Albert Bridge
in Chelsea, London.
General
Visually, the Albert Bridge is distinctive for its Gothic iron towers. From an engineering point of view, the interest lies in its structure, which combines catenary chains with stay bars to support the continuous plate-girder deck. A support was added below the deck at mid-span in the early 1970s.
The sharing of duty between the stay bars and the chains is by no means obvious, and will be discussed later.
History
The bridge was opened in 1873 after a long period of negotiations.
Rowland Mason Ordish was appointed to design the new bridge. Ordish was a leading architectural engineer who had worked on the Royal Albert Hall, St Pancras railway station, the Crystal Palace and Holborn Viaduct.
The bridge was built using the Ordish–Lefeuvre Principle, an early form of cable-stayed bridge design which Ordish had patented in 1858. Ordish's design resembled a conventional suspension bridge in employing a parabolic cable to support the centre of the bridge, but differed in its use of 32 inclined stays to support the remainder of the load.
The bridge bore a general resemblance to the larger Franz Josef Bridge, Prague.
The nain contractor's Engineer was F. W. Bryant. Arthur Thomas Walmisley was the Engineer for the Albert Bridge Co. Ewing Matheson was the Resident Engineer for Andrew Handyside and Co.
Robinson and Cottam produced the large cast iron cylindrical sections (some being 21 ft diameter, others 15 ft) in 1871 for the piers.[1]
From the Press
1872 From Engineering 1872/09/20:-
'The bridge, when completed, will have a total length of 710 ft., and a width of 41ft. between the parapets. These will be formed of the main girders, which are of wrought iron, 8ft. deep and continuous, the upper portions being ornamentally perforated in order to lighten and improve the appearance of the structure. The main girders will be connected transversely by cross girders placed 8ft. apart, and on these will be laid the planking for the carriage roadway, which will be formed of blocks of wood placed with the grain vertically on the planking. The roadway will be 27 ft. in width, and on either side will be a footway 7ft. in width, which will be paved with diamond-shaped slabs of Ransome stone, 12 in. square and 1 1/2 in. thick, laid on planking with a layer of tar and asphalted felt interposed. The slabs in the centre of the footpath will be of a grey colour, whilst the border will be of an ornamental character.
'The course of the roadway and the two pathways will be uninterrupted throughout the whole length of the bridge, as the towers will be placed outside the parapet girders. There will be four towers carrying the main chains of the bridge, and they will be placed in pairs, each pair being connected at a height of 60 ft. from the platform level by an ornamental iron arch. The towers are of cast iron, and consist each of an inner column 4ft. in external diameter, and surrounded by eight 12-in. octagonal columns placed 12 in. from the central shaft, the whole group being connected together at intervals by disc pieces or collars of cast iron . The design carries with it somewhat the effect of Gothic clustered columns, the difference being that in the present case the surrounding columns are detached from, instead of being attached to the central shaft.
'The straight chains are composed of rolled iron bars, united end to end by rivetted joints, and having swelled heads only at the extreme ends. The curved cable, from which the straight chains are suspended to preserve their equilibrium, is of steel wire, and is 6 in. in diameter. It is composed of a series of strands of straight wires bound together by a coiled wire of smaller diameter.
'The bridge will be divided into a centre and two side openings, the former a span of 400ft., and the latter 155 ft . each. There will be a clear headway of 21 ft. at the centre of the bridge from the underside of the platform to Trinity high-water mark, the height being reduced to 10 ft. at the abutments. The piers carrying the four towers are formed of cast-iron cylinders, sunk down to the London clay and filled in with concrete. The foundations of the piers consist also of cast-iron cylinders, the bottom or cutting ring being 21 ft. in diameter, 4 ft. 6 in. high, and 1 3/8 in. thick. The next ring above this is 5 ft. high, and tapers from 21 ft., at its junction with the cutting ring, to 15 ft. at the top, from which point the pier will be constructed with cylinders 15 ft. in diameter up to the level at which the towers commence. The thickness of the metal in the coned and upper rings is 1 1/4 in. The bottom or cutting rings are noticeable as being the largest cylindrical castings ever made in one piece, to which circumstance we directed attention at the time the first ring was cast.
