Você está na página 1de 8

Preliminary Estimation of Seismic Vulnerability of a Database of Bridges

Located at the Mexican Pacific Coast


Consuelo Gmez-Sobern
Professor
U. Autnoma Metropolitana
Mxico DF, Mxico

Luis A Barrera Bautista


Civil Engineer
U. Autnoma Metropolitana
Mxico DF, Mxico

cgomez@correo.azc.uam.mx
Consuelo Gmez Sobern, received
her Civil Engineering Degree from
the BUAP university (Mxico) and
her Master and Doctorate Degree
from the UNAM (Mxico) and
UPC (Spain) universities.

Luis A Barrera Bautista received


his Civil Engineering Degree
from the Universidad Autnoma
Metropolitana in 2006.

Summary
The Mexican Secretary for Communications and Transports (SCT) periodically applies an
evaluation procedure to define maintenance politics of highway bridges, named SIPUMEX. This is
a very general method which does not consider some important parameters needed to characterize
the seismic behavior in bridges. A modification of the SIPUMEX process is presented in this paper,
considering the addition of new evaluation parameters, based on the results of surveys made with
Mexican experts in this theme, or by eliminating some parameters used in the original methodology
that do not provide information about the bridges seismic vulnerability. The proposed procedure
was applied to a database of 79 highway bridges located next to the Mexican Pacific Coast. The
obtained results were compared with the ones calculated with a procedure available in the literature.
Keywords: Bridge seismic behavior; seismic vulnerability; maintenance procedure; RC bridges;
preliminary evaluations.

1. Introduction
Bridges are essential facilities in the highway systems, thus its correct and continued behavior is an
important aspect, specially after a mayor earthquake. The bridge collapse or its bad performance
could cause vehicular jamming, interruption of diverse economical activities, human lives decease,
or important economic lost. Because of these aspects, the maintenance procedures of bridges are a
topic considered in many researches in the last years.
The necessity of adequate maintenance and rehabilitation procedures was evident after the
Northridge (1994) and Kobe (1995) earthquakes. Due to the action of these earthquakes, between
3% and 5% of the affected bridges suffer some degree of damage. In addition, the partial and total
collapse of some structures produce direct and indirect millionaire economic lost [1, 2, 3, and 4].
Some of the failures reported after these earthquakes were due to seismic resistant erroneous
configurations, deficiencies of design codes and structures with reduced capacity. With periodic and
reliable maintenance procedures, these possible failures can be detected before earthquakes and
strategies to reduce damage and economic lost could be defined.
The maintenance or rehabilitation procedures of bridges consider three principal aspects: (1)
preliminary evaluations, (2) detailed evaluations, and (3) definition of the actions to be realized.
The objective of the preliminary evaluation is to detect, between a numerous group of systems, the
structures with important reduction of its capacity. In the detailed or secondary evaluation,
meticulous analyses are realized to define the degradation degree of the elements characterized in
the preliminary evaluation. With the information defined in the above steps, decision tools are
available to propose specific maintenance, rehabilitation or reposition procedures.

There are various preliminary evaluation procedures available in the literature to define the seismic
vulnerability of bridges [5]. In general, these methods are based on the assignation of the seismic
susceptibility of different parameters and in their combination with a predefined model. In this form,
a global vulnerability index is assigned to each bridge in order to classify all the elements and to
detect the structures with important degradation. In this section, only two of these procedures are
described, one available in literature and the other used by the Secretary for Communications and
Transports (SCT) of Mxico.
Table 1 Parameters of the method of Kim

