When it all goes wrong
This picture shows maintenance work on the Clifton suspension bridge.
"But remember, please, the Law by which we live,
We are not built to comprehend a lie,
We can neither love nor pity nor forgive.
If you make a slip in handling us you die!"
Rudyard Kipling - "The secret of the machines"
"It appears that there are enormous differences of opinion as to the probability of a failure with loss of vehicle and of human life. The estimates range from roughly 1 in 100 to 1 in 100,000. The higher figures come from the working engineers, and the very low figures from management. . . . . . . . . . For a successful technology, reality must take precedence over public relations, for nature cannot be fooled."
Richard P Feynman, after the explosion of the Challenger space shuttle.
"For the want of a nail, the horse-shoe was lost . . . . "
The first picture reminds us that reliability begins with foundations. This building was built near a beach, and the ground has proved inadequate; perhaps a hurricane contributed. The second picture reminds us that a structure can be affected by forces beyond the control of the builder, in this example subsidence due to mining works.
This picture shows pillars and a fence that have been moved in a remarkably uniform manner by soil movement on a slope. The leaning is the result of differential movement, the speed of movement being greater near the surface. Similar effects are seen in the flow of glaciers and rivers, but we should not necessarily attribute similar causes to similar effects. Clues as to vertical are the pillars at the top right of the picture, and the step near the bottom left of centre.
Bridge-building has generally been a safe discipline, but no activity can progress without a few disasters, unless we are prepared never to tackle the unknown. But the unknown should be attempted with as much investigation as possible before the step is taken.
The Tay railway bridge was designed with a badly underestimated resistance to high winds, and there were instances of faulty materials and inadequate responses to detected problems.
Several suspension bridges have succumbed to induced oscillation. In the year 2000, the Millennium bridge across the river Thames was closed after only two days of use because of oscillation of an unforeseen type.
One way to reduce the probability of trouble from unknown causes is to make tests on scale models. Computer programs can be used to simulate many structures. In both cases the results will of course be no more accurate than the faithfulness of the simulation.
Accidents and problems during construction
Some early box-bridge spans fell down during construction, a period during which many structures are more vulnerable than when they are complete. These bridges were designed as continuous hollow beams, with diaphragms at intervals. This is a strong and light design. But during construction, the growing beams were cantilevers, supported only at one end. Until they were joined to their partners, the parts near the piers had to withstand bending moments that would not occur after completion. In addition to this, there was also the weight of cranes and other equipment, of necessity at the tip of the cantilever. The Milford Haven bridge in Wales and the West Gate bridge over the river Yarra near Melbourne, Australia, collapsed during construction, with loss of life.
The incomplete spans of cable-stayed bridges, or the towers of incomplete suspension bridges are obvious candidates for wind-induced oscillation.
Moving and lifting spans or parts of spans are particularly dangerous operations for bridge builders. Large and heavy objects in motion have huge momentum if they move quickly. Lifting them puts them in positions where they are not yet secured in their final places. Wind, tide and river flow are among the factors that can cause trouble at these times.
Operations underground and under water are particularly dangerous, and great care is taken to minimise risk to safety.
This picture shows a part of the south side of Tewkesbury abbey, built on a flood plain. Although the abbey is rarely flooded, ground close by is frequently under water, and the water table is not far below the surface. This wall has moved, either during construction, or later.
Problems during use
Scouring by rivers is a powerfully destructive force, which can destroy the most solid masonry. Piers from early times have been shaped to reduce resistance to the flow of water, and have often been surrounded by a pavement of stone. Their foundations have to go so far down that scouring cannot remove a dangerous amount of material.
Maintenance is a huge task on big structures - "painting the Forth bridge" is in the UK a well-known metaphor. Ideally a structure would be designed so as to require no attention once complete. In practice this is an impossible ideal. At the very least, any structure needs inspection, however well it was built. On any large structure you can see inspection hatches for access to the inside.
The collapse of the Tacoma Narrows bridge is well known. Spare a thought for the designer. Books do not record a general outcry against the design, which was presumably accepted as an example of progress in design economy. It is the designer's bad luck that he took his design too far for the current state of knowledge. But other suspension bridges had been damaged or destroyed by oscillations. History is not bunk. Download
Modes of Failure
What are the commonest modes of bridge failure? Because a bridge was "not strong enough"? The straightforward failure of a strut or tie because it is not strong enough is quite rare. Ties rarely snap in two, and struts are seldom crushed by pressure. What is more likely is that fatigue can start cracks, resulting in stress concentration, and at a certain moment, the crack can propagate with lightning speed, leading to a broken tie. Struts and plates are more likely to fail by buckling than by straight crushing.
