Seismic performance of a bridge subjected to far-field ground motions by a Mw 9.0 earthquake and near-field ground motions by a Mw 6.9 earthquake

(1)

i

Seismic performance of a bridge subjected to far-field ground motions by a Mw 9.0 earthquake and near-field ground motions by a Mw 6.9

earthquake

REINA GOTO

Master of Science Thesis

Stockholm, Sweden 2012

(2)

(3)

Seismic performance of a bridge

subjected to far-field ground motions by a Mw 9.0 earthquake and near-field

ground motions by a Mw 6.9 earthquake

Reina Goto

June 2012

TRITA-BKN. Master Thesis 358 ISSN 1103-4297

ISRN KTH/BKN/EX-358-SE

(4)

©Reina Goto, 2012

Royal Institute of Technology (KTH)

Department of Civil and Architectural Engineering Division of Structural Engineering and Bridges Stockholm, Sweden, 2012

(5)

i

Preface

This Master’s thesis was initiated with the help of Kawashima Research Group at the Department of Civil Engineering at the Tokyo Institute of Technology, Tokyo Tech, and was carried out at the Department of Civil and Architectural Engineering at the Royal Institute of Technology, KTH.

First and foremost, I would like to give my sincere gratitude to Professor Kazuhiko Kawashima, Tokyo Tech, and Assistant Professor Hiroshi Matsuzaki, Tohoku University. I would like to thank Professor Kawashima for his great support and advices. I would like to thank Assistant Professor Matsuzaki for helping me with all my questions and to understand the seismic design methods and the dynamic response analysis better.

I would like to thank my supervisor Post Doctor Nora Ann Nolan, KTH, for her great support and giving me valuable advices.

I would also like to thank my examiner Professor Raid Karoumi, KTH, for showing great interest and being very helpful during the course of the thesis.

Special thanks to the members of the Kawashima Research Group, Tokyo Tech, for all their help and for sending me papers and files needed for the research.

Stockholm, June 2012 Reina Goto

(6)

(7)

iii

Abstract

In the last two decades, two major earthquakes have occurred in Japan: the 1995 Kobe earthquake and the 2011 Great East Japan earthquake. In the 2011 Great East Japan earthquake, many bridge structures were destroyed by the tsunamis, but it is interesting to study the ground motion induced damage and also how this earthquake differed from the one in 1995. In this thesis, the seismic response of a bridge designed according to the current Japanese Design Specifications was evaluated when it is subjected to near-field ground motions recorded during the 1995 Kobe earthquake and far-field ground motions recorded during the 2011 Great East Japan earthquake. For this purpose, a series of nonlinear dynamic response analysis was conducted and the seismic performance of the bridge was verified in terms of its displacement and ductility demand.

It was found from the dynamic response analysis that the seismic response of the target bridge when subjected to the ground motions from the 2011 Great East Japan earthquake was smaller than during the 1995 Kobe earthquake. Although the ground motions from the 2011 Great East Japan earthquake were very strong, they were not as strong as the ground motions from the 1995 Kobe earthquake. The results obtained in this thesis clarify the validity of the Type I and Type II design ground motions. The target bridge used in this thesis was designed according to the post-1990 design specifications and showed limited nonlinear response when subjected to the different ground motions which shows how efficient the enhancement of the seismic performance of bridges has been since the 1990’s.

Keywords: seismic performance, dynamic response analysis, far-field ground motions, near-field ground motions

(8)

(9)

v

Notations

Main notations

Ah Sectional area of each lateral confining reinforcement Aw Sectional area of reinforcing bars

b Width of the column section cc Cyclic loading effect factor cD Damping modification factor ce Effective height factor

cpt Modification factor depending on the longitudinal tensile reinforcement ratio cR Factor depending on the bilinear factor

cs Response modification factor cZ Zone modification factor

d Effective length of lateral confining reinforcement D Effective height of the column section

dR Residual displacement developed at a column dRa Allowable residual displacement at a column

dRa,LG Allowable residual displacement in the longitudinal direction dRa,TR Allowable residual displacement in the transverse direction du Ultimate displacement of column

dy Yield displacement of column Ec Young’s modulus of concrete Edes Descending gradient

fc Strength of concrete

fcc Strength of confined concrete fck Design strength of concrete

fsy Yield strength of reinforcement bars gal Measure of acceleration (1 gal = 1 cm/s²)

h Height of the column/effective height of the column section

(10)

vi khc0 Standard modification coefficient khc Design horizontal seismic coefficient Lp Plastic hinge length of column My0 Initial yield moment

Mu Ultimate moment P Lateral strength

Pa Lateral capacity of a column Ps Shear strength of a column

Ps0 Shear strength under static loading of a column Pu Ultimate lateral strength of a column

r Bilinear factor

s Spacings of lateral confining reinforcement

S Response acceleration spectrum for the Level 1 earthquake ground motion S0 Standard acceleration spectra for Level 1 earthquake ground motion SI Response acceleration spectra for Type I ground motion

SII Response acceleration spectra for Type II ground motion Sc Shear capacity resisted by concrete

Si Standard acceleration response spectra

Ss Shear capacity resisted by transverse reinforcement T Fundamental Period

Ti Natural periods W Equivalent weight Wp Weight of the column

WU Weight of part of the superstructure supported by the column concerned

 Shape factor/safety factor



Shape factor

c Strain of concrete

cc Strain of concrete under the maximum compressive stress

s Volumetric ratio of lateral confining reinforcements

sy Yield point of the reinforcements

 Damping ratio

a Design displacement ductility factor of a column

(11)

vii

R Response displacement ductility factor of a column

c Average shear stress that can be borne by concrete

u Ultimate curvature

y Yield curvature

Abbreviations

AIS Arc Information Systems

EW “East-West” horizontal component of ground motion JRA Japan Road Association

LG Longitudinal

NIED National Research Institute for Earth Science and Disaster Prevention NS “North-South” horizontal component of ground motion

PGA Peak ground acceleration RC Reinforced concrete

SPL Seismic Performance Level TR Transverse

UD “Up-Down” vertical component of ground motion WSJ The Wall Street Journal

(12)

(13)

ix

Chapter 1 Introduction

Japan is situated on a region where several tectonic plates meet, which is why Japan is extremely prone to earthquakes. There have been many earthquakes in the past and many lessons to be learnt alongside it. Japan has made huge investments to improve buildings and infrastructures to mitigate seismic damage. The Japanese seismic design codes have been revised several times and revisions are sure to appear in the future.

