Engineering and Construction

Analysis - Tower Structures.

In the analysis of towers the largest uncertainty is accurate knowledge of the wind loads. Highly sophisticated methods of analysis cannot improve this.A static linear three-dimensional structural analysis is sufﬁcient for almost all lattice tower structures.

For transmission towers, line break conditions can also be critical. Line breakage will in general induce dynamic loads in addition to any residual static loads.Detailed loading is outside the scope of this section.

For lattice towers with large complicated panel bracing, the secondary bracing forces can be signiﬁcantly altered by non-linear effects caused by curvature of the panels under the inﬂuence of the design loads. Generally the rules in the codes are sufﬁcient, but where structures are of particular importance or where there is much repetition of a design, a non-linear analysis may be necessary.

Dynamic analyses of self-supporting lattice towers are rarely necessary unless there are special circumstances such as high masses at the top, use as a viewing platform, or circular or almost solid sections of mast which could be responsive to vortex shedding or galloping. Knowledge of the dynamic response is also necessary for assessment of fatigue of joints if this is signiﬁcant. For guyed towers the non-linear behaviour of the guys is a primary inﬂuence and cannot be ignored. The choice of initial tension, for example. can have a very great effect on the deﬂections (and dynamic behaviour).

The effects of the axial loads in the mast on column stiffness can be signiﬁcant.Methods of static analysis are given in the main international codes for the design of guyed towers. Guyed towers can also be particularly sensitive to dynamic wind effects especially those with cylin- drical or solid sections.

Environmental Loading - Tower Structures.

The primary environmental loads on tower structures are usually due to wind and ice, sometimes in combination. Earthquakes can be important in some parts of the world for structures of high mass, such as water towers. Loading from climatic temperature variations is not normally signiﬁcant but solar radiation may induce local stresses or cause signiﬁcant deﬂections, and temperatures can inﬂuence the choice of ancillary materials.

Most wind codes use a simple quasi-static method of assessing the wind loads, which has some limitations for calculating the along-wind responses but is adequate for the majority of structures. Tower structures with aerodynamically solid sections and some individual members can be subject to aeroelastic wind forces caused by vortex shedding, galloping, ﬂutter and a variety of other mechanisms which are either poorly covered or ignored by current codes of practice. Such factors have been responsible for more tower collapses and serviceability failures worldwide than any shortfall in resistance to along-wind loads.

Most national and international design codes now specify wind loads in terms of design wind speeds, either mean hourly or gust, that will recur on average once in a 50 year period (i.e. with an annual probability of 2%). Guidance is sometimes given on wind shape factors for typical sections and lattices. Consideration of dynamic response to the wind is not always covered in depth and there is still a mixture of limit state and working stress codes. BS 81001 is a recent code in a limit state format speciﬁcally written for towers. Wind loads are speciﬁed in terms of a ‘50 year return’ mean hourly wind pressure together with gust factors which convert the forces to an equivalent static gust.The overall wind forces calculated using BS 8100 are substantially similar to those that would be obtained using earlier codes such as CP32 but forces near the tops of towers are relatively higher due to an allowance for dynamic response. The code also gives guidance on means of allowing for the importance of particular structures by adjusting the partial factor on the design wind speed.

Guidance is limited for structures that have a signiﬁcant dynamic response at their natural frequencies, and gust factors for guyed towers are speciﬁcally excluded from the scope of the code. Part 4 of BS 8100 is intended to address these aspects.

The inﬂuence of height and topography on wind speed can be signiﬁcant; this is covered in some detail in both codes. Ice loads and types of ice are also covered but neither mentions the very signiﬁcant inﬂuence of topography on the formation of ice. This has not yet been subject to systematic study but some hill sites are known to be subject to icing well in excess of the code requirements. The combination of wind and ice loads is even less well understood although some guidance is given.

Towers - Structural Types.

Steel towers can be constructed in a number of ways but the most efﬁcient use of material is achieved by using an open steel lattice. Typical arrangements for microwave radio and transmission towers are shown in Fig. 5.1. The use of an open lattice avoids presenting the full width of structure to the wind but enables the construction of extremely lightweight and stiff structures.Most power transmission, telecommunication and broadcasting structures fall into this class.

