Anda di halaman 1dari 11

Seismic performance of transportation system during 1995 Kobe earthquake

Abstract
Mobility has always been a major concern to human society. Right from Indus Valley
civilization to Egyptians and Romans all built roads valuing efficient mobility for communication and
trade, which are pillars of the functioning of societies. In a modern society need of efficient and safe
transportation has grown limitless, directly correlating to socio economic progress and development.
Problem of significant economic disruption has clearly surfaced during the recent earthquake disasters
causing major damages to transportation networks. Transportation system is a conglomerate of various
modes like highways, railways, tunnels, interchange facilities, bridges, airways etc. While this suggests
the need to evaluate total system performance in transportation risk assessment, in addition to
examining the vulnerability of individual components such as bridges, no appropriate measures
currently exist.
The paper discuses multidimensional approach of system evaluation, first at individual level,
for example structural performance of bridges, roads etc; and then evaluation of the entire network is
done in terms of its functional performance. Indices of transportation system performance are proposed
to evaluate network coverage and accessibility. Post-disaster system performance measures are applied
to the transportation system devastated by the 1995 Hyogokem Nanbu earthquake in Kobe. Kobe
earthquake provides the precious opportunity to observe and learn the urban earthquake vulnerability.
Introduction
Evaluating transportation system performance can greatly facilitate in understanding the
effects of historic disasters and preparing for future hazard events. Many events in the last decade
including earthquakes, hurricanes, tornadoes, floods, terrorist attacks, fires, power blackouts etc
have increased our awareness of how our critical infrastructure systems such as transportation,
water, power, fuel, communication, and so on affect our well-being. Expert believe that one of the
major reason of rescue failure in Kobe earthquake was unclear roads because of which emergency
equipment and firefighters could not make their way to the city.
Any infrastructure system is a conglomerate of engineered elements; hence any constructed
system will function as an engineered element of a parent infrastructure system. Hence
performance of the system will depend on interactions and interdependencies among engineered,
natural, and human systems that are components of the same infrastructure system. Hence a
multidimensional approach to define performance of infrastructure is inevitable.

Sr.
no.
Performance/ Health
Category
Issues under the category Type of analysis
1 Operational & utility limit
states
Network coverage /
Accessibility
Total Infrastructure system
(complete network)
2 Engineering/Structural
Limit State
Structural health /
Serviceability and durability
Individual system (bridge,
port etc)
Table1: Performance criterion for Infrastructure system
Health of a system can be defined as the ability of a system to perform as promised, so it
should be seen as a mean to evaluate and assure its future performance. As discussed seismic
performance of transportation system can be evaluated by broadly categorizing it in terms of
structural performance and functional performance, which can again be segregated as performance
of individual components or performance of transportation system as whole. Economic loss can be
better evaluated if we can measure the total system performance in transportation risk assessment,
in addition to examining the vulnerability of individual components. But unfortunately due to the
simplicity much attention has been given to understanding and predicting the performance of
individual system.

Structural Performance of bridges
Dynamic Response of bridges during earthquake:
Bridges are unique in their structural response. They are longitudinally lengthy, and consist
of many structural components which contribute to the overall resistance capability of the system.

Various points that can be noted in bridges are
1) Decks of bridges are often skewed and curved, and intermediate expansion joints divide a
bridge system into several structural segments with different natural periods.
2) There are various structural types with complex geometries and dynamic response
characteristics.
3) Suspension bridges and cable stayed bridges generally display a very complex structural
response with long natural periods, often exceeding 10 seconds. Many modes with closely
spaced natural periods contribute to the complexity of the structural response.
4) Bridges are generally constructed at soft soil sites such as rivers and bay areas. Because
ground motions are amplified at these sites, greater attention should be paid to seismic
design for large ground motion. Failure of foundations associated with the instability of
surrounding ground is a common occurrence.
5) The degree of statically indeterminacy is smaller in bridges than in buildings, and therefore
ductility of piers or columns needs to be carefully examined to prevent failure during
strong earthquakes.

For dynamic response analysis, the stiffness, mass and damping properties of each element
must be realistically defined. Since a typical deck is extremely stiff and strong in comparison with
its supporting columns and abutments, the high amplitude bridge response produced during severe
ground shaking will be caused primarily by deformations in the columns, abutments and expansion
joints. The deck will remain elastic and, therefore, can be modeled by linear elastic elements.
Nonlinear elements must, however, be used for columns, abutments and expansion joints.

