlan Shah Bridge model was
performed using three-dimensional models with time history of
PGA = 0.161g (Figure 5.5). Maximum response of structures can
be seen in Table 5.6.
Figure 5.12 Maximum displacement is 0.051m
Table 5.6 Maximum applied force for 3 dimensional time history

analysis
Component
P(kN)
V(kN)
BM(kNM)
Deck
Pier
835
749
616
1615
10390
20225

The applied design acceleration response spectrum is shown in
Figure 5.6. The result of analysis is presented in Table 5.7. Figure
5.13 shows the displacement value at the bridge deck.
71
Table 5.7 Maximum applied force of bridge for 3 dimensional response

spectrum analyses
Component
P(kN)
V(kN)
BM(kNM)
Deck
Pier
1731
1665
1478
3431
23409
60106
Figure 5.13 Maximum displacement is 0.056m
NONLINEAR SEISMIC DAMAGE INELASTIC ANALYSES

The seismic analysis of Sultan Azlan Shah Bridge in Perak was
carried out using IDARC for 2 dimensional modeling. From the
analysis, the bridge started to crack at 0.25g and collapsed at
0.32g. The sequence of segment cracking or yielding can be seen
at Figure 5.14 to 5.18. Figure 5.18 presents the location of the first
beam yield at T = 4.505 second.
72
Figure 5.14 Flexural crack initiated at first span at PGA = 0.25g
Figure 5.15 The flexural crack initiated at the first span and bottom of all
piers at PGA = 0.27g
Figure 5.16 Flexural crack initiated at first, third and fourth span and
bottom of all piers at PGA = 0.29g
Figure 5.17 Flexural crack initiated at first, second, third and fourth span
and bottom of all piers at PGA = 0.31g
73
= Initial Cracking
= Plastic Hinge Develop
= Local Failure
Figure 5.18 Location of first beam and column yielding at PGA = 0.32g
CAPACITY OF STRUCTURES
To be able to know whether Sultan Azlan Shah Bridge could resist
the force from external load, the response should not be not more
than the structural capacity of the bridge. In this study the structure
capacity was divided into 2 parts; deck and pier section. For the
deck capacity, the forces considered were bending moment and
shear stress capacity. While for the pier, the forces considered were
bending moments and axial forces. It should be noted that due to
lack of field strength testing of the concrete, increase in concrete
strength due to aging was not considered and on the other hand, no
strength reduction factors were applied for capacity calculations.
The column moment and axial force interaction diagram can be
seen in Figure 5.19. By using strain compatibility method, we
found that the bending moment resistance of prestressed concrete
boxgirder of Sultan Azlan Shah Bridge was 24854 kNM and the
ultimate shear force is 5730 kN.
74
Figure 5.19 P-M interaction diagram
CONCLUSION
Based on the comparison between the maximum response and the
capacity of Sultan Azlan Shah Bridge in Perak, it can be seen from
Table 5.8 that all the responses for deck have not reached the
capacity of the bridge, however, the response of the column for
three dimensional response spectrum analyses more than capacity.
Table 5.8 The comparison of maximum applied and capacity force for
bridge deck
DECK
TH-2D
RS-2D
TH-3D
RS-3D
Capacity
Max Shear (kN)
794
852
616
1478
5730
Max BM (kNM)
12985
13728
10390
23409
24854
75
the bridge pier
PIER
TH-2D
RS-2D
TH-3D
RS-3D
CAPACITY
Max Axial (kN)
2049
2113
749
1665
Refer to
Max BM (kNM)
987
1465
20225
60106
Figure 19
From the analysis, it can be concluded that the column of Sultan

Azlan Shah Bridge is under the assigned earthquake loadings but
the column response for three dimensional response spectrum
analyses is higher than the capacity when subjected to local site
effect of PSA = 0.161g. The bridge started to initiate cracks at 0.25
g and collapsed at 0.32g.
REFERENCES
Chopra,A.K.1995. Dynamics of Structures: Theory and
Application to Earthquake Engineering. New Jersey,
Prentice Hall , Inc
Adnan,A.1998. Low Intensity Earthquake Effects on Steel Girder
Bridges. PhD Theses, Universiti Technology Malaysia.
Yazdani-Motlagh A.2002.Inelastic Seismic Behaviour of
Stiffening Systems Multi Span Simply Supported (MSSS)
Bridges. PhD Theses, New Jersey Institute of Technology.
Adnan,A. Taib, I.M & Suhatril,M. 2008. Seismic performance of
sungai merang bridge in Terengganu under low earthquake
ground motion. Proceeding of International conference on
earthquake engineering and disaster mitigation, Jakarta
Hendriyawan. 2007. Macrozonaton of Peninsular Malaysia and
microzonation of Kuala Lumpur and Putrajaya. PhD theses.
6
SEISMIC PERFORMANCE OF RAPID KL
ELEVATED SPAN BRIDGE UNDER LOW
EARTHQUAKE GROUND MOTION
Azlan Adnan
Meldi Suhatril
Ismail Mohd Taib
INTRODUCTION
Earthquake factor is not taken into account in structural design
consideration in Malaysia. The seismic response of existing
structures like bridges should be investigated due to recently the
effects of earthquake are often felt in peninsular Malaysia. . The
seismic analysis for bridges will be conducted in linear and
nonlinear. In this paper, the seismic performance of Rapid KL
elevated span Bridge in Kuala Lumpur which consists 3 spans (81
meters) was modeled under two dimensional and three
dimensional concept using SAP2000 and IDARC. The site specific
analysis will be performed to determine the earthquake loading
(e.g. surface time history and design response spectrum) using
borehole data at the bridge locations. Design response spectrum
will be constructed using IBC2000 modification. The seismic
analyses conducted were vibration analysis, time history analysis,
response spectrum analysis and damage inelastic analysis. Free
vibration analysis presented the periods and mode shapes of the
structure while time history and response spectrum analyses
considered the applied forces or stresses on the deck and piers.
Damage inelastic analysis showed the critical part of the bridge
78
structural failure under several peak ground accelerations (PGA).

It can be concluded that the Column of Rapid KL elevated span
Bridge is safe under earthquake loading at 0.161g. For 2
dimensional analyses, the deck applied force is lower than capacity
which subject to local site effect of surface acceleration and for 3dimensional analysis, the tensile stress at deck was more than
stress limit for class 2. The bridge started to show initial cracking
at 0.13g and collapsed at 0.44g.
BACKGROUND
In recent years, Malaysia is more aware to the seismic effect on
their structures because the tremors were repeatedly felt over the
centuries from the earthquake events around Malaysia. Peninsular
Malaysia has felt tremors several times from some of the large
earthquake originating from the intersection areas of Eurasian plate
and Indo-Australian plate near Sumatra , and some of the moderate
to large earthquake originating from the great Sumatran fault. Thus
the structures safety and adequacy in resisting earthquake effects
have been questioned. Detailed research should be conducted in
the future to determine the exact performance of the structure
before and after earthquake. The effects of earthquake for critical
structures like bridges should be considered. In this paper, the
performance of Rapid KL elevated span bridge in Kuala Lumpur
was investigated.
Rapid KL elevated span Railway Bridge is a prestressed
concrete deck structure that uses single cell box girder as
prestressed girder. The prestressed concrete bridge consists of 3
spans supported by four intermediate piers. The bridge is
approximately 81 Meters long and 5.2 Meters wide. The cross
section of Bridge deck and pier components can be seen in Figure
6.1.
79
Figure 6.1 Cross section of bridge pier and deck

RESEARCH OBJECTIVES
There are three (3) primarily objectives in this study;
(i)
To perform 2D and 3D modeling analysis to investigate

the seismic response of the actual rapid KL elevated
span Bridge in the longitudinal direction
(ii) To determine the time history and design response
spectrum at the location surface for Rapid KL elevated
span Bridge in Kuala Lumpur, and
(iii) To determine the seismic response of bridge under
earthquake ground motion from initial
failure up to
collapse. In this stage, the results show the critical parts
of the bridge for every earthquake ground motion
(PGA).
SITE SPECIFIC ANALYSIS
Soil data at bridge location in Kuala Lumpur (BH-7) were
collected from existing soil investigation (SI) of Rapid KL
elevated span Bridge. The shear wave velocity (VS) were obtained
by converting the N-SPT value from Standard Penetration Test to
80
shear wave velocity using empirical formula proposed by Ohsaki

and Iwasaki (1973), Imai and Tonouchi (1982) and by averaging
those two formulas. Based on analysis, the VS-30 of BH-7 is 262
m/s, respectively. Based on these results, generally the site can be
classified as stiff soil or site class D (SD) in accordance with 1997
UBC. Based on Macrozonation study (Figure 6.2), the peak ground
acceleration (PGA) at the bedrock of Kuala Lumpur for 500-year
return period of earthquake is 0.073g (73 gal). The results of
analysis can be seen in Figures 6.3. Generally, the accelerations at
the bedrock were amplified on the surface in the range of 2.2-2.4.
The predominant periods of the spectra generally occur in the
range of 0.3 0.8 second. Time history at the surface can be seen
at the Figure 6.4.
Figure 6.2 Macrozonation map for 500 years return period at T=1.0 sec
81
Figure 6.3 Result of ground response analysis
Figure 6.4 Time history at surface (0.072g) for Sungai Merang Bridge
Response spectrum analysis used IBC2000 to construct the

design response spectrum. The design response spectrum can be
seen in Figure 6.5.
82
Figure 6.5 Design response spectrum using IBC2000 (modified)
SEISMIC ANALYSIS
A number of computer models of the RapidKL elevated span
Bridge were created, analyzed and compared to evaluate the
structural response of the bridge under earthquake loading. All
models were linear elastic simulation in SAP2000 and nonlinear
inelastic in IDARC.
TWO DIMENSIONAL ANALYSIS OF RAPID KL
ELEVATED SPAN BRIDGE
The seismic analysis of RapidKL elevated span Bridge used
SAP2000 for two-dimensional modeling (Figure 6.6). The deck
and pier of the bridge was modeled by using beam elements. There
were three types of seismic analysis that were implemented in this
study. Namely: Free vibration, Time History and Response
Spectrum analysis
83