'One of the chief peculiarities in the Albert Bridge is the method introduced by Mr. Ordish in forming the anchorages, which, as far as we are aware, has never been adopted before. The arrangement is perfectly independent of the great mass of masonry generally employed in anchorages, the anchorage being contained within an iron structure. It consists of a cast-iron cylinder 20 ft. 6 in. deep and 3 ft. internal diameter, enlarged at the bottom into a chamber 5 ft. diameter for anchoring the chains. The cylinder is water-tight, and is provided with a manhole and steps, so that the anchorage can be examined at any time, and cleaned and pain ed when necessary. This cylinder is set vertically in a surrounding bed of concrete, the bottom being 26 ft. below the roadway level. From this proceeds a vertical anchorage chain, connected to the end of the main girder, to which is also connected the principal back chain and the wire cable. The horizontal strain is thus taken through the main girders, and the vertical lift by the mass of concrete in which the cylinder is embedded, and which is about one-tenth the quantity required in ordinary anchorages. ........
'.....Messrs. Williamson and Company are the contractors for the bridge, Mr. F. W. Bryant being their engineer. The cylinders for the piers were cast by Messrs. Robinson and Cottam, of Battersea; the cast and wrought ironwork for the superstructure is being supplied by Messrs. A. Handyside and Company, of Derby and London, and the steel wire cables by the Cardigan Iron and Steel Works, Sheffield.'
1875 From the Building News, Friday 29 January 1875:-
'SUSPENSION BRIDGES. On Friday week, January 15, Mr. A. T. WALMISLEY read a paper before the Civil and Mechanical Engineers' Society. The following is an abstract of paper :—
'A suspension-bridge consists, in its simplest form, of a platform suspended from inverted bows by means of rods (made in the present day invariably of wrought iron), the bows (or chains, as they are called), being supported by and hung from towers built at each abutment, or, in the case of multiple spans, over towers erected on each pier, and anchored down at each end by masses of masonry calculated to balance the fixed load and strain resulting from a moving load, such action tending to pull the abutments towards each other, which tendency is resisted by the weight and arrangement of material in the abutments. It is universally admitted that a suspension-bridge is that which requires much less material to support a given load than any other type of bridge with equal rigidity; and as no centering is necessary for its construction, it is well adapted for crossing deep valleys, rivers, &c.
'Though generally suitable for large spans, the suspension principle has generally been considered unfit for fast and heavy traffic, on account of the variations in the strain tending to distort the natural curve of the main chains when an unequally-distributed load passes over the roadway. The greatest amount of vibration is felt when only one or two heavy vehicles travel over the roadway. The more the platform is loaded the less the vibration experienced; and the faster the load travels, the greater are the effects of agitation or undulation of the main chains reacting in a vertical direction on the platform, the depression causing a wave to traverse the whole length of the platform. and the roadway consequently rising and falling the same in a shorter period of time. The effect of oscillation or transverse swing has been partially overcome by inserting diagonal bracing under the platform of the bridge. It is now held that this instability arises more through misapplication of, rather than for any defect in, the principle, and that when properly carried out, the suspension type will prove the most economical application of iron for bridges of large span, adapted, on the rigid suspension principle, for railway traffic, where the weight of the train is distributed over a girder platform and girder parapet, the main chains which take the load being hung in such a manner as to act as braces also, and made of a sufficient sectional area to bear any due strain per square inch which may come upon them. Where, as in the case of the Albert Bridge, the distance between the chains at towers is greater than the width of roadway, the suspending rods may be made to slope inwards ; and this has a great tendency to prevent the rocking motion, thereby rendering the structure tolerably proof against side winds. Inclined rods radiating from piers have been used in America. Parallel sloping rods may be found to give increased stiffness; but the great objection to them is their liability to sag when they are of great length, so that a load coming upon the platform straightens the rod immediately over it; but after the load has passed, they resume their original position, and thus undulation is started.