Joints are a frequent source of failure, and if we extend the meaning of joints to include the connection to the ground and river beds, then we can include effects such as the scouring mentioned earlier. Some joints consist in a span resting on a support, held only by its own weight, and there have been cases where flooding has created enough upthrust for a span to be moved off its pier by the flowing water.
This road sign has buckled, probably as a result of thermal expansion. The bending is small, and is visible only because of the glancing angle of the light. It has no effect on the function of the sign, but this much deflection in a compression member in a structure would be a cause for concern,
Accidents during use
Some events are outside the control of the designers and constructors. Some earthquakes are so powerful that no structure can withstand them. Ships have been known to collide with bridges, bringing spans crashing into the river. This happened to the Sharpness railway bridge over the river Severn. It has never been replaced. The Sunshine Skyway bridge in Florida was struck by a ship in 1986, bringing down a span.
A tragic consequence of building a big bridge at a dramatic location is the attraction to suicidal people. The Clifton bridge is well equipped with cameras, but it has proved necessary to add curved fences to make climbing over the side very difficult. Some people survive bridge jumps, often horribly injured. People sometimes jump off quite small bridges over roads or railways.
People have occasionally decided to fly aeroplanes under bridges. The Tower bridge in London has been the subject of such behaviour. A Vampire was flown under Clifton bridge, but the pilot misjudged the manoeuvre needed to escape from the curving gorge. He didn't make it. This type of exploit is extremely dangerous because it cannot be practised - the pilot can only study maps and models.
This signpost has received an impact for which it was not designed. No structure can be made resistant to all conceivable forces: all that can be done is to find a sensible compromise between cost and resilience. The calculation must include an assessment of the effects of any damage, including danger to people. One problem in many calculations is that there is no cut-off in the size of events. For example, large sea waves are less frequent than small ones, but there is no known limit. The same is true of the size of meteorites and the magnitude of earthquakes. A graph of frequency of occurrence might look rather like this, which is an idealized example -
Notice how the logarthmic vertical scale in the lower graph reveals detail for the small frequencies which the linear scale obscures.
Sometimes the frequency roughly follows a simple mathematical formula. If F is the frequency, and S is the size, some examples are as follows -
F = A S-N F = A log(S) where A and N are constants. The graphs above show events distributed linearly by the log of the size, an exponential distribution. A more realistic example is shown below: it includes the effects of randomness on a finite sample of data.
We must remember that with a finite number of events, the accuracy of prediction will be limited. When the numbers are very small, and the potential effects are important, arguments can become very heated. Some events are correlated, such as earthquakes and their aftershocks: others appear to occur at random times.
diagrams above represent a simple beam bridge, which has been affected by
subsidence (exaggerated). One response is for the beam to remain so
straight that it is only supported in two places, leading to a bigger
effective span. Another is for it to bend. A third would be to break, if
either of the first two conditions were unsustainable by the
In practice the designers might include jacks at the base of the piers, to allow for adjustment.
What happens as a result of the movement is that the beam suffers stresses which were not in the design. In fact the problem exists from the start.
The four support points can never be perfectly aligned, but the alignment is of course made so small that the beam can adjust its shape without absorbing too much energy.
The penalty for a through beam is the over-determination. The benefits are the spreading and controlling of loads and stresses.
diagram above represents the response of a simple cantilever bridge to
subsidence. In this case the joints allow stress-free movement, so nothing
Jacking might still be provided. In a very slender foot-bridge, the slightest change in position could be noticeable, and so some adjustment may be needed.
After a bridge has been completed, jacks may be concreted over, or they may be left as usable adjusters. The Eiffel tower is a good example of the jacking requirement. The stresses, and therefore the strains, at the base, changed markedly during construction. Jacking enabled the builders to compensate as the work progressed. This subject is developed further in Indeterminacy.
Even if nothing apparently goes wrong, without maintenance, most structures will degenerate. Painting the Forth bridge is a well-known metaphor. On the whole, it is rare for a bridge to be left to rot, unless it is disused. Here are pictures that show how plants can seize the smallest opportunity to grow where they are not wanted. Once they have taken root, they can generate more soil, and they can generate pressure that widens the cracks. This allows more water in, increasing the risk of damage by the expansion caused by freezing.
The top of the wall shown in the last picture has since been rebuilt, as we see here. The second picture shows another part of the same wall, which has metal braces to keep it more or less upright.
Plants, of course, are welcome in the right place. Many shops, pubs and other buildings have hanging baskets and window boxes, providing colour during the summer. Some towns even have hanging baskets along the centre line of major roads. Derer, in The Nativity, Paumgartner altar, shows some substantial plants growing from a ruined arch.
