The Royal Albert Bridge at Saltash

Back to Bridges     Back to Home Page

RABridge1846S.jpg (231998 bytes)     RABridgeIKB1849.jpg (142062 bytes)     RABig35.jpg (461141 bytes)

The Royal Albert Bridge is an arch bridge.

The Royal Albert Bridge is a suspension bridge.

Both statements are true. The unusual design of this bridge was a result of the constraints imposed by the geographical situation and the demands of navigation, together with the range of materials available at the time of construction.  

Large parts of Devon and Cornwall are very hilly, and in addition there are some wide drowned valleys which even today form significant obstacles to transport. In creating the south western part of the Great Western Railway, I K Brunel had to design a large number of bridges and viaducts. In the early days of construction, many of the viaducts were built in timber; the last of these survived until 1934, so they were by no means thought of as short term solutions.

The Tamar estuary at Saltash is about 1100 feet wide, and the navigational requirement was for a clearance of 100 feet over most of that width. The estuary had a depth of about 70 feet of water, under which there was a further 15 feet of mud and clay. Thus from solid rock to the bridge spans, the height was about 185 feet. Brunel eventually decided on two spans of 455 feet, after having considered a single span of 850 feet. For navigational reasons, the estuary could not be blocked by falsework, so the two spans were built on land and floated to the piers on pontoons. The rising tide was used to lift them on the pontoons, and the falling tide was used to lower them on to their first supports.

In 1849, when preparatory investigations were begun, the main constructional material was wrought iron: the age of cheap steel was many years away.  A span of over 400 feet would be a tremendous feat, as were those of Robert Stepehenson's tubular Britannia bridge. The rectangular tubes of the Britannia bridge were tall enough to contain the railway loading gauge, but the ratio of depth to span was such that Stephenson provided extra height in the towers in case suspension cables were needed. Brunel realised that if the depth of the spans were increased, the forces would be reduced, and the spans could be lighter. He also determined through tests that round sections are more efficient than rectangular ones, and settled on an oval section whose height is about three quarters of its width. The width provides lateral stability against wind pressure, and also enables vertical hangers to be attached to the sides with space between for the loading gauge.

RABridge1849Sm.jpg (228902 bytes)The apparently curved tubes are in fact composed of straight sections, so that the outline is a small part of a huge polygonal torus. Within the tubes are flanges around the circumference and longitudinal stiffeners also. The tubes are in compression, being in the form of shallow arches. These hollow structures and the flanges can be likened to the bones of human legs, which comprise an outer layer of compact bone, and an inner structure of cancellous or spongy bone, which consists of many small struts. The shapes of bones and their are exceedingly subtle, reflecting the variety of stresses and attachments that they have to bear. To design an artificial structure to such a degree of refinement would be incredibly costly, and would achieve nothing. Whereas a minute small degree of superiority can be significant under natural selection, where the type of engineering costs are completely different from ours, human engineering refinement is halted by cost and time at a much simpler level.

Because there are no abutments, the outward thrust has to be contained, and this is achieved by the inward pull of the suspension chains. Thus the spans can be seen as tied arches or as self-anchored suspension bridges. Can they be seen as trusses? There are indeed numerous diagonal members to brace against wind pressure and longitudinal movements, but these are very slender compared with the vertical hangers, emphasising that the spans are not trusses in the usual sense of the term.

RAPlate1849.jpg (84938 bytes)RATrain1856S.jpg (110117 bytes)RATrain35S.jpg (306951 bytes)The deck employs deep plate girders on each side of the railway track, helping to stiffen the bridge against the localised weights of locomotives, which have greatly increased since the bridge was completed in 1859, though the second photograph shows a rather light example. The tubes and the chains are as they were on completion, but a number of modifications have been necessary on account of corrosion, wear and tear, and increased loads.

RABridgePier1849.jpg (196240 bytes)The central pier comprises four octagonal cast iron tubes, internally and externally braced, all resting on masonry which in turn rests on an early form of pneumatic caisson which was taken down to bedrock.

It is often a good idea to look for a question to ask, on seeing a structure. A good question here is - "How did they attach the chains to the tubes?" The forces in the tubes are in principle diffused all round the circumference, but the forces in the chains and the hangers are highly localised. The provision of attachments and the passing of forces between structures is often one of the main problems. Unsustainable stress concentrations have to avoided at all costs.

Here is another question. How is the weight of the deck shared between the arch and the chains? Is this even a meaningful question? At the ends of both, the horizontal forces must be equal, since there is no horizontal force from outside the span. The horizontal forces in the arch and chains must remain about the same throughout, as the main connecting members are vertical. Does the sharing of the load between arch and chains depend on (a) the lengths of the hangers between these, and (b) the lengths of the hangers between the chains and the deck girders?  Are the spans indeterminate in some way?

Let's now look at the answers to these questions.

SaltashAttach1849.jpg (69074 bytes)RABridgeAttach1856.jpg (125787 bytes)RAMiddle35B.jpg (165942 bytes)In this bridge, the chain attachments are shrouded in the portals at the ends of the spans. The tubes are completed at each end by flanged plates. The attachment of the tubes is achieved by an elaborate system of plates which spread the chain forces into the tubes. The tubes are also locally strengthened at the hanger attachments. The portals have no structural significance, but the bridge would be much diminished without them, partly because of the abrupt termination of the tubes by large flat surfaces. The tubes could in principle have been tapered towards the ends, but the cost in calculation and building complexity would have been significant. Like the pylons of Sydney Harbour bridge and Hell Gate bridge, these portals serve as a kind of punctuation mark. Many buildings have porticos and porches, steps and ramps, to signify the transition from exterior to interior. Bridges, especially in towns and cities, cannot be considered solely as utilitarian objects: people have to live with them and see them every day. The appearance of bridges is discussed in a dedicated page.