'In the Albert Bridge, which consists of a continuous girder of three spans, a load coming on one half of the centre would produce a strain on the links of the central portion, and a tendency to raise the side span if unloaded ; or if a heavier load came on the centre half than on the side span, would cause the links of the side span to sag. This is, however, prevented by the suspension rods from the curved wire-rope being connected to each link, and by the vertical anchorage fixed at end of girder. In Mr. Ordish's principle the curved chain is used to support the weight only of the other and straight chains which bear the weight of the platform and its live load. The roadway may be completely supported at any convenient number of points by these inclined chains, which are thus subject only to direct tensile strains conveyed directly to the pier ; and the only deflection in the platform, exclusive of the bending of each portion of the girder between its points of suspension, produced by a travelling load is that due to the expansion of the straight chains. In a bridge on this system consisting of three spans, the centre span may be supposed to consist of two cantilevers connected at ends or hinged to centre of curved chain, but in reality each half centre girder or cantilever is balanced by the adjoining side spans. Hence the centre span should be about twice the length of each side span. By this means the hitherto objectionable feature in a curved chain, viz., its constant alteration in form by the application of weights at various points, is to a great extent overcome by subjecting it to an equal and constant strain. It might be urged that an extraordinary rise in temperature would cause the curved chain to drop relatively more than the roadway, thus causing the straight inclined chains to sag ; and that a load coming on the platform would cause it to be depressed by straightening the chains, which, when the load has passed, will resume its curved shape, and so produce that objectionable undulation which it is the object of the principle as much as possible to avoid ; but such variation in temperature would have to be very great to exercise any injurious effect on the structure, as the strain would come on gradually, and the form of the curved chain (which has only a constant weight to support) would be independent of the load on the bridge. In Mr. Dredge's principle of suspension, one advantage of the sloping over the vertical rods may be that they convey the strain to the main chains in a direction nearly tangential to the curve, and thus transmit the strain more directly to the pier ; but in Mr. Ordish's principle the strain is sent direct to the pier by means of the inclined straight chains. Deflection, though more observable in suspension-bridges, is common to all bridges, particularly iron ones. It is, however, small compared to the vertical motion due to the flexibility of the chains of ordinary suspension-bridges. Among the various improvements adopted to preserve the true catenary curve of the chains, and thus render the system sufficiently rigid for fast and heavy traffic. We find in some cases two chains used, one after the other, forming similar curves and connected by diagonal bracing designed to withstand alternate compression and tensile strain, but leaving the platform as much subject to undulation as before. It has been attempted to make the two flexible portions of the bridge, the platform and the chain, share the action of the moving load, by bracing them together and thus lessen the undulation. This method of stiffening does not prove efficient unless the fixed load bears such a ratio to the moving load as will prevent any suspending rod from being subject to thrust This system, adopted at Lambeth, as a braced iron arch inverted, and is open to the same objections as the suspended girder with regard to the effects of heat and cold. In 1823, Brunel designed some bridges for the Isle of Bourbon, in the East Indies, upon the suspension principle, in which he placed two inverted curved chains beneath the roadway platform, with the the planes of the curves in an inclined position, with the view of reducing the oscillation and undulation of the roadway ; but the arrangement was not so effectual as he anticipated, inasmuch as the lower chains being curved, they would be equally acted upon with the upper chains by the disturbing influences. It has also been proposed to substitute for the flexible chain or cable, a stiff wrought-iron rib in the shape of an inverted arch made in the form of the catenary, of the usual section for girders, and sufficiently deep to admit of the tensile strain following a curve within the depth of the web, whether the structure be loaded or not. This method, would, however, run into time, labour, and expense, for in order to annul the straining action produced by expansion by heat and central depression, the centre and points of connection should be hinged. A more successful plan would consist in connecting with short straight chains the two points of the catenary most subject to deformation with the girders at the piers, as in the Prague footbridge. Mr. Ordish converts the supension bridge from a structure having no rigidity in itself, into one sufficiently firm for locomotive traffic, by supporting the platform from a girder bridge of a number of spans, the weight of the fixed and moving loads being transmitted to the saddles over piers by inclined straight chains subject to a tensile strain only, and maintained in straight lines by vertical rods connected to a curved chain.
'Care must be taken to allow play under the saddles for the rollers, and the details of the whole bridge must be so arranged that every part of the ironwork can be thoroughly examined and painted when necessary. The versed sine or degree of curvature usually given to suspension-bridges varies with the span, the ratio generally being from 1 in 9 to I in 15. The greater the deflection, the longer the chains; on the other hand, the less the depression, the greater the strain on the chain ; and as it is impossible to keep a chain of any length possessing weight in a perfectly horizontal position, the nearer it approaches the horizontal the greater the strain. It is advisable to bring the platform as near the lowest part of the chain as possible, as it stiffens the bridge, rendering it less liable to undulation. This is effected by giving the bridge a slight camber or rise towards the centre of the span, which not only brings the platform closer to the centre of chains, but allows the rain-water freely to run along the gutters, and, if the gradient be not too great to affect the traffic, improves the general appearance of the bridge, besides giving additional headway under the centre spans. The curved chain may either consist of a number of round or square rods with eyes at either end, or flat bar links widened at head to give the requisite sectional area at connection.
