Load Rating of Impaired Bridges Using A Dynamic Method: B. Samali, J. Li, K.I. Crews M. Al-Dawod

EJSE Special Issue: Loading on Structures (2007)
Load Rating of Impaired Bridges Using a Dynamic Method

B. Samali, J. Li, K.I. Crews
University of Technology Sydney, Australia
M. Al-dawod
Maunsell and Partners Consulting Engineers, Sydney, Australia
ABSTRACT: Local Government in Australia is responsible for the operational management and maintenance
of over 20,000 bridges. More than 70% of these bridges comprise aging timber bridges, the load capacity and
structural adequacy of many of which have been impaired over time. This is partly due to increased vehicular
loads with little attention to consequence of such increases. It is now necessary to determine the load carrying
capacity of these bridges using simple yet reliable methods to allow local authorities to upgrade, replace or
sign post at-risk bridges. In this paper a novel dynamic based method is presented by which the in-service
stiffness of the bridge is estimated first. From this stiffness the load carrying capacity of the bridge is
estimated following a statistically based analysis.
dynamic responses are measured with uniaxial
accelerometers which are robust and simple to
1 INTRODUCTION
attach. The data is logged and the bridge deck
A major challenge facing Local Government in
properties evaluated, using a dynamic signal
Australia is to develop effective strategies for the
analyser or a standard computer with special
maintenance and rehabilitation of the extensive
software.
timber bridge stocks which form a key component of
Two sets of bending frequencies are measured for
the road network under its control. Raising the
the bridge, as is, and when loaded by the extra
efficiency and reliability of bridge maintenance
weight.
By loading the bridge, the bending
practices of local government has the potential not
frequency of the bridge decreases. From the
only to minimize costly unscheduled emergency
resulting frequency shift due to added weight, the
repairs, but also to reduce the overall maintenance
flexural stiffness of the bridge can be calculated.
costs, whilst improving the operational effectiveness
User friendly software has also been developed
of its road network.
which allows the estimation of bridge load carrying
The field testing of several timber bridge spans in
capacity from calculated stiffness, adopting a
NSW has been undertaken successfully using a
statistically based approach. The proposed test does
novel and simple dynamic method to estimate the innot require the precise measurement of
service stiffness of the bridge, from which its load
deformations, as is the case for static load tests. It is
carrying capacity is estimated. Coupled with
also much quicker to conduct compared with load
specially developed analysis software, the method
testing, and hence less expensive and much more
provides a measure of the structural adequacy of the
affordable than load testing. It is also safer than load
bridge and a reliable basis for devising appropriate
testing, particularly with respect to old bridges
maintenance or remedial measures.
where applying a large load may further jeopardise
the integrity of the bridge.
2 BASIC CONCEPTS UNDERPINNING THE
PROPOSED TESTING PROCEDURE
3 TEST RESULTS FOR A TYPICAL BRIDGE
The proposed dynamic bridge assessment procedure
As a direct result of the modal analysis, the dynamic
involves the attachment of a few accelerometers
properties of a bridge, such as the natural
underneath the bridge girders and the measurement
frequencies, damping ratio and mode shapes, can be
of the vibration response of the bridge superstructure
obtained. However, the proposed dynamic method
unloaded and with one or more loads (such as a
requires only the first flexural natural frequency for
truck, water tanker, grader, concrete blocks, etc, of
both with and without added mass cases. Figures 1
known weight) applied at midspan. The excitation is
and 2 show the comparison of Frequency Response
generated by a modal impact hammer. The resulting
66
Functions (FRFs) with and without added mass for a

two span tested bridge, respectively.
2.6
10.4 Hz
4.50
4.00
No added mass
Added mass
7.6 Hz
3.00
2.50
2.00
1.50
1.00
0.00
5.00
10.00
15.00
20.00
25.00
Frequency (Hz)
Figure 1. Comparison of sum FRFs for span 1 with and without

added mass
BLACK FLAT ROAD BRIDGE
5.00
4.50
8.65 Hz
4.00
No added mass
Added mass
3.50
20.2
Current best practice in Australia generally

