ANALYSIS OF THE MEELIYADDA BRIDGE FAILURE IN KURUNEGALA

Abstract

This paper investigates the failure of the Meeliyadda Bridge, which occurred, when two trucks carrying metal were crossing the bridge. It was reported that some of the lateral bracings of the top chord had removed with the failure of the bridge. Finite element (FE) analysis was carried out using the SAP2000 FE software, where two different FE models were developed: one with all the lateral bracings and the other without three consecutive lateral bracings at the entering end of the trucks. Results from the FE analysis show that removal of the top lateral bracings increases the unbraced length of the top chord members in the lateral direction, reducing their compressive load capacity. This initiates a compression failure in the top chord which could be the reason for the failure in the Meeliyadda Bridge.

Full text

- 21 -
International Research Symposium on Engineering Advancements 2016 (IRSEA 2016)
SAITM, Malabe, Sri Lanka
ANALYSIS OF THE MEELIYADDA BRIDGE FAILURE IN KURUNEGALA
S. I. S Atapattu1, A. Amaraweera1, S. D. L. H. Siyambalapitiya1, H. U. A. Weeraratne1, T. L. Nabrees1
and H. D. Hidallana-Gamage2
1 Undergraduate student, Department of Civil & Infrastructure Engineering, Faculty of Engineering, South Asian
Institute of Technology and Medicine (SAITM), Sri Lanka, Email: shashika24@gmail.com
2 Senior Lecturer, Department of Civil & Infrastructure Engineering, Faculty of Engineering, South Asian Institute
of Technology and Medicine (SAITM), Sri Lanka, Email: hasitha.saitm@gmail.com
ABSTRACT
This paper investigates the failure of the Meeliyadda Bridge, which occurred, when two trucks carrying
metal were crossing the bridge. It was reported that some of the lateral bracings of the top chord had
removed with the failure of the bridge. Finite element (FE) analysis was carried out using the SAP2000
FE software, where two different FE models were developed: one with all the lateral bracings and the
other without three consecutive lateral bracings at the entering end of the trucks. Results from the FE
analysis show that removal of the top lateral bracings increases the unbraced length of the top chord
members in the lateral direction, reducing their compressive load capacity. This initiates a compression
failure in the top chord which could be the reason for the failure in the Meeliyadda Bridge.
Key words: Steel trusses, Lateral bracings, Unbraced length, Finite element analysis
1. INTRODUCTION
According to the RDA database 2015 there are
4210 local highway bridges in Sri Lanka. Out of
these 218 are steel bridges, which include about
100 truss type steel bridges. Truss is a simple
structure made out of pin ended elements, where
the primary forces in the members are axial
forces. Since the truss is an open web structure,
they can be designed for higher depths with less
self-weight when compared with the solid web
systems. It could be also noted that steel has high
strength to weight compared to concrete and
most of the other materials.
Truss type steel bridges can be classified into
different types according to the structural
systems. Warren, Pratt and Howe are some of the
basic truss types. Different truss types have come
up later with some modifications to the basic
truss types. Varied height truss types became
popular with the aesthetical appearance and can
be used for higher spans effectively. Warren,
Pratt, Double warren and varied height truss are
some of the common truss types used in Sri
Lanka. Veralgastotupola Bridge, Muwagama
Bridge, Nattupana Bridge are some recently built
truss type steel bridges in Sri Lanka. Figure 1
shows the Muwagama Bridge located in
Muwagama, Rathnapura.
Most of the existing truss type steel bridges in Sri
Lanka were constructed during colonial period
about 100 years ago. Most of these bridges are
outdated for the modern traffic conditions, where
the bridges such as Allawa Bridge, Gampola
Brige and Katugasthota Bridge have been
replaced with newer concrete/steel bridges.
Figure 2 shows the old steel bridge located in
Gampola [1].
Figure 1: Muwagama Steel Bridge [1]
Figure 2: Gampola Steel Bridge