'Cables of wire are very common in America. They possess the advantage over a chain in being more easily hoisted and stronger than wrought-iron bars of an equal sectional area, and the implements needed for the work are cheaper; but wire cables present favourable opportunities for rusting unperceived, and the wires may have bends in them and not be all equally strained. There is, too, less difficulty in repairing a chain of links than a cable of wires, as every link and pin may be carefully removed in turn if necessary, and fresh ones substituted. In the Albert Bridge great care was taken to cover each wire well with oil, and to paint the rope when in position. Where care has been Observed in these particulars, there are instances in America of cables that have been exposed to the air for some years, and when examined the oil on the interior has been found sufficiently moist to have preserved the wires uninjured. The wire cables of Lambeth Bridge pass over the piers unaltered in form. On the Albert Bridge a spring is fastened in the saddle in towers, extra wires are inserted at the bearing, and the wires are spread out over the end casting connecting the anchorage with the main girder. The wires in the cables of Continental suspension-bridges are almost invariably spread out and pass over friction rollers in a flat band, so that the under ones are not worn by friction, and the strain upon each is more equalised, the rollers being, of course, well supplied with oily matter to reduce the friction.
'In the Albert Bridge the wire-rope saddle is independent of the saddle connecting the straight chains. The wire-rope, though continuous, is pulled down to a point in the centre of the platform-girder. and fastened with a strong spring fixed in the connection, the object being to prevent the wires travelling through, and enabling the wire rope to act in the same way as if an ordinary link chain took its place and held up the centre of bridge. The fixing of chains at either end of a suspension-bridge is ordinarily effected by carrying them down a sloping tunnel, and finally fixing them by keys or wedges to strong cast-iron girders called anchor-plates, firmly embedded in the masonry or concrete of the abutment. If piles are used in the foundations they ought to be driven at an angle formed by the line of the resultant pressure with the line of outward thrust produced by the inclination of the saddles on the abutments. The chamber in which the anchor-plates are fastened should be well drained. On the Continent the practice of carrying chains abruptly and generally perpendicularly into the earth is more the custom than in this country, where we generally find they are carried in a sloping direction some distance inland down to the anchorage-plates; In the case of wire cables, they are seldom carried down to the anchorage - plates, but are fixed to chains in the abutments made of iron bars, the wires generally spreading out and passing round a crupper through which an iron bar passes. Great care must be taken in proportioning the strength of the chains and their curvature. The old, slow, and expensive process of hammering the heads of suspension links to their required form, or the alternative insecure and unreliable method of welding the heads to a parallel rolled bar, have now both been superseded by rolling the bars at once into their necessary shape with an eye or hole carefully centred and drilled to receive the connecting pins, the heat necessary for obtaining a good weld being injurious to the fibrous texture that the bar acquired by rolling. Particular attention must also be paid to the manufacture of iron or steel selected for suspension-bridges, and also to the connection and bracing of the roadway platform to obtain the required strength combined with rigidity of construction.
'The Franz Josefs Bridge, over the Moldau, at Prague, was tested previous to opening with a load equally disributed over the foot-paths of 80 lb. per sq. ft., and the deflection under this test approached witiin one-eighth of an inch of the 8in. calculated deflection, the permanent set upon the removal of the load after remaining ten minutes amounting to seven-eighths of an inch. The limit of variation for expansion and contraction between extremes of temperature is about equal to the deflection registered under this test. In conclusion, Mr. Walmisley, with the aid of numerous diagrams and working drawings, entered somewhat minutely into the details of the suspension foot-bridge erected over the Moldau in 1869 ; of a bridge on Mr. Ordish's rigid suspension principle erected at Singapore ; and of the Albert Bridge at Chelsea. The latter structure was fully described in the BUILDING NEWS, for May, 23, 1873. A discussion ensued, in which the President (Mr. W. F. Butler), and Messrs. Perrett, Haughton, Bancroft, Burrell, Morrison, and Kingsford took part, and the customary vote of thanks was accorded to the author of the paper.'
