Cumberland Basin Bridges (Bristol)
From Graces Guide
There are four notable bridges at the Cumberland Basin, at the western end of Bristol Harbour. Two of these date from the 19th century, while two were built in the 20th:-
Merchants Road swing bridge, built by John Lysaght in 1925
Plimsoll Bridge: a large 1960s swing bridge carrying heavy traffic over the entrance locks to Cumberland Basin. Built by Head, Wrightson and Co
19th Century Bridges
Below the Plimsoll Bridge are two old wrought iron bridges, which could once be swung across the two entrance locks. These are of similar design, having tubular top members. One bridge is fixed, and spans the now-sealed Brunel (south) Lock. The other, longer bridge is in the location where it ended its days as a swing bridge over the north lock ('Howards Lock').
There was a third wrought iron bridge of similar design. The three bridges were built in 1849, 1864 and 1875.
One of these bridges was designed by Isambard Kingdom Brunel. The other survivor, of very similar design, was built much later. There was some uncertainty about which was the earlier bridge. The situation was complicated because of major alterations to the two locks which the bridges spanned. However, the question was thoroughly addressed in 2006, and the outcome is summarised below.
In 2006 Dr David Greenfield undertook a thorough analysis, and his excellent report is available online .
The author concluded that the bridge now sitting parallel to the north lock (Thomas Howard's Lock) is 'Brunel's Swivel Bridge', i.e. the same bridge that was originally installed over the south lock ('Brunel's Lock') in 1849. He concluded that it was adapted for use at Howard's Lock around 1873.
He also concluded that the bridge now permanently closed across the sealed south lock (Brunel's Lock), and known as the 'Replica Bridge' is the same bridge that was installed there in 1875-1876.
The following information is drawn from Dr Greenfield's Paper:
- In 1844 Brunel was commissioned to design a replacement South Entrance Lock. The ironwork was sub-contracted to George Hennet of Bristol and Bridgwater. To span the lock, it was decided in 1849 to install a swivel bridge, designed in Brunel's office in Westminster. Robert Brereton, Brunel's chief assistant, signed off the drawings and letters, and was probably responsible for the detailed design, with Brunel closely monitoring the design and execution of the work. The bridge was first operated on 8 October 1849. Overall length 121 ft 9".
- In 1863 a new 'swivel' bridge was ordered, to traverse Jessop's North Entrance Lock. This was of the same general design as Brunel's bridge. The bridge was built by Hennet, Spink and Else of Bridgwater. Its length was 77ft overall.
- In the 1860s the restriction imposed by the remaining Jessop locks was proving intolerable, and work started on a new, larger North Lock in 1868 ('Howard's Lock'). At this time, the 1863 bridge was removed, and reputedly shortened and re-erected over the entrance to Bathurst Basin.
- In 1872 Howard was authorised to adapt Brunel's bridge and install it to cross the north (Howard's) Lock. The bridge needed to be shortened by 10 ft, and the counterweights adjusted. The lock was opened for traffic on 19 July 1873. At some point the bridge was converted to hydraulic operation.
- A new bridge was needed to cross Brunel's Lock, replacing a temporary wooden bridge. This was the so-called Replica Bridge built in 1875-6 by Edward Finch & Co Ltd, of Bridge Works, Chepstow
Additional Observations (2012)
The '1849' bridge is of particular interest because it was the first of Brunel's 'tubular' bridges. It is also an early example of a wrought iron plate girder bridge.
Some constructional and design aspects are discussed below (J Ditchfield), together with some additional information supporting Dr Greenfield's conclusions.
The 2006 report referred to the bridge having been shortened by 10 ft. Referring to Photo 3, there are three vertical 'web stiffeners' near the outboard end of the 'northern bridge' before the last vertical plate at the end. Logically, these vertical plates would have been spaced symmetrically, but in fact while the typical spacing is approx 19 ft 6", the end plate is only 9 ft 3" from the adjacent web, consistent with a shortening of just over 10 ft.
The vertical web stiffeners on the 'northern' bridge use flat plates, whereas the 'southern' bridge uses tee bulb sections. Such sections were not available in 1849. Note that these stiffeners (on the southern bridge) are spaced at 10 ft intervals. The southern bridge shows neater workmanship and is in better condition, consistent with later construction.
