The Hawkesbury Bridge, New South Wales.” by CHARLESORMSBY BURGE,M

2 BURGE ON THE HAWKESBURY BRIDGE, N.S.W. minutes Of

Associate iMernbers-continued. HERBERTSPON~ HAWKINS. HAROLDWILLIAM PAULING, Stud. Inst. JADlES ALLANHENDERSON. C.E. PATRICKNICHOLAS HILL JOXES, Stud. ALFRED CECIL PEREIRA,Stud. Inst. Inst. C.E. C.E. FREDERICKJORGEN~EN. CHARLESROBERTS, F.C.H. JOHSKNIGHT. ARTHUR SAXUELFRANCIS ROBINSOS. TOMGALLON LUMB. CHARLESTALLENT SPENCER, Stud. Inst. WILLIAWMCKELVIE. C.E. HENRY ALLANNoss. ROBERTANTHONP SYNGE, B.E.. JOSEPHMOYNAN, B.C.E. ALEXANDERTHOPSON WALKER. JOHNJAMES MYRES. HORACEREGINALD WATERS, B.E. ADOLPHEERNEST ORR.

(Paper No. 2421.) The Hawkesbury Bridge, New South Wales.” By CHARLESORMSBY BURGE,M. Inst. C.E. THErailway systemof New South Wales until recentlyconsisted of two separate divisions,the one starting from Sydney, and branching in a westerly, southerly, and south-westerly direction, while the otheroriginated at Newcastleon the sea-coast, about 100 miles north of Sydney, and communicated with the northern part of the Colony, and with Queensland. The Hawkesbury Bridge is situatedupon the connecting link of railway which was designed to unite these separate divisions. At the site of the bridge, about 7 miles from the sea, the estuary of theHawkesbury has a totalwidth of about 6,600 feet, and is dividedinto two channels by Long Island (Plate 1, Fig. l). A solid embankmentcarries the railway across thecomparatively shallow southern channel, on leaving which the line is tunnelled through a high promontory forming the eastern end of Long Island (Plate l, Fig. 2), andthen crosses the mainchannel to Mullet Point by the bridge which is the subjectof this Paper. The Colonial Government, after a considerable amount of dis- cussion, decided to invite notonly tendersfor the construction of the work, but also designs for the same, and took the necessary steps t,o make the competition world-wide. A committee, consisting of Xr. W. H. Barlow, PastPresident Inst. O.E., &h.G. Berkley, Vice PresidentInst. C.E., andCaptain DouglasGalton, R.E., .lssoc. Inst. C.E., was appointed to adjlldicate on the tenders; and when the report of this committee hadbeen supplemented by others from Sir John Fowler,Past President Inst. C.E., and Mr. John

Downloaded by [ UNIVERSITY OF CAMBRIDGE] on [23/09/16]. Copyright © ICE Publishing, all rights reserved. Proceedings.] BURGE ON THE HAWKESBURYBRIDGE, N.S.W. 3 Whitton, M. Inst. C.E., Engineer-in-Chief for Railways, New South Wales, it was finally decided to accept the tender of the Union Bridge Company, of New York, who undertook to complete the bridge within two-and-a-half years for the sum of S327,OOO. The accepted design consisted of seven spans of 416 feet each, from centre to centre of the piers, the foundations for the latter being of concrete encased in steel caissons, while the upper portions of the piers and the whole of the abutments, are of masonry. The girders are formed of built steel compression-members, and solid steel eye-bar tension-rods, all the connections being made by steel pins. The cross-girders and rail-bearers are of riveted steel plate. The two main girders of each span are 410 feet 03 inch long, from end pin to end pin, and 58 feet deep at centre, and are placed 28 feet apart, from centre to centre, the bridge carrying two lines of railway. The general elevation of the bridge is shown in Plate l, Fig. 3.

LOCATION. The line and spans of the bridge were located by a system of triangulation carried out by Mr. F. Small, and continued by the Author who succeeded him as representing the Government, and by Mr. 0. Schulze on thepart of the contractors. The results obtainedfrom the firstbase-line on the shore wereafterwards supplemented and checked byusing a portion of thestructure itself, when fixed, as a secondary base-line.

