Corus Student Guide to Steel Bridge Design

Student guide to steel bridge design

Corus Construction Services & Development

02

Contents

1 Introduction1.1 General1.2 Basic features of bridges

2 Forms of steel bridge construction2.1 Beam bridges2.2 Arch bridges2.3 Suspension bridges2.4 Stayed girder bridges2.5 Advantages of steel bridges

3 Composite plate girder highway bridges3.1 General layout3.2 Girder construction3.3 Girder erection and slab

construction3.4 Scheme design3.5 Design code checks

4 Material properties andspecifications

5 Corrosion protection

6 Concluding remarks

7 References and further reading

Corus gratefully acknowledges the assistance given by the Steel Construction Institute in compiling this publication.

Introduction

1 Introduction

Bridges are an essential part of thetransport infrastructure.

03

1.1 GeneralA bridge is a means by which a road,railway or other service is carried overan obstacle such as a river, valley, otherroad or railway line, either with nointermediate support or with only alimited number of supports atconvenient locations.

Bridges range in size from very modestshort spans over, say, a small river tothe extreme examples of suspensionbridges crossing wide estuaries.Appearance is naturally less crucial forthe smaller bridges, but in all cases thedesigner will consider the appearance of the basic elements, which make uphis bridge, the superstructure and thesubstructure, and choose proportionswhich are appropriate to the particularcircumstances considered. The use of steel often helps the designer toselect proportions that are aesthetically pleasing.

Bridges are an essential part of thetransport infrastructure. For example,there are more than 15,000 highwaybridges in the UK, with approximately300 being constructed each year asreplacements or additions. Many ofthese new bridges use steel as theprincipal structural elements because itis an economic and speedy form ofconstruction. On average, around 35,000 tonnes of steel have been usedannually in the UK for the constructionof highway and railway bridges.

The guide describes the general featuresof bridges, outlines the various forms ofsteel bridge construction in commonuse, and discusses the considerationsto be made in designing them. Itdescribes the steps in the designprocedure for a composite plate girderhighway bridge superstructure,explaining how to choose an initialoutline arrangement and then how toapply design rules to analyse it anddetail the individual elements of thebridge. Reference is made to simplifiedversions of the Structural Eurocodes forbridge design, which are available forstudent use (see Ref.1 on page 31). Inaddition, the guide outlines materialspecification issues and the variousapproaches to corrosion protection.

Above: Renaissance Bridge (Photo courtesyof Angle Ring Co.), Bedford, England

Opposite: Clyde Arc Bridge, Glasgow,Scotland

Front cover: Hulme Arch, Manchester,England

04

1.2 Basic features of bridgesSuperstructureThe superstructure of a bridge is thepart directly responsible for carrying theroad or other service. Its layout isdetermined largely by the disposition of the service to be carried. In mostcases, there is a deck structure thatcarries the loads from the individualwheels and distributes the loads to theprincipal structural elements, such asbeams spanning between thesubstructure supports.

Road bridges carry a number of trafficlanes, in one or two directions, and mayalso carry footways. At the edge of thebridge, parapets are provided for theprotection of vehicles and people. Thearrangement of traffic lanes andfootways is usually decided by thehighway engineer. Traffic lane and

footpath widths along with clear heightabove the carriageway are usuallyspecified by the highway authority.Whilst the bridge designer has littleinfluence over selecting the layout andgeometry of the running surface, hedoes determine the structural form ofthe superstructure. In doing so, he mustbalance requirements for thesubstructure and superstructure, whilstachieving necessary clearances aboveand across the obstacle below.

Rail bridges typically carry two tracks,laid on ballast, although separatesuperstructures are often provided foreach track. Railway gradients are muchmore limited than roadway gradientsand because of this the constructiondepth of the superstructure (from raillevel to the underside or soffit of thebridge) is often very tightly constrained.

This limitation frequently results in ‘half through’ construction (see Section 2.1). Railway loading is greaterthan highway loading and consequently the superstructures for railway bridgesare usually much heavier than forhighway bridges.

Footbridges are smaller lighterstructures. They are narrow (about 2mwide) and are usually single spanstructures that rarely span more than40m. There are a number of forms ofsteel footbridge (see Ref.4 on page 31),although they are outside the scope ofthis guidance publication.

SubstructureThe substructure of a bridge isresponsible for supporting thesuperstructure and carrying the loads tothe ground through foundations.

05

To support the superstructure, singlespan bridges require two ‘abutments’,one at each end of the bridge. Wherethe bearing strength of the soil is good,these abutments can be quite small, forexample a strip foundation on anembankment. Foundations on poor soilsmust either be broad spread footings orbe piled. The abutments may also act asretaining walls, for example to hold backthe end of an approach embankment.

Multiple span bridges requireintermediate supports, often called‘piers’, to provide additional support tothe superstructure. The locations ofthese supports are usually constrainedby the topography of the ground, thoughwhere the superstructure is long thedesigner may be able to choose thenumber and spacing of piers for overalleconomy or appearance. Intermediate

supports are generally constructed ofreinforced concrete.

Integral constructionTraditionally, movement (expansion)joints have been provided at the ends ofthe superstructure, to accommodateexpansion/contraction. Experience inrecent years has been that such jointsrequire on-going maintenance, yet theyinevitably leak and result in deteriorationof the substructure below the joint. Forbridges of modest overall length, it isnow common to use integralconstruction, with no movement joint. Inits simplest form, the ends of thesuperstructure are cast into the tops ofthe abutments. Integral constructionrequires the consideration of soil-structure interaction and is likely to bebeyond the scope of a student project.

Introduction

Above: Docklands Light Rail Bridge, London, England.

Forms of steel bridge construction

2 Forms of steel bridge construction

Structural steelwork is used in thesuperstructures of bridges from thesmallest to the greatest.

Steel is a most versatile and effectivematerial for bridge construction, able tocarry loads in tension, compression andshear. Structural steelwork is used in thesuperstructures of bridges from thesmallest to the greatest.

There is a wide variety of structuralforms available to the designer but eachessentially falls into one of four groups:• beam bridges• arch bridges• suspension bridges• stayed girder bridges

The fourth group is, in many ways, ahybrid between a suspension bridge anda beam bridge but it does have featuresthat merit separate classification.

The following sections describe therange of forms of steel and composite(steel/concrete) bridge that are in currentuse, explaining the concept, layout andkey design issues for each type.

06

Below left: Trent Rail Bridge, Gainsborough, England.

Opposite: Severn Bridge, Bristol, England.


08

2.1 Beam bridgesBeam and slab bridgesA beam and slab bridge is one where areinforced concrete deck slab sits ontop of steel I-beams, and actscompositely with them in bending. Thereare two principal forms of this beam andslab construction – multi-girderconstruction and ladder deckconstruction. Between them, theyaccount for the majority of medium spanhighway bridges currently being built inthe UK, and are suitable for spansranging from 13m up to 100m. Thechoice between the two forms dependson economic considerations and site-specific factors such as form of

intermediate supports and access forconstruction.

Multi-girder decksIn multi-girder construction a number ofsimilarly sized longitudinal plate girdersare arranged at uniform spacing acrossthe width of the bridge, as shown in thetypical cross section in Figure 1 below.The girders and slab effectively form aseries of composite T-beams side-by-side. The girders are braced together atsupports and at some intermediatepositions.

