testing during the operation has been dispensed with. It is true that our early results were not all satisfactory, but latterly since we have used a mixture containing about C.C. 40 per cent., Si 5 per cent., S2 per cent., P 1-4 per cent., Mn 10 per cent. in the converter, we have succeeded very well, as the results below will show. During the week ending April 10th, twelve test pieces, taken at random from the week's work, contained of phosphorus as follows:

And in the week ending April 17th, when not a single sample was taken during the operation, except in the case of the experimental blow 748, the average amount of phosphorus contained in 36 blows, all of which were analyzed, was 056 per cent., the highest being 101 per cent., and the lowest 019 per cent. The composition of this quality of steel has been in other respects very regular, the analyses and results of a test piece, 2 inches long and 533 inches in diameter, being as follows:

•40

40 ·085

Since the sampling during the operation has been dispensed with there has been comparatively little trouble with the slag adhering to the sides of the converter, and the wear of the lining has been practically uniform. As many as eighty-seven blows, representing about 630 tons of steel, have been produced from one lining without any repairs whatever, except, after the fiftieth blow, new fireclay bricks to the nose. Thirty-seven more blows, equal to about 270 tons of steel, were converted in the same lining after renewing the front or blowing side and putting in a third fireclay nose. At the end of our last week's work (April 17th), during which bricks made of the best pot clay were used for this purpose, instead of being scoured away as had been invariably the case with the commoner ones, they were little the worse

for wear, and would, we feel sure, have run for a second week. The vessel bottoms (all made with a mixture of tar and lime rammed round pins) during this same week, averaged eight blows each, the maximum being twelve and the minimum four; but this bottom was taken out on completion of the week's work, and was very little worn. With respect to output, during the eight weeks ending April 17th (omitting Easter week, which was a broken one), 3380 tons of steel were made, or an average of 422 tons per week, the largest week being that ending March 27th, when 541 tons 7 cwt. were produced. In conclusion, we think it will be apparent to all that there are no difficulties in the working of the process. We are satisfied that as good steel can be produced by it from phosphoric pig, and quite as regularly as that obtained from hematite in the ordinary gannister-lined converter. We have shown that it is an easy matter to produce a malleable ingot iron, containing practically no carbon and very little manganese by this method, which, to say the least, it is rather difficult to do from hematite pig-iron by the old process; and although our only experiment (results of which we have laid before you) to produce a harder quality of steel from phosphoric pig high in manganese, without the addition of any manganiferous pig, at the end of the operation did not turn out as satisfactorily as we could have wished, there appears to us little doubt but that this too will be accomplished, as well as the production of soft steels, suitable for boilers and ships' plates, etc., in a similar manner. It is true that our production from one pair of converters has never yet exceeded 541 tons 7 cwt. in one week, but it must not be forgotten that the process is still a new one, and that the plant at our disposal is not of modern construction, nor well adapted to its requirements. These requirements are, in our opinion (on account of the lining), either additional fixed converters, or duplicates with proper facilities for changing, as well as suitable arrangements for the speedy removal of the large quantity of slag. Under these conditions, in a well-arranged shop, we feel sure that not only as great an output can be obtained by the process as is now being produced in the best English practice in the conversion of hematite, but also that it may be made to equal anything that has ever been accomplished by our friends on the other side of the Atlantic.

ON THE STEEL-COMPRESSING ARRANGEMENTS AT THE BARROW WORKS.

By ALFRED DAVIS, of London.

Paper read before the Institution of Mechanical Engineers, at Barrow.

The unsoundness of steel castings, particularly in the case of ingots made by the Bessemer or Siemens-Martin process, has given manufacturers considerable trouble, and occasioned much waste of material.

A good deal has been stated and written of late as to the cause of this unsoundness, which occurs principally at the upper end of the ingot; but it appears now to be pretty generally conceded that the defects proceed from two distinct causes: First, the existence of gases, generated at the point of transition from the fluid to the solid state, which are imprisoned in the form of bubbles when the surrounding metal becomes solid; and secondly, the existence of spaces formed by the natural contraction of the metal in cooling, by reason of the outer skin first becoming solid, and refusing to follow up the interior portion of the ingot, which subsequently cools, and consequently occupies a smaller space.

Various systems, designed to cure this evil, have already been discussed before this Institution. The system of compressing fluid steel by the direct application of high pressure steam, has recently been adopted by the Barrow Hematite Steel Works, and by Messrs. Bolckow, Vaughan & Co., and has the merit of simplicity combined with efficiency. The arrangements adopted for the purpose are founded upon those used by Mr. H. R. Jones of the Edgar Thomson Steel Works at Pittsburgh, where the system has been used for some years.

