Professional Documents
Culture Documents
4)
ORDINARY MEETING
A paper to be presented and discussed at a meeting
of the Institution of Structural Engineers, I1 Upper Belgrave
Street, London SWlX %BH, on Thursday I1 April 1985, at 6 p m.
.I
Kinlochbarvia
Lochinver- 3
Kyle of Lochalsh
The S~ructural.BngineerNolume
6 3 m o .3hlarch..1985. 69
Paper: NissedFalbe-HansenlStears
There has been a ferry service at Kylesku for a long time. It is known would be 8.4 m; the total length would be about 262 m.
that rowing boats were regularly carrying passengers across the waters in The bridge would be setin an exceptionally beautiful landscape and in a
the early 1800s, and later horses and carts were also carried. A car-ferry remoteand veryexposed marineenvironment.The designshould be
servicewas introduced between the wars, using a varietyof turntable suited to the character of the site and aim at low maintenance. The client
boats, and in 1975 a roll-on roll-off ferry was bought as part of the route had said that he would prefer a concrete bridge.
improvements. This ferry, theMaid of Glencoul, was capable of carrying Foundations would not be difficult. Bedrock is at shallow depth and is
up to 12 cars at a time. It operated from dawnto dusk. After dark, which generally strong gneiss which is freshor slightly weathered. Several sets of
comes early during the winter, travellers had to make the long detour if widely-spaced joints are present, and the massive, blocky structure would
they wanted to cross the loch. give good bearing capacity.
A bridge was clearly a better long-term proposition than maintaining a Designs were prepared for a number of possible schemes in order to
ferry boat at Kylesku. A bridgewouldhelp to unify the communities find the most suitable structural form for the bridge. Rough estimates
along the road by making their travel easier and available around the were made for both concrete and steel bridges.
clock. This in turn should lead to greater opportunities for providing A bridge on vertical supports would have a main span of about 130 m.
services on a wider basis and with a greater choice of service centres. It The cost of a bridge is controlled to some extent by the length of its main
would,for example,bepractical to considerproviding secondary span, and severalwaysofreducing the span wereexaminedwith the
education in the area. A bridge would also encourage tourism and thus purpose of lowering the cost of the bridge.
act as a much-needed stimulus to the local economy. A cable-stayed structure would allow savings to be made in the deck
The Highland Regional Council made an initial feasibility study for a structure, but would require costly towers given the height of the bridge.
bridge early in 1978 and, in June of that year, appointed Ove Arup & It would not cope easily with the curved deck and would generally be
Partners to design the bridge and the approach roads. more expensive than a bridge with a substantial structurebelow the deck.
It seemed natural to use the heightof the bridge to advantage by
Route studies introducing structural elements below the deck in such a way that the
The Council hadidentified four possible routes in the initial study (Fig 2). main span of the deck was reduced. There are anumber of ways in which
It was clear from rough estimates that the cost of the approach roads this can be done, and three of these were considered. They are shown in
would be relatively low compared to the bridging costs, and the most Fig 3, together with a bridge on vertical supports.
economic bridge location was chosen therefore in order to achieve the The first of these structures is a deck-stiffened arch. The construction
most economic scheme overall. For this reason one of the routes (route B of the arch would involve major temporary works, and the curvature of
which was the most direct route but which required a longer bridge) was the deck would need to be dealt with. The centre of gravity of the whole
abandoned. bridge would lieoutside aline between the natural centres of the supports
The remaining three routes were now considered in more detail. They and, to resist overturning, it would be necessary to widen the arch on the
all used the same site for the crossing to give the shortest possible bridge. outside of the curve towards the supports. The line of the road would
One of the routes followed the existing road as closely as possible and need to be raised to provide the required navigation clearance.
would require a second short bridge. The other two crossed the island, Theothertwostructuresare frames; one is acentralframe with
Garbh Eilean, and returned to themainland on a low causeway. Route A inclined legsand side spans onslender columns or walls, the otherhas two
passed through the peak on the island and would require a very deep cut inclined legs at each support to form a V-shape.
at this point. Route C followed the contours closely and would avoid
substantial rock excavation and preserve the character of the landscape.
