You are on page 1of 12

Revised Design Specifications for Highway Bridges in Japan

and Design Earthquake Motions


Keiichi Tamura
ABSTRACT
The Design Specifications for Highway Bridges in Japan were revised in March 2002. The
most important change regarding seismic design in this revision is the adoption of performancebased design criteria. Although we had similar design concepts in the previous specifications,
they are systematically arranged in the newly revised specifications. Two levels of design ground
motions, which are further classified into three kinds of ground motions, three levels of seismic
performance criteria and corresponding limit states are prescribed. A new concept has also been
introduced to the provision of design earthquake motion, and a site-specific ground motion can
be conditionally employed for design practice.

INTRODUCTION
The first requirements for seismic design of highway bridges in Japan were included in the
Details of Road Structures (draft), which were issued in 1926, following the 1923 Kanto
earthquake. Since then, the seismic design regulations for highway bridges have been repeatedly
revised, based on earthquake disaster experience and progress of research. Among them, the
most comprehensive revision was made after the 1995 Hyogo-ken Nanbu (Kobe) Earthquake,
which caused the worst damage to various structures including highway bridges since the 1923
Kanto earthquake [1, 2]. After this earthquake, the Design Specifications for Highway Bridges
were revised in 1996, in which a number of new design techniques were incorporated [3, 4].
They include:
1) Seismic design force that represents destructive near-field ground motion caused by an
inland earthquake,
2) Ductility design method that is applicable to bridge pier, foundation, bearing support and
unseating prevention system,
3) Seismic isolation design,
4) Seismic design of unseating prevention system.
The Design Specifications for Highway Bridges in Japan were revised in March 2002 [5].
This revision is not large in scale, comparing to the previous one in 1996, however, several
important additions and modifications are included. Among them, the most important change
regarding seismic design is the adoption of performance-based design criteria. This revision also
allows the designer to develop a site-specific design ground motion, in addition to using the
standard design ground motion or design spectrum.
_____________
Keiichi Tamura, Team Leader, Ground Vibration Research Team, Public Works Research Institute
1-6 Minamihara, Tsukuba-shi, Ibaraki-ken 305-8516 Japan

This paper first summarizes the basic concepts and principles of current seismic design of
highway bridges, where an emphasis is put on the new design philosophies introduced to the
specifications. Also presented in this paper are the prediction of site-specific earthquake ground
motion and a numerical example.

BASES AND PRINCIPLES OF SEISMIC DESIGN


Basic Concepts of Seismic Design
A Bridge shall be designed to achieve the required seismic performance in accordance with
the level of design ground motion and the importance of the bridge. In designing a bridge, the
earthquake-resistant structural type shall be selected based on topographical, geological and site
conditions, and both individual members and the bridge as a whole system shall be ensured to be
earthquake-resistant.
Highway bridges play key rolls for evacuation, rescuer, first aid, fire fighting and
transporting relief supplies after an earthquake. Hence, it is most essential to secure the safety of
bridge against an earthquake and minimize the influence of its functional deterioration on social
activities. Reflecting such important rolls of bridge, a bridge shall be designed to achieve the
required seismic performance according to the level of design ground motion and the importance
classification.
Principles of Seismic Design
Design Ground Motion
In designing bridges, two levels of design ground motions, i.e., ground motion likely to occur
during the service period of bridge and destructive ground motion less likely to occur during the
service period shall be considered. These design ground motions are termed Level 1 ground
motion, and Level 2 ground motion, respectively. Level 2 ground motion contains the ground
motion resulting from a large plate-boundary earthquake and that from an inland earthquake
which occurs at short distance from a construction site. The former and latter ground motions are
designated as Type I and Type II ground motions, respectively. These design ground motions are
the same as adopted in the previous specifications that were issued in 1996. Type I ground
motion is characterized by large amplitude and large number of cycles, while Type II ground
motion has short duration and destructive strength.
Importance of Bridge
In the specifications, each bridge is classified either ordinary bridge or important bridge,
according to the road classes, function and structure of the bridge. Table 1 indicates this
classification.
Seismic Performance