assumes that the fibre strength of any girder is 80 to
100 MPa (depending upon the species). Bending
capacity is predicted by multiplying the assumed
section modulus "Z" (based on the gross section) by
the assumed fibre strength.
Proof loading of timber bridges is expensive and
inherently risky, since it is a well established fact
that high load levels cause permanent and
irrecoverable damage to the wood fibres. This may
result in subsequent failure of a timber girder at load
levels significantly less than that indicated by the
proof test. It is for this reason that most rating
procedures of timber bridges have been based on
applying serviceability load levels (such as from a
water tanker), measuring the deflections in order to
estimate the stiffness and then using an assumed
relationship between strength and stiffness to predict
the load carrying capacity of each girder.
BLACK FLAT ROAD BRIDGE

5.00
3.50
8.65
10.1 Hz
3.00
2.50
2.00
200
1.50
180
1.00
0.00
5.00
10.00
15.00
20.00
25.00
160
Frequency (Hz)
140
Figure 2. Comparison of sum FRFs for span 2 with and without
120
MOR (MPa)
added mass
R2 = 0.0421
100
80
3.1 Flexural Stiffness of the Tested Bridge
60
With added mass given, and frequencies with and

without the mass known, the flexural stiffness of the
tested bridge can be easily calculated. Table 1 shows
the amount of mass added, the first natural
frequency with and without mass and prediction of
the flexural stiffness of tested bridge.
The first span stiffness is calculated as 12.7
kN/mm and the corresponding calculated stiffness
for span 2 is 20.2 kN/mm. It can be seen that the
stiffness results for both spans are different, partly
due to varying span length.
40
20
AS 1720.1 relationship
0
0
5000
10000
15000
20000
25000
MOE (MPa)
Figure 3. MOR vs MOE for round timbers
The relationship between strength and stiffness

used in current load assessment methods is based on
the assumed relationship between Modulus of
Rupture (MOR) and Modulus of Elasticity (MOE)
defined in the Australian Timber Structures Code,
AS1720.1. However, it is not commonly understood
that this relationship is both idealized and
theoretical. Figure 3, illustrates the problems
associated with this approach. This figure presents a
plot of MOR vs MOE data obtained from full scale
testing of round timbers extracted from service with
an average life of 30 years and also compares this
with the AS1720 relationship. It is obvious from the
linear regression co-efficient for the test data that the
relationship between strength and stiffness for aged
poles/girders is not statistically significant.
Furthermore, the theoretical relationship assumed in
AS1720 is not reliable for these timbers, as many
round timbers have a rupture strength significantly
3.2 Load Capacity of the Tested Bridge

The determination of strength of in-service bridge
girders is extremely difficult and complex, unless of
course the girder is broken and the failure load and
loading pattern is known.
Table 1 - Results of predicted stiffness using the proposed
dynamic method
Predicted
First natural
Mass
Span
stiffness
frequency
added
No.
(kN/mm)
(Hz)
(tonnes)
0
10.4
1
2.6
7.6
12.7
2
0
10.1
67
so a modal hammer is usually incapable of imparting

enough energy to excite the bridge, hence a larger
exciter in the form of a drop mass was designed by
researchers at the University of Technology Sydney
(Figure 5). This was undertaken to increase the
precision of the measured modal frequencies.
As mentioned, the modal hammer has been
proven to excite timber bridges adequately but in the
case of concrete, steel or other stiff bridges, did not
provide enough energy to adequately excite the
lower frequency spectrum of 5 to 30 Hz. Thus, the
first vibration modes of the structures have not been
as clear, and as definable, as desired.
The new exciter consists of a PCB 200C50
Quartz Impact Force Sensor to sense the force and
frequency applied to the bridge from the exciter; a
safe lifting mass of 20kg to provide the impact;
replaceable rubber tips to provide low frequency
excitation and reduce second impact; and a steel
frame and base plate. The mass is guided by three
vertical rods to where it impacts the sensor at the
bottom of the guides. The force is then transferred to
the structure through the base plate. The mass can be
dropped from any height up to 1.5m depending on
the level of excitation needed.
lower than that predicted by the Code relationship.