Parameter
Peak ground
acceleration
Design
specifications
Type of
superstructure

Shape of the
superstructure
Internal
hinges
Type of pier

Type of
foundation
Material of
the subsest.
Structural
irregularity

Soil condition
Liquefaction
Seat length

Category
1: PGA <0.1g;
2: 0.1g < PGA
< 0.2g;
3: 0.2g < PGA < 0.3g; 4: PGA > 0.3g
1: after 1981; 2: 1972-1980; 3: 19401971; 4: before 1940
1: cable-stayed, suspension, single span;
2: arc, monolithic girder, trusses;
3: continuous girder, trusses;
4: simple-supported girder and trusses,
multiple spans, elevated structures
1: straight;
2: skewed 20 - 45 or curved 40 - 90;
3: skewed 45 - 60 or curved 90 - 180;
4: > skewed 60 or curved 180
1: none; 2: yes, with cable restrainers or
seat length 12 ;
3: yes, with 6 < seat length < 12;
4: yes, with seat length < 6
1: monolithic multi-pier bent or solid;
2: pinned multi-pier bent;
3: monolithic single pier;
4: pinned single pier
1: single pier shaft; 2: spread footing;
3: piled footing; 4: pile bent
1: steel; 2: ductile concrete; 3: no-ductile
concrete; 4: timber, masonry, other old
materials
1: none; 2: any heights of 2 piers
1.25 times; 3: any adjacent heights of 2
piers more than 1.25 times; 4: any
adjacent pier heights more than 1.5
times
1, 2, 3 o 4 for different soil types
1: LSI* < 5; 2: 5 < LSI < 25;
3: 25 < LSI < 100; 4: LSI > 100
1: good; 2: fair; 3: poor; 4: very poor

W
0.141
0.456

0.114

0.437

0.089

0.029
0.024
0.034

0.278

0.188
0.932
0.512

*LSI = Youd and Perkins factor that characterize the effect of liquefaction
[6]

N d = 8 + 0.02 L + 0.08 L p
N d = 12 + 0.03 L + 0.12 L p
S p = Nd

PGA / 0.19

PGA 0.19 g
PGA > 0.19 g

1.1 Method of Kim


The Seong Hoon Kim method [6] is a
statistical procedure of simplified
evaluation
for
bridges
seismic
vulnerability. This method evaluates
twelve parameters to define the
vulnerability index, one of theme is the
seismic hazard of a bridge site location.
For each of the analyses parameters,
Kim method uses four seismic
susceptibility categories. In Table 1, the
parameters, categories and importance
weights of each parameter of the Kim
model are described. The proposed
weights were defined as a subjective
form, in function of the influence of the
parameter in the bridge seismic
response and due to calibration process
with damaged structures. To evaluate
the parameter Internal hinge, Kim
method define the hinge number of the
bridge using
N art = N span n p 1

(1)

where Nspan is the number of spans and


np is the number of piers. Also for the
same parameter and the Seat length
parameter, in the case where the
minimum support length is unknown,
Kim proposes to use Equation 2, where
Nd is the support length of reference, L
is the superstructure length between
internal hinges, Lp is the average high
of piers between internal hinges, PGA is
the peak ground acceleration at the
bridge site, and Sp is the minimum
support length. In Equation 2, the
variables are in inches.

(2)

The parameter Superstructure form is evaluated in the Kim method using the skewed, s, and
curvature, cur, angles. The variables are defined as is described in Equation 3, where i and p
are the initial and final skewed angles of the bridge, and CR is the curvature radio.

s = i f

(3)

cur = Lp / CR
Starting from the parameters of Table 1, Kim method proposes the two vulnerability index of
Equation 4. For these indexes, C = 3.84 is a constant defined for statistical calibrations of reports
of damaged bridges
(4)

12

IVK 1 = Yi Wi + C
i =1

IVK 2 = Y1 *

12

Y Wi

i =1 i

Table 2 Degradation degree of the method of Kim

IVK1 Damage

IVK2

Damage degree

0
1

Without damage
Minor damage

< 1.5
1.5<IV<2.5

Minor damage

2
3

Moderate damage
Severe Damage

Collapse

Moderate damage
> 2.5

where IVK1 and IVK2 are the linear


and non linear vulnerability index of
Kim, Yi and Wi are the categories and
weights assigned to parameter i,
conform the values of Table 1. The
obtained values of the vulnerability
index are classified as show in Table 2.