Extracts from an article in The Engineer 1885/12/04:-
'The chief drawback to the ordinary type of suspension bridges, where the platform is suspended from a catenary chain, is the undulating nature of the platform under a moving load, and several methods have been proposed and adopted for meeting this difficulty. Thus Roebling, in some of his American bridges, reduced the deflection of the platform and the consequent distortion of the chain by assisting to support the moving load at various points by straight ropes or chains extending from the towers to the platform, as in Fig. 2. These ropes are light and quite subordinate to the main rope. Another plan is that adopted at a railway bridge in Vienna of rendering the chain rigid by dividing it into two parts, placed one above the other, and bracing them together, .... Another plan is that adopted by Mr. P. Barlow in the Lambeth Bridge, where he reduced the alterations in the form of the catenary by placing diagonal bracing between it and the platform, ..... .... If it were possible to maintain in straight lines inclined ropes, such as Roebling adopted, they might be used for carrying the weight of the bridge, and the catenary might be dispensed with or become of secondary importance; but this did not appear feasible till Mr. Ordish, in his Prague bridge above referred to, held the inclined chains or bars - in a straight line - as in Fig. 6 by the catenary, which, being used for this purpose only and not to carry any of the bridge load, has always the same weight upon it - namely, that of the inclined bars, the proper curve being always maintained both with equal and unequal loads on the platform, without any of the distortion usual in suspension bridges. This plan was adopted at the Albert Bridge, with the modifications that the catenary rope carrying the weight of the inclined bars or chains also supports a portion of the weight of the platform and the moving load of the bridge at points 20ft. apart, as well as the whole of the weight and the moving load of the central section of the bridge, which is more or less than 40ft. in length, according to the arrangement of the moving load. From the fact that the catenary cannot alter its form without affecting the action of the straight chains, there will always be a certain proportion between the weight supported at the apex of the catenary and that supported by it at each of its suspending points, its form also being so calculated as to prevent any change in the proportion. The present difficulty in regard to the bridge, which is the cause of the alterations now proposed by the Metropolitan Board of Works, has arisen from the defective condition of the wire rope. Mr. Ordish desired to make the catenary chains of links as he had done at Prague, and as he subsequently did at a bridge of a similar kind at Singapore; but the directors of the company having purchased some of the wire for Captain Roberts' foolish scheme above referred to, and having only a limited amount of money at their disposal, preferred that this wire be used for making the catenary, and as this plan, if properly carried out, is quite feasible, Mr. Ordish consented. The rope is not twisted, but consists of parallel steel wires about 1/5 in. diameter. Great care was taken to lay the wires so that each maintained its proper position in the rope, and the rope is clipped at intervals of about 7ft. But in order to protect the cable thus formed from rust, and from bulging between the clips, it ought to have been wrapped closely with small wire, as, for instance is done at the suspension bridge over the East River at New York, and this was, of course, the intention of the engineer of the Albert Bridge. But the wrapping of the cable is an expensive operation, requiring a special machine to do it effectually; and when this point in the construction of the bridge was reached, the funds of the company were exhausted, and, despite the engineer's protests, the cable has remained unwrapped to the present day. The necessity for replacing the rope by links, as is now to be proposed, was foreseen, and it would have been done when sufficient funds had accumulated, but for the sale of the bridge to the Metropolitan Board of Works. Thus unprotected, the cable has bulged between the clips, as was expected, leaving wide gaps into which the rain can enter between the wires, so that the rust is having a serious and rapidly deteriorating effect. If the bridge were of the ordinary suspension type, depending entirely on the catenary rope for its stabil1ty, this deterioration would be a much more serious matter; but as it is, no harm would occur under ordinary traffic if the ropes were entirely taken away. As the Albert Bridge is at the present moment probably the strongest suspension road bridge in Europe it is greatly to be regretted that the neglect of the cable renders an expensive alteration necessary. It is now proposed to remove the wire cable, and to substitute a chain made in the ordinal way of links. These links are all 6in. wide but vary in length between 21ft. and 23ft., and they afford a total sectional area for each chain of 30 square inches. The links interlace at their junction in the chains, as in Fig. 6, alternating from four to five in number; and in order to maintain the same sectional area. of 30in., the four links are 1 1/4in. thick and the five links are 1in. thick. ..... In order to take all strain from the present wire rope, and to allow the substitution of the new chain, the bridge is to be supported temporarily by timber piers, as shown in Fig. 1, each pier consisting of 13in. piles driven into the bed of the river and suitably framed together. Upon the bridge a scaffolding will be erected to support the inclined chains and the new catenary chain during the operation. The substitution of the chain will require some alteration in the upper saddles on the towers which receive the present wire rope'.