On March 11^th 2011, Japan was hit by a huge earthquake called 2011 Great East Japan earthquake. It was the biggest earthquake ever recorded in Japan and it was apparent that the country was not prepared for the kind of damages that followed the earthquake. Not only did this earthquake cause immense damage and casualties, but it also caused the biggest nuclear disaster since Chernobyl in 1986 to further grieve the people of Japan. Many bridge structures were destroyed by the tsunamis, but it is interesting to see how the ground motions of the earthquake damaged these bridge structures. In the context of seismic design of bridges, perhaps there are lessons to be learnt from this earthquake. To mitigate seismic damage of bridges, it is important to find out how this earthquake differed from other earthquakes in the past and whether or not the Japanese Seismic Design Specifications for bridges need to be revised.

1.1 Aim and scope of thesis

The aim of this thesis is to evaluate the seismic response of a bridge designed by the current Japanese seismic design codes when it is subjected to ground motions recorded during the two major earthquakes that have occurred in Japan in the last two decades:

the 2011 Great East Japan earthquake and the 1995 Kobe earthquake. For this purpose, a series of nonlinear dynamic response analysis of a bridge is conducted. The seismic performance of the bridge is then verified in terms of its displacement and ductility demand.

First, the seismic history of Japan will be studied to understand the damages that have occurred in the past. The 1995 Kobe earthquake and the 2011 Great East Japan

(18)

2

earthquake will be studied in more detail, since the ground motion records from these earthquakes will be used in this thesis. The current Design Specifications and how it has changed over the years since its first publication will also be studied. A literature review to find any information that is relevant to this thesis will be conducted and presented.

Ground motion records from the 1995 Kobe earthquake and the 2011 Great East Japan earthquake will be evaluated to see differences in the ground motion characteristics.

Also response acceleration spectra for the ground motion records will be analyzed to see and compare the intensity and predominant period of each ground motion. A bridge based on the Japanese Seismic Design Specifications is used to conduct a dynamic response analysis using a Japanese finite element analysis program called TDAP III. The seismic response and seismic performance of the bridge when subjected to different ground motions will then be evaluated.

General steps in this thesis are:

1. Evaluate how the two earthquakes differ in character. The ground motion characteristics will be compared as well as its response acceleration spectra.

2. Conduct dynamic response analysis using TDAP III.

3. Compare the seismic response of the bridge and evaluate its seismic performance based on the results from the dynamic response analysis.

4. Discuss whether or not the current Japanese Seismic Design Specifications for bridges are sufficient for an earthquake with a different character than the 1995 Kobe earthquake such as the 2011 Great East Japan earthquake.

Several assumptions and simplifications were made in this study. The target bridge was taken from an example book issued by the Japan Road Association and it was designed based on nonlinear static analysis. In reality nonlinear dynamic response analysis should be conducted when designing a bridge, but in this case the bridge was designed based on only nonlinear static analysis for simplicity.

In the dynamic response analysis, the difference in the arrival time of the earthquake ground motions were not considered since the length of the target bridge is only 0.2 km.

The difference in the arrival time should be considered for longer bridges. Also, torsion and shear deformation were not considered in this analysis.

The damping ratios of the elastomeric bearings, soil springs, and structural components of the bridge were assumed using values from the Japanese Design Specifications (JRA, 2002). These damping ratios were assumed since the aim of this study is to compare the seismic response of the target bridge when subjected to

(19)

3

different ground motions and to find a precise damping is not of interest. Damping of the bridge structure was idealized using Rayleigh damping (see Section 3.5.2 for the calculations of the Rayleigh coefficients and the damping curve).

1.2 Organization of thesis

Chapter 2 gives some background information necessary for this thesis. A short summary of the seismic history of Japan is presented and the 2011 Great East Japan earthquake and 1995 Kobe earthquake are presented in detail. The history of seismic design in Japan is summarized linking the revisions of the Seismic Design Specifications to damages that were observed after some of the major earthquakes.

Also the current Japanese Seismic Design Specifications are presented. Previous studies of relevance are discussed. The methods of analysis, the target bridge, and the bridge model are presented in detail in Chapter 3. The results of the analysis are presented in Chapter 4 and are analyzed. In Chapter 5, the conclusions that were deduced from this study are presented and any suggestions for future research are discussed.

(20)

4

(21)

5

Chapter 2 Background and previous studies

2.1 Seismic history of Japan

Japan has a long history of earthquakes and some of the more significant earthquakes, in terms of seismic design, will be presented here. In the early 1900’s when seismic effects were either not or poorly considered in design, the 1923 Kanto earthquake with a moment magnitude of 7.9 occurred in the Tokyo-Yokohama area (Kawashima, 2000).

This earthquake caused large scale damage to buildings and infrastructure, where bridges collapsed due to tilting, overturning, and settlement of the foundations. Due to this earthquake, the importance of considering seismic effects in design was recognized for the first time (Kawashima, 2011).

In 1964, an earthquake with a moment magnitude of 7.5 occurred in Niigata which came to be called the “1964 Niigata earthquake”. Many bridges were damaged or collapsed due to soil liquefaction and it was at this time that the actual term

“liquefaction” was first coined (Kawashima, 2011). It became evident after this earthquake that soil liquefaction needed to be considered in seismic design. However, at that time, further research to understand the mechanism of liquefaction was needed before implementing any countermeasures. Bridges were also damaged by large relative displacements of the decks, which inspired the development and implementation of unseating prevention devices.