Mayor damage

1.2 SIPUMEX method


The highway bridges evaluation in Mexico is conducted by the SCT, in free toll system.
Specifically, this evaluation is developed by the General Direction of Highway Conservation
(DGCC, for its name in Spanish). The DGCC use two evaluation methods, one for bridges and the
other for the highway system [7]. The evaluation method used in bridge structures, named
SIPUMEX, is based on two techniques, one principal and simple, and another secondary o special.
One of the final objectives of the SIPUMEX procedure is to analyses the results of the two
evaluation techniques to design, in function of available resources, the strategies of maintenance
and rehabilitation of the damaged elements.
Almost every two years, the personnel of the SCT realize the evaluation of the studied parameters.
After the inspection, the information of each structure is actualizing to detect degradation before a
partial o total failure is presented. By this form, the probability of failure in normal operations is
intended to be reduced. However, the bridge situation is not analysed for hydro meteorological or
seismological conditions. The first step of the method applied by the SCT is a subjective evaluation
of each bridge of the system, based on the quantification of different parameters. With these
periodically visual inspections, the SCT creates a database of more than 6 800 bridges, with
geometric and structural basic characteristics, location of the structure, physical actual condition
and traffic over the bridge. Using the qualification of the considered parameters, a global index
between cero and five is assigned to the structure, as it is described in Table 3.

Table 3 Qualification of the SCT procedure


IVSCT
0
1
2
3
4
5

Qualitative qualification
New or recent repaired bridge. Without problems
Bridge in well condition. Do not require attention
Bridge with minor problems, undefined period of
attention
Important damage, necessary reparation in a period of
3 to 5 years
Important damage, necessary rehabilitation in a period
of 1 to 2 years
Extreme damage or possible total collapse,
immediately rehabilitation

substructure elements or hydraulic capacity.

The SIPUMEX method considered next


parameters: bridge surface, expansion
hinges, parapets, embankments, piers,
abutments, bearings, superstructure, girders,
caudal and general bridge. Every of these
parameters
are
evaluated
with
a
qualification between cero and five using a
visual estimation. For the evaluation are
considered aspects like: average diary
traffic, percent of cracking, crushing of
concrete material, reinforced steel corrosion
and oxidation, verticality loss of the

2. Modified procedure
However
the
SIPUMEX
methodology has the objective to
W
prevent degradation of the general
Year of construction
1: >1993; 2: 1985-1993; 3: 1977condition of bridges for normal
1985;
0.1154
operation, the procedure does not
4: 1959-1977; 5: 1959 <
evaluate the specific vulnerability of
Substructure
1: none;
seismic or hydraulic phenomenon.
irregularity
3: piers height 1.5 a 2 times;
Precise evaluations like these ones
5: piers height 5 times
0.1154
would produce more reliability
Superstructure
1: skewed: < 10;
estimations for better classification
irregularity
2: skewed: 10 < 30;
and regulation of rehabilitation and
3: skewed: 30 < 45;
maintenance programs. For these
5: skewed: > 45
0.1154
reasons, the method used by the SCT
Foundation
1: rock; 3: clay; 5: granular
0.1154
is modified to only evaluate the
Bridge importance
1: D road ; 2: C road;
seismic vulnerability of highway
3: B road ; 5: A road
0.0962
bridges of Mexico.
Seismic hazard
1: A zone; 2: B zone;
4: C zone; 5: D zone
0.1346
Initially, the parameters used by
Hinge
1: number of spans < 1;
SIPUMEX procedure, without direct
2: number of spans between 2 - 3; 0.0192
influence on bridges seismic
3: number of spans between 4 -5;
behavior, were eliminated; like
5: number of spans > 5
parapet,
embankments,
bridge
Type of abutments
1: with concrete wing wall;
surface, hydraulic capacity, etc. The
2: with concrete masonry wall;
former parameter is very important to
3: piles with concrete head;
characterize hydrological problems,
4: piles with masonry head;
but it does not have influence in the
5: no monolithic masonry or
estimation
of
the
seismic
footing over embankment
0.0384
vulnerability.
As
a
second
step,
new
Type of piers
1: solid concrete pier;
parameters
were
proposed,
starting
2: solid masonry pier;
from enquires to expert engineers in
3: single hammer column;
bridge behavior, from practice,
4: single column; 5: others
0.0769
academic
and
governmental
Types of bearings
1: slip bearing; 2: neoprene
institutions. In general, the experts
bearing;
0.0769
recommend a procedure with few
4: steel bearing; 5: steel plate
parameters and aspects of easy
Type of
1: unique span; 2: monolithic
superstructures
girders;
evaluation, to not produce a large and
3: continuous girder; 4: simply0.0769
difficult process.
supported girders; 5: multi-spans
The new parameters to be evaluated
Conservation state
1: maintenance action < 5 years;
include aged of the structure,
3: maintenance action > 5 years;
foundation type, bridge importance,
5: without maintenance
0.0577
conservation condition and seismic
hazard. For these parameters, the experts also indicate the seismic susceptibility categories and
Table 4 Parameters of the modified method
Parameter
Category

