40 C I V I L E N G I N E E R I N G

KeywordsBridges; history

Barry Mawson

Proceedings of ICECivil Engineering 138 February 2000

Pages 40-48 Paper 12086

is divisional director of GwentConsultancy in Newport,Wales

Bob Lark

is an engineering lecturer atCardiff University

‘A giant, with the grace of Apolloand the strength of Hercules’(Alderman Canning, 12September 1906)

The Newport Transporter Bridge hasbeen both loved and loathed by manyover its 90 year life, but the importanceof this unusual structure (Fig. 1) cannotbe denied and is recognized by the grade

I listed building status conferred on itfollowing its refurbishment.

Essentially an aerial ferry and original-ly designed to carry two four-wheeledvehicles, each with axle loads of 7·5 t,and a footway load of 3 kN/m2, thebridge is today capable of carrying up tosix light road vehicles and 120 pedestri-ans across the river Usk at 15 min.intervals.

Exactly 100 years ago French suspension bridge expertFerdinand Arnodin was commissioned to design the NewportTransporter Bridge in south Wales, one of the world’s fewsurviving aerial ferries. Completed in 1906, the crossing wassubstantially refurbished in 1995 and is now a grade I listedstructure.This paper reviews the historical development of thisrare and unusual bridge—which uses both cable-stayed andparabolic suspension systems—and shows that Arnodin was rightat the cutting edge of civil engineering development at thebeginning of the last century.

Newport Transporter Bridge—an historical perspectiveB. R. Mawson, MSc, CEng, FIStructE, MICE, and R. J. Lark, BSc(Eng), PhD,CEng, MICE

Fig. 1. Newport Transporter Bridge following refurbishment

41C I V I L E N G I N E E R I N G

NEWPORT TRANSPORTER BRIDGE

At a maximum speed of 11 km/h,journey time for the 196·6 m crossing istypically about 5 min and operations areonly suspended if winds exceed galeforce 6 (50 km/h). The height to theunderside of the boom is 53·95 m andthe towers reach a further 19·65 mabove the boom. The ferry or ‘gondola’measures 10 m long by 12·2 m wide andis suspended by cables from a high-leveltravelling frame. This spreads the liveload over a 32 m length of the stiffeningboom which, in the central portion ofthe bridge, was designed in such a wayso as to distribute the load as evenly aspossible to the 16 suspension cables.Closer to the towers, oblique stay cablesfan out from the tower top saddles toprovide a more direct support, and it isthe interaction between these two sup-port systems which characterizes theresponse of the structure (Fig. 2).

At the towers, both the main suspen-sion cables and the stay cables areattached to roller-mounted saddles.These are in turn held in place by a fur-ther 16 main anchor cables connected to

masonry anchorages on both sides of theriver, each of which have a designweight of over 2200 t.

The towers are of lattice construction,giving an open portal, through which thegondola and traveller can pass (Fig. 3),and the foundations for the towers areon bed rock some 25 m below the riverbank, with each pier containing around550 m3 of masonry and concrete.

OriginsThe town of Newport in south Wales

is divided into two parts by the relative-ly wide, swiftly flowing River Usk,renowned for its exceptionally high tide.Crossed at the northern end of the townfrom as early as the 12th century, and instone from 1800, the adequacy of theTown Bridge was forever being ques-tioned. It was widened in 1866 and thenpartly reconstructed in 1892–3 follow-ing the rapid development of the eastbank of the river along the then newCorporation Road (Fig. 4).

Between 1869 and 1889, variousschemes were proposed for joining the

two sides of the river at the southernend of the town but, although a ferryand foot-passenger subway were autho-rized, the latter was never startedbecause of a lack of finance and the for-mer ended following a number of fatali-ties on what was undoubtedly anextremely risky crossing.1

In 1896 John Lysaght ofWolverhampton announced his intentionto build a steelworks in the south westand, so as to attract him to a site on theeast bank of the river, proposals for acrossing were again resurrected. Tunnelsand high-level bridges were rejected forfinancial reasons, and traditional movingbridges were considered unsuitablebecause of the obstruction they offeredto the free use of the waterway. The bor-ough engineer of Newport, R. H.Haynes, had however heard of the workof the French engineer FerdinandArnodin and his ‘aerial ferry’, whichappeared to meet Newport’s needs. Thiswas therefore the proposal and, follow-ing a visit to Rouen to inspect a similarTransbordeur designed by Arnodin, theborough elected to proceed withoutdelay.