After the 1964 Niigata earthquake and up to the early 1990’s, several big earthquakes occurred. However, the damages in these earthquakes were quite limited due to changes in seismic design practices. It was not until 1995, that a big earthquake that would greatly change the seismic design practices in Japan occurred. This was the 1995 Kobe earthquake and it had a great impact on the seismic design of bridges. Even to this day, ground motions from this earthquake are used for dynamic response analysis of bridges. Since the ground motions from the 1995 Kobe earthquake are used in this thesis, it is presented more in detail in Section 2.1.1. Similarly, the recent 2011 Great East Japan earthquake is presented in detail in Section 2.1.2, since ground motions

(22)

6

from this earthquake are also used in this thesis and the damages that occurred need to be thoroughly described.

2.1.1 1995 Kobe earthquake

In 1995, 17^th of January, the 1995 Kobe earthquake occurred at Kobe and Awaji Island, in southern Japan. This earthquake had a moment magnitude of 6.9 and near-field ground motions were recorded. Thousands of deaths and extensive damage to buildings and infrastructures were reported. Many bridges suffered damage, where 9 highway bridges collapsed or nearly collapsed and 16 bridges were severely damaged. The four major types of damages that were observed are summarized below, based on the lecture notes from “Seismic design of urban infrastructure” (Kawashima, 2011).

2.1.1.1 Shear failure of RC columns

Figure 2.1 shows the collapse of the 18-span Fukae Viaduct of the Hanshin Expressway in Kobe. This viaduct was designed based on the 1964 Design Specifications that will be presented later in Section 2.2. During the earthquake, the RC columns which were 9.9 m to 12.4 m tall with a diameter of 3.1 m to 3.3 m were damaged by severe flexural and diagonal cracks that developed 2.5 m above the footing. This was where one third of the longitudinal reinforcement bars terminated. Since the amount of tie bars were not enough, premature shear failure shown in Figure 2.2 also occurred in the columns.

These damages occurred due to the deficiencies in design. For instance, the allowable shear stress was overestimated and the development length of the longitudinal bars was insufficient. This kind of failure occurred in several other bridges as well such as in the Takashio Viaduct which was built according to the 1971 Design Specifications.

Figure 2.1: The collapse of the Fukae Viaduct. (Kawashima, 2011)

(23)

7 2.1.1.2 Collapse of steel columns

Steel columns collapsed in numerous bridges and an example of this is Tateishi Viaduct at the Hanshin Expressway. A picture of a collapsed column from this viaduct is shown in Figure 2.3. This viaduct was also built based on the seismic coefficient method from the 1964 Design Specifications. The steel columns were built between two RC columns at the sides and lateral beams were constructed to support two side decks. To reduce damage of the steel columns in the event of an automobile accident, the inside of the columns were filled with weak concrete from the bottom up to a height of 2.3 m.

During the earthquake, local buckling of web and flange plates and rupture of the welded corners at the bottom of the columns occurred. This caused the bearing capacity of the columns to decrease in the lateral and vertical directions. The columns became vulnerable to the dead weight of the decks and started to settle. When this happened, the decks in the center started to buckle and in the end the steel columns collapsed.

Figure 2.3: Collapse of a steel column of the Tateishi Viaduct. (Kawashima, 2011) Figure 2.2: Premature shear failure of column of the Fukae Viaduct.

(Kawashima, 2011)

(24)

8 2.1.1.3 Damage to unseating prevention devices

Damage to various types of unseating prevention devices was observed. This happened since the design force of the devices was too small. The design force was calculated by multiplying the static reaction force by a seismic coefficient of 0.3 to 0.4. In the Nishinomiya Bridge of the Hanshin Expressway, one of the approach spans collapsed (Figure 2.4.a). The main bridge and the approach spans were connected by plate-type restrainers (Figure 2.4.b). During the earthquake, the fixed bearings of the main bridge failed and caused the bridge to displace, pulling the approach span. Eventually the approach spans dislodged from its supports and collapsed since the unseating prevention devices could not support it without the help of the supports.

a) Collapse of an approach span

(Nishinomiya Bridge) b) Failure of a plate-type restrainer Figure 2.4: Damage to unseating prevention devices (Kawashima, 2011).

2.1.1.4 Damage to steel bearings

Extensive damage to steel bearings was also observed in this earthquake (Figure 2.5).

Prior to the 1995 Kobe earthquake, steel bearings were thought to restrict extensive damage to the bridge substructures. However, after observing the damage caused by the failure of steel bearings, it became apparent that steel bearing were one of the main causes of the extensive damage that occurred (Kawashima, 2011). This is because steel bearings are weak for shock and have insufficient strength and length of movement.

Apart from the three above mentioned damages, damage to bridge foundations were also observed. However these damages were minor compared to the rest of the structural components. Damage caused by soil liquefaction was also observed in the form of settlements and tilting of foundations and bridge substructures. Foundations were also damaged by large lateral spreading which was caused by soil liquefaction.

The Japanese Design Specifications were revised in 1996 due to the poor seismic performance of bridges in this earthquake. The revisions that were made in the Design Specifications will be presented in Section 2.2.

(25)

9

a) Failure of steel pin bearing b) Failure of steel bearing Figure 2.5: Failure of steel bearings (Kawashima, 2011).

2.1.2 2011 Great East Japan earthquake

On March 11, 2011 a devastating earthquake of moment magnitude 9.0 occurred off the Pacific coast, northeast of Japan. This was the biggest earthquake ever recorded in Japan and was named the “2011 Great East Japan earthquake”. This earthquake lasted for more than 300s and strong ground motion accelerations were recorded in several areas. The coastal regions of northeast Japan were hit by tsunamis after the earthquake which caused severe damage to buildings and infrastructures, human injuries, and casualties. The earthquake was felt all the way down to the Kanto region and extensive soil liquefaction occurred in the Tokyo Bay area as well as in Chiba Prefecture where damage such as settlements of buildings and uplift of sewage manholes were observed (Ishihara, 2012).

