Salford Journal of Bridge Engineering Jan-18

Salford Journal of Bridge Engineering
Volume 1, Issue 6, January 2018
Editorial Panel Contents

Jonathan Haynes BEng(Hons) MSc CEng
FIStructE FICE PgC FHEA Learning from Bridge Failures
I-35W Mississippi Bridge Collapse ………………………………… 1

Cook, Bailey, Shazali, Marton, Tate,Osei-Asibey, Chiegboka
Tacoma Narrows Bridge Failure ……………………………………. 5
Bekas, Olivera, Garada, Reed, Robinson, Sousa Deusdado, Torres Martinez
Hyatt Regency Walkway Collapse …………………………………. 11
Math, Farjawi, Ojoko, Kazadi, Rajabi, Yaghobian, Mohammed Ali
The 1846 Dee Rail Bridge, Chester ………………………………… 15
Edwards, Thompson, Fraine, Musafer, Khalel, Nur Azhar
Aims & Scope Barton Lifting Bridge over The Manchester Ship Canal …………… 20
Widlak, Esenel, Sauffisseau, Chowdary, Shah, Zahoor, Soname
Salford Journal of Bridge
Engineering is a peer reviewed
electronic journal which publishes
short academic papers written by
civil engineering students.
Papers must present material of use
to a student of civil engineering and
be of approximately 2000 words
length, excluding figures.
Student authors are engaged in
researching the structural form,
construction methods, cost and
durability problems of a bridge as
part of a taught module assessment.
excellent education since 1896

Salford Journal of Bridge Engineering Volume 1, Issue 6, December 2018 1
I-35W Mississippi River Bridge

Edward Cook, MEng(Hons) Civil Engineering
Martin Bailey, MSc Structural Engineering
Shafiq Shazali, MSc Structural Engineering
Jack Marton, MSc Structural Engineering
Alec Tate, MSc Structural Engineering
Nana Osei-Asibey, MSc Transport Engineering & Planning
Anene Chiegboka, MSc Transport Engineering & Planning
1.0 Abstract deficient and annual assessments of the bridge were carried
The I-35W was a steel truss arch bridge in out starting in 1993.
Minneapolis, USA that spanned the Mississippi River.
This paper documents the conditions that induced the
collapse of the I-35W on 1st August 2007. These
include poor maintenance, insufficient design of
structural elements and changes in the loading Figure 0.2 - Side Elevation (National Transportation Safety
conditions. These compounded and resulted in the Boards, 2008).
death of 13 people. The bridge was designed
specifically for vehicle traffic in downtown The structural form of the bridge was a continuous truss
Minneapolis – allowing the interstate highway to pass arch with three main spans across the Mississippi River, as
over the Mississippi River. shown in Figure 0.2. The deck of the I-35W consisted of
After the collapse, inspections were carried out on all hot rolled asphalt atop reinforced concrete, while the truss
steel truss bridges in the state and the disaster is an arch was constructed from steel. The deck and trusses sit
example of the poor maintenance of the American atop concrete piers which transfer the loading to reinforced
infrastructure network. A replacement bridge was concrete pile foundations. When constructed, the bridge
completed on the interstate within a year. had fixed steel bearings at the pier 7 and the remaining
three piers all had steel roller bearings.
2.0 Introduction It is evident that the roller bearings suffered from
2.1 General Information significant corrosion before the collapse; mainly due to the
Officially known as Bridge 9340, the I-35W proximity of the bridge to water. Moreover, examination of
Mississippi River Bridge was located in Minneapolis. the roller bearings discovered roller wear – meaning the
Construction work started in 1964 and was completed deck had been moving relative to the piers (National
three years later – meaning at the time of collapse the Transportation Safety Boards, 2008). Roller wear is an
bridge had stood for 40 years. The bridge was indication of thermal expansion and contraction which is
designed to American Association of State Highway discussed further in section Error! Reference source not
Officials (AASHO) design codes by Sverdrup and found..
Parcel. The geometry of the bridge was a series of right angled and
scalene triangles – some of which are pitched to give the
impression of a curved arch. With a centre span of 139 m
and external spans of 80 m, after which piers were
arranged to raise the carriageways to the bridge level.
2.2 Collapse
On 1st August 2007, road maintenance on the bridge was
underway and materials were being delivered to the site,
shown in Figure 0.4. Two lanes were closed on the
Figure 0.1 - I-35W (National Transportation Safety Boards, northbound carriageway and two lanes were closed on the
2008). southbound carriageway – to store materials. During rush
hour, traffic was at standstill on the bridge and at
I-35W, shown in Figure 0.1, carried up to 140,000
approximately 6.05 pm, the three main spans across the
vehicles per day across two carriageways of four
Mississippi River experienced catastrophic failure
lanes. Minnesota Department of Transport (MnDOT)
(National Transportation Safety Boards, 2008).
constructed the bridge with contracts worth $5.2
million and completed all necessary inspections.
MnDOT also owned the bridge. In 1990 the
government identified the bridge as structurally
I-35W Mississippi River Bridge Cook, Bailey, Shazali, Marton, Tate,Osei-Asibey, Chiegboka
the least corroded bearings. The other at the time of and

after the collapse was in worse disrepair. Figure 0.9 and
Figure 0.10 show the extent of the roller bearing corrosion
I-35W was suffering from.
Figure 0.3 - I-35W Collapse (National Transportation

Safety Boards, 2008).
This combined with how retrofitting the deck had
increased the dead load acting on the superstructure,
seen between the 1995 and 1997 BARS analyses
indicated a rise in approximately 1.73 kN/m on some Figure 0.6 - Roller Bearing (National Transportation Safety
deck spans (National Transportation Safety Boards, Boards, 2008).
2008).
The National Transportation Safety Board (NTSB)
considered other attributing factors to the collapse, such as
the fracture of the floor truss and movement of the piers
(Cardino, 2009). However, the investigation undertaken by
the NTSB determined the thin gusset plates were the cause
of the failure which was amplified by the rush hour traffic
and construction loading (Hao, 2010).
2.3 Technical Drawings

Figure 0.4 - Plan Loading (National Transportation
Safety Boards, 2008).
A report produced after the collapse determined that
the bridge was insufficiently designed, suffered from a
lack of guidance regarding the placement of
construction loads during repair, and inadequate
technology available for assessing the condition of
gusset plates. The thin gusset plate is shown in Figure
0.5, where it is evidently buckling before the bridge
collapsed. The gusset plate shown is only 12.7 mm
thick, whereas most other gusset plates were designed Figure 0.7 - Gusset Plate Fracture (Barker & Puckett,
to be in excess of 25.4 mm thick. 2013).
Due to the time period that the bridge was designed and
detailed, there are limited digital copies of technical
drawings for the bridge. The gusset plate fracture in Figure
0.7 was only completed post collapse due to its implication
in the collapse. There is a distinct lack of technical
drawings for the I-35W Mississippi River bridge.
3.0 Structural Form