The seismic performance of bridge is categorized into the following three levels based on the
seismic behavior of a whole bridge system:
- Seismic performance 1; to secure integrity.
- Seismic performance 2; to limit damage and secure rapid restoration of function.
- Seismic performance 3; to prevent fatal damage.
Table 2 shows the seismic performance objectives, which is prescribed by the combination of
design ground motion and importance of bridge.
The seismic performance is established from the three different standpoints, i.e., safety,
serviceability and reparability, which are further classified into three levels, respectively, as
summarized in Table 3. Safety is the seismic performance to prevent loss of human lives due to
unseating of superstructure. Serviceability represents the performance to maintain the original
traffic function after an earthquake and serve as a route for evacuation, rescuer, first aid, fire
fighting and transporting relief supplies. Reparability represents the performance to be able to
repair the damage caused by an earthquake.
Requirement for Preventing Unseating of Superstructure
An additional requirement is to prevent the unseating of superstructure due to unexpected
seismic behavior of the bridge and ground failure. As Type II ground motion, those recorded in
the 1995 Kobe earthquake that influenced the most destructive effects on structures were
incorporated into the specifications, whereas even greater ground motions could occur. There
still remain large uncertainties to predict such ground motion characteristics and reflect them to
the seismic design of bridges. Furthermore, ground failure and unexpected response of structural
members may cause unpredictable effects to the bridge structure. Even under such circumstances,
it is intended to secure safety against the unseating of superstructure.

DESIGN GROUND MOTION


General
The standard design response spectrum, which is presented later in this paper, may be used as
design ground motion, whereas a site-specific design ground motion shall be developed when the
ground motion at a construction site can be appropriately predicted, based on the information of
past earthquakes, active faults, plate-boundary earthquakes, geological structure, local site
condition, recorded ground motions, and so forth. Reflecting the recent research and
development in the area of ground motion prediction, the site-specific design ground motion
shall be established, if enough information is available to appropriately predict ground motion at
the site. Otherwise, the standard design response spectra may be used as design ground motions.
Standard Design Response Spectra
Level 1 Ground Motion
Level 1 ground motion in terms of acceleration response spectrum S may be determined from
the following equations:

S=cZcDS0

(1)

cD=1.5/(40h+1)+0.5

(2)

where cZ is the zone factor (=1.0, 0.85, 0.7), cD is the modification factor by damping ratio h, and
S0 is the standard acceleration response spectrum. This standard response spectrum was
established from attenuation relations of spectral acceleration, characteristics of past earthquake
damage and ground vibration, and so on. Since ground motion characteristics and resultant
structural damage are closely related the soil condition, the standard acceleration response
spectra are defined for the three different soil classes that are given in Table 4. Table 5 and Figure
1 show the standard acceleration response spectra of Level 1 ground motion.
Level 2 Ground Motion
Similar to the case of Level 1 ground motion, Level 2 motion is also prescribed by
acceleration response spectrum, and Type I and Type II ground motions, which are denoted by SI
and SII, are expressed as
SI=cZcDSI0

(3)

SII=cZcDSII0

(4)

where SI0 and SII0 indicate the standard acceleration response spectra of Type I and Type II
ground motions, respectively. Table 6 and Figure 2 give these standard response spectra. Type I
ground motion stands for ground motions in Tokyo by the 1923 Kanto earthquake, for instance.
The acceleration response spectrum of this ground motion is estimated from attenuation relations
and past experiences. The acceleration response spectrum of Type II ground motion, which
represents ground motion generated by an inland earthquake at short distance, was developed by
smoothing the response spectra that are computed from the ground motions records obtained in
the 1995 Kobe earthquake.

VERIFICATION OF SEISMIC PERFORMANCE


General
In order to verify the seismic performance of bridge, the limit states of individual structural
members shall be first established, based on the limit state of bridge as a whole, which is
described in the following section. Then, the seismic response of individual members shall be
verified not to exceed the corresponding limit state, in which an appropriate method depending
on design ground motion, structural type and limit state should be selected. In the specifications,
standard static and dynamic verification methods are prepared for the bridges whose seismic
behavior is uncomplicated and complicated, respectively. In addition to this, it is necessary to
verify to prevent unseating of superstructure due to unexpected structural response and ground
failure, for which the standard method is also provided in the specifications.