The free vibration analysis considered five modes. The periods of
structure are shown in Table 6.1. Figure 6.7 shows the mode shape
1 of the bridge structure.
Figure 6.6 SAP2000 2-D frame element model
Figure 6.7 Mode shape 1 of bridge structure
Table 6.1 First 5 natural periods of Rapid KL elevated span Bridge

No
Period(s)
0.28953
0.08647
0.10850
0.08647
0.04165

In this analysis, the effect of time history at surface with PGA of
0.161g (Figure 6.4) was applied to two dimensional analyses. The
maximum axial, shear and bending moment forces at bridge pier
and deck can be seen in table 6.2. The maximum displacements are
presented by U1=7.3mm and U3=4mm as horizontal and vertical
displacement.
84
Table 6.2 The results of 2 dimensional time history analysis

Component
Maximum axial force pier
Maximum shear force pier
Maximum BM pier
Maximum axial force deck
Maximum shear force deck
Maximum BM deck
Value
3463 kN
890 kN
5638 kNM
99 kN
1471 kN
8425 kNM

Response spectrum analysis used IBC2000 to construct the design
response spectrum. The applied design acceleration response
spectrum is shown in Figure 6.5.The maximum forces result of
response spectrum analysis can be seen in Table 6.3. The deck and
piers maximum displacement were u1 = 21mm, u3 = 5.5mm.
Table 6.3 The result of 2 dimensional response spectrum analysis
Component
Value
Maximum axial force pier

Maximum shear force pier
Maximum BM pier
Maximum axial deck
Maximum shear force deck
Maximum BM deck
3685 kN
2383 kN
15466 kNM
190 kN
1939 kN
13625 kNM
THREE DIMENSIONAL ANALYSIS OF RAPID KL

ELEVATED SPAN BRIDGE
For three dimensional analyses, the maximum applied stresses
were performed under response spectrum analysis. The deck,
prestressing tendon and pier of the structure was modeled by shell,
85
tendon and beam elements. Three dimensional model of Rapid KL

elevated span Bridge can be seen in Figure 6.8
Figure 6.8 The three dimensional modeling of Raid KL elevated span

Bridge

This analysis considered five modes. The period of structure are
shown in Table 6.4. The first and fifth mode shapes of bridge
structure can be seen in Figure 6.9 and 6.10.
Table 6.4 First 5 Natural periods of Rapid KL elevated span Bridge
No
Period(s)
0.18412
0.11081
0.08477
0.07710
0.04477
86
Figure 6.9 Mode shape 1 of three-dimensional analysis
Figure 6.10 Mode shape 5 of three-dimensional analysis

Time history analysis of Rapid KL elevated span Bridge model
was performed using three-dimensional models. It used the time
history with PGA of 0.161g (Figure 6.4), The Maximum
displacement of time history analysis was about 0.074m at xdirection (Figure 6.11). The maximum applied force of Rapid KL
elevated span bridge was presented at Table 6.5.
87
Figure 6.11 Maximum displacement of time history analysis is about

0.074m
Table 6.5 Maximum applied force of pier for Rapid KL elevated span
Bridge for 3-D time history analysis
Max axial force
2857 kN
Max BM
1904 kNM

Response spectrum analysis used IBC2000 to construct the design
response spectrum. The applied design acceleration response
spectrum is shown in Figure 6.5. From the analysis, the
displacement can be seen in Figure 6.12. The stresses at the top
and bottom of the box girder deck is shown in Figure 6.13.Table
6.6 shows the maximum applied forces (axial and bending moment
forces) at the column.
88
Figure 6.12 Maximum displacement of response spectrum is about

0.385m
Figure 6.13 maximum stresses of top and bottom of deck at transfer and
service stage for response spectrum analysis
Table 6.6 Maximum applied forces at the column for three

dimensional response spectrum analyses
Max axial force
Max BM
3063 kN
9968 kNm
89
IDARC NONLINEAR SEISMIC DAMAGE INELASTIC

ANALYSIS
The seismic analysis of Rapid KL elevated span Bridge in Kuala
Lumpur used IDARC for 2 dimensional modeling. From the
analysis, the bridge started to crack at 0.11g and collapsed at
0.44g. The sequence of segment cracking or yielding can be seen
at Figure 6.14.
THE CAPACITY OF STRUCTURE
To know the ability of the Rapid KL elevated span Bridge to resist
the force from external load, the external force should be not
greater than structure capacity of the bridge. In this study the
structure capacity will be divided into 2 parts; deck and pier
section. For the deck capacity, the forces in consideration are
bending moment and shear stress capacity. While for the pier
capacity, the forces in consideration are bending moment and axial
force. It should be noted that due to lack of field strength testing of
the concrete, no increase in concrete strength due to aging is
considered and on the other hand, no strength reduction factors
were applied for capacity calculations.
By using strain compatibility method, we found the ultimate
moment resistance of the deck was about 20745 kNM and the
ultimate shear resistance of the deck was 6019 kN. The moment
and axial force interaction diagram can be seen in Figure 6.15. The
top and bottom stress limit for class 2 at transfer and service was
presented at Figure 6.16.
90
= Initial Cracking
= Plastic Hinge Develop
= Local Failure
Figure 6.14 The sequence of IDARC analysis under earthquake ground
motion for initial failure up to collapse
91
Figure 6.15 M-P interaction diagram for Rapid KL elevated span pier
Figure 6.16 Stress limit for class 2 at transfer and service stage based on
BS8110

Based on the comparison between the maximum applied force and
capacity of Rapid KL elevated span Bridge in Kuala Lumpur, for
two dimensional analysis, it has shown that all applied forces for
deck and pier is not more than capacity of the bridge (Table 6.7
and 6.9). For three dimensional analyses under response spectrum
analysis, was shown the tensile stress at transfer and service was
more than stress limit for class 2.
92
bridge deck
DECK
TH-2D
RS-2D
CAPACITY
Max Shear (kN)
1471
1939
6019
Max BM (kNM)
8425
13625
20013
Table 6.8 The comparison of stress limits and applied stress at the top
and bottom of the deck
Stage
Transfer
Service
Stress at top and bottom of

the deck ( N/mm2)
-3.956
6
0.745
-6.619
Stress limit
( N/mm2)
-2.28
20
18.15
-2.67
the bridge pier
PIER
TH-2D
RS-2D
TH-3D
RS-3D
CAPACITY
Max Axial (kN)
3463
3685
2857
3063
Refer to
Max BM (kNM)
5638
15466
1904
9968
Figure 15
CONCLUSIONS
It can be concluded that the Column of Rapid KL elevated span
Bridge is safe under earthquake loading at 0.161g. For 2
dimensional analyses, the deck applied force is lower than capacity
which subject to local site effect of surface acceleration and for 3
93
dimensional analyses, the tensile stress at deck was more than

stress limit for class 2. The bridge started to show initial cracking
at 0.13g and collapsed at 0.44g.
REFERENCES
Chang, J.B &Robertson, I.N.2003. modeling of the proposed
Kealakaha stream bridge .Research Report UHM/CEE/03-03,
State of Hawai
Adnan,A. Taib, I.M & Suhatril,M. 2008. Seismic performance of
sungai merang bridge in Terengganu under low earthquake
ground motion. Proceeding of International conference on
earthquake engineering and disaster mitigation, Jakarta
Chopra,A.K.1995. Dynamics of Structures: Theory and
Application to Earthquake Engineering. New Jersey,
Prentice Hall , Inc
Adnan,A.1998. Low Intensity Earthquake Effects on Steel Girder
Bridges. PhD Theses, Universiti Technology Malaysia.
Yazdani-Motlagh A.2002.Inelastic Seismic Behaviour of
Stiffening Systems Multi Span Simply Supported (MSSS)
Bridges. PhD Theses, New Jersey Institute of Technology.
Hendriyawan. 2007. Macrozonaton of Peninsular Malaysia and
microzonation of Kuala Lumpur and Putrajaya. PhD theses.
Megally,S.H., Garg,M., Seible,F., Dowell. 2002. Seismic
Performance of Precast Segmental Bridge Superstructures.
Report No.SSRR-2001/24. Department of Structural
Engineering. University of California.
7
ANALYSIS OF PRESTRESS CONCRETE
HIGHWAY BRIDGES WITH AND
WITHOUT RUBBER BEARING
Mohd Zamri Ramli
Azlan Adnan
INTRODUCTION
Malaysia has been using bridge rubber bearing in their practices to

