Case
Study
Watergate
Bridge
-
Arch profile investigations
-
Linear and nonlinear buckling analysis
-
Global modelling of structure
and local modelling of
connections
The new Watergate Bridge across the
River Boyne in Trim, County Meath, in the Republic of Ireland is a low
rise, parabolic arch road bridge with cantilevered pedestrian
walkways. The designers, Roughan
& ODonovan, used LUSAS Bridge to assist with all
aspects of its analysis including the preliminary design and
selection of the structural form, linear and geometric nonlinear
buckling analysis of the superstructure, and also for local analysis
of selected complex fabricated connections.
Overview
The new bridge spans 35m and comprises tapering pentagonal arch
chord sections with pairs of Macalloy hanger bars connecting to
Universal Beam tie members. Universal Column transverse members
and end diaphragms support a 225mm reinforced concrete deck which
acts compositely with superstructure steelwork. It replaces a
previous
composite multi-span structure that had a weight restriction and was
in a bad state of repair. A
clear-span replacement structure was required in order to
eliminate the regular upstream flooding problems caused by the three
wide piers and deck superstructure of the existing structure. The
result for the
client Meath County Council is a landmark modern structure that
complements the heritage of the town and its surroundings.
|
|
|
|
2D preliminary
design
Preliminary design with LUSAS
involved the creation of 2D line beam models to investigate
alternative arch profiles, hanger configurations and relative
section sizes. These models quickly showed the advantages and
disadvantages of each arrangement. A low rise arch with vertical
hanger rods was chosen in preference to a higher rise arch for
aesthetic reasons. However, this meant that greater bending
effects would occur in the arch chord and longitudinal members,
and also the arches could not be braced horizontally, so careful
consideration of their buckling capacity would be required.
Having chosen the arch type the
next step was to carry out a comparative analysis of the relative
section sizes of the arch chord member against the longitudinal
tie member because a stiff arch would reduce the effects in the
deck and vice versa. This process was very much simplified by
using the section library built into LUSAS. This led to an
influence line analysis which showed that the worst effects in the
arch are caused by quarter span loading and by high intensity,
short loaded lengths.
|
|
The final and possibly most
important stage of the preliminary design was the buckling
analysis. UK design code BS5400 Part 3 gives a table of effective
lengths for compression members, all of which are rather
conservative, but allows the designer to determine the effective
length by an elastic critical buckling analysis. LUSAS calculated
the first few mode shapes and the effective length and critical
buckling load was found.
3D detailed
analysis
A complete elastic analysis was
carried out on a 3D line beam model with the deck modelled with
thick shell elements. Live loading, temperature loading, and the
effect of a collision on the arch rib were all analysed. From
this, it was found that the tensile forces in the tie beam were
shed to the composite concrete deck and as a result the girder
size was reduced and additional reinforcing steel was added to the
deck to take the tensile stresses. This made the deck monolithic
with the tie beam, and torsionally stiffened the girder as a
result.
It was considered that the linear
buckling analysis from the two dimensional model had
over-estimated the critical buckling load so this was recalculated
for the 3D model with more accurate modelling of the end fixity. A
full geometric nonlinear buckling analysis investigated 2nd order
effects of loading the deck and from the load cases considered it
was seen that loading the full span with a Uniformly Distributed
Load caused the greatest lateral deflections of the arches. A
comparison of the effective lengths from the 3D linear and
nonlinear analyses gave very similar results of 12.7m and 12.5m
respectively. Based on these values a maximum allowable
compressive stress of 212N/mm2 was obtained for the steel type
used.
Local modelling
While the 3D line beam model was
satisfactory for the design of individual members, more detailed
local models were required for complex fabricated connections such
as the arch/girder connection above the bearing and for the
arch/tie rod connection. Forces and moments derived from the 3D
global model were applied to the local model allowing stresses and
shear distribution in these connections to be easily seen and
economical weld sizes obtained.
"LUSAS was used throughout the design, from preliminary
sizing of members right through to the detailed design of the more
complex fabricated elements. It helped us to provide the client
with an efficient, aesthetically pleasing structure."
David Doyle, Structural Engineer, Roughan & ODonovan
Find out more
Other LUSAS Bridge case studies:
|