Limit State of Bridge against Seismic Performance 1


The limit state of bridge against seismic performance 1 shall be appropriately determined so
that the seismic response of whole bridge system remains within the elastic range. This limit
state is determined to maintain function of bridge after an earthquake and limit structural damage
to be minor. Corresponding to this limit state of the bridge as a whole, the limit states of
individual structural members may generally be established so that seismic response of
individual members remains within the elastic range.
Limit State of Bridge against Seismic Performance 2
The limit state of bridge against seismic performance 2 shall be appropriately determined so
that the plastic deformation is limited to the structural members that are allowed to be plastic and
secure reparability. This limit state is determined to ensure rapid restoration of bridge function
after an earthquake. As the structural members that are allowed to be plastic, members that can
reliably absorb energy and be presently repaired shall be selected. The designer shall
appropriately combine members that are allowed to be plastic and properly establish the limit
states of individual members. For general bridges, bridge piers may be regarded as the structural
members that can absorb energy and be easily repaired. In case of the bridges equipped with
isolation bearings, the limit states of individual members shall be established so that energy can
reliably be absorbed at the bearings.
Limit State of Bridge against Seismic Performance 3
The limit state of bridge against seismic performance 3 shall be appropriately determined so
that the plastic deformation is limited to the structural members that are allowed to be plastic and
plastic deformation does not exceed the plastic deformation capacity. This limit state is
determined to prevent fatal damage or collapse of bridge. As the structural members that are
allowed to be plastic, members that can reliably absorb energy shall be selected. The designer
shall appropriately combine members that are allowed to be plastic and properly establish the
limit states of individual members.
Verification Method of Seismic Performance
The seismic performance of bridge shall be verified by an appropriate method, based on
design ground motion, structural type and limit state. For the bridges whose seismic behavior is
uncomplicated, the static verification method prescribed in the specifications may be employed
as a verification method of seismic performance. The conventional seismic coefficient method
and ductility design method are applicable as static verification methods. In case of the bridges
that have complicated seismic behavior, the dynamic verification method in the specifications
may be applicable to verify the seismic performance. The bridges that have complicated seismic
behavior generally correspond to the following cases:
1) The principal vibration mode is definitely different from that assumed in the static
verification method of seismic performance.
2) Two or more vibration modes which dominate the seismic response of bridge exist.

3) Plural plastic hinges are assumed to be formed in the verification of seismic performance
against Level 2 ground motion or location of plastic hinges cannot be identified.
4) The applicability of energy-constant rule based on the nonlinear characteristics of
structural member or whole bridge system is not sufficiently confirmed.
Figure 3 illustrates the standard procedure of seismic design.

DEVELOPMENT OF SITE-SPECIFIC DESIGN GROUND MOTION


Ground Motion Prediction Method
As previously mentioned, in the newly revised design specifications for highway bridges, a
site-specific design ground motion shall be developed when the ground motion at a construction
site can be appropriately predicted, based on the information of past earthquakes, active faults,
plate-boundary earthquakes, geological structure, local site condition, recorded ground motions,
and so forth. This provision does not restrict methods of ground motion prediction, and the
recorded ground motion data and various techniques, such as attenuation relations, semiempirical and theoretical ground motion simulation methods may be used for this purpose.
The semi-empirical ground motion synthesizing technique, in which ground motion record
from a small event is used as a Green's function, has been applied to ground motion prediction [6,
7]. This technique has an advantage of automatically incorporating the complicated earthquake
source mechanism and seismic wave path effects into calculation. On the other hand, this method
requires an actual small event record, and it is obvious that an appropriate record is not always
available at a construction site of new structure. To compensate this disadvantage, a stochastic
Green's function technique has been proposed, in which the stochastically simulated small event
motion is used as a Green's function [8, 9]. As an example of site-specific ground motion
prediction, ground motion predicted by a stochastic Green's function technique is presented in
the following section.
Ground Motion Prediction for the Kanto Earthquake
We herein simulate ground motion at Kannonzaki, which is located at the mouth of Tokyo
Bay, from a Kanto earthquake [10, 11]. Kanto earthquake recurrently occurs off the coast of
Tokyo, and the latest one occurred in 1923 caused destructive damage to Tokyo metropolitan
area. We assume the fault plane model proposed for the 1923 Kanto earthquake, and
systematically change the location of hypocenter and asperities to examine the effects of
uncertainties of source parameters on the calculation results. This is because it is hardly possible
to predict the location of these quantities for a future earthquake, which is essentially important
for seismic design of structures. Figures 4 and 5 indicate the fault plane model and the changed
locations of hypocenter and asperities. The length and width of the fault plane are assumed to be
130km and 70km, respectively. The following numerical results correspond to those estimated on
the outcropping layer with shear wave velocity vS=700m/s.
Figure 6 shows the synthesized acceleration time history that has the largest peak
acceleration among the computed results and the corresponding velocity time history. The 5 %
damped acceleration response spectra obtained from all the numerical simulations are plotted in
Figure 7. We see from this figure that the spectral amplitude varies about four times for an

arbitrary natural period due to uncertainty of hypocenter and asperity locations. The thick line in
this figure indicates the spectral level that has a 90 % probability of not being exceeded for each
natural period. This spectral line exceeds 1G over the wide natural period range 0.1<T<2 (s), and
reaches 2G for 0.1<T<0.6 (s). Although further detailed study is necessary, the presented result
seems to be consistent with ground motion characteristics from a large earthquake.

