A major objective of design is to avoid failures; We shall start by considering failure in its broadest sense and illustrate this by a story related to one of the authors by his father, who was for many years concerned with construction in various parts of the world. The supply of water to an industrial plant for which his company acted as consultant required the construction of a water tower in order to provide sufficient water pressure at the plant. The tower had to be fairly tall and was to be built of reinforced concrete. A small local contractor was given the job of building the tower. He had no mechanical equipment and the concrete was all mixed by hand and taken to the tower in baskets carried on the heads of women. The quality of the placing and compaction of the concrete was very poor and, by the time that the tower was about half built, it was clear that the tower could never be built to its full height and be filled with water. The resident engineer acting for the consultant condemned the tower and asked that it be demolished and a new tower built. The local government officials were not prepared to lose face by admitting that the local firms were incapable of constructing the tower, so political pressure was brought to bear and this led to a compromise. The tower would not be built any higher and the tank would be installed at the existing level. This was duly done and the tower was handed over by the contractor with, no doubt, great pride. It was, however, totally useless to the client because it was not high enough to do what it was required to do. The tower did not fall down; nevertheless, it was a failure as far as the client was concerned as it was just as useless to him as if it had fallen down. This suggests a broad definition of failure along the following lines:
A structure has failed if it is, or becomes, incapable of fulfilling its required purpose.
Clearly not all failures are of equal seriousness. The example of the water tower above is clearly a case where the failure was total as the resulting tower was useless. However, many failures, though they inhibit the proper functioning of a structure, either do not make it completely useless or the fault can be rectified without excessive expenditure.
We can now look at various ways in which this may occur.
The above example comes into this category. This is where the structure is simply unsuitable for its required function. There is no failure of the materials from which the structure is built, and the structure probably meets all the requirements of the appropriate design rules and regulations. Failures of this type, or at least partial failures, are less uncommon than we might like to think. They are not necessarily the fault of the structural engineer or constructor; it is actually more likely that the fault lies in the client either not having analyzed sufficiently carefully what he needed or not having communicated his requirements to the design team with sufficient clarity. Drains not placed at the lowest point of the area they are intended to drain, bridges with insufficient clearance for what is intended to pass under them, have all happened and fit into this category. The way to avoid such failures is for all parties to analyze the problem being solved and for all to be absolutely clear about what is required. This is probably the first task that should be carried out in any design.
Serviceability failures are where the behavior of the structure under the service loads inhibits the proper functioning of the structure. Examples of this type of failure are very frequently due to excessive deflection. For example a beam that deflects to the extent that it distorts doorframes and makes it impossible to open the doors would be a serviceability failure. Another example is where vibration of the structure, possibly due to machinery, makes it impossible to operate sensitive scientific equipment. Most serviceability failures are related to a lack of stiffness of the structure rather than a lack of strength.
If you take a set of children’s wooden bricks and build a tower, it eventually reaches a height where it becomes unstable and falls down. This is a failure of the tower but the materials from which the tower is built (the bricks) are completely undamaged and can immediately be used again to build another tower. This type of failure, where there is no overstress or failure of the materials but failure arises simply because the members of the structure become separated, is a disconnection failure. Another example of such a failure is the collapse of a house of cards. It will be seen that disconnection failures tend to be sudden and catastrophic and therefore, in the real world of structures, need to be avoided.
There are practical examples of such failures, frequently called ‘progressive collapses’. Probably the most famous is Ronan Point, where a gas explosion in a multistorey residential structure made of precast panels resulted in a wall being blown out. Loss of this wall removed support from the structure above, which fell. The resulting loads from the falling panels caused the lower levels to collapse and the whole corner of the building fell down. This failure is discussed again in a later chapter. It is arguable that the collapse of masonry buildings where the masonry is laid in lime mortar are also disconnection failures as the masonry units (stone, brick) are often not failed. This type of failure probably kills more people than any other in earthquakes in less developed countries.
In large buildings, the possibility of this type of failure is nowadays guarded against by ensuring that structural members are connected together with connections capable of resisting at least some minimum force.
These are failures where there is no failure of the structure (at least initially). They are generally the result of a failure of the foundation or soil around a structure. The classic example of an incipient translation failure is the Leaning Tower of Pisa. The Tower itself is not failing, there is simply excessive differential settlement between one side of the tower and the other, causing it to lean. Should the tower fall down, the prime cause will be a rotational translation of the tower. Of course, as the tower falls, and certainly when it hits the ground, it will fall to pieces, but the breaking up of the tower would be a secondary issue and not the primary mechanism leading to failure. With luck, and the application of engineering expertise, the tower will not collapse but will be straightened somewhat.
Actual translation failures do occur, and are probably the normal way in which one would expect a retaining wall to fail.
When you take a straight piece of thin wire, hold it vertically and try to apply a vertical load to the top, you will find that the load that can be applied is small, and the wire will bend. Take the load off and the wire will return to its original position, showing that the material from which the wire was made has not been overstressed. Failure of the ‘structure’ in this case is the result of the deflection and is generally known as ‘buckling’ or, more correctly, as elastic instability due to second-order effects. The mechanisms involved will be discussed further in Chapter 3. For the present, it is only necessary to see that the failure does not arise primarily from overstress of the material but is the result of the deflection, which is a function of the stiffness of the member rather than its strength. Elastic instability can lead to failure in almost any type of thin member subjected to compressive stress. Examples of cases where this possibility has to be considered are:
Failures in this category are where the member or structure is overloaded to the extent that the material fails. Guarding against this type of failure probably forms the major part of the detailed design of a structure. Since the next chapter is largely concerned with understanding this type of failure, nothing further will be said here.