For example, extensive testing of some 1200 round
timber poles indicates that the actual 5th percentile
strengths for strength group 1 & 2 timbers range
between 30 and 55 MPa, not 80 to 100 MPa as
previously assumed.
In order to assess the strength of timber bridge
girders with any degree of reliability, it is necessary
to develop strength models, which reflect the actual
bending capacity of timber. This should take into
account the uncertainties associated with
determination of the geometric section properties
and the actual strength properties. Such a model has
been developed to form the basis of the proposed
load testing system developed in this paper. Using
test data obtained from extensive testing of full scale
round timbers, a relationship between actual
measured stiffness (EI) and actual bending capacity
has been derived, with correlation coefficients in the
range 0.4 to 0.5 for lower bound 5th percentiles, as
indicated in Figure 4.
400
350
BM / Girder (kNm)
300
250
200
150
100
50
0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
EI (1e12)
Figure 4.
timbers
Bending strength vs gross stiffness for round
Using a probabilistic approach, this relationship

can be used in reliability-based models to predict the
load capacity of a deck from the stiffness data
obtained from the dynamic frequency method, with
acceptable and transparent degrees of uncertainty.
Figure 5. The new exciter and its schematic
Applying the probabilistic approach described above,

the estimated live load factor (defined as the ratio of the
net factored moment capacity and the moments,
including live load allowance, caused by a T44 truck
weighing 42.5 tonnes per lane) is 1.2 for span 1 and 1.5
for span 2. This means that the maximum load carrying
capacity of the bridge is estimated at 50 tonnes.
4.2 Alternate locations for added mass

Based on the Dynamic Frequency Analysis (DFA)
procedure, prediction of the flexural stiffness of the
bridge requires determination of the frequency shift
produced as a result of added mass, which is
distributed at midspan of the bridge. Theoretically
speaking, the DFA procedure should produce
accurate prediction of the flexural stiffness of any
given bridge. However, in practical situations, a few
problems do arise in field testing which may
severely impact on the accuracy of the predictions
using a DFA procedure. In addition, the possibility
of varying the added mass locations may have
4 PRACTICAL CONSIDERATIONS
4.1 A new exciter for larger bridges
The method proposed above is not limited to timber
bridges. The method has been extended to test other
and larger bridges made of concrete and steel. To do
68
significant benefits in terms of reducing the testing

time and easing the traffic flows during testing. In
the following, a few possible field testing situations
are described, which may adversely impact on the
results obtained from the standard midspan mass
addition procedure:
where k is the flexural stiffness of the bridge,

M is the total added mass and 1 and 2 are,
respectively, the natural frequencies before and after
adding mass in rad/sec.
In this case, because the added mass, M, is
located at the midspan of a bridge, and as it
contributes directly to modal mass of the first mode,
there is no need for mass compensation.
(i). Due to the existence of the kurbs, added mass

cannot be completely and uniformly
distributed along the midspan;
(ii). The physical size of the blocks precludes them
as being point masses as assumed in the
theory; in addition, for practical convenience it
may be required to arrange the blocks in a
more distributed fashion;
(iii). Some extreme situations, such as when the test
bridge is in a very poor condition (eg, some old
and deteriorating timber bridge), may prohibit
the use of a crane truck to carry mass blocks to
the midspan. The only possible location for
adding mass may be quite far from the
midspan of the bridge;
(iv). As using a trailer as added mass has practical
advantages. The mass of a trailer cannot be
distributed evenly along the midspan, and
hence, the impact of this on the accuracy of
results needs to be investigated.
(v). There are possible arrangements for adding
mass on the bridge which may have
advantages in terms of easing traffic flow.
These need to be also considered.
4.4 Calculation of Flexural Stiffness with Modal Mass