3.1 Aesthetic Assessment
The bridge was an under-deck truss which was painted
Figure 0.5 - Gusset Plate (National Transportation green to provide corrosion resistance for durability and to
Safety Boards, 2008). suit the low-profile of the low-rise building and green
Further investigation showed that the roller bearings, environment. The arrangement of the truss was a tapered
which allowed for thermal expansion of the bridge warren truss, being deepest at the piers and reducing its
deck, were heavily corroded. The roller bearing shown profile at the centre of the spans. The form of which was
in Figure 0.6, taken before the collapse, was one of consistent throughout providing appealing aesthetics;
however, the depth could be considered stocky. Four 3.3 Fatigue and Dynamic Behaviour
slender piers supported the triangulated truss arch and Temperatures in the region ranged from an average low of
raise the deck to the same level as the highway. The -13.6oC in January to an average high of 28.6oC in July
open deck suited its function as a low-profile bridge with the maximum temperatures regularly reaching over
for travel through downtown Minneapolis – with 30oC (CBS Minnesota, 2007). On the day of the collapse,
views from the bridge barely impeded by the minimal the bridge had undergone a temperature change of 11.1oC.
barriers. The proportion of the truss elements gave the Large temperature changes resulted in the bridge
impression of a sturdy bridge while maintaining superstructure expanding and contracting beyond
significant sightlines through the bridge when viewed allowable limits; leading to immediate bending stresses
from an elevation. Abutments were set within a and long-term axial stresses through the deck. To mitigate
concrete spillway area which was not visible from this, roller joints were present and de-icing salts were used
most angles so appeared suitably understated. to combat reportedly dangerous black ice build-up on the
deck. However, due to the poor maintenance of the bridge
3.2 Loading and Analysis Assumptions described before, these two aspects induced corrosion in
Simula created a finite element analysis (FEA) model, the superstructure.
partially shown in Figure 0.8, on behalf of NTSB and
determined that the areas of the thin 12.7 mm gusset 4.0 Critical Review
plates were beyond their yield stress even at initial 4.1 Structural Performance
construction. The inadequate design of the gusset plate thickness
combined with the load increasing over time led to
buckling. This was reported through several assessments
with no remedial action undertaken to stiffen the
connection. Signs of cracking in other truss elements were
present due to the redistribution of the stresses from the
deficient gusset plate; resulting in the bridge being
considered structurally deficient since the 1991 inspection
from the Federal Highway Administration (FHWA). These
combined with the dead and live loads increasing over the
design life caused the bridge to collapse after 40 years.
This would not be close to meeting a modern-day design
Figure 0.8 - Gusset Plate Finite Element Analysis life of 120 years.
(Tomasetti, 2007)
4.2 Durability and Sustainability Issues
It was concluded by Simula that the elastic
Painted to protect the steel against corrosion but due to
components surrounding the gusset plates had
neglected maintenance and no action after failed checks,
prevented large deflections. A replacement deck and
the roller bearings performance was affected. Fatigue
barriers were added in 1977 and 1998, respectively.
cracking was present from changing temperatures – with
Assessments were undertaken due to the dead load
the roller bearings’ reduced performance being the primary
being increased and it was determined the bridge was
cause. One of the inspections in 1993 noted a loss of metal
still sufficient. However, the gusset plates were not
in at least one gusset plate due to corrosion (National
analysed and it was not known they were overstressed
Transportation Safety Boards, 2008). Moreover, one of the
(National Transportation Safety Boards, 2008).
12.7 mm thick gusset plates had lost 37% of its thickness
The I-35W was initially designed to resist imposed
from corrosion over a length of 457.2 mm.
loads, variable loads and an impact load. Heavy trucks
across the span were used to determine a variable load
of which 9% was used to determine the impact load.
On the day of the collapse, the variable loads slightly
exceeded the allowable loading which was adequate
for most members except the thin gusset plates. The
neglected gusset plates had been inadequately
designed and analysed during loading assessments
conducted in 1977 and 1998 – compulsory
assessments when loading changed (National
Transportation Safety Boards, 2008).
Figure 0.9 - Roller Bearing (Retro Report, 2007).
Figure 0.9 shows the roller bearing after the side plate https://www.retroreport.org/video/when-a-bridge-
was removed post collapse, showing the extent of the falls/
corrosion. Tomasetti, T. (2007). I-35W Bridge Collapse. Retrieved
At the time of construction, sustainability strategies from
were not considered paramount as they are nowadays, http://www.thorntontomasetti.com/projects/i35_bridge
thus no significant action or technologies were _collapse/
incorporated into the bridge design.
Figure 0.10 - Roller Bearing (Retro Report, 2007).
5.0 Conclusion
Poor design of gusset plates ultimately resulted in the
collapse of I-35W – while all other structural elements
performed as designed. A lack of thorough inspection
and analysis after bridge modifications, that increased
the loading, resulted in further buckling of gusset
plates. This, combined with neglected maintenance
reports of structural deficiency and corrosion
concerns, led to failure.
The collapse incentivised repair work on all streel
truss arch bridges in the state in addition to
highlighting the disrepair throughout the entire
American infrastructure network.
6.0 References
Barker, R. M., & Puckett, J. A. (2013). Design of
Highway Bridges: An LRFD Approach. New
Jersey; John Wiley and Sons
Cardino, C. A. (2009). Inadequate Gusset Plates
Probable Cause of I-35W Bridge Collapse. Civil
Engineering, 79(1), 17.
CBS Minnesota. (2007). Minnesota Weather Records.
Retrieved from
http://minnesota.cbslocal.com/minnesota-
weather-records/
Hao, S. (2010). I-35W Bridge Collapse. Journal of
Bridge Engineering, 15(5), 608-614.
National Transportation Safety Boards. (2008).
Highway Accident Report: Collapse of I-35W
Highway Bridge. Retrieved from
https://www.ntsb.gov/investigations/AccidentRep
orts/Reports/HAR0803.pdf
Retro Report. (2007). When a Bridge Falls.
Retrieved from
Tacoma Narrows Bridge Failure

Ioannis Bekas, MSc Structural Engineering
Omar Garada, MSc Structural Engineering
Aitor Oliveras Ibanez, MSc Structural Engineering
Alysia Reed, MEng(Hons) Civil & Architectural Engineering
Isabelle Robinson, MSc Structural Engineering
Barbara Sousa Deusdado, MEng Civil & Architectural Engineering
Cristobal Torres Martinez, MSc Structural Engineering
1.0 Abstract insufficient to withstand the dynamic forces acting on

On the 7th of November 1940, with a wind of 64km/h the structure.
(40mph) the first Tacoma Narrows bridge started to 3.1 Loading and Analysis Assumptions
twist and vibrate. The bridge was closed to traffic and Self-weight, permanent loads, imposed loads, thermal
one hour later the oscillation was a torsional and wind loads were applied as a static pressure, and
movement of double waves with both sides not in the bridge dimensions were designed accordingly. It
synchrony. This led to the failure of the central span is now known that torsional effects were not
and the connecting cables. Despite the false media considered during the design process, therefore
coverage, vertical resonance was not the cause of second order effects caused by wind force induced
collapse. Incorrect design assumptions resulted in twisting on the bridge deck and ultimately, failure.
unsuitable dimensions for the deck and the bridge’s In theory, the cables would be sufficiently stiff to
main span. absorb the energy from the wind pressure, if
2.0 Introduction aerodynamic forces were only taken into
Tacoma Narrows Bridge was a suspension bridge consideration as vertical actions. The analysis was
situated in Washington, USA. It was designed to limited to how the cables would lengthen and change
carry road traffic over the Tacoma Narrows Strait and shape, regardless of the lack of deck rigidity.
was to be the third longest suspension bridge in the Although torsional resistance is now seen as
world. Instead, just 4 months after it opened on July fundamental analysis for long spanning bridges, for
1st, 1940, it became known as the world’s most early 20th century structural engineers this concept
famous bridge failure. was still not developed.
The Structural Engineer for this bridge was Leon 3.2 Structural Form and Assessment
Moisseif whose work included the Manhattan Bridge The final design chosen for the Tacoma Narrows
in New York, the Bay Bridge as well as being a project was heavily influenced by consultant Leon
consultant for the Golden Gate. The Chief Engineer Moisseif, who had previously worked on the Golden
was Lacey Murrow. Both engineers had long Gate bridge and would bring his reputation to the
experience in the design of suspension bridges. development.
Costing approximately $8 million to build, Tacoma In addition, his design was chosen because he
Narrows was intended to be the ‘little brother’ of the proposed a change of the truss to a 2.4-meter-high
Golden Gate bridge. Tacoma Narrows had two plate girder, seen in Figure 0.11 and Figure 0.12,
towers that carried a central span on 853m. The deck lowering the cost to 8 million dollars from the
had a width of 12.5m and a depth of 2.4m and was previous design of 11 million dollars. The main issue
built using two longitudinal steel girders and that was caused by this cheaper design is that it
transverse ties. The girders used the full web without decreased its torsional resistance and increased wind
clearance holes. Although it was aesthetically loading.
pleasing, with a slenderness span/depth ratio of This deck design had a 1:72 depth-to-width ratio with
1/350, serviceability problems were apparent even a slenderness (span-to-depth) ratio of 1:350.
during the construction phase. The structure was
quickly given the nickname ‘Galloping Gertie’ by the
site labour when construction started on September
27th1938.
This paper explores the reasons behind the Tacoma
Narrows bridge collapse and the events leading up to
it.
3.0 Structural Form and Construction
The failure of this bridge was a product of incorrect
analysis and assumptions during the design stage.
Consequently, the deck dimensions used were
Tacoma Narrows Bridge Failure Bekas, Olivera, Garada, Reed, Robinson, Sousa Deusdado, Torres Martinez
3.3 Construction
During its construction phase problems emerged, the
deck was showing a lack of resistance to vertical
forces which alarmed the engineers. There were
several attempts to reduce the bridge’s movement;
adding tie-down cables anchoring the plate girders to
45-tonnes of concrete blocks, and hydraulic buffers to
dampen the longitudinal motion in the main span
from the wind acting sideways.
These solutions were ineffective, because they did
not eliminate the unknown primary issue at hand. The
Figure 0.11 - Cross-section of Bridge Deck (Aguilo engineers of the project hired a professor from
Alonso, 2010) Washington University, Frederick Farquharson to
investigate further. He modelled the bridge in 2D and
3D in a wind tunnel study and noticed a lack or
torsional resistance. This was the first-time torsion
was considered to be an issue in the Tacoma Narrows
bridge. The finished bridge is shown below in Figure
0.14
Figure 0.14 - Tacoma Narrows Bridge (HERNANDEZ,