FOUNDATIONS. The borings showed a bed of mud extending to a depth varying from 60 to 170 feet below high-water mark,and overlying the sand ; the greatest depth of water being 77 feet, and the range of tide 7 feet. Thegreatest depth of the foundations occurs at pier No. 6, which is carried down 162 feet below water, this being, as far as the Author isaware, the deepest bridge-foundation yet sunk. The caisson foreach pier isrectangular, with rounded ends, and its construction is shown inPlate l, Figs. 4 and 4a. It is 48 feet long transversely tothe bridge, and 20 feet wide,splaying out in the lowest 20 feet,, so as to form a tapered shoe which is 2 feet wider all round at the bottom. In the centre line on plan, parallel withits length, are three wrought-iron dredging-tubes, 8 feet in diameter, and 14 feet apart from centre to centre. These are connected to the outerskin and to each other by steel strutting of tees and angles. B2 Downloaded by [ UNIVERSITY OF CAMBRIDGE] on [23/09/16]. Copyright © ICE Publishing, all rights reserved. 4 BURGE ON THE HAWKESBURP BRIDGE, N.S.W. [Minutes of

At thebottom, the dredging-wells splay out in a trumpet mouth so as to meet the outer skin, andalso each other, in a strong cutting edge formed of heavy steel plates. The method of sinking the caissons was as follows :-The shoe, having been built on shore at Dangar Island, and provided with a timber false bottom, was floated out to position and sunk to the bottom of the river, by removing the temporary bottom, and par- tially loading the caisson with concrete. The caisson was then sunk through the mud by dredging the material from the bottom of the wells and by loading the space between the wells and the skin with concrete,more steel being built up as the caisson went down. As soon as the structure was firmly in the sand, the dredging- wells were filled with concrete, and the masonry was then begun at a level somewhat below low-water. The concrete was composed of l part of Portland cement, 3 parts of sand, and 6 parts of stone, broken to 2fr-inch gauge. The stone was what is locally known as Kiama blue stone, the material being mixed by Jamieson’s concrete-mixers, each passing through about 5 cubic yards per hour. The concrete in the shoe was made stronger by the addition of one-fifth of a cask of cement per cubic yard. The caisson for No. 5 pier was the first one started, and, having undergone greater vicissitudes in its downward progress than any other, was the last to be completed. The sinking was begun on the 9th of December, 1886, and the foundation was only ready for the masonry on the 9th of October, 1888. Shortly after it had well entered the mud, the caisson showed a tendencyto work eastwards, that is to say,transversely to the direction of the bridge. Efforts to recover its position were first made by endeavouring to cantit eastward at thetop, by excavating the eastern well in advance of the others, thus pointing the central vertical axis downward in the westward direction required; but this was not successful, as even when the eastern well excavation was 15 feet deeper than the western one, the cant was still westwards. Dredging outside was then resorted to without effect, also dumpingthe excavatedmud outside on the easternend. When the caisson had reached 75 feet below the river bed, the divergence had amounted to 5 feet at thebottom and 3 feet at thetop, the cant still continuing westward, notwithstanding constant extra sinking of the eastern wells. The margin for lateral divergence allowed by the specification for the caissons was 2 feet. A recovery of about 18 inches at the bottom was affected about