Compensation
Considering the theory of DFA, it is clear that the

first modal mass is the key for improving the
accuracy of the method. The current DFA procedure
requires the added mass to be positioned at
midspan to ensure the first modal mass of the
structure is increased by the total added mass. In
Equation 1, if M is considered as equivalent first
modal mass instead of physical added mass, it is
then possible to increase the accuracy of the stiffness
prediction for various scenarios mentioned above.
To determine the modal mass for each case of added
mass, the added mass matrix is normalized with
respect to the mode shape. The modal mass is
therefore calculated from,

mi = iT Mi
where M is the added lumped mass matrix and

i the mode shape corresponding to the added mass
mode.
In Equation 1, the added mass M can then be
replaced with the modal mass mi to yield the
following,
In summary, an investigation of the resulting

modal mass for each arrangement of added mass is
required to assess the accuracy of stiffness
predictions, due to these variations, and to find
possible solutions to further improve the accuracy
for various practical situations. The following
section attempts to address this particular issue.
k=
at Midspan
f 12 f 22
(3)
4.5 Case Study - a previously tested bridge
The principle of the DFA procedure in bridge testing

has been discussed extensively in previous papers by
authors (eg, Champion, et al. 2002; Crews, et al.
2004a,b,c; Crews et al. 2005; Li et al. 2004; Li et, al.
2005; Samali, et al. 2002) and hence this paper will
avoid unnecessary details. In summary, by knowing
the natural frequencies before and after adding mass,
the flexural stiffness of the bridge can be estimated
as follows:
12 22
M
12 22
(2 )2 f 12 f 22 m
Equation 3 can predict the flexural stiffness with

the added mass, compensated with the modal
mass m i . In Equation 3, f1 and f2 are, respectively,
the natural frequencies of the bridge before and after
adding mass but expressed in Hertz.
4.3 Calculation of Flexural Stiffness with Mass Located
k=
(2)
To confirm and evaluate the proposed modal

compensation method, a five girder composite
steel/concrete bridge was selected for modelling and
calculation. The bridge over Redbank creek in New
South Wales near Sydney was chosen for this
purpose.
The composite steel/concrete bridge structure was
built in 1945, and consists of three simply supported
spans of length 10.46m, 10.67m and 10.46m,
(2.8)
respectively,
with a carriageway width of 6.1m. The
concrete slab, with an average thickness of 160mm,
is supported by five RSJ (22x7) girders spaced at
(1)
69
1.37 meter centres. Transverse reinforced concrete

diaphragms are located at each support and at
midspan (Figure 6).
Figure 7. Microstran model of Redbank Creek bridge
The concrete deck is modelled using a grid of

beams in the X-Z plane. The total area of these
beams equals the total area of the concrete deck.
This is important in the dynamic analysis as it
ensures that the mass of the model is equal to the
actual mass of the concrete deck. The grid work
assumes the relative stiffness the deck provides is
equal in both the X and Z directions. This is a
reasonable assumption as the concrete deck is
uniform in depth across the entire bridge.
The structure is assumed to be simply supported
and, therefore, the supports are modelled with
simple pinned supports. This ignores any
contributing effects from adjoining spans as only
one typical span is modelled.
Figure 6. Underside of Redbank Creek bridge
Microstran computer program was chosen to

model the bridge structure due to its relative
simplicity. The bridge structure was modelled using
space frame elements with six degrees of freedom at
each node. The space frame elements were chosen
over the grillage elements as they provide greater
flexibility in modelling the composite bridge
structure.
The structural elements modelled included the
Rolled Steel Joist (RSJ) girders, the reinforced
concrete deck and the transverse reinforced concrete
diaphragms. The concrete kerbs and barriers were
not modelled as it was assumed that their
contribution to the overall global stiffness is
negligible.
The same model was later used to estimate the
load carrying capacity of the bridge using current
code provisions and after calibrating it using the
experimental stiffness results.
4.5.2 Material Properties

The material properties adopted for the model were:
Concrete: E = 30,000 MPa
v = 0.15
= 2.4 tonne/m3
Modulus of Elasticity
Poissons ratio
Mass Density
Steel: E = 200,000 MPa Modulus of Elasticity

v = 0.33
Poissons ratio
= 7.85 tonne/m3 Mass Density
4.5.3 Static analysis
A linear elastic static analysis of Redbank Creek
Bridge by Microstran will determine a value for the
global flexural stiffness of the structure. Static
loading was modelled as point loads over each
girder at midspan (Figure 8).
4.5.1 Modelling Assumptions