The road was designed to the standards set by the client with a design
speedof 60 km/h and a 5 5m-widecarriageway. The standards were
generallysimilar to those setby ‘Layout of roads in rural areas’ but
allowed some local relaxations in sight distance and gradient. A digital a
ground modelwas prepared from a1 : 2500 topographical survey, and the
routes were studied using the MOSS computer program. All the routes
involved rock cuts and steep and curving alignments, and great care was
exercised to ensure that the integrity of the landscape was not broken.
Thestudiesconfirmedthe client’soriginalpreference forroute C,
although it was found that this alignment required a curved bridge to
meet the road standards. Itgave an economical scheme and was the route
most sympathetic to the general ground form. This routewas selected for
further detailed study. A new digital ground model was prepared from a
detailed survey at 1 : 500 scale, and the road design was completed, again b.
using the MOSS program.Interrain such asthis,comparison of
alternativesanddetail designby handmethods wouldhavebeen
extremely laborious.
currents are strong. The water is deep, and the sides of the channel are
steep. The bridge should therefore be founded clear of the water and
would have to be built without temporary supports in the water.
The bridge would be at high level in an area of high winds. Winds in
excess of 100 mile/h were recorded during construction, and prolonged
periods of gale force winds are not uncommon. The Department of Trade
had indicated that a headroom of 30 m would be ideal but they would
accept a 24m-high and 80m-wide clearance. The alignment studies had 05.0 132.0
v -*
shown that a headroombetween 20m and 24 m could be provided with an
economic line that would fit the ground form on the approaches. d
The bridge deck would be narrow in width compared with its length
and it would be curved in plan. In addition to the carriageway, the client Fig 3 . (a) structure on vertical supports; (b) arch structure;
wanted to provide one l 5 m footway on the bridge and the overall width (c) frame with inclined legs; (d) frame with V-supports
of the supports
MRES 8.4-9.28
k
0.5, * 1.5
h T
Fig 7. Cross-sections
-.-
,
m, 52-5 m, and 46.5 m. The legs are cast into the base and are rigidly modified to accommodate these variations is shown in Fig 7.
connected to the deck without joints. Thedeck is continuous throughout The deck and the legs are built in 52.5 N/mmz concrete. Ordinary
its length, and bearings and expansion joints, which are the most likely Portland cement is used in the deck and sulphate-resisting cement is used
items to lead to maintenance problems, are used only at the abutments in the legs which are exposed to salt and water spray. A smooth finish is
where they are relatively easy to inspect and renew if necessary. used for all visible concrete.
As the bridge is curved in plan, it is necessaryto vary the superelevation The parapets are aluminium with three rails. The black top surfacingis
and the width of the carriageway. The way in which the deck structure is 100 mm thick and laid on top of awaterproofingmembrane.The
NORTH SOUTH
STAGE 0
TEMPORARY RESTRAINT
STAGE 1
STAGE 2
STAGE 3
Fig 9. Stages of construction as assumed in the design
carriageway is drained to gullies at the inside kerb and from there to Structural analysis
catchpits behind the abutments via pipes inside the box. Electricity and Thestructure wasdesigned to the Technical Memorandafor Bridges
telephone cables are also carried inside the box. issuedby the ScottishDevelopment Departmentand, in particular
The abutments aresimple wing wallsfounded on the rock at high level. SB6/74 ‘Standard highwayloading’,SB4/74‘Prestressedconcrete for
Small embankments are required at both ends. highway structures’. The structure was designed for full HA loading and
checked for 37.5 units of HB loading.
Stages of construction The structural analysis was carried out on the consultant’s computer.
The deck was designed to be built in four principal stages (Fig g), and it The idealised structure used in the analysis wasa space frame. Thecurved
was assumed that all concrete would be cast in situ. The detailing of the deckwasrepresentedby a single string of straight membersbetween
reinforcement and the prestressing was compatible with this sequence. A nodes positioned on the curved centreline of the deck. The supports were
number of experienced contractors in Scotland were consulted before the represented by members along the centreline of each inclined leg. Deck
construction sequence was finally decided. and support members were connected by horizontal cross-members. The
The construction stages were: structure was pinned at the bases under the supports and free to rotate
Stage 0: north and south: and move longitudinally at the abutments. A plane frame representation
constructsupport legs to soffit ofdeck and
construct of the structure was also used in the design and as a useful check on the
abutments. more complicatedspace frame model. The consultant’splane frame
Stage l : north and south: computer program could automatically generate equivalent loads from
construct endspan and small cantilever of the following span. cable profiles which the space frame program could not, and the basic
Stress two-thirds of the cables in stage 1. prestressingdesignwas therefore carried out using the plane frame
Stage 2: north and south: program.