Parliamentary approval was obtainedin 1900; Haynes and Arnodin wereappointed joint engineers and detaileddesign was undertaken during 1901. In1902 a contract was let to AlfredThorne of Westminster and the bridgewas opened on 12 September 1906 byViscount Tredegar at a cost of £98,124.

From construction to closure

Construction 1902–1906Prior to construction, two separate

and independent site investigations werecarried out, the agreement betweenwhich was remarkably good. Theanchorages were founded on timberpiles driven into the underlying sandsand gravels, whereas the tower founda-tions penetrate the marl and reach adepth of 26·2 m. Because of the tenden-cy for this material to be soft in itsupper layers, particularly on the eastshore, it was also thought necessary toexamine these foundations carefullybefore the piers were capped.

Caissons were therefore used, excavat-ed by hand under compressed air at up

Boom anchor cables

Tower

Gondola

Traveller

Main boom orstiffener girder

Main suspension cablesMain anchor cables

Oblique stay cablesTower saddle

Fig. 2. Newport Transporter Bridge—part elevation

‘‘ ’’Tunnels and high-level bridges were rejected for financial reasons, and traditionalmoving bridges were considered unsuitable because of the obstruction theyoffered to the free use of the waterway.

42 C I V I L E N G I N E E R I N G

MAWSON AND LARK

Fig. 3. Bridge towers and gondola

Line of thetransporterbridge onthat of the former ferry

Town bridge

NEWPORTCorporationRoad

East banksteelworks

0 km 1

Fig. 4. Newport, circa 1900

‘‘’’

Arnodin used ‘witnesscables’, a technique onwhich he wrote severaltheses.A small cable ofsimilar material was hungalongside the main cableand its tension adjusteduntil it assumed the sameprofile.

43C I V I L E N G I N E E R I N G

NEWPORT TRANSPORTER BRIDGE

to 2·5 atmospheres, the working cham-ber being de-pressurized at the end ofeach shift to aid sinking of the steelshoe.

The towers were constructed using aconsiderable amount of temporary brac-ing and electrically-driven derrick cranes(Fig. 5), the economy of which wasproudly boasted when it was noted that

‘Less than 1,000 units of electric-ity, costing under £8 10s, weresufficient to lift and place inposition the whole of the tempo-rary scaffolding and staging, aswell as the permanent steelworkin both towers, a total weight ofabout 880 tons’.2

On completion of the towers, the mainand anchor cables were installedbetween the tower saddles and theanchorages, and the lower part of theboom incorporating the rail–bearinggirder was lifted into position piecemealfrom barges. The remaining elements ofthe boom were then simply placed on thislower part so as to establish a uniformdead load profile for the suspension systembefore assembly of the girder as a whole.

Commissioning August–September 1906Tests to prove the load carrying capac-

ity of the bridge were carried out on thecompleted structure between 29 Augustand 5 September 1906. Supervised byArnodin, they consisted of applying aseries of test loads to the gondola whilestresses were measured in the cables. Todo this Arnodin used ‘witness cables’, atechnique on which he wrote severaltheses. A small cable of similar materialwas hung alongside the main cable andits tension adjusted until it assumed thesame profile. This tension was then mea-sured and the stress in the main cablewas assumed to be the same as that inthe witness cable. No mention is madeof the accuracy of this technique, theresults simply being recorded as satisfac-tory despite an apparent standard devia-tion of as much as 30%.

As well as Haynes and Arnodin, thetests were attended by the contractorAlfred Thorne, his son John (who hadsupervised the construction) andArnodin’s assistant and son-in-law

Fig. 5. Tower construction

44 C I V I L E N G I N E E R I N G

MAWSON AND LARK

Gaston Leineklugel-le-Cocq. On thesecond day, however, it was reportedthat

‘the platform of the bridgeweighed in round figures 23 tonsmore than originally calculatedthrough the necessity of adoptingEnglish steel sections.’