For smaller spans it is possible to userolled section beams (UKBs), but these

are rarely used today for bridges: plategirders are almost always used.Typically, plate girders are spacedbetween about 3m and 3.5m, aparttransversely and thus, for an ordinarytwo-lane overbridge, four girders areprovided. This suits an economicthickness of the deck slab thatdistributes the direct loads from thewheels by bending transversely.

Ladder decksAn alternative arrangement with onlytwo main girders is often used. Then the slab is supported on crossbeams atabout 3.5m spacing; the slab spanslongitudinally between crossbeams andthe crossbeams span transverselybetween the two main girders. Thisarrangement is referred to as ‘ladderdeck’ construction, because of the plan configuration of the steelwork,which resembles the stringers and rungsof a ladder.

A typical cross-section of a ladder deckbridge is shown in Figure 2. Thearrangement with two main girders isappropriate (and economic) for a bridgewidth up to that for a dual two-laneFigure 1: Cross-section of a typical multi-girder deck bridge.

Footway

Steelgirder

SurfacingWaterproofing

Concrete slab

FootwayTraffic lanes

carriageway. Wider decks can be carriedon a pair of ladder decks.

For both deck types, the use of plategirders gives scope to vary the flangeand web sizes to suit the loads carriedat different positions along the bridge.However, the resulting economies mustbe weighed against the cost of splices.Designers can also choose to vary thedepth of the girder along its length. Forexample, it is quite common to increasethe girder depth over intermediatesupports or to reduce it in midspan. The variation in depth can be achievedeither by straight haunching (taperedgirders) or by curving the bottom flange upwards. The shaped web, eitherfor a variable depth girder or for aconstant depth girder with a verticalcamber, is easily achieved by profilecutting during fabrication.

Half-through plate girder bridgesIn some situations, notably for railwaybridges, the depth between thetrafficked surface (or rails) and the underside of the bridge is severelyconstrained and there is little depthavailable for the structure. In these

circumstances, ‘half through’construction is used. In this form thereare two main girders, one either side ofthe roadway or railway and the slab issupported on crossbeams connected tothe inner faces at the bottom of thewebs. The half-through form is perhapsmore familiar in older railway bridges,where the girders are of rivetedconstruction, but it is still used for newwelded railway bridges and occasionallyfor highway bridges.

In half-through construction using I-beams, the top flange, which is in compression, has to be provided with lateral stability by some means. The two

main girders together with the deck andtransverse beams form a rectangular U shape and this generates so-called ‘U-frame action’ to restrain the topflange. There has to be a momentconnection between the cross-membersand the main girders to achieve this.Under railway loading, the connection issubjected to onerous fatigue loadingand an alternative using box girders hasbeen developed.

09

Top: M4/M25 Poyle Interchange, Slough, England.

Figure 2: Cross section of typical ladder deck bridge.

Footway

Steelgirder

SurfacingWaterproofing

Concrete slab

FootwayTraffic lanes

Box girder bridgesBox girders are in effect a particular formof plate girder, where two webs arejoined top and bottom by a commonflange. Box girders perform primarily inbending, but also offer very goodtorsional stiffness and strength. Boxgirders are often used for large and verylarge spans, sometimes as a cablestayed bridge. They can also be used formore modest spans, especially when thetorsional stiffness is advantageous, suchas for curved bridges.

In beam and slab bridges, box girdersare an alternative to plate girders when spans exceed 40-50m. They can show economies over plate girders,though fabrication cost rates aresomewhat higher for box girders. Twoforms are used:• multiple closed steel boxes, with the

deck slab over the top• an open top trapezoidal box, closed

by the deck slab, which is connectedto small flanges on top of each web

Spans of 100 to 200m typically useeither a single box or a pair of boxeswith crossbeams. Boxes are often variedin depth, in the same way as plategirders, as mentioned earlier.

For very long spans and for bridges suchas lifting bridges, where minimisingstructural weight is very important, anall-steel orthotropic deck may be usedinstead of a reinforced concrete slab.The form of deck has fairly thin flangeplate (typically 14mm) to the undersideof which steel stiffeners have beenwelded; the stiffened plate is then ableto span both transversely andlongitudinally (to internal diaphragms) todistribute the local wheel loads.

Above about 200m, box girders arelikely to be part of a cable stayedbridge or a suspension bridge. The boxgirders used in suspension bridges arespecially shaped for optimumaerodynamic performance; theyinvariably use an orthotropic steel deckfor economy of weight.

The principal advantages of box girdersderive from the torsional rigidity of theclosed cell. This is particularly importantas spans increase and the naturalfrequencies of a bridge tend to reduce;stiffness in torsion maintains areasonably high torsional frequency.

Torsional stiffness also makes boxesmore efficient in their use of material to

resist bending, especially whenasymmetrical loading is considered.Comparing a single box with a twin plategirder solution it can be seen that thewhole of the bottom flange of the boxresists vertical bending wherever theload is placed transversely.

The aesthetic appeal of box girders, withtheir clean lines, is especially importantwhere the underside of the bridge isclearly visible.

Although the fabrication of box girders ismore expensive than plate girders, themargin is not so great as to discouragetheir use for modest spans. For largespans, the relative simplicity of largeplated elements may well lead to moreeconomical solutions than other forms.Erection is facilitated by the integrity ofindividual lengths of the box girders.Sections are usually preassembled atground level then lifted into position andwelded to the previous section.

Box girders are also used for railwaybridges in half-through construction, asan alternative to plate girders. Two boxgirders are used, with the deck simplysupported between them. With thisarrangement, there is no need to achieve

10

U-frame action, because of the torsionalstiffness and stability provided by thebox sections themselves.

Truss bridgesA truss is a triangulated framework ofindividual elements or members. A trussis sometimes referred to as an ‘openweb girder’, because its overallstructural action is still as a memberresisting bending but the open nature of the framework results in its elements(‘chords’ in place of flanges and ‘posts’and diagonals’ in place of webs) beingprimarily in tension or compression.Bending of the individual elements is asecondary effect, except where loadsare applied away from the nodepositions, such as loads from closely-spaced crossbeams that span betweena pair of trusses.

Trusses were common in the earlierperiods of steel construction, sincewelding had not been developed andthe sizes of rolled section and platewere limited; every piece had to bejoined by riveting. Although very labourintensive, both in the shop and on site,this form offered great flexibility in theshapes, sizes, and capacity of bridges.As well as being used as beams, trusses were also used as arches, ascantilevers and as stiffening girders tosuspension bridges.

A typical configuration of a truss bridgeis a ‘through truss’ configuration. Thereis a pair of truss girders connected atbottom chord level by a deck that alsocarries the traffic, spanning between thetwo trusses. At top chord level thegirders are braced together, again with atriangulated framework of members,creating an ‘open box’ through whichthe traffic runs. Where clearance belowthe truss is not a problem, the deckstructure is often supported on top ofthe truss; sometimes a slab is made toact compositely with the top chords, ina similar way to an ordinary beam andslab bridge.

Today, the truss girder form ofconstruction usually proves expensive tofabricate because of the large amount oflabour-intensive work in building up themembers and making the connections. Trusses have little advantage over plategirders for ordinary highway bridges. Onthe other hand, they do offer a very lightyet stiff form of construction forfootbridges, gantries and demountablebridges (Bailey bridges).

Trusses are still considered a viablesolution in the UK for railway bridges,especially where the spans are greaterthan 50m. A high degree of stiffness canbe provided by deep trusses, yet the useof through trusses minimises the

effective construction depth (between raillevel and the bridge soffit), which is veryoften crucially important to railways. Theconstruction depth is dictated only bythe cross members spanning betweenthe main truss girders.