The northern bridge is located with its pivot on the 'island' between the two locks. An 1860s photograph  shows that the bridge over the original northern lock (Jessop's Lock) had its pivot on the opposite side, i.e. the 'mainland' (northern) side. The bridge had only a short overhang at the counterweight end, dictated by the limited width between from the edge of the lock and the street. The photo also shows symmetrically-spaced web stiffeners. Taking these as 10 ft apart (scaled from the photo), the length of the bridge would have been approx 70 ft. Incidentally, the same photo shows another swing bridge, over the locks at the eastern end of Cumberland Basin.
Further support for Dr Greenfield's conclusions comes from an 1868 photograph . This shows that the design of the bridge over the south lock ('Brunel's Lock')in 1868 is entirely consistent with the bridge now fixed next to the north lock. Specifically, it has just three intermediate web stiffeners on the outside of each girder, biased to the outboard (non-counterweight) end. By scaling, these are spaced at 19.5 ft intervals, consistent with the measurements taken on the bridge now alongside the north lock. Another photograph showing the two bridges c.1865 - 70 may be found in 'Bristol's Floating Harbour: The First 200 Years'
Given the great unsupported span of the Brunel bridge, its girders seem surprisingly shallow. This may reflect the limited size of wrought iron plates available from rolling mills that time, although the height could of course have been increased by riveting on extra plates. By the time the '1875' bridge was built, deeper girders could have been readily adopted. Presumably no need was seen to alter this aspect of the design when the newer bridges were built.
Although the length of rolled plates available in 1849 was limited to approximately 10 ft, it will be noted that one plate on the tubular part of each girder is 20 ft long. This is above the reinforced pivot area. In fact these plates were 'fire-welded' from two 10 ft plates.
A curious aspect of the 1849 bridge is that the top tubes contain longitudinal tie bars. Evidence can be seen in Photos 11, 12 & 13. Note the rectangular bars. Those visible in the hole in Photo 13 are tapered cotters. These are in the form of opposing pairs of wedges, inserted from either side and driven in to the tension the tie rods. Reference to Photo 13 shows the presence of square headed bolts adjacent to the hole and the supposed 'cotter'. These indicate the presence of an internal sleeve against which the cotters can react. Presumably the purpose of the tie bars was to prevent excessive strain or droop when the bridge was in the 'open' position, when the top tubes would be under tension. The tapered cotters would allow the top tubes to be pre-tensioned.
It is not known how the 'balloon' section top tubes were riveted. Specifically, it is not known how the hot rivets were inserted from inside the tubes and then held while the heads were closed by hammering from the outside. The possibility of a small boy being sent along the tubes can be discounted, not least because of the internal obstructions. One possibility is that the girder was laid on its side, and the hot rivet was inserted while held in a long 'forceps'-like tool. One leg of the 'forceps' would hold the hot rivet in a simple clip, and this leg would go inside the tube. The corresponding end of the other leg would simply have a sighting hole in the end - an 'eye'. This 'eye' would be lined up with the required rivet hole, and it follows that the rivet would be lined up, too. After inserting the rivet, the setting tool would be quickly withdrawn and a wheeled jack-like dolly pushed into place.
Another surprising aspect of the 1849 bridge design is the sparsity of web stiffeners. The '1875' bridge is much more generously provided in this respect. However, a degree of extra stability would be conferred to the torsional stiffness of the main beams by the presence of the cross beams which support the deck.
Regarding the possibility that the 1863 bridge was shortened and removed to Bathurst Basin, a c.1900 photograph does indeed show a bridge of similar design located between the two lock gates at the junction of Bathurst Basin and the New Cut. Based on the distance of the last web stiffener to the end of the bridge, this bridge had almost certainly been shortened, consistent with it having previously been installed elsewhere. A larger steel girder bridge now occupies this location.
