DANKS FURNACE AT THE MILLVALE WORKS, PITTSBURG, PA. 327

and finds that it is a circle struck from a centre, which is situated vertically above the centre of gravity of the weight, when in its lowest position. The equation to the curveisy=

√n2 (1 — w )2 — (x − xo) ;

where the y axis of the co-ordinate system coincides with the centre line of the crane post and the x axis is placed horizontally; n is a constant; 7 is the radius of the crane, measured from the centre of the post to the centre of the load; 2 w is the wheel base, or width between the centres of the rails, being always the smallest of these two dimensions; x, is the distance of the centre of the weight from the y axis when in the lowest position. The weight would naturally move in the same path if attached to a lever hinged at a point coinciding with the centre of the previous circle, and with a short arm attached to the end of the hoisting chain. This is the plan adopted by Michaud and Jay.

The most theoretically general case is contained in an 'arrangement by Appleby. Here two weights are employed, one hung from a lever hinged near the top of the side frames, and the second running on a curved path behind the crane, but connected by rods to the first weight and moved by it. The Author in this case finds as equation to the path of the second weight—

where p equals the length of the lever, x, is the same as above, and C is a constant. This equation represents an ellipse with dimensions depending on the construction. Appleby uses, however, a straight horizontal path.

If the first weight be omitted, but the lever be retained to move the second one, the construction patented by Schnabel and Henning is obtained. A second design, by Appleby, is somewhat similar to that of Jambille, but the pull is exerted by a chain attached to the tie-rods instead of the hoisting chain, part of the strain being taken by a spiral spring. In practice the weight never runs out quite to the theoretical point, in consequence of friction, and this has to be allowed for by taking a rather larger margin of safety in the calculations. W. P.

The Danks Furnace at the Millvale Works, Pittsburg, Pa.
By JOHN I. WILLIAMS.

(The Metallurgical Review, New York, vol. i., pp. 37-58.)

The Author, who is superintendent of the Millvale Works of Messrs. Graff, Bennett and Co., at Pittsburg, Pa., U.S.A., has had much experience in the management of the Danks puddling furnace, and considers that by the alterations he has carried out in many of the details of its construction, and by the mode of working he adopts, it has now been rendered a practical success.

In the quality of the iron produced it has no equal; but as originally brought forward, by Mr. Danks, it was liable to very frequent breakages in work, which involved difficult and costly repairs. The Author has chiefly endeavoured to avoid these breakages, and to render the repairs less expensive.

The general construction of the furnace is so well known that it need not be described here. Some of the chief alterations made in it, which are illustrated in the Paper by two woodcuts, are the following:

(1) An increase of 50 per cent. in the grate surface.

(2) Placing the body of the furnace at right angles to the fireplace, the two being connected by a movable curved flue-piece, so that the fire end of the furnace may be more readily got at for repairs.

(3) The adoption of simpler forms (and of a better quality) of bricks, for the arches over the fire-bridge and in the flue-pieces, and of a firmer mode of keying up the arches.

(4) Alterations in the forms of several of the castings, so as in some cases to render them stronger, and in others to give more freedom for expansion, or to allow of the more ready replacement of the pieces when broken.

(5) An alteration of the flue end of the furnace, so as to admit of a more free expansion of the fettling. In Mr. Danks' design, the ends were so nearly at right angles to the axis of the furnace, that the end rings (through the expansion of the fettling, when repeatedly cooled down and reheated) were constantly being broken, or the rivets started. The Author now makes the end ring next to the flue so much more conical, that the fettling, in expanding, slides up it, and the ends are thus relieved of a great part of the strain to which they were formerly exposed. The more tapered form of the end admits also of keeping it more thoroughly cooled, by water, without