How Structures Work

Design and Behaviour from Bridges to Buildings

David Yeomans

• Éditeur: Wiley-Blackwell
• Année: 2015
• ISBN:
  9781119012283
• Édition: 2
• Sujet: Architecture
• Sous-sujet: Architecture Design

Structural engineering is central to the design of a building. How the building behaves when subjected to various forces – the weight of the materials used to build it, the weight of the occupants or the traffic it carries, the force of the wind etc – is fundamental to its stability. The alliance between architecture and structural engineering is therefore critical to the successful design and completion of the buildings and infrastructure that surrounds us. Yet structure is often cloaked in mathematics which many architects and surveyors find difficult to understand.

How Structures Work has been written to explain the behaviour of structures in a clear way without resorting to complex mathematics. This new edition includes a new chapter on construction materials, and significant revisions to, and reordering of the existing chapters. It is aimed at all who require a good qualitative understanding of structures and their behaviour, and as such will be of benefit to students of architecture, architectural history, building surveying and civil engineering. The straightforward, non-mathematical approach ensures it will also be suitable for a wider audience including building administrators, archaeologists and the interested layman.

1
Brackets and Bridges

Figure 1.1 Failure of the Dee railway bridge, 1846.

When civil engineers in Canada graduate, they are given a steel ring, to be worn on the little finger of their writing hand, which identifies them as graduate engineers. No doubt those who wear it see it as a kind of badge of honour, but they should see it as a constant warning because traditionally it was made from the steel of the first Quebec railway bridge, which collapsed during construction. The warning is simple; no matter how experienced you might be as an engineer – and Theodore Cooper, the consulting engineer for the Quebec Bridge, was very experienced – you can still have a failure. When the Dee railway bridge collapsed in England in 1846 (Figure 1.1), it might well have destroyed the reputation of Robert Stephenson, then Britain’s leading railway engineer. Instead, it led to a major enquiry because at that time the behaviour of iron railway structures was little understood. The same cannot be said of the Quebec Bridge – the story of its collapse is a human drama and a human tragedy. It involved no complex technical issues, as some engineering failures have, and Cooper’s career ended in disaster rather than triumph. Most of the technical issues were very simple, with a structural type similar to the successful Forth Bridge, built just a few years earlier, so what could have gone wrong? To understand this involves understanding something of the process of design and having some very simple technical understanding of the structural principles involved. This collapse is an excellent introduction to the latter.

Cooper’s tragedy

Theodore Cooper began his career as an assistant to another famous engineer, James Eades, on the St Louis Bridge, another very significant structure in the history of engineering. This elegant three-span arch bridge over the Mississippi River, completed in 1874, was the first all steel bridge; bridges up to then had been in cast and wrought iron. Cooper then had a long and successful career as a bridge engineer, but he had never built a really long span, so that his appointment by the Quebec Bridge Company as consulting engineer for a railway bridge over the St Lawrence River must have suited both him and the bridge company. The bridge company had the services of the leading North American bridge engineer, while Cooper had an opportunity to build what he would surely have seen as his crowning achievement. Instead, it was to be a disaster, and his career ended with a major failure instead of a record-breaking span. This was to be a cantilever bridge rather like the Firth of Forth Bridge, with cantilevers extending from either shore and a suspended span between them. With the south side cantilever completed and the suspended span under construction, the cantilever suddenly collapsed and over 80 men working on the bridge at the time lost their lives.

By then in his early 60s, Cooper had been reluctant to travel frequently from his New York office to Quebec to inspect the work in progress and instead relied upon the services of a young assistant, Norman McClure, just as Eades had relied upon him in the construction of the St Louis Bridge when he was a young man. Unfortunately, this assistant proved too inexperienced, or perhaps lacked sufficient self-confidence, and when things began to go wrong, there were conflicts of opinion in which his voice was insufficiently firm. But at the same time, all was not well within Cooper’s office. He was understaffed for the magnitude of the task so that the true situation was not appreciated. The result was fatal delays in taking the action that might have prevented the collapse and saved those lives.