Figure 0.12 - Plate Girder, (WSDOT - Tacoma Narrows NIETO & MOSQUERA, 2011)
Bridge, 1940) 4.0 Structural Performance
The general arrangement of the Tacoma Narrows On November 7th, 1940, professor Farquharson
bridge and its replacement are shown in Figure 0.13. observed the twisting phenomenon that he studied.
With a wind of 64km/h (40mph) the bridge started to
twist and oscillate with a 0.90m amplitude and with
36 cycles per minute for three hours. Soon after,
traffic was diverted away from the bridge.
Oscillations caused a torsional movement with the
main span in two segments, with the node in the span
at midpoint, where both sides of the deck are out of
phase, seen in Figure 0.15.
Figure 0.13 - General Arrangements (WSDOT - Tacoma

Narrows Bridge, 1940)
5.0 Critical Review

The failure of the bridge was a direct result of several
engineering errors that combined to create the
unexpected motion.
Torsion was never assumed to be an issue because, in
theory, the cables would absorb the static load
pressures of the wind. However, wind forces on the
deck acted dynamically and investigations proved
that there was a lack of stiffness that added the
unexpected vertical oscillations. Not only the slim
deck, but also the long spans create a slender
structure, more liable to have low torsional
Figure 0.15 - Oscillations During Bridge Failure resistance.
(HERNANDEZ, NIETO & MOSQUERA, 2011) The next issue that caused the bridge to collapse were
The amplitude of the vertical displacement increased the cables. Even though the cables were over
from 0.90m to 8.40m (changing to 20 cycles per designed for 2.45kN/m2, they were not flexible
minute) so the deck was twisting 35 to 45 degrees enough to compensate for the extreme flexibility of
along the main axis in a sequence of 20 cycles per the deck. When one of the main cables in the centre
minute. span of the bridge slipped, it created two unequal
The steel cables snapped around 11a.m and 183m of segments that contributed to the change from vertical
bridge deck collapsed into the water followed by the into torsional forces. Figure 0.18 locates the elements
centre span, leaving only the two towers standing, of the bridge.
show in Figure 0.16 and Figure 0.17.
The New York Times was the first of many to falsely
report that the reason for the Tacoma bridge failing
was resonance caused by wind, when it was the lack
of rigidity of the deck that allowed for an in-phase
movement with the wind.
Figure 0.18 - Tacoma Narrows Bridge Labelled

(Warraich, 2015)
However, although the bridge was designed to resist
a 2.45kN/m2 static wind pressure, on the day of
collapse, dynamic loadings induced large
Figure 0.16 - Centre- Span Collapse (HERNANDEZ, displacements in the deck, despite a wind speed of
NIETO & MOSQUERA, 2011)
0.25kN/m2.
Second order effects played an important part that
day. As seen in Figure 0.19 while the deck moved
due to wind, it shifts the weight of the deck,
concentrating it to the edge, and therefore increasing
its deflection and oscillations.
Figure 0.17 - The Two Towers and End-Spans Left

(HERNANDEZ, NIETO & MOSQUERA, 2011)
structure. Figure 0.22, shows these wind lapses being

separated and creating vortexes, vortex shedding.
Finally, Figure 0.23, shows the immense amplitude of
the oscillations as the bridge moves in phase with the
wind vortexes. By this point the shedding vortex
phenomenon becomes torsional fluttering, a self-
induced harmonic vibration pattern.
Figure 0.19 - Second Order Effects on the Deck

The most severe consequence of the lack of torsional
resistance, is that it became incredibly easy for wind
to act upon the deck and created an increasing
oscillation. Figure 0.20 shows how the deck started to
oscillate due to the wind forces.
However, this figure demonstrates the out of plane
behaviour of the deck, while ignoring the in-plane
behaviour. During the bridge failure, the deck was
not rigid and deflected across its width. Figure 0.20
shows the deck stiff throughout its motion, but its
deflections was a contributor to the out of phase wave Figure 0.21 - Wind Vibrating the Deck (PASTERNACK,
on both sides of the deck. 2015)
Figure 0.22 - Wind starts to Oscillate the Deck and Create

Wind Vortexes (PASTERNACK, 2015)
Figure 0.20 - Wind Acting Upon the Deck (PASTERNACK,

2015)
As the deck was twisting, the wind flow separated,
forming wind vortexes. These further lifted the deck.
Figure 0.21 shows the wind is starting to vibrate the
Figure 0.23 - Deck is in phase with wind vortexes

(PASTERNACK, 2015)
Torsional fluttering is the reason why the deck
collapsed, the torsional motion to build upon its own
self-induced energy. However, it was not only due to
the large oscillations, but also because the two sides
of the deck were moving out of phase, as shown in
Figure 0.24, due to the deflection across the width.
Figure 0.25 shows how some of the components
failed. On the left, the plate girder’s beams have
clearly buckled and on the right, there are parts of the
snapped cables and the sides dislocated from the
main deck.
Figure 0.25 - Components of the Failed Bridge

6.0 The Aftermath
Due to the collapse of the first bridge a second one
was commissioned. This was a long process that
required investigation of the site and its structure to
correctly locate the problem. Testing of the new
bridge design was appointed to Professor
Farquharson, who confirmed the new bridge could
stand up to 127mph of wind. The replacement bridge,
Tacoma Narrows II, was finished in 1950, ten years
after the collapse of the first one and cost a total of
$14,011,384. Financing the bridge proved to be more
difficult. After the second world war, construction
projects were given funding based on priority, and
Figure 0.24- Sides of the Deck Oscillating Out of Phase steel was in short supply. In addition to this, the State
(Blaschke, 2015) found it challenging to find insurance for the bridge.
Figure 0.26 shows a clear change in the deck’s
thickness. The previous deck was 2.4m wide
compared to the 10m one in the new design. They
also implemented the trusses that were not previously
included in the first bridge due to funding.
Construction began on the bridge in 1948 and took 29
months to complete, despite the challenges that came
from having one of the coldest winters on record. The
bridge was given the new nickname ‘Sturdy Gertie’
and was thought to be a milestone in aerodynamic
bridge engineering.
University. Txstate.edu.
http://www.txstate.edu/news/news_releases/news_
archive/2015/November-
2015/TacomaNarrows110315.html
Hernandez, S., Nieto, F., & Mosquera, A.
(2011). Bridge Aeroelasticity. Coruna, Spain: WIT
Press.
Pasternack, A. (2015). The Strangest, Most
Spectacular Bridge Collapse (And How We Got It
Wrong). Motherboard.
https://motherboard.vice.com/en_us/article/kb78w
3/themyth-of-galloping-gertie
Voss, D. (2016). November 7, 1940: Collapse of the
Tacoma Narrows Bridge. This Month In Physics
Figure 0.26 - Deck Cross-section of Tacoma Narrow II History, 25(10).
(Aguilo Alonso, 2010)
Warraich, M. (2015). Tacoma Narrows Suspension
Bridge. Slideshare.net.
7.0 Conclusion https://www.slideshare.net/asadwarraichc/tacoma-
The Tacoma Narrows collapse revolutionized narrows-suspension-bridge
aerodynamic analysis and its effects in bridge design. WSDOT - Tacoma Narrows Bridge: Extreme
Vortex shedding and torsional fluttering became key History. Wsdot.wa.gov. Retrieved from
considerations in suspension bridge design. https://www.wsdot.wa.gov/TNBhistory/
Speculations of a resonance problem are widely Young, D. (2013). Exit133 - 73 Years Ago Today:
reported in media still today, however is has been Epic Fail of Tacoma Narrows
proven that the deck’s design was too slender to Bridge. Exit133.com.
withstand the unidentified dynamic loads. http://www.exit133.com/articles/view/73-years-ago-
epic-fail-of-tacoma-narrows-
8.0 References bridge#.Wnx5GducaAw
Aguilo Alonso, M. (2010). Forma y tipo en el arte de
construir puentes. Madrid: Abada
Blaschke, J. (2015). 75 years after infamous Tacoma
Narrows Bridge collapse: Not so fast! : Office of
Media Relations : Texas State
Hyatt Regency Walkway Collapse

