Back to Bridges Back to Home Page
Testing bridges, like testing any other structure, and any mechanism, begins long before anyone starts to dig a hole in the ground.
With the advent of powerful computers, it became possible to make accurate simulations of behaviour. These calculations are only possible if the properties of the materials are accurately known, so we are still dependent on testing actual objects.
Before the computer age, all testing was done in the material world, sometimes using models. The first stage is the testing of raw materials, which is universal, and not specifically related to bridges. Raw materials are converted into structural materials such as steel, concrete and other composites. The conversion process must be rigorously controlled, and samples of the products must be tested at prescribed intervals. Engineers who order metals, alloys and other materials need to know that the quoted specifications are absolutely reliable as a guide to the actual behaviour of the product as received.
Nevertheless, engineers may decide to test samples of the materials they receive, in case of some problem that the makers have not detected. And, of course, sample testing cannot detect a problem in an example that was not tested. The testing regime must be designed to achieve a given level of probability that all is well.
In order for a system to work, the following steps are needed -
Precise definition of what is required of the product.
Precise adherence to the specification.
Adequate testing of the product.
Adequate supervision of the testing.
Use of the product only within the envelope of conditions under which it is specified. Depending on the material, these conditions may include temperature, pressure, humidity, corrosive environment, paint, electric fields, magnetic fields, electromagnetic radiation, ionizing radiation, sound waves, wind, vibration, continuous stress, varying stress, and even time.
At the heart of all this are human behaviours such as honesty, discipline and communication. As Rudyard Kipling wrote in a poem, if you cheat, you die (or someone else dies). The laws of physics will not be fooled.
A good example can be found in the superb book "The Great Bridge", by David McCullough, which explains how rejected wire was recycled into the system and was built into the bridge. The book also explains what was done after the fraud was discovered. Another form of fraud is the deliberate manufacture of parts from inferior materials. Forgery of certificates and other documents is
The collapse of the the first railway bridge was, as often happens, the result of a number of inadequacies. The design was far too weak, but acceptance of poor quality materials and attempts to hide the defects, allied by poor supervision and communication, probably contributed.
A lower chord of the first Quebec bridge had shown signs of buckling a long time before the final collapse, but poor communication resulted in the deaths of many men.
The box-girder bridge over the estuary near Milford Haven collapsed during construction, because the stresses were not sufficiently understood. At some points in the bridge, these were greater than they would have been after completion. Welding is an important consideration in box-girder bridges. In fact, joints are a major source of concern in all structures, not only because of the high stress, but because they may allow the ingress of water or corrosive gases or liquids, which may remain in intimate contact with the material for long periods.
In the 19th century, the design of the Britannia tubular bridge over the Menai Strait was assisted by the use of scale models, which were tested to destruction by Walter Fairbairn. As a result of these tests, rectangular tubes were preferred over elliptical ones. Robert Stephenson also employed a mathematician, Professor Hodgkinson.
The Millennium bridge in London was closed after two days, because the amplitude of oscillation was unacceptable to many pedestrians. The design had been thoroughly tested by computation, but the interaction between the walkers and the bridge was of a kind that had not been anticipated.
To do . . . .
Testing during construction - deflections
Ultrasonics, acoustic emission, etc
Testing with heavy trucks or trains after completion.
Testing materials may be conveniently divided into destructive testing and non-destructive testing, meaning crudely that we break the object or we don't.
Destructive testing is very often performed on samples of material to find out their properties or to check that they are within the bounds of a specification. But it is also performed on complete assemblies, to ensure that they can withstand the required stresses. On the other hand, it is not feasible to destroy a bridge in order to be sure that it was correctly built. A more suitable technique is to load it with heavy vehicles, while measuring the deflections at a number of points. The results can be compared with the calculated values.
Nevertheless, complete assemblies have been destroyed in the interests of safety. After the loss of several Comet 1 airliners, a complete fuselage was immersed in water, and filled with water. The pressure difference between the inside and the outside was cycled many times, to simulate climbing and descent. Eventually, cracks occurred at the corners of rectangular holes. The wings, outside the tank, were bent up and down repeatedly. Why water and not air?
To increase the pressure of air to 1.5 times the original, we need to reduce the volume to about two thirds of the original. Since the work done in a small change in volume is equal to the pressure multiplied by the change in volume, the air stores a lot of energy, which is why the Comets exploded. But the same change in water pressure makes only a tiny difference in volume. The bulk modulus of water is around 2 109 Newtons/square metre.
Nowadays, one example of each aircraft is sacrificed in fatigue tests, and is taken through more cycles than any actual flying aircraft of the same type.
Back to Bridges Back to Home Page