Fig. 5.1 Lattice towers: (a) microwave tower. (b), (c) and (d) transmission towers

Lattice towers are typically square or triangular and have low redundancy. The legs are braced by the main bracings: both of these are often propped by additional secondary bracing to reduce the effective buckling lengths.The most common forms of main bracing are shown in Fig. 5.2.

Fig. 5.2 Main bracing arrangements

Lattice towers for most purposes are made of bolted angles. Tubular legs and bracings can be economic, especially when the stresses are low enough to allow relatively simple connections. Towers with tubular members may be less than half the weight of angle towers because of the reduced wind load on circular sections.

However, the extra cost of the tube and the more complicated connection details can exceed the saving of steel weight and foundations.

Connections are usually arranged to allow site bolting and erection of relatively small components. Angles can be cut to length and bolt holes punched by machines as part of the same operation. Where heavy-lift cranes are available much larger segments of a tower can be erected but often even these are site bolted together.

Guyed towers provide height at a much lower material cost than self-supporting towers due to the efﬁcient use of high-strength steel in the guys. Guyed towers are normally guyed in three directions over an anchor radius of typically 2 – 3 of the tower height and have a triangular lattice section for the central mast. Tubular masts are also used, especially where icing is very heavy and lattice sections would ice up fully.

A typical example of a guyed tower is shown in Fig. 5.3.

The range of structural forms is wide and varied. Other examples are illustrated in Figs 5.4 and 5.5. Figure 5.4 is a modular tower arrangement capable of extension for an increased number of antennas.The arrangement shown in Fig. 5.5 is adopted for supporting ﬂare risers where maintenance of the ﬂare tip is carried out at ground level.

A signiﬁcant inﬂuence on the economics of tower construction is the method of erection, which should be carefully considered at the design stage.

Fig. 5.3 Guyed tower

Fig. 5.4 Modular tower

Fig. 5.5 Flare tower

Initial Design: Cable-Stayed Bridges.

Cable-stayed bridges virtually behave as continuous beams with elastic supports at the cable anchorage points. Cable lengths are adjusted during construction so that the effect of cable extension under dead load conditions is cancelled out. Provided that high tensile material is used for the cables then non-linear effects due to self- weight can be shown to be negligible.

For initial design the deck may be proportioned as a continuous beam, as shown in Fig. 4.16, to which are added:

(a) deck deﬂections at the cable positions due to extension.
(b) axial forces due to horizontal component of the cable forces.

Where a single plane of cables is used then an initial design check must be made to ensure that sufﬁcient torsional rigidity is provided (using single or twin box girders) to control transverse tilt of the deck at mid-span when trafﬁc loading occupies one carriageway.

Fig. 4.16 Cable-stayed bridges, initial analysis. (a) Anchored side spans (harp shown); (b)
harp; (c) fan

Initial Design - Suspension Bridges.

For initial design Fig. 4.15 gives approximate formulae for making ﬁrst estimates of cable size and bending of the stiffening girders. The most severe condition for the stiffening girder is with approximately one half of the main span loaded asymmetrically

Fig. 4.15 Suspension bridges, approximate formulae (Pugsley). (a) Neglecting stiffness of
deck; (b) including deck stiffness

Steel Grades - Bridges - Weathering Steel.

Steels to BS EN 10025 of grade S355 are usual for bridges as they offer a better cost-to-strength ratio than grade S275.All parts subject to tensile stress are required to achieve a speciﬁed notch toughness, depending upon design minimum temperature, stress level and material thickness.

Special consideration may be needed if tensile stresses occur only during erection, for example arising from lateral bending in girders due to wind loading, or during lifting of components. Judgement is necessary but it is often accepted that modest values of tensile stress can be permitted in such cases without the need for speciﬁc notch toughness provided that the work is not carried out during very low temperatures. In all other cases appropriate grades having speciﬁed notch toughness are necessary.

Weathering steel
To eliminate the need for painting, weathering grades in BS EN 10155 should be considered. Although it can be shown that the commuted cost of repainting steel bridges is not of great signiﬁcance compared with the initial bridge cost, weathering steel is particularly useful in eliminating maintenance where access is difﬁcult – over a railway and increasingly so over highways.Weathering steel is not suitable at or near the coast (i.e. within about 2km of the sea). The Highways Agency requires sacriﬁcial thickness to be added to all exposed surfaces for possible long-term corrosion of 1.5mm per face in a severe industrial environment, 1mm otherwise.