CONCLUDING REMARKS
The Design Specifications for Highway Bridges were revised in March 2002. The major
modification in this revision is the introduction of performance-based design criteria. This paper
presented the basic concepts and principles of seismic design of highway bridges including
design ground motion, seismic performance and limit states. In the revised specifications, a sitespecific ground motion can be conditionally employed as design ground motion. To predict a
site-specific ground motion, we adopted a stochastic Green's function technique. The ground
motion at the mouth of Tokyo Bay was simulated, in which a Kanto earthquake was assumed.
Based on numerical results, effects of source parameter uncertainties on the simulated ground
motions were examined.

ACKNOWLEDGEMENT
The draft of Design Specifications for Highway Bridges, Part V Seismic Design was developed by the Working
Sub-Committee on Seismic Design (Chair: Dr. Shigeki Unjoh) and approved by the Committee on Bridges (Chair:
Dr. Shoichi Saeki) under the auspices of the Japan Road Association.

REFERENCES
[1] Japan Society of Civil Engineers (1996). The 1995 Hyogoken-Nanbu Earthquake Investigation into Damage
to Civil Engineering Structures .
[2] Public Works Research Institute (1997). "Report on the Disaster Caused by the 1995 Hyogoken Nanbu
Earthquake", Journal of Research, Public Works Research Institute, Vol.33.
[3] Yokoyama, K. and Unjoh, S. (1997). "Seismic Design and Retrofit of Highway Bridges in Japan", Second
National Seismic Conference on Bridges and Highways, Sacramento, CA, U.S.A.
[4] Japan Road Association (2000), Design Specifications for Highway Bridges, Part V Seismic Design (English
edition).
[5] Japan Road Association (2002), Design Specifications for Highway Bridges, Part V Seismic Design (in
Japanese).
[6] Hartzell, S. (1978). "Earthquake Aftershocks as Green's Functions", Geophysical Research Letters, Vol.5, No.1.
[7] Irikura, K. (1983). "Semi-Empirical Estimation of Strong Ground Motions during Large Earthquakes", Bulletin
of Disaster Prevention Research Institute, Kyoto University, Vol.33, No.298.
[8] Boore, D. M. (1983). "Stochastic Simulation of High-frequency Ground Motions based on Seismological
Models of the Radiated Spectra", Bulletin of Seismological Society of America, Vol.73, No.6.
[9] Kamae, K., Irikura, K. and Fukuchi, Y. (1991). "Prediction of Strong Ground Motion based on Scaling Law of
Earthquake", Journal of Structural and Construction Engineering, Architectural Institute of Japan, No.430 (in
Japanese).
[10] Yasuda, M. et al. (2000). "Seismic Design Ground Motions for Strait-crossing Projects in Japan", Thirty-second
Joint Meeting, U.S.-Japan Panel on Wind and Seismic Effects, UJNR, Tsukuba, Japan.
[11] Tamura, K. and Kataoka, S. (2000). "A Procedure for Setting up Level-2 Earthquake Motions for Seismic
Design of Bridges", Sixteenth U.S.-Japan Bridge Engineering Workshop, U.S.-Japan Panel on Wind and
Seismic Effects, UJNR, Lake Tahoe, NV, U.S.A.

Bridge class
Class-A bridges
Class-B bridges

TABLE 1. IMPORTANCE CLASSIFICATION


Bridges included
- Bridges other than Class-B bridges
- Bridges on national expressways, urban expressways, designated city expressways, HonshuShikoku bridge highway and general national highways
- Double-section and overpass bridges on prefectural highways and municipal roads, and other
bridges and viaducts that are important in view of regional disaster prevention plans, traffic
flow volume, etc.