Haty Math, MSc Structural Engineering
Saif Farjawi, MSc Structural Engineering
Hilda Ojoko, MSc Structural Engineering
Clinton Kazadi, MSc Structural Engineering
Masoud Rajabi, MSc Structural Engineering
Masoud Yaghobian, MSc Structural Engineering
Hemen Mohammad Ali, MSc Structural Engineering
 1978 – Construction began and in December with

1.0 Abstract Eldridge Construction Company as the lead contractor
In 1981, two suspended concrete walkways designed and Haven Steel Company was subcontracted to
directly under each other collapsed at the Hyatt fabricate and erect the steelwork atrium.
Regency Hotel in Kansas City, USA. The collapse  1979 (Jan) – Havens proposed to change Drawing 30
was caused by the change from a one rod supporting and Erection E-3 which involves altering the
system to two-rods system, altering the load path. continuous rod hanger to two separate rods ending and
Failure to check and redesign the connection of the starting at level 4. The reason for alteration was due to
new two-rod configuration between box beam and the long length of continuous rod would possibly be
supporting rod of the 4th floor (higher level) walkway damaged during bridge deck lifting.
lead to excessive load on the connection and collapse  1979 (February) – G.C.E reviewed 42 drawings
of the walkways. This was due to the negligence and including the changed Drawing 30 and erection E-3,
incompetence of the responsible engineer. The and stamped for approval.
catastrophic collapse was the deadliest structural  1980 – Hotel opened for business and one year later,
failure in the U.S. at the time where 114 people lost collapse of level 4 and level 2 bridges occurred during
their lives and 200 more people were injured. a Tea Dance event which hosted 1600 people killing
114 and leaving over 200 people injured (The
2.0 Introduction Engineer, 2006).
In Kansas City, a tea dance party took place at the
new Hyatt Regency Hotel on the 17th July 1981. Due 3.0 Structural Form
to the large number of guests attending the event, the The walkways consisted of RC decks loaded on cross I-
Beams supported by box beams (Figure 0-1). A hanger
entrance/reception area was filled with guests as well
rod connected at the roof truss holds the box beams in
as the three suspended concrete walkways on 2nd, 3rd place.
and 4th levels which were used to connect both sides
of the previous levels. The walkway bridges were of
an equal dimension of 2.3m wide x 36.58m long, the
length of the bridges were split into four equal spans
of 9.145m. The 2nd and 4th walkways were hung
directly on top of each other, also a 30mm diameter
rod was designed to connect the walkway to the roof
truss (Randall, P. and Bernhardt, P.E. 2016).
During the tea party, whilst crowded with guests the
2nd and 4th walkways collapsed onto the lobby packed
with guests causing a horrific tragedy.
2.1 Brief Project Timeline

 1976 – Owner, Crown Centre Redevelopment
Corporation commence the project of Hyatt
Regency Hotel Kansas City and Gillum-Colaco
Inc. was chosen as consulting engineer for the
project.
 1977 (July) – Project sub-contracted to G.C.E Figure 0-1 – Final Section of Box Beams (Marshall R.D et al.,
International, Inc. and along with architect 1982)
PBNDML Architects were selected.
Hyatt Regency Walkway Collapse Math, Farjawi, Ojoko, Kazadi, Rajabi, Yaghobian, Mohammed Ali
3.1 Loading and analysis assumptions other. However, the load increased at the 4th floor level
The walkways of concern were those of 2nd and 4th when the fabricator’s changed the rod arrangement to two
floors which lie directly above each other. The deck rod configuration causing the fourth floor beam taking load
was analysed and designed as a simply supported slab of second floor deck as well as its own deck loading
(Marshall, R.D., et.al, 1982.) producing a total load of 180 kN (The Engineer, 2006).
Initially a single rod was designed to support both
walkway decks bolted at each level (Figure 0-2). 3.3 Aesthetics
Therefore, the load path of each deck would be The architectural design objective for the Hyatt Regency
independent of each other. hotel bridges was to connect each side of the hotel with a
glass railing bridge to allow for clean, unobstructed views
to the lobby and as much transparency as possible. Also
the glass in general is a great material for its ability to
blend with the surrounding material used for construction
of the bridge. The Hyatt Regency bridge was a signature
design element in 1981.
In terms of integration, the bridge was very well integrated
within the hotel environment, the concrete deck was
constructed as one continuous slab that connect both sides
of the building beautifully, also the glass railing
transparency made the bridge look like it is within one
open big space, especially with the views to the ground
floor lobby. The colours of the finished slab are in contrast
with the surrounding colours of the concrete structure
which clearly indicates that the simplicity used in terms of
Figure 0-2 - Initial Designed Single Rod Configuration colours choice and material made the bridge look
(Marshall R.D et al., 1982) stunningly elegant.
The as – built design was altered from a single rod to In terms of texture, the Hyatt regency hotel bridges were
a two-rods system. At the 4th floor level, a second rod mainly constructed of welded steel which gives the bridge
was introduced where the first rod ends at an offset a smooth surface texture. Finally, the aesthetic summary of
100mm inward along the axis of the box beam (Figure the bridge could be concluded as the structure and function
0-3). Consequently, the load path has changed from of the Hyatt Regency bridge are represented in a simple
independent between decks to the 4th floor beam and obvious way which adds to the bridge aesthetic appeal.
supporting the 2nd floor deck. Due to the offset, the
effect of eccentricity produced extra moment in the 3.4 Sustainability
box beam that lead to failure. In terms of sustainability the lightweight concrete provides
many advantages over normal weight concrete, the reduced
dead load is significant in which both the client and
contractor could save plenty of time and money on less
construction material and formwork used.
The fabrication of the box beam has greater costs than
normal steel beam as there are more welding and coating
operations involved. Also, risks could be an issue
especially during operations with the box beam within
surrounded spaces.
Just like steel, the glass is a sustainable material, however
it is very costly and especially when the Hyatt Regency
hotel was built, there were less glass fabricators which
indicates the cost of the using such fancy materials is very
high. Yet again, the glass is a feasible option for this type
of walkway as it does produce less carbon emissions when
compared against all other walkway railing systems except
Figure 0-3 - As built hanger rods configuration (Marshall for timber.
R.D et al., 1982)
Initial design loads were specified as 90 kN where the
deck loadings at each level are independent of each
4.0 Structural Performance