weights. For example, for the parameter antiquity, the experts select such as susceptibility category
the year of construction, the design code, or another parameter with a better description of the
bridge deterioration by its aged. The fragility values of the categories were limited between cero to
five, conform to the SIPUMEX format. Different criteria of experts were unified to define a unique
group of weights, using the solution of an equation of multiple variables.
The final procedure considers 12 parameters, with their categories and associated importance
weight, as they are shown in Table 4. In the assignation of the categories of some parameters, the
following aspects were applied:
o The Bridge importance parameter was evaluated by means of the type of road defined by
the SCT, considering A for highway, B for basic roads, C for secondary roads, and D for
communitarian roads.
o The categories of the Seismic hazard parameter are defined accord the four seismic hazard
zone of Mexico
o In the Type of pier parameter, the category others, with fragility of five, is assigned to
bridges were no data information is available. To one-span bridges, a fragility category of
cero is considered.
o If the necessary information for a parameter is not available, the greater fragility category
was assigned
o For all the parameters when some visual damage degree is observed in the visual inspection,
one or two additional fragility values are considered. For example, if there is a pier
evaluated with a fragility category of one and it shows concrete crushing or exposition of the
reinforced steel, a fragility category of two or three is assigned.
With the parameters, categories and weighs defined in Table 4, the fragility index of the modified
method is defined by
12

IVMG = Yi *Wi

(5)

i =1

where Yi is the category assigned to parameter i and Wi is its weight. The vulnerability index of
Equation 5 is grouping in five qualitative parameters, conform the SIPUMEX format. Vulnerability
indexes of 1, 2, 3, 4, and 5 are considered for non vulnerable, slight vulnerable, moderate
vulnerable, vulnerable and very vulnerable, respectively.

3. Applications to a bridges database


The Kim and modified SIPUMEX
methods were applied to a database of
79 bridges situated at the Mexican
Pacific Coast, specifically located in the
States of Colima, 36 structures, and
Michoacn, 43 structures. In Figures 1
and 2, the geographic localization of the
bridge in Colima and Michoacn areas
are shown.
To indicate the general characteristics of
the bridges considered in this evaluation,
the distribution of the categories of some
parameters is presented in Figure 3 in
form of frequencies (left scale) and accumulated frequencies (right scale). In Fig. 3 (a) it is
observed that most of the bridges are structures designed between 1970 and 1980, with 18% of
them with a minimum age of 36 years; only seven bridges were builder before 2000 year. As it is
Fig. 1 Geographic localization of the Michoacn bridges