The tsunamis swept away and damaged several bridges along the coast, but damage to bridges which was induced by ground motions was less extensive. However, according to a study by Kawashima et al. (2011) and Kawashima (2012), bridges that were designed based on the design codes prior to the 1990 and 1995 Design Specifications and were not retrofitted were damaged due to the ground motions. Bridges that had been retrofitted or built according to the post 1990 Design Specifications showed only minor damage or no damage at all. This showed that seismic retrofitting and the improvements that had been made in the Design Specifications were efficient. The bridge damage that was observed after the 2011 Great East Japan earthquake is presented within two categories: bridges designed pre-1990 and post-1990.

(26)

10 2.1.2.1 Bridges designed before 1990

The same type of damage to RC columns as in the 1995 Kobe earthquake occurred, which was mentioned in Section 2.1.1. In the Esaki Ohashi Bridge, damage to the RC columns was observed which can be seen in Figure 2.6.a. This type of damage occurred in bridges that were designed before the 1990’s and had an overestimated shear capacity and not enough development length of the longitudinal bars. Kunita Ohashi Bridge was also designed prior to the 1990’s and had not been retrofitted at the time of the earthquake. This bridge was closed for service after the earthquake, since its steel bearings were damaged (Figure 2.6.b) and shear cracks had developed in the RC columns. The information on the damage on the Esaki Ohashi Bridge and the Kunita Ohashi Bridge were obtained from a study by Hoshikuma et al. (2012).

a) RC columns (Esaki Ohashi Bridge)

b) Steel bearings (Kunita Ohashi Bridge) Figure 2.6: Damage to bridges designed prior to 1990-design codes

(Hoshikuma et al., 2012).

2.1.2.1 Bridges that had been retrofitted or designed after 1990

Bridges that had been retrofitted after the 1995 Kobe earthquake, by for example steel jacketing of RC columns and replacing steel bearings with elastomeric bearings, showed in most cases no signs of damage. Bridges that were designed according to the post-1990 Design Specifications were not damaged or showed only minor damages.

However, some bridges suffered severe damage to its elastomeric bearings and dampers.

One of these bridges was the Tobu Viaduct in Sendai, where elastomeric bearings ruptured (Kawashima, 2012). Figure 2.7.a show how the rupture of the bearings caused the bridge deck to offset in the transverse direction by 0.5 m and Figure 2.7.b show that the rubber layers detached from the steel plates and ruptured. Some possible reasons to why this damage occurred could be because of a design miss or that the interaction of adjacent decks was not properly considered (Takahashi, 2012 and Kawashima, 2012). Damage was also observed in the attachments and anchors of dampers (Figure 2.8).

(27)

11 a) Offset due to rupture of

elastomeric bearings b) Rupture of elastomeric bearing Figure 2.7: Damage of elastomeric bearings in the Tobu Viaduct (Kawashima, 2012).

a) Damage of anchors b) Damage of attachment

Figure 2.8: Damage of attachments and anchors of dampers (Takahashi, 2012).

2.2 History of seismic design of bridges in Japan

The revisions and history of the Japanese Design Specifications for seismic design of bridges will be presented in this subsection, based on lecture notes from “Seismic Design of Urban Infrastructures” (Kawashima, 2011) and papers by Professor Kawashima (Kawashima, 2000 and Kawashima, 2006).

In 1926, three years after the 1923 Great Kanto earthquake, the first Japanese seismic provisions for highway bridges were published. In these specifications, the seismic coefficient method using a seismic coefficient of 0.1 to 0.3 was included and only the requirement of seismic lateral force of 20% gravity force was presented. Design specifications of steel highway bridges were included in 1939 and were revised twice afterwards in 1956 and 1964. At this time, earthquake engineering was still something new and under progress, so the seismic design requirements in these specifications were far from what they are now. It was not until the 1964 Niigata earthquake that engineers realized the need for major improvements of the seismic provisions. After

NEXCO East

(28)

12

observing the damages caused by the 1964 Niigata earthquake, a completely renewed seismic design provisions, “Guide Specifications for Seismic Design of Highway Bridges”, were issued in 1971. Some of the improvements and changes that were made are presented below:

 The lateral force should be calculated by considering the zone, importance of the bridge, and ground condition in the seismic coefficient method. Also the structural response should be considered in the modified seismic coefficient method.

 Since many bridges were damaged by soil liquefaction in the 1964 Niigata earthquake, the evaluation of soil liquefaction was included. However, the mechanism of soil liquefaction was unknown at that time so design procedure for liquefaction could not be included in 1971.

 The need for unseating prevention devices were recognized in this earthquake so several types of unseating prevention devices such as steel plate connectors and cable restrainers were developed.

 Many independent methods for the design of substructures had been developed and these methods were unified as “Guide Specifications of Substructures”

between 1964 and 1971. This resulted in the development of new types of foundations which helped reduce the damage of the bridge foundations.

In 1980, the above Guide Specifications for seismic design and substructures were revised. These specifications were written as “Part V Seismic Design” and “Part IV Substructures” in the “Design Specifications of Highway Bridges”. Parts I to III were the “General Aspects”, “Steel Bridges”, and “Concrete Bridges” respectively. A method for the design of foundations in liquefying soils and an updated version of the evaluation method for predicting soil liquefaction were added in Part V. In Part IV, the allowable shear stress for concrete was reduced since this was overestimated in the past. The anchoring length of the reinforcement bars from the footings was increased to 20 times the diameter of the bars and the length equivalent to the effective width of the column.

The Design Specifications were revised again in 1990. In this revision, the following changes were made:

 The seismic coefficient method and the modified seismic coefficient method were unified.

