Construction

Construction of Bridges

Construction of Arch Beam Cable-Stayed Cantilever Suspension Truss

It is one thing to design a bridge. It is another thing to build it. Planning and executing the construction of a bridge is often very complicated, and in fact may be the most ingenious (ingénieur - engineer) parts of the entire enterprise. An incomplete structure is often subjected to stresses and oscillations that would not arise after completion. The construction work is potentially a grave hindrance to existing traffic and to normal life in the area, especially when large local fabrication works have to be installed.

Even before any actual construction is done, substantial work may needed in the form of tests. Boreholes will be made to check the condition of the ground, in conjunction with any available geological maps. Records of wind speed and direction will be consulted, and new measurements made if necessary. In the case of a river or sea crossing, records of water levels and velocities will be needed. Models of the bridge or of parts may be tested aerodynamically and hydrodynamically, and of course mechanical tests will be made. Computer simulations will supplement these tests, enabling a great variety of applied forces to be investigated. There may also be investigations into the effects on people and on the natural environment. It may even be necessary to overcome opposition to the construction, from a variety of objectors. There was much opposition, for example, to the construction of the Skye bridge. In older times, ferry operators could be extremely vociferous about a bridge proposal.

This web-site includes very little material about bridge construction, because the subject is so varied and so specialist. As soon as possible, some useful links will be provided, and if possible, some simple explanations as well.

Meanwhile, let's look at the living world, to see how things are done there. Starting at the beginning, an immature form is likely to be vulnerable. A half built medieval castle must have been hard to defend. In modern times, a half built tank or aircraft is vulnerable if bombs can be dropped on the factory. In the animal world, there are many solutions.

One is to produce so many offspring that an enormous mortality rate is acceptable. All that matters is that on average, enough survive to produce an equally large next generation.

Another way is to nourish and protect young until their probability of survival is sufficiently high. From eagle chicks to tiger cubs to baby crocodiles, even the fiercest animals can be vulnerable when young. Many animals, from earwigs to apes, protect their young. Nevertheless, mortality can be very high once the young leave their parents. In the mammals, this method is taken furthest, by retaining the young inside until they are quite large. Question, why are baby mammals more round and cuddly looking than adults?

If the young have already some defence, as in vipers, they can be left alone from the start. But even there, some species retain the eggs inside until they have developed into little versions of the adult.

A common strategy to cope with immaturity is to undergo metamorphosis. By adopting very different forms, behaviours and habitats at different stages of growth, animals can inhabit ecological niches that are optimal at each stage. This is most easily seen in many insect orders, though prevalent in other classes which are less commonly seen. Sometimes there is a passive pupal stage, during which the tissues are so radically reorganized that the adult, or imago, is almost unrecognizably different from the larva. The larval stages, which are devoted mainly to eating and growing, are very often not very mobile, whereas the adult stage, or imago, is very frequently mobile, because males and females have to meet. Mating with nearby examples from the same brood is deleterious genetically, so the emerging adults make at least some attempt to move. There are species in which the females are wingless, leaving the males to do the travelling.

The bridge builder, like the mammals, has to nurse the embryo structure through difficult stages. Very often, the stresses differ considerably from those of the complete structure, and can be more concentrated. The collapse of several box girder bridges in the 1960s damaged the reputation of this type for a period, until the stresses were better understood. Note the paradox that an apparently simple structure, a set of boxes, can be hard to understand, while a complicated looking truss can be solved, at least in principle, using a set of equations. The variation in stresses during construction may be so severe that jacking must be provided. The four legs of the Tour Eiffel were provided with jack which were adjusted from time to time as the structure grew. Even after completion, a concrete structure may be subject to creep, and the ground may settle; jacks are therefore provided for later adjustment when this behaviour is foreseen.

Excavation for foundations may have to be taken to great depths, through unsuitable ground, often below water level, before solid rock is reached. Keeping out water and preventing diggings from collapsing can require major feats of engineering in themselves. Tunnels in particular are subject to unforeseen problems and disasters. Tunnelling under the river Thames, under the river Severn, and under the Alps, was achieved at cost of life and limb.