complete span between legsand construct 18 m cantilever of the The following procedure was used:
central span over the water. Stress the remaining cables in stage (1) Design the prestress for bending in the vertical plane:
1 and stress all thecables in stage2 apart from continuity cables Analyse the plane frame model by the equivalentload method.
for stage 3. Allow for extra frictionlosses due tothe horizontal curvatureof the
Stage 3: construct the 43m-long central part of the deck between stage 2 deck.
north and stage 2 south. Stress continuity cables. Find the secondary moments at the frame corners using the plane
Two restraints had to be specified in order to achieve an efficient design: frame analysis. Transfer thesemoments to the space frameas
- Thelandward crosshead should be supported until thecentral restraint forces at the pier crosshead nodes and release the space
sectionof the deck(stage 3) was cast. This restraint shouldbe frame.
removed before the continuity cables for stage 3 were stressed. (2) Design the prestress for bending in the horizontal plane:
- The maximum deflection
of the seaward crosshead during - Analyse the space frame using the equivalent load method.
construction should not exceed 50 mm. The structureis a highly redundant frame, so special consideration had to
The first restraint would fix the position of the landwardcrosshead and begiven to creep,shrinkage, andtemperature effects. The Technical
thusensurethattheframe would
be stable in the intermediate Memoranda didnotcoverthesefully and the design for these strain-
construction stages.Thisrestraintcould be providedby atemporary inducedforces wasbased on BS5400’: Part 2 and Part 4 and ACI
vertical support which would be in compression during the construction Publication SP272.
of stages 0, l , and 2, and in tension during the construction of stage 3. The analysisofcreepisof some interest. Thestructure progresses
The second restraint was introduced to control the bending moments in duringconstructionthrough various statical systemsofincreasing
the pier legs and couldbe provided bya temporary horizontaltie between redundancy. This change in statical system will generate a redistribution
the two crossheads. of forces due to the creep of the concrete. This redistribution has to be
taken into account in the design. and to prepare the working drawings before the independent check was
A method was developed that allowed for the variations in the statical carried out. G. Maunsell & Partners were appointed to make the
system and the differences in age of the concrete in each construction independent check in November 1981, working in parallel with, but
stage. It was based on the following set of simplifying assumptions: independently of, the consultants.
- a 2 months' timelag between successive stages This arrangement worked well, and has some advantages when unusual
- cables in aparticular stage would be stressed 15 days afterthe structures have to be designed. The design engineer can have a second,
concrete had been cast independent opinion when it matters, and differences in interpretation
- a linear relation between therate of creep and the rate of can be discussed and a solution agreed when it is still relatively easy to
redistribution make modifications.
The following time-dependent creep function was used2: Six contractors were invited to tender on 29 January 1982, and tenders
were receivedon 30 March 1982. The tenders were reasonably close, three
t 0.60 being within 10 9'0 of the lowest tender. Two tenderers proposed
4 (t,to) = 1.25t0-('.1'* 0 (a,7)-
10+t0.60 alternative construction methods, one proposing to build the legs using
precast units and the deck in cantilevering, the other proposing to precast
where the centre section of the deck.
to is the age at time of loading The successful tender was submitted by Morrison Construction Ltd. in
t is time after loading conjunction with Lehane, McKenzie & Shand, with a contract price of
4 (m,7) is the ultimate creep coefficient for loading at 7 days E2.75M. They started work on site on 9 August 1982. The time for
completion was 2 years.
The ultimate creep coefficient depends on the concrete properties, the
dimensions of the member, and the climatic conditions. To allow for the
Construction
In his tender, thecontractor offered to construct the bridge in the
uncertainty in estimating 4 (m,7) and the effects of the simplifying
sequence proposed in the tender documents. Two changes were, however,
assumptions, two extreme values which were f 25 9'0 of the best estimate
made later:
were used in the design.