Foundations
A simple model can be defined consisting of three translational and three rotational soil
springs to connect the base of each column and abutment to a rigid foundation where the seismic
excitation is fully prescribed, as shown in fig.
For smaller amplitudes linear analysis may be adopted where the stiffness of these soil springs may
be evaluated using linear elastic half-space theory. But for the large amplitude response, the
foundation soils may undergo inelastic deformation of the hysteretic type. In this case, the six soil
springs should be nonlinear hysteretic springs. Their stiffness can only be established through
extensive experimental studies on the dynamic properties of foundation soils.
Since J apan is located in the monsoon area, soft and unstable soils sediment. Major cities
are located on such thick sedimentation. Since bridges are constructed at those sites, failure of
unstable soils as well as scouring always resulted in damage of foundations.



Pre -Kobe earthquakes
Importance of taking account of the seismic effects in design of engineering structures was
first recognized when the destructive damage occurred in 1923 Kanto Earthquake. Seismic
Coefficient Method was fist incorporated in design of Highway Bridge in 1927. The method
considered equivalent static lateral force using a seismic coefficient of 0.1-0.3 based on the
allowable stress design approach.
In the 1964 Design Specifications (J RA 1964) also important requirements such as realistic
near-field ground motions, ductility, dynamic response, liquefaction and unseating prevention
devices were not included. The only seismic requirement was seismic coefficients of 0.2 in
horizontal and 0.1 in vertical directions. As a result most bridges which suffered damage in the
1995 Kobe Earthquake were designed and constructed in accordance with the 1964 Design
Specifications.
It was only after Niigata Earthquake, phenomenon of brewing of sand in water during
earthquake was termed liquefaction and scientific researches started worldwide. But the
mechanism of liquefaction was still unclear, lateral spreading was still confused with slippage of
soft clayed soils and so attention was limited only to soil liquefaction, not to the lateral spreading.
It was 1990s when the lateral spreading was first recognized.
In the 1971 Guide specifications (J RA 1971) no clear design procedure for liquefaction was
included, but an assessment of liquefaction potential based on N-value of standard penetration test
depending on depth was first incorporated.
In the Design Specifications (1980 J RA) a procedure for decreasing the stiffness of soil
springs which connect soils and foundations depending on FL-value was first included. Where FL-
value was defined as a ratio of lateral force and soil strength, which gave a better assessment of
soil liquefaction.
As discussed earlier rigid and large concrete sections were used due to the thick soft soil
and scoring problem, the same design practice was also used in reinforced concrete piers. This was
the reason why even though the tie reinforcement was insufficient there were limited damages in
the past earthquakes. Although reinforced concrete columns with smaller sections as well as steel
columns started to be constructed from 1970s for urban highway viaducts, the design practice was
practically unchanged which resulted in the extensive damage in the 1995 Kobe Earthquake.

It was only the 1990 Design Specifications (J RA 1990), ductility studies were incorporated.
In addition to the countermeasures to soil liquefaction, restrainers, or unseating prevention devices
in a broader sense, were first developed and implemented in bridges after the Niigata Earthquake.
Extensive damage of bridges that resulted from excessive relative displacements between
superstructures and substructures created an inspiration for developing unseating prevention
devices.
1995 Kobe Earthquake
1995 Kobe Earthquake had destructive damage to bridges. Reinforced concrete columns
suffered failure in shear. Premature shear failure at terminations of longitudinal bars with
insufficient development lengths resulted in collapse of many bridges. Extensive soil liquefaction
occurred, and this resulted in settlements and tilting of foundations and substructures. Lateral
spreading of ground associated with soil liquefaction caused movements of foundations.

Premature shear failure of reinforced concrete columns
Eighteen spans of Fukae Viaduct, Hanshin Expressway collapsed as shown in figure. The
viaduct was designed in accordance with the 1964 Design Specifications using 0.2 horizontal and
0.1 vertical seismic coefficients based on the allowable stress design approach. It was completed in
1969.

Figure: Collapse of Fukae Viaduct on Hanshin Expressway Figure: Uplift of Deck, Higashi-Kobe
Bridge
Major problems in the design were:
1) Overestimated allowable shear stress. The allowable shear stress that is required in the
current codes is less than 60 % of the value used in design of this viaduct.
2) Insufficient development length of longitudinal bars terminated at mid-height.
3) Insufficient amount of tie bars.

Subjecting to a strong ground motion, the columns suffered extensive flexural and diagonal cracks
at 2.5m above the footing. Since the amount of tie bars was insufficient, this caused premature
shear failure in the columns. This was the major reason for causing the destructive damage of
bridges in the Kobe Earthquake.

