Diaphragm 21 22 x22x4

Figs 6-12. POTTER

Figs 8.

NEW RAIL BEARER

RIVER

METHOD OF SUPPORTING FLOOR-SYSTEM
RIVER

SPAN

OLD SYSTEM OF JACK ARCHES

B

15.6"

CROSS SECTION

AT

AA.

THOS KELL & SON, LITH,40 KING ST, COVENT GARDEN.

(Paper No. 3554.)

"The Strengthening and Maintenance of Early
Iron Bridges.'

By WILLIAM MARRIOTT, M. Inst. C.E.

Ar the present time serious attention is being paid to the strengthening of bridges built in the middle of the nineteenth century, necessitated not only by the deterioration of these structures through ordinary wear and tear, but also by the greater stresses which they are now called upon to withstand, owing to the increased weight and higher speed of the trains passing over them. In this Paper the Author describes the measures which have been adopted for the strengthening of two bridges built in the sixties and the seventies respectively, in order to adapt them to modern requirements, and for their maintenance as regards protection from rust. An example is also given of a case in which the stresses on an old bridge have been reduced by rearranging the flooring so as to diminish the dead load.

It is a comparatively simple matter to condemn an old structure and replace it by a new one; but it is more to the credit of the engineer if he can, by careful scheming, adapt the old structure at moderate cost to the requirements of modern conditions. The difficulties of renewing parts or adding strengthening-pieces, without interrupting the traffic, are, in many cases, very great, calling for extreme care for the safety of the traffic and of the men employed in the alteration, and consequently throwing great responsibility on those in charge. All these difficulties are naturally absent in the case of new works. The addition of new plates also involves the removal of rivets in the old structure, the rivet-area of which in most cases is already deficient. The time for carrying out this work must therefore be well judged, and every rivet cut out must be temporarily replaced by a well-fitting bolt before a train is allowed to pass.

The two principal cases which have come under the Author's supervision are the West Lynn Bridge over the River Ouse, near

King's Lynn, and the Potter Heigham Bridge over the River Thurne, near Yarmouth.

West Lynn Bridge. This bridge, which spans the Ouse, a tidal river having a rise and fall of about 16 feet, with a minimum headway of about 14 feet, was built in 1866 by Messrs. Waring and Eckersley under the Norwich and Spalding Act, 1861, and consists of five spans--three central spans of 117 feet, and two end spans of 70 feet each. The girders are shallow lattice-girders on screw-pile foundations (Fig. 1, Plate 4). The bridge came under the Author's supervision a few years ago, and soon showed the necessity of strong measures being taken to render it secure for the increasing traffic. The piers, some of which were 50 feet in height from the bed of the river to rail-level, were examined by a diver, who found the iron piles in good condition, although, owing to the height of the structure, the vibration was excessive. Each pier is formed of two groups of five screw-piles, 18 inches in diameter, under the main girders, the heads of the piles being braced together by cast-iron distributing-girders (Figs. 5, Plate 4). The bed of the river consists of clay, and a bar driven close to the piles failed to locate the depth of the screwed end. Soundings taken indicated considerable scouring, and it was therefore determined to lay concrete in bags round the piles. As many as 1,668 bags of concrete were laid round one pier. The concrete was gauged experimentally at 4 to 1 and 6 to 1, and was mixed dry, but the greater part was mixed 6 to 1. Recent examination shows that scour is still going on, but beyond one rearrangement of some of the bags round the piers, little alteration has been found necessary.

The main girders are of the double lattice type, with verticals (Figs. 2, Plate 4). The longer spans, 117 feet in length, are 7 feet 6 inches in depth, and 2 feet in width of flange. It will be seen that the ratio of length to depth, namely 15 to 1, is one rarely adopted in modern practice. These girders weigh 32 tons each, as compared with 28 tons for a modern girder of the same span, but having a ratio of 7 to 1. It is supposed that the girders were originally overstrained when subjected to the Board of Trade inspection-test, giving them all a permanent reverse camber of 11

inch.

Whether this was due to their having been built without camber or to sag during construction is not known, but it is known that they had to be strengthened after the original Board of Trade inspection. From the stress-diagram the cross-sectional area required in the flanges, allowing for a stress of 5 tons per square inch, was found to be 58.8 square inches. The available area was

53.76 square inches, and, owing to the riveting being deficient, the diagonals had been re-riveted, and the rivets had so continually worked loose and had been replaced so often that rivets 11 inch and 1 inch in diameter were found in bad holes. It was therefore imperative to carry out the work in such a way as to obtain increased rivet-area and to stiffen the girders so as to reduce the deflection. It being impracticable to add a plate to the flange and keep trains running, two continuous vertical plates 1 foot 9 inches in depth were introduced, butting against the main angle-bars of the top and bottom flanges, and bearing close up to the diagonals (Figs. 2, Plate 4). In order to get the plates into position all the rivets in the main angle-bars and diagonal members had to be cut out, a length of 15 feet 6 inches, or two bays, being taken at a time. As each rivet was cut out, a well-fitting bolt was inserted, until the plate was ready to be placed in position; the bolts were then removed, and the plate, which had already been drilled, was bolted in position, and the drilling and riveting were completed. The greater part of the strengthening of the main girders was completed before commencing on the floor. Although these operations were very carefully carried out, the effect of the "undoing" was apparent. This was expected, and arrangements were made to keep a careful watch on the sag and deflection. A wire was stretched over pulleys at each end of the bridge and weighted. The sag of the wire was measured with a theodolite, and the sag of the girders below the wire was taken with callipers at the centre of each girder. The deflections were taken regularly throughout the alterations, during the passage of trains, sometimes by the method described, and at other times with the level. The deflections during the passage of a train were found to be: before the alteration, 0.87 inch; after the alteration, 0.54 inch. The total weight of the superstructure of one of the long spans was 100 tons; after the alteration it was 126 tons. The temporary extra load on each span during the alteration was 34 tons.

Alternate vertical members had to be entirely removed and refixed outside the new plates, and the diagonals had to be strengthened, having been fixed by means of only two rivets at each connection, as against six rivets required. The new vertical flange-plates were deep enough to enable these additional rivets to be provided (Figs. 2).

The old cross girders were of the box type (Fig. 3, Plate 4), and were in bad condition owing to the impossibility of access to the interior for painting. One of the webs on being weighed showed a loss of 23 per cent. on the original section, which was