Fig 10 shows the effect of creep on the vertical and horizontal bending - Anadditionalconstructionjoint and prestressing stage was
moments in the deck due to its own weight. introduced over the crosshead of the seaward legs, thus dividing
At the design report stage it was assumed that the bottom slab of the stage 2 into two stages, 2A and 2B. The prestress was modified to
box would be left out in theendspans. This reduced their bending accommodate the extra stage.
stiffness and therefore the shrinkage and temperature stresses. Detailed - The centre section of the deck was precast and lifted into place.
analysis showed, however, that the endspans were now too flexible in Some redesign of the continuity cables was made.
torsion and the full box section was used for the full length of the bridge. The actual construction stages are shown in Figs 1 1 to 14.
To compensate for their increased stiffness, the endspanswere lengthened Fig 1 1 shows the construction of the supports. The landward legs were
by about 7 m. This also reduced the size of the abutments and the amount constructed first on a framework of steel beams which was supported by
of earthworks and improved the appearance of the bridge. military trestling towers. The steel framework was also used to construct
the seaward legs, this time held by horizontal ties which were anchored
Works design and tender back into the landward legs. The ties were made from heavy H-sections
The works design of the approach roads was carried out first, and the designed to be left in place to provide the required restraint of the seaward
roads were constructed between July 1981 and August 1982. crosshead.
The works design of the bridge started in June 1981, and itwas Fig 12 shows stage 1 completed and stage 2A in progress. Stage l was
proposed that tenders would be called in January 1982. The client wished built in Bailey bridging and stage 2A on Mabey universal bridging beams.
thatan independent check should be carried out. Given the tight Alm diameter steel tube has been installed to provide the vertical restraint
programme, it was not possible to design and check the design in detail of the landward crosshead. It had to carry acompression force (max.9000
M O M N T S AT TRANSFER
- 50
- 10
10
a.
b.
Fig 12. Stage I completed and stage 2A being cast Fig 14. The lift of the centre section
The 1985 75
Paper: NissedFalbe-HansedStears Book reviews
kN) in stages 0, 1, and 2 and a tension force ( m a . 7500 kN) in stage3. It 1393 m3 in supports
was prestressed using six tendons coupled to rock anchors at the bottom 1894 m3 in deck
and anchored into the deck structureat the top. The prestress was applied Reinforcement: 675 t
in two stages in order to limit the deformation of the pier head. of
which 70in
abutments
t
Fig 13 shows stage2B being built in incremental cantilevering, using a 300 t in supports
gantry which was specially designed and built for use on this project. 305 t in deck
Fig 14 shows the600 t centresection being lifted into place. It was cast
on a platformwhich the contractor had prepared on the shoreof the loch Prestressing tendons: 85 t (5073 m) in deck.
adjacent to thesite, and was prestressed temporarily beforebeing moved 15-2mm Dyform strands. Number of strands in
onto a floating barge. The barge was then manoeuvred into position cables varies from 12 to 19. t
under the bridge and accurately held there using four mooring lines
anchoredtopointsalongthe shore. Thecentrespan was lifted into Acknowledgements
position using a strand jacking technique. The section was held on the Client: Highland Regional Council,
cables, while the continuitycables were installed and the twosmall gaps, G . K. M. MacFarlane,
each 1 5 wide, at each end of the precast centre section were concreted. Director of Roads& Transport
The vertical restraints at the landwardpier heads were then released and,
finally,thecontinuity cables were stressed andthetemporary cables Road & bridge design: Ove Arup & Partners
removed. G . K. S. Rotne was the consultant
architect
Completion
The bridge was completed in July1984 and was officially opened by Her Independent check: G . Maunsell & Partners
Majesty The Queen on8 August 1984.The completedbridge is shown in
Figs 15 and 16. Contractor:
Morrison
Construction
Ltd.
Main quantities References
Length of bridge: 277.8 m 1. BS 5400 Steel, concrete and composite bridges: Part 2: 1978
Width of bridge varies from 8.4 m to 9-28m Specification for loads;Part 4 1978 Code of practicefor the design
of concrete bridges,London, British Standards Institution
Concrete: 3844 m3 2. Designingfor effects of creep,shrinkage and temperaturein concrete
of
which 557 m3 in
abutments stuctures, American Concrete Institute, Special Publication 27