isolate vertical vibration from the movement of traffics. The
bearings placed between the bridge decks and the upper side of
columns are constructed to make sure the loading couldnt affect
the overall structure of bridge and other structures. In structural
earthquake engineering, natural rubber and steel are used mainly in
base isolator that is one of the energy dissipation systems to reduce
earthquake effect on structures. With laminated natural rubber and
steel, the bearing can provide very high vertical stiffness and very
flexible in horizontal direction to make sure the mounts can
support the loading from the structure and prevent excessive
sideways from any horizontal loading especially when earthquake
occur.
Most bridges in Malaysia are multi-span simply supported
where due to impact at the joints. In the design guidelines clearly
state that, in Malaysia, no seismic consideration in bridges
designing especially for JKR bridges. This study try to analyse the
common JKR Bridge fewer than two types of Earthquake ground
motion (San Fernando and El Centro Earthquake) and also Truck
load.
95
The overall objective of this study was to evaluate the linear

response spectrum analysis of actual bridge:
i.
ii.
iii.
Evaluate the combined effect of longitudinal and

transverse earthquake excitation on the response of
actual JKR bridge using 2-D and 3-D models,
Determine the capacity/ demand ratios for various
bridge components using JKR guidelines, and
Recommended possible modifications to the design of
new bridges and response possible retrofit procedures
for existing bridges.
METHODOLOGY
The method and theory that were practised to obtain successful

results in this research will be briefly explained.
a) The Finite element software is used to create a model of a bridge
structure that will be studied. The vertical and horizontal
stiffness of rubber bearing are calculated. The equations used to
obtain these stiffness values are given in equations below:
(i) Vertical Stiffness, K C =
3 AG
[1 + 2 S 2]
h
Shape factor for a square shape:

(ii) Horizontal Stiffness, K S =
S=
r
2h
AG
h
b) Free Vibration Analysis and Response Spectrum Analysis is

done on the model bridge structure in the Finite element
software. From the Free Vibration Analysis various Periods and
Mode Shapes of the model are obtained where else we are able
96
to obtain the Moment, Shear Force, Axial Force and

Displacement using the Response Spectrum Analysis.
c) Comparisons are done for the bridge with rubber bearing and
without rubber bearing on displacement, shear and moment for
various earthquake intensity
DESCRIPTION OF BRIDGE
A simple span and simply supported bridges, representative of
typical JKR bridges, are evaluated under this study. For these
bridges the numbers of spans are equal to three (Bridge F0094,
Kulai-Kota Tinggi highway). They have concrete slab on prestress
girder decks and reinforce concrete pier bents and abutments. The
pier columns are all circular. The geometry and boundary
condition of these bridge models are shown in Figure 7.1.
i) Decks and Pier Bents
The bridge decks are composite concrete slab on prestress concrete
girders and pier bents consist of reinforced concrete circular
columns. The cross section of deck prestress girders varies along
the length (Figure 7.1 and 7.2). In these studies the simplified
modelled are used by using the mid-span cross section girders
(Figure 7.3). The girder dimensions properties in the mid-span and
adjacent to the end-span segment are presented in Table 7.1.
Figure 7.1 Bridge Cross Section
97
Figure 7.2 Bridge Plan View
Figure 7.3 Mid span cross-section prestress girder
ii) Column
Three span bridges have all circular columns with reinforcements.
Figure 7.4 shows the geometric properties of the column. Table 7.1
shows the cross-sectional properties and properties of the columns
as an input data in the finite element modelling.
98
Figure 7.4 Bridge columns for 3 span bridges
Table 7.1 Beam and column element dimension and materials properties
Beam
Column
0.3322
0.5027
Moment of Inertia on X direction (m )
3.740E-2
3.260E-2
Moment of Inertia on Y direction (m4)
5.857E-3
3.260E-2
Modulus Of Elasticity (kN/m)
33.7E6
25.0E6
Poisson ratio
0.2
0.2
Weight (kN/m )
23.5616
23.5616
Mass
2.402
2.402
Composite Equivalent sec. Area (m)

4
iii) Bearings
A typical fixed bearing is shown in Figure 7.5, which consists of
rubber and steel layer with each part welded to the top and bottom
99
steel plates. The top steel plates are connected to the deck prestress
girder by connecting bolts and the bottom steel plate is connected
to the concrete support by anchor bolts. Rubber and steel layer
bearing are used to transfer the vertical and horizontal forces from
the superstructure to the substructure. Table 7.2 shows the bearing
properties used in this study (ELB 400).
L
B
h
t
Figure 7.5 Typical dimension of rubber bearing
Thickness
ts (mm)
Shape factor, S
2.1
Nos.
457
Steel
Plate
Outer Slab
t0 (mm)
Shear Stiffness
(kN/mm)
250x280x39
Natural
Rubber
Thickness For
Inner Slab
t1 (mm)
Compressive Stiffness
(kN/mm)
ELB
400
Dimension
LxBxt
(mm)
Type
Table 7.2 Elastomeric laminated Bearings properties
7.34
100
INPUT GROUND MOTION

Two seismic ground motions were used as inputs for the 2
dimensional and 3 dimensional simulation analyses. There were
recorded during the San Fernando Earthquake and El Centro
Earthquake. The horizontal components of the original ground
acceleration were linearly scaled so that the peak ground
acceleration (PGAs) is 1.17g and accordance with the maximum
PGA for San Fernando Earthquake.
i) San Fernando Earthquake
Ground motion records of the S16E component recorded at the
Pacoima Dam during the February 9, 1971 San Fernando
Earthquake was obtained and used in the study. The actual ground
acceleration for this earthquake is 1.17g and its classified under
most strongly earthquake with 65 is deaths. Figure 7.6 shows the
response spectrum curves of San Fernando Earthquake
Unit in mm / s / s
Pseudo Spectral Accelerations vs Time Period <Seconds>
Figure 7.6 Response Spectrum Curves of San Fernando Earthquake
101
ii) El Centro Earthquake

Ground motion records at 270 Degrees, Caltech IIA001, during the
May 18, 1940 Imperial Valley Earthquake (El Centro) with the
actual ground acceleration recorded are 0.35g were obtained and
used in the study. Nine are deaths with this earthquake. Figure 7.7
shows the response spectrum curves of Imperial Valley Earthquake
Unit in mm / s / s
Pseudo Spectral Accelerations vs Time Period <Seconds>
Figure 7.7 Response Spectrum Curves of Imperial Valley Earthquake
iii) Earthquake Intensity scaling factor

The 5% damping are taken for both earthquake ground motion.
With the actual ground acceleration, the scaling factors are used to
get the uniform intensities for comparison between both ground
motions. The intensities vary from 1.17g to 0.02g (Table 7.3).
COMPUTER MODELS
The bridges are analyses using SAP computer programs, where
beam-column elements are used to model the columns and simple
102
connection elements are employ in modelling bearings. In this

study 5% damping is considered in the bridge models. This is
consistent with AASHTOs response spectra and it is commonly
used value for others analyses. This level of analysis can be looked
upon as radiation damping at the foundations. Two earthquake
records along with their orthogonal components, as described in
previous section, are considered for the analyses. The two and
three-dimensional analysis is done to get the clear behaviour of
bridge with and without bearing under earthquake ground motion.
Following are the description of the 2D and 3-D models of the
bridges.
Table 7.3 Earthquake Intensity scaling factor
El Centro
(0.35g)
San Fernando
(1.17g)
Ground
Motion
Ground
Acc. (g)
Scaling
Factor
0.020
0.017
0.075
0.064
0.150
0.128
Low Intensity (SPC B)
0.250
0.214
0.350
0.299
Requirement from ASSHTO (SPC C and

D)
1.170
0.020
0.057
0.075
0.214
0.150
0.429
Low Intensity (SPC B)
0.250
0.714
0.350
Requirement from ASSHTO (SPC C and

D) and the Normal ground acc.
1.170
3.343
Most Strongly Earthquake
1.1.1 Description
Expected ground acc. from Sumatra
earthquake for Peninsular Malaysia
The ground acc. used by Chin Fan Kee to
design Penang Bridge
Most Strongly Earthquake and Normal

Ground acc
Expected ground acc. to Penisular from
Sumatra earthquake
The ground acc. used by Chin Fan Kee to
design Penang Bridge
103
i) 2-D Models
Sap were employ for modelling bridges in longitudinal direction.
For multi span simply supported bridges it is reported that
longitudinal seismic response is more important (Zimmerman and
Brittain). Figure 7.8 shows the analytical models for the three span
bridge model. Two-dimensional models are not able to represent
the bridge skewness. Beam-column element can be oriented
arbitrarily in XY plane.
Figure 7.8 Joint and Element Number for 2D analysis
ii) 3-D Models