Of course, failures are not unknown. There had been bridge failures before, and there were to be failures to follow, but of the most famous failures, that of the Quebec Bridge stands out because it did not involve any principles that were unknown at the time, something that has not been true of some other significant collapses. The Dee Bridge on the Chester and Holyhead Railway collapsed when a train was crossing it and there was some loss of life. This bridge used a combination of wrought and cast iron in a form that had been used on other bridges but was in principle simply wrong. What was significant about that collapse was that it led to a government enquiry into the use of cast iron in railway bridges. Railway construction then involved longer spans than had been built before and much heavier loads so that engineers were stepping into unknown territory. While the design of this bridge was wrong in principle, it was not obvious to engineers at the time that this was so, and the report of the commission of enquiry provides a good insight into contemporary engineering knowledge.¹ Stephenson’s career survived this failure, and he went on to complete the spectacular Conway and Britannia tubular bridges on the railway line to Holyhead and the Victoria Bridge over the St Lawrence at Montreal.

Sir Thomas Bouch was not so lucky when his Tay Bridge collapsed in a storm in 1879, taking with it a train and the lives of all those on board. The bridge had scarcely been completed, and the inquiry into that disaster revealed a story that included poor construction and irresponsible use by the engine drivers who were in the habit of ignoring the speed limit and racing the ferry across the river.² But it also became clear that Bouch had not taken wind loads into proper account in his design. This was not something that bridge engineers had taken much note of until then, but the failure now concentrated minds on the issue, and it was certainly taken into account in the design of the Forth Bridge.

It was wind that was also to bring down the Tacoma Narrows Bridge in 1940, although it was not a particularly severe wind that caused the problem, nor was the bridge a particularly long span. It was, however, the most slender suspension bridge that had been built up to that time, and the wind produced aerodynamic effects that had not been previously recognised. Because of its slenderness, it twisted in the wind (Figure 1.2) in a way that exaggerated the effect of what became known as vortex shedding, until the bridge was destroyed. The replacement bridge was made much stiffer, and since then engineers have been aware of the need to guard against this kind of failure, sometimes using wind tunnel tests to explore the behaviour of their bridge decks.

Figure 1.2 Tacoma Narrows Bridge twisting in the wind.

These three failures all occurred when engineers were pushing at the boundaries of what was then known. When buildings are built higher, or bridges have longer spans, there is always the possibility that one will discover some aspect of the structure’s behaviour that was not previously recognised. But the span of the Quebec Bridge was not that much greater than that of the Forth Bridge, less than 10% longer, so that should not have been an issue. What Cooper also did, however, was to design for higher stresses in the steel than had been used previously, something that will be dealt with in more detail later in the chapter, and this in part was his undoing. But the structural issues come down to too much load on members that were too weak – and it is the reason for this that needs to be explained.

The Forth Bridge

In building a bridge across a wide body of water, an engineer has to balance two difficulties: the difficulty of having a long span (or several long spans) and the difficulty of making foundations in the river. The simplest iron or steel bridge structure is just a series of girders resting on piers. The ill-fated Tay Bridge had been just that, and so is its replacement that is still in service today, but these designs required a great many piers and so a great many underwater foundations. When Baker came to bridge the Firth of Forth, he chose to limit his foundations to shallow water and an island and build gigantic cantilever structures to span across the gaps between them (Figure 1.3). A cantilever is a structure that is supported at one end only, unlike a beam, which is supported at both ends. A simple example is a shelf bracket, but Baker’s design involved cantilevering out in both directions from each of the supports, double cantilevers, like brackets fixed back to back.

Figure 1.3 The Forth Bridge.

To explain this design, Baker made a now famous demonstration to show exactly how it would work. He had his assistant, Kaichi Watanabe, sit on a board to represent the weight of one of the suspended spans supported between the cantilevers. The board was supported by a pair of inclined wooden struts while two men sitting on chairs with their arms held out grasped the ends of these struts. The men, their arms, the struts and the chairs they sat on represented these massive double cantilever struct...