Plants are sometimes welcome in a wall, even planted deliberately. That they are wanted in the wall does not, of course, reduce their eventual effects on the wall. But the owners of a wall that could last for hundreds of years are not likely to worry about a possible reduction of its life by a very small amount. Furthermore, a dry stone wall, includes relatively large cavities where roots can grow without exerting much pressure. But a brick wall relies on the mortar to space the bricks, and any penetration by roots is potentially damaging.
Plants are growing in this arch bridge over the disused canal near Brimscombe, creating a charming effect, but the long term effects will be bad. The second picture, squeezed horizontally, shows the slight distortions that are so often found in old arches.
Domestic gardeners too, often grow plants on the tops and outsides of their walls, where others can enjoy them.
During the summer of 2000, the central reservation of the Barnwood bypass in Gloucester was alive with colour because someone had had the idea of sowing thousands of wild-flower seeds. In Birmingham, central fences have been decorated with flower baskets, and in many towns, spring flowering bulbs give pleasure to thousands of people.
Plants and fungi can exert tremendous pressure. In these two pictures, parts of a fence have been displaced by the roots of trees. In the first picture, the pavement has been cracked by the roots as well.
In the background of this picture is Gloucester cathedral. Unlike the Forth railway bridge, which is continually painted to keep it from rusting, cathedrals have to cope with wind and weather without surface coatings. The Cotswold limestone, used extensively in this cathedral, and in places like Bath and many Cotswold towns and villages, is vulnerable to acidic impurities in the air. The next pictures illustrate examples of erosion on the outside of Gloucester cathedral and nearby buildings.
Erosion by wind-borne dust adds to the damage. Masonry at the top of the cathedral tower became so dangerous that it had to be replaced. The outside of the cathedral is continually being checked and repaired. Many buildings in Bath were almost black until a vigorous cleaning program restored them to the normal pale colour.
This example shows a part the bridge, built in old red sandstone, over the river Wye.
Exfoliation can be a danger with certain types of stone. Significant thickness can be taken from a block of stone. In the second and third pictures we see that the pattern of damage on one headstone is mimicked by the dark shape (damp?) on the other. Do you think this is significant?
These two pictures show exfoliation on the parapet of a small bridge, less than a year after the blocks were placed as part of a repair job. The material is oolitic limestone, which contributes to the appeal of Cotswold villages and towns. Some samples of it have a tendency to split or flake, a process which is accelerated by damp and frost. These examples have no structural significance, but splitting of reinforced concrete is very significant if steel reinforcing bars are exposed to damp air, with the probability that rust will occur.
An unusual maintenance activity takes place on the Tour Eiffel, which was designed for a very short life. Now over a hundred years old, it has had many parts replaced, and you can buy rivets that have been grossly deformed by shear.
The way to make things last is this - design it right - build it right - use it right - maintain it right.
These pictures were kindly donated by Gilberto Castaeeda. They show the magnificent Puente Tampico across the Rio Panuco in Mexico, with damage to a cable sheath, caused by exposure to sun and wind. The main span is 360 metres. A new sheath has been provided.
This is a poor picture of a bridge. That is because it isn't a picture of a bridge: it's a picture of a piece of wood - the one that's a different colour from the rest. This bridge was built in 2016, on two steel I-beams, after the previous wooden bridge had been deliberately damaged to the point where it was unusable. The pale piece of wood repairs the first damage to the new bridge, done after less than a year of use. In time of war, bridges are often among the early targets. They are often among the later ones, too, when retreating troops destroy their own structures behind them. But this example is damage for its own sake. Many objects in public places are now built in steel, or even stainless steel, to reduce the susceptibility to damage. The requirement to resist deliberate damage adds one more difficulty to the engineer's burden.
A Bridge Needing Repair
These pictures show an arch bridge after some voussoirs had fallen off. If even one voussoir falls out, the compressive force on all the others in its row are to some extent relieved, and those nearest to the gap are held in place only by the adhesion of the mortar. The release of stress means a release of strain, and that implies that the relieved row of voussoirs tends to become slightly longer than the neighbouring row. There will thus be shear stress between the broken row and the unbroken row. This shear stress will in fact tend to share the compressive stress between the two rows, especially far from the gap.
There are implications for a repair. If any structure contains more than the minimal set of parts that will keep it standing, loss of a part may not be disastrous: the stresses are simply shared among other parts. But when a repair is attempted, simply fitting a new part is not enough: in theory, the gap should be jacked until the stresses are similar to their values before the break. Sometimes the loss of stress is irreparable. When the Britannia bridge over the Menai Straits caught fire, the internal stressing put in by Stephenson was lost, and the spans now have arches to support them. In the example of the skew bridge, any one row of voussoirs is so lightly stressed that simple cementing new bricks is probably an adequate repair. An example where repair is difficult is a large piece of pottery that breaks, and the pieces are found not to match when held together. This happens when internal stresses and strains were locked in during the cooling process after firing. They are released by the breaking, and the parts take slightly different shapes.