Sap2000 are used in this study, Figure 7.9 show the analytical 3-D
models for 3 Span JKR Bridge. For major mode shapes in the tree
translational directions (i.e., longitudinal, transverse and vertical)
corresponding mass proportional damping was assigned to the
bridge modelled.
Figure 7.9 Joint Number for 3D analysis
104
Free-Vibration Analysis
These bridges have dominant horizontal free vibration modes with
periods ranging 0.1 to 0.5 Sec. The bridges were modelled using
the SAP computer code for finite element analysis. Simplified 2
dimensional and 3 dimensional linear models were established.
The substructure columns were assumed to be fixed in their footing
and the abutments were modelled as pinned and roller supports.
The beam models are used in the analysis. Table 7.4 summarised
the natural frequencies of dominant modes.
Table 7.4 Natural frequencies of dominant modes
Natural Frequencies (Hz)
2-D
3-D
Periods (sec)
Without Bearing
7.559
0.1323
With bearing
3.679
0.2718
Without Bearing
3.039
0.3291
With bearing
2.159
0.4632
Free vibration can be expressed by means of the motion of

structure without any dynamic excitation-external force of support
motion. The analyses on the first 4 modes were studied to find the
initial motion modes of the bridges structure. The study shows that,
the bearing are lengthens the period (Figure 7.10), increases the
displacement and reduced the axial forces for beam and column
element. This finding are follows the flexibility concept for base
isolation system (Figure 7.11).
Period
105
Mode 1
Mode 2
Mode 3
Mode 4
2-D
With
Bearing
Without
Bearing
With
Bearing
Without
Bearing
0.6
0.5
0.4
0.3
0.2
0.1
0
-0.1
3-D
Types of analysis
Figure 7.10 The structure period profiles
Figure 7.11 Displacement Response Spectrum
This analysis is concerned with procedures to compute the peak

response of a structure during an earthquake directly from the
response spectrum. When increased the flexibility may make
106
structure more responsive to wind and service loads. Figure 7.12

Show the theory force distribution for isolated and conventional
structure.
In this study, the moment and shear analysis for the beamcolumn element, the forces are increases but there are still lower
compare than truck analysis. With the result, the bearing must be
modified to sustain the earthquake loading to minimise the forces.
The analysis with bearing factor were minimised the axial forces
for beam element (Figure 7.13) but still increased in threedimensional analysis for column element (Figure 7.14). This is
because the effected from bearing location and earthquake forces
distribution from bottom to top (Figure 7.15)
Figure 7.12 Theory force distribution for isolated and conventional

structure
600
500
400
300
200
100
0
Moment (kNm)
Shear (kN)
Axial (kN)
2-D
With
Bearing
Without
Bearing
3-D
With
Bearing
Without
Bearing
2-D
With
Bearing
Without
Bearing
With
Bearing
Without
Bearing
Internal Forces
107
3-D
San Fernando
El-Centro
Types of Analysis
Internal Forces
Figure 7.13 Maximum Forces for Beam (0.15g)
800
700
600
500
400
300
200
100
0
Moment (kNm)
Shear (kN)
Axial (kN)
2-D
With
Bearing
Without
Bearing
San Fernando
With
Bearing
3-D
Without
Bearing
With
Bearing
Without
Bearing
With
Bearing
Without
Bearing
2-D
3-D
El-Centro
Types of Analysis
Figure 7.14 Maximum Internal Forces For column (0.15g)
108
Bearing
Forces
Figure 7.15 Bridge bearing and location and earthquake forces
distribution
1200
1000
800
600
400
200
0
Verticle (kN)
Horizontal (kN)
2-D
With
Bearing
Without
Bearing
San Fernando
With
Bearing
3-D
Without
Bearing
With
Bearing
2-D
Without
Bearing
With
Bearing
Without
Bearing
External Forces
The result from external forces analysis give the maximum

vertical forces come from truckload but the different situation for
the horizontal direction. This is the right condition because the
seismic effects come from the horizontal direction. Thats mean;
the bridge horizontal design must be taken seriously especially for
those countries are exposed to direct earthquake.
3-D
El-Centro
Types of Analysis
Figure 7.16 Maximum External Forces for Beam (0.15g)
External Forces
109
800
700
600
500
400
300
200
100
0
Moment (kNm)
Verticle (kN)
Horizontal (kN)
2-D
San Fernando
With
Bearing
Without
Bearing
3-D
With
Bearing
Without
Bearing
With
Bearing
Without
Bearing
With
Bearing
Without
Bearing
2-D
3-D
El-Centro
Types of Analysis
Figure 7.17 Maximum External Forces for Column (0.15g)
DISCUSSION AND CONCLUSION

In Malaysia situation, the effect from Sumatera earthquake, the
seismic intensity is assuming not more than 0.075g. When we lock
for the result, this bridge can sustain the seismic loading. Based on
the result, the following specific conclusions can be made:
i.
ii.
iii.
The bridge can sustain the seismic load base on the

requirement from ASSHTO (SPC C and D).
The effect from both type of ground motion may not
give big impact for this bridge especially for our
country with low expected intensities (about 0.02g).
The bridge has more effect from El-Centro earthquake
compare than San Fernando earthquake.
110
Although our country is free from direct earthquake, at least we

have a minimum requirement to be considered in earthquake effect
for our structural design procedure.
REFERENCES
Ali, H.M. and Abdel-Ghaffar A.M. (1995). Modelling the
nonlinear seismic behaviour of cable-stayed bridges with
passive control bearings. Computers & Structures, 54(3),
461-492.
Ballard, T.A., Krimotat ,A., Mutobe. R. and Treyger .S. (1997).
Nonlinear seismic analysis of carquinez strait bridges.
Computers & Structures, 64(5/6), 1041-1052.
British Standards Institution (1982). Guide to Selection and Use of
Elastomeric bearings For Vibration Isolation of
Buildings~BS6177 :1982. BSI Ltd.: London
Derham, C.J. (1983) Proceedings, Natural Rubber for Earthquake
Protection of Buildings and Vibration Isolation, Proc.,
International Conference, Kuala Lumpur. Direct Art Co:
Kuala Lumpur.
Edward L. Wilson and Ashraf Habibullah. (1989) SAP90 User
Manual: A Series of Computer Programs for the Static and
Dynamic Finite Element Analysis of Structures. California.
Imbimbo, M. and De Luca, A. (1998). F.E. stress analysis of
rubber bearings under axial loads. Nonlinear seismic analysis
of carquinez strait bridges. Computers & Structures, 68(1/3),
31-39.
Jun, Y.M. and Eric, M. L. (2000). Torsional effects on short-span
highway bridges. Computers & Structures, 75(5), 619-629.
8
PREDICTING OF BRIDGE CONDITION
BASED ON SEISMIC ZONATION BY
USING ARTIFICIAL NEURAL
NETWORK
Azlan Adnan
Sophia C. Alih
Rozaina Ismail
INTRODUCTION
In this study, artificial intelligent methodology is applied to bridge
inspection system. Artificial neural network (ANN) is developed
to predict bridge condition rating based on different intensity of
seismic zonation.
Inspection results from nondestructive
evaluation are used as an indicator to the structural condition.
Numbers of systems are developed to determine the effective
parameters and neural network structure in order to build the most
predictive ANN system. Backpropagation algorithm with one
hidden layer is used to develop the neural network and Borland
C++ is used as the programming language. 75 concrete bridges
under the supervision of Public Works Department, PWD
(Malaysia) have been selected for further inspection using
nondestructive evaluation technique which includes the rebound
hammer test, Ultrasonic Pulse Velocity, and electromagnetic cover
meter. These tests were conducted to determine the bridge
strength, structural damages, and level of the damages.
Assessing the condition of a structure is necessary to determine
its safety and reliability. Ideally, structural health monitoring
113
should be similar to medical health monitoring of the body. In

medical health monitoring, the life signs such as pulse and blood
pressure give an overall indication of the overall health of the
body. This is analogous to global health monitoring, in which
damage to the structure can be identified by measuring changes in
the global properties of the structure. Once the body signs show an
anomaly, we do a battery tests to determine the cause of the
anomaly.
Analogously in structural health monitoring,
nondestructive evaluation can be used to determine the nature of
the damage.
Concrete bridges are exposed to numerous environmental
stressors and traffic loads which increases from time to time.
These can cause a reduction in overall strength and lead to
eventual failure of the bridge. Periodic bridge inspections are
therefore necessary to assess the extension, implications, and
current state of the deterioration process. Inspections not only help
to prevent failure but also deliver information necessary to
effective administration of the bridge network.
Thus, the
administrative bodies can further define priorities and establish
programs to apply available resources to the most critical bridges.
Currently bridges are evaluated through either a visual
inspection or structural analysis. Visual inspections are commonly
used nowadays. When bridge evaluation is conducted using this
method, rating will be assigned to the bridge components by a
responsible inspector. Major problem with visual inspection is the
inherent variability that naturally occurs when subjective
observations are converted to a numerical rating.
Bridge
evaluation based on this method may vary according to personal
judgment. Thus, large uncertainties exist in the interpretation of
inspection data.
Nondestructive evaluations are one of the techniques suggested
by researchers to overcome the limitations faced by the existing
rating system. This method has gained interests among researchers
due to its effective ability in determining damages inside the
structure that are not visible. Previous researches show a good
potential of nondestructive testing to be used in evaluating
114
structural condition of existing structure. Thus several trials were