RAForcesS2S.jpg (111845 bytes)Here is a picture that tells us something about the forces in the main members. The vertical red arrow represents one half the weight of the span. We know that this force at the pier must be vertical because the whole span acts as a beam, with no external horizontal reaction. The thick red and blue arrows represent the compressive force in the tube and the tension in the chain. Because the three forces are in equilibrium, the three vectors must form a closed triangle. This means that if the slopes of the tube and the chain are equal, the forces in them must be equal, and so they must be sharing the load equally. Even if the slopes differ slightly, the effects on the ratio of the lengths of the vectors will be small. From the diagram we see that the forces at the ends of the tube and the chains are about twice the half-weight of the span, ie about equal to the weight of the span. Since there are two chains and only one span, the tension in each chain at the attachments is about half the weight of the span. The weight of each span is about 1060 tons.

RALandSpans35.jpg (339499 bytes)SaltashBig10OctS.jpg (330349 bytes)Here are pictures of the eastern land spans, which are on a curve, like the western ones. Assuming the height and width of the piers and the radius of the curve, it should be easy to calculate a rough value for the outward force caused by a locomotive of a given weight and speed. Hence it should be possible to calculate the maximum speed for which the line of thrust remains in the allowed small region at the bases of the piers. The maximum speed for the crossing is now 10 mph or 16 kph.

The lateral force FL needed to accelerate a mass m that moves with a velocity v on a curve of radius r is as follows.

FL = mv2/r

The vertical force FV is simply the weight mg, where g is the acceleration due to gravity. Since the resultant has to be almost vertical for stability, the angle A of the resultant from the vertical is given by

A = FL / FV = mv2/rmg = v2/rg.

Whether this speed is greater or less than the speed restriction imposed by considerations such as vibration depends on the details of the structure. The tighter the curve, the smaller v must be.

The completion of this magnificent bridge in 1859 was overshadowed by the severe ill health of Brunel, who was only able to see the complete structure by lying on a specially modified wagon which was pulled slowly over the bridge. He died soon after.  Ask most English people to name an engineer, and they will almost certainly not name a Stephenson, or Locke, Smeaton, Rennie, Telford, Rastrick, Hazeldine or indeed many others (and this list includes only British engineers). They will almost certainly name I K Brunel. 

It is idle to compare different people who had to solve different problems: what is indisputable is that Brunel's solution to one of his biggest problems is a remarkable and unique structure.  In its use of both suspension and arch principles it reminds us of the importance of the funicular, and the opposition of the two ideas might even remind us of the symmetries found in poetry and music, such as the use of inversion. The huge arching tubes are not unlike the back of a great animal, with the chains representing the tensed belly muscles holding the internal organs.

These spans are easily understood in principle, as are many great constructions, but it is characteristic of many recent developments that the layman can no longer be expected to know what is going on. This has of course always been true of falsework and of underground and underwater work, but the advent of reinforced concrete and pre-stressed concrete means that many subtleties are hidden. In architecture too, the structure is often hidden behind walls of glass that are obviously not structural. The entire world of technology has become vastly more complicated, and whereas you can understand the main ideas behind a steam engine in a museum without a knowledge of thermodynamics, a museum of communications and computers could not possibly hope to display and explain integrated circuits containing hundreds of thousands of parts or the complexities of the software algorithms that control them.

Excellent accounts of the design and construction of the Royal Saltash bridge can be found in these books -

Track Topics - A book of railway engineering for boys of all ages, by W G Chapman

The Works of Isambard Kingdom Brunel, edited by Sir Alfred Pugsley.

Track Topics also describes the ingenious way in which the plate girders of the land spans were replaced in 1928. Careful preparation and the use of a special wagon enabled a span to be renewed in an average time of about three and a half hours, without having to remove the track or the deck.

Near the east end of the bridge, the track passes over one of the many railway viaducts found in Devon and Cornwall; it is seen in one of these photographs.  The track slopes down to this viaduct, as the first picture shows, and the viaduct itself is on a slight slope.

TamarBridges1838.jpg (313827 bytes)  RABridge1847S.jpg (214667 bytes)  RABridgeTies1854.jpg (101296 bytes)  RAViaduct1843.jpg (417859 bytes)

RANoTrain1849Sm.jpg (54241 bytes)RAWithTrain35S.jpg (125889 bytes)Lastly, here are two pictures that show clearly the deflection of a span when loaded by a train. There is no such thing as a rigid body: every body deflects under the action of external forces until it generates within itself exactly the right forces to balance the external ones. These pictures would have been more convincing had they been taken from the same viewpoint with the same equipment, as the movement of the bridge against the background would have been conclusive. One function of the designer is to achieve an acceptably small deflection with the most economical arrangement of material.

Click here for a series of superb photographs of the nearby suspension bridge, many of which include the railway bridge.

Construction of the Royal Albert bridge

Back to Bridges     Back to Home Page