Structural Failure in Pavements
Asphalt concrete: The design of an asphalt concrete road typically doesnt include any factors to
account for earthquake conditions. The reason is that usually the surface course, base, and subbase
are in a compact state and are not affected by the ground shaking. In addition, the cumulative
impact and vibration effect of cars and trucks tend to have greater impact than the shaking due to
earthquakes.
Concrete pavement: Concrete pavement and concrete median barriers are often damaged at their
joints, or they are literally buckled upward. This damage frequently develops because the concrete
sections are so rigid and there are insufficient joint opening to allow for lateral movement during
the earthquake.
Common causes of damage to roads are as follows: (a) Surface Rupture (b) Slope instability (c)
Liquefaction flow slides or lateral spreading (d) Settlement of soft soils. (e) Collapse of underlying
structures.
Failure of pavements due to Surface Rupture:
Seismic study maps are developed delineating the approximate location of active fault
zones. Various local authorities have written codes for construction near these areas. Most of the
structures would not be able to resist the surface rupture shear movement it is recommended not to
do any construction in the area.
Since transportation routes are most of the time inevitable the following recommendations
can be used:
1. Constructing the road such that they cross the fault in the perpendicular direction.
2. Cross the surface rupture zone at the level ground location, so that it doesnt warrant need
of any interchange or bridge.
3. Use asphalt concrete in the region, which are flexible in nature and easy to repair.

Failure of pavements due to Liquefaction:
One of the major problems in geotechnical earthquake engineering is the phenomenon of
liquefaction of loose to medium dense sands. Liquefaction can be defined as the reduction of the
soil shear strength when subjected to undrained that is constant volume loading, whether that
loading is monotonic, cyclic, or dynamic. As expected from the definition liquefaction will be a
typical phenomenon during earthquakes since during shaking, the sand tends to compact. The
water in the pores cannot escape quickly enough, at least in the finer sands, to accommodate
instantaneously the compaction. This increases the pore pressure, reducing the effective
intergranular stresses.

Problem during Kobe Earthquake:
Since Kobe is near to water body and is on lose and filled sediment, it is more susceptible
to liquefaction. Liquefaction could be a serious problem for ports especially due to their location
disadvantage and since they carry very high load due to the stacked containers.
Remedial Measures:
In case of important pavement construction design for liquefaction should be incorporated.
Liquefaction potential under earthquake loading should be evaluated.
Simplified methods of evaluating liquefaction potential under earthquake loading have been
presented by Seed & Idriss (1971), Ishihara (1977), Iwasaki et al. (1984), Seed et al. (1983, 1984)
& Robertson and Campanella (1985). The first three approaches employ data from laboratory tests
to assess the stress conditions necessary to cause liquefaction, whereas the latter three employ data
from in-situ penetration tests. The liquefaction risk evaluation procedure is based on SPT N-
values.
The methods involve three main steps:
1. Estimation of the cyclic shear stress induced at various depths within the soil by the
earthquake, and the number of significant stress cycles.
,
max
,
max
max
) (
vo
d vo
vo
r a
CSR
o
o
o
t
= =
MSF CSR CSR
5 . 7
) ( ) ( =

) ( 65 . 0 ) (
,
max
max
MSF
r a
CSR
vo
d vo
o
o
=
2. Estimation of the cyclic shear strength of the soil, i.e. the cyclic shear stress ratio which is
required to cause initial liquefaction of the soil in the specified number of cycles.
200
1
] 45 ) ( 10 [
50
135
) (
) ( 34
1
) (
2
60 1
60 1
60 1
5 . 7

+
+ +

=
N
N
N
CRR
3. Comparison between the induced cyclic shear stress and the cyclic shear strength; at
locations where the induced shear stress exceeds the shear stress required to cause initial
liquefaction, there is a potential for liquefaction.
FS =(CRR7.5/CSR)MSF
From the above three steps design curves similar to that shown in the figure can be obtained. For a
selected range of PGA value a curve can be recommended for the minimum required density of the
filled material.
Figure: Kobe port after 1994 earthquake. White sand
after soil boil can be seen.
Figure: Design chart for considering liquefaction
potential of soil in design
Deep soil improvement and densification are recommended to minimize liquefaction
settlements and pavement failure. A minimum N-value can be specified for any given depth.
Various soil improvement methods such as Dynamic compaction, vibroflotation, sand piles can be
recommended for deep soil improvement.
Figure: Localized Compression failure (Northridege
earthquake 1994)
Figure: Damage caused due to underlying collapse of the
Daikai subway station. Kobe earthquake 1994
Functional performance
The operational or functional performance can be evaluated in terms of network coverage
and transport accessibility, traffic delay and loss modeling etc. The restoration of highway system
performance can be correlated closely with the recovery of highway traffic volumes.