Pedestrian Bridges.

A minimum clear deck width of 1.8m is usual, increased to 3.0m or 4.0m in busy areas or if a cycleway is also present. Steel provides an efﬁcient solution because the entire cross-section including parapets may be fabricated and erected in one unit. For this reason multiple spans tend to be simply supported and not continuous. Staircase and sloping or stepped ramp spans are also erected as complete units, and columns are often of steel.

Half-through cross-sections are often used because the shallow construction depth is able to provide the shortest lengths of staircase or ramp approaches in urban areas. Half-through rolled beams, Warren truss or Vierendeel girders are used with rolled hollow section (square or rectangular) members. Floors are steel stiffened plate, often surfaced with a factory-applied epoxy non-slip surfacing ap- proximately 5mm thick. Rigid connections between ﬂoor and girders provide for U-frame stability. Staircase approaches may be either steel stringers supporting steel plate treads or a central spine box with cantilever treads of steel or precast concrete. Ramps may be similarly formed, but where the span exceeds about 10m then half-through construction tends to be used. Spiral approach ramps can be used, formed in steel using twin rolled sections curved in plan supporting a composite or steel plate ﬂoor.Where adequate construction depth is available, for example when a pedestrian bridge is required across a motorway or railway cutting, then deck cross-sections are appropriate with a composite or steel plate ﬂoor. Precast concrete ﬂoor units are also suitable.
Primary members may be a single box girder, twin plate girders or twin rolled beams.A single box girder provides a structure of neat appearance.

Fig. 4.12 Pedestrian bridges. (a) Twin universal beam; (b) box; (c) Vierendeel or Warren
girder; (d) half-through universal beam; (e) half-through plate girder; (f) through
lattice

Railway Bridges – Girder Type.

Many existing small-span railway bridges weather cast or wrought iron girders to which rails are ﬁxed directly without use of ballast or half-through girders with trough ﬂoors.As these bridges have reached the end of their lives they have been replaced in steelwork in modern form with ballasted track. The legacy of the original decks with their very shallow construction depth has inﬂuenced modern underline bridge practice in replacements and new bridges.The majority of railway underline bridges therefore tend to be of half-through type. The direct fastening of track without ballast gives the cheapest possible form of railway bridge and is appro- priate for rapid transit or tramway bridges where speeds are not high.

Fig. 4.11 Railway bridges. (a) Half-through plate girders; (b) half-through box girders; (c)
composite deck type (spans > 30m)

Girder depths are generally greater than for highway bridges due to the heavier loading and because limits for deformation are necessary. A span-to-girder depth ratio of 12 to 15 is typical. For short spans half-through plate girders are used with composite steel cross girders forming rigid U-frames and supporting a concrete ﬂoor with ballasted single or double track.Through or half-through construction is appro- priate for medium spans exceeding 50m using trusses. For spans up to 39m, the railway authorities use a standard box girder design with a steel ribbed ﬂoor of minimal depth which is achieved by spanning between the inner webs of trapezoidal box girders proportioned so as to ﬁt closely within the station platform space. The type has advantages in using components entirely of steel,which are bolted together at site and commissioned during temporary possession of existing tracks.Where suf- ﬁcient depth is available then deck construction is preferable and more economic, with either twin plate girders or a box girder beneath each rail track.

Simply-supported spans are widely used for railway bridges because:

(a) individual spans can be erected or replaced quickly during temporary track possession,
(b) uplift is more likely to occur if spans are unequal under heavy railway loading,
(c) fatigue is potentially less critical. Fatigue still tends to govern the design of steel elements having spans less than about 24m. Thus cross girders, railbearers and other short-span members need to use lower working stresses.

Highway Bridges – Composite Deck Construction.

Composite deck construction (Fig. 4.9) should be used wherever the construction depth will permit. If possible, multiple spans should be made continuous over the intermediate supports, so reducing the number of bearings and expansion joints.

Continuity gives economies throughout the structure and reduces trafﬁc disruption arising from the maintenance needs of these vulnerable elements. A number of options are available for maintaining continuity over intermediate supports. If ground conditions are poor such that the predicted differential settlement of the supports is signiﬁcant (say, exceeding (span/1000)) then to avoid overstress the structure should be made statically determinate by use of simply-supported spans.