Very many footbridges are built usingtrusses made from steel hollowsections. Profile cutting and welding ofthe hollow sections is straightforwardand economic. Half through or throughconstruction is usually employed – thefloor of the bridge is made at the bottomchord level between two truss girders.

11


Opposite page: A9 Bridge, Pitlochry, Scotland.

Below left: Nene Bridge, Peterborough, England.

Below right: Brinnington Rail Bridge, Manchester, England.

2.2 Arch bridges In an arch bridge, the principal structuralelements (‘ribs’) are curved membersthat carry loads principally incompression. A simple arch ‘springs’from two foundations and imposeshorizontal thrusts upon them. Althoughthe arch ribs are primarily incompression, arch bridges also have tocarry asymmetric loading and pointloading and the ribs carry this partly bybending. This is more conventionallyseen (in masonry bridges, for example)as the displacement of the line of thrustfrom its mean path under dead load.

In masonry bridges, load is imposed onthe arch from above; the roadway (orrailway) is on top of fill above the arch. A steel arch can have a similarconfiguration, with a steel or concretedeck above the arch, supported onstruts to the arch below, or the arch canbe above the roadway, with the decksuspended from it by hangers.

One situation where the arch is stillfavoured is in deep ravines, where asingle span is required; the ribs can bebuilt out without the need forintermediate support. In such cases, thedeck is usually above the arch.

Perhaps the most familiar arch is that ofthe Sydney Harbour Bridge. In that

bridge, much of the deck is hung fromthe heavy arch truss, although the deckpasses through the arch near the endsand is then supported above it.

One form of arch which is sometimesused for more modest spans is the tiedarch. Instead of springing fromfoundations, the two ends of the archare tied by the deck itself (this avoidshorizontal reactions on the foundations).The deck is supported vertically byhangers from the arch ribs.

In recent years, arches and tied archeshave become a little more common,partly because the use of an arch fromwhich to hang the deck allows theconstruction depth of a suspended deckto be kept shallow, even at longerspans, and partly because the archesmake a clear architectural statement.Arches are sometimes skew to the lineof the deck and sometimes the archplanes are inclined (inclined arch planeshave been used in many recentfootbridges, for dramatic visual effect).

2.3 Suspension bridges In a suspension bridge, the principalstructural elements are purely in tension.A suspension bridge is fundamentallysimple in action: two cables (or ropes orchains) are suspended between twosupports (‘towers’ or ‘pylons’), hanging

in a shallow curve, and a deck issupported from the two cables by aseries of hangers along their length. Thecables and hangers are in simpletension and the deck spans transverselyand longitudinally between the hangers.In most cases the cables are anchoredat ground level, either side of the maintowers; often the sidespans are hungfrom these portions of the cables.

In the mid 19th century, wrought ironlinks were used to make suspension‘chains’; by the end of that century, highstrength wire was being used forsuspension ‘cables’. Steel wire is stillbeing used today. Sometimes, for moremodest spans, wire ropes (spirallywound wires) have been used.

In addition to its action in carryingtraffic, the deck acts as a stiffeninggirder running the length of each span.The stiffening girder spreadsconcentrated loads and providesstiffness against oscillation; suchstiffness is needed against both bendingand twisting actions.

Because of their fundamental simplicityand economy of structural action,suspension bridges have been used forthe longest bridge spans. The gracefulcurve of the suspension cable combinedwith the strong line of the deck and

12


13

stiffening girder generally give a verypleasing appearance. The combinationof grace and grandeur in such situationsleads to the acknowledged view thatmany of the world’s most excitingbridges are suspension bridges.

In American suspension bridges, whichpioneered long span construction, trussgirders have been used almostexclusively. They are particularly suitablefor wide and deep girders – some USbridges carry six lanes of traffic on eachof two levels of a truss girder! Japanesesuspension bridges have also favouredthe use of trusses, again because of theheavy loads carried – some carryrailways as well as highways. Thelongest suspension bridge span is thatof Akashi-Kaikyo (1991m) and there thedeck is of truss construction, carryingsix lanes of traffic.

Box girders have been used for thestiffening girders of many suspensionbridges. They provide stiffness inbending and in torsion with minimumweight. Some of the longest spans,such as the Humber Bridge (1410m),Runyang Bridge (1490m) and theStorebælt East Bridge (1624m) havesteel box girder decks.

Left: Forth Road Bridge, Edinburgh, Scotland

Right: River Usk Crossing, Newport, Wales

14


2.4 Stayed girder bridgesIn this form of bridge, the main girdersare given extra support at intervalsalong their length by inclined tensionmembers (stays) connected to a highmast or pylon. The girders thus sustainboth bending and compression forces.The deck is ‘suspended’, in the sensethat it relies on the tensile stays, but thestays cannot be constructedindependently of the deck, unlike asuspension bridge, so it is a distinctlydifferent structural form of bridge.

Stayed girder bridges were developed in Germany during the reconstructionperiod after 1945, for major river bridges such as those over the lowerRhine. Stayed bridges using plategirders and simple cable stays of hightensile wire have proved to be muchcheaper than trusses and have thereforedisplaced them for longer spans (overabout 200m).

Recent developments have extendedthe realm of the cable stayed bridge tovery long spans, which had previouslybeen the almost exclusive domain ofsuspension bridges. Several cablestayed bridges have been built withspans over 800m and Sutong Bridge,due to be completed in 2008, has aclear span of 1088m. Such

development has only been madepossible by the facility to carry outextensive analysis of dynamic behaviourand by using sophisticated dampingagainst oscillation.

The visual appearance of stayedstructures can be very effective, evendramatic. They are frequentlyconsidered appealing or eye-catching.

On a more modest scale, cable stayedconstruction is sometimes used forfootbridges (spans of 40m and above),to give support and stiffness to anotherwise very light structure.

2.5 Advantages of steelbridgesRegardless of the form of bridgeconstruction, a material with goodtensile strength is essential and steel iseffective and economical in fulfilling thatrole. The advantages of steel in bridgesare outlined below.

High strength to weight ratioThe lightweight nature of steelconstruction combined with its strengthis particularly advantageous in longspan bridges where self-weight iscrucial. Even on more modest spans thereduced weight minimises substructureand foundation costs, which is beneficial

in poor ground conditions. Minimumself-weight is also an important factorfor lift and swing bridges, as it reducesthe size of counter-weights and leads tolower mechanical plant costs.

The high strength of steel allowsconstruction depths to be reduced,overcoming problems with headroomclearances, and minimising the lengthand height of approach ramps. This can also result in a pleasing low-profile appearance.

High quality prefabricationPrefabrication in controlled shopconditions has benefits in terms ofquality, and trial erection can be done at the works to avoid fit-up problems on site.

Speed of erectionConstruction time on site in hostileenvironments is minimised, resulting ineconomic and safety benefits.

The lightweight nature of steel permitsthe speedy erection of largecomponents, which minimises disruptionto the public where rail possessions orroad closures are required. In specialcircumstances complete bridges can beinstalled overnight.

15

VersatilitySteelwork can be constructed by a widerange of methods and sequences. Forexample the main girders can beinstalled by crane, by slide-intechniques or using transporters. Steelgives the contractor flexibility in terms oferection sequence and constructionprogramme. Girders can be erectedeither singly or in pairs, depending onplant constraints, and components canbe sized to overcome particular accessproblems at the site. Once erected, thesteel girders provide a platform forsubsequent operations.