Downloaded by [ UNIVERSITY OF CAMBRIDGE] on [23/09/16]. Copyright © ICE Publishing, all rights reserved. froceedings.1 BURGE ON THE HAWHESBURY BRIDGE, N.S.W. 5 this time, and when the west cutting edge first felt the indications of sand, the caisson, as had been expected, commenced a righting movement towards the vertical, but of course at the expense of still further lateraldivergence at the top. At this juncture, the 7-feet grab working in the ea,stern well, got caughtin thebottom splay, and as aconsiderable period elapsed before it was recovered, work was suspended at this caisson for a time. Later on, twograbs were similarly caught in the centre well, one of which was never recovered-diving operations, at the depth then attained, being attended with greatdifficulty. The contractorsnow began driving piles, at the eastend, to sustain a cribwork, which was loadedwith stone and was intended to forma buttressagainst the structure and prevent further movement ; and, subsequently, a similar cribwork was putin about 400 feet upstream or westwards, with the intention of anchoring the top of the caissonfirmly to it; andthen, by digging per- sistently in the eastern well, it was hoped that the bottom would be brought back to its position. But as to all these proposals, it was always the Author’s opinion, that, as the purchase, in each case, was in the same unstable medium of mud, any horizontal movement in the caisson could not be corrected by either tying it to, or resting it against, any other structure not founded in the sand. This opinion was verified, as, on further sinking, the top, in righting itself, pushed over the eastern cribwork. Operations, as regards No. 5 caisson, werenow suspended for some months, except for the recovery of the lost grabs, and the construction of the cribwork upstream, as mentioned above. This experiment having failed, a proposal was made to cease sinking, as the caisson was well in the sand at 144 feet below high-water of ordinary spring-tides, and then it was proposed to rectify matters by sinking an additionalcaisson at thewest end. This addition was, in plan, something in the form of a crescent with rounded ends, the concave sideof the crescent being meant tofit the rounded west end of theoriginal caisson. It was made of steel plates. Two wells were provided in the new caisson, one at each extremity, the spacebetween being loaded with concrete, andthe bottom finishing in cutting edges similar to the original one. The ill luck of the pier attended this effort also, for when about 28 feet from the bottom, the wells caved in, under the pressure of the mud on the one sideand of theoriginal caisson on theother, SO that further sinkingbecame impossible. As the additional caisson could not now be got either down or up, there was no alternative left but tocommence the masonry at the

Downloaded by [ UNIVERSITY OF CAMBRIDGE] on [23/09/16]. Copyright © ICE Publishing, all rights reserved. 6 BURGE ON THE HAWKESBURY BRIDGE, N.S.W. CMinutes of west end, at as low a level as possible, viz., 12 feet 6 inches under the original masonry level, and to corbel out. With the aid of a cofferdam thiscorbelling was carefully carried out with solid stones of 7 to S feet in length, with a 9-inch overhang in each course, and, though adopted as a last resource, the centre of the column of masonry above, coinciding with the centre of the west girder, is well within the base of the original caisson; and the resultant line of the pressures of the pier and load passes very closely to the centreof the bottom foundation area. The sub-contractor for the work attributed the eastern divergence to the fact that themud, at theeast end, was more consistent than at the west; hence the material displaced by the splay at the former end did not cave in immediately, but stood for a considerable depth without filling the cavity. The west displacement on thecontrary being immediately refilled bythe softermud falling into it, greater resistancewas felt at that end, and the caisson gradually tended towards the other end, where the chief resistance was only at itsprojecting cutting-edge. This theory is supported by the fact already noted, that per- sistent dredging of the east well in advance was not followed by faster sinkingof that end, which appears to showthat the material at the west end was softer and therefore more easily pierced by the cutting-edge.There is no doubtthat the outward splay of the caisson at thebottom wasthe cause of nearly all thedifficulties connected with the sinking; as incaissons Nos. 1, 2, and 3, where an additional outer vertical steel wall was subsequently adopted, no difficulty in sinking correctly intoplace was experienced. This addition, which was in the form of a pocket, filled with concrete all round the caisson, carried up from the cutting-edge 24 feet high, is indicated by a broken line P on the drawing of the shoe in Plate 1, Figs. 5 and 5a. About t.hree months after the launch of No. 5 caisson, No. 4 was towed out and sunk withmuch less difficulty to a depth of 147 feet below high-water of spring-tides, and with comparatively slight divergence from its correct position. No. 6 was launched about six weeks after No. 4, and gave some trouble. Its site coincided with a sudden declivity in thebottom of the river. The caisson tended northward, following the declivity, and continued to do so, even when the sinking hadprogressed to a great depth, and in spiteof the fact that large quantities of stone were dumped to the bottom, at the north side of the structure. The caisson also took a skew position, though not to such a degree as to interfere with the correct location of the masonry pier upon