Above ground, until spans are joined, wind can be a great hazard. Before the Pont du Normandie was completed, there was serious discussion about the use of active stabilisers to keep the long thin cantilevers in place. Large weights would have been moved by rams, in response to amplified and processed signals from accelerometers. A few tall buildings actually use this technique to reduce wind induced swaying.

The pillars of the towers of big suspension bridges may have to be stabilized by temporary cables until they are completed and joined at the top.

Arches generally need to be supported on falsework until they are complete. Perhaps this is the origin of the word "keystone", the last block without which the structure cannot hold up, though in the finished arch, the keystone is no more important than any other voussoir. So the value of the word keystone is to remind us that until the structure is complete, we have to keep thinking.

Completeness includes completeness of communication. The chain of command and communication must be designed to cope with every foreseeable situation, and it must include rules for dealing with emergencies and unforeseen problems. Designers and builders have ranged from those who have overly interfered in small details that should have been delegated, I K Brunel, for example, and those in which day to day disconnection has contributed to eventual failure, and indeed total disaster, as in the first attempt at the Quebec bridge.

Some very large arches have been built by treating the halves as cantilevers until they meet in the middle. Whether or not this method is used, joining the parts of any big structure is a very serious matter. The stresses in the separate parts are different from those that will apply in the complete system. If the parts are just left to rest against each other, or joined as they meet, the resulting stresses may be far from those that are required. Some form of jacking will often be required.

Furthermore, the temperature and wind may make life difficult for the builders. There have been occasions when heaters or ice-packs have been used. Spinning the cables of the Forth road bridge was not possible on many days because of high winds.

Back to Home Page back to Bridges

Construction of Arches

An arch is in compression throughout, and it cannot stand except as a whole. It therefore requires temporary support, or falsework, until it is complete. The type of falsework depends very much on the material of which the bridge is made, and on the size of the bridge.

Masonry arches, being made of relatively small voussoirs joined by mortar cannot take tension, need continuous support during construction from below. This type of falsework is called centring, and is often of the general form shown below.

These two pictures show the corbels upon which the centring was erected. When the centring has been removed, or struck, the arch will inevitably settle slightly. This is inevitable, because it can only generate the required compressive forces by undergoing some strain. All structures, in fact, must deflect when temporary support is removed.

If the ground is weaker than was anticipated, the settling may be severe, as in the case of Telford's Over bridge, which sank about ten inches (25 cm) at the crown. Telford had wanted to build one of his standard 150 foot cast iron arches, which would have been a very economical solution, but some important people in Gloucester decided that they did not want iron, even though Over is not in the city, and not even visible from it. These people should have known better, because the fact that Gloucester is built entirely on the east side of the river Severn is a constant reminder that the ground on the west side is soft and unsuitable for construction work. Apparently the information that Telford was given about the soil conditions were faulty. Nevertheless, the bridge was in use until the second half of the 20th century, and still stands as one of many testimonies to Telford's skill, and also as one of many testimonies to the results of powerful people making decisions on matters of which they know little or nothing. In this case the result was a very heavy and very expensive bridge.

Concrete bridges may be supported in a similar way, except that the centring will support the formwork in which the concrete is poured. For large bridges, the centring will be a substantial structure in its own right, and will be expensive. For a multi-arch bridge, it is desirable to re-use the centring for each span. In the case of a bridge across a river, the centring may be floated from one arch to the next, and will be in the form of a tied arch to maintain it integrity. The centring for the illustrated arch would have been impressive.

The procedure for steel arches is usually different, because their spans are often so large. The next picture shows how a steel arch with side spans may be cantilevered out from each side.

This approach cannot be used for bridges like the Sydney Harbour bridge. One solution is to tie the arms back using cables. This method was pioneered by James Eads for his triple arch over the Mississippi river.

In both cases the stresses will be significantly different once the gap has been closed. For example, in the first example, the top chord is in tension, but after closure, it may be made to take half the compression force away from the bottom chord. Jacks may be used to distribute the stresses according to the designed values. The next two pictures give a rough idea of the centre of the span before and after closure.