carried out to correlate data from nondestructive testing with visual
inspection in order to enhance the existing evaluation process.
Despite of all the advantages of using the nondestructive testing,
this method is not always readily available and there may be
problems occur with the lack of experienced inspectors to conduct
the test. Hence, the implementation of this method in routine
inspection may be limited. The strong capability of artificial
neural networks in predicting fuzzy data and the success
applications of this approach in various fields give an idea to
implement ANN to predict bridge condition based on
nondestructive testing data and visual inspection. In other words,
nondestructive tests may not be necessarily conducted in each
routine inspection; previous nondestructive testing results will be
used to predict the condition rating of a bridge. It is hope that this
system will assess the current inspection process and thus lead to a
more thorough yet uncomplicated evaluation.
PROBLEM STATEMENT
Existing practice in evaluating bridge condition through visual
inspection has been identified to have few limitations. Despite of
their role as the first step of any condition assessment procedure,
this type of evaluation is subjected to large uncertainties and
depends primarily on a personal judgment of responsible inspector.
Ratings assigned to the bridge component are subjective and may
vary according to the visual observation. Due to these limitations,
few researches have been conducted to support assessment made
using visual inspection.
In recent years, researchers and industrial practitioners has turn
to nondestructive testing (NDT) method to evaluate their structure
due to the ability of this method in determining invisible defects
inside the structure that is not possible to be evaluated through
visual inspection. Therefore, the NDT method has been chosen in
this research to support evaluation made in the existing practice.
115
However, despite of their masses advantages, this method is not

always readily available and there may be problems occur with the
lack of experienced inspectors to conduct the test. Hence, the
implementation of this method in routine inspection may be
limited.
If the NDT results can be predicted, the bridge condition can
still be assessed without even conducting the test during
inspection. The strong capability of artificial neural networks
(ANN) in predicting fuzzy data and the success applications of this
approach in various fields give an idea to implement ANN to
predict bridge condition based on previous inspection data. If this
approach success, there will be less works need to be done during
inspection and yet the evaluation is still thorough. This will
benefit lots of people involved in bridge inspection especially the
bridge authority. This system can help the authority to forecast
bridge condition at any given time. Critical bridges can be short
listed and prioritized for the allocation of maintenance budget.
OBJECTIVES
This study is conducted to comply with the following objectives:
i)
ii)
iii)
iv)
To conduct nondestructive testing (NDT) on selected bridges

to evaluate the bridge condition
To find the correlation between NDT results and visual
inspection (VI) ratings
To determine NDT results and VI ratings using Artificial
Neural Network (ANN)
To find the correlation between NDT results and VI ratings
from field test (manual process) and ANN
116
BACKGROUND OF RESEARCH
Artificial Neural Network
From the time of the first primitive computing machine, their
designers and users have been trying to push computers beyond the
role of automatic calculators and into the realm of thinking
machines. Thus, the emergence of artificial intelligence (AI) has
brought up new transformation to the computer applications.
Different approaches using the concept of AI have becoming more
popular and beneficial to industrial practitioners and researchers.
The human senses detect stimuli, and send this input
information (via neurons) to the brain. Within the brain, neurons
are exited and interact with each other. Based on the input, a
conclusion is drawn, and an output is sent from the brain in the
form of an answer or response. The interaction between neurons is
not seen by anyone, but manifests itself as identifiable intelligent
behavior.
The same type of structure can be developed for a computer
modeling of intelligent behavior. Neurologists and AI researchers
have proposed a highly interconnected network of neurons or
nodes for this purpose. Information is applied as an input to a
network of nodes. The nodes mathematically interact with each
other in a manner unknown by the users. Eventually, based on the
input, an output arises that maps the expected, macroscopic inputoutput pattern.
In other words, the development of ANN are intended to mimic
the behavior of biological learning and the decision making
process without being biologically realistic, in detail (Kim et al.
2003). Neural networks represent simplified methods of a human
brain and may be used to solve problems that conventional
methods with traditional computations find difficult to solve.
117
METHODOLOGY
Programming Phase
The aim of this phase is to develop the artificial neural network
system to determine conditional ratings. The development consists
of few stages which may vary according to the programmer.
However, each set of stages shared the same basic fundamental in
ANN development. For instance, Wu and Lim (1993) applied five
stages in developing a neural network model which include: data
acquisition, architecture determination, learning process
determination, training the network, and finally testing the trained
network for generalization evaluation. Timothy (1993) classified
ANN development phase in three steps; training or learning phase,
testing phase, and validation phase. Whereby, Elazouni et al.
(1997) classified ANNs modeling into three main phases; design,
implementation, and recall or use for problem solving. The design
phase consists of two aspects; problem analysis and problem
structuring. The implementation includes three main tasks; data
collection, selecting the network configuration, and training and
testing the network.
In this study, three main phases are applied in the application of
neural network model, in which each phase consists of few steps as
shown in Figure 4.32. The phases are as listed below:
i)
ii)
iii)
Analyzing data
Developing ANN structure
Operating the ANN
In the first phase; data involved in developing the network are

analyzed to determine their characteristics and correlation between
input and output data are evaluate. Data samples are divided into
two groups; one group is to be used in training and testing phase,
and another one is for the validation phase. Normalization process
is conducted to the data before it is applied to the network.
118
In developing the ANN structure (phase ii), parameters involved

in the network need to be finalized. In this phase numbers of trial
and error process are conducted to determine the effective input
variables and number of hidden neurons needed in developing the
neural network. Other parameters such as weights and biases,
functional forms, learning rate, and momentum of coefficient will
also be outlined.
When the network structure has been concluded, operating
process (phase iii) can now be proceed. This phase consists of
three process; training, testing, and validation.
Network
performances in each process are analyzed to determine the
networks ability in providing output based on the data it is
subjected to. The above mentioned phases are discussed further in
the following sub-topic. Figure 8.1 graphically illustrates the
phases involved in the programming phase.
ANALYSING DATAFOR SEISMIC ZONE
The data used in developing the neural network is limited to the
data gathered during site survey phase. These include the data
from bridge inventories, visual inspection report, and
nondestructive testing. There are nine parameters applied to the
ANN model which includes the parameter based on the seismic
intensity at the bridge site. Peak ground acceleration, PGA value
used is based on 2% probability of accidence in 50 years (return
period equal to 2500 years). The PGA value for each bridge
sample is determined using the macrozonation map shown in
Figure 8.2. Road maps for each district as attached in Appendix
4D are referred to get the exact location of the bridge samples.
The above parameters are classified into specific groups to
characterize their types and variety.
Each group will be
represented by a particular code to be used in the ANN
development. Table 8.1 to 8.5 shows the classification made to the
parameters.
119
As mentioned, the PGA values shown in Table 8.1 are

determined from the macrozonation map shown in Figure 8.2.
Referring to the map, PGA values for Peninsular Malaysia ranges
from 40 to 200 gal, whereas for Johor state it ranges from 60 to
140 gal. For programming purposes, the PGA values are classified
in ten groups starting from 0 to 200 gal with an interval of 20 gal.
Figure 8.1 Stages involved in the programming phase
120
Table 8.1 Seismic zonation

PGA Value (gal)
Group Code
0-20
21-40
41-60
61-80
81-100
101-120
121-140
141-160
161-180
181-200
1
2
3
4
5
6
7
8
9
10
Figure 8.2: Macrozonation map at 2% PE in 50 years on rock site

conditions for the Peninsular Malaysia (TR=2500year),
Adnan et. al. (2006).
121
Although bridges can be classified according to different aspects

such as their function, form of superstructure, as well as material
and method used in construction, interspan relation is used to
classified bridge types in this study. There are three major
classification made as shown in Table 8.2.
Table 8.2 Classification of system type
System Types
Simply supported
Continuous
Cantilever
Group Code
1
2
3
Bridge samples used in this study are limited to concrete bridge

only. Basically, deck type for the bridge samples can be classified
in six groups as shown in Table 8.3. Abutment types are
categorized in four groups as shown in Table 8.4. As for the pier
structure, six groups are identified to classify the pier type. This
classification is shown in Table 8.5.
Table 8.3 Classification of bridge deck
Bridge Decks
Group Code
Prestressed concrete I-beam

Reinforced concrete beam
Prestressed inverted T-beam
Prestressed concrete M-beam
Concrete box-girder
Reinforced concrete slab
1
2
3
4
5
6
122
Table 8.4 Classification of abutment

Abutment Types
Bank seat
Wall abutment
Skeleton abutment
Wing wall
Group Code
1
2
3
4
Table 5: Classification of piers

Pier Types
Solid pier
Leaf pier
Single leg pier
Multi column pier
Portal pier
V-pier
Group Code
1
2
3
4
5
6
As mentioned earlier, data involved in developing the neural

network needs to be analyzed first. This analysis consists of three
stages;
i)
analyzing the characteristics of data involved
ii) classification of data
iii) data normalization
RESULT AND DISCUSSIONS
Data Characteristic
Characteristic of each data used in developing the neural network
needs to be defined due to the random process applied to extract
the data. The distributions of each inputs and outputs are defined
through histograms and the relationships between these two
123
variables are determined using linear correlation coefficient, r.

