In the early summer of 1837, after Brunel's skew bridge at Hanwell was complete but before the railway proper was open, one of the main girders failed. The girder weighed 8.5 tons and was supporting a dead weight of 50 tons of bridge, plus the weight of any trains passing over. Fortunately the bridge did not collapse; the grid of girders was fixed together using mortise and tenon joints and these, together with some tie-bars, and the substantial timbers which supported the rails held the bridge deck together. The broken beam was quickly shored up with timber, and the bridge reopened. On 8 June, Brunel wrote to Grissell & Peto, the contractors who had built the bridge, telling them ‘No time must be lost ... to remedy this serious defect’. It was a very serious defect, and there were two possible causes: either the girder was faulty, or it had been incorrectly designed—and the designer was I K Brunel. Subsequent events showed that the casting certainly was faulty; whether the design was also marginal is a moot point.
The Commissioner of Roads, Sir James McAdam, with whom Brunel had probably negotiated the significant concessions on the bridge specifications, had become alarmed by the failure and retained a consultant to investigate. Sir Robert Smirke had extensive experience with cast iron beams, having introduced their use into domestic architecture. His younger brother, Sidney Smirke, took over from him when he retired, and designed the famous circular readers-room at the British Museum. Robert Smirke visited the Hanwell bridge in the company of McAdam, and on 20 November 1837 wrote to the commissioners saying that in his opinion the design of the bridge was perfectly good ‘if sufficient means have been taken to prove the soundness of all the castings’. He had spoken to ‘Mr Brunel’ who told him that before the timber framing supporting the ironwork was removed (having been in place for five months at this point), this would be done. The entire bridge, including the new girder, would be proved by Passing over it carriages loaded with a much greater weight than will ever be required at other times, and as the work is now completed, I do not know how any other or better proof can be obtained. Smirke’s words suggest that he was not entirely happy with the situation. Evidently Brunel managed to satisfy him that the bridge was safe, and the Great Western Railway opened for commercial traffic from Paddington to Maidenhead on Monday 4 June 1838. All was well for nine months, but in 1839, the West Kent Guardian reported that on the morning of 18 March just as a train was passing over the bridge ‘passengers were thrown into the utmost consternation by hearing a report resembling that of a heavy cannon...’ The main girder on the other side had broken in half. This time it was a passenger-carrying train that had been passing over the bridge. A tragedy was averted by the interconnected structure of the bridge and again it did not collapse, but McAdam must have been furious given the previous tests and assurances, and Brunel would have been very seriously concerned. Wasting no time, the company had a gang at the bridge that evening to shore up the broken girder. To compound Brunel’s misery, Daniel Wood, a gravel-digger who appears to have volunteered as a helper to the gang, was killed when a two hundredweight baulk of timber being lifted into place fell ten feet on top of him; the inquest into his death was widely reported in the newspapers. Two days after the incident Brunel wrote to Grissell and Peto: The fracture is ... in the corresponding girder and nearly at the same place as in the former case and apparently from exactly the same cause. I have had a piece drilled off and it consists of a mass of half sand half iron. That casting too had been faulty; eight-and-a-half tons of molten iron contains a considerable amount of heat, and part of the mould, which was made of sand, must have collapsed into the liquid iron under the thermal shock leaving no outward sign. There was a flurry of correspondence. A Mr Bramah, probably the subcontractor to Grissell and Peto who had cast the girder, had written to Brunel expressing doubts regarding its design. Brunel ‘returned’ the letter to Grissell and Peto writing: Had Mr Bramah or you expressed any doubt previous to the commencement of the work and had you declined being responsible it might be right for you to repeat the grounds of your fears—but the result of the accidents that have happened have proved that there is no cause for alarm. That the unsound girder should have stood what it has done is the strongest proof possible that a sound one of the same dimensions would safely stand anything that comes upon it. Here is Isambard keeping his head, making a very good point regarding the design, and clarifying contractual (and moral) responsibilities should it come to an argument as to ‘who pays’. He, of course, had far more to lose than the cost of replacing a girder. The bridge failure was public knowledge, and it was vitally important to establish that although the casting was faulty, the design was not. A new girder was duly cast and proved at a different foundry, and evidently transported to Hanwell with great despatch. Thus on 18 September Brunel wrote to Sir James McAdam: It was quite accidental that the [new] girder was proved and carried down to Hanwell without informing you. The obtaining the casting had occupied so much time that everybody was anxious to lose no more ... the casting has all the appearance of being an excellent one, it has been carefully proved, we have no means of proving it at Hanwell and if this be required it must be transported [back] to Limehouse. But I can furnish you with certificates of the proof to which it was subjected, signed by highly responsible parties, the contractors Grissell & Peto, the foundry, Messrs Seawards and my assistant and this I trust you will consider sufficient as proving the fact Reading between the lines, it is clear that Sir James had wanted an independent inspection of the proving process before the new girder was committed to the bridge. McAdam decided to get his consultant to press Brunel on the matter, and Smirke demanded further proof of the soundness of the bridge. On 4 November Brunel wrote to him: The new girder has been fitted into its place at the Metropolitan Road bridge at Hanwell and is now ready for fixing. In consequence of your letter of 4th Ult I have considered what method could be devised for increasing the security of the work and propose for that purpose to lessen the weight to which the girder is subjected very considerably and to prevent any vibration from the passing of the trains being communicated to the cast iron by the following arrangement. 