The arch becomes an option.

From Engineering Heritage New South Wales

    The bridge across Sydney Harbour was almost always going to have a very long span, though some earlier designs included piers in the waterway to keep the main span within what engineering could safely build. These piers would be problematic for shipping, and the bridge may not survive in case of a collision. It would also be difficult to build as the depth to rock would be too great for the construction techniques available in the 1920s.
    The cantilever design which had become the accepted option was eminently buildable. The Firth of Forth railway bridge had been built in Scotland a quarter of a century earlier and it had two spans longer than the Sydney Harbour Bridge would require. The Quebec Bridge, opened in 1917, was even longer, though it had collapsed twice during construction - in 1907 and 1916. Granted that any form of falsework to support any bridge was inconceivable because of the depth of water and the need to keep the harbour open for shipping, the technique of building out from both shores until the halves met in the middle or got close enough to allow a suspended span to be floated out and lifted into position, was attractive.
    The drawback of the cantilever was that much of its steel structure, the balancing back spans, and therefore cost, was over the foreshores which didn’t need a bridge at all. An arch on the other hand placed all of its structure over the water which was the intent of the whole project.
    Arch bridges had been built in years preceding Sydney by cantilevering out from both shores, with the half arches held back by cables secured to rock anchorages – but the Sydney Harbour span would be much longer, wider and heavier.     

The Zambesi River Bridge at Victoria Falls, about 1905. The half arches are tied back as cantilevers, but the high walls of the deep gorge mean that the cables are not so readily noticed.

   The Zambesi River had been crossed at Victoria Falls in 1905 using this technique. Ralph Freeman had contributed as an employee of the consulting engineers Sir Douglas Fox and Partners, and was later to have the same role in Sydney. The Zambesi bridge had been fabricated by the Cleveland Bridge Company of Middlesborough, England and their resident engineer was Georges Camille Imbault who was later also Freeman’s associate at Sydney. While these people knew how to do it, the Zambesi bridge had a span of just 513 feet, less than one third of Sydney, and just two railway tracks, so the force in its tie-back cables was one fiftieth of that needed in Sydney.

Hell Gate Bridge under construction 30 September 1915. The half arches are held back by tension members passing over high towers. A method like this was envisaged for Sydney.

    The East River in New York had been crossed at Hell Gate in 1916 by an arch bridge using the tie-back method, and looks very much like Sydney The bridge carried four railway tracks and spanned 978 feet. Certainly, a step up from Victoria Falls, but still a lot less bridge than Sydney. Could the work be scaled up, reliably?

The cable anchorage scheme as first envisaged for Sydney. The derivation from Hell Gate is obvious. This is a very late drawing as the RL of the main pins is shown as 27.5 feet, reduced from 40 feet in the tender documents. The half arches would have been lowered to meet by shortening the post rather than lengthening the cables. The difficulty of designing the joints at the panel points where the ties met the arch top chords eventually led Freeman to abandon this concept. Freeman and Ennis 1934 ICE Journal

    The firm for which Bradfield possibly held out most hope in his 1922 trip was the Cleveland Bridge Company. Imbault was their chief engineer. Sir Douglas Fox and Partners were retained by them as consultants, and Ralph Freeman was leading the work. The Chairman of the firm, C F Dixon, was enthusiastic. Dorman Long and Company, the eventual bridge constructors were not really in the picture as they were largely a steel production and fabrication business, not contractors for the whole job. Bradfield did not even meet with them in his 1922 visit to England.

    Dixon, and other tenderers, told Bradfield that they could build an arch of the required span, and so he went back to Australia and changed the tender specifications to include the arch.

    Sadly, in September 1923 C F Dixon died suddenly, and the Cleveland Bridge Company decided that they could not continue despite the design and the costing being in an advanced state. They came to an arrangement for Dorman Long and Company to take over the tender and the personnel – Freeman, Imbault and others. Dorman Long were well experienced in steel manufacture and heavy fabrication and were probably on the point of moving into full construction anyway, and in February 1924 they were announced as the successful tenderers.

    G C Imbault was in Sydney as one of their representatives at that time, though he seems to have not remained with the project. Freeman credited him as having prepared the erection technique on which the tender was based, though it was later evolved by Freeman.

    The other bridge which is often cited as a precursor to the Sydney Harbour Bridge is that over the Tyne River in Newcastle, UK. This much smaller bridge was not designed by the same engineers, has significant differences to Sydney's bridge, was started later but finished earlier because of its smaller size, and so cannot be seen as a precursor. It was however built by Dorman Long and Company using similar techniques so was no doubt a valuable test bed for some construction details.

The Newcastle on Tyne Bridge has both chords meeting at the bearing so is a crescent shaped arch, very unlike Sydney. The tie-back technique using a tall post is also different.
Cookies help us deliver our services. By using our services, you agree to our use of cookies.