In the first part of this article the development of concrete was described and the concrete culverts and earthenware pipes on the Pichi Richi Railway discussed. In this part the girder bridges of our railway are described.

The S-Bend Bridge spanned an unsealed road back in 1962 when NM17 hauled derelict NMs through the Pass to Port Augusta for scrapping. (Jack Babbage)

Ancient discovery and development of iron and steel

Iron had come into use at various times in different locations around the ancient world. Its use is known in the Pyramids by the Egyptians around 2000 BC, and in India as early as about 1500 BC. By 700 BC iron working techniques had spread throughout the world into Europe to the west and China to the east. The smelting techniques of that era produced a metal with a low carbon content around 0.1% (wrought iron) which was suited to many uses, but required considerable labour as the iron had to be frequently hammered and also consumed prodigious quantities of fuel in the process. The Chinese, due to superior furnaces, were able to develop an iron with a high carbon content of around 3% (cast iron) which melted at a lower temperature (about 1150°C) than wrought iron (1537°C). Thus it could be readily cast and furthermore was resistant to corrosion.

The rediscovery of cast and wrought iron in the Middle Ages

As Europe emerged from the dark ages the revival of economic activity brought about increased demands for the production of iron. The methods of producing both wrought iron and cast iron demanded large quantities of charcoal, the only fuel then suitable for smelting iron ore. In England the Weald of Kent, which had in early days been heavily forested, was by the sixteenth century dotted with blast furnaces and forges alongside streams, each surrounded by an ever-increasing patch of cleared land. So serious had the shortage of timber for shipbuilding become that in 1558 "the felling of trees to make coals for the burning of iron" was banned.

By 1700 Britain, an island rich in both coal and iron ore, was facing industrial stagnation due to the shortage of iron. The British economy could not grow unless iron could be smelted using coal as a fuel. Using his experience in the brewing industry, which had successfully substituted coke prepared from coal in place of charcoal for drying malt, Abraham Darby was able to develop just such a technique. Darby apparently realised that the hardness and purity of coke would make it suitable for smelting iron. He established his works in Coalbrookdale using a disused blast furnace to produce cheap cast iron pots, and on 4 January 1709 he tapped the first iron of quality to be made using coal. At first news of his success was slow to spread but when it did Coalbrookdale became the centre of a revitalised English iron industry. The quantities of cast iron now being produced made possible the subsequent development of at first a mechanical source of power (the steam engine) and later the use of iron for improved civil engineering works.

Cast iron is said to have first been used by Smeaton for floor beams in a factory in 1755. Smeaton had undertaken tests to determine the suitability of the new material for structural purposes. The first large scale use of cast iron in civil engineering was the Iron Bridge built in 1779 at Coalbrookdale over the River Severn. It has a main arch span of 30 m and was cast over a six month period in the yard of Abraham Darby by his grandson Abraham Darby III to the design of T. F. Pritchard. It was assembled on site in 3 months and opened to the public on New Year's Day 1781. Still standing, it is the first of many iron bridges around the world.

Whilst cast iron is cheap to produce and easy to mould, its brittle behaviour is a serious drawback to its widespread use. The more ductile properties of wrought iron were needed but it was not until 1783 and 1784 that Henry Cort took out two patents covering the production of wrought iron from cast iron by puddling and then its subsequent working by rolling between grooved rolls. The puddling process was the reinvention of an old Chinese method of the first century AD. Pigs of cast iron are placed in a furnace, heated and stirred so that oxygen from the atmosphere combines with the carbon in the cast iron. As the carbon content of the iron is reduced it becomes a spongy mass known as bloom iron and contains considerable quantitites of slag from the furnace. Bloom iron has a much lower carbon content and greater ductility than cast iron. By hammering it much of the slag is expelled to form wrought iron. Cort's second invention replaced the hammering. By passing the still hot bloom through a succession of rolls each slightly smaller than the previous, a length of almost any desired shape could be produced. Cort's rolls made possible the production of 15 tonnes of wrought iron in the 12 hours which it had formerly taken the forging or tilt hammer to produce 1 tonne.

