Dee Railway Bridge

Collapse of the bridge. Picture published in 1894.

The first railway bridge over the River Dee in Chester was built in 1846, and collapsed on 24 May 1847, with the loss of five lives.

The bridge was built by the Chester and Holyhead Railway, and was initially used by contractors' traffic for the Chester and Shrewsbury Railway.

Robert Stephenson was responsible for the design of the bridge, as Chief Engineer. He is unlikely to have been involved in the detail design, but he would have had to approve it.

The bridge had three spans of cast iron girders, each made of three lengths bolted together with complex joints. In recognition of the inferior strength of cast iron in tension, the bottom flanges were much larger than those at the top, in accordance with the general recommendations of Eaton Hodgkinson.

Trusses, in the form of flat wrought iron bars (tendons), were attached to the assembled beams. These are just visible in the drawing. On each side of the central section of the beam, a tendon was attached at low level. From each end another tendon sloped up to the end of the beam, where it was connected to a bracket bolted to the top of the beam. This was a fatal mistake, as it caused the tension in the tendon to act at a point well above the neutral axis of the beam. So, when weight was applied to the beam, the increased tension in the tendons applied a bending moment to the top of the beam, thereby increasing the compressive stress in the top flange and the tensile stress in the bottom flange of the cast iron beam.

On the morning of 24 May 1847 Robert Stephenson inspected the bridge and was concerned that the timber deck might be set on fire by hot cinders. He ordered 5" of ballast to be laid for protection. This would have increased the static load on the bridge.

That evening, the 6.15 pm train was accelerating out of Chester, when one of the girders collapsed. Several carriages fell, but the locomotive had enough momentum to reach the abutment. Five lives were lost, and an inquest found that the girder failed because it had insufficient strength to support the passage of a fast train, and not because of derailment of the train, or other fault.

An inquest was opened on 25 May. The Chester and Shrewsbury Railway Co and the Chester and Holyhead railway Co both offered 'every facility'. Robert Stephenson and Mr Betts would be summoned to give evidence. The inquest was adjourned until 28 May.

At the reopened inquest, General Pasley criticised the arrangement of the tendons. Robert Stephenson handed the coroner a report which he had made to his directors, stating that he had found nothing wrong with the bridge, and that the fracture was caused by a sideways blow and not a downward force, and that the cause was the derailment of the train. He disagreed with Gen. Pasley's opinion. Messrs Locke, Vignoles, and Gooch concurred with Stephenson's conclusions. Sir Edward Walker, foreman of the jury, asked Stephenson whether he intended to make any alteration in the construction of the bridge. Stephenson replied that he proposed to make certain alterations with a view to appeasing the public mind. ! Sir Edward Walker responded that he would not travel over it.

Capt. Simmons R.E. and James Walker (Admiralty Engineer) quickly carried out an investigation and reported on the failure. See Report here.

On the final day of the inquest, 17 June, Henry Robertson presented his calculations on the strength of the bridge, and the following extract is perhaps worth quoting:-