The collapse of the Hyatt Regency walkway clearly
indicates the walkway performed poorly, however by
examining the circumstances that led to the horrific
collapse, it can be concluded that the design errors and
modifications would certainly lead to failure
eventuality.
The initial bolted box beam shown in (Figure 0-2)
meant to carry the weight of the 4th floor walkway
only, however as connections were modified by the
fabricator, the same bolted box beam is carrying the
load of the 2nd floor walkway too, which resulted in
the bolt carrying twice the load as originally designed Figure 0-2 - Post failure box beam (Randall, P. and Bernhardt,
(Figure 0-1). Without any modification in the design P.E., 2016)
of the bolts, in effect, the design modification resulted
in new load path which presented a compounding Finally, assuming the connection weren’t modified and
shear stress on the box beam. placed in the correct positions as stated in the design
(Figure 0-2) the connections would still perform poorly
and fail as the investigation data clearly states the load
capacity of the connections that were constructed were
only 60% below the required load capacity for this type of
connection by Kansas City Building Code’s (The Engineer,
2006).
41 Material & Durability
The walkways in the Hyatt Regency hotel were constructed
using lightweight concrete, illuminated glass railing and
steel box beam and connection.
The lightweight concrete used for the walkway deck is
located inside the building, thus it is not exposed to
weather conditions. Moreover, deterioration is not likely to
happen as the concrete is not exposed to water. Illuminated
glass railings that were used for the walkway are very
durable and it is resilient to moisture and weather
conditions. Also, size of the glass will affect the durability
Figure 0-1 - Elevation of as-built configuration and roof of it, as the larger glass railing used will increase the
support probability of finding critical flaw in it.
During the erection of the walkways, a noticeable Typical box steel beam that was used for the walkway is a
deflection in the box beam was observed by the site common feature for footbridges, the weathering steel used
personnel, however no follow up checks were carried for the project provide improved durability as weathering
although the deflection concern was reported to site steel prevent the need for corrosion protection. Durability
engineers (The Engineer, 2006). of the steel elements were not significant and they did not
National Bureau of Standards (NBS) carried an contribute to the failure of the beam nor the steel
investigation to determine the causes of the failure, connections.
the main cause of the Hyatt Regency walkways It could be concluded that the durability of the structural
collapse was a failure of the box beam carrying 4th elements used for the Hyatt Regency walkways does not
floor walkway. The bolts are placed directly through have any significant impact on the collapse of the
the welded joint connecting two c-channels, the walkways. Also taking into consideration the fact that
weakest structural point in the box beam. Thus, the walkways were only in operation for one year before the
failure initiated at the support rod connection where incident, thus the materials are fairly new and nowhere
the beam torn apart along the weld. As a result, the close to their design life.
nut and washer on the rod connection slid through the 5.0 Critical Review
box beam (Figure 0-2). Also, the box beam The communications between both the consultants and
experienced further yielding due to the rupture of the contractor were ineffective, inconsistent and
beam weld (Marshall, R.D., et.al, 1982.). unprofessional. Since the project was pushed to finish as
quickly as possible, changes to the design of the support
structure of the walkway were undertaken by the fabricator
and verbally reported to the consultant engineer which revoked of its engineering license and several firms went
later on approved on paper (Slayton, E. 2007). As a bankrupt (The Engineer, 2006).
result, a catastrophic structural collapse occurred by 7.0 References
downgrading the structural integrity of the structural Department of Philosophy and Department of Mechanical
elements. During the construction and erection of the Engineering. (2009) Negligence And The Professional
walkway structural elements, site personnel observed "Debate" Over Responsibility For Design. [Online]
a bending in the box beam due to the doubled load Available from: http://ethics.tamu.edu/wp-
applied on it and the bending concern was reported to content/uploads/sites/7/2017/04/HyattRegency.pdf .
the site engineers, however no follow up inspection McFadden, C. (2017) Understanding the Tragic Hyatt
were carried out at all. This shows the lack of Regency Walkway Collapse. ThinkReliability. [Online]
communication and how regular check-ups are Available from:
important. Also from this, it should be noted that the https://interestingengineering.com/understanding-hyatt-
consultant should be attentive to the contractors since regency-walkway-collapse
they are the one with hands on. NASA. (2005). Single rods vs modified two rods. [Online
Insufficient design verification: despite the fact that Image], Available from:
the design of the connection was changed by the https://sma.nasa.gov/docs/default-source/safety-
fabricator, however even if the original design was messages/safetymessage-2008-05-01-
constructed the original designs did not meet the hyattregencywalkwaycollapse.pdf?sfvrsn=4
Kansas City Building Code, based on the AISC Pilkington Ltd. (2017) Durability of Glass. [Online]
specifications. This issue clearly indicates that the Available from: https://www.pilkington.com/en-
ulterior motive such as saving time or costs or simply gb/uk/architects/glass-information/functions-of-
neglecting full checks can lead to severe glass/mechanicalfunctionsofglass/durability
consequences. Again, questioning of professional Randall, P. and Bernhardt, P.E. (2016) Hyatt Regency
standards in the industry. Obviously public safety Skywalk Collapse Remembered. Journal of Structural
should be first and foremost and this was neglected Failures. [Online] Available from:
which questions the engineer’s ethical judgment. It is http://www.structuremag.org/?p=10274
a serious issue when a complete design is passed R. D.Marshall, E. O.Pfrang, E. V.Leyendecker, K.
without being code compliant. A.Woodward. (May 1982). Investigation of the Kansas
Responsibility: After the incident, the consultant City Hyatt Regency Walkways Collapse: final report.
engineer stated that the structural engineers do not Available from:
have to produce any connection design as it is the http://ws680.nist.gov/publication/get_pdf.cfm?pub_id=90
fabricator’s responsibility to design steel to steel 8286
connections and the drawings provided by the Slayton, E. (2007) The Hyatt Regency Walkway Collapse.
structural engineers were only conceptual. As a result [Online] Available from: https://www.asce.org/question-
of this misunderstanding and ineffective of-ethics-articles/jan-2007/
communication the walkways collapsed. After the Steel construction. (2018) Box girder bridges. [Online]
incident, the American Society of Civil Engineering Available from:
(ASCE) produced a report which states that the https://www.steelconstruction.info/Box_girder_bridges
structural engineers have the full responsibility of the The Concrete Centre. (2015). Lightweight concrete.
design aspects (The Engineer, 2006). [Online] Available from:
6.0 Conclusion https://www.concretecentre.com/Performance-
In summary, it can be concluded that the Hyatt Sustainability-(1)/Special-Concrete/lightweight-
Regency Walkway failed as a result of negligence and concrete.aspx
inadequacy in design by the responsible structural The Engineer (2006). Hyatt Regency Walkway Collapse.
engineer. The main cause of failure was due to the Engineering.com. [Online] Available
change in design from one continuous hanging rod to from:https://www.engineering.com/Library/ArticlesPage/
two separate rods at different floor levels which in tabid/85/articleType/ArticleView/ArticleID/175/PageID/
effect changed the load path. Without the redesign of 199/Default.aspx
the connection bolts, the double load in the design
alteration caused the tragic incident on 17th of July
1981 which took the lives of 114 people and left over
200 injured.
In turn, two structural engineers, Danial M. Duncan
and Jack D. Gillum lost their engineering license in
the state of Missouri and Texas. G.C.E has been
The 1846 Dee Rail Bridge, Chester

Phillip Edwards, MSc Structural Engineering
Liam Thompson, MSc Structural Engineering
William Fraine, MSc Structural Engineering
Dhanish Musafer, MSc Structural Engineering
Hammad Khalel, MSc Structural Engineering
Shamsul Nur Azhar, MSc Structural Engineering
1.0 Abstract
In the midst of ‘railway mania’, six months after
opening, Dee bridge suddenly suffered a partial
collapse as a train was travelling over it; this resulted
in five fatalities and in conjunction with other failures
of cast iron structures led to a severe lack of
confidence in such structures. This paper looks at the
construction and collapse of this structure.
2.0 Introduction
Dee Bridge was a cast iron/wrought iron underbridge
constructed by Robert Stephenson on behalf of the Figure 2: Location of structure, Grid Ref: SJ 39649 65890
Chester to Holyhead Railway Company in 1846. It 3.0 Construction
carried two tracks of the aforementioned line over the 3.1 Introduction
River Dee in Chester, Cheshire. Figure 1 shows an The bridge was initially to be constructed with five
illustration of the collapsed bridge, Figure 2 shows its masonry arches spanning between masonry piers, with
location and Table 1 lists some salient facts about its the scheme getting as far as commencing the piling
construction. for this arrangement (Walker & Captain Simmons,
1847). However, there were doubts over the suitability
of the ground to support a heavy masonry structure
and there were also objections from river users to the
proposed reduced navigable width of the river, hence
the design was amended.
The final design was to use flat cast iron ‘compound
trussed’ girders. This had the advantage over masonry
arches of being significantly lighter, reducing the
vertical load on the foundations and eliminating the
Figure 1. View of disaster site in etching from Illustrated horizontal forces that an arch would transmit to the
London News of 12 June 1847 (Peter R. Lewis & Gagg,
substructures. This reduction in loading allowed the
2004)
Table 1: List of bridge details
number of spans to be reduced to three, increasing the
resultant navigable width of the river. The increased
Name Dee Bridge span was significantly larger than had ever been
Location Chester, Cheshire. attempted with this form of construction (Petroski,
Date of Construction 1846 1994).
Date of demolishment 1871 Compound trussed beams were originally developed
Owner Chester and Holyhead Railway in France (NASH, 1844) and were used often by
Contractor Robert Stephenson
Robert Stephenson for overbridges in the 1840’s. On
Cost Unknown
completion of a similar structure, the 86ft span
No. of spans 3
Type of structure Compound Trussed Beams
Stockton bridge, he described it as ‘one of the finest
Clear Span 98ft (29.8m) specimens of architecture of this description in the
Skew 51° north of England, or perhaps in the known world’ (P.
Carried Up and down lines of the R. Lewis, 2007) and this bridge would have been
Chester and Holyhead Railway aesthetically pleasing with its curved ironwork and
Spanned Over River Dee slender profile.
Grid Reference SJ 39649 65890 2.2 Substructures and Foundations
The substructure consisted of two masonry abutments
and two masonry piers. These were faced in Runcorn
The 1846 Dee Rail Bridge, Chester Edwards, Thompson, Fraine, Musafer, Khalel, Nur Azhar
sandstone with the bed-stones being in the same

material. The piers and abutments were built upon
timber piles. b.
3.2 Superstructure
The superstructure consisted of two pairs of a.
compound beams, each pair carrying one track each.
This arrangement was repeated to each identical span,
using twelve beams in total. The distance between two
beams, forming a pair, was 12 ft. Although the two
pairs were adjacent to each other, they were not
physically connected in any way, and each beam had Figure 4. Part elevation and cross section of beam end
no continuity over the piers. The beams simply sat on ("Railway Details of Girders Over the River Dee," 1846)
the sandstone bed-stones. Two beams of a deck were connected transversely by
Each compound beam was 98ft long and comprised of a wrought iron tie bar. Oak joists 10” deep spanned
three equally sized cast iron beams spliced together, transversely between a pair of beams, resting on the
these lengths being the limit for casting. The sections inside bottom flanges. On top of this was longitudinal
were connected by bolts which passed through cast-in timber planking, 4” deep. The cast iron chairs were
end plates and were strengthened by semi-circular fixed on the planking and carried two rails each. One
castings (as shown as ‘a’ in Figure 3) which were rail was a running rail while the inside rail was a
bolted to the top flange of the beam overlapping the guard rail, designed to pick up a wheel in the event of
connection. a derailment.
a.
b.
Figure 5. Section through deck (Walker & Captain