The impetus of the railway boom

The quantity of iron which could be puddled at a time was limited by the need to stir it constantly, hot and hard manual work. Thus only small sizes of plates or lengths of sections could be rolled. This restricted the wider use of wrought iron and the more easily produced cast iron remained popular as a structural material for almost another 100 years. The railway construction boom which was to last for half a century began in Britain during the 1830s. Railways demanded good alignments and this brought a need for improved bridging. As the earlier lines were laid through comparatively easy terrain, only small and medium span bridges were required. These usually had cast iron girders, and were built in great numbers. However the brittle nature of cast iron, particularly under shock loading, soon proved its undoing. Following the failure of a trussed cast iron bridge of 33 m span under a train crossing the bridge over the River Dee near Chester, the use of the material for the girders of railway bridges fell out of favour.

Shortly after the failure of the River Dee bridge, Stephenson was asked to build a railway bridge over the Menai Straits. As the two main spans of 140 m were much greater than any other British railway bridge of that era, a new approach was required. At that time Stephenson was engaged in a series of experiments with Prof. Hodgkinson and William Fairbairn, a ship builder, to determine the properties of wrought iron beams. The restrictions imposed by the small size of rolled wrought iron plates and sections then available had been overcome in shipbuilding and boilermaking by the development of rivetting. In this process, the plates or sections to be joined are lapped and holes drilled right through. A pin or rivet of a slightly smaller diameter than the hole, heated to a cherry-red colour, is inserted and the ends then hammered over. The rivet expands in the hole due to the hammering and as it cools, clamps the two sections together to make a rigid joint and fuses the individual plates into one effective whole.

Stephenson's experiments established the criteria to be used in fabricating wrought iron beams and led to the development of the I-girder and the box girder as the most efficient sections. Stephenson chose a rectangular tubular box girder for the Britannia Bridge which was completed in March 1850. Subsequently many wrought iron girder bridges, fabricated by assembling small plates and rivetting them together, were built. Between 1853 and 1857, a bridge which was to become a pattern for the next 25 years was built by T. W. Kennard over the Ebbw valley in Wales. Known as the Crumlin Viaduct, it comprised 10 spans of wrought iron lattice girders (Warren trusses), pin jointed, which were supported on trestle piers of hollow cast iron columns filled with concrete and braced together. To give stability, the columns of the tall piers spread outwards towards the base. Appropriate materials had been chosen: wrought iron where tension or shock loadings could be expected and the brittle cast iron only where it was subjected to compressive loads. So successful was the design that the pattern became almost standard for the next thirty years. This then was the state of the art of bridge building when the Pichi Richi bridges were designed and built.

The Australian iron and steel industry

The first Australian iron ore deposits to be exploited are believed to have been those discovered at Mittagong NSW in 1833. They were not mined for fifteen years until in 1848 the Fitzroy Iron Works established a smelter. The smelter flourished for a while but then foundered. Activity started again in 1859, when iron was smelted for the piers of two bridges over the Murrumbidgee at Gundagai and over the Macquarie at Bathurst. This is possibly the first use of Australian produced iron in an Australian bridge. However the fledgling industry foundered again in 1884 as high labour costs made it cheaper to import iron from Great Britain. In that year not only the Mittagong Ironworks closed but also works at Lal Lal in Victoria and at Tamar in Tasmania resulting in a tenfold drop in production of pig iron. In South Australia a proposal to smelt iron ore at Cox's Creek was made in 1849, but the venture was not pursued. Again in 1851 Mr Neales moved in the Legislative Council, that a reward of £500 be paid to the person who before July of the next year produced iron from local ore as good as imported iron. The first smelting in this state was done by James Martin of Gawler who smelted some ore from deposits in the Barossa district in the late 1870s. The venture was not commercially viable as coal had to be brought long distances, making the cost of production too high. Some of the iron from this batch, which was the first to be smelted in the state, forms part of the railing in front of the Gawler Institute.