The modeling of the composite action between the
steel girders and the concrete deck was through
additional dummy members. These dummy
members model the composite action through a rigid
vertical member that connects the steel girder to the
concrete deck rigidly. These members have a
relatively large moment of inertia compared to
surrounding members. This ensures compatibility of
deformation is maintained through the rigid
connection.
70
4.6 Added mass cases

The determination of flexural stiffness of the bridge
structure is through a shift in the natural frequency.
This frequency shift is due to the additional mass on
the structure. The standard DFA procedure requires
the added mass to be evenly distributed at midspan.
The possibility of varying the location of the added
mass offers significant and practical benefits to the
testing procedure. A total of ten varying patterns of
added mass have been investigated. The common
factor to each type of added mass is that the total
additional mass remains constant to ensure the
results are comparable. For the case investigated
here the total added mass is 10.6 tonnes.
Figure 8. Static load applied to the model at midspan
The determination of the flexural stiffness of the

bridge structure through static means is the familiar
force displacement relationship where
k=
Load
= 98,355kN / m
Displacement
4.6.1 Type 1 added mass

The added mass is lumped at midspan over each
girder. This is based on the current theory where the
total added mass increases the modal mass of the
structure (Figure 10).
(4)
The results of the static analysis are shown in Table

2.
Table 2. Results of the static analysis
Girder
1
2
3
4
5
Total
Load
kN
1,000
1,000
1,000
1,000
1,000
5,000
Vertical
Displacement
mm
50.8
50.9
50.9
50.9
50.8
50.8
Figure 10. Type 1 added mass

The added mass is lumped at midspan but only on
half the width with girders 1, 2 and 3 loaded. The
advantage of such a configuration is the ability to
allow traffic to continue through one lane as the
structure is loaded and unloaded. This reduces the
bridge closure time, hence allowing bridges with
large traffic volumes to be tested with minimal
disruption to traffic (Figure 11).
4.5.4 Dynamic analysis

The dynamic analysis requires the stiffness and mass
matrices to be determined from the Microstran
model. Forty five (n = 45) nodes or DOF (Figure 9)
were used in the analysis.
- Nodes
- Support
- Added Mass
Figure 9. Bridge model with n = 45 DOF
71

The added mass is identical to that of Type 2, only
mirrored. The position of the added mass is on the
opposite side (Figure 12).
lumped at every DOF. This is only considered for

theoretical benchmarking (Figure 15).


The added mass is distributed to each girder as in
Type 1, but offset from midspan. This type of
loading has practical implications in that the lumped
mass may not be able to be placed exactly at
midspan; therefore, this type of loading will
investigate the sensitivity of the current method, to
the loading location (Figure 16).

The added mass is located at spans on girders 2
and 4. This type of loading represents a vehicle
which may be used as the added mass. The total
lumped mass is proportioned to each wheel location.
This greatly reduces the loading and unloading time
of the bridge (Figure 13).

The added mass is identical to that of Type 7 only
with a greater offset from midspan. This case has a
theoretical value when investigating the results
(Figure 17).

The added mass is located at midspan but only on
the central girders 2, 3 and 4. This type of loading is
considered for bridge structures that do not allow a
lumped mass on the exterior girders. This may be
due to the location of the barriers or kerbs on the
bridge (Figure 14).

The added mass is located with two masses on
girders 2 and 4 on either side of midspan. They are

The added mass is distributed over the entire
structure. The total added mass is distributed and
72
closely spaced and resemble the added mass of

wheels from a trailer or vehicle. The advantage of
this type of added mass configuration is in the
reduction of time to load and unload the bridge
(Figure 18).
Table 3. Calculated natural frequencies for Redbank Creek

bridge model
4.8 Natural Frequency and Mode Shapes with Added
Mode
1
2
3
4
5
6
7
8
9
10
Natural Frequency, Hz
12.00
12.98
31.66
39.40
41.34
47.46
57.33
59.05
76.50
79.56
Mass

The added mass is located at midspan but only on
the exterior girders. This configuration would limit
the disruption to traffic, allowing a single lane
through the centre of the bridge to remain open
(Figure 19).
The natural frequencies and mode shapes are

calculated using the MATLAB program for each
type of added mass. The results for only the first ten
modes are shown in Table 4.
Table 4. Natural frequencies for ten cases of added mass
Natural Frequency (Hz)