Fig. 2 Geographic localization of the Colima bridges

Frecuency

(a) Year of construction


60

100.0%

50

80.0%

40

60.0%

30
40.0%

20

20.0%

10
0

0.0%
1940

1960

1970

1980

1990

2000

observed in Fig. 3 (b), 58% of the bridges


have a total length minor to 50 m and only
two structures overcome 400 m of total
length. In addition, 33% of the structures
are one-span bridges and one of them has
14 spans, as it can be observed in Fig. 3(c).
Also, in Fig. 3(d) it can be observed the
type of pier distribution. In this last figure,
10, 20, 30, 32, 33 and 91 are solid pier,
single column, hammer pier, frame piers,
hammer columns with diaphragm, and no
piers (33% as in Fig. 33 c), respectively.
In other aspects, 80% of the bridges have
concrete superstructure, most have
abutments and piers of concrete, masonry
or combinations of these materials. The
bridges of the state of Colima are simplysupported, with constant section and 20%
of maximum curvature. Most of the bridges
located at the state of Michoacn are
simply-supported with constant section,
and have straight superstructures or with
little curvature.

Class

Frecuency

(b) Total length


50

100.0%

40

80.0%

30

60.0%

20

40.0%

10

20.0%

0.0%
10.0

50.0

100.0

200.0

300.0

400.0

> 400

Class

Frecuency

(c) Span number


30

100.0%

25

80.0%

20

60.0%

15
40.0%

10

20.0%

5
0

0.0%
1

14

Class
(d) Type
of pier
100.0%

Frecuency

40

The 79 bridges of the database were


classified by seismic vulnerability class
using the Kim and modified SIPUMEX
methods. Tables 5 and 6 present the
frequency of the bridges assigned to each
qualitative vulnerability class. As these
tables show, there are important
discrepancies
between
the
bridge
classifications obtained with the two
methods. This is due to the fact that Kim
method
considers
more
important
parameters like: Design specifications,
Type of superstructure, Liquefaction,
and Seat length. For Liquefaction
parameter, a high weight is considered,
principally by the collapses reported by this
problem in earthquakes as the one in Kobe,
1995. Also, this method considers a
constant adjusted with characteristics
bridges and damage reported by structures
of the United States.

80.0%

30

60.0%
20
40.0%
10

20.0%

0.0%
10

20

30

32

33

91

Class

Fig. 3 General characteristics of the studied bridges

results of Table 6 are considered reliable.

The modified SIPUMEX method gives


more importance to Seismic hazard,
Aged, Substructure irregularity, and
Foundation. Considering that the
evaluation of bridge fragility with this
method is based on visual inspections, and
the little damage reported in Mxico by
seismic action in this type of structures; the

4. Final commentaries
Table 5 Bridge classification. Kim method
Type of damage Frequency
Colima Michoacn

All

Without damage
Minor damage
Moderate damage
Severe damage
Collapse

9
10
32
28
0

6
4
13
13
0

3
6
19
15
0

The determination of the actual conditions of


bridges is a necessary step to define
maintenance and rehabilitation procedures of
these structures, in particular with limited
economic resources.

To prevent a not desired behavior in bridges, it


is possible to applied simplified and detailed
techniques. The first are considered only to
classify a group of structures, while more rigorous methods are useful to determine the damage
degree of the structures. The simplify methods have to be both, easy application and reliable.
Table 6 Bridge classification. Modified
SIPUMEX method
Type of damage Frequency
Colima Michoacn All
Without damage
Minor damage
Moderate damage
Severe damage
Collapse

3
33
0
0
0

4
38
1
0
0

7
71
1
0
0

The Secretary for Communications and


Transports of Mxico applies periodically a
simple procedure to evaluate highway bridges.
This procedure analyses different aspects, like
pavement condition, average daily traffic, river
caudal, etc. The method is very general, so a
modification was proposed to estimate
specifically the seismic vulnerability of the
structures. Similar modifications are actually
considered
to
evaluate
hydrological
vulnerability.