Downloaded by [ UNIVERSITY OF CAMBRIDGE] on [23/09/16]. Copyright © ICE Publishing, all rights reserved. Proceedings.1 BURGE ON THE HAWKESBURY BRIDGE, N.S.W. 7 it, at righhngles to the bridge. When the caissonreached the bottom, at 162 feet below high-water of spring-tides, it hada slightlean to the southward at the top, notwithstandingthat its proximityto thenorthern bank of theriver had afforded facilitiesfor counteracting this tendencyby anchoring thetop northwards. Being unable to cope with the general northern movement of caisson No. 6, the contractorshad to get the consent of the Government to increase the span of the girders, between the fifth and sixth piers, by 4 feet 3 inches. As the north abutment had not yet been commenced, its site was shifted4 feet 3 inches north, SO that the width of span No. 7 remains the same as originally intended. The maximum pressure on the base of No. 6 caisson, which is the heaviest, is about 9 tons per square foot. About three months after the launch of No. 6 caisson, No. 1 was moored in place, and, being constructedwith vertical sides without any splay at the bottom, it was sunk without difficulty through the mud, which was here of no great depth. Before the, caisson reached the bottom of the water, however, it met with an accident similarto that which had befallen the caissons of theJubilee Bridge over the Hooghly,' and of the Poughkeepsie Bridge over the Hudson River. Each of the Hawkesbury caissons, in sinking, had been lashed to a 700-ton barque, which was used as a, store, workmen's quarters, &c. ; and on the night of the 18th of August, 1887, the wind and tide becoming too strong for the anchorages, the ship and caisson with the derrickpunts attached, were all carriedaway some distancedown theriver, where, with some difficulty and risk of wreck, they were re-anchored. No attempt totow the structure backwas made till the weathersettled, which was about a week subsequently. No. 1 was at last founded at about 101 feet below high-water of spring-tides. No. 2 waslaunched about two months after No. 1, and,four months later, was founded accuratelyat 155 feet below high-water of spring-tides, being nearly as deepas No. 6. Twomonths later, or at the end of November1887, the re- maining caisson, No. 3, wastowed out, and founded, without incident, on the 16th of February, 1888, at a depth of 146 feet below high-water of spring-tides. The foundation work was carried out by sub-contractors,Messrs. Anderson and Barr, of Jersey City, U.S.A.

Minutes of Procedings Inst. C.E., vol. xcii. p. 79.

There can be no doubt, the Authorthinks, that, having in view the recorded experience of the Hooghly, Poughkeepsie, and Hawkesbury bridges, this class of foundation is well suited for excessively deep work, and there is no reason why even the great depth attained at the Hawkesbury should not be exceeded if required; but this experience, especially at the latter work, points to certain essential features of design, the want of which gave much trouble. In the first place, the bottom outward splay should be avoided. If it is desired to diminish skin-friction in sinking, the increase of size in the caisson shoe should bemade by an offset, as in the amended formof caisson for piers Nos. 1,2, and 3, shown in Plate1, Figs. 6,6u, 64 6c, and 64and not by tapering, as in the first three launched. Secondly, as the wells, or dredging-tubes, are the only means by which the descent, in such large and deep caissons, can be rep- lated, they should be so distributed in plan as to facilitate this object. Fortunately,there was no serious deviationfrom the vertical in the ultimate position of the caissons described, but the tendency to cant longitudinally as regards the bridge, and transversely to the caisson, was found to be very difficult of control. The three wells, being in one line east and west, it was generally easy enough to depress the caisson to the eastwardor westward, by dredging the east or west well ; but there was no corresponding means of counteracting a northward or southward inclination of' the caisson. When sinking had proceeded to a certain depth, the caisson could not be righted by unequal loading; while the digging in any of the wells tended only to increase the cant, the dredger swinging over to the vertical and excavating below, and under- mining, the cutting-edge on that side which was already too low. The wells, therefore, should be at least four in number, and dis- posed in a diamond or quadrangular plan, so as to allow of each well acting separately for regulating the descent. Finally,notwithstanding these precautions, the proportion between the size, in plan, of the top of the caisson, and that of the finished structure or pier to be raised upon it, should be sufficient to allow of a moderate amount of deviation from the true location in theformer, without affecting the correct position of the latter. The masonry, which was of a massive character,does not call for special remark. It was begun as each caisson was complete, and was out of hand in January 1889. It was carried out by a sub- contractor, Mr. Louis Samuel, whose lamented death, during the currency of the contract, was widely deplored.