In 1875 J. Rutherford and D. Williams began processing iron ore at Lithgow NSW, and on 25 April 1900 William Sandford produced Australia's first open hearth steel there. However volume production of steel in Australia did not start until 1915, when on 8 March the Broken Hill Proprietary Co. Ltd (BHP) blew in a 350 tonne blast furnace at Port Kembla, NSW. Thus at the time of the construction of the Pichi Richi bridges, practically all iron was supplied from Britain.

Construction of the Pichi Richi girder bridges

The late 1870s were years of rapid expansion of the railway network in South Australia. New lines were pushed out in all directions. The most ambitious of these was the Great Northern Railway being built out from Port Augusta. Standardised designs were used for the bridges on this line to reduce the amount of design work and to facilitate fabrication and construction. The designs comprised rivetted wrought iron plate girders of 20, 40 or 60 ft spans supported on piers of braced cast iron columns of similar design to the Crumlin Viaduct. Alternative 60 ft span lattice truss girders were also designed but only used at one location (Lattice Girder Bridge). The plans allowed for either a square bridge or a skew of 22½ degrees to either left or right. To minimise costs, the bridges were designed for a very light loading, a fact which caused considerable inconvenience and expense during their service life. Subsequent lines, including the Peterborough to Quorn line, were built to a heavier loading.

The Lattice Girder Bridge design

The wrought iron girders were rivetted in twenty foot lengths by their English manufacturers to save as much labour as possible whilst still allowing relatively easy transport after their arrival in the colony. Upon arrival at the bridge site the lengths were rivetted together into complete girders. The cast iron columns were cast in ten foot lengths with flanges at each end to connect the next length, and lugs to connect the cross bracing.

Bridges on the Quorn – Stirling North section have masonry abutments, whilst those on the Quorn – Hawker line near Quorn have concrete abutments. The stone for the masonry abutments was obtained from two quarries alongside the track, one located just east of the "dipper" road crossing near Saltia and the other on the eastern side of the main road next to curve 29. This latter quarry was a poor one, the stone coming out in "bad shape", and no other convenient quarry, including a trial one, one and a half miles south west of Quorn, was found. Therefore the Resident Engineer for the construction of the line, Mr J.W. James, authorised the use of lime concrete for abutments less than 12 feet in height on the line north of Quorn.

The ends of the main girders at the abutments rest on cast iron bed plates which are fixed to large bed stones by "Lewis Bolts". These have a conical shaft which is set into holes drilled in the bed stone, and a protruding threaded shaft for a nut to hold down the bed plate. The taper of the conical shaft prevents pulling out of the bolt from the bed stone. One intriguing detail to emerge from inspections of several girder bridges was the use of sulphur to grout the Lewis bolts!

Detail of screw piles

The design concept of the bridges was simple and intended to give rapid construction times. The pier columns had a cast iron screw fitted to the bottom section and were installed by turning the shaft with a capstan while at the same time loading the top of the column. As the screw turned, the shaft was pulled into the ground. To ensure that the screw pile was kept in its correct position, a heavy guiding frame was used to locate the column whilst it was turned. The designer provided a sufficient length of pile to ensure that embedment was obtained to carry the weight from the bridge above. Once the pile had been screwed into the correct position and depth, the remaining ten foot sections of the column were bolted above and the cross bracing installed. The columns were hollow and the internal space was filled with concrete. To complete the piers, the tops of the columns were connected by crossheads to provide a support for the main girders. Meanwhile the girder sections were transported to site and rivetted together into the individual girders. They were then lifted to the top of the completed piers and bolted into position. To stablilise the girders, longitudinal and cross bracing was installed. With the addition of the sleepers and the running rails, the bridge could be opened to traffic. So straightforward did this giant meccano set seem that the Superintendent of Works, Henry Parker, who had supervised almost the whole of the erection of Murray Bridge, considered that the two large bridges near Woolshed Flat, the Woolshed Flat Bridge and the Lattice Girder Bridge could have been erected in two weeks "had the material been on the ground and alterations required been made".