The French drawings had called formetric sections, and the nearest equiva-lent oversize section had been used. Itwas therefore discussed whether to pro-ceed to the full test load or to reduce itby the 23 tons. Two more quotes thenfollow

‘The contractors expressed theiropinion that it was not wise toload a structure with a weightsuperior to that for which it hadbeen designed’‘The Engineers, sharing the con-tractors’ opinion ...’.

A wise decision no doubt. What ismore, on the following day it was notedthat testing up until then had beenbased on English tons of 1016 kg,whereas the calculations were based onFrench tonnes of 1000 kg. The problemsof communication it would seem were

just as bad in 1906 as many believethem to be 90 years on. It was howeverfinally agreed that the structure shouldbe tested with an applied load of 45tons, which it carried satisfactorily.

Operation 1906–1985Arnodin was a great believer in main-

tainable structures and, with the design,produced what amounts to a mainte-nance manual for the bridge. In thismanual he proposed a programme ofreplacement of parts so that the life ofthe structure could be extended indefi-nitely, including the systematic replace-ment of all cables such that the costcould be spread over time. He acknowl-edged that this meant initially replacingparts that still had a useful life.

We can only conjecture as to whether,had his advice been taken, there wouldhave been no problem with the structuretoday, or whether the whole life costswould have been significantly different.Arnodin was 60 when the NewportTransporter Bridge was opened in 1906and he died in 1924. As the bridge lostmoney from the beginning his pro-gramme was not, and perhaps couldnot, be followed and certainly a lowerstandard of maintenance than he expect-ed was applied.

From 1906, the bridge generally oper-ated from dawn to dusk every day, withSunday morning reserved for routinemaintenance. The bridge was mannedon a shift system, each shift requiring adriver, conductor and at least one main-tenance engineer, all of whom wereresponsible to the bridge superinten-dent. By the time of its closure in 1985the total workforce on the bridge num-bered eleven. Although tolls werecharged the bridge never paid its wayand by 1919 was costing the borougharound £6000 a year.

To reduce the cost of collection thetolls were therefore dropped in 1946,although a charge for pedestrians whowished to cross by the high-level walk-way and tower stairs was continueduntil the bridge was closed. As thebridge was classed as a transport facility,the costs of routine maintenance and theoccasional repaint have been borne bythe local highway authority throughoutits life.

The last full repaint before the 1990srefurbishment work was in 1957, whenthe cables supporting the gondola fromthe traveller were also replaced. Movingparts needed regular repair and the wind-ing ropes were changed approximatelyevery three years. The main suspensionand anchor cables received little attentionapart from some paint and an inspectionof the anchor bolts in the 1930s.

The need for refurbishment asopposed to maintenance was firstaddressed in the early 1960s when bro-ken wires in the cables were identified.In 1961 Sir Alfred Pugsley inspected thebridge and commented on the slacken-ing of the main cables as well as recom-mending replacement of the boomanchor cables, which was eventuallyundertaken in 1969. Further concernwas then expressed in 1979 over thecondition of the main suspension cablesand the oblique stay cables. The latterwere judged to be in such bad conditionthat their renewal was undertakenimmediately.

In 1985 yet more wire breakages werereported (Fig. 6) and non-destructivetesting suggested that there may also beproblems within the body of the cables.Closure was therefore the only sensiblecourse of action.

Fig. 6. Main cable wire breakages

‘‘ ’’Although tolls were charged the bridge never paid its way and by 1919 wascosting the borough around £6000 a year.

45C I V I L E N G I N E E R I N G

NEWPORT TRANSPORTER BRIDGE

Background of suspension andtransporter bridges

In order to appreciate the historicalsignificance of the Newport TransporterBridge it is appropriate to view its designagainst the background of the develop-ment of suspension bridges and trans-porter bridges during the last century.