Figure 8.3 shows the distribution chart for each parameter.
Figure 8.3 shows the distribution of seismic zone at bridge sites.
There are four zones involved with peak ground acceleration
ranges from 61 to 140 gal. The distributions are quite even
between the bridge samples. Zone class 1 shows the highest
readings with 20 bridge samples, followed by class 3 and 4 with 18
and 15 samples respectively. Zone 4 has the least samples with 14
bridges.
Figure 8.3 Distribution of seismic zonation on bridge site
Data Classification
After knowing the distribution of input and output parameters used
in this study, it is important to evaluate the classification made to
these parameters. Data involved are classified in two groups; 1)
training and testing phase, 2) validation phase. From 75 samples
of concrete bridges involved in this study, 52 samples are used in
the training and testing phase while 15 samples are used in the
124
validation phase. These validation samples can never be applied in

the training and testing phase. This selection is based on a random
process.
Another 8 samples have incomplete details and
inadequate to be used in the ANN system. Appendix 6A shows the
whole input and output data used in training and testing phase, as
well as in the validation phase. Figure 8.4 graphically shows the
distribution of data used in both categories.
Figure 8.4 shows the distribution of seismic zonation for training
and validation phase. Classifications of seismic zones in this
figure are the same with Figure 8.3. Training and validation data
are represented in all categories. Since the data distributions are
uniform, the differences between training and validation data set
are uniform as well. Seismic zone class 1 with PGA between 6180 gal, has the highest number of samples in both phases.
Figure 8.4 Distribution of training and validation data
125

In this section, result from the application of ANN in each rating
prediction is discussed. In this stage, the final ANN structure is
used in which all seven input parameters and 15 hidden neurons
are applied to the network. Mean square error between the actual
output and output given by the ANN are discussed to evaluate the
training phase. Linear correlation coefficient between the actual
output and output given by the ANN is analyzed.
Rating Prediction for Bridge Deck

Figure 8.5 shows the plotted values of actual rating and rating
predicted by the neural network in the testind phase for bridge
deck. The VI rating gives r equal to 0.758.Out of 52 samples used
in the testing phase, 48 samples were predicted accurately for the
VI rating which presents 92% from the total data used. Rating 1
dominated the samples for VI rating and this rating had the highest
number of samples predicted without error. Rating 4 had a lowest
number of samples and it was predicted with more than 1 rating
difference.
Figure 8.5 Plotted actual rating and predicted rating by ANN during
testing phase for deck ratings, VI
126
CONCLUSIONS
This study involved two major aspects; 1) conducting
nondestructive testing to evaluate bridge condition, and 2)
applying artificial neural network methodology to predict the
bridge condition. Conclusion from the bridge inspection conducted
in this study and ANN that had been developed can be summarized
as follows:
1) Based on the visual inspection conducted by PWD, most of
the bridge samples were rated 1 and 2. It can be observed that
the VI rating ranged differently based on each district. In other
words, the rating assigned during visual inspection was
dependent on the inspectors evaluation and judgment.
2) Based on the artificial neural network developed in this study,
it can be concluded that the ANN predictions were very much
affected by the data applied to the network. These included the
total samples used in the training and validation phase, number
of samples in ach rating type, and the distribution pattern of
the data. Increasing the number of output parameter will
improve the ANN performances.
3) The ANN models used in the rating prediction were capable to
predict the condition ratings with high accuracy. Results show
that the VI rating can be predicted between 67% correct. The
linear correlation coefficients between the actual rating and
rating predicted by the network were up to 0.9 which shows a
very strong relationship between these two ratings. These
results prove the capability of the network to recognized data
pattern and provide accurate prediction even when the
correlation between input and output data is very weak.
4) However, the development of the model needs extra
consideration on the most effective parameters and number of
neurons to be applied to the network. Both of these aspects are
important and affect the network prediction significantly.
Based on findings obtained in this study, the ANN can best
127
perform when all of the bridge parameters are applied as input

to the network.
REFERENCES
Adnan, A., Hendriyawan,. Marto, A., and Irsyam, M. (2006).
Development of Seismic Hazard Map for Peninsular
Malaysia. Proceedings on Malaysian Science and
Technology Congress. PWTC Kuala Lumpur, Malaysia. 1-26
September.
Elazouni, A. M., Nosair, I. A., Moheildin, Y. A., and Mohamed,
A.G. (1997). Estimating Resouce Requirements at
Conceptual Design Stage Using Neural Networks. Journal of
Computing in Civil Engineering. Proc. Paper 11485, 11, 4:
217-223.
Kim, J. I., D. K., Feng, M. Q., and Yazdanhi, F. (2004).
Application of Neural Network for Estimation of Concrete
Strength. Journal of Materials in Civil Engineering. Vol. 16
(3): 257-264.
Sophia C.Alih (2007). The Application of Artificial Neural
Network in Nondestructive Testing for concrete Bridge
Inspection Rating System. Master Degree Thesis: University
of Technology Malaysia.
Wu, X. and Lim, S. Y. (1993). Prediction of Maximum Scour
Depth at Spur Dikes with Adaptive Neural Networks. Neural
Networks and Combinatorial Optimization in Civil and
Structural Engineering. Vol. 10 (4): 263-268.
Timothy Masters (1993). Practical Neural Network Recipes in
C++. San Diego CA: Academic Press, Inc.
9
DATABASE SYSTEM AND DIGITAL
EARTHQUAKE EVALUATION OF
BUILDINGS
Azlan Adnan.
Rozaina Ismail
INTRODUCTION
This paper presents the development of database system for digital
earthquake evaluation of building. With the present understanding
of earthquake event and existing study, it shows that the public
now is getting aware to this natural phenomenon that can be
classified as an environmental load. Besides of this awareness, the
information of the existing data and study is still hard to be
accessed by public. Therefore, the database system for digital
earthquake evaluation of our building needs to be developed. The
objectives of this study are to develop the digital form of ATC-21
and development of database system. Based on these objectives,
the significance of this study is when the evaluation process can
be performed in much more easier way and quickly rather than the
manually evaluation process. The analysis of this study covers the
evaluation procedure for building, development of database
system and development of user interface. As a result of this
study, the database system can be performed as a medium to
provide information to public such as the data and result from
existing earthquake evaluation of our building and also provide
the digital form of evaluation.
Earthquake evaluation of building must be done in order to
identify those building that might pose a risk of loss of life or
injury in the event of damaging earthquake (Farzad Naeim, 1989).
Besides, the existing data and study regarding this evaluation of
129
building, there must be some medium to act as the platform for

public to access the data and result of existing study. Thus, the
development of database system for the purpose of making easily
to retrieve the data will be very useful to public especially. Since
the evaluation process needs to refer quite a lot of separate sheet,
there is also a need to simplify the evaluation process.
The evaluation of earthquake start with the ATC-21 procedure
which is to assign the final score. The existing preliminary
evaluation by ATC-21 based on sidewalk survey of a building and
data collection which am inspector completes based on visual
observation of the building (ATC, 1971). This typical procedure
takes long time to be evaluated, thus better evaluation system must
be challenged to make sure the evaluation process is easier.
THE RESEARCH AREA AND SCOPE
The evaluation of building exactly base on ATC-21 procedures as
shown in Figure 9.1, which is focus on tree type of building under
different seismic zones they are low, moderate and high zone. The
development of digital form for ATC-21 and database system used
the Microsoft Visual Basic 6.0 programme and Microsoft Access.
Visual Basic can be, and is being, used very effectively in a variety
of ways (Mary R. Sweeney, 2000).
THE METHODOLOGY
The methodology begins with learning the procedure of building
evaluation based on ATC-21 procedure where by identifying the
primary structural lateral load resisting system and materials of the
building. The Rapid Screening Procedure (ATC-21) method
consists of inspecting a building by using digital form in order to
quickly find out whether our building should be subjected to more
detailed investigation as to its seismic adequacy. Basic structural
hazard scores for various building types are provided on form. In
130
the form, there are modification factor needs to be added or

subtracting from the basic score, which relate to significant
seismic-related defects in order to arrive at a final structural score
S. The basic structural hazard score, performance modification
factors and final structural score S all relate to the probability of
the building sustaining major life-threatening structural damage.
Final scores typically range from 0 to about 6, with higher S scores
corresponding to better seismic performance.
Figure 9.1 Procedures of ATC-21
The result is a ranking of surveyed buildings, which may be