"The fracture at the centre, from the position of the fallen portions, and of the middle tension rod wrapped over the girder, and especially from the form of the fracture, appears to me to have first taken place. This fracture I consider to have resulted from the weakness of the top flange, which was compressed and broken by the strain, arising from the rolling weight of the engine and tender, and the vibratory weight of the structure itself, increased to a large extent by the deposit of twenty-five tons of ballast on the roadway immediately before the accident. This compression is remarkably evident by the bulging out of the metal at the point of the parting at the top of the web or vertical portion of the girder. In estimating the strength of the girder, I am of opinion that the tension rods, from the form of the section of the girder, weakened it, and so threw an undue strain, by compression on the top flange ; but assuming that they did not weaken it, and applying the formula as given by Eaton Hodkinson, F.R S., to the girders — by one formula the breaking weight is equal to 61 1/4 tons, by the other the breaking weight is equal to 16 tons. Now it has been an established rule in practice that one-third or one-fourth or the breaking weight is the safe working weight to which a girder should be subjected, and the larger the size the smaller ought to be the proportion; taken, therefore, one-quarter of 56 (the breaking weight), it follows that the safe weight to which one of the girders ought to be subjected is 18 1/2, and the two girders 37 tons. The weight of the timber platform, beams, rails, chains, &c, exclusive of the girder, according to an approximate calculation I made, is 19 tons 6 cwt., and adopting the rule that a uniform weight, diffused over the beam, is equivalent to one-half that weight suspended at the centre, this becomes equal to a weight suspended at the centre of 9 tons 13 cwt. The equivalent weight of an engine and tender of 33 tons 10 cwt. 2 qrs., suspended at the centre of the beam, I estimate at 32 tons, making a strain of 41 tons 13 cwt. against 37 tons, the safe working strain to which the bridge ought to be subjected. However, on the afternoon of the accident, immediately previous to the passing of the train, the bridge was subjected to an additional strain, by the laying on of five inches of broken red sandstone ballast, amounting to a weight over the bridge of 25 tons, which is equivalent to a weight suspended at the centre of 12 tons 10 cwt. This makes a total of 37 tons against the safe strain of 37 tons formerly stated, and this last addition appears to me to be the immediate cause of the accident. In these calculations, however, it is assumed that everything is at rest, and that the forces applied are those resulting from direct pressure, while the evidence shows that there is a vibratory movement of the whole structure to a large extent, and there is besides a percussive movement of the engine and tender, which, with a heavy long boiler engine with outside cylinders, is considerable. The weight of the structure and of the train in motion will be about 164 tons in all, and the strain from this cause must be added to that formerly stated. This strain, although it cannot be ascertained by accuracy of calculation founded on experiment, experience shows to be great, and I am of opinion that it formed a large element in the strain which broke the bridge down. There is also the whole gross strain arising from the pressure and the percussion of the structure and its load, the apportionment of that strain between the girders ; for I am of opinion that from the loose and independent connexion of the girders, and the giving of the structure, the strain may have been unequally divided between the girders. These investigations, independently of the evidence of the eye witnesses, lead me to the conclusion that the girder broke in the middle from its weakness to resist the strain increased by the laying on of the ballast.
"The opinions of Mr. Stephenson and Mr. Locke, founded on the alleged facts as to the paint on the tender, the broken carriage wheel, and the snips in the chairs, appear to fall to the ground, as they must have been misinformed on those particulars, which can all be disproved."

A statement was also made by Capt Simmons, R.E.

After returning their verdict on the cause of the individual deaths, the jury added the following:-

'We are further unanimously of opinion that the aforesaid girder did not break from any lateral blow of the engine, tender, carriages, or vans, or from any fault or defect in the masonry of the piers or abutments, but from its being made of a strength insufficient to bear the pressure of quick trains passing over it.

"We feel that the eleven remaining girders having been cast from the same pattern and of the same strength, are equally weak, and consequently equally dangerous for quick passenger trains as was the broken one.

"We consider we should not be doing our duty towards the public if we separated without expressing our unanimous opinion that no girder bridge of so brittle and treacherous a metal as cast iron alone, even though trussed with cast iron rods, is safe for quick or passenger trains. And we have it in evidence before us, that there are upwards of one hundred bridges similar in principle and form to the late one over the river Dee, either in use or in the course of being constructed, on various lines of railways. We consider all these unsafe, more or less in proportion to the span, still all unsafe.

"We therefore call upon her Majesty's Government to institute such an inquiry into the merits or demerits of these bridges as shall either condemn the principle, or establish their safety to such a degree that passengers may rest fully satisfied there is no danger, although such bridges may deflect from one and a half to five inches."

'The Coroner stated that that portion which related to the death of the deceased could only be taken as their verdict. Their recommendations, however, he would forward to the railway department of her Majesty's Government, and no doubt the press would give them due publicity. Sir E. Walker trusted they would. They materially affected the security of the public. The inquiry, which has extended over several weeks, was then closed'

The use of relatively slender cast iron girders for cast railway iron bridges was questionable from the outset, although such bridges, with and without trusses, were widely used, albeit with shorter spans. The girders would be subject to additional unquantified strains due to eccentric loading imposed by the deck on the lower flanges. In addition, cast iron girders were vulnerable to residual stresses and hidden casting defects. In the case of the Dee Bridge, the wrought iron tendons not only made no contribution to the strength, but actively increased the stresses, resulting in the factor of safety being marginal.

A Royal Commission was appointed to investigate the application of iron to railway structures, and reported in 1849.

Stephenson was heavily criticized for the design, even before the collapse, particularly for the use of cast iron for such a long span, and the unsatisfactory disposition of the wrought iron trusses.