Fig. 4.9 Highway bridges – composite deck construction

Cantilever and suspended spans are alternative options which retain some of the advantages of continuity. A girder depth of (span/20) (girder depth excludes ﬂoor slab) is generally economic although shallow girders can be used down to a depth of (span/30) or less.

Rolled sections are appropriate for short spans up to 25m span. For continuous spans exceeding about 22m, fabricated plate girders will show economy because lighter ﬂanges and webs can be inserted in the mid-span regions.Automated manu- facture of plate girders means that they are highly economic when compared with box girders. Normally a girder spacing of 2.5–3.5m is optimum with a ﬂoor slab of about 220–250mm thick. Edge cantilevers should not exceed 50% of the beam spacing and to simplify falsework should where possible be less than 1.5m.An even number of girders (i.e. 2,4, 6, 8, etc.) achieves better optimization for material ordering and permits girders to be braced in pairs for erection.

For medium spans exceeding 40m where adequate construction depth is available it may be economic to use twin girders only. A number of variants are available as shown in Fig. 4.9, using a thickened haunched slab, longitudinal stringers and cross girders to support the slab intermediately. For narrow bridges the complete precasting of composite ﬂoors may offer advantages in speed of construction. Use of girders with a curved sofﬁt becomes economical for medium spans exceeding 45m and efﬁciently achieves maximum headroom if required over the central portions of a span (see Fig. 4.9). Plate girder ﬂanges should be proportioned so as to be as wide as possible consistent with outstand limit to reduce the number of intermediate bracings. For practical reasons a desirable minimum ﬂange width is about 40mm to accommodate shear connections and to permit the possible use of permanent formwork. A maximum ﬂange thickness of 75mm is recommended as a guide to avoid heavy butt welds, but thicker ﬂanges can be used where necessary, to 100mm or greater.

For long spans, box girders (see Fig. 4.10) are more suitable than plate girders, for which ﬂange sizes would be excessive. Other reasons for using box girders for long spans may include a need to improve aerodynamic stability, the presence of severe plan curvature, a requirement for single column supports or very limited construction depth.However, for short and medium spans box girders are generally less economic because, although a reduction in ﬂange sizes may be possible due to superior load distribution properties, this is more than offset by the amount of internal diaphragms and stiffening and the extra costs in manufacture. Fabrication costs are higher because the assembly and welding processes are less amenable to automation than with plate girders. Also access must be permitted inside box girders for welding, protective treatment processes, and permanent inspection. However, erection of box girders is often easier because they require minimal external bracing to maintain overall stability.Multiple compact section box girders can be economic for spans of up to 50m in particular situations, and enable longitudinal stiffeners to be eliminated.Open-top trapezoidal box girders (known as ‘bath tubs’) are widely used in North America and possess some of the advantages of plate girders. They have seen same use in the UK but temporary bracing is required during construction to maintain shape and relative twist of the sections until the concrete slab is placed and the full torsional rigidity achieved.

Fig. 4.10 Highway bridges – box girders. (a) Twin box and cross girders (spans 40–150m);
(b) open top box (spans 40–100m); (c) multiple box (spans 30–60m)

Bridges - Girder Type - Structural Steel.

Girder bridges (Fig. 4.8) predominate over the previously described types for short and medium spans, and generally provide the most economic solution. For highway bridges composite deck construction is generally used unless the depth is very critical in which case half-through girders or through trusses may be necessary. Railway bridges frequently require to be of half-through or through form because of depth limitations; these are more suitable, being generally narrower. Pedestrian bridges are similarly suited to a half-through form so as to reduce the length of staircases and ramps, which can often exceed the length of the span itself. Highway, railway and pedestrian bridges of girder type for short and medium spans are further described below.

1 Highway bridges – composite deck construction Composite deck construction (Fig. 4.9) should be used wherever the construction depth will permit. If possible, multiple spans should be made continuous over the...

2 Railway bridges – girder type Many existing small-span railway bridges weather cast or wrought iron girders to which rails are ﬁxed directly without use of ballast or half-through girders with trough ﬂoors.As...