1st To remove the brick parapet walls which now rest on the main girder and which together weigh about 30 tons and to substitute cast iron plates weighing only about 6.5 tons and of sufficient strength to carry themselves as well as hold up the girder in the event of its breaking. 2nd To remove the small brick arches now forming the floor of the bridge and which with the concrete constitutes a load of about 70 tons now resting on the two main girders and to substitute a timber flooring weighing only 16 tons and upon which the rail timbers would be laid so that any vibration upon the rails could not be transmitted to the girders. The total reduction in dead weight will then be 78 tons 10 cwt or 39 tons 5 cwt upon each of the main girders having consequently a very great deal of excess strength while at the same time from the introduction of the timber and the cast iron parapet the girders will be exposed to less injury and would be supported in case of failure of such could then be considered possible. I trust you will approve of these proposals and that you will have the goodness to communicate such approval to Sir James Mc Adam as early as possible. I am under promise to him not to proceed with the fixing of the girder. The brick arches and their cement infill contributed 70 tons of deadweight to the two main bridge girders. In his letter to the contractors Brunel said that the new girder (the second replacement) had been proved to ‘78 tons, 11 tons more than his maximum.’ Thus the maximum design load per girder was 67 tons. The function of the brick arches was to provide a flat and continuous bridge deck for the rails as well as a walkway for maintenance staff; was it really good design practice for this relatively minor feature of the bridge to consume almost half of its load-carrying capacity? Furthermore when pressed, Brunel cheerfully replaced it with wood saving a weight of 54 tons! Likewise, a straightforward alteration to the parapets saved more than 20 tons. In his letter to Smirke of 4 November, Brunel had said that the new girder ‘has been fitted into place ... and is ready for fixing’. Exactly what this meant was clarified by a letter to Sir James McAdam dated 21 November, and judging by Brunel’s words, Sir James was far from happy. Brunel began his letter: My Dear Sir James, I have received yours of the 18th. I am sorry it does not show that courtesy and confidence which you have uniformly displayed towards me in former correspondence and which I feel I deserve at your hands. I regret that before assuming that I was about to do something ‘in violation’ of any understanding or request, you had not applied to me to ascertain if such were the case as you would at once have been satisfied that it was not so. There is more to be done to the girder before fixing than can now be completed by Sir R Smirke’s return; it has to be raised to its [final position] and marked then lowered or removed to be chipped and filed before it is ready to be fixed and this alone I am about to proceed with. That no further delay that can possibly be avoided shall take place when it is definitely determined what is to be done with it up to the present time the girder remains in the lane close by. This is not quite what he had said to Smirke nearly three weeks earlier. Then he said that the girder had been ‘fitted into its place’; now he appeared to be saying that that still needed to be done. He went on: I should add that your surveyor has informed my assistant that they shall not move the girder from where it is, that he should prevent it by constables. That is going rather far and not only threatening that which he has no authority to enforce but adopting a course which if he had the authority is totally uncalled for. It seems that there had been an unpleasant confrontation at Hanwell, and it is not difficult to deduce what had happened. A few days after the failure of the first girder Smirke and McAdam had inspected the bridge, and either by then or shortly afterwards, the faulty girder had been replaced before either of them could witness the proving. Subsequently Smirke had had a meeting with Brunel and was presented with a fait accompli in that only by accepting Brunel’s proposed method—using heavily loaded waggons—could the girder, by default, be proven to be good. McAdam was obviously suspicious that Brunel once again was trying to out-manoeuvre him. Brunel continued: You have requested that the girder shall not be fixed until Sir Robert Smirke is satisfied of its soundness or shall have arranged with me on the subject. I have no intention of neglecting that request but waiting Sir R Smirke’s return I am going on with all the preparation required and I informed Mr Brouse [McAdam’s surveyor?] that such was my intention He finished by appearing to invite Sir James to withdraw his letter, otherwise Brunel would have to make a formal response ‘entered upon the official correspondence of the office’—was this letter then not a formal response? Evidently ruffled feathers on both sides were eventually smoothed over, the new girder was fitted, the brick arches removed, the other alterations made, and the bridge was opened for business. But the law of unexpected consequences reigned supreme; in substantially lightening the bridge, Brunel had opened it up to a new and fairly obvious danger that was eventually to spell disaster, and not just at Hanwell.
0 Comments
Leave a Reply. |
AuthorWelcome to the Mirli Books blog written by Peter Maggs Archives
March 2024
Categories |