 For the first time, to enhance the ductility of bridge columns, the check of the strength and ductility of the reinforced concrete columns was included. The nonlinear behavior of a bridge was to be checked after the structural members yielded. The Type I ground motion of the standard lateral force coefficient in

(29)

13

Figure 2.9 was introduced for the ductility check. This ground motion represents the ground motions that are assumed to have occurred in the 1923 Kanto earthquake. Type II ground motion was included in the later revisions.

 The static frame method was introduced so that the lateral force distribution of a multi-span continuous bridge could be evaluated. Through this method, the three dimensional behavior of a bridge could be considered in the equivalent static analysis.

Figure 2.9: The standard lateral force coefficient (Kawashima, 2000).

As previously mentioned, even though strong earthquakes occurred several times in the 1980’s and the beginning of 1990’s, the damages were quite limited due to the improvements that had been made in seismic design. Therefore, the damages that resulted from the 1995 Kobe earthquake were somewhat shocking. 40 days after this earthquake, the “Guide Specifications for reconstruction and repair of highway bridges which suffered damage in the 1995 Kobe earthquake” was issued to guide the reconstructions of the bridges that were damaged in this earthquake. This Guide Specifications came to be used in new constructions of bridges as well, until a revised version of the Design Specifications came out in 1996. In this Guide Specifications, a requirement for the design of a plastic hinge at the bottom of columns and the effect of lateral confinement was included. Also, the Type II ground motion in Figure 2.9 was included which represents the ground motions recorded in the 1995 Kobe earthquake.

In 1996, the Design Specifications from 1990 were fully revised and included the above mentioned 1995 Guide Specifications. Some of the major changes that were made are the following:

 The previous check of the ductility of the reinforced concrete columns was improved to the “ductility design method”. Although the seismic coefficient method was still in use, revisions in the Design Specifications were made so that all the structural components that are vulnerable to seismic effects are to be checked with the ductility design method.

(30)

14

 The type of ground motion (Type I and Type II) is to be considered in determining the design ductility factor and shear strength of a bridge column, and also in determining the soil strength for liquefaction.

 Specifications for the dynamic analysis were revised, where revisions were made in the analytical models and methods, and safety checks. Also the input earthquake ground motions to be used in dynamic analysis were specified.

 Requirements for the residual displacement of a column after an earthquake were included and this had to be checked for bridges in the important bridge category.

 An unseating prevention system was introduced and design loads and methods were specified. The function of the unseating prevention devices was also clarified.

 Elastomeric bearings were recommended to be used as opposed to steel bearings which have several deficiencies.

 The seismic design treatment of soil liquefaction was reviewed and is to be used as a seismic design method in places where liquefaction is likely to occur. The seismic design treatment of lateral spreading caused by soil liquefaction was also defined.

Since 1996, the Design Specifications have been revised in 2002. Revisions were made based on the “Performance-based” design concept, where requirements of the necessary performance and verification of policies are clearly stated. Some of the changes that were made are summarized in the following points:

 Seismic performance requirements of highway bridges, principles of seismic performance verifications, and the determination concept of design earthquake ground motion were clearly defined. These specifications were based on concepts from the performance-based design.

 The methods of verifying seismic performance were rearranged to two design methods: “Static analysis” and “Dynamic analysis”. The verification method for the latter analysis was defined in detail and its applicability was improved.

 A method to verify the seismic performance of abutment foundations on liquefiable grounds was included for the first time. Similarly, a method to verify the seismic performance of steel and concrete superstructures was introduced.

(31)

15

The current Design Specifications will be presented in detail below in Section 2.3. The Design Specifications have been revised again in March 2012, but this has not been published yet. Therefore the revisions that were made in 2012 will not be discussed in this study and the Design Specifications from 2002 will be used.

2.3 Current seismic design

In this section, the Design Specifications from 2002 (JRA, 2002) will be presented. The

“Design Specification of Highway Bridges” is issued by the Japan Road Association (JRA) and consists of five parts: Part I Common, Part II Steel Bridges, Part III Concrete Bridges, Part IV Substructures, and Part V Seismic Design. Some key parts of the Part V Seismic Design will be presented in this section based on lecture notes from “Seismic Design of Urban Infrastructures” (Kawashima, 2011), papers by Professor Kawashima (Kawashima, 2004 and Kawashima, 2006), and the English translation of the Part V Seismic Design by JRA (JRA, 2002).

2.3.1 Basic principles

In seismic design, a bridge must be designed so that its required seismic performance is satisfied during an earthquake. The seismic performance of a bridge is determined by the importance of the bridge and also the levels of design ground motion that is likely to occur at the site of construction. Furthermore, the topographical-, geological-, soil-, and site conditions must be considered in seismic design.

Table 2.1 shows the seismic performance matrix. Bridges are categorized into two types; either Type A or Type B. Type A are bridges with standard importance and Type B are bridges with high importance. The importance of the bridge is classified by using Table 2.2. The type of design ground motions is divided into two levels: Level 1 Earthquake which is ground motions with a high probability occurrence and the Level 2 Earthquake which is ground motions with a low probability occurrence. The design response acceleration spectra for these design ground motions can be seen in Figure 2.10. The Level 1 Earthquake is the ground motions that are developed in moderate earthquakes and the ground motion used in conventional elastic design method. The Level 2 Earthquake includes two types of ground motions: Type I and Type II. Type I represents ground motions developed in interplate-type earthquakes with a large magnitude, which targets the ground motions that most likely occurred in the 1923 Kanto earthquake. Type II represents ground motions developed in inland-nearfield- type earthquakes and the ground motions from the 1995 Kobe earthquakes are typical targets of this type. Type I ground motion is characterized as having a large amplitude and longer duration, while Type II is characterized as having strong accelerations and shorter duration.

(32)

16

Depending on the bridge type and design ground motions, the Seismic Performance Level (SPL) needs to be ensured. SPL 1 requires bridge damage to be prevented, which means that the main functions of the bridge must be maintained during an earthquake.