- 22 -
International Research Symposium on Engineering Advancements 2016 (IRSEA 2016)
SAITM, Malabe, Sri Lanka
However, some of the old truss type steel bridges
such as Peradeniya Bridge, Kochchikade Bridge
and Moratuwa Bridge are still on operation with
the modern traffic. Figure 3 shows the steel truss
bridge in Peradeniya. It is important to
investigate the present condition and suitability
of these existing bridges for the modern traffic as
a bridge failure could be devastating causing
damage to the vehicles and loss of human lives.
Figure 3: Peradeniya Steel bridge
This paper investigates the failure of the
Meeliyadda Bridge, which has been occurred
recently, when two trucks carrying metal were
crossing the bridge. A view of the Meeliyadda
Bridge before the failure is shown in Figure 4,
while a view of the failed bridge is shown in
Figure 5.
Figure 4: Meeliyadda Bridge before the failure
Figure 5: Meeliyadda Bridge after the failure
The finite element (FE) modelling and analysis
were carried out using the SAP2000 FE software
[2]. Results from this study indicate that the
initial failure of the top lateral bracings reduce
the compressive load capacity of the top chord
members causing a failure in the top chord.
Results from the FE analysis for failure load and
failure mode agreed reasonably well with those
seen during the actual bridge failure. This paper
therefore highlights the importance of carrying
out a detailed study about the existing old steel
bridges in Sri Lanka. This will enable engineers
to investigate their suitability for modern traffic
conditions, avoiding their failure, saving damage
to vehicles and loss of human lives.
2. LITERATURE SURVEY
2.1. History
The history of truss bridges dates back in 1779
when the first truss bridge was built in
Coalbrookdale, U.K. Several bridge failures were
happened in the history of steel bridges.
Ludendorff Bridge failure, San Francisco
Oakland Bay Bridge collapse, Ashtabula River
Railroad disaster were some examples for the
truss type steel bridge failures in the world [3,4].
By observing large number of bridge failures, the
common reasons can be noted as,
 Accidental over load and impact
 Force majeure (flood, earthquake etc)
 Structural and Design errors
 Construction and supervision mistakes
 Lack of maintenance
Bridge designers could learn lessons from those
failures and new design techniques and
improvements were implemented later as a result.
2.2. Bridge Failures in Sri Lanka
Paragasthota Bridge failure, Ehelakanda Bridge
failure and Meeliyadda Bridge failure were some
of the steel bridge failures happened in Sri
Lanka. Both Paragasthota and Ehelakanda Bridge
failures were analyzed and the results are
presented in a previous paper [1].
Ehelakanda Bridge failed due to its self-weight
while it was constructing in 2006. The reason for
this failure was identified as the lateral torsional
buckling of the top chord members close to the
support. This was a pedestrian bridge where the
lateral bracings could not be provided for the top
chord members close to supports to meet the
minimum headroom requirement of 2m.

- 23 -
International Research Symposium on Engineering Advancements 2016 (IRSEA 2016)
SAITM, Malabe, Sri Lanka
Paragasthota Bridge failed when a 10 wheel
tipper with 5 cubes of metal (weight of about 20-
25 tons) was crossing the bridge. The movement
of the tipper towards the middle of the bridge
increases the axial forces in both top and bottom
chords. Results from a previous study [1] showed
that the progressive collapse of the bridge was
started with the tensile failure of a bottom chord
member at the middle of the truss.
Figure 6: Ehelakanda Bridge Failure
Figure 7: Paragasthota Bridge Failure
These bridge failures were predicted reasonably
well by using the SAP2000 FE software in a
previous paper [1] and hence a similar study has
been carried out in this paper to investigate the
failure of the Meeliyadda Bridge.
2.3. Meeliyadda Bridge Failure
The Meeliyadda Bridge was constructed in about
1820 by the British government in order to
transport rubber to a nearby factory. The bridge
was constructed across the Deduru Oya in the
Miliyadda area on the road connecting
Keppettigala and Kurunegala. This is a Pratt type
truss bridge having a length of of 37.2m, width of
4.2m and height of 3.6m.
On 25th of July 2015 this bridge failed when 2
trucks carrying metal was crossing the bridge. It
was evident from the villages that the lateral
bracings to the top chord were subjected to
corrosion and were also damaged time to time by
hitting the tippers crossing the bridge. However,
exact cause for the failure was not known and
hence it was studied in this paper.
3. FINITE ELEMENT ANALYSIS
The failure investigation of the Meeliyadda
Bridge was carried out in this paper by using the
SAP2000 FE software [2]. By observing the
failed bridge it was identified that some of the
lateral bracings to the top chord were removed
during the failure. By observing the failed bridge
and analyzing the information gathered from the
villages, it can be predicted that the top lateral
bracings would have been broken initially, which
might be the reason for the bridge failure.
Two finite element models were prepared in this
study, one model with the lateral supports and the
other without lateral supports to analyze the
effect of lateral supports. The finite element
model of the Meeliyadda Bridge with the top
lateral supports is shown in Figure 8. When there
are no lateral supports, unbraced length of the top
chord members in the lateral direction is around
18.6m as illustrated in Figure 9.
Figure 8: FE model of the Keppettigala Bridge
Figure 9: Unbraced length of the bridge
3.1. Material Properties
Following material properties were used in the
analysis by assuming all the truss members are
Unbraced Length
(3.1mx6 = 18.6m)

