The Tay Bridge carries rail traffic across the Firth of Tay in Scotland between Dundee and the suburb of Wormit in Fife. Its span is . It is the second bridge to occupy the site.
Plans for a bridge over the Tay to replace the train ferry service emerged in 1854, but the first Tay Bridge did not open until 1878. It was a lightweight lattice design of relatively low cost with a single track. On 28 December 1879, the bridge suddenly collapsed in high winds while a train was crossing, killing everybody on board. The incident is one of the worst bridge-related engineering disasters in history. An enquiry determined that the bridge was insufficiently engineered to cope with high winds.
It was replaced by a second bridge constructed of iron and steel, with a double track, parallel to the remains of the first bridge. Work commenced on 6 July 1883 and the bridge opened in 1887. The new bridge was subject to extensive testing by the Board of Trade, which resulted in a favourable report. In 2003, the bridge was strengthened and refurbished, winning a British Construction Industry Engineering Award to mark the scale and difficulty of the project.
First bridge Origins and concept Proposals to build a bridge across the Tay date to 1854 but it was not until 15 July 1870 that the (33 & 34 Vict. c. cxxxv) received royal assent. On 22 July 1871, the foundation stone of the bridge was laid.
The bridge was designed by engineer Thomas Bouch, who received a knighthood following the bridge's completion. The bridge was a lattice-grid design, combining cast and wrought iron. The design had been used by Thomas W. Kennard in the Crumlin Viaduct in South Wales in 1858, after the use of cast iron in the Crystal Palace. The Crystal Palace was not as heavily loaded as a railway bridge. An earlier cast-iron design, the Dee bridge collapsed in 1847, having failed because of poor use of cast-iron girders. Gustave Eiffel used a similar design to create several large viaducts in the Massif Central in 1867.
The original design was for lattice girders supported by brick piers resting on the bedrock, shown by trial borings to lie at no great depth under the river. At either end of the bridge, the single track ran on top of the bridge girder, most of which lay below the pier tops. At the centre section of the bridge (the high girders), the railway ran inside the bridge girder, which was above the pier tops to give clearance for the passage of sailing ships. To accommodate thermal expansion, there were non-rigid connections between girders and piers.
As the bridge extended out into the river, by December 1873, it became clear that the bedrock lay much deeper, too deep to act as a foundation for the bridge piers.Minutes of Evidence p. 402. – evidence of Sir T Bouch. Bouch redesigned the bridge to reduce the number of piers and increase the span of the girders. The pier foundations were no longer resting on bedrock; instead they were constructed by sinking brick-lined wrought-iron caissons onto the riverbed, removing sand until they rested on the consolidated gravel layer which had been misreported as rock, and then filling the caissons with concrete.
To reduce the weight that the ground underneath the caissons would have to support, the brick piers were replaced by open lattice iron skeleton piers. Each pier had multiple cast-iron columns taking the weight of the bridging girders, with wrought iron horizontal braces and diagonal tiebars linking the columns to give rigidity and stability. The basic concept was well known, having been used by Kennard in the Crumlin Viaduct in South Wales in 1858.Maynard, H. M. "Handbook to the Crumlin Viaduct." J M Wilson: London, 1862. retrievable via Google Books Maynard claims the loading on the Crumlin viaduct foundations to be 1/5th what they would have been had brick piers been used. Bouch had used the technique for viaducts, including the Belah Viaduct (1860) on the South Durham & Lancashire Union Railway line over Stainmore, but for the Tay Bridge, even with the largest practicable caissons, the pier dimensions were constrained by their size. Bouch's pier design set six columns in a hexagon maximising the pier width but not the number of diagonal braces directly resisting sideways forces. Structure Crumlin viaduct Belah viaductmost data from . Retrieved 22 January 2012 (the quoted height between joints/bracing is clearly inconsistent with the accompanying photo so the number of column sections is taken from another reference;) Tay Bridgeall data from Annexe to Report of Court of InquiryEngineer (year of opening) Kennard (1858) Bouch (1860) Bouch (1878) Single span ) Pier height Pier width at top Pier width at base Columns per pier 14 (1-3-3-3-3-1) 6 (2-2-2) 6 (1-2-2-1) Sections per column 10 11From diagram 276 in History of Bridge Engineering HG Tyrell Chicago 1911 7 Diagonal tiebars giving lateral bracing (per pier)calculable from previous 2 rows 180 88 28 FateDemolished 1966-7 Demolished 1963 Failed in service 1879
Design details
The engineering details on the Tay Bridge were considerably simpler, lighter, and cheaper than on the earlier viaducts. The machined base of each column section docked securely into a machined enlarged section of the top of the section below."like the spigot and faucet of a domestic water pipe" was felt be a useful analogy at the Tay Bridge Court of Inquiry but would probably baffle modern householders The joint was then secured by bolts through matching holes on lugs (Crumlin) or flanges (Belah) on the two sections. This 'spigot and faucet' configuration was used, apparently without machining, on some Tay Bridge pier columns, but on some the bolts were relied upon to ensure correct alignment. In the event, the joints were made using undersized bolts, of a smaller diameter than that which would just go through the hole. This made assembling the column easier, as the bolt holes would not need to align exactly before inserting the bolt. However, this allowed the two members, so joined, to move relative to each other under load, weakening the column.