The modification of the SIPUMEX method is based on including important parameters to estimate
the seismic vulnerability, such as seismic hazard, and to eliminate parameters that do not describe
the fragility of the structures during a seismic action. The new parameters included in the method
were defined by means of experts enquires. With this information a preliminary procedure to define
the seismic fragility of bridges is proposed.
With a database of 79 bridges, located in a Mexican seismic hazard zone, two simplify methods of
evaluation of the seismic fragility were applied. The methods considered are the proposed one, the
modified SIPUMEX method, and the Kim procedure (methodology available in the literature). The
results show that important discrepancies with these two methods are obtained in the bridge seismic
susceptibility classification. This is due to the fact that each method considers different parameters
and weights assigned to the parameters. Keeping in mind the limited reports of damaged bridges in
Mexico in last earthquakes, the results obtained with the SIPUMEX modified method are
considered more consistent. However, it is known that these methods do not have physical rigorous
analyses.
In the application of preliminary methods for the evaluation of the bridges seismic vulnerability, it
is important to consider simplified procedures, with easy evaluations, reliable and invariable results.
The Kim and modified SIPUMEX methods are based on subjective estimations and different results
could be obtained, as it is presented in this paper. So, it is important to consider some modifications
of this type of methods without change its simple application. Actually, for some parameters of the
modified SIPUMEX procedure, elastic evaluations were realized to define the importance of the
parameters categories in the structure seismic behavior. The importance of each category of each
parameter is defined analyzing the statistical variation of maximum displacement of the bridge
elements. With these analyses, is considered that the same modified SIPUMEX methodology could

be applied; only changing the category and weight values. So a simple procedure, based on the
physics of the problem will be formalized.

5. References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]

[9]

MOHELE, J. P., Northridge earthquake of January 17, 1994: reconnaissance report Volume 1
Highway bridges and traffic management, Earthquake Spectra, Vol. 11, supplement c, 1996,
pp: 557-566.
HOUSNER, G. W. and THEIR Jr. c. c., The continuing challenge: report of the performance
of state bridges in the Northridge earthquake, Earthquake Spectra, Vol. 11, No. 4, 1995, pp:
607-636.
NAKAJIMA, H., Seismic performance and repair of mayor steel bridges on the Hanshing
Expressway, Japan, 11th World Conference on Earthquake Engineering, Paper No. 2024,
1996, Mxico.
PINTO, A. V., The Kobe earthquake (January 17, 1995). Damage on RC bridges. Technical
report: Semi-rigid behavior of Civil Engineering Structural Conections COST C1, ISSN
1018-5593, 1999.
GMEZ SOBERN, C., Riesgo ssmico de puentes de hormign armado de seccin cajn y
vanos mltiples, PhD Thesis, Escuela de Caminos, Canales y Puertos de Barcelona,
Universidad Politcnica de Catalua, 2002.
REN, S. and. GAUS, M. P., GIS tools for regional bridge seismic risk assessment,
Technical report GIS-2 Department of Civil Engineering, State University of New York at
Buffalo, 1996.
DGCC, General Direction of Highways Conservation of the Secretary of Communications
and Transports of Mxico, Web page http://dgcc.sct.gob.mx., 2006.
GMEZ SOBERN, C., BARRERA BAUTISTA L. A. and MIRANDA CID D.,
Metodologa de estimacin preliminar de la vulnerabilidad de puentes basada en
procedimientos de la Secretara de Comunicaciones y Transportes. Aplicacin a puentes
carreteros del Pacfico, XV Congreso Nacional de Ingeniera Estructural, Paper No. 139,
2006, Puerto Vallarta, Mxico (in Spanish).
HALDAR, A. and MAHADEVAN, S., Probability, reliability and statistical methods in
engineering design, John Wiley and sons, 2002.

Você também pode gostar