- 24 -
International Research Symposium on Engineering Advancements 2016 (IRSEA 2016)
SAITM, Malabe, Sri Lanka
constructed with structural steel.
 Weight per unit volume =76.97 kN/m3
 Modulus of Elasticity, E =205 kN/mm2
 Minimum Yield Stress =275 N/mm2
 Minimum Tensile Stress =448.1 N/mm2
3.2. Loading
Dead load, Superimposed Dead load and live
load of the trucks were accounted in the analysis.
Dead loads are the self-weight of the structural
elements including the weight of the two trusses
and the lateral supports. Loads due to Reinforced
concrete deck slab and the asphalt layer was
considered as the super imposed load, where total
thickness of the concrete deck was assumed as
150mm in the analysis. According to RDA details
two trucks with metal load was going on the
bridge at the failure time.
It was evident from the villages that the truck
entered the bridge firstly was close to the remote
end of the bridge when the second truck was
entering the bridge. This arrangement of the two
trucks is illustrated in Figure 10. During the
analysis, front truck was kept close to the remote
support (without moving), while the rear truck
was moved along the bridge. Different load cases
were created in the analysis depending on the
position of the rear truck.
Figure 10: Side view of the two trucks on the
bridge
It was estimated that the total weight of a truck
with the metal load is 16000kg (W=16tons).
About 70% of the load (0.7W) was distributed to
the rear axle while 30% (0.3W) was distributed
to the front axle in the analysis. The load was
inserted as point loads to the lateral cross girders
at the bottom chord level, where half of the axle
load is distributed through each wheel as shown
in Figure 11.
Figure 11: Load distribution through the axles
3.3. Results
Results from the FE analysis for the two FE
models: with top lateral bracings and without top
lateral bracings are discussed below.
FE model with all the top lateral bracings
Axial force variation of the truss members were
analyzed depending of the movement of the rear
truck. The load case envelope was developed by
considering all the load cases to identify the
maximum load in each truss member. Figure 12
illustrates the axial force variation of the truss
members for the load envelope in the FE model
with all the top lateral bracings. It is clear that the
bottom chord members at the middle of the truss
have higher tensile forces while those at the
middle of the top chord have higher compressive
forces.
Figure 12: Axial Force Diagram for the load case
envelope
Figure 13 illustrates demand/capacity ratios of
the truss members for the load case envelope.
Demand/ capacity ratio is the ratio between the
applied axial force and the axial load capacity of
a given member. When the member is under
tension, tensile load capacity of the member is
used to calculate the demand/capacity ratio,
while the compression load capacity is used
when the member is under tension.
Member 7