SPL 2 requires limited damage in order to recover its function, meaning that the bridge should only suffer limited damage and be able to recover within a short time. In SPL 3, critical damage of the bridge must be prevented.

a) Level 1 Earthquake

b) Level 2 Earthquake (Type I) c) Level 2 Earthquake (Type II) Figure 2.10: Design acceleration spectra (JRA 2002, Kawashima 2004).

(33)

17

Table 2.1: Classification of importance of bridges (JRA, 2002).

Type Definitions

A

bridges Bridges other than Type B bridges

B bridges

 Bridges of National expressways, urban expressways, designated city expressways, Honshu-Shikoku highways, and general national highways.

 Double-deck bridges and overbridges of

prefectural highways and municipal roads, and other bridges, highway viaducts, etc., especially important in view of regional disaster prevention plans, traffic strategy, etc.

Table 2.2: Seismic performance matrix (JRA, 2002).

Type of Design Ground Motions Standard Bridges (Type-A)

Important Bridges (Type-B) Level 1 Earthquake: Ground Motions with

High Probability to Occur SPL 1: Prevent Damage

Level 2 Earthquake:

Ground Motions with

Low Probability to Occur

Interplate Earthquakes

(Type-I) SPL 3: Prevent

Critical Damage

SPL 2: Limited Damage for Function Recovery Inland Earthquakes

(Type-II)

The loads and load combinations that need to be considered in the seismic design of bridges are the primary and the secondary loads. These loads are shown below in Table 2.3. The combination of the loads should be: primary loads + effects of earthquake.

The loads and its combinations should be determined to give the most unfavorable condition. Depending on the site of construction, not all loads will be considered.

According to JRA, the live load does not need to be considered in seismic design. This is because the live load varies temporally and spatially and during an earthquake, the probability of a full live load occurring is small.

(34)

18

Table 2.3: Primary and secondary loads to be considered in design (JRA, 2002).

Primary loads Secondary loads

Dead load Effects of earthquake

Prestress force

Effect of creep of concrete Effect of drying shrinkage of concrete

Earth pressure Hydraulic pressure Buoyancy or uplift The effects of earthquake include:

 Inertia force due to the dead weight of the structure

 Earth- and hydrodynamic pressure during an earthquake

 Effects of liquefaction and liquefaction-induced ground flow

 Ground displacement during an earthquake

2.3.2 Analytical methods to verify the seismic performance

In the Japanese Design Specifications, to verify the seismic performance of a bridge, the limit state of each structural member should be defined considering the limit states of the bridge. If the response of the structural members due to the design ground motions does not exceed the determined limits, the seismic performance is verified. The limit states of the bridge are the Seismic Performance Levels 1, 2, and 3 which were briefly mentioned in Section 2.3.1. These limit states are determined considering the requirements summarized in Table 2.4 from the Design Specifications.

Table 2.4: Establishing the Seismic Performance Levels (JRA, 2002).

Seismic Performance

Level Limit States

SPL 1 “Mechanical properties of the bridges maintained within the elastic ranges”

SPL 2

“Only the structural member in which the generations of plastic hinges are allowed deforms plastically within a range of easy functional recovery”

SPL 3

“Only the structural member in which the generations of plastic hinges are allowed deforms plastically within a range of the ductility limit of the member”

The design earthquake ground motions, and structural type and limit states of the bridge must be considered when choosing the appropriate analytical method to verify

(35)

19

the seismic performance. The appropriate analytical method is either a static or dynamic analysis. For a proper evaluation of the seismic performance, the nonlinear behaviors of a member might need to be considered so an appropriate analytical method must be chosen to account for these properties. See Table 2.5 for the required analytical method depending on the complexity of seismic behavior and the SPL’s.

When determining the seismic performance by a static analysis, the loads that are caused by an earthquake are added statically to the bridge. The dynamic structural characteristics in the elastic range are considered in the seismic coefficient method when verifying for SPL 1. In the seismic coefficient method, loads that have been calculated by using the seismic coefficient are applied to the bridge statically. From this, the resultant deformations and sectional forces are evaluated. In ductility design method, the deformation properties and dynamic strength of the nonlinear zone of a structure are considered. This method is used for the verification of SPL 2 and SPL 3.

In both the seismic coefficient method and design ductility method, the dynamic seismic forces are changed to a static force by using the seismic coefficient.

When a dynamic method is used for seismic performance verification, the maximum response values of the bridge obtained from the dynamic analysis must be smaller than the allowable values. The response spectrum or time-history response analysis methods are commonly used in dynamic analysis. The most suitable method and model are chosen considering the purpose of the analysis and the earthquake ground motion level.

(36)

20

Table 2.5: Relation between Complexity of Seismic Behavior and Design Methods Applicable to Seismic Performance Verification (JRA, 2002).

2.3.3 Design of RC columns

RC columns are designed so that it satisfies the following requirement in Equation 2.1.

W k

P_a  _hc (2.1)

p p

U c W

W

W   (2.2)

where, Pa is the lateral capacity of a column, khc is the design horizontal seismic coefficient, W is the equivalent weight, WU is the weight of part of the superstructure supported by the column concerned, Wp is the weight of the column, and cp is the equivalent weight coefficient (0.5 for bending failure or shear failure after flexural yielding and 1.0 for shear failure).

Dynamic characteristics of bridges

Bridges without complicated seismic behavior

Bridges with plastic hinges & yielded sections, and bridges not

applicable of the Energy Conservation Principle

Bridges of likely importance of higher modes

Bridges not applicable of the Static Analysis Methods Seismic

Performance to be verified

SPL 1 Static

analysis Static analysis Dynamic

analysis Dynamic analysis SPL 2 & SPL 3 Static

analysis

Dynamic analysis Dynamic analysis

Dynamic analysis Examples of

applicable bridges

Other than bridges shown in the right columns

 Bridges with rubber bearing to disperse seismic lateral forces

 Seismically-isolated bridges

 Reinforced Concrete rigid-frame bridges

 Bridges with steel piers likely to generate plastic hinges

 Bridges with long natural periods

 Bridges with high piers

 Long-span bridges such as cable- stayed bridges and suspension bridges

 Deck-type

& half through- type arch bridges

 curved bridges

(37)

21

The design horizontal seismic coefficient is calculated using Equation 2.3:

z hc

z s

hc cc k c

k  ₀ 0.4 (2.3)

where, cs is the response modification factor, cz is the zone modification factor (= 0.7, 0.85, or 1.0 depending on the zone), and khc0 is the standard modification coefficient.