Theoretical work on the behaviour ofsuspension bridges appears to have beenfirst addressed at the end of the 18thcentury. The problem then was the formand design of the hanging cable orchain, with the decking simply beingseen as a beam-like structure suspendedfrom it. Many problems associated withexcessive deflections were thereforeexperienced and even Telford’s MenaiBridge suffered from vertical oscillationsexcited by the wind until it was stiffenedby Rendel after damage in 1839.3,4

There was also a number of disastrouscollapses, most notably in the context ofArnodin’s future career, at Angers. Here,on 16 April 1850, two battalions ofFrench troops were crossing the 102 mspan of Pont Suspendu de la Basse-Chaine when it failed, sending 487 ofthem into the River Maine with the lossof 226 lives.5 Such events led to whatwas effectively a moratorium on theconstruction of major suspension

bridges in Europe and, although workcontinued under the direction ofRoebling and others throughoutAmerica, an improved understanding ofsuch structures was required before theywould again really be popular inEurope.6,7

To reduce the flexibility which wascharacteristic of early suspension bridgesit was recognized fairly early on that thedecks of such structures were requiredto be stiffened in some way, but that itwas not required to such an extent thatthe cables and suspension system mightbe dispensed with altogether.8 Therewas, however, no real evidence of thisuntil Barlow demonstrated it by a seriesof model tests and Rankine published hisapproximate theory for two- and three-hinged stiffening girders in 1858.9

The development of suspensionbridges in the 19th century, like many ofthe engineering developments of thattime, was therefore due both to engi-neers who were progressing their knowl-edge on the basis of this theoreticalwork and those who relied on theirexperience developed under the supervi-sion of the pioneers from the early partof the century.

Ferdinand Arnodin was born in 1845and lived in Chateau Neuf sur Loire,

where his father worked for MarcSeguin and his brothers.10 He studied atOrleans and then Paris, before joininghis father as an inspector in theCompanie Seguin. Then, in 1872, he setup his own business with a view toreviving suspension bridge constructionin France with the development of hisSysteme Arnodin. This was firstemployed on the Pont de Saint Ilpize(70 m main span) in 18795 and with ithe pioneered the use of spirally woundsteel cables arranged so that the centralportion of the main span was hung fromgroups of parabolic cables while theouter portions were supported directlyfrom the towers by oblique stay cables(Fig. 2). This enabled him to feel muchmore confident about assuming that themain cables were carrying a uniformlydistributed load.

In France, it was the work of Navierthat laid the foundations for the elastictheory of suspension bridges, and it wasin 1886 that the theoretical basis of thesystem proposed by Arnodin was exam-ined by Professor M. Levy, details ofwhich are set out in his treatise LaStatique Graphique et ses Applicationsaux Constructions.11,12

Initially, Arnodin used timber for thedecks of his structures but later turned

Fig. 7. Stiffening girder

Diagonal bracing cables 2m 2m

Fig. 8. Part elevation of stiffening girder

46 C I V I L E N G I N E E R I N G

MAWSON AND LARK

to steel, experimenting with variousforms. The boom or stiffening girderused at Newport is a double-intersectionWhipple truss (Figs 7 and 8). InAmerica it was as early as 1847 that forlong-span truss bridges Whipple recog-nized that a deep truss was required tominimize the material used.13,14 He wasalso aware that for the diagonals to bemost effective they should be placed at45º and that if the panel length was toolong, then the cost and flexibility of thedeck structure would be excessive. Tomeet all of these criteria he thereforecame up with the idea of having thediagonals cross over two panels.

It is uncertain just how much ofWhipple’s work Arnodin was familiarwith, or how much was being developedconcurrently in France, but what is clearis that Arnodin was at the forefront ofthe engineering developments of histime.