Steel also has broad architecturalpossibilities. The high surface quality ofsteel creates clean sharp lines andallows attention to detail. Modernfabrication methods facilitate curvaturein both plan and elevation. The paintingof steelwork introduces colour andcontrast, whilst repainting can change orrefresh the appearance of the bridge.

DurabilitySteel bridges now have a proven lifespan extending to well over 100 years.Indeed, the life of a steel bridge that iscarefully designed, properly built, well-maintained and not seriouslyoverloaded, is indefinitely long.

The structural elements of a steel bridgeare visible and accessible, so any signsof deterioration are readily apparent,without extensive investigations, andmay be swiftly and easily addressed byrepainting the affected areas. Mostmajor structures are now designed withfuture maintenance in mind, by theprovision of permanent access platformsand travelling gantries, and modernprotective coating systems have lives inexcess of 30 years.

Modification, demolition and repairSteel bridges are adaptable and canreadily be altered for a change in use.They can be widened to accommodateextra lanes of traffic, and strengthenedto carry heavier traffic loads. When thebridge is no longer required, the steelgirders can easily be cut intomanageable sizes and recycled, which isa benefit in terms of sustainability.Should the bridge be damaged, theaffected areas may be cut out and newsections welded in. Alternatively, girderscan be repaired by heat straightening, atechnique pioneered in the US, andrecently introduced to the UK.

Top: Forth Rail Bridge, Edinburgh, Scotland.

Below: Top: QE2 Bridge, Dartford, England.

Bottom: Festival Park Flyover, Stoke,England.

Composite plate girder highway bridges

3 Composite plate girder highway bridges

This section of the guide deals principally with beam and slab bridges using fabricated plate girders.

This section of the guide dealsprincipally with beam and slab bridgesusing fabricated plate girders. Itprovides guidance that may help with anundergraduate bridge design project.Following a brief summary of the generallayout, the construction aspects thatneed to be considered are described.Advice is given on scheme or conceptdesign and an explanation of the designcode checks that need to be made isoffered. Advice on more detailedaspects of material specification aregiven in Section 4, and an introductionto corrosion protection is given inSection 5.

3.1 General layoutThe cross-sectional layouts of bridgesdiscussed in this section are the multi-girder deck shown in Figure 1 on page 8,and the ladder deck shown in Figure 2on page 9. The guidance offered relatesboth to constant depth girders (parallelflanged beams) and to beams withvariable depth, although the designcode checks of the latter may bebeyond the scope of an undergraduateproject. For these bridges, theproportions of the girder section (depth,width of tension and compression

flanges and web thickness) are chosenby the designer to suit both the in-service condition (carrying traffic loads)and the loadings at the various stages ofconstruction. The girders are continuousover intermediate supports (when thereis more than one span) and are bracedtogether at supports and at someintermediate positions.

Composite action between the slab andgirder is usually achieved by using studconnectors (headed dowel bars) weldedon the top flange; the number andspacing of studs depends on the level ofshear flow between steel girder andconcrete slab.

In continuous construction, the slab is intension in the hogging moment regionsover the intermediate supports. It isnecessary to provide sufficientreinforcement to the slab in theseregions to share the tensile forces andto limit the consequent crack widths toan acceptable level.

At abutments and intermediatesupports, the girders sit on bearingsfastened to the bottom flange. Thegirders need to be stiffened to carry the

16

Top: Simon de Montfort Bridge, Evesham, England.

Above: Robotic welder(Photo courtesy of Fairfield-Mabey).

Opposite: T&I machine(Photo courtesy of Fairfield-Mabey).

17

bearing loads at these points. Inaddition, transverse bracing is requiredbetween girders at supports to providetorsional restraint and to carry lateralforces (chiefly wind) from the plane ofthe deck slab to the bearings.

At other positions in the span, transversebracing may be required to providelateral restraint to slender compressionflanges. This is required in midspanregions during construction, before thedeck slab concrete has cured, andadjacent to intermediate supports bothduring construction and in service. Inmulti-girder decks transverse bracingnormally takes the form of triangulatedframes between pairs of beams. Inladder decks, the moment-connectionsto the cross girders provide the restraintto the main beams.

For longer spans, the depth of girdersproduces rather slender webs and it is customary to provide vertical webstiffeners in the regions of high shearnear the supports to improve theirresistance to shear buckling. (Withladder deck construction, the stiffenersto which the cross girders are attached perform this function). For a

neater appearance, the web stiffenersfor the outer girders of multi-girderdecks are usually only on the insidefaces, where they cannot be seen,except at bearing positions.

3.2 Girder constructionFabricated I-girders are assembled from three plates, two flanges and aweb. These are normally cut from alarger plate (plates from the rolling millsare typically 2.5m wide x 18m long).

The cutting of flanges and web from a larger plate is achieved by usingcomputer controlled cutting equipment.In cutting a web plate, it is easy to cut to a required camber with very little wastage.

When the three plates have been cut,they are then fillet welded into the I-section. Traditionally this was carriedout by manually assembling the piecesin a jig, tack welding them and thenwelding alternately on each side of theweb. Now available in some fabricatingshops are machines, which can locateand press a web onto a flange in aninverted T, then weld automatically andcontinuously on both sides from one

end to the other. Repeating the processwith the second flange creates the I-girder. This obviously saves labour andconsequently reduces cost.

Vertical web stiffeners are fittedmanually but in some fabrication shopsthe welding is carried out by computer-controlled equipment. It is often cheaperto choose a thicker web than tointroduce a large amount of stiffening.

Each main girder is fabricated in severallong pieces, which will be joined end-to-end or ‘spliced’ on site. Thelengths of these pieces are chosen tosuit the configuration of the bridge, with the fabricated length of eachusually restricted to a maximum of 27m, since girders longer than thisrequire special permission to travel onpublic roads.

Painting is almost always done in thefabrication shop, with the exception ofthe final coat, which is usually appliedon site. Refer to Section 5 for details ofcorrosion protection systems.


18

3.3 Girder erection and slabconstructionWhen the substructure is ready to receivethem, the bare steel girders are erectedfirst, usually by mobile crane. In someinstances, partially assembled steelwork(braced pairs of multi-girders or ladderdeck beams with cross girders) are‘launched’, i.e. they are pushed out alongthe bridge axis from one abutment,though this is not common.

During erection, the consecutive girdersections are joined at the site splicepositions. Such splices, which arenormally arranged near the point ofcontraflexure, are most easily made usingfriction grip bolts, though welded jointsare also used. Bolted splices use coverplates which lap over the ends of bothgirders. Covers are normally on bothfaces of flanges and web, thussandwiching the girder material.

In multi-girder decks, the slenderindividual I-sections are rather unstableon their own, spanning across the fullspan; adjacent girder sections aretherefore frequently braced together andlifted in pairs. In ladder deckconstruction, the main girders usuallyhave to be erected individually and mayrequire some temporary restraint beforethe cross girders are connected.(Alternatively, pairs of main girders withcross beams already connected can belifted in one go, if a sufficiently largecrane is available.)

With the girders in position, the nextstage is for the concrete deck slab to becast. During casting, the concrete needsto be supported and this is normallyachieved using formwork supported bythe steel girders. The traditional formworkcomprises sheets of plywood laid oncross members, which are removed afterthe concrete has hardened. Alternativesto timber include glass reinforced plastic(GRP) panels which can be left in place,and precast reinforced concrete plankswhich become a structural part of thedeck slab, acting with the in-situconcrete above them. The advantage ofsuch permanent formwork is that the removal operation is eliminated. The use of precast planks is now becomingvery common.