Type
Type
Type
Type
Type
9.47
8.77
8.77
10.94
9.41
10.26
11.83
11.83
12.35
11.81
25.69
25.65
25.65
29.79
28.16
39.40
39.40
39.40
30.55
39.40
41.34
41.34
41.34
31.99
41.34
47.46
47.46
47.46
41.12
47.46
50.54
51.48
51.48
46.29
49.35
59.04
59.04
59.04
48.49
59.04
71.25
68.79
68.79
56.45
69.06
10
72.96
71.86
71.86
65.61
70.88
Mode
4.7 Natural Frequency and Mode Shapes

The natural frequencies and mode shapes are derived
using the MATLAB program. The number of natural
frequencies calculated will equal the number of
DOF. Therefore, for the case n = 45 there will be
forty five natural frequencies and corresponding
mode shapes calculated. The first ten frequencies are
of most interest to the analysis and hence are shown
in Table 3. The higher frequencies and
corresponding mode shapes become too complex
and somewhat irrelevant.
The first frequency represents the first flexural or
bending mode, the second frequency represents the
first torsion or twisting mode and the third frequency
represents the transverse flexural or bending mode.
These are the basic representations of the modes
with all higher modes exhibiting some correlation to
the fundamental modes.
Natural Frequency (Hz)

Type
Type
Type
Type
Type
10
10.42
9.64
10.15
9.63
8.74
11.30
10.45
11.00
11.46
9.33
Mode
73
27.49
25.30
25.39
30.25
26.59
34.04
36.92
33.35
34.90
39.40
35.76
38.71
34.95
37.50
41.34
Added
10.6
10.6
10.6
10.6
10.6
185
110
151
109
92.2
85.5%
10.2%
51.6%
9.4%
-7.5%
Mass (t)
Stiffness
(MN/m)
41.03
44.84
40.07
44.61
47.46
49.59
44.95
41.79
46.64
54.61
50.99
58.44
52.67
50.90
59.04
66.01
60.97
58.14
67.65
69.28
10
68.70
77.95
78.12
69.11
70.22
Error (%)
The MATLAB program performs the required

transformation and calculation to determine the FRF
for the case of bridge as is and for each case of
added mass. It is from these FRFs that the frequency
shift required to determine the flexural stiffness is
measured.
The predictions of the flexural stiffness for each
added mass case are presented in Table 5. The
frequency shift error is expressed considering Type
1 added mass as the benchmark. This is also the case
for the stiffness error calculation.
It can be seen that added mass cases 5, 9 and 10,
in addition to benchmark case 1, can produce
accurate results with errors of less than 7%. This
provides lots of flexibility when choosing the most
appropriate and convenient mass locations without
compromising the overall accuracy.
4.9 Frequency Response Function of the Bridge

The transient analysis of the bridge model with
the impact force imparted by a tuned modal impact
hammer, allows the calculation of the acceleration
responses of the structure. The acceleration response
data, when processed using the fast Fourier
transform (FFT) allows the Frequency Response
Function (FRF) to be obtained. This method
replicates the Dynamic Frequency Analysis (DFA)
procedure applied to bridge structures and allows the
determination of the first flexural frequencies from
the appropriate FRF.
5 CONCLUSION
A new method, based on dynamic response of