It was because his reputation as anexpert on suspension bridges thatArnodin was first approached in con-nection with a transporter bridge.Although the idea of a person-carrying‘basket ferry’ operating on the cable-carprinciple had been in use for many cen-turies, the first recorded proposal for amodern vehicle carrying structure isthat of Charles Smith of Hartlepool. In1873 he published his proposal to crossthe Tees at Middlesbrough in the jour-nal Engineering15 and, although the ideawas not immediately taken up, thebridge finally erected at Middlesbroughin 1911 is similar.16

By the end of the 19th century thegrowth in traffic, both on land and water,had generated a need that appeared to bemet by the ability of the transporterbridge to provide a low-level vehicularcrossing while providing the clearancerequired for shipping. The period from1893 to 1911 saw the construction of aconsiderable number of these structuresbefore traffic demands reached a pointwell in excess of the essentially limitedcapacity they provided.17,18

The first to be opened was in 1893 atPortugalete in Spain.19 A leadingSpanish architect, Alberto del Palacio,was commissioned in 1885 to provide acrossing of the Nervion to be known asthe Puente Vizcaya. Palacio considered

Fig. 9. Boom refurbishment

47C I V I L E N G I N E E R I N G

NEWPORT TRANSPORTER BRIDGE

the span large enough to warrant usinga suspended structure and so engagedArnodin as a specialist. Arnodin pro-duced the successful design and, realiz-ing its potential, went on to promote theform, taking out a number of patents.He designed another seven, although thelast at Bordeaux was never completedand is believed to have proposed a verylong span crossing of the River Seinenot far from the current location of thePont de Normandie.10,20,21 The NewportTransporter Bridge was the last of hisdesigns to be realized and was one ofhis largest.

Refurbishment 1991–1995Another full condition survey of the

Newport Transporter Bridge in March1991 identified the stairs and walkwaysas being in a critical condition, with the

result that access had to be severelyrestricted and even holding maintenancebecame virtually impossible. It was there-fore recognized that a decision on thefuture of this much loved, listed, locallandmark could be deferred no longer.

A £3 million financial package wastherefore assembled, with contributionscoming from Gwent County Council, theEuropean Regional Development Fund,Cadw (Welsh Historic Monuments), theWelsh Development Agency and theEuropean Architectural Heritage Fund.Work on phase I, using a bill of rates,began in the summer of 1992. Due to thecritical condition of the stairs and accessways, it was these items which were tack-led first, as were the towers to ensure thatthey were in good condition before dis-turbing the equilibrium of the cable loads.Details of this and subsequent phases of

the work are described in full elsewhere.22

A fixed-price contract for phase 2 waslet in the spring of 1994 and involvedreplacing all the main anchor and suspen-sion cables and refurbishing the cableanchorages. The cables were replaced, oneat a time, by transferring the load from thecable to be replaced to a jacking assemblyand then detensioning it to allow the cableto be removed. Replacement cables wereinstalled by reversing this operation. Morecorrectly described as spiral strand, therewas evidence that they had originally beenprotected by using a coal tar/pitch oil mix-ture, but the amount which remained wasminimal. Replacement cables are basicallyof the same construction but are com-posed of a higher grade of galvanized wire(1570 N/mm2, class A zinc coated) pro-tected both internally and externally withMetalcoat, a proprietary protective treat-

Fig. 10. Vehicles using the bridge

‘‘ ’’The cables were replaced, one at a time, by transferring the load from the cableto be replaced to a jacking assembly and then detensioning it to allow the cableto be removed.

MAWSON AND LARK

48 C I V I L E N G I N E E R I N G

ment supplied by the cable manufacturers.Prior to commencing the replacement ofthe cables, all damaged and broken cablehangers were renewed, as were some ofthe hanger pins and all the U-bolts con-necting the cables to the hanger pins.

Phase 3 included structural steelworkrepairs, surface protection, and majorrefurbishment of the main boom, gondolaand motor house. Work to the boom wascarried out from an enclosure built on thetraveller from which the gondola is nor-mally suspended (Fig. 9), but as extensiverefurbishment was also required to thegondola itself, it was removed from itscables and stationed on temporary sup-ports on the west bank. A major overhaulof the electrical and mechanical systemsand the renewal of the unusual counter-weighted fender system at the east andwest gondola docking points were alsoundertaken in this phase.