Simmons, 1847)
Figure 3. Part elevation and cross section of typical beam at
3.3 Construction Sequence
splice position ('Railway Details of Girders Over the River A full detailed construction sequence is not recorded,
Dee,' 1846) however from similar structures constructed around
Additionally, there were quadrant shaped castings that time an indicative construction method can be
again bolted to the top flange at the end of the beams determined:
(a, Figure 4), from which wrought iron tension bars  Install cofferdams, using wrought iron tubing or
(b, Figure 4) were attached. The tension bars which timber sheet piles.
consisted of eight plates, were attached at this location  Pump out water from foundation using steam
with bolts passing through eye sockets at the end of driven pumps
the tension bars and the aforementioned casting. The  Drive timber piles through riverbed silt to clay.
bars were then inclined where they ended at the Piles were likely driven using manual labour or
bottom flange of the connection between castings (36’ horsepower using a pile derrick.
9” from the ends), with a bolt passing though eye  Construct masonry pier and abutments off piled
sockets at the end of the bars and a hole in the beam. foundation
Through these bolts an identical number of tension  Cast ironwork in foundry and deliver direct to site
bars were located horizontally at the bottom flange along partially constructed railway
fixed in the same way. As the holes in the girder were  Assemble compound beams and tension bars. All
bigger than the bolts, for fitting tolerances, a series of cast iron sections bolted together, tension bars
screws were incorporated in order to tension the rods. attached with bolts and tensioned where possible
The rods were heated before installation in order to using screws at back of eye socket.
incorporate some pre-tension.  Lift individual beams using block and tackle on a
trestle from a high level, on the opposite side of the
span.
 Place transverse joists onto bottom inner flanges of passed over the structure causing one of the wrought
main beams by hand iron tie bars to be fractured by impact. Eyewitnesses
 Fix longitudinal planking in place over transverse stated that the train had not derailed and that the
joists. collapse started with a fracture propagating from the
 Install permanent way directly to longitudinal bottom flange at midspan, with excessive deflections
planking before the collapse. The coroner and jury absolved
The structure was opened to freight traffic in 1846, Stephenson of any negligence however it was
however just before the bridge was due to open to requested that a government inquiry was set up in
passenger traffic it was noted that a beam had cracked order to investigate the safety of iron bridges.
near a joint (P. R. Lewis, 2007). This was assumed by
Stephenson to be a casting defect and the beam was
temporally propped with piles before a new section
was cast and installed.
3.4 Load path
Rail loading was transferred through the longitudinal Figure 6. Sketch of girder showing location of fractures.
(Walker & Captain Simmons, 1847)
planks, through the joists, to the bottom flange of the
The Royal Commission appointed Captain Simmons
compound beams. The beams carried the load by
of the Railway Inspectorate to investigate the incident.
bending, whilst additional support was provided by
Simmons, assisted by civil engineer James Walker
tension in the tie rods to transfer vertical loading to
produced a report, which was delivered on 30th June
the piers and abutments. Horizontal loading was
1847.
considered to be absorbed into the substructures via
Simmons visited the structure and ran several
friction between the compound sections and bed-
experiments. When passing a load over the remaining
stones.
structure, it was noted that as the beam deflected, the
Small horizontal loads at deck level would be shared
ends of the beam rotated considerably. This had the
by tie bars between beams, and transmitted to the
effect of reducing the distance between the end
substructures by friction between the bearing plate
supports of the tension bars which in turn reduced the
and masonry. The resultant horizontal force would be
tension in the rods negating their purpose.
insignificant compared to the self-weight of the piers.
It was noted that the beams noticeably rotated inwards
4.0 Collapse under load, an inward movement of 2” to the top
The line was regularly worked by traffic for six flange was recorded under a 48 ton loading with the
months after it was opened. Although the Chester to beam being lively under locomotive loading.
Holyhead line was not in operation the Shrewsbury The inquiry concluded:
and Chester Railway (S&CR) shared the use of this  The bridge could support the loading if the beams
structure and a small length of line either side. On 24th and tension bars did act together, carrying an equal
May 1847, six trains had passed over the structure proportion of strain,
without any incident, before a local contractor (under  There was difficulty in ensuring joint action,
Stephenson’s instruction) laid 5” of red sandstone  The beams and tension bars, when taken separately
ballast over the planking. This was due to the fears of could not safely carry the load therefore the cast
cinders from the engine igniting the bridge, which had iron should have been designed to carry load on its
been the cause of a recent bridge collapse elsewhere. own,
After the ballast had been placed, a train departed  The curvature of the track, solar gain, dynamic
Chester station and passed over the structure on the impact and loading the beams eccentrically on the
Southern line. bottom flanges should have all been considered
The train in total weighed around 60 tons with the when determining the capacity of the structure.
engine and tender weight of 25 tons spread over seven 6.0 Remaining use of bridge
axles (Hayward, 2014). The engine passed over the
first two spans and was at around midspan of the third After the collapse the structure was used for the
span when the southern beam gave way. There were construction of the remaining line. However engines
25 passengers on the train four of which were killed, were forbidden from traversing the structure in the
along with the fireman who had been thrown violently short term, with a stationary engine hauling goods and
to the bank. carriages across the structure. To enable the line to
5.0 Inquiry fully reopen the structure was propped at third points.
As a more permanent solution shallow arches were
An inquest began ten days after the collapse. Stephenson, constructed under the girders to support them along
backed up by several eminent engineers was adamant their full length.
that the collapse was caused by the train derailing as it
In 1871 the structure was completely replaced by a