Introduction
Due to the complex large network system and their high demand and utility urban
transportation systems are vulnerable to earthquake disasters.
System performance measures are important for several reasons such as.
1. They enable comparisons of system conditions across disaster events in different urban
areas.
2. They allow comparisons across scenario disaster events for a single study region.
3. System performance measures can be used in designing efficient post-disaster restoration
strategies by prioritizing damage repair such that overall system performance can be
optimized. This aids mitigation prioritization under budget constraints.
4. System performance measures can be implemented for estimating economic impacts in the
context of real-time earthquake loss models for emergency response and recovery planning.
5. System performance measures have proven useful in evaluating the seismic performance of
other lifeline infrastructure systems.

There are many traditional transportation quality and accessibility indicators but its very
difficult to implement them in the post earthquake cases because they require data that are difficult
to obtain after a disaster. Some of the traditional methods are total travel time on the network in
vehicle-hours, or the sum of vehicle count over all system links multiplied by travel time on each
link. This warrants the need of new measures or indices which should use commonly available
data and easy in application.
Three system performance measures are proposed here1:
1. Total length of highway open, measure parameter L;
2. Total distance-based accessibility, measure parameter D;
3. Areal distance-based accessibility, measure parameter Ds.
Each is estimated as the ratio of post-earthquake to pre-earthquake conditions and ranges
from 0 (system non-functional) to 1 (system fully functional). The first two proposed measures
pertain to the overall performance of the system. In contrast, the last measure is specific to
individual subareas within the study region, such as neighborhoods, and can indicate spatial
disparities in transportation performance. The measures are specific to time t after the earthquake.
Network has been defined as a series of nodes and the links connecting them. For the
highway mode, nodes represent on- and off-ramps and junctions between highways. For the
railway mode, nodes are stations.

Measure parameter L(t)
x
t x
t L
) (
) ( =
Where, x is pre-earthquake open length
x(t) is the post earthquake open length at any time t
1 ) ( 0 s s t L
0 value signifies that the system is system non-functional and 1 is that the system is fully
functional.
Measure parameter D(t) and A(t)
The measure parameter D(t) considers the minimum network travel distances and hence
tries to measure changes in accessibility at all nodes on the network. Since the users travelling
between two nodes tends to use the minimum cost path, the change in the minimum cost path will
depends on the extend as well as the location of damage.
1
) (
) (

=
f
t A f
t D



=
j
j i
i
j
j i
i
t d
t d
t A
) (
) (
) (

f A s s 1
Where,
ij
d is the minimum travel distance between nodes i and j on damaged network
ij
d is the minimum travel distance between nodes i and j on undamaged network
f =effective distance multiplier
A =total network accessibility ratio

Similar to L(t) 0 value of D(t) represent no network accessibility and 1 means network
accessibility is intact.

Measure parameter Ds
Ds shows the accessibility of subareas. Subareas considered for Kobe earthquake case
study are the city wards.
1
) (
) (

=
f
t A f
t D
s
s

=
=
s
N i
i
s
s
t A
n
t A ) (
1
) (
,

=
=
=
s
s
i j
ij ij
i j
ij ij
i
t d w
t d w
t A
) (
) (
) (

=
p
sp
sr
r r
ij
v
v
n
w
o
1

s
N i = ,
r
N j =

1 =

j
ij
w
Where, Ds(t) =accessibility performance measure for area s at time t
As(t) =transport accessibility ratio for area s at time t
Ai(t) =transport accessibility ratio for node i at time t
ns =number of nodes in area s
Ns =set of nodes in area s
dij(t) =minimum distance on damaged network from node i to node j at time t
) (t d
ij
=minimum distance on intact network from node i to node j
wij =destination weight for node j for commuters originating from node i
Traffic ratio T can be calculated by normalizing the total traffic volume with the pre-earthquake
average traffic volume.



Issues arise in the application of the preceding system performance measures
1. The definition of the study network is fundamental yet difficult to maintain in a consistent
manner across events and urban areas, the definition of nodes and links.
2. Issues include the geographic boundaries of the study area, road classes and routes to be
included in the study network.
3. The specification of damage (functionality) states, including the definition of damage
states.

Aggregate System Performance
Performance of Highway system
Highways that were severely damaged were Hanshin Expressway, Meishin National
Expressway and Chugoku National Expressway. The most significant damage occurred to Hanshin
Expressway Route 3.