- 25 -
International Research Symposium on Engineering Advancements 2016 (IRSEA 2016)
SAITM, Malabe, Sri Lanka
Figure 13: Demand/capacity ratios in the model
with lateral bracings
Top lateral bracings have demand/capacity ratio
in the range of 0.7-0.9. However, according to
Figure 13, none of the truss members have
exceeded their demand/capacity ratios over 1,
indicating no failure will occur in the truss
members with the existence of the top lateral
bracings. It is therefore evident that the bridge is
safe under the given loading if the lateral
supports could securely withstand.
FE model without the first three top lateral
bracings at the entering end
By observing the failed bridge it could be seen
that the first three top lateral bracings from the
entering end of the bridge were completely
removed during the failure. FE analysis was
therefore carried out by removing the first three
top lateral bracings from the entering end. This
increases the unbraced length of the top chord
members in the lateral direction from 3.1m to
18.6m as illustrated in Figure 9.
The increase in the unbraced length reduces the
compressive load capacity of the top chord
members from 2403kN to 674kN. The maximum
compressive forces could be seen in the members
at the middle of the top chord, reaching a value
of about 720kN. These members exceed their
compressive load capacity of 674kN. This could
be seen in the FE model without lateral bracings
as shown in Figure 14. However it could be noted
that the top chord members do not fail with
lateral torsional buckling as the slenderness ratio
of the members are less than 180.
SAP2000 FE software also provides design check
of the structural members based on the selected
design code. In this study design check of the
steel truss members were carried out using the
BS 5950-1:2000 [5] design standard. Design
check data sheet for the critical top chord
member (member 7) in the FE model without
lateral bracings is illustrated in Figure 15. It is
clear that the FE model accurately predicts the
failure of the top chord member by giving a
warning message as seen in Figure 15. However
this warning message could not be seen in the
critical top chord members of the FE model
without top lateral bracings as illustrated in
Figure 16.
Figure 14: Demand/capacity ratios in the FE
model without first three lateral bracings from
the entering end
Figure 15: Design check data sheet of member 7
in the FE model without top lateral bracings
Figure 16: Design check data sheet of member 7
in the FE model with all the top lateral bracings

- 26 -
International Research Symposium on Engineering Advancements 2016 (IRSEA 2016)
SAITM, Malabe, Sri Lanka
Axial force variation of the critical top chord
member (member 7) was plotted with the
movement of the rear truck while placing the
front truck close to the remote support of the
bridge. Figure 17 illustrates the axial force
variation of the member 7 with the movement of
the rear truck in the FE model without top lateral
bracings. When the truck reaches 5-10m distance
from the entering end, the member reaches its
compressive load capacity of 674kN. Maximum
axial force can be seen when the truck is at the
middle region of the bridge giving a peak force
of about 818kN. This is well over 674kN
confirming the failure of the top chord
Figure 18 illustrates the axial force variation of
member 7 with the movement of the rear truck in
the FE model with all the top lateral bracings.
Axial force variation of the member is very
similar to that observed in the previous FE
model. However, it is clear that the maximum
axial force in the member 7, which is about
824kN is well below the compressive load
capacity of the member which is about 2403kN.
Figure 17: Axial force variation of member 7 in
the FE model without top lateral bracings
Figure 18: Axial force variation of member 7 in
the FE model with all the top lateral bracings
4. CONCLUSION
By analyzing the results from FE analysis it can
be concluded that the bridge could withstand the
loads of the two trucks if all the top lateral
supports were held in their positions without
failure. However, failure of the top lateral
bracings increases the unbraced length of the top
chord members in the lateral direction
considerably, reducing the compressive load
capacity of the top chord. This could be the
reason for the failure of the bridge as the removal
of the top lateral bracings was observed in the
failed bridge. This paper therefore highlights the
importance of the lateral supports in a truss type
steel bridge, where their failure can cause the
failure of the entire bridge.
5. REFERENCES
[1] K. Baskaran, H. D. Hidallana Gamage, and
W. W. N. Karunarathna, “Finite Element Analysis
of Truss Type Steel Bridges” IESL Annual
journal, October 2011.
[2] “SAP2000 Version 11,” Computer software,
Computers & Structures INC.
[3] M. V. Bioezma, and F. Schanak, “Collapse of
Steel Bridges,” Journal of performance of
constructed facilities, /ACSE/21:5, pp. 398-405,
2007
[4] A. Astaneh-Asl,“ Progressive Collapse of
Steel Truss Bridges, the Case of I-35W Collapse,”
invited Keynote paper, Proceedings, 7th
International conference on Steel Bridges,
Guimaraes, Portugal, 4-6 June 2008.
[5] BS 5950-1, “Structural Use of Steel Work in
Buildings,” British Standard institution, p. 213,
2000.
ACKNOWLEDGEMENT
The authors wish to thank the South Asian
Institute of Technology and Medicine (SAITM)
for their provision of financial assistance and all
the other facilities for this research. We would
also acknowledge Mr. J.M.W.K Hunukumbura,
Executive Engineer –Kurunegala for facilitating
the authors to find all the necessary information
about the Meeliyadda Bridge failure