ERECTIONOF THE SUPERSTRUCTURE.

Having in view the exposed situation of the bridge, and the strength of the tidal currents, the contractors decided to construct a floating pontoon, sufficient in size to carry a complete span, and toerect the girders upon it atan elevationhigh enough to command the piers. The pontoon, when launched, was towed, with the stagingupon it, to a sheltered bay at Dangar Island, and there scuttled on a gridiron in shallow water. In this position it remained resting on the gridiron while the spanwas erected upon it ; then, at a favourable conjunction of wind and tide, the pontoon, with its load, was floated by closing the valves at low-water, and thus preventing the re-entry of the water when the tide rose. The wholewas then hauled out to the bridge, about 4 mile away, by 6-inch hawsers (Plate 2, Fig. l), extending from the island to the bridge, running throughsteam winding-engines on the pontoon and on shore, whilethe propulsionwas assisted by tug-steamersand bythe flowing tide, Dangar Island being below the bridge. Being thus moored at high-water between the piers, the pontoon was allowed tofall with the tide (or in somecases assisted bypartial scuttling),thus leaving the span in place. The pontoonwas then taken back to the gridiron, for the erection of the next span. The chief risk, therefore, was confined to the navigation of the pontoonfrom theisland to the bridge and back, an operation which required only a very few hours, andfor which the time was carefully chosen for wind, tide, and weather. As the abutments of the bridge were set back from the water’s edge (Plate 2, Fig. 2), it was necessarythat thepontoon for carrying the shore spanshould be made shorter than the girders, which overhung it at one end by a projection of 63 feet; and, as all arrangements for flotation and stability were provided for on this supposition, theremaining spans were dealt with inthe same manner and were built on the staging with the same overhang. The overhanging end wasin allcases intended to become the fixed end of the girder, and was first secured into placeon the pedestals fixed on the bed-plates. Theother end of thesuperstructure, which was about 5 inches higher, was then gradually let down upon the roller bearings and guided intoposition by vertical steel bars. Details of the pontoon and staging, on which the superstructure was put together, complete,including floor-beams, wind-bracing, &C.,

Downloaded by [ UNIVERSITY OF CAMBRIDGE] on [23/09/16]. Copyright © ICE Publishing, all rights reserved. 10 BURGE ON THE HAWKESBURY BRIDGE, N.S.W. minutes Of exceptingonly the floor-beams of the projectingtwo panels, are given in Plate 2, Figs. 3 and 4. The two panels, or 63 feet of overhang, were temporarily strengthened, during launching,by timber struts and iron ties, tomeet the contrary stresses to ,which, under the influence of the overhang, they weresubject. A fixed staging, notshown in thePlate, was erectedat the overhang end of the pontoon, andseparate from it, to support the cantilever end during erection, and a traveller, on rails 36 feet G inches apart, on the top of the staging, and commanding two complete panels, spannedthe truss transversely. This traveller was run on to thefixed staging when a spanwas complete, leaving thelatter free for launching. On thetop of this fixed staging also, were placed the hoisting engines for lifting the members of the girders from the deck of the pontoon to their places. The pontoon and staging were constructed mainly of Oregon pine timber, and the former was divided into forty-four water- tight compartments connected by valves. The whole was laterally trussed on the Bollnlansystem, by fourl+-inch steel cables secured tothe four corners, andworking freely about upright round posts at the opposite sides, as shown in Plate 2, Fig. 4. The whole of the submerged portion of the pontoon was covered with galvanized sheet-iron. Five large anchors were provided in case of emergency. The gridiron, which was parallel to the shore line, was constructedwith a transverse declivity riverwards of 3 inches in a width of 61 feet, to facilitate floating off, and with a grade of 11 inches in its length higher at the overhangend. This longitudinal inclination of the pontoon, which, of course, followed that of its bed, was reversed whenafloat, the overhang depressing that end of the girder 5 inches below the other end. The girders were built on the staging with a camber of 12 inches, which was reduced to about 34 inches when they were left on the piers. The first span set up was No. 4, this being considered the easiest to navigate, and as the currents were moderate' and all the con- ditions favourable, this span was easily placed in position in the course of a fewhours on themorning of the25th ofMay, 1888. In floating out span No. 1, Plate 2, Figs. 5, 5a, and 5b, great difficulties had to be encountered owing to the action of violent currentsand the proximity of the rocky beach. Thespan was launched on JulyIlth, and notwithstandingevery effort, the pontoon grounded upon the rocks on a falling tide, and for some hours the vessel and her load were in danger of going to pieces.