A high pressure steam boiler is provided, and communicates with a receiver, which is attached to the side of the ingot crane, and which is furnished with a row of cocks corresponding with the number of ingot moulds. From these cocks strong india-rubber pipes convey the steam to the ingot moulds, which are arranged in the arc of a circle round the ladle crane. The metal from the ladle is poured through a loose pouring cup, which rests on a conical seat at the top of the ingot mould. As soon as the pouring is finished, this cup is removed, and a lid having the steam pipe ready coupled to it, is placed on the top of

the mould, and secured to it by a steel cover. The cock on the receiver is then opened and the steam allowed to act upon the metal until it has completely set. The result of this pressure is to make the ingot sensibly shorter than when cast in the ordinary manner, the difference, according to experiments made at the Edgar Thomson Works, being from 13 inch to 2 inches in a 5 feet or 6 feet ingot. The ingots when cold are perfectly level at the top, and there is no porous head requiring to be cut off.

The arrangements adopted by the Barrow Steel Company differ somewhat from those in operation at the Edgar Thomson Works. These arrangements require only a very brief explanation.

The ingot moulds, which are of similar construction to those used by the Edgar Thomson Company, are placed in a row, within a dock or siding, the centre line of which runs to the centre of the pit. The metal flows from the ladle into a trough mounted upon wheels and provided with runners at points corresponding with the centres of the ingot moulds when the trough is in position. This trough runs upon rails, placed on either side of the row of ingot moulds, and can readily be removed after the moulds are charged. Each mould is provided with a steam-tight cover, having a wrought-iron pipe attached to it, furnished with a stop-cock. This pipe communicates at right angles with the main steam pipe, which runs parallel with the side of the dock. The junction of the branch steam pipes with the main is formed by means of a cast-iron sleeve piece, with stuffing boxes, to enable the covers, with their respective cocks and pipes, to be thrown back out of the way when not in use.

The boiler for supplying the steam has been constructed by Messrs. Daniel Adamson & Co. It is 3 feet 6 inches in diameter and 9 feet high, and is intended to be worked at a pressure of 200 pounds per square inch.

At the Cambria Steel Works in Pennsylvania an attempt was made two or three years ago to inject water through the cover of the ingot mould after the metal had been poured. The heat of the molten steel of course generated steam, which acted as a compressing medium; a safety valve being provided and loaded to the pressure required. The disadvantages of this system, as compared with that now described, are › sufficiently obvious; the complication of parts and the danger from explosions being very great.

The results obtained by the process of casting ingots under steam

compression are highly satisfactory. Not merely is the ingot perfectly sound, but the action of the steam is such as to enable the men to work it earlier and in a hotter state than with the ordinary method, so that there is an appreciable increase in the output. The presence of the steam also acts beneficially on the sides of the mould, and causes it to last longer.

The pressure necessary to produce a perfectly sound ingot will depend upon the quality of steel to which it is applied. At the Edgar Thomson Works it is found that for ordinary rail metal 100 pounds per square inch is sufficient. But for milder steel a higher pressure is needed; and since experience has proved that steam is readily dealt with at very high pressures, there does not appear to be any reason why 1000 pounds or 1500 pounds per square inch should not be applied if required. It is only a question of giving sufficient strength to those parts which are exposed to the pressure. As a matter of fact the boilers designed by Mr. Loftus Perkins will carry a steam pressure of 2000 pounds per square inch with perfect safety. The question of making tight joints between the ingot moulds and covers with such high pressure is one of considerable importance; but there are several ways in which this difficulty may be overcome. In using steam at a very high pressure, the size of the supply pipe may be considerably reduced, and the mode of attachment greatly simplified; and since the amount of steam used is inconsiderable, the size of the boiler would be correspondingly small. As an alternative, in cases where high pressures are needed for the consolidation of fluid metals, the author proposes the use of compressed air. With this system a pressure up to 1500 pounds or 2000 pounds per square inch may be obtained without danger or difficulty, as is completely demonstrated by the torpedo practice at Woolwich, and by the experiments carried out by Colonel Beaumont, in connection with the use of compressed air for tramway locomotion.

The advantages of an elastic compressing medium in the consolidation of fluid metals, as compared with the hydraulic process, scarcely need to be dwelt upon. In applying hydraulic pressure a rigid piston is necessary; and the outer portions of the cooling mass (which are the first to set) must be crushed down, before the interior portions, which are still liquid, are reached by the pressure. A considerable amount of power is wasted in consequence. In addition the fluid. metal is forced against the sides of the mould, and in a contrary direcWHOLE NO. VOL. CX.-(THIRD SERIES, Vol. lxxx.)

14