3 Pedestrian bridges A minimum clear deck width of 1.8m is usual, increased to 3.0m or 4.0m in busy areas or if a cycleway is also present. Steel provides an efﬁcient solution because the entire cross-section...

Fig. 4.8 Girder type bridges. (a) Single span; (b) simply-supported spans; (c) cantilever and
suspended span; (d) continuous; (e) continuous – curved sofﬁt

Truss Bridges - Structural Steel.

Through trusses are used for medium spans where a limited construction depth precludes use of a composite deck bridge.They are suitable in ﬂat terrain to reduce the height and length of approach embankments and for railway bridges where existing gradients cannot be modiﬁed.A truss may be unacceptable visually; a bowstring truss is an alternative solution. For short spans and medium spans up to 50m, trusses are generally less economic than plate girders because of higher fabrication cost.

They are therefore adopted only where the available construction depth is not sufﬁcient for composite beams. (See Fig. 4.7.)

Fig. 4.7 Truss bridge: (a) elevation (b) cross section (c) arrangements for sliding into position.

Portal Frame Bridges - Structural Steel.

Portal frame bridges (Fig. 4.6) are mainly suitable for short or medium spans. In a three-span form with sloping legs they can provide an economic solution by offering a reduction in span and have an attractive appearance.The risk of shipping collision with sloping legs must be considered for bridges over navigable rivers. Portal bridges tend to be less economic than beam bridges because the foundations must be designed to resist horizontal thrust, complex details can be required at the joints and erection is more complicated.

Fig. 4.6 Portal frame bridges. (a) Three-span inclined legs; (b) single span

Arch bridges - Structural Steel.

Arch bridges (Fig. 4.5) are suitable in particular site conditions. An example is a medium single span over a ravine where an arch with spandrel columns will efﬁciently carry a deck with the horizontal thrust taken directly to rock.A tied arch is suitable for a single span where construction depth is limited and presence of curved highway geometry or other obstruction to the approaches conﬂicts with the back stays of a cable-stayed bridge.

Fig. 4.5 Arch bridges. (a) Spandrel post arch; (b) tied arch; (c) part tied arch

Cable-stayed bridges - Structural Steel.

Cable-stayed bridges (Fig. 4.4) are of a suspension form using straight cables which are directly connected to the deck.The structure is self-anchoring and therefore less dependent upon good foundation conditions, but the deck must be designed for the signiﬁcant axial stress from the horizontal component of the cable forces. The construction process is quicker than for a suspension bridge because the cables and deck are erected at the same time and the amount of temporary works is reduced.

Fig. 4.4 Cable-stayed bridges. (a) Harp; (b) fan; (c) single tower; (d) single plane; (e) single
plane; (f) twin plane

Either twin sets of cables are used or alternatively for dual carriageways a single plane of cables and tower can be located in the central reserve space. Two basic forms of cable conﬁgurations are used, either ‘fan’ or ‘harp’.A fan layout minimizes bending effects in the structure due to its better triangulation but anchorages can be less easy to incorporate into the towers. The harp form is often preferred where there are more than, say, four cables. The number of cables depends on the span and cable size, which is often selected such that each fabricated length of deck (say, 20m) contains an anchorage at one end to suit a cantilever erection method.Bridges either have two towers and are symmetrical in elevation or have a single tower as suited to the site.

Floors are generally an orthotropic steel plate but composite slabs can be used for spans up to about 250m. A box girder is essential for bridges having a single plane of cables to achieve torsional stability, but otherwise either box girders or twin plate girders are suitable. Aerodynamic oscillation is a much less serious problem than with suspension bridges but must be considered. Some bridges with plate girders incorporate non-structural aerodynamic edge fairings.

It is essential to use cables of maximum strength and modulus at a high working stress so that sag due to self-weight, which produces non-linear effects, is negligible.

Cables are normally of parallel wires or prestretched locked coil wire rope. During erection the cable lengths are adjusted or prestressed so as to counteract the dead load deﬂections of the deck arising from extension of the cables.

Suspension Bridges - Structural Steel.