Downloaded by [ UNIVERSITY OF CAMBRIDGE] on [23/09/16]. Copyright © ICE Publishing, all rights reserved. Proceedings.] BURGE ON THE HAWKESBURY BRIDGE, N.S.W. 11 Fortunately, however, the pontoon was floated off the rocks without any considerable damage, and the span was afterwards safely deposited in position. Spans Nos. 3 and 2 werelaunched without incident on the 16th of August and the 8th of September, respectively, the latter having occupied only eighteen working days in erection on the pontoon. Span No. 7, being that next the north abutment, was started from the gridiron earlyon the 5th of October, but an adverse wind rising, the main hawser parted when half way across, and it was considered advisableto anchor for the night, and telegraph to Sydney formore steamers. These arrived during the afternoon, and were lashed alongside so as to prevent risk of the anchors draggingduring the night. The wind, however, fortunately moderated, and the operation was successfully completed at high tide the following morning. Span No. 5 was put successfully into place on the29th of January 1889. Span No. 6 was started on the 1st of March. The weather being somewhatunsettled, and, owing to the parting of the hawser during a squallof wind, the pontoon swung round into a dangerous position, from whioh it appeared impossible to avoid a collision with the structure already fixed. But by a dexterous manceuvre, adopted on thespur of the moment, the spa-n was saved from destruction by bringing it into place in the reverse position to that originally intended, and as the girders were in all respects quite symmetrical, the change in the plan of operations made no difference to the finished structure. The superstructure was erected and launched byMessrs. Ryland and Morse, of Chicago, assub-contractors tothe Union Bridge Company. It is a curious circumstance connected with this bridge that, though the successful tender was made by an American firm, the whole of the steel andiron, except only that of the eye-bar heads, was provided by the United Eingdom, where also it was manu- factured. The Authorwas in charge of the workfor the Colonial Governmentfrom nearly its commencement, under Mr. John Whitton,Engineer-in-Chief for Railways, and the contractors were represented successively byMr. M. Van Brocklin, and Messrs. Charles and Frederick Macdonald. The bridge wasopened for traffic on the 1stof May, 1889, by His Excellency Lord Carrington, Governor of the Colony, in presence

of betweenseven hundred and eight hundred guests, including representatives from all the Australian Colonies.

The communication is accompanied by three sheets of tracings and a framed photograph, from which Plates l and 2 have been engraved.

r ...... :'v, ...... _..._.._...~ ~ ...... _....;....__._._...... _..._...... I.: CS~U:~:. . .-i ._.___.______._ ...... 1...... :. ...

--. PLAN *

......

!I ...... - ...... -. . .. **:a?...... ' ...... 1

PLAN AHO SECTIOH. SHEWIHC CROSSIND.

, ,...... 4 ...... u4: a ..... CROSS SEOTIOH. LONCffUdlNAL SECTION OF SHOE.

6 l i c t 1---7g$ HALF TRANSVERSESECTION, ! ! ! I l l ! ! ! I I

I ! I I

I-

4 9

HALF PLAN OF PONTOON.