The response modification factor, needed to calculate the above mentioned requirement, may be calculated using Equation 2.4, which assumes the equal energy principle. The equal energy principle is more conservative than the equal displacement principle.

1 2

1

 

a

cS

 ^(2.4)

where, _ais the design displacement ductility factor of a column.

In order for the RC column to perform according to its expected seismic performance, the response displacement ductility factor, _r , should be smaller than the design displacement ductility factor, _a. However it may not be greatly smaller, since this would result in an overestimation of the response modification factor.

A plastic hinge, which can show ductile behavior when it is subjected to repeated alternate deformations, can be defined at the bottom of each RC column. The plastic hinge region dissipates energy through plastic deformation without collapsing the remaining structural members and by designing these plastic hinges in a proper way can allow the damage that occurs after an earthquake to be localized and repaired more easily (Long and Bergad, 2004). The plastic hinge length is determined in the Japanese Design Specifications using Equation 2.5, however it must be in the interval

D L

D _P 0.5 1

.

0   . In analytical purposes, the plastic hinge is a virtual concept which allows the displacement due to plastic deformation at the defined plastic hinge region to be evaluated more easily (Kawashima, 2011).

D h

L_P 0.2 0.1 (2.5)

where, h is the height of the column and D is the effective height of the column section.

For every column which has a defined plastic hinge, a fiber element analysis is performed at the plastic hinge regions assuming a stress vs. strain relationship for concrete and reinforcing bars. An elastic-perfect plastic model is used to idealize the stress vs. strain relationship of reinforcing bars. The stress vs. strain relationship of

(38)

22

confined concrete is based on Hoshikuma et al. (1997) which is presented below in Equation 2.6 to Equation 2.11.

 

























 



 





cc c des cc

n

cc c c

c c

E f E n f



 

1 1

1



0c cc



(2.6)



_cc _c _cu





E_c _cc^c ^ccf_cc



n E

 



 (2.7)

sy s ck

cc f f

f  3.8 (2.8)

ck sy s

cc f

 f



 0.0020.033 (2.9)

sy s

ck

des f

E f

 



2

2 .

11 (2.10)

018 . 4 0

 sd A_h

s ^(2.11)

where, fc is the strength of concrete, fcc is the strength of confined concrete, fck is the design strength of concrete, fsy is the yield strength of reinforcement bars, _cis the strain of concrete, _ccis the strain of concrete under the maximum compressive stress, Ec is the Young’s modulus of concrete, Edes is the descending gradient,



and  are shape factors, and _s is the volumetric ratio of lateral confining reinforcements, Ah is the sectional area of each lateral confining reinforcement, and s and d are the spacings and effective length of lateral confining reinforcement. The shape factors are obtained by the following Table 2.6.

Table 2.6: Shape factors for circular and rectangular columns.

Circular Rectangular



1.0 0.2

 1.0 0.4

According to the Design Specifications, the volumetric ratio of lateral confining reinforcements should be smaller than 1.8%. This recommendation was proposed since there must be a limitation to how much the ductility capacity of a column should be enhanced by just increasing the amount of lateral reinforcement. If the restraining force of concrete is too high, the plastic hinge region will generally become smaller when the column is subjected to repeated plastic deformations. This can cause the longitudinal reinforcements to fracture leading the column to reach the ultimate state.

(39)

23

The ultimate displacement, du, is defined as the displacement at the gravity center of a superstructure when the compression strain of the concrete at the out-most reinforcements reaches the ultimate strain,_cu, in Equation 2.12. The ultimate strain is dependent on the type of ground motion.







 

des cc cc

cc

cu

E f 2 .

 0



Type I ground motion

(2.12) Type II ground motion

The ultimate displacement of a column, du, is found using Equation 2.13 (Priestly and Park 1987 and Priestly et al. 1996).

 

_



 



 





 2

p p

y u y u

h L L d

d   (2.13)

where, dy is the yield displacement, ^u is the ultimate curvature, _y is the yield curvature, h is the height of the column, and Lp is the length of the plastic hinge.

The shear strength of a RC column, Ps, is evaluated according to the following equations:

^P^s ^^S^c^^S^s (2.14)

d b c

c c

S_c  _c _e _pt_c  _(2.15)

 

a d S_s A^w ^sy

15 . 1

cos sin 

 

 (2.16)

where, Ps is the shear strength, Sc and Ss is the shear capacity resisted by concrete and transverse reinforcement, _c is the average shear stress that can be borne by concrete, cc is the cyclic loading effect factor which can be obtained from Table 2.7, ce is the effective height factor, cpt is the modification factor depending on the longitudinal tensile reinforcement ratio, b is the width of the column section, h is the effective height of the column section, Aw is the sectional area of reinforcing bars with interval a and angle and ^sy is the yield point of the reinforcements.

(40)

24

Table 2.7: The cyclic loading effect factor, kc.

Load Type kc

Static loading 1.0 Type I ground motion 0.6

Type II ground

motion 0.8

The failure mode of a column is classified as either flexural failure, shear failure after flexural yielding, or shear failure and this is to be evaluated using Equation 2.21. The failure mode is decided based on the ultimate lateral strength Pu, shear strength Ps, and shear strength under static loading Ps0 of a RC column.

s

u P

P  : Flexural failure

(2.17)

0 s u

s P P

P   : Shear failure after flexural yielding

u

s P

P₀  : Shear failure

The lateral strength of the RC column, Pa, is calculated depending on the failure mode using Equation 2.18:







0 s u

a P

P P : Flexural failure + shear failure after flexural yielding (2.18) : Shear failure

The ductility capacity of the RC column, _a, is also calculated depending on the failure mode:









 

 1 1

y y u

a d

d d

  : Flexural failure

(2.19) : Shear failure after flexural yielding + shear failure

where, du and dy are the ultimate and yield displacement,



is the safety factor which is determined based on the Seismic Performance Level and the type of ground motion.