Fig. Study Highway network for study Fig. Railway network for study

Route 3 Characteristics: (a) It carries 40% east west corridor traffic, (b) 252,800 average daily
traffic (ADT), (c) Connects Osaka & Kobe, (d) approx 28 km closed was closed after earthquake,
(e) It was completely opened after 20 months.
Route 5 Characteristics: (a) It runs at Ashiya River screen line, (b) 28,300 average daily traffic
(ADT), (c) Nishinomiya-ko Bridge was collapsed, (d) Alternate to route 3 together with route 7 &
16, (e) No major physical damage.
Meishin National Expressway: (a)Toyonaka Interchange damaged,(b)Nishinomiya Interchange
damaged (c) 70,000 average daily traffic (ADT) (d) direct connection to route 3 was lost.
Chugoku National Expressway: (a) Toyonaka Interchange damaged (b) Nishinomiya Interchange
damaged (c) 98,700 average daily traffic (ADT) (d) Alternate to Hanshin Expressway (e) 4 of 6
lane open in 1month.
The case study focus on Hanshin Expressway Routes 3, 5, 7 and 16, Daini Shinmei
Expressway from Tsukimiyama to Akashi-Nishi IC, Chugoku National Expressway from Yokawa
J ct. to Suita J ct., and Meishin National Expressway from Suita J ct. to Nishinomiya IC.
To perform the study, data required on highway status like length, pre and post earthquake
monthly ADT, and status of reopening of damaged sections during the reconstruction period were
obtained from the concerned authorities. Total network consisted of 63 nodes and 62 links. Each
measure was evaluated on a monthly basis.
Values of L, D, & T have been measured on monthly basis using the above equations. The
restoration plot is as shown in the figure. Performance dropped to 0.14 from 1 immediately after
the earthquake.
Fig. System performance and traffic restoration of highway
network.
Fig. Highway accessibility by Kobe City Ward,
J uly 1995.

Performance of Rail system


Similarly performance of the railway system was gauged. The system contains 185 stations
and 202 links. Since railway lines are much more redundant than the highway system the recovery
was rapid. Immediately after earthquake L value for rail system dropped to 0.29 and D value 0.22
indicating only 30% functional rail length, but by the end of 1 month system recovered to 0.80
Areal performance
For areal investigation study regions used were cities of Kobe, Akashi, Ashiya,
Nishinomiya, Amagasaki, Itami, Sanda, Takarazuka, and Kawanishi in Hyogo Prefecture and
Osaka, Toyonaka, Ikeda, and Suita in Osaka Prefecture. The study area accounts for some 96% of
work and school destinations for Kobe City commuters
Accessibility Ds was evaluated for each of Kobes city wards for highway and rail transportation
over the post-disaster restoration period, respectively.
Fig. Highway performance restoration, selected Kobe
City Wards.
Fig. Rail performance restoration, selected Kobe City
Wards.


Conclusion:
Seismic performance of transportation system is interlinked to the performance of the other
lifeline infrastructure systems. Poor performance of transportation system may poof in failure of
the entire rescue operations and considerable economic loss. Performance of the transportation
system under individual level that is structural damage of bridges, roads, geotechnical performance
has thrown light over various short comings in the codes. 1995 Kobe earthquake case study
suggest that the major reason of bridge failure was due to premature shear failure of reinforced
concrete columns because of the practice of using thicker concrete section with insufficient tie bars
and development length. Ignorance of liquefaction consideration in design procedures was also
viewed as the major cause of bride and pavement failure. A design procedure for designing of
pavements considering the liquefaction effect has been suggested.
Functional performance of the system at the network level using the simpler performance
indices have shown a complete picture of damage caused by the earthquake and the performance
restoration. The performance restoration was quick in rail network compare to the highway
network.

References:

1. Kazuhiko Kawashima, Department of Civil Engineering Tokyo Institute of Technology, Seismic
Design of Bridges After 1995 Kobe Earthquake.
2. Kazuhiko Kawashima, Tokyo Institute of Technology, Seismic Design of Urban
Infrastructures.
3. Offshore Geohazard University, Offshore geohazards, Evaluation of Liquefaction
Potential.
4. Greenstein J. and Poran J. C., Rigid Pavement Design for Ports in Chili, Transportation
Research Record 1388.
5. Stephanie E. Chang and Nobuoto Nojima, Measuring Post-Disaster Transportation System
Performance.
6. PacificEarthquakeEngineeringResearchCenterPEER,
http://peer.berkeley.edu/research/research_goals.html.
7. Chakroborty,P.andDas.A.,PrinciplesofTransportationEngineering,PRENTICEHALL
8. MayA.D.,Trafficflowfundamentals,PrinceHall.
9. RobertW.D.,GeotechnicalEarthquakeEngineeringHandbook,McGrawHill.
10. Shanjiang Zhu and David Levinson (2007) An agent-based Route Choice Model.

Anda mungkin juga menyukai