Poor Decision Making
We all have to live with the work of engineers and architects, so they cannot always (ever?) be allowed to make things as they please. Some decisions, however, must be left to those who know. What to build and where to build it, what to do and when to do it, cannot be decided entirely by people who have no technical knowledge: a "can do" spirit is a fine thing only when tempered by good advice on the technically feasible.
This picture shows Thomas Telford's fine bridge at Over. When the centring was removed, the crown sank by about ten inches, a matter of life-long regret for a man of integrity who experienced very few failures. Telford freely admitted that he had provided inadequate restraint for the great thrust of this very flat and very heavy arch, but behind the facts lay a decision made by non-technical people - the "important" men of Gloucester, who did not want an iron bridge. An iron bridge would have been the logical choice, as Telford was already using a standard 150 foot design. It would have been cheaper, and much lighter, than the bridge that was actually built.
A related problem is the outward movement of walls, caused by heavy pitched roofs.
When you see the sign "Weak Bridge", the reason may be one or more from a very long list of possibilities. Whether the bridge be of wood, cast iron, wrought iron, steel, reinforced concrete or pre-stressed concrete, it may be "weak" from one of two main causes. The bridge may have retained its original integrity, while being overtaken by the huge increase in axle weights of railway locomotives and wagons and of road vehicles, or it may have lost some strength through any of a variety of causes. These causes may be related to the intensity of loading over the years, to poor maintenance, or simply to age. The ancient Romans built much that still survives, but most engineers are constrained by economics and practicality to build for a foreseeable span of years.
Many a structure that looked so good in the artist's impression, depicting the scene on a sunny spring day, with imaginary trees in leaf, later looks dowdy, as a result of fading or peeling paint, wrinkled or cracked surfaces, corrosion, creep, dirt, erosion, graffiti, and other forms of entropy. Land, too, will revert to a more natural state, and given enough time, the climax vegetation may be reached, though, because of changes made to the soil, it may not be the original.
On the other hand, the designer of this sculpture must have known that it would rust. The use of copper on roofs to form an attractive green patina was once quite common, but is now quite rare. The bridge was also designed to acquire a stable coat of oxide, requiring no subsequent protection. Aluminium forms a film of oxide which prevents corrosion, but if it is used in contact with some other metals, such as iron, in damp conditions, corrosion can occur because of electrolytic effects. This phenomenon can be turned around to protect iron or steel by using a metal such as aluminium or zinc, with a source of electric current. Galvanizing works on a related principle.
Some changes in the appearance of structures are deliberate.
These pictures above show the effects of time on a multi-arch railway bridge. The two left hand pictures shows a crack from top to bottom, where the retaining wall of the embankment meets the bridge. The two right hand pictures show a number of small stalactites, formed by dripping rainwater.
After these pictures were taken, the large crack was grouted. This has no structural effect, but it keeps out the weather and wet.
In a limestone cave, we marvel at the size of the stalactites and the curtains of limestone. We forget the sheer arithmetic of time. If a drop of water runs down every ten seconds, that's three million drops in a year. Three thousand million drops in a millennium. And what is a millennium in geological time? Not much. Even on the bridge last mentioned, dry for most days of the year, and less than a hundred years old, we see measurable progress in stalactite production in one of the pictures. Other pictures show the effects of water running down the sides of the arch.
It is possible to cope with relative movement without cracks. Half timbered buildings comprise frames filled in with materials like brick. The areas of brick are so small that they are unlikely to crack, and any cracks cannot get past the beams. This example has clearly sagged in the middle. On ground that is not of good quality, a large reinforced concrete raft will spare the building any deformation, though it is an expensive solution.
From Gloucestershire to Yorkshire and beyond, stone walls delineate the landscape. Most have stood for hundreds of years, but as their utility has declined, so has the care invested in them. Cotswold limestone in particular, soft and susceptible to cracks, frost, colonization by plants, damage by people and animals, is suffering badly. This wall is typical of many that remain. As soon as any part starts to collapse, it is used as a way through by animals, whose continual passage hastens the process of decay. The positions of many other walls are marked by little more than elongated ridges in the ground.
Intermediate stages can form useful habitats for snakes, lizards, insects, mammals and plants. The picture at left shows an adder basking on the wreckage of an ancient Cotswold wall. The adders find refuge and prey within these walls.
The end is nigh . . . .