Over Bridge near Gloucester

These pictures show the old bridge at Over, west of Gloucester, which was used from 1829 to 1974. This 150 foot span, designed by Telford after the previous bridge was damaged by ice in 1818, uses the cornes de vache technique, which may improve the flow in times of flood, previously used by Perronet for a bridge over the River Seine at Neuilly. It also gives the illusion of a flatter arch. The actual profile of the arch along its centre-line is not changed by this technique – what is changed is the sharpness of the transition from the full width of the river in spate to the narrow arch. With the horn shapes cut out there is some hope of reducing turbulence as the water is funnelled into the opening.

Telford’s plan was to build a 150 foot cast iron arch like those at Mythe and Holt Fleet: by using a standard design he saved money, but he varied the treatment of the abutments and flood arches to suit the local conditions, and perhaps to try out ideas.  But some important people in the city of Gloucester would not allow the use of cast iron, even though the bridge is not visible from Gloucester, which is why a heavy masonry arch was built in a rather unsuitable place.

The crown of the arch dropped by about ten inches, about 25 cm, when the centring was removed. This is shown in two of the pictures. Telford freely admitted that he was at fault in providing inadequate support for the thrust, and this error was probably a source of regret throughout his life. The line of thrust was still within the deep voussoirs, and so the bridge continued to stand, and it is still there, over 170 years later. The ground here is composed of soft alluvial material, and in fact Gloucester, like nearby Tewkesbury and Worcester, has been built entirely on or near the left side of the River Severn, where the ground is better. In fact Tewkesbury is not very close to the Severn at all. The first recorded bridge in the neighbourhood of Over was recorded in Domesday Book as existing in 1089. Another bridge was started in the sixteenth century, and it was this one that was damaged by ice carried down by the river.

To illustrate the difficulty of preventing movement of an arch, we should consider the formula dR/dH = H / R, where R is the rise of the arch and H is half the span. dR/dH is ratio of the drop in the arch to the increase of the half span as a result of movement. Thus if the rise is one tenth of the span, and one abutment moves by 1 cm, the crown will drop 5 cm. This calculation assumes that the arch behaves as a simple two hinged arch, which may not be the case.

The next picture (compressed laterally) shows the south side of the left bank abutment, showing distortion in the courses of blocks, some of which are denoted by red and blue lines. Do you think this could be the result of the movement of the arch, which is to the left of the picture? The other pictures, normal and exaggerated, show the appearance of the parapets from the west end of the bridge. Note that the roadway does not reflect the sag, having been built up to compensate.

These pictures show the Over bridge at a time of high flood. We can imagine that the shape of the arch with cornes de vache might make the flow a little easier when the water reaches the arch, especially as the water flows fastest at the surface, firstly because it experiences little friction, and secondly because it is being forced through the narrowing arch. In these pictures the water has begun to encroach upon the arch, as the third picture shows. The first two pictures show the vertical chamfers that the cornes de vache provide.

Some Severn bridges upstream have much less clearance, and are submerged more often. This picture shows Maisemore bridge, the next masonry bridge upstream from Over, during the same flood period. We can see how much of the arch has been engulfed. This is a disadvantage of the single span flat arch. It avoids piers in the river, and the dangers of scouring, but it is vulnerable to pressure from floods. In this example, some water flowed across fields on the left bank of the Severn, and flowed back to the Severn downstream of Maisemore.

The diagram below shows a rough idea of the horizontal cross section of the bridge at a normal water level and at very high flood level. Perhaps Telford wisely provided the clearance because the Severn estuary has a very high tidal range. Flooding can be severe when spring tides coincide with flood water coming down river, though of course the incoming tide reduces the speed of the water. But when the tide goes out the high level spurs the water to higher speeds than before. The Severn bore, one of the biggest in the world, also swirls through the arch.

Just downstream of Over bridge is a railway bridge, probably the ugliest bridge depicted in this web-site. The clearance above the water is very small compared with that offered by the old road bridge. Why is this? In the eighteenth century, the river Severn and the nearby Sharpness canal were two of the busiest channels in Europe, and all bridges had to offer navigational clearance for quite large boats. By the time the railway bridge was built, the river traffic was no longer important.

Many boats were towed along rivers and canals by people or horses, using towpaths. To take the towpaths under a wide flat arch such as that at Over would have required an expensive increase in the span, so at this and other locations, the towpaths went through subsidiary openings in the abutments. Two questions can be asked about these. Does the thrust of the arch go over or under the openings? How did the towing people or animals manage when they reached the bridge, going upstream, and going downstream? The next pictures show the opening in the abutment on the left bank.

Note the distressed stonework, including many examples of exfoliation. More examples of exfoliation are shown in the next pictures, of blocks in the parapets and under the soffit. You might expect a north-south difference in these effects. Why?