TABLE 2. DESIGN GROUND MOTION AND SEISMIC PERFORMANCE


Class-A bridges
Class-B bridges
(Ordinary bridges)
(Important bridges)
Level 1 ground motion
Secure integrity
Level 2
Type I ground motion
Prevent fatal damage
Limit damage and secure
ground motion Type II ground motion
rapid restoration of function

Seismic performance
criteria
Seismic
performance 1
Seismic
performance 2
Seismic
performance 3

TABLE 3. SEISMIC PERFORMANCE CRITERIA


Reparability
Serviceability
Short-term
Secure safety against Secure preNeed no repair for
collapse
earthquake function
restoration of
function
Secure safety against Secure rapid
Emergency repair
collapse
restoration of
enables restoration
function
of function
Secure safety against

collapse
Safety

Long-term
Need minor repair
Possible to perform
permanent repair
easily

TABLE 4. SOIL CONDITION CLASSIFICATION


Soil classification Natural period (s) Geological description
Group-1
TG<0.2
Rock or shallow soil deposits
Group-2
0.2TG<0.6
Diluvium or alluvium
Group-3
0.6TG
Soft alluvium

Soil classification
Group-1
Group-2
Group-3

TABLE 5. LEVEL 1 STANDARD DESIGN RESPONSE SPECTRA


Spectral acceleration S0 (cm/s2) at natural period T (s)
S0=431T 1/3 for T<0.1
S0=200 for 0.1 T 1.1
S0=220/T for 1.1< T
(S0160)
S0=427T 1/3 for T<0.2
S0=250 for 0.2 T 1.3
S0=325/T for 1.3< T
(S0200)
S0=430T 1/3 for T<0.34
S0=300 for 0.34 T 1.5
S0=450/T for 1.5< T
(S0240)

TABLE 6(1). LEVEL 2 STANDARD DESIGN RESPONSE SPECTRA


(a) Type I ground motion
Soil classification Spectral acceleration SI0 (cm/s2) at natural period T (s)
Group-1
SI0=700 for T 1.4
SI0=980/T for 1.4< T
Group-2
Group-3

SI0=1505T 1/3 for T<0.18


(SI0700)
SI0=1511T 1/3 for T<0.29
(SI0700)

SI0=850 for 0.18 T 1.6

SI0=1360/T for 1.6< T

SI0=1000 for 0.29 T 2.0

SI0=2000/T for 2.0< T

TABLE 6(2). LEVEL 2 STANDARD DESIGN RESPONSE SPECTRA


(b) Type II ground motion
Soil classification Spectral acceleration SII0 (cm/s2) at natural period T (s)
Group-1
SII0=4463T 2/3 for T<0.3
SII0=2000 for 0.3 T 0.7
SII0=1104T -5/3 for 0.7< T
Group-2

SII0=3224T 2/3 for T<0.4

SII0=1750 for 0.4 T 1.2

SII0=2371T -5/3 for 1.2< T

Group-3

SII0=2381T 2/3 for T<0.5

SII0=1500 for 0.5 T 1.5

SII0=2948T -5/3 for 1.5< T

Response Spectral Acceleration(cm/s )

500
300
200

100

h=0.05
Group1
Group2
Group3

50
30
20
0.1

0.2

0.3

0.5

Natural Period(s)

3000

3000

2000

2000

Response Spectral Acceleration(cm/s )

Response Spectral Acceleration(cm/s )

Figure 1. Standard acceleration response spectra of Level 1 ground motion

1000

500
300

h=0.05
Group1
Group2
Group3

200

100
0.1

0.2

0.3

0.5

Natural Period(s)

1000

500
300

h=0.05
Group1
Group2
Group3

200

100
0.1

0.2

0.3

0.5

Natural Period(s)

(b) Type II ground motion


(a) Type I ground motion
Figure 2. Standard acceleration response spectra of Level 2 ground motion

Start

No

Is seismic response to Level 1


ground motion complicated?

Determine lateral seismic force


coefficient and inertia force

Yes

Compute response values by


dynamic analysis

Determine allowable values


(Allowable stress, etc.)

Compute sectional forces and


displacements by static analysis

Verification of seismic performance for Level 1 ground motion

No

Is seismic response to Level 2


ground motion complicated?

Determine lateral seismic force


coefficient and inertia force

Yes

Compute response values by


dynamic analysis

Determine allowable values


(Horizontal capacity,
allowable displacement, etc.)

Compute sectional forces and


displacements by static analysis

Verification of seismic performance for Level 2 ground motion


Design of unseating prevention system
End
Figure 3. Standard procedure of seismic design

Figures 4. Fault plane model of Kanto earthquake

Asperity Kannonzaki Hypocenter

Figure 5. Locations of hypocenter and asperities

Figure 6. Simulated ground motion

Figure 7. Acceleration response spectra

You might also like