Elevation of the pier of the Saltia bridge as built
Detail of the piers for the Lattice Girder Bridge

However there were many difficulties encountered in practice. The construction site was at the end of a materials supply line which stretched six months and 12,000 miles to England. It was located in a remote part of the colony with difficult access and no local supply of labour. Transport from the advancing railhead to the site was by dray or cart over rudimentary roads. The only source of power on site was human or animal and the heavy weights of the girders and piles had to be hauled up into position by stiffleg or derrick and block and tackle. Considerable advance planning and organisation was required to ensure that the multitude of required materials in the correct quantities, and the equipment and labour were all assembled and available on site when needed. Not unnaturally there were occasional shortages which would cause considerable delay unless ingenuity was used to overcome the problem.

On 17 February 1879, the Resident Engineer, Mr James, informed the Engineer-in-Chief, Mr H.C. Mais, that the contractors had received eight lengths too few of the column sections needed to complete the piers of the Woolshed Flat Bridge. So that construction would not be delayed, Barry Brooks & Fraser, the contractors, substituted other pile lengths and adapted them by special castings to provide a flange at each end and by fitting extra lugs for the cross bracing. However some concern had also been felt about the expertise available on site, and Henry Parker was sent to the contractor's camp at Woolshed Flat, arriving there on Friday 7 February. He was told by the Foreman that the Lattice Girder Bridge would not be finished for five weeks and even then only if good hands were brought in. Parker was critical of the skill and equipment of the contractor and immediately took over the direction of operations. Within a week all the girders of the Lattice Girder Bridge had been fixed and three of the four piers of the Woolshed Flat Bridge erected. By the time Parker left on 24 February, the Lattice Girder Bridge was almost ready for traffic and in fact the first engine crossed four days later on the 28th. The site inspector, J.J. Frisk, wrote to Parker on 3 March informing him that a deflection of 3/8 inch had been measured in each span with no permanent set, a result considered very satisfactory. Completion of the Woolshed Flat Bridge was being delayed by the shortage of pier columns. Three piers had been completed and three spans landed. No more could be done until the last pier could be completed using the pile lengths adapted for the purpose. On 4 May the line was completed to Summit and on 6 June to Quorn.

The subsequent demise of wrought and cast iron and the rise of steel

Late in the railway construction era the Firth of Tay in Scotland was bridged. Sir Thomas Bouch's design adopted the concept of the Crumlin Viaduct of 1853 but not the same care in design and construction—wrought iron lattice girders supported on trestle piers of cast iron columns. The thirteen central navigation spans were at a high level where the railway ran between the trusses whilst the approach spans had the trusses underneath the track. The bridge was completed in May 1879. During a high wind in December 1879, the central thirteen spans collapsed under a train with the loss of 75 lives. The loss stunned Britain, which had grown complacent about the extensive engineering achievements of the time. The immediate effect of the disaster was to halt the construction of the next large railway bridge, the Firth of Forth, and in the long term it led to the demise of cast iron and wrought iron as structural materials and hastened the introduction of steel.

In the 1850s William Kelly of Kentucky USA and Thomas Bessemer in the UK independently developed a steel-making process. However many problems remained before good reliable steel could be commercially produced. Bessemer had experimented with a low-phosphorus ore and once producers started using the more common ores with higher phosphorus contents, the steel produced was inferior to wrought iron! In the early 1860s the Dutch built several railway bridges of Bessemer steel, but the results were not satisfactory and a distrust of Bessemer's steel became general. A solution to the use of high-phosphorus ores was not found until 1879 when Sidney Gilchrist experimented with a dolomite lining in the blast furnace. Only then did the use of steel become widespread.