On the Tay Bridge the diagonal bracing was by means of flat bars running from the top of one column-section diagonally down to the bottom of the adjacent column section. The top connection was to a lug that was an integral part of the column casting."..although cast-iron lugs are peculiarly liable to fail from shock, they have been used in precisely this way successfully in tens of thousands of yards of viaduct; therefore, Sir Thomas Bouch was only following precedent in using them here. Mr Barlow will know that on the Bombay and Baroda line they were used for a great many viaducts" Minutes of Evidence – evidence of B Baker, p. 507 The bottom connection was to two sling plates bolted to the base of the equivalent section on an adjacent column. The bar and sling plates all had matching longitudinal slots in them. The tie bar was placed between the sling plates with all three slots aligned and overlapping. A gib was driven through all three slots and secured. Two cotters, metal wedges, were then positioned to fill the rest of the slot overlap, and driven in hard to put the tie under tension. Horizontal bracing was provided by wrought iron channel iron.Minutes of Evidence – evidence of H Law, p. 246 The various bolt heads were too close to each other, and to the column for easy tightening up with spanners; this coupled with lack of precision in the preparation of the channel iron braces led to various on site fitting expedients (one of them described by a witness to the enquiry as "about as slovenly a piece of work as ever I saw in my life".Minutes of Evidence – evidence of H Law, p. 331
On the Crumlin and Belah Viaducts, however, horizontal bracing was provided by substantial fitted cast-iron girders securely attached to the columns, with the diagonal braces then being attached to the girders.Details of the bracing of the Crumlin viaduct are given in Maynard, see also a photograph of painters at work on a pier in 1914 to be found at Crumlin Viaduct which will probably be found of considerable assistance The chairman of the Court of Inquiry quoted at length from a contemporary book praising the detailed engineering of the Belah viaduct piers, and describing the viaduct as one of the lightest and cheapest of the kind that had ever been erected. ... It is a distinguishing feature in this viaduct that the cross, or distance girdersmade of cast iron Minutes of Evidence – evidence of Sir T Bouch, p. 430 of the piers encircle the columns, which are turned up at that point, the girders being bored out to fit the turned part with great accuracy. No cement of any kind was used in the whole structure, and the piers when completed, and the vertical and horizontal wrought-iron bracings keyed up, are nearly as rigid as though they were one solid piece... .... The fitting was all done by machines, which were specially designed for the purpose, and finished the work with mathematical accuracy The flanges of the column were all faced up and their edges turned, and every column was stepped into the one below it with a lip of about 5/8 of an inchie c 16 mm in depth, the lip and socket for it being actually turned and bored. That portion of the column against which the cross girders rested was also turned. The whole of these operations were performed at one time, the column being centred in a hollow mandril-lathe. After being turned the columns passed on to a drilling machine, in which all the holes in each flange were drilled out of the solid simultaneously. And as this was done with them all in the same machine, the holes of course, perfectly coincided when the columns were placed one on the other in the progress of erection. Similar care was taken with the cross-girders, which were bored out at the ends by machines designed for that purpose. Thus, when the pieces of the viaduct had to be put together at the place of erection there was literally not a tool required, and neither chipping or filing to retard the progress of the work.Rothery in Court of Inquiry report pp 43–44 quoting Humber op cit p 224-5 Either, said the chairman, the Belah viaduct had been over-engineered, or the Tay Bridge had been under engineered.Rothery in Court of Inquiry report pp 43–44
Construction Whilst Bouch was in the process of revising his design, the company which had been awarded the contract for the bridge's construction, Messrs De Bergue of Cardiff, went out of business.Reportedly, the company's owner, Mr De Bergue, had gone insane and then died – Minutes of Court of Inquiry. During June 1874, a replacement contract for the work was issued to Hopkin Gilkes and Company, successors to the Middlesbrough company which had previously provided the ironwork for the Belah viaduct.Minutes of Evidence – evidence of Sir T Bouch, p. 406Bouch’s brother William was Locomotive Superintendent of the Stockton and Darlington Railway and also a director of Gilkes which made locomotives for the S&D. On his brother's death in January 1876, Bouch as his sole executor effectively inherited his shares, and also his personal guarantee of the firm's borrowings. The firm was not in good enough shape for Bouch to dispose of his interest, and he lost a large amount of money when it went bankrupt during 1880. In principle, there was therefore a conflict of interest with Bouch's duty to advise the railway company on the acceptability of the contractor's work. (Minutes of Evidence – evidence of Sir T Bouch, p. 440) Gilkes had originally intended to produce all the bridge ironwork on Teesside, but in the event continued to use a foundry at Wormit to produce the cast-iron components, and to carry out limited post-casting machining operations.