The Forth Bridge/Sinking of the Caissons

by Wilhelm Westhofen

Sinking of the Caissons.

The operation of sinking a caisson through mud, silt, or clay differs somewhat from that pursued in sinking through solid rock, at any rate in the mode of attack.

In the Queensferry caissons the first thing to be done was to fix pipes in the air-shaft, one for the admission of water, the other for the removal of mud diluted with water. As this process of ejecting matter will be described further on, it need only be mentioned that by degrees the shaft first, and next a space underneath it in the air-chamber, was cleared, and access had to the latter. As soon as the ejector pipes described above could be reached, flexible hose was attached to them, and a larger number of men employed to clear the caisson of the semi-fluid mud. Great care was required during this early stage of the work, and the weight of the caisson regulated to a nicety, for there was nothing but its buoyancy to prevent it from suddenly descending and smothering the men below. At low water was naturally the most dangerous time, there being the least displacement in action then, while the weight was greatest, and moreover the cutting edge resting upon a treacherous ground. The men were generally withdrawn at this time, and the air pressure diminished, allowing the caisson to descend as far as it would go. On one of these occasions the caisson suddenly descended some 7 ft., and not only the air-chamber but part of the ascending shaft became filled again with mud and silt. It is satisfactory to be able to say that not a single life was lost here from this cause, which in other works had proved fatal to so many.

Fig. 47. Section of Tilted Caisson after Completion.

HYDRAULIC SPADES
Fig. 48. Sinking the Queensferry Caissons.

In using the ejectors the following mode of working was employed. (See Fig. 48.) A sort of sump or hollow in the ground was formed, and into this the water from an overhead tank was allowed to flow in any quantity desired, and mixed with the more solid material excavated. A man who held the end of the flexible hose attached to the ejector manipulated the same in such a manner that a certain amount of the air inside the chamber was allowed to enter into the ejector pipe with the mixture in the sump. This air in escaping carried with it a certain amount of liquid, and forced it out at one of the openings in the caissons above water level. The operation is shown in the cut figured here. It is somewhat puzzling at first to understand why the air pressure, which is only just equal, or only slightly exceeding that which is due to the head of water outside, should lift and discharge a quantity of semi-solid material at a level above the water; the explanation will, however, lie found in the fact that the velocity of the air due to this pressure is very much higher than that of water flowing under the same head, and that it carried mechanically along with it as much of the fluid as it could convey. Of course the man who manipulated the end of the hose in the sump had to feel his way into the solution of this problem, and had to vary the quantities or proportions of air and liquid according to the requirements of the moment. Outside the caisson the flow from this ejector was not continuous, but in gulps, like that from a single plunger pump, now in a large mass, now in a thin stream, and at times nothing at all.

As soon as all the soft material had been removed, and a bed of stiffer clay reached, the ejector could be used only for removing the superfluous water, and picks and heavy spades had to be resorted to and the material charged into the buckets and hoisted out by the locks. Lines of narrow-gauge railway were laid down to run small bogies, on which the skips were carried in all directions within the chamber. Presently the hard boulder clay was reached, which was nearly dry, and which proved tougher and harder than anything the existing tools could work in. Various means were tried, and even blasting by powder or dynamite resorted to, but with little result. The rate of progress became exceedingly slow, and the men's energies became quite exhausted. Here Mr. Arrol's ingenuity fortunately came to the rescue, and he devised a most simple, yet a most efficient, tool to grapple with this difficulty namely, a hydraulic spade shown in Figs. 49 and 50. It consisted in the main of a ram, to which a spade is attached, the ram fitting into a cylinder, which represents the handle of the spade. On the top of the cylinder was fitted a head-piece, which could be set against a piece of batten or any other convenient object. The ceiling of the air-chamber furnished the resistance, and the projecting rivet heads were useful in preventing side slip of the head-piece. The spade was lifted by two men, a third attending to the fixing of the head and the turning of the cock admitting water pressure. This was of the ordinary amount, namely, 1000 lb. per square inch, less the amount of air pressure inside the caisson.

The spade was set on the ground, the pressure turned on, and as soon as the head-piece had been firmly set against the ceiling the full pressure was given, and the spade forced down into the clay. The water was then exhausted and the spade brought forward, thereby detaching a slab of clay from 16 in. to 18 in. deep, and from 2 in. to 4 in. thick. The spade was then set up again, and the clay thus cut away in long ridges all over the area. Many of these slabs of clay had to be cut to allow them to be charged into the skips. Any boulders too large to pass through the air-shaft were either drilled and blasted, or left in the chamber to be hereafter incorporated with the concrete filling. The water discharged by the spades had to be collected in a sump and discharged through an ejector.

Fig. 51. Section of Caissons With Air-locks and Working Chamber.

TABLE No. V.—Progress of Work with Queensferry Caissons and Piers.