The 1885 article begs the question of why the corroded rope was not replaced by a new one with improved corrosion resistance? Replacement by chains was a major undertaking. There does seem to have been a general distrust of suspension ropes on the part of British engineers at that time, despite their large scale use in Europe and the USA.
Aspects of Construction
It is difficult to visualise how load is shared between the diagonal stays and the chain.
Referring to a somewhat similar bridge, the Franz Josef Bridge, Prague, an article in The Engineer 1868/11/06 stated that:
'The difference between Mr. Ordish's principle and the ordinary system is, that there is no curve in the main chains or bars of the former, as in those of the latter. Hence there is no motion or vertical wave in the roadway from any passing load, except that due to direct tension, and consequent extensibility of the iron by the strain induced by that load. The platform of a bridge constructed on the ordinary principle will deflect or become depressed under the load, from the disturbance of the equilibrium in the curved chains, as the load passes along the bridge. In the Prague bridge this cannot occur, inasmuch as the principle gives rigidity similar to that present in a trussed bridge, where the bottom flanges of the main girders are nearly as long as the tension chains for carrying the platform and moving loads. The main chains being of considerable length are not of themselves able to keep a straight line. They are, therefore, carried by an upper curved chain, from which they are suspended at intervals of about 14ft., this chain having no work whatever to do in carrying the platform. The curved chain was preferred by Mr. Ordish to the use of struts, as being lighter and cheaper, besides which it unquestionably adds to the graceful appearance of the structure. The adoption of this alternative was also judicious, inasmuch as a curved chain never alters its form after it is permanently loaded, that is, it can have no wave nor motion unless a moving load is carried directly along it. In the Prague bridge the straight chains are a permanent and unchangable load, and, therefore, cannot change the form of the curved chain after they have been suspended in their proper position.'
Repeating an extract above, from The Engineer 1885/12/04:
'Mr. Ordish, in his Prague bridge above referred to, held the inclined chains or bars - in a straight line ..... by the catenary, which, being used for this purpose only and not to carry any of the bridge load, has always the same weight upon it - namely, that of the inclined bars, the proper curve being always maintained both with equal and unequal loads on the platform, without any of the distortion usual in suspension bridges. This plan was adopted at the Albert Bridge, with the modifications that the catenary rope carrying the weight of the inclined bars or chains also supports a portion of the weight of the platform and the moving load of the bridge at points 20ft. apart, as well as the whole of the weight and the moving load of the central section of the bridge, which is more or less than 40ft. in length, according to the arrangement of the moving load. From the fact that the catenary cannot alter its form without affecting the action of the straight chains, there will always be a certain proportion between the weight supported at the apex of the catenary and that supported by it at each of its suspending points, its form also being so calculated as to prevent any change in the proportion.'
A key point from the 1868 article is that 'The main chains [bars in the case of the Albert Bridge] being of considerable length are not of themselves able to keep a straight line. They are, therefore, carried by an upper curved chain, from which they are suspended at intervals of about 14ft., this chain having no work whatever to do in carrying the platform.' Referring to Fig. 6, it may be apparent that the rods coming down to the deck from the catenary chain, nearest the tower, are attached to each of diagonal stays through which they pass, so they are serving to support the stays, and cannot be assumed to be providing support for the deck. On the other hand the short rod near the camera in Fig. 12 is evidently supporting the deck while also being attached to the stay bars by a stirrup.
Fig.20 shows a connection between a stay and a rod. The picture is somewhat confused by the presence of the lighting bracket.
The main girders are made of wrought iron plates joined by riveted angle iron and T-iron. Fig. 14 shows the reinforced connection between a stay and girder plates.
Figs. 16 & 17 show the central support added in the early 1970s.
See Also
Sources of Information
- [1] Wikipedia