The Dee Bridge was rebuilt and strenghtened with additional cast iron components, designed by Thomas Gooch. [?]

1849 'THE (CHESTER) DEE RAILWAY VIADUCT. THE CHRONICLE. We stated in our last that the Railway Board had ordered an inspection and report of the Dee Railway Viaduct, on the Chester and Holyhead Railway, the scene, in May last, of a most dreadful catastrophe, now admitted to have been the consequence of insufficient cast iron girders being used for the duty performed. The structure, as our readers are aware, was a three span iron girder bridge, each span being ninety feet. When the trains used to pass over it a deflection was often obtained of five or six inches, and the wonder only was, that a most serious accident had not occurred much sooner, as hardly ever a train passed over it, but the breaking point was obtained within a hair's breadth. It will also be recollected how stoutly Mr. Robert Stephenson fought (against we believe his own conviction) for his favourite iron girder principle ; and we might say much upon what took place at the inquest — but we forbear, mainly from the unhappy state into which oneof his main authorities has since fallen.
However, although tbe iron girders yet remain, the bridge is not now a cast iron girder bridge- That principle, as well as the famous 'lateral blow,' is consigned to 'the tomb of all the Capulets', the bridge is now essentially and actually a wooden bridge, with this drawback, that the wooden framework has to do double duty, is, not only to sustain the traffic, but also the cumbrous cast iron platform roadway. By the principle now adopted each girder of ninety feet is now reduced to three lengths of thirty feet each. But it must be obvious, that as all rest upon tbe wooden frame work, it would bave been more scientific, that the cast iron girders should have been totally removed, and replaced by a substantial wooden roadway. We must however, now deal with the bridge as we find it.
The inspection by Captain Simmons took place on Tuesday last. We have no authorised statements to make, nor shall we have, until Captain Simmons makes his report to the Railway Board. We have however gleaned some facts, which we are enabled to state, and which may be relied on. The bridge was repeatedly tested by heavy luggage trains purposely loaded, and moved by two heavy locomotives. The speed at times was as high as forty miles an hour. The utmost deflection obtained was three-sixteenths of an inch, in the span of thirty feet; and vibration was scarcely perceptible. This result is so far satisfactory as establishing the superiority of the present structure, over the one that it has undoubtedly replaced — that is the bridge, being now wood, and with a divided risk of span. We however, have a strong opinion, and which we shall at all times express, that the cast iron girders should be removed. Instead of adding to the strength of the bridge, they diminish its safety. They are an unnecessary and useless weight upon the amply sufficient frame work, with this additional drawback, are continually liable to fractures from trivial causes. We have it on record that Mr. Robert Stephenson considers that they cannot stand "lateral blows." Not many weeks ago one girder was very seriously fractured with blow of a hammer in driving a bolt; and so convinced are tbe parties connected with the line that the principle can not be relied on, that they have cased and protected the girder in all directions. All that wood can do has been done, but it would be much safer if the girders were removed.
Mr. Robert Stephenson promised at tbe Chester inquest that he would re-construct the bridge so as to satisfy the public mind. There can be no question but that bridge in its present state is infinitely superior to the construction at the time of the accident. The wooden supports are every thing that can be desired in point of strength. If it was an entirely wooden bridge, it would have our unreserved approval. But knowing as we do that an useless duty is imposed upon the wooden structure, that is of sustaining heavy iron girders of no utility as to the strength of bridge, but being instead useless weight, cannot conciously say that Mr. Stephenson has redeemed his pledge.'

1870 'THE VIADUCT BRIDGE OVER THE DEE.— One of the Corporation Committees having suggested to the London and North-Western Railway Company that in reconstructing the bridge over the Dee, forming part of the viaduct which supports the Chester and Holyhead line of rails, and replacing the old wooden fabric by a substantial one of iron, the desirability of making a public footpath along one side of the bridge, the reply of Company is that no alteration, with a view of giving effect to the suggestion, can be made in their plans.'

In 1870-1 a replacement bridge was constructed with wrought iron girders. One source states that the work was undertaken by W. S. Woodall of Dudley. Windmill End Works, Dudley? This was supplemented by another lattice girder bridge constructed immediately alongside between 1899 and 1904. The 1870-1 bridge is now disused. The bridge is now known as the Roodee Viaduct.

See Also

Loading...

Sources of Information

• 'The Britannia & Other Tubular Bridges' by John Rapley, Tempus Publishing, 2003