three span wrought iron, lattice half-through truss
bridge. In 1902 an additional steel lattice half-through
superstructures was constructed to the south of the
existing deck. In the late 1970s the wrought iron
structure was taken out of use due to an assessment
failure of the cross girder in fatigue and is now in very
poor condition. The steel structure now carries the up
and down lines of the Chester South Jn. - Holyhead Figure 9. Cross section of beam showing ‘cavetto’
moulding pattern (P. R. Lewis, 2007)
(CNH3) line.
8.0 Critical Review
It is clear that the collapse of this structure was due to
a limited understanding of the structural behaviour of
the bridge. The beam suffered a brittle fracture under
live loading, partially due an increase in dead loading.
While the collapse mechanism is not clear it is likely
that the principal cause was as the beams bent, tension
was released in the tie roads meaning that the beam
took the full load and failed under lateral torsional
buckling, rather than tensile failure that would usually
Figure 7. Bird’s eyes view of current structure. govern shorter spans.
9.0 Conclusion
The Dee Bridge was a structure installed at the edge
of the available technology operating at an unproven
scale. Calculations carried out on the safe capacity of
the structure where unreliable and fatigue influence on
the capacity of cast iron was not understood. The
calculations carried out were more suitable for smaller
spans and did not accommodate lateral torsional
buckling. Stephenson did not take heed of previous
Figure 8. Southern Elevation of current structure. warnings over this type of construction and hence
7.0 Contemporary Studies further use of this type of construction was
discontinued, with new cast iron rail bridges being
Surprisingly some later references to this incident banned in 1883 (Hayward, 2014).
state that the collapse was caused by a train 10.0 References
derailment seemingly taking Stephenson’s inquiry
testimony at face value (Jeaffreson; Collings, D. (2008). Lessons from historical bridge
QUARTERMAINE et al., 1944) failures. Proceedings of the Institution of Civil
Lateral torsional buckling has been identified as the Engineers - Civil Engineering, 161(6), 20-27.
primary reason for failure, exacerbated by an increase doi:10.1680/cien.2008.161.6.20
in dead loading on the day of the incident (Collings, Hayward, A. C. G. (2014). Train loading on bridges
2008; Hayward, 2014; Petroski, 1994). since Stephenson's Rocket. Proceedings of the
Further studies have noted that the beams had a Institution of Civil Engineers - Bridge
cavetto moulding pattern at the fillet between the Engineering, 167(4), 326-337.
bottom flange and web(P. R. Lewis, 2007). This is a doi:10.1680/bren.11.00029
detail that would have been used in timber as a Jeaffreson, J. C. (2015) Chapter II.* Iron Bridges. In
decorative feature. This would have inadvertently The Life of Robert Stephenson, F.R.S Late
added a stress raiser to the beam, highlighting the President of the Institution of Civil Engineers. (pp.
possibility of a fatigue induced failure. 30-72). London: Thomas Telford.
Lewis, P. R. (2007). Disaster on the Dee : Robert
Stephenson's nemesis of 1847. Stroud: Tempus.
Lewis, P. R., & Gagg, C. (2004). Aesthetics versus
function: the fall of the Dee bridge, 1847. Institute
of Materials, Minerals & Mining, p 177-191.
Nash, F. (1844). Description of a new system of Railway Details of Girders Over the River Dee.
trussed girder of wrought and cast iron, for (1846). In WCRM300362 (Ed.): NETWORK
bridges. Minutes of the Proceedings of the RAIL.
Institution of Civil Engineers, 3(1844), 102-103. Walker, J., & Captain Simmons, R. (1847). Report to
doi:10.1680/imotp.1844.24526 the Commissioners of Railways, by Mr. Walker
Petroski, H. (1994). Design paradigms: case histories Captain Simmons, R.E., on the Fatal Accident on
of error and judgment in engineering. Cambridge: the 24th day of May 1847, by the Falling of the
Cambridge University Press. Bridge over the River Dee, on the Chester and
Quartermaine, A. S., Gribble, C., Fuller, F. M., Holyhead Railway; together with any Minutes of
Stratton, H. E., Pearce, J. G., Bridge, E. K., . . . the Commissioners thereupon. Retrieved from
Mitchell, G. R. (1944). Discussion. The strength of http://www.railwaysarchive.co.uk/documents/BoT
cast-iron girder bridges. Journal of the Institution _DeeBridge1847.pdf
of Civil Engineers, 22(8), 292-307.
doi:10.1680/ijoti.1944.14061
Barton Lifting Bridge over The Manchester Ship Canal

Artur Widlak, MEng(Hons) Civil & Architectural Engineering
Tanyel Esenel, MEng(Hons) Civil Engineering
Rémi Sauffisseau, MEng(Hons) Civil Engineering
Maryam Chowdary, MEng(Hons) Civil Engineering
Arooj Shah, MEng(Hons) Civil Engineering
Ehsan Zahoor, MEng(Hons) Civil Engineering
James Soname, MSc Structural Engineering
1.0 Abstract Work had started in May 2015 and completion was
The Barton Lifting Bridge over the Manchester Ship planned for the second half of 2016 but due to the
Canal is known by the people of Greater Manchester failure of the lifting mechanism during a test run
for collapsing months before its opening to traffic. (Figure 2) and subsequent delays, the bridge only
Collapse was due to the failure of the lifting system opened to the public on the 19th of December 2017
when a test run was conducted during construction, (MEN, 2017; NCE, 2017).
fortunately no workers were harmed. This tower drive The design of the bridge was completed with the
vertical lift bridge is part of a larger scheme for the possibility of future developments in mind such as the
development of the port of Salford. It rises to allow installation of tram lines (contingent on TfGM
for the passage of vessels and structurally resembles funding) (PeelLand&Property, 2018). With the AJ
the Centenary Bridge in Salford- which also crosses Bell stadium situated nearby this is a likely
the Manchester Ship Canal. possibility.
This bridge can be categorised as a tower drive
2.0 Introduction vertical lift bridge as it moves along a fixed upright
Barton Lifting Bridge, owned by Peel Land & axis (Birnstiel, 2015; Parke & Hewson, 2008).
Property, is located at Barton-upon-Irwell in the City
of Salford, Greater Manchester on the east side of the
existing M60 bridge (Figure 1). It is part of the £32m
Western Gateway Infrastructure Scheme (initial cost
of the bridge: £1.1m). Cass Hayward were the
designers and the construction was executed by the
joint venture between Hochtief and the Buckingham
Group. Civil works were undertaken by Howard Civil
Engineering. The design was based on the Centenary
bridge which links Eccles with Trafford Park (NCE,
2016).
Figure 2. Bridge after failure. (MEN, 2017)
3.0 Technical details of the bridge

3.1 Load paths
Under the main loading conditions (traffic loading on
a lowered bridge), the bridge deck supports gravity
loads through the slab (which acts as a plate) spanning
across the grillage of primary and secondary beams.
The beams are mostly in bending but transmit shear as
a reaction down to the abutments (shear can be high
when vehicles are near supports, entering/exiting the
bridge). Compressive forces are then transmitted
through the pile cap to the ground through friction.
When the bridge is lifted, the same actions occur
within the deck but the reactions lead to the cables
instead of directly to the abutments. The cables in the
lifting system go into tension (ties) and the
Figure 1. Site location. (GoogleMaps, 2018) mechanical system lifts the deck vertically. At the top,
Barton Lifting Bridge over The Manchester Ship Canal Widlak, Esenel, Sauffisseau, Chowdary, Shah, Zahoor, Soname
the cables run back down inside the tower creating a chosen due to its ability to sufficiently resist
vertical reaction at the top of each tower. Each tower compression. Each corner of the deck is secured into a
supports a compressive force of a quarter of the load rail to prevent sway when the deck is raised.
coming from the deck. Wind loading can become an
issue especially when the deck is in the air and
therefore, the reinforced concrete towers must also
resist sway with bending moments (combined axial
and bending).
Each of the four towers are 31m high and built from
reinforced concrete. These towers allow for a 20m
clearance below the bridge deck once lifted to allow
vessels to continue along the river Irwell
(PeelLand&Property, 2018).
3.1 The deck

Figure 5. Four towers with hollow trapezoidal shape.
The composite deck sits on the ladder of secondary
(LiverpoolEcho, 2016)
beams which span onto the two longitudinal girders
(which rest on the bearings). The cables support the 3.3 Abutments
shorter primary beams at each end as shown in Figure The abutments on each side of the Canal sit on the
3 and observed in Figure 4. To increase efficiency, the pile cap located on the pre-made platform. They are
beam sections are shaped like their bending moment stepped to allow for the inspection of the bearings as
diagrams by becoming deeper at mid-span where the well as enabling their repair if necessary in the future.
moment is greatest. Figure 6 shows the cantilevered Fill was placed behind the abutments to raise the
segment of the cross beams being restrained by the ground level and align the road with the surface of the
top flange of the longitudinal girder. The plate bridge deck. The fill applies some lateral earth
running between the cross girders then reduces their pressure onto the abutment which is transferred into
lateral movement. the pile caps and piles.
3.4 Foundations
The four concrete towers were designed to be located
inside the canal therefore the ground had to be
extended into the water. To achieve this, sheet piles
were driven into a part of the canal to create a
cofferdam-like effect and enable de-watering. This
was then backfilled with granular material to create a
Figure 3. Bridge deck grillage. (Widlak, A.C.)
safe working platform from which to build the
foundations. Piles were chosen due to poor existing
ground conditions at shallow depths and were driven
into the ground (high water table and low frictional
resistance of the soil (Craig, 2004). The piles
supported a pile cap upon which lay the abutments
and towers.
3.5 Bearings
Two main girders run across the river on each side of
the bridge and sit on bearings. In total, four bearings
are used. In a video recorded by a member of the
Figure 4. Bridge deck. (BBCNews, 2017) public, one of the bearings can be seen with the steel
plate connected to the girder. It seems that at the
3.2 Four concrete towers failure stage, the elastomeric fragment was removed
The lifting system structure is made of four hollow from the bearing and it can be observed sitting to the
concrete towers of trapezoidal cross-sections. The left of the bearing in Figure 6.
choice of shape contributes to the aesthetic of the
bridge. The hollow sections allow for the concealment
of the lifting systems yet still allows access for
maintenance and inspection. Reinforced concrete was
Fail-safes used in elevators could be applied to tower