divided into two categories: those acceptable as to risk life safety
or those that may be seismically hazardous and should be studied
further. If a building has an S score of 2 or less, it should be
further study.
These procedures have been digitalized into digital form by
using Microsoft Visual Basic 6.0. The friendly user interface was
developed from the visual basic environment. Microsoft Visual
131
Basic 6.0 is a programming application for Microsoft Windows

operating system and also for variety of Web browser. Visual
Basic is the interaction between user and system event. It is
program called event-driven programming. Furthermore, this
program is fast interacting with the user by giving a command
language; where else a user can control the activities in the
programme. A Visual Basic application or project typically
requires some input, processing and output. This is a modern
programming language that combines the use of a graphical user
interface with the easy-to-learn BASIC programming language
(Bishop-Clark, 1998). For interaction with user the project will
typically use several control such as forms text and picture boxes,
labels and command buttons. Using these controls, the users
control the programme execution. Any graphic or mouse control
events required quite complex programming skills. In contrast, VB
consists primarily of individual blank pages (called forms), upon
which various common controls are presented, such as buttons,
text boxes and picture frames (Francisco Cabello et al., 2002 ).
The interface of database system also was developed by using
Microsoft Visual Basic 6.0. Then the database was using Microsoft
Access to store and retrieved all the data that has been store.
ATC-21: RAPID SCREENING PROCEDURE (RSP)
SURVEY
Rapid Screening Procedure (RSP) described in ATC-21 is a tool to
establish the level of risk to building structures and to perform
general assessment of the expected performance of these building
structures due to earthquakes. It is a method whereby buildings can
be rapidly identified via a visual screening as either seismically
acceptable or potentially seismically hazardous. Rapid Screening
Procedure (RSP) utilized a methodology based on a sidewalk
survey of a building and a Data Collection Form, which an
inspector completes, based on visual observation of the building.
The Data Collection Form includes space for sketch and -
132
photograph for the building, as well as pertinent earthquake safetyrelated data.

This ATC-21 procedure begins with identifying the primary
structural lateral load resisting system and materials for the
building. Basic structural hazard score for various building types
are provided on the form, and the inspector needs to select the
appropriate one. This basic structural hazard score will be modified
by the inspector by adding or subtracting the performance
modification factors, which relate to significant seismic-related
defects that the inspector may observe, in order to obtain a final
score, called S. Figure 9.2 shows the sample manual form of
ATC21.
The Basic Structural Hazard score, Performance Modification
Factors and final Structural Score S all relate to the probability of
the building sustaining major life-threatening structural damage. A
high S score indicates that the building is probably adequate. A
score S of 2 is suggested as a cut-off based on present seismic
design criteria. That is, if a building has a score of 2.0 or less, it
should be investigated further (ATC21, 1988).
133
Figure 9.2 Sample manual form of ATC 21
134
SURVEY IMPLEMENTATION INSTRUCTION

There are two survey implementation instruction involved in
ATC-21. The first survey implementation instruction is the
sequence of survey implementation. There are several steps
involved in collecting data, planning and performing a rapid
screening of potential seismically hazardous buildings. The general
sequence of implementing the survey methodology starts with
selection of area to be surveyed and supplementary data to be
included in the survey and used in decision making. Then followed
by development of record keeping system, training of survey
personal, selection and review of data collection.
The next sequence is pre-field data collection, identification of
structure and walking around and sketching building.
Photographing the building for identification purposes will be
conducted. Then the basic data such as number of stories,
structural system and material and major problem will be noted as
well. Before we go to the final steps of survey implementation
which is determining the structural score, S for the building, we
must determining the basic structural hazard score and modifiers
due to major problems, building height and soil condition.
The second survey implementation instruction is selection and
review of data collection form. In carrying out this survey, data
collection form of low seismicity is to be selected. Low seismicity
data collection from corresponds to NEHRP Map areas 1 and 2 in
National Earthquake Hazards Reduction Program (NEHRP)
Recommended Provisions for the development of Seismic
Regulations for new Buildings (BSSC, 1985).
RSP METHOD AND DATA COLLECTION FORM
i) Overview of the RSP Method
Briefly, the RSP method consists of inspectiong a building from
the exterior (termed a sidewalk survey) in order to quickly
135
determine if the building is probably adequate for the earthquake

force it is likely to experience, or whether there may be reasonable
doubts as to the buildings seismic performance. The result of the
rapid visual screening procedure is a finding as to whether the
building should or should not be subjected to more detailed
investigation as to its seismic adequacy.
A basic concept of the RSP is to identify, for the building under
review, which of several typical building types it corresponds to.
These typical building types as defined for use in this method are
listed as below:
Table 9.1 Building Type Identifier (ATC-21, 1988)
Building
Identifier
W
S1
S2
S3
Description
PC1
PC2
RM
Wood buildings for all types

Steel moment resisting frame
Braced steel frame
Light metal building
Steel frame with cast-in-place concrete
shear walls
Concrete moment resisting frames
Concrete shear walls building
Concrete or steel frame building with
unreinforced masonry infill walls
Tilt-up buildings
Precast concrete frame buildings
Reinforced masonry
URM
Unreinforced masonry
S4
C1
C2
C3/C5
However, all buildings selected for this study are found to be of

reinforced concrete frame construction. They are selected based on
the availability of data. Therefore, the building identifier
representing type of building is type C1 (see Table 9.1 Concrete
moment resistance frame).
ii) Structural Score S
136
Having determined the basic structural material and system, the

corresponding Basic Structural Hazard should be circled.
Performance Modification Factor (PMFs, or Modifier) that account
for major defects are determined based on visual inspection carried
out. These are generally subtracted from the Basic Structural
Hazards score, to determine the final Structural Score S.
iii) Modifiers
The severity of the impact on structural performance varies with
the type of lateral forces resisting system, thus the assignment
PMFs depend on building type. Performance Modification Factor
for each building type are indicated in columns on the Data
Collection Form. PMFs for the building being surveyed can be
circled, and the column summed to get the final Structural Score S.
iv) Poor Maintenance

Generally, the effect of poor condition or maintenance on seismic
behavior is one of the most difficult aspects of a building to
quantify. Poor condition or maintenance impacts seismic behavior
when it results in materials weaker than called for in the structural
design. A prime example is excessive and/or uneven ground
settlement or movement, typically detected by diagonal (shear)
cracking perpendicular to the principal tensile stress.
For concrete buildings, poor condition include as follow:
Poor concrete placement (e.g., segregation, where the

coarse and fine aggregates are separated, giving a honeycomb like appearance)
Deteriorated concrete (due to chemical attack by agents
such as salt or sulphates, or due to poor use of materials in
137
the original concrete, such as use of beach sand or high salt

content aggregates)
Exposed and/or corroding reinforcing steel (usually
indicated by rust coloured stains on the exterior of the
building).
These conditions obviously will often require a detailed inspection

of the building in order to be observed. This modifier should be
applied only when definite evidence poor condition, say about
70%, (ATC-21, 1988) that clearly impacts structural strength is
detected.
v) Vertical Irregularity
This modifier applies to all types. If the building is regularly
shaped in elevation, or if the walls are not perpendicular to the
ground, the modifier should be applied. Irregularity is a difficult
characteristic to define, and considerable judgment and experience
are required for identification purposes. Basically, if the building
appears unusual in the vertical dimension in some path or has
major setback, then this PMF should be subtracted.
vi) Soft Story

This is referring to a major discontinuity in a building, where the
stiffness of one story is dramatically less than that of most of the
others. A soft story can be considered as one in which the lateral
stiffness is less than 70% of that in the story immediately above or
less than 40% of the combined stiffness of the three stories above.
(ATC-21, 1988).
The ground story often soft due to large window opening for
display purposes or due to open multi-storey car parks below.
Assume that, if one story is particularly tall and/ or has windows
on all sides, it is soft storey if the stories above have many fewer
138
windows. If two to three adjacent walls at that story height are

solid or only have small openings, then the torsion PMF should be
used.
A building may be adequate in one direction but be soft in the
perpendicular direction. For example, the front and back walls may
be open but the sidewalls may be solid.
vii) Torsion
These PMFs applies in the situation where significant torsion or
twisting may occur in the building due to major eccentricities in
the lateral force resisting system. This situation is especially
prevalent among corner and wedge shape building, in which the
two adjacent street of the building will be largely windowed and
open, whereas the other sides are generally solid.
viii) Plan Irregularity

This involving sharp re-entrant corner, where increased damage is
likely to occur. This problem is prevalent in buildings with long
wings that are L, E, or T shaped. If the length of a wing
projection beyond a re-entrant corner is greater than 15% of the
plan dimension in the given direction, the structure can be
considered to have plan irregularity. (ATC-21, 1988)
ix) Pounding
Pounding occurs when there is a little or no clearance between
adjacent buildings, and the building impact or pound against
each other as they deflect in earthquake. If the building floors are
at different elevations, this particularly significant, as one
buildings floors may pound against the others columns,
destroying the column and resulting in partial or total collapse.
139
Pounding applies only to frame buildings (S1, S2, S4, C1, and
PC2). (ATC-21, 1988)
Pounding is a problem only if there is an insufficient gap
between adjacent buildings. Preferable, there should be 4 inches of
separation per story to prevent pounding. (ATC-21, 1988)
x) Large Heavy Cladding