bridges to an impact load, is proposed to measure
Table 5. Flexural stiffness prediction for different mass types
the in-service flexural stiffness of bridges. Utilizing
a statistically based analysis, the knowledge of
Type
Type
Type
Type
Type
flexural stiffness can be converted into an estimate
as is
of the load carrying capacity of the bridge (for
1
2
3
4
5
timber bridges). The reliability and simplicity of the
Frequency 12.00 9.47
proposed methodology has been demonstrated by
8.77
8.77
10.94 9.41
testing over 200 bridge spans covering a wide range
(Hz)
of single and multi-span timber bridges. The results
Shift (%)
0.0% 26.6% 36.8% 36.8% 9.7%
27.5% pertaining to two spans of one of these bridges are
reported in this paper, along with the underlying
N/A
0.0%
38.0% 38.0% -63% 3.2%
Error (%)
principles and methodology adopted.
The methodology was refined and extended to
Added
10.6
10.6
10.6
10.6
10.6
larger concrete and steel bridges but using the same
Mass (t)
principles.
This necessitated the design and
fabrication of a larger exciter capable of imparting
Stiffness
99.7
69.2
69.2
296
96.3
larger impacts to the bridge to excite it.
(MN/m)
The sensitivity of varying mass locations on the
accuracy of measured frequency and hence the
N/A
0.0%
-30.% -30.% 197% -3.5%
Error (%)
predicted stiffness was also investigated. It is found
that in addition to the optimal location to add mass,
Type
Type
Type
Type
Type
other alternatives also exist which allow the addition
6
7
8
9
10
of mass with more ease and less obstruction to
traffic without sacrificing the accuracy of results.
Frequency
10.42 9.64
10.15 9.63
9.33
For the case of larger concrete and steel bridges,
(Hz)
the developed computer model is also used to
estimate the load carrying capacity of the bridge
Shift (%)
15.1% 24.4% 18.2% 24.6% 28.6%
using current code provisions and after calibrating it
using the experimental stiffness results.
Error (%)
-43%
-8.4%
-31%
-7.8%
7.2%
74
6 REFERENCES
Champion, C., Samali, B., Li, J., Crews, K.I., and Bakoss, S.L.,
"Assessing the Load Carrying Capacity of Timber Bridges
Using Dynamic Methods", IPWEA Queensland Division
Annual Conference, Qld, October 2002.
Crews, K.I., Samali, B., Bakoss, S. L., and Champion, C.,
Overview of Assessing the Load Carrying Capacity of
Timber Bridges using Dynamic Methods, Austroads 5th
Bridge Conference, May 2004, Hobart, Australia.
Crews, K.I., Samali, B., and Li, J., Reliable Assessment of
Aged Timber Bridges Using Dynamic Methods,
Proceedings of 8th World Conference on Timber
Engineering, June 2004, Finland.
Crews, K.I., Samali, B., Bakoss, S.L., and Champion, C.,
Testing and Assessment Procedures to Facilitate the
Management of Timber Bridge Assets, the 3rd CECAR
Civil Engineering Conference in the Asian Region, August
2004, Seoul, Korea.
Crews K.I., Samali, B., Li, J. and Al-Dawod, M. Reliable
Assessment of Road Bridges using Dynamic Procedures.
Australian Structural Engineering Conference 2005:
Structural Engineering - Preserving and Building into the
Future, 11-14 September 2005, Newcastle City Hall,
CDROM. Australia: Structural College of EA.
Li, J., Samali, B., and Crews, K.I., Determining Individual
Member Stiffness of Bridge Structures Using a Simple
Dynamic Procedure, Acoustics Australia, Vol.32, No.1,
April 2004, pp 9-12.
Li J., Samali B., Crews K.I., Choi, F. and Sherestha R.,
Theoretical and Experimental Studies on Assessment of
Bridges Using Simple Dynamic Procedures. Australian
Structural Engineering Conference 2005: Structural
Engineering - Preserving and Building into the Future, 1114 September 2005, Newcastle City Hall, CDROM.
Australia: Structural College of EA.
Samali, B., Crews, K.I., Bakoss, S.L., Li, J., and Champion, C.,
"Assessing the Structural Adequacy of Timber Bridges
Using Dynamic Methods", IPWEA NSW Division Annual
Conference, Coffs Harbour, November 2002.
75

Load Rating of Impaired Bridges Using A Dynamic Method: B. Samali, J. Li, K.I. Crews M. Al-Dawod

Uploaded by

Copyright:

Available Formats

Load Rating of Impaired Bridges Using A Dynamic Method: B. Samali, J. Li, K.I. Crews M. Al-Dawod

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Load Rating of Impaired Bridges Using A Dynamic Method: B. Samali, J. Li, K.I. Crews M. Al-Dawod

Uploaded by

Copyright:

Available Formats

EJSE Special Issue: Loading on Structures (2007)