Suspension bridges (Fig. 4.3) are used for the longest spans across river estuaries where intermediate piers are not feasible. The cables form catenaries supporting both sides of the deck and are tied to the ground usually by gravity foundations sometimes combined with rock anchors. Thus ground conditions with ﬁrm strata at or close to the surface of the ground are essential. Towers are usually twin steel or concrete box members which are braced together above the roadway level. They are designed so as to be freestanding under wind loading during construction until the cables are installed.Cables are either a compacted bundle of parallel high tensile steel strands (commonly 5mm diameter) installed progressively by ‘spinning’ or may be formed from a group of wire ropes. Deck hangers are wire ropes (or round steel rods for light loading as for a footbridge) clamped to the cable and connected to the deck at a spacing equal to the length of each deck unit erected, typically 18m.The construction process for suspension bridges is more time consuming than for other types because the deck cannot be installed until the towers, anchorages, cable and hangers are constructed.

Suspension bridges. (a) Three-span suspended; (b) straight back stays, viaduct side spans; (c) straight back stays, no side spans; (d) truss deck; (e) aerofoil box deck; (f) inclined hangers

Fig. 4.3 Suspension bridges. (a) Three-span suspended; (b) straight back stays, viaduct side
spans; (c) straight back stays, no side spans; (d) truss deck; (e) aerofoil box deck; (f)
inclined hangers

Depending upon ground conditions, the cables can be catenaries supporting side spans. Cables may alternatively be straight from tower top to the ground anchorages and merely support a main span, side spans being non-existent or formed as short-span viaducts. Decks are either trusses with a steel orthotropic plate ﬂoor spanning between or an aerofoil box girder. Footways are often cantilevered outside the two sets of cables.

Aerodynamic behaviour must be considered in design because of the tendency for the deck and cables to oscillate in ﬂexure and torsion under ‘vortex shedding’ and other wind effects. This is due to the ﬂexible nature and light weight of suspension bridges illustrated by the collapse of the USA Tacoma Narrows Bridge in 1940, which had a very ﬂexible narrow deck consisting of twin plate girders forming a torsionally weak deck of ‘bluff’ shape prone to wind vortex shedding. Aerodynamic considerations usually justify wind tunnel testing of models. The advantage of an aerofoil box girder such as used on the Severn and Humber bridges is that it discourages vortex shedding and reduces the wind forces to be resisted by the towers and substructures. The Severn (1966) and Bosporus Bridge (1970) use inclined hangers that form a truss system which helps to reduce any tendency to oscillate. Suspension bridges behave as non-linear structures under asymmetric deck loading so that deﬂections may be signiﬁcant. Behaviour under such loading depends upon the combined gravity stiffness and the ﬂexural rigidity of the deck or stiffening girder. The type is less suitable for heavy loading such as railway trafﬁc, especially for short spans. Suspension bridges are sometimes suitable for medium spans carrying pedestrian or light trafﬁc.

Selection of Type - Structural Steel in Bridges.

Suspension or cable-stayed bridges are suitable for the longest spans, but are less suitable to support heavy loading across short or medium spans.At the same time medium-span footbridges can appropriately be suspension, cable-stayed or arch types because concentrated loading is absent. Some of the considerations given to long spans such as aerodynamics need to be applied to footbridges. For footbridges the phenomenon of pedestrian-excited vibrations needs to be considered and tends to affect the design of spans exceeding about 25m. The possibility of horizontal oscillation needs consideration if the structure is laterally ﬂexible or is mounted on slender supports. Other bridge types such as arches or portals may be suitable in special locations. For example, an arch is a logical solution for a medium span across a steep-sided ravine.

Through trusses are suitable for medium spans where the available construction depth is limited. The vast majority of short- and medium-span highway bridges are formed with composite construction because the highway proﬁle can be arranged to suit the depth available.

For short and medium spans the most important factor which inﬂuences the type is the available construction depth.This particularly affects railway bridges because it is rarely feasible to modify existing track levels.Where depth is limited then types such as arch, truss or half-through plate girder offer an alternative solution. Types of steel bridge are shown in Fig. 4.2 with their normal economic span range and the world’s longest. Each is brieﬂy described below.

1 Suspension bridges   Suspension bridges (Fig. 4.3) are used for the longest spans across river estuaries where intermediate piers are not feasible. The cables form catenaries supporting both sides of the deck and are tied to the ground usually by gravity...

2 Cable-stayed bridges   Cable-stayed bridges (Fig. 4.4) are of a suspension form using straight cables which are directly connected to the deck.The structure is self-anchoring and therefore less dependent upon...