Concrete arches also may be built using the cantilever method, as in the schematic example shown below. For very wide and exposed sites, this is preferable to lightweight formwork which could be vulnerable to high winds.

The stresses in the arch during construction would be different from those expected after completion, and the tensions in the temporary stays need to be adjusted as new sections are added. An advanced version of this method was used for the Tilos bridge on la Palma, which is based on a concrete arch. The verticals used during the construction were the same ones that form spandrel piers in the completed bridge. These were connected at the top by horizontal members during construction. Unlike the deck of the completed bridge, these members were clearly in tension, illustrating very clearly how stresses in completed and uncompleted structures may differ greatly.

Change of stress of course implies change in strain, and this influenced the design of the formwork for the concrte arch. Over the months following completion of construction, any concrete element in compression will be slightly reduced in size. Because the Tilos arch has no hinges, the change in length would induce bending moments. Massive jacks were provided to allow fro adjustments. Because of the differing stresses in the various phases, this bridge was designed to be very light.

The bowstring or tied arch is exceptional in that it can be treated as a unit which creates no outward thrust. In principle such an arch can be built away from the final site and then lifted or slid into place. The arch illustrated here was built close to the M42 motorway and slid into place during a single night, with minimal disruption of traffic.

Construction of Beam Bridges

Beam bridges are generally in the form of plate girders, box girders or trusses. In all cases, a common construction method is to build the beam away from the final position and slide it or lift it into place as a complete unit. Bowstring or tied arches may be built in the same way.

There is nothing especially complicated about the lifting process, and the stresses in the beam are more or less as they will be in the final position. Nevertheless, a heavy object suspended in space is potentially dangerous, and accidents do happen. During the lifting of the suspended span of the Quebec bridge, something broke, and the span fell into the river and was destroyed. During the lifting of one span of the Britannia bridge, a jack burst, and the end of the span fell. Fortunately, Stephenson had given strict orders to insert packing after every few inches of lifting. Nevertheless, a slight distortion of the beam did occur. The incident shows the importance, not only of correct instructions, but of good communications. Accidents have happened because instructions were not received, or if received, modified or ignored.

An exception to the method of lifting may be adopted in the case of the suspended span of a cantilever bridge, as in the case of the Forth railway bridge. Here, the beams were built out from the ends of the cantilever arms until they could be joined in the middle. Only then were the temporary rigid connections removed. Such an operation has to be done with great care, because large stresses are released and created. The bending moment at the join completely disappears, for example, while the stress in the top chord of the beam changes from tension to compression. The faked picture shows the general idea.

Some continuous box girder beams and concrete beams are constructed as cantilevers until they are joined up. New parts may be taken along the existing part of the bridge, or they may be lifted up from below. Sections of a segmented prestressed concrete bridge may be lifted by a crane which rests on previously attached segments, or they may be manipulated by means of a launching girder which is slowly advanced along the already built part of the span. Another method is to use moving formwork which progresses with the construction. Because the cantilevered part span is subject to stresses for which the complete span is not designed, temporary stays are sometimes attached between the segments and a temporary tower, turning the span into a cable-stayed one until closure is complete.

Yet another technique, when the lower surface of the span is horizontal, is to rest the formwork on a steel girder which reaches to the next pier.

The casting of each segment is often done using the previous segment as part of the formwork. This technique, called match-casting, ensures an almost perfect fit at assembly time, and greatly reduces the tendency for misalignments to occur during erection. A film of epoxy may be spread between segments.

The general idea is shown in the diagrams below, with a diagram of a lattice truss for comparison. After each segment is placed, steel pre-stressing wires, or tendons, are fitted to hold it in position. The pre-stressing tension prevents the concrete ever reaching a tensile state with any foreseeable load. This method of pre-stressing is called post-tensioning, because the stress is added after placement. When the wires are tensioned before the concrete is poured, the process is called pre-tensioning. The tendons are equivalent to the ties in the lattice truss, while the concrete is roughly equivalent to the struts. Thus the concrete and the steel are both used in the roles for which they are best suited. Blue represents tension: red represents compression.

The joints between the concrete segments are subject to vertical shear forces, which can be resisted by the use of matching corrugated surfaces.