A new beginning—1995The bridge was reopened on the 15

December 1995, operates from 8 a.m. to 6p.m. Monday to Saturday and from 1 p.m.to 5 p.m. on a Sunday, and since April1996 there has been a charge of 50p pervehicle (Fig. 10). The re-opening attracteda considerable number of enthusiasts, butits use is never likely to be great and itscontribution to the highway network willalways be insignificant. Nevertheless, theunusual nature of this historic structureand the difficulties posed by its restorationhave led the authors to research the worksof Arnodin, and his contemporaries, andthis has generated a tremendous respectfor those engineers who were undoubtedlyworking close to the edge of their engi-neering knowledge. The structure hastherefore already served as a teaching toolfor many—surely a fitting tribute.

AcknowledgementsThe authors wish to thank the

Gwent Consultancy and NewportCounty Borough for permission topublish this paper. They also thankeveryone who contributed to the refur-bishment project, in particular theprincipal contractors Structures(Teeside) (Phases I and III), LaingIndustrial Engineering andConstruction (Phase II) and BridonInternational (Cable Manufacture).

References

1. JAMES L. Newport Transporter Bridge.Newport Museum and Art GalleryPublication, 1980.

2. The Transporter Bridge atNewport. The Engineer, 14 Sept,1906, 263–265.

3. PASLEY C.W. Description of thestate of the suspension bridge atMontrose. Transactions of theInstitution of Civil Engineers, 1840,III, Pt III, 219.

4. RENDEL J. M. Memoirs of theMontrose suspension bridge.Proceedings of the Institution of CivilEngineers, 1841, 122.

5. PRADE M. Ponts et Viaducs au XIXes.Paris, 1988.

6. TYRRELL H. G. History of BridgeEngineering. Chicago, 1911.

7. BROWN D. J. Bridges—Three Thou-sand Years of Defying Nature.Mitchell Beazley, London, 1993.

8. PUGSLEY A. G. The Theory ofSuspension Bridges. Edward Arnold,London, 1957.

9. RANKINE W. J. M. A Manual ofApplied Mechanics, 1st edn., CharlesGriffin & Co., London, 1858.

10. FARDET M. Le Pont de Martrou(Denier Transbordeur de France).Centre International de la Mer, LaRochelle, 1987.

11. LEVY M., Mémoires sur le Calculdes Ponts Suspendus Rigides, 2ndedn., Annales de Ponts etChaussés, Paris, 1886.

12. LEVY M. La Statique Graphique et sesApplications aux Constructions—IIIPartie - Arcs. Ponts Suspendus 3rdedn. , Corps de Révolution, Paris,1918.

13. WADDELL J.A. L. Bridge Engineering.John Wiley, New York, 1916.

14. GRIGGS JR. F. E., It’s a pratt! It’s ahowe! It’s a long! No, it’s a whippletruss. Civil Engineering Practice,Spring/Summer 1995, 67–85.

15. Proposed bridge ferry atMiddlesbrough. Engineering, 25 July,1873, 62–63.

16. PHILLIPS M. The Majestic Dinosaur (AHistory of the Middlesbrough

DiscussionIf you would like to contribute to

the discussion please post/fax/e-mail

your comments (500 words maximum)

to the editor by

30 June 2000.

Transporter Bridge). Lyndhurst,Billingham1992.

17. FORBES N. N. Transporter Bridges.Light Transport Railway League,Cardiff, 1972.

18. DUFFY, M. C.The evolution of thetransporter bridge.The NewcomenBulletin, No. 137,April, 1987.

19. TRIMINO A. P. Puente Vizcaya.Portugalete, 1990.

20. BOUYER J. P. Vie et Mort d’unTransbordeur. Presse del’Imprimerie, Chiffoieau a Nantes,Nantes, 1989.

21. Le Pont Transbordeur et la VisionModerniste. Musees de Marseilles,1991.

22. LARK, R. J., MAWSON B. R., SMITH A.K.,The refurbishment of Newporttransporter bridge. The StructuralEngineer, 1999, 77, No. 16, 15-21August.

‘‘ ’’this has generated a tremendous respect for those engineers who wereundoubtedly working close to the edge of their engineering knowledge.