—	South-West.	South-East.	North-East.	North-West.
Launched and towed to berth	May 26, 1884	Aug. 24, 1884	Nov. 24, 1884	Dec. 3, 1884. Submerged Jan. 1, 1884. Floated again Oct. 19, 1885
Commenced sinking	Sept. 1, 1884	Nov. 24, 1884	Jan. 28, 1885	Nov. 25, 1885
Level of cutting edge at commencement below high water	33 ft.	33 ft.	37 ft.	52 ft.
In final position	Dec. 6, 1884	Feb. 4, 1885	April 10, 1885	Feb. 4, 1886
Level of cutting edge below high water	71 ft.	73 ft.	89 ft.	85 ft.
Commenced concreting of air chamber	Dec. 8, 1884	Feb. 7, 1885	April 14, 1885	Feb. 5, 1886
Finished concreting of air chamber	Dec. 17, 1884	Feb. 18, 1885	April 25, 1885	Feb 10, 1886
Concrete up to granite level	Feb., 1885	April, 1885	June, 1885	March, 1886
Pier completed	July, 1885	Sept., 1885	Sept., 1885	June, 1886

In digging or undercutting the edge of the caisson all round, the spades had to be worked at an angle varying from 30 deg. to 60 deg., the rivet-heads coming in usefully hero to provide a good hold for the headpiece of the spade.

In undercutting round the sides, portions were left standing to carry the weight of the caisson while excavation was still going on, and these portions were by degrees removed until the bearing surface became too small to sustain the weight, and the caisson settled down into them. In proportion as the caisson descended, more weight was added on the top; but this was also required to be done with care, for fear of displacements taking place. After entering into the hard clay, the caisson edge was nearly sealed by the pressure of the water upon the clay outside, and instead of the pressure of air corresponding to the head of water outside, a few pounds sufficed to keep the caisson dry.

Thus, in the case of the Queensferry north-east caisson, founded at 89 ft. below high water, the outside head is equal to fully 42 lb. of air pressure inside, but it was actually worked and finished with a pressure of from 15 lb. to 18 lb. per square inch, after once the hard clay had been entered. This made it, of course, much easier for the men to work in, apart from a considerable saving of wear and tear of the machinery.

It will be readily understood that it was of paramount importance that the caissons, when sunk to their final depth, should be in the correct positions laid down for them; yet it is equally easy to understand that in sinking such a mass it would not take a great deal to cause a displacement in one way or other, since there is nothing to hold it in its proper centre.

The position of each caisson was therefore checked nearly every day, and if it deviated in any way from its right course, steps were at once taken to remedy this. To do this it was only necessary to undercut the caisson edge on the side to which it had moved, and to gradually tilt it to a small degree in the direction of that side. When this had been done the caisson was sunk further down for a distance with this tilt on it, and was then righted by undercutting on the high side. This was repeated until the centre of the caisson had got back to its proper position. The slight upward taper in the cylindrical portion of the caisson added some little help in that direction, and the writer thinks that a little more taper than that given to the Forth Bridge caissons would be found very advantageous in future works of this kind. Another plan is to set up inside the air-chamber and against the top of the sloping plates of the shoe, a number of strong timber struts at an angle to the perpendicular, and well secured against some timber or stone in the solid ground. This should be done previous to the caisson being allowed to descend. When it does so all these struts have a tendency in descending to become longer and to force thereby the caisson over in the direction in which it is intended that it should go. This should also be done at high tide, when the buoyancy of the caisson is largest. As in all things, so here, prevention is better than cure, and it is best to keep a sharp eye upon the workmen in the chamber, to insist that the edge be undercut on the outside to the extent of 6 in. to 8 in. all round before it is allowed to descend, and to have the position checked as frequently as possible, in order to find out a movement to any side at once, and not allow it to become too great to be remedied. In working on a sloping face, whether of rock, silt, or clay, it is good practice to keep the caisson tilted very slightly to the lower side, which has the effect of forcing it up, but this should not amount to more than the natural batter of the outside shell. A few inches one way or the other is of no consequence in so large a caisson, as the pier raised upon it could easily be centred as much as that to either side, but it is just as easy and much better to keep it right from the first. During the sinking of No. 4 caisson, owing to an error by one of the surveying staff, the caisson got out of position to the extent of 10 in., but it was brought back to its true position in less than three days, though as much as that could not be done in every case.

FLOATING OUT CAISSON FOR QUEENSFERRY PIER, MARCH 26, 1884.

The excavation and the rate of progress varied naturally with each caisson, and while for instance the first, sunk through a depth of 38 ft. of material, took ninety-six days to finish, the last was done in sixty-nine days, through over 30 ft. of nearly all stiff material. Table No. V., on the preceding page, gives principal dates and figures of interest in connection with these four Queensferry caissons.