If large heavy cladding elements are present, usually precast
concrete or cut stone, significant numbers of these panels may fall
off the building during an earthquake, posing a life-safety hazard
especially over the exit doors.
For this PMF to apply, two key criteria must be met:
1) There must be more than one panel.
2) The connections must have been designed for seismic
resistance of 200% of gravity loads. (ATC-21, 1988)
xi) Short Column

Also termed sub-columns, this is primarily a defect in concrete
buildings, and consists of columns that were designed as having
full story height but which, because of lateral addition of spandrel
beams or wall sections between the columns, having an effective
height substantially less than the fully story height. The effect of
the wall or spandrel beams is to increase the bending stiffness of
the column, so that it attracts much more lateral load than it was
designed for. Often this results in brittle shear failure of the
column. This is an especially severe problem when the stiffening
beams or walls are not placed symmetrically, so that the shear
forces are even more concentrated.
140
xii) Interpretation of Results

The result of the RSP from the ATC-21 of surveyed buildings,
which may be divided into two categories: those acceptable as to
risk to life safety or those that may be seismically hazardous and
should be studied further. Buildings identified by this procedure as
potentially hazardous must be analyses in more detailed. Because
this rapid visual screening is designed to be performed from the
street without benefit of entry to the building, in some cases
hazardous may prove to be adequate.
Database System
The objectives of this research can be achieved when all the data
and analysis study has been compiled and can be referred in the
form of database system. This database can be used by public to
evaluate the new building in efficient ways to do seismic
evaluation and also can review the existing building that has been
evaluated. The analysis result and raw data also can be retrieved.
Figure 9.3 shows the main interface for the database system. This
database is quite restricted where the user cannot edit or do any
changes to the data that has been stored except the administration
which is only allowed.
141
Figure 9.3 Main Interfaces for Database
142
Figure 9.4 Database System
The database system stored the data from building that has been
analyzed by ATC-21. There are 65 building that has been
evaluated. Database also contains building map and picture as
shown in Figure 9.4.
Digital Form for ATC-21

The digital form of ATC-21 has also been developed to simplify
the processes and procedures become easier. Figure 9.5 shows the
interface of digital form for ATC-21. The digital form was built
similar to the original form, so that the user maybe engineer or
public can easily understand.
143
Figure 9.5 Digital Form for ATC-21
CONCLUSION
From this study, the digital input method was introduced to
evaluate the existing building. This evaluation assigned the final
score for the particular building and this shows whether the
buildings need to further detail analysis or not. Basically, this study
was a preliminary stage of evaluation. The digital form of ATC-21
would be very useful to the respective parties in determining
whether our buildings need to do the detail analysis for seismic
loading or not. The compiled data and the result in database can be
easily retrieved by public for their awareness and references.
144
REFERENCES
Applied Technology Council (ATC), ATC21 and ATC22, 1971
Bishop-Clark, Comparing Understanding of programming Design
Concept Using Visual Basic and Traditional Basic. Journal of
Educational Computing Research, 1998.
Farzad Naeim. The Seismic Design Handbook, Van Nostrand
Reinhold, 1989
Rozaina Binti Ismail (2008). The Application of Artificial Neural
Network in Seismic Evaluation of Buildings. Master Degree
Thesis: University of Technology Malaysia.
Francisco Cabello, Dermot Barnes-Holmes, Denis OHora and Ian
Stewast, Using Visual Basic in the experimental analysis of
Human Behavior: A Brief Introduction, Maynooth, 2002.
Mary R. Sweeney, Automation Testing Using Visual Basic 6.0,
2000.
127
Indeks
INDEX
Artificial Neural
Network(ANN),
111, 113-117, 122-125
Backpropagation
algorithm, 111
Base Isolator, 93
Bridge Inspection, 5, 11, 16
Nondestructive
testing
(NDT),
112-114, 117, 124, 126
Visual inspection, 112114, 117, 124, 135
Building Evaluation, 128
Applied Technology
Council (ATC), 35, 47, 49
59, 127-134, 136-139, 141
143
Rapid Screening Procedure
(RSP), 51, 128, 130, 131,
133, 134, 139
Dynamic Analysis, 37, 47
El Centro Earthquake, 37,
93,
98, 99
Free Vibration Analysis
(FVA), 37, 40, 61, 67, 69,
77,
83, 85, 94
Natural period, 37, 40, 41,
Earthquake Intensity, 100, 101

Macrozonation, 14, 16, 64, 80,
117, 118, 119
Attenuation, 2, 11, 13, 14,
16, 17
Regional Seismicity, 6, 9,
Seismotectonic, 2, 4
Seismic Hazard, 1, 2, 8,
11, 14, 17, 19, 20, 24, 25,
35, 51, 59, 125
Seismic Hazard Map, 1, 2,
24, 25, 125
Seismic Source, 8, 9, 39
Microzonation, 19, 20, 32, 33,
192
Ground Response, 20-22,
26, 27, 32, 33, 65, 81
Ground Motion, 8, 11, 13,
16, 17, 20, 24, 25, 27, 61,
62, 75, 71, 79, 89, 92, 93,
98-101, 108
Prestress Concrete, 93, 95
Index
67,
69, 83, 85
Response Spectrum
Analysis(RSA), 47, 61, 68,
70, 74, 77, 81, 82, 84, 90,
94,
95, 104,
San Fernando Earthquake,
98,
99
Time History
Analysis(THA),
40, 61, 67, 68, 70, 79, 83,
86
Programming, 111, 116-118
Database, 6, 127, 128, 130,
139
142
Rubber Bearing, 93,94,109

Advanced in Earthquake Engineering Applications

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advanced in Earthquake Engineering Applications

Uploaded by

Copyright:

Available Formats

First Edition 2008

MOHD. ZAMRI RAMLI, ROZAINA ISMAL & MELDI SUHATRIL 2008

Advances in earthquake engineering applications / editor Mohd Zamri Ramli,

Fakulti Kejuruteraan Awam

34 38, Jalan Kebudayaan 1, Taman Universiti,

Lot 47 & 48, Jalan SR 1/9, Seksyen 9

Development of Seismic Hazard Maps of 1

Vulnerability Study of Public Buildings 35

Buildings Classification using

Seismic Performance of Sultan Azlan Shah 61

Seismic Performance of Rapid KL Elevated 77

Analysis of Prestress Concrete Highway 93

Predicting of Bridge Condition Based on 111

Database System and Digital Earthquake 127

This book talks about applications of advances earthquake

finite element modeling. The vulnerability and analysis of

Mohd Zamri Ramli

Advances in Earthquake Engineering Applications

for occupancy. The earthquake of magnitude 4.8 that occurred on

Advances in Earthquake Engineering Applications

and active. According to the Seismo-Tectonic Map published by

Generally tectonic features that affected East Malaysia can be

Advances in Earthquake Engineering Applications

subducting an island of Sulawesi on the north side as shown in

Figure 1.2 Seismotectonic setting around East Malaysia (JPKM, 1994)

Figure 1.3 Cross section of North Sulawesi Subduction Fault (Cardwell,

Advances in Earthquake Engineering Applications

The Sangihe plate is subducting down to more than 300 km on

Advances in Earthquake Engineering Applications

Figure 1.4 Fault sources in Borneo Island (McClay, 2000)

Advances in Earthquake Engineering Applications

Figure 1.5 Distribution of epicenters around project Location

Some analyses have been performed to the earthquake catalogues

Advances in Earthquake Engineering Applications

completely reported only during the most recent 41 year

Advances in Earthquake Engineering Applications

Figure 1.6 Seismic source models for East Malaysia.

The NSS was investigated by plotting the spatial distribution of

Advances in Earthquake Engineering Applications

Figure 1.7 Earthquake spatial distributions

Figure 1.8: The hypocentral profiles of NSS

Advances in Earthquake Engineering Applications

SEISMIC HAZARD PARAMETERS

Advances in Earthquake Engineering Applications

b) Other shallow crustal

(c) N-S Megathrust

Figure 1.9 Recurrence relationship chart for East Malaysia

Advances in Earthquake Engineering Applications

There are several attenuation relationships derived for

Advances in Earthquake Engineering Applications

Figure 1.10 Simple logic tree for incorporation of model uncertainty

THE RESULTS OF ANALYSIS

Advances in Earthquake Engineering Applications

Advances in Earthquake Engineering Applications

Advances in Earthquake Engineering Applications

America. Bulletin of the Seismological Society of America.

Advances in Earthquake Engineering Applications

response and damage were varying significantly due to variation of

Advances in Earthquake Engineering Applications

approaches to include the effect of nonlinearity of soil material into

Advances in Earthquake Engineering Applications

the studies were performed in the 1970s in Japan. Since then,

Advances in Earthquake Engineering Applications

Figure 2.1 The result of seismic down-hole test