The first large steel bridge to be constructed in the United States was Capt. Eads' St Louis bridge built across the Mississipi between 1867 and 1874. The British however were much slower to use the new material for bridge building. Whilst shipbuilders and boilermakers rapidly adopted steel, the first British steel bridge was not built until 1883—a railway bridge on the Chester-Holyhead line. The first large British steel bridge was the Forth Bridge near Edinburgh, constructed between 1882 and 1890. In the three decades between 1870 and 1900, steel almost completely replaced wrought iron as a structural material. Larger quantities of steel could be produced from one heat, which could be rolled by more powerful mills into larger girders.

In 1880 the South Australian Railways decided to try rolled joist girders made from Belgium Iron, which was in fact steel. Tests showed that the new material was about 50% stronger than wrought iron. Girders were ordered for the Jamestown to Yongala railway and the size specified was 12" x 6" x 51.2 lb per foot (300 mm x 150 mm x 76 kg/m). However, when weighed, the girders were found to be about 60% heavier than specified, a fortunate circumstance, as it had been decided after ordering to run heavier locomotives on the line and it would have been necessary to strengthen the specified girders! The heavier girders supplied were adequate without strengthening.

Subsequent history of the Pichi Richi bridges

NT66 on is delivery run from Sydney, via Broken Hill, Peterborough, Quorn and Pichi Richi Pass, to Port Augusta in 1965. It is pictured on French's Bridge, with the timber strengthening clearly visible. (Jack Babbage)

As noted earlier, the bridges on the Port Augusta and Government Gums line were designed for a very light loading. As early as July 1880, only one year after the line was opened, the bridges were checked for their capacity to carry K class locos, by the Engineer-in-Chief, Mr H.C. Mais, and found to have insufficient strength. However, the bridges on the Terowie to Quorn line were found to be strong enough and indeed appear to have had little strengthening added during their lifetime, unlike the Pichi Richi bridges which required strengthening at two separate stages, in 1895 and again in 1924 when heavier locomotives were introduced. Additionally, the bridges were checked in 1908 for the introduction of the Yx class and in 1951 for the NSU class diesels. The load capacity of these bridges would have been one limit (besides the lack of traffic and the poor track alignment) on the size of engine used in the Pass.

The 1895 work involved rivetting extra plates to strengthen the girders. On single span bridges these were added over the midspan section where the stress is greatest, whilst on multispan bridges the plates were added over the piers to make the girders continuous. These latter plates were cut through with a cold chisel to make individual spans again in 1924 when the bridges were restrengthened. The spans of the girders were reduced in most cases with timber props such as exist at the S-Bend Bridge. The spans of the Woolshed Flat Bridge were reduced in length by moving the support from the ends of the girders to the outside columns of the piers. The Lattice Girder Bridge was strengthened by adding a complex system of timber trestles and steel undergirders. The 20 ft girder bridge had its girders replaced by new steel girders in 1924, a fate shared for different reasons half a century later by the twin span bridge at Saltia.

The elegant lines of the Lattice Girder Bridge as designed (below) had been somewhat marred by strengthening when the ARHS operated the 'Pichi Richi Pathfinder' over it on 14 October 1962. (Jack Babbage)

Cattle crossings

Cattle crossings were provided in the original plans. It appears that originally double-headed rail was used to span between the abutments, whilst later cattle crossings used rolled steel girder sections, which had come into general use at about the time of the First World War. Quite a few of these structures appear to have been removed in the past when the need for them was no longer perceived.

References

Hopkins, H.J., "A Span of Bridges", Praeger Publishers, New York, 1970.

Patterson, R.C., "On the best methods of Railway Construction for the development of New Countries, as illustrated by the Railway Systems of South Australia", minutes of proceedings of the Institution of Civil Engineers, London, paper presented January 14, 1879.

South Australian Government, Commissioner of Public Works, docket nos: 376/1879, 515/1879, 612/1879, 677/1879, 1190/1879, 727/1882.


This article by Bill Stacy was published in Pichi Richi Patter Volume 12 No. 3, Autumn 1985 and Volume 12 No. 4, Winter 1985.