vertical lift bridges in the future where emergency
brakes lock the position of the deck.
5.0 Aesthetics
As this bridge was constructed adjacent to an existing
one, its aesthetic had to conform to its predecessor’s.
The Barton Lifting Bridge assimilates well into its
surroundings with its dusty white hue reflected in the
existing bridge. The deepest section of the deck is at
mid-span whereas the bridge behind it becomes
deeper at the piers. This shape inversion visually pairs
the bridges and brings them further into harmony. The
Figure 6. Bearing on abutment and steel plate on girder. tall towers jut out above the water and their hard
(@irlamcadviews, 2017) edges draw the eye to the lifting platform whilst their
slenderness provides a striking reflection in the water.
3.6 Lifting System The tapering of the tower peaks further accentuates
As mentioned in section 3.2, each concrete tower the bridge, dividing it from its surroundings to stand
houses the lifting system. At the top, a sheave (Parke proudly at the centre of the canal.
& Hewson, 2008) facilitates the movement of the
cable from the outside (visible) to the inside of the 6.0 Durability
hollow concrete tower where the counterweight and The main problem to tackle in terms of durability for
lifting systems are located. Monitoring of the this bridge is the humid environment in which it is
mechanism is essential mainly for ensuring that all of located. For the piles, located above and below the
the cables share the same load and are not ground water level (which are likely to be made of
overstressed. They should also be lifting/lowering the concrete) sufficient cover and a concrete of low water
deck at a rate such that all four corners are at equal to cement ratio is expected. Accordingly, sufficient
heights at any given time. cover for the humid environment is expected to have
been used for the concrete towers with a drainage
4.0 Bridge failure during construction system to avoid accumulation of water.
During construction, contractors tested the lifting For steel elements, corrosion is the main issue. It can
system. This led to the progressive collapse of the be observed that the girders and beams were painted
bridge deck. It was observed that each of the corners off-site which guarantees a higher quality coating.
fell progressively, pointing to a problem with the A waterproofing membrane was placed between the
mechanical system (NCE, 2016). Simon Bourne concrete slab and asphalt layer to avoid water and
(2016) stated that the lifting ropes on the side away cement contact. Asphalt is usually waterproof, but
from the existing bridge went down slowly while the cracks can occur and let water through.
ropes on the other side still did not move- suggesting
a system or motor failure instead of a structural fault. 7.0 Critical Review
By building this new bridge, congestion due to traffic
4.1 Lessons learned
in other areas is reduced. It also allows vessels up to a
Allan Mann (NCE, 2016) highlighted the importance
certain height and width to sail through by closing to
of monitoring potential modes of failure like cable
traffic and vertically lifting the bridge. There is a
jam or disproportionate lifting and sensors should be
planned major work for the development of Salford
in place to spot any anomalies and prevent this from
Port for which the newly built bridge will not be an
happening in the future. It was useful to have tested
issue.
the lifting mechanism during construction as members
The delays meant subsequent increases in embedded
of the public were not in danger. It is important to
carbon due to the use of additional materials and a
note that the mechanical system may be just as
lengthier use of plant to build another deck. The
important if not more critical for the successful
continued debate over the responsibility of
functioning of the bridge. Perhaps the mechanical
maintenance payments further delayed the opening of
system could be calibrated to measure the movement
the bridge. This caused the local community
of each corner more sensitively and feed that back to
exasperation due to severe congestion in that area
lifting the deck more evenly.
(QuaysNews, 2017).
The structural forms used are simple and the
Manchester Ship Canal enabled for the sections to be
transported by water therefore avoiding road Craig, R. F. (2004). Craig's soil mechanics. London:
congestion. CRC press.
As this bridge opened recently (December 2017) GoogleMaps. (2018). Modified satellite view of
durability issues are yet to be discovered but the main Barton Lifting Bridge site. Retrieved from
hindrance to longevity in movable bridges is the https://www.google.co.uk/maps/search/barton+lift
mechanical system. +bridge+manchester/@53.4695727,-
This is a recent build therefore public information 2.3707242,698m/data=!3m1!1e3
such as drawings and structural characteristics are LiverpoolEcho. (2016). Collapse of bridge leads to
difficult to procure. The majority of the references in popular Mersey Ferries cruises being scrapped.
this report are sourced from the news rather than Retrieved from
published papers or official documentation and care https://www.liverpoolecho.co.uk/news/liverpool-
must be taken when using the information. news/collapse-bridge-leads-popular-mersey-
11378239
8.0 Conclusion MEN. (2017). Construction of giant lifting bridge next
Barton Lifting Bridge belongs to the new to M60 back on track after collapse last year.
development scheme which improves traffic in the Retrieved from
Barton area between Trafford Park and the AJ Bell https://www.manchestereveningnews.co.uk/news/g
Stadium. Its lifting system enables vessels to sail reater-manchester-news/construction-giant-lifting-
through the Manchester Ship Canal to the Salford Port bridge-next-13503026
dock which is in planning for future development. NCE. (2016). Lift bridge collapse points to systems
Architecturally, the bridge conforms to its failure. Retrieved from
environment however, it is not particularly visually https://www.newcivilengineer.com/latest/lift-
inspiring. Aesthetic durability issues could involve bridge-collapse-points-to-systems-
cleaning the towers as the concrete may darken over failure/10006938.article
time. The lifting mechanism may need parts replacing, NCE. (2017). Deck collapse bridge finally opens.
the rate of deterioration will depend on the frequency Retrieved from
of lifting. https://www.newcivilengineer.com/tech-
excellence/deck-collapse-bridge-finally-
9.0 References opens/10026553.article
@irlamcadviews (Producer). (2017). Barton Lift Parke, G. A., & Hewson, N. (2008). ICE manual of
Bridge Collapse 2016. Retrieved from bridge engineering. London: Thomas Telford.
https://www.youtube.com/watch?v=bIblbAMO4D PeelLand&Property. (2018). Port Salford. Retrieved
s from http://www.portsalford.co.uk/rd-
BBCNews. (2017). Cash row could hold up new improvements/#new-page-2
£35m bridge opening in Salford. Retrieved from QuaysNews. (2017). Pressure grows on Peel Holdings
http://www.bbc.co.uk/news/uk-england- as community calls for Barton Lifting Bridge to be
manchester-41838636 opened. Retrieved from
Birnstiel, C. (2015). Movable bridge design standards. http://quaysnews.net/index.php/2017/12/15/calls-
In Movable Bridge Design (pp. 115-121): ICE for-barton-lifting-bridge-opened/
Publishing.

Salford Journal of Bridge Engineering Jan-18

Uploaded by

Copyright:

Available Formats

Salford Journal of Bridge Engineering Jan-18

Uploaded by

Document Information

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

Salford Journal of Bridge Engineering Jan-18

Uploaded by

Copyright:

Available Formats

What caused the collapse of the I-35W Mississippi River Bridge?

What caused the collapse of the I-35W Mississippi River Bridge?

What structural form did the I-35W Mississippi River Bridge have?

What structural form did the I-35W Mississippi River Bridge have?

Salford Journal of Bridge Engineering

Volume 1, Issue 6, January 2018

Editorial Panel Contents

I-35W Mississippi Bridge Collapse ………………………………… 1

excellent education since 1896

I-35W Mississippi River Bridge

the least corroded bearings. The other at the time of and

Figure 0.3 - I-35W Collapse (National Transportation

2.3 Technical Drawings

3.0 Structural Form

Figure 0.9 - Roller Bearing (Retro Report, 2007).

Figure 0.10 - Roller Bearing (Retro Report, 2007).

Tacoma Narrows Bridge Failure

1.0 Abstract insufficient to withstand the dynamic forces acting on

Figure 0.14 - Tacoma Narrows Bridge (HERNANDEZ,

Figure 0.13 - General Arrangements (WSDOT - Tacoma

5.0 Critical Review

Figure 0.18 - Tacoma Narrows Bridge Labelled

Figure 0.17 - The Two Towers and End-Spans Left

structure. Figure 0.22, shows these wind lapses being

Figure 0.19 - Second Order Effects on the Deck

Figure 0.22 - Wind starts to Oscillate the Deck and Create

Figure 0.20 - Wind Acting Upon the Deck (PASTERNACK,

Figure 0.23 - Deck is in phase with wind vortexes

Figure 0.25 - Components of the Failed Bridge

Hyatt Regency Walkway Collapse

 1978 – Construction began and in December with

2.1 Brief Project Timeline

4.0 Structural Performance

The 1846 Dee Rail Bridge, Chester

sandstone with the bed-stones being in the same

Figure 5. Section through deck (Walker & Captain

In 1871 the structure was completely replaced by a

Barton Lifting Bridge over The Manchester Ship Canal

Figure 2. Bridge after failure. (MEN, 2017)

3.0 Technical details of the bridge

3.1 The deck

Fail-safes used in elevators could be applied to tower

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.