See Table 2.8 for the safety factors.

Table 2.8: Safety factor



. Seismic Performance

Level

Type I ground motions

Type II ground motions

SPL 2 3.0 1.5

SPL 3 2.4 1.2

(41)

25

The residual displacement dR developed at a column, should satisfy the requirement in Equation 2.20, which states that the residual displacement should be smaller than the allowable residual displacement dRa. The allowable residual displacement is 1% of the distance from the bottom of the column to the height of inertia force of the superstructure.

Ra

R d

d  (2.20)



R

 

y

R

R c r d

d   1 1 _(2.21)

 

_^_







 









 

 1

2

1 ²

a s

R g P

 S _(2.22)

where, dRa is the allowable residual displacement, r is the bilinear factor (ratio of yield stiffness and post-yield stiffness), cR is a factor depending on the bilinear factor, _Ris the response displacement ductility factor.

2.4 Previous studies

An evaluation of the seismic performance of RC bridge piers designed by the pre- and post- 1995 Kobe earthquake was conducted by Kawashima (2000). A cantilever RC column of a four-span continuous bridge was used in this study which was designed by the 1964, 1980, 1990, and 1995 Design Specifications respectively for comparison. The four columns were assumed to have the same conditions except for changes in the size and reinforcement of the column. After evaluating the columns based on the 1995 Design Specifications, it was found that the column designed by the 1964 Specifications was the only one that failed in shear. However the column of the 1980 Specifications failed in flexure and the 1990 column suffered extensive flexural damage. The 1995 column did not suffer damage or fail in either of the two above-mentioned ways. To evaluate the seismic performance of the columns, a dynamic response analysis using a ground motion record from the 1995 Kobe earthquake was conducted for all the columns except for the 1964 column since it failed in shear. Through this study, it has been found that the column based on the 1964 Design Specifications overestimates the allowable shear stress and together with the inadequate anchorage of the reinforcements, makes this column vulnerable to damage. The 1980 column was also vulnerable to flexural damage when subjected to the ground motion type recorded in the 1995 Kobe earthquake.

A study by Matsuzaki (2012) evaluating the intensity of the ground motions from the 2011 Great East Japan earthquake based on nonlinear seismic response of standard

(42)

26

bridges was recently published. Response acceleration spectra are usually used to evaluate the intensity of ground motions, however intensity evaluated by nonlinear response of bridges can be more reliable. In this study, three bridges designed by the Japanese Design Specifications were used. All three bridges were a three-span continuous plate girder bridge with four RC columns and five elastomeric bearings on each column, but the dimensions of the RC column are different in each of the bridges.

In the nonlinear response analysis, the bridges were subjected to four ground motion records from the 2011 Great East Japan earthquake. Also, one ground motion from the 2008 Iwate-Miyagi earthquake and two ground motions from the 1995 Kobe earthquake were used for comparison. It was found from the analysis that although the response acceleration of the ground motions recorded in the 2011 Great East Japan earthquake were high at a period shorter than 0.3 s, the seismic response of the bridges was small when subjected to these ground motions. JMA Furukawa and JR Takatori had similar response accelerations at the natural period of the target bridges, but the peak deck displacement under JR Takatori was much larger than under JMA Furukawa. This type of difference in response cannot be predicted by only looking at the response acceleration spectra. JMA Furukawa ground motion record from the 2011 Great East Japan earthquake developed the largest response out of the ground motions from this earthquake, but the response was smaller than that of the JR Takatori ground motion record from the 1995 Kobe earthquake. According to this study, the ground motion records from the 2011 Great East Japan earthquake can be said to be smaller than the Type II design ground motion. Finally it was concluded that the seismic response of bridges evaluated by nonlinear response analysis are different than the expected response based on only response acceleration spectra, so the intensity of ground motions should be evaluated based on nonlinear response analysis of several target bridges with different structural properties.

(43)

27

Chapter 3 Methodology

In this chapter, the ground motions from the 2011 Great East Japan earthquake and the 1995 Kobe earthquake that will be used in this study will be presented. The ground motion characteristics will be evaluated as well as the response acceleration spectra.

The target bridge that will be used in this study and a detailed design process of the RC columns will be presented to understand how the Seismic Design Specifications introduced in Section 2.3 are used in the design process. The idealizations that were made in modeling the target bridge will be presented. To find the seismic response of the target bridge when subjected to the ground motions, three types of analysis (self- weight, Eigen value, and dynamic response analysis) are conducted using a Japanese finite element analysis program which will be explained later in this chapter. At the end of this chapter, a sensitivity analysis and convergence study that was conducted when choosing certain values that will be used in the analysis will be presented.

3.1 Ground motions

In this study, three far-field ground motions recorded during the 2011 Great East Japan earthquake and two near-field ground motions recorded during the 1995 Kobe earthquake are used in the dynamic response analysis. The ground motion records from the 2011 Great East Japan earthquake were obtained from the Kyoshin Net, abbreviated “K-Net”, which is a Japanese strong motion seismograph network that provides access to strong motion data. K-Net can be accessed online from the National Research Institute for Earth Science and Disaster Prevention’s website (NIED, 2011).

The strong motion data are obtained from 1000 observatories set up throughout Japan.

In this study, since the ground motion records from K-Net did not start from zero acceleration, the data was adjusted to zero start. The ground motions from the 2011 Great East Japan earthquake has a long duration and in order to reduce the computational time for the analysis, the first 10 s were cut since cutting the beginning by this amount only caused an insignificantly small change in the response of the target bridge. See Section 3.8 for the “Sensitivity analysis”.