3 Arch bridges   Arch bridges (Fig. 4.5) are suitable in particular site conditions. An example is a medium single span over a ravine where an arch with spandrel columns will efﬁciently carry a deck with...

4 Portal frame bridges   Portal frame bridges (Fig. 4.6) are mainly suitable for short or medium spans. In a three-span form with sloping legs they can provide an economic solution by offering a reduction in...

5 Truss bridges   Through trusses are used for medium spans where a limited construction depth precludes use of a composite deck bridge.They are suitable in ﬂat terrain to reduce the height...

6 Girder type   Girder bridges (Fig. 4.8) predominate over the previously described types for short and medium spans, and generally provide the most economic solution. For highway bridges...

Fig. 4.2 Normal span range of bridge types

Selection of Span - Structural Steel in Bridges.

The majority of bridges fall within the category of short span because for many crossings of rivers, railways or secondary highways a single span of less than 30m is sufﬁcient. For multiple-span viaducts a decision on span length must be made,which depends on factors shown in Table 4.4.

For long viaducts it is necessary to carry out comparative estimates for different spans to determine the optimum choice, as shown in Fig. 4.1.

Typical optimum spans are shown in Table 4.5.

Long spans are usually adopted only when a larger number of shorter spans are precluded by restrictions of the site. This is because the material content and cost per unit length of long spans is much greater.Therefore the bridging of a wide river or estuary should generally use a number of medium spans. Only if there is a high risk of shipping collision, or if the depth and speed of water ﬂow is such as to make foundation construction very difﬁcult, should a long-span bridge be chosen.

Table 4.4 Factors which decide choice of span for viaducts

Fig. 4.1 Choice of optimum span for viaducts

Table 4.5 Typical optimum span ranges (m) for viaducts

Use of Structural Steel in Bridges.

Introduction
Use of structural steel in bridges exploits its advantageous properties of economically carrying heavy loads over long spans with the minimum dead weight. Steel is however suitable for all span ranges, categorized in Table 4.1.

For long spans steel has been the natural solution since 1890 when the Firth of Forth cantilever railway bridge, the world’s ﬁrst major steel bridge, was completed.

For short and medium spans concrete bridges held a monopoly from 1950 to 1980 because of the introduction of prestressing and precasting. Developments in steel during this period such as higher tensile strength and improved welding techniques were applied mainly to long spans.However, improvements in construction methods from 1980 have enabled steel to improve its market share within Europe and other continents to more than 50% for short and medium spans. Contributing factors to this trend are shown in Table 4.2.

Table 4.2 Factors contributing to the improved market share for steel in short and medium spans
from 1980

Where trafﬁc disruption during construction must be minimized then steel is always suitable. Most bridgework is now carried out under these conditions so that structural steel will continue as a primary choice. Steel has an advantage where speed of construction is vital; it is no coincidence that this usually results in cost economies. If rapidly erected steelwork is used as a skeleton from which the slab and ﬁnishes can be carried out without need for falsework then the advantages (summarized in Table 4.3) are fully realized.

Table 4.3 Advantages of steel bridges

For short and medium span highway bridges composite deck construction is economic because the slab contributes to the capacity of the primary members.

For continuous spans it uses the attributes of steel and concrete to best advantage. In cases where construction depth is restricted, for example in developed areas, then half-through or through construction is convenient; this is common for railway and pedestrian bridges.

For long spans, including suspension or cable-stayed bridges, all-steel orthotropic plate ﬂoors are used. Although the intrinsic costs of a steel orthotropic plate are higher (often up to about four times more) than an equivalent concrete slab, the advantage in dead weight reduction (approximately 1 : 3 ratio) may more than offset this when the overall economy is considered. Steel decks are also employed when erection must be completed in limited occupations, such as for railway bridges on existing routes. Movable bridges of swing, lift, rolling-lift (‘bascule’) or retractable type usually employ steel ﬂoors to rationalize the amount of counterweighting and

the mechanical equipment.Costs of this equipment usually exceed that of the bridge superstructure. For long spans of suspension or cabled-stayed form then special considerations affect the design including aerodynamic behaviour, the feasibility of deep foundations in estuarial conditions, cables with anchorages, non-linear structural behaviour and the absolute necessity to include the effects of the erection procedure in the design process.