The steel girders alone carry the weight of wet concrete and all temporaryworks. As formwork cannot be reliedupon to stabilise the top flange of thegirders, the designer must ensure thatgirders and bracing are adequate for thisloading condition. Temporary cross-bracing is sometimes provided (forexample in midspan regions of simplysupported girders) to stabilise thecompression flange. This bracing isremoved after the concrete has hardened.

Once hardened, slab and girders form a composite section that carries allfurther loads imposed on the bridge. It should be remembered that the stress distribution due to the weight of wet concrete on the bare steel girders remains unchanged; that weight is not carried by the strongercomposite section.

With continuous girders, the bridge issubject to negative or hogging momentsover the intermediate supports, puttingthe deck slab into tension. To minimisethe built-in tensile forces in the slabreinforcement, it is usual to concrete themidspan lengths of slab first and then fillthe lengths over supports.

When the concrete deck is complete, thesurfacing is laid over the whole bridge.The weight of surfacing is thereforecarried on the composite beams.

Below: M20 Road Bridge, Folkstone, England.

19

Figure 3: Typical elevation of composite bridges.

3.4 Scheme design GeneralFor the designer to show his preferencesand exercise his judgement, the greatestscope is at the initial stage when the firstlines are drawn on paper. The roadlayout will have been determined by thehighway engineer. There will be someconstraints in placing the substructure,but the designer is often free to modifyspan lengths and support arrangementsto achieve economy, appearance orother requirements.

At the same time as considering thelongitudinal elevation of the bridge, thecross-section arrangement must also beconsidered. For a plate girder bridge,this means the number and spacing ofthe girders.

ElevationThe dominant parameter, whichinfluences the elevation of the bridge, isgirder depth. Girders may be deep orshallow; they may have parallel flangesor taper to a greater depth atintermediate supports (haunched); thesoffit (bottom flange) may be curved inelevation, like an arch. In deciding uponan appropriate girder depth there aresome useful rules of thumb that may beemployed to produce an initial outline ofthe bridge. Typical arrangements andproportions are shown in Figure 3 above.

Girders with constant depth are,naturally, the basic form and thestarting point from which to considerthe elevation. These parallel flangegirders are cheaper per tonne tofabricate than variable depth girders.For shorter spans, below about 35m,there is little advantage in choosingvariable depth girders, and parallelflanges are usually selected.Continuous viaducts (many similarspans) use parallel flange girders, theappearance of numerous variations ofgirder depth being generally consideredunattractive. Girders with haunches ora curved soffit are most suited to athree-span bridge or the major andadjacent spans of a viaduct. Curvedsoffit girders look particularly pleasingwith a fairly low level bridge such as ariver. Girders with tapered haunches areused in crossing motorways or largerspans over railways.

The design of variable depth girders isgenerally beyond the scope of thesimplified design rules appropriate tostudent use (Ref.1 on page 31), so careshould be exercised in applying them toproject designs other than parallelflange girders.

The depth of a parallel flange bridgeshould normally lie between span/20and span/30, the depth being measuredfrom top of slab to underside of girder.Simply supported spans will usually betowards the deeper end of this range,with continuous spans toward theshallow end. Ladder deck bridges areoften a little deeper, particularly wherethe deck is wide.

Variable depth girders allow reducedconstruction depth in midspan at theexpense of greater depth over theintermediate supports. A central depthof between span/30 and span/40 can beachieved with a depth of about span/18at the adjacent supports. A tapered endspan needs a depth of about span/15 atthe first intermediate support.

In selecting span sizes, it should benoted that where there are many spans,uniformity looks better than irregularspacing. It is better not to vary toogreatly the spacing of adjacent spans:an end span of 80 per cent to the lengthof the next span is structurally wellproportioned. For a bridge with variabledepth girders, the spans either side ofthe major span should be betweenabout 60 per cent and 80 per cent of themajor span.

20

Cross section – multi-girder decksThe basic cross section appropriate tomulti-girder composite plate girderbridges is shown in Figure 1, on page 8.The reinforced concrete slab sits on topof four steel girders. The spacing ofgirders is uniform and on each side ofthe bridge are cantilevers supported bythe continuity of the deck slab.

Girder spacing is influenced by thedesign of the deck slab, which acts bothas a top flange in longitudinal bendingand as a slab in traverse bending. Forpresent purposes, it is sufficient to notethat girder spacing is normally between2.5m and 4m and slab thicknessbetween 240mm and 260mm, the actualvalue depending largely on theconfiguration necessary to suit the deckwidth. Cantilevers should not exceedabout 2m and should certainly be less ifthey carry vehicle loading (evenfootways have to be designed foraccidental vehicle loading unlessprotected by a crash barrier).

An even number of girders is to bepreferred. This allows girders to bepaired together by transverse bracing

for lateral stability of the compressionflanges; there is then no bracingbetween adjacent pairs.

If transverse bracing is continuousacross many girders, it participates inthe global bending of the bridge andbecomes prone to fatigue damage –such continuity is best avoided.

In a multi-girder bridge, the webs areusually thin and require intermediatetransverse web stiffeners to enhanceshear resistance. In hogging momentregions (adjacent to intermediatesupports) most of the web depth is in compression – the thin webs then limit the cross section to its elasticbending resistance (unlike sections with thick webs, which may developplastic resistance).

Cross section - ladder decksThe basic cross section appropriate toladder deck-girder composite plategirder bridges is shown in Figure 2, onpage 9. The reinforced concrete slabsits on top of the cross girders and themain girders, and spans longitudinallybetween cross girders. The spacing of

the cross girders is generally uniform(there is some variation local to thesupports of skew decks, to suit theskew angle). The deck slab outside thelines of the main girders is notsupported on beams, it cantilevers inthe same way as in multi-girder decks.

The cross girder spacing is usuallybetween about 3m and 4m and the slabthickness is between 240mm and260mm, as for multi-girder decks. Crossgirders can span up to about 18m.

Initial sizingTo make a start on detailed design, it isnecessary to select some preliminarymember sizes so that analysis can becarried out. Such initial selection can bebased on fairly crude estimation ofbending moments. In simple spans,overall moments can obviously becalculated quickly; in continuous spans,moments can be estimated as aproportion of the values calculated for afixed-ended beam. In both cases, loadscan be shared between girders bystatics or by simple rule of thumb.

21

For continuous spans, dead andsuperimposed load moments should betaken as fixed-ended beam moments,possibly modified by the momentdistribution method or other simplemanual calculation if the spans aresignificantly unequal in length. Live load moments over an intermediatesupport should be taken as about 90 per cent of wL2/12 and at midspanthey should be about 120 per cent ofwL2/24 (where w is the load/unit lengthcarried by an individual composite beamand L is the span).

For ladder decks, the traffic loads canbe proportioned between the two maingirders on the basis of a ‘static’distribution. For multi-girder decks a‘static’ distribution of the load from eachlane between the two girders under thatlane can lead to significant over-estimate of the load on an individualgirder, because the slab spreads theloads transversely between all thegirders. On the other hand, equalsharing of the total load between all thegirders will give an underestimate. A value midway between these twoalternatives could be used as a first

guess, and a little experience with thesubsequent analysis would aid futureinitial judgements in similarcircumstances. For loads on the crossgirders of ladder decks, the self weightand the UDL component of traffic load,(see the simplified Eurocodes document,Ref.1 on page 31) is shared equallybetween cross girders; the TandemSystem (TS) component may be takenconservatively as being whollysupported on one cross girder.