An enthusiastic team of volunteers is currently engaged in a restoration project on the 1849 bridge, which they refer to as 'Brunel's Other Bridge'. They have an excellent website here which includes progress reports with photographs
Information Relating to Photographs
Photo 1 & 2: This excellent 1929 aerial photograph appears on a display board by the locks near Plimsoll Bridge. The steamship in the foreground is in the north ('Howard's) lock. Behind the ship is the Brunel bridge, and to the right is the 1875 bridge. The 1925 road swing bridge can be seen at the far end of the basin, and beyond that is a rail swingbridge in the open position.
Photo 3: Note the considerable span between the near end of the bridge and the pivot point
Photos 4 & 5: The large outboard span dictates the need for a heavy counterweight at the other end, and this was made from iron castings. These castings also incorporate the axles for the counterweight's rollers. It is difficult to argue that the bridge is a thing of beauty, but clearly great effort has been taken to blend the curves of the cast and wrought iron components. No such care for aesthetics when the extra weights were bolted on in the 1960s!
Photo 6: Brunel bridge, showing cross members which supported the wooden decking. Note weld repairs on nearest vertical web
Photo 7: More weld repairs (including replacement of part of the tubular section)
Photo 8: This shows where the 'counterweight end' of the 1849 bridge met the roadway on the island between the locks. Note the track for the counterweight's wheels. An iron stop is provided at the end of the track
Photos 9 & 10: These show the pivot pin and one of four rollers, and one of the two cable connectors for slewing the bridge. The fixed pivot and the base for the rollers was an iron casting fixed to the masonry. Obviously levelling of this casting would be very important. Note: Latterly the bridge was slewed by cables worked by hydraulic rams, but originally it was hand cranked. The crank's spindle passed through one of the girder's vertical plates (the hole can still be seen). On the end of the spindle was a small bevel gear, which turned a vertical spindle via a larger bevel gear. At the bottom of the vertical spindle was a small spur gear, which engaged with a curved rack. This rack was fixed to the inside of the circular iron casting fixed in the masonry. Although not apparent externally, the axles and bearings for the rollers are very unusual, having a double conical profile, tapering down from the ends to the middle.
Photo 11: Counterweight end, looking from the deck. Note the rectangular bar in the tubular section. This secures the end of the internal tie bar.
Photo 12: The other side of the bar seen in Photo 11. Note that the square bolt heads are stamped with numbers. In 1849, inconsistencies in threaded fasteners and holes probably made it important to ensure that bolts were used in the holes for which they were made.
Photo 13: Example of a 'cottered joint' at an intermediate position along the top tube.
Photo 14: Lying on the bridge deck is a shaft with two cams and support bearings. It is clear that these bearings were fixed to the black and yellow painted pads in Photo 14. It is assumed that after the bridge had been swung closed, the cams were rotated to raise the deck to match the road, and also to support the weight. Small rollers are fitted to the end of the bridge to reduce friction when turning the cams
Photo 15: This shows where the Brunel bridge met the road on the north bank. Note the replaceable 'ramp' on the cast iron column, and the notch to receive a simple latch fitted to the end of the bridge
Photo 16: Looking up inside the 'latch' post. This shows the screw which raised and lowered two 1.5" dia bars to rotate the cam shaft.
Photo 17: The 'latch' post showing the squared shaft near the top for a handle to turn the leadscrew to operate the cam shaft.
Photo 18: Latch on the end of the Brunel bridge
Photo 19: The '1875' bridge, at the non-counterweighted end. Note the larger number of vertical web stiffeners compared with the Brunel bridge. Also, these are made from 'bulb' section rather than plain flat plates.
Photo 20: The '1875' bridge at the counterweight end. The counterweight has been removed to avoid jutting out into the later road and pavement.
Sources of Information
-  'BRUNEL'S SWIVEL BRIDGE', CUMBERLAND BASIN, BRISTOL: REPORT 01 - AUTHENTICATION OF THE BRIDGE: Author: David Greenfield: Revised: 10 May 2006
- 'Hotwells and the City Docks on old postcards' by Brian Lewis: Janet and Derek Fisher
- 'Bristol As It Was 1845 - 1900' by Reece Winstone, 1983: Plate 417
- 'Bristol's Floating Harbour: The First 200 Years' by Peter Malpass and Andy King, Redcliffe Press, 2009: p.52
- Postcard entitled 'General Hospital, Bristol'. Published by Max Ettlinger