This chapter on bridges gives emphasis to initial design, an important stage of the process, because the basic decisions as to member proportions, spacings and splice positions vitally affect economy of the structure. It is essential that the detailed analysis is based upon optimized sizes which are as accurate as possible. If this is
not achieved then the detailed design will be inefﬁcient because time consuming repetitive work will have been expended, adversely affecting the economy of the design and the construction costs.Guidance is given on the initial design of highway bridges using composite deck construction, which is a signiﬁcant proportion of the number of steel bridges built, although since 1990 a signiﬁcant number of railway bridges have been constructed for new-build schemes such as the Channel Tunnel

Rail Link and in replacement of life expired structures.

Steel is particularly suitable for the strengthening and repair of existing bridges arising from increases in highway trafﬁc loadings and the incidence of accidental impact damage from road vehicles. Steel is suitable for such work using welded or bolted strengthening.

Industrial - Structure in its Wider Context.

Industrial structural steelwork is inherently inﬂexible, being purpose-designed for a particular function or process, and indeed often being detailed to suit quite speciﬁc items of major plant and equipment. Nevertheless, it is important to try to cater for at least local ﬂexibility to allow minor alterations in layout, upgrading or replacement of plant items.The most appropriate way to ensure this is to repeat the advice that has been given on numerous occasions already in this chapter. The designer must understand the industrial process involved and be aware of both structural and layout solutions that have been adopted elsewhere for similar processes. Previous structural solutions may not be right, but it is preferable to be aware of them and positively to reject them for a logical reason, than to reinvent the wheel at regular intervals.

General robustness in industrial buildings may be difﬁcult to achieve by the normal route of adopting simple, logical shapes and structural forms, with well- deﬁned load-paths and frequent effective bracing or other stability provisions.

Instead of these provisions, then, it is sensible to ensure that a reasonable margin exists on element and connection design. Typically, planning for a 60–80% capacity utilization at the initial design stages will be appropriate, so that even when these allowances are reduced, as so frequently occurs, during the ﬁnal design and checking stages, adequate spare capacity still exists to ensure that no individual element or joint can weaken disproportionately the overall structural strength of the building.

Blast Loadings - Industrial Buildings.

This section deals only with blast loadings from industrial processes and not with any generalized design requirements to survive blast loadings from unspeciﬁed sources or of unspeciﬁed values.

Varying requirements exist for blast loadings. Typical examples are:

(1) transformers, where the requirement is usually to deﬂect any blast away from other vulnerable pieces of plant but where frequently one or more walls and the roof are open, serving to dissipate much of the energy discharged
(2) dust or ﬁne particle enclosures, which are often wholly inside enclosed buildings.

The decisions to be made are as follows.

(1) Can the potential source of the blast be relocated outside the building altogether, in a separate enclosure?
(2) If not, can it be placed against the external wall with arrangements to have a major permanent vented area or a specially designed blow-off panel, both of which will limit loadings on the remaining structure?
(3) Where the location cannot be controlled, it must be established which direction or directions require full protection against damage and which can tolerate certain degrees of damage.

Loading data given by plant designers are usually stated in terms of peak pres- sures to be applied to projected areas in line with the potential source of the blast.

The validity of the data must be treated as being highly suspect since blast loadings are classic examples of true dynamic loading where the time-dependent response of the structure actually determines the loading that is imposed. Quoted blast pressures, which are probably derived from theoretical considerations of high-rigidity high-mass targets fully enclosing the source, may be invalid for steel structures which have tremendous capacity to deform rapidly, absorbing energy and thereby reducing and smoothing out peak blast pressures. Whatever results are obtained from analysis or calculation, it is good practice to use a grade of steel which has good ductility, a lower yield stress from which to commence ductile behaviour and a long and reliable extensibility prior to fracture.While this will ensure reasonable material behaviour, overall ductility of the structure also depends on stability against premature buckling and the ductile behaviour of connections.

Steelwork designers should acknowledge the very imprecise nature of most blast loading data, even when the potential source of the blast is precisely located and speciﬁed. They should thus direct their attention to ensuring that collapse does not occur until major deﬂections and rotations have occurred, following normal guidelines for achieving plastic behaviour.