From these simply calculated values ofmoments and shears, flange and websizes can be selected using theprinciples of limit state design (seeSection 4.5.2 of Ref.1 on page 31).Tension flanges may presume a designresistance based on yield strength.Compression flanges may conservativelypresume a resistance based on 90 per cent of yield strength for the in-service condition. Webs may presumea resistance based on 60 per cent of theshear yield strength.


This page: A1(M), North Yorkshire, England.

Left: Thelwall Viaduct, M6, Warrington, England.

22

3.5 Design code checks GeneralDesign code checks are sometimescalled ‘detailed design’, but it is more ofa checking process than originalcreative design. The selected structuralarrangement is analysed for the variousloading conditions and then thestrengths of the members are checkedin detail to ensure that they areadequate to carry the moments andforces. Details such as stiffener sizesand bracing member sizes, etc, arechosen at this stage to suit the globalactions of the main members.

Before commencing design checks, thedesigner should confirm and record thenecessary parameters. He should know:• the geometrical configuration to be

achieved• the loading to be applied• the design standards to be observed• the properties of materials to be used

For the purposes of student projectdesign, the simplified versions of theStructual Eurocodes should be used.(Ref.1 on page 31). The project brief willdefine the geometrical configuration andperhaps the loading to be considered (ifit is other than the simplified highwayloading given in Ref.1 on page 31).

Limit state designModern design is based on limit stateprinciples. Under this philosophy,structural adequacy is verified at twolimit states, referred to as ultimate limit state (ULS) and serviceability limitstate (SLS).

At each limit state the effects of nominalor ‘characteristic’ values of appliedloads are evaluated and multiplied by a‘partial factor’ to determine ‘design loadeffects’ that have a high level ofreliability (i.e. a very low level ofprobability that they would be exceededduring the life of the bridge). These

effects are the internal forces, momentsand stresses within the structure.

The designer then verifies that theeffects are acceptable throughout thestructure. In practice The designeridentifies the small number of positions where the effects will begreatest or most critical and evaluatesthe effects and the limits at thesepositions. Typically the designer willneed to consider:

At supports:• Maximum moment with coexisting

shear• Maximum shear with coexistent

moment• Maximum reactions (for bearing

stiffener design and bearing selection)

In midspan regions:• Maximum moment with coexisting

shear

23

At changes of beam section:• Maximum moments and shears

For cross girders of ladder decks:• Maximum moment in midspan regions• Maximum shear at the ends (for

design of bolted connections)

At each position where adequacy isverified, the design value of theresistance is determined, based on thenominal strength of the material and thegeometrical proportions of the memberand its cross section. In the calculationof the resistance, the ‘nominal’ strengthis reduced by dividing by another ‘partialfactor’, again to ensure a high level ofreliability (a high probability that thestrength would be at least this value).

Adequacy is achieved when the designresistance is at least equal to the designload effects.

The ultimate limit state (ULS) is reachedwhen a member or part just fails,through rupture, buckling or fracture.

The serviceability limit state (SLS) isreached when damage becomesapparent, necessitates remedial action,or where the condition causes publicconcern, for example because ofexcessive vibration or deflection.

Partial factors on loads are normallygreater for ULS than for SLS because agreater margin is demanded againstfailure than against first occurrence ofdamage.

In many instances, for example whereULS is deemed to be reached whenyield occurs in an extreme fibre of thesection, the lesser requirements for SLSneed not be checked, since they willautomatically be satisfied.

Top: Slochd Beag Bridge, Inverness, Scotland.

Above: Westgate Bridge, Gloucester, England.

24

LoadingThe loading carried by a bridgecomprises dead load (the weight of thestructure itself and any permanentfixtures to it) and the variable trafficload. The actual traffic loadingexperienced by bridges is of courseextremely variable, and it would beimpossible to examine any design for allpossible vehicle combinations.

Instead, ‘load models’ are used torepresent the traffic loading. To representnormal traffic a uniformly distributed load(UDL) is applied within each traffic laneover an appropriate length. Additionally,to represent the non-uniform nature ofactual loading, a pair of heavy axles areapplied at a position along the lane thatcauses the worst effect. In theEurocodes, this loading is known as‘Load Model 1’ (LM1).

To represent abnormal traffic – the heavy multi-axle commercial vehiclesthat are permitted within certainlimitations on their movement – aseparate load model is used. In theEurocodes, a number of different axlearrangements and axle loads are definedfor this load model (LM3).

The loading given in the simplifiedEurocodes document (Ref.1 on page 31),are based on LM1, as interpreted by theUK national annex; a simplified LM3 isalso shown, for information, but wouldnot normally need to be considered for astudent project.

Loading on footways is also modelledusing a UDL – use either the value in thesimplified Eurocodes document or thevalue specified in the project brief.

AnalysisTo carry out the detailed design of anelement of the structure, a globalanalysis is necessary to determine theforces and moments in the structureunder the variety of loading conditions.

Moments and shears in the steel beamsdue to dead loads can be calculated byanalysis of line beam models. For simplespans, manual calculations arestraightforward.

Calculations of forces on the compositestructure require global analysis that

takes account of longitudinal andtransverse stiffness throughout thestructure. For a multi-girder deck, thisanalysis is usually carried out bycomputer using a grillage model inwhich the structure is idealised as anumber of longitudinal and transversebeam elements in a single plane, rigidlyinter-connected at nodes. Loads areapplied, normal to the plane, at the nodepoints. or the level of analysisappropriate to a student project, it isadequate to use a model with six equalspacings along the main beams in eachspan and one line of elements for eachlongitudinal beam. A typical grillage isshown in Figure 4 below.

Each beam element represents either acomposite section (a main girder withassociated slab) or a width of slab. Slabwidth should be calculated midway tothe node on either side, or to the end ofa cantilever.

For a ladder deck, beam elements arerequired for each transverse beam andfor each longitudinal beam. Since thedesign of the slab itself is outside thescope of a student project, a finer meshis not normally needed.

Gross section properties are used forglobal analysis. Properties for compositebeams should include the full area of theappropriate slab width, except that thefirst longitudinal elements adjacent to anintermediate support in each span(about 15 per cent of the span) shouldbe given cracked section properties

(i.e. ignore the area of concretebecause it is in tension and will becracked, but include the area ofreinforcement). Properties for a width ofslab should include the full sectionalarea of the slab.

The short-term modulus for concreteshould be used throughout. Divide theconcrete area by the modular ratio togive equivalent steel areas forcalculation of composite properties.

Strength checkingThe strength check of the critical partsof the bridge is the heart of the designcode checking process. To verify the adequacy at key positions, total design load effects need to bedetermined. Combinations of the effectsdue to the various loads, each multipliedby its appropriate partial factor, shouldbe clearly set out in tabular form toavoid errors.

Note that the weight of the wet concreteis carried by the girders alone, not bythe composite section. The calculationof moments and stresses must therefore be made separately for the twostages of construction and the effectsadded. For a simple design it is notnecessary to consider a succession ofstages representing sequential castingof the slab.

The design resistances must bedetermined for key positions in the structure. The Eurocodes provide rulesfor determining the various resistances

Figure 4: Examples of typical grillages

(a) Orthogonal grillage

Span 1 Span 2 Span 3

Main beamMain beamMain beamMain beam

(b) Grillage for spans with small skew (<20º)

(c) Grillage for spans with large skew (>20º)


25

of steel and composite sections, and forthe design of the reinforced concretedeck slab. For detailed guidanceappropriate to student project designsrefer to the simplified version of theEurocodes (see Ref.1 on page 31).

The method of determining resistance isto treat each local area of the bridgeseparately and assess separately thebending and shear strengths. Allowanceis made in the rules for buckling, both ofthe beam members and of web panelsin shear. The Eurocodes and thesimplified version contain tables, figuresand formulae for this purpose. Wherethere is interaction, for example betweenshear and bending, interactionrelationships (limiting values for each incombination with the other) are defined.

Certain beam cross sections candevelop plastic bending resistance,which is greater than elastic bendingresistance. A classification system isgiven for deciding when plastic

resistance may be relied upon (referredto as ‘class 2’ cross sections). Theclassification depends on the actualproportions of the steel elements thatare in compression, but as a roughguide it may be noted that saggingregions of composite beams will usuallybe class 2, whilst hogging regions ofcomposite beams will in a lower class. Itshould be noted that whilst the finalcomposite section may be class 2, thesteel girder alone (i.e. before the slab iscast) may be a lower class.

Special details such as bearing stiffenersare covered by rules that determine boththe share of forces carried by theparticular detail and its resistance.

For each element or part, the designresistance must at least equal thedesign load effects. Where this is notachieved, the design must be modifiedby increasing the flange size, reducingthe spacing of bracing or in some otherappropriate manner and rechecked.

Above: Chieveley, A34/M4 Junction 13,England

Material properties and specifications

26

For structural use in bridges, steelproducts (plates, hot rolled sections andtubes) are cut to size and welded. In thestructure, the material is subject totensile and compressive forces. Thesteel generally responds in a linearelastic manner, up to a ‘yield point’, andthereafter has a significant capacity forplastic straining before failure.

Steel derives its mechanical propertiesfrom a combination of chemicalcomposition, mechanical working andheat treatment. The chemicalcomposition is essentially a balancebetween achieving the required strengththrough alloy additions, whilstmaintaining other properties (i.e. ductility,toughness and weldability). Mechanicalworking is effectively rolling the steel; themore steel is rolled, the stronger itbecomes, but this is at the expense ofductility. ‘Heat treatment’ covers thecontrol of cooling as the steel is rolled,as well as reheating and coolingprocesses that can be employed toinfluence a range of material properties.

All new structural steel for bridges is‘hot-rolled’ to one of the followingEuropean standards.• BS EN 10025-2

Non-alloy steels• BS EN 10025-3 & 4

Fine grain steels• BS EN 10025-5

Weather resistant steels• BS EN 10025-6

Quenched and tempered steels • BS EN 10210

Structural hollow sections

In these material standards, thedesignation system uses the prefix “S”to denote structural steels, followed by athree digit reference that corresponds tothe specified minimum yield strength (inN/mm2). Various other letters andnumerals may be appended to indicateother properties or manufacturingprocess routes. The most commonlyspecified steel for bridges is gradeS355J2+N to BS EN 10025-2; the “J2”indicates a certain level of toughnessand “+N” indicates the process route(i.e. which combination of heattreatment and rolling are used).

The principal properties of interest to thedesigner are:• Yield strength• Ductility• Toughness• Weldability

Yield strengthThe yield strength is the mostsignificant property that the designerwill need to use or specify. The strengthgrades covered by the materialstandards include; S235, S275, S355,S420 and S460, all of which relate tothe strength of material up to 16mmthick. Yield strength reduces slightlywith increasing plate thickness, but forstudent design projects, the basicnominal yield strength may be assumedirrespective of thickness.

Steels of 355 N/mm2 yield strength arepredominantly used in bridgeapplications in the UK because the cost-to-strength ratio of this material is

4 Material properties and specifications

Steel derives its mechanical properties from acombination of chemical composition, mechanicalworking and heat treatment.

Above: Hallen Rail Bridge, Avonmouth, England.

Opposite: Jackfield Bridge, Shropshire, England.

27

lower than for other grades. Higherstrength steels may offer otheradvantages, but they are less readilyavailable and the additional strength isof little benefit if fatigue or maximumdeflection governs.

DuctilityDuctility is a measure of the degree towhich the material can strain or elongatebetween the onset of yield and theeventual fracture under tensile loading.Good ductility offers the ability toredistribute localised high stresseswithout failure and to develop plasticmoment capacity of sections. Whether itis appreciated or not, the designer relieson ductility for a number of aspects ofdesign and fabrication. It is therefore ofparamount importance to all steels instructural applications.

Notch toughnessThe nature of steel material is that itcontains some imperfections, albeit ofvery small size. When subject to tensilestress these imperfections tend to open.If the steel were insufficiently tough, the‘crack’ would propagate rapidly, withoutplastic deformation, and failure wouldresult. This is called ‘brittle fracture’ andis of particular concern because of thesudden nature of failure. The toughnessof the steel, and its ability to resist thisbehaviour, decreases as the temperaturedecreases. The requirement fortoughness increases with the thicknessof the material. Hence, thick plates incold climates need to be much tougherthan thin plates in moderate climates.

Toughness is specified by requiringminimum energy absorption in a CharpyV-notch impact test, which is carriedout with the specimen at a specified(low) temperature and the requirementis given as part of the gradedesignation. For typical bridgesteelwork, to BS EN 10025-2, the usualdesignation letters are J0, J2 or K2 (inincreasing level of toughness).

BS EN 1993-1-10 describes therequirements for notch toughness in theform of a table, which gives a limitingthickness of a steel part, depending onthe reference temperature, the steelgrade (yield strength and toughness),and the stress in the element. From thistable, the limiting thicknesses for a typical UK bridge (using grade S355steel, at a reference temperature of -20oC with a tensile stress under thedesign loading at that temperature of 75 per cent of the yield strength)would be approximately:

Toughness subgrade Limiting thickness(BS EN 10025) (mm)J0 35

J2 50

K2 60

If steels thicker than 60mm are needed,other grades, to BS EN 10025-3 & 4,would be needed.

WeldabilityAll structural steels are essentiallyweldable. However, welding involveslaying down molten metal and localheating of the steel material. The weldmetal cools quickly, because thematerial offers a large ‘heat sink’ and theweld is relatively small. This can lead tohardening of the ‘heat affected zone’ ofmaterial adjacent to the weld pool andto reduced toughness (often calledembrittlement). The greater thethickness of material, the greater thereduction of toughness.

The susceptibility to embrittlement alsodepends on the quantity and nature ofthe alloying elements, principally thecarbon content. This susceptibility canbe expressed as the ‘Carbon EquivalentValue’ (CEV), and the material standardsgive an expression for determining thisvalue. The higher the CEV, the moredifficult it is to weld.

Weld procedure specifications are drawnup that set out the necessary weldingparameters for any particular steel gradeand weld type, to avoid embrittlement.For the purposes of a student project, itmay be assumed that any thickness ofstructural steel to the standardsmentioned above are weldable.

Corrosion protection

28

In the UK, the ‘design life’ of a newbridge is usually taken to be 120 years.Because ordinary structural steel willrust if exposed to the elements,corrosion protection is an importantissue to consider when designing anddetailing steel bridges. Corrosionprotection is usually achieved by theapplication of coatings (although thereare some alternatives) which, withsuitable maintenance, are capable ofachieving the required design life.Various coating systems are currentlyavailable, and paint technology isadvancing at a rapid pace with lives tofirst major maintenance in excess of 30 years anticipated for the latest systems.

The following remarks provide a generalintroduction, but for more detailedadvice on the corrosion protection ofsteel bridges, (see Ref.6 on page 31).

Paint coatingsConventional painting systems involvethe application of several coats –typically a primer, undercoats and afinishing coat, usually by spraying.Before painting, the steel surface mustbe abrasive blast cleaned to remove millscale, dirt, etc. and to achieve a suitablestandard of surface cleanliness andprofile to which the paint can adhere.Various paint systems are specified byhighway and railway authorities, basedupon the environment, accessibility formaintenance, and the life untilmaintenance of the coating is necessary.

Recently, high-build paint systemsdeveloped for the offshore industryhave been introduced to bridgeconstruction. These systems achieve athick and durable coating in only one ortwo coat applications.

Thermally sprayed metallic coatingsA coating of aluminium can be appliedby heating the metal (in wire form) in aspecial ‘gun’ that sprays the molten metal onto the steel surface. Thermallysprayed aluminium coatings have beenapplied for many years to provide long-term corrosion protection to steelbridges. The aluminium acts as a barrierand is usually over-painted to form a‘duplex’ coating system (see Figure 5,page 29). Such ‘duplex’ systems arefrequently specified for HighwaysAgency and Railtrack bridges, becausethey provide a high level of corrosionprotection, and long life to first majormaintenance.

Hot-dip galvanizingHot dip galvanizing is a process wherethe steel component to be coated isimmersed in a bath of molten zinc andthen withdrawn. The steel surfaces areuniformly coated with zinc, which ismetallurgically bonded to the structuralsteel. The zinc weathers at a slow rategiving a long and predictable life. Inaddition, if any small areas of steel areexposed (say through accidentaldamage), then the coating providesgalvanic (sacrificial) protection bycorroding preferentially. However, thereare limitations on the size ofcomponents that can be galvanized dueto the size of the zinc bath and there arepotential complications whengalvanizing welded fabrications.

5 Corrosion protection

Corrosion protection is an important issue to considerwhen designing and detailing steel bridges.

Top: Hot-dip galvanised steel bridge, (Photo courtesy of Forestry Civil Engineering)Scotland.

Above: Hardy Lane Bridge, (Example ofenclosure system), Gloucestershire, England.

Opposite page: Shanks Millennium Bridge,(Example of weathering steel), Peterborough,England.

29

Weathering steelsWeathering steel is a low alloy steel thatforms an adherent, protective oxide filmor ‘patina’ that, in a suitableenvironment, seals the surface andinhibits further corrosion. Weathering steelbridges do not require painting. Periodicinspection and cleaning should be theonly maintenance required to ensure thebridge continues to perform satisfactorily.Weathering steel bridges are ideal whereaccess is difficult or dangerous and wherefuture disruption needs to be minimisedbut they are not suitable in saltyenvironments, such as near the coast. Forfurther details on weathering steel bridges(see Ref.5 on page 31).

Enclosure systemsEnclosure systems offer an alternativemethod of protection for the structuralsteelwork of composite bridges, whilstat the same time providing a permanentaccess platform for inspection andmaintenance. The concept of enclosingthe structural steelwork on compositebridges is based on the fact that cleansteel does not corrode significantly ifenvironmental contaminants are absent.Nevertheless, the steel within anenclosure is usually painted, but with avery modest system. Typically,enclosures are formed from light weightdurable materials such as GRP.However, enclosures have not beenwidely used in the UK as they haveproved to be relatively expensive.

Two Pack Polyurethane Finish

HB Epoxy MIO Undercoat

HB Zinc Phosphate Epoxy Undercoat

Sealer Coat

Sprayed Aluminium

Steel SubstrateBlast Cleaned: Sa 3

50µm

150µm

100µm

100µm

Total (Paint)300µm Min.

Siteapplied

Shopapplied

Figure 5: Schematic cross-section through atypical modern high performance paintsystem.

Concluding remarks

6 Concluding remarks

Steel is an ideal material for bridges, and is widely usedfor all forms of bridge construction around the world.

Steel is an ideal material for bridges.The many advantages of steel have ledto it being widely used for all forms ofbridge construction around the world,from simple beam bridges up to thelongest suspension bridges. However,its most widespread use in the UK overrecent years has been on steelcomposite highway bridges.

Composite construction is aneconomical and popular form ofconstruction for highway bridges. Itcombines high quality, factory-madeproducts (the steel girders) with a castin-situ reinforced concrete deck slab,utilising each element where it is mosteconomic. It is appropriate for the greatmajority of spans, from 13m up to 100mor more.

Familiarity with the method ofconstruction, an understanding of thepart each element plays and the

interaction of the elements is to beencouraged as a means to goodeffective design.

The design principles for a compositebridge are quite straightforward andreadily understandable. The codified requirements of the design process aremore complex, reflecting the fact thatthe structural behaviour of a bridgeinvolves the interaction of manydifferent effects. However, theessentials have been condensed into a‘user-friendly’ document suitable forstudents (see Ref.1 on page 31).

By following the steps in this elementaryguide and using the simplifiedEurocodes document, students shouldbe able to produce a basic project typedesign. In doing so, they should acquirevaluable experience which can leadeasily into full-scale design.

Below: Swansea Sail Bridge, Swansea, Wales.

Opposite above: Puente del Alamillo, Seville, Spain.

Opposite below: Puente de la Barquetta,Seville, Spain.

30

31

Design code checks forundergraduate projects1 Bridge design to the Eurocodes –

Simplified rules for use in student projects

Published by SCI the Steel ConstructionInstitute, this document contains‘simplified versions’ of sections of theStructural Eurocodes that are relevant tothe design of composite highwaybridges for the use of undergraduatestudents. It is written to explain both theEurocode code provisions andbackground concepts at easilyunderstood levels. It is emphasised thata bridge designed to this simplifiedversion will not necessarily meet all themore detailed requirements of theEurocodes, but it will provide areasonable solution for undergraduatedesign projects.

Corus BrochuresA number of bridge related publications are available from Corus,giving introductory information on arange of issues. These may bedownloaded in ‘pdf’ format fromwww.corusconstruction.com

2 Bringing steel to life – A comprehensive range of bridgerelated products and services

3 Composite steel highway bridges4 The design of steel footbridges5 Weathering steel bridges6 Corrosion protection of steel

bridges

Steel Construction InstitutepublicationsFor more detailed guidance on thedesign of steel bridges in fullaccordance with BS 5400, the SCI havea range of publications including:

7 Composite highway bridges –design to the Eurocodes (P356)

8 Composite highway bridges –worked examples using theEurocodes (P357)

9 Design guide for composite boxgirder bridges (P140)

10 Steel Bridge Group, Guidance noteson best practice in steel bridgeconstruction (P185)

Other publications11 Steel Bridges, The practical aspects

of fabrication which influenceefficient design (published by BritishConstructional SteelworkAssociation, 2002)

12 BSCA Guide to the erection of steelbridges (published by BritishConstructional SteelworkAssociation, 2005)

7 References and further reading

References and further reading

www.corusgroup.com

Care has been taken to ensure that thisinformation is accurate, but Corus Group Ltd,and its subsidiaries, does not acceptresponsibility or liability for errors orinformation which is found to be misleading.

Copyright 2007Corus

Designed by Orchard Resourcebase

Corus Construction Services & DevelopmentPO Box 1Brigg RoadScunthorpeNorth LincolnshireDN16 1BPT: +44 (0) 1724 405060F: +44 (0) 1724 404224E: [email protected]

English version

Corus Student Guide to Steel Bridge Design

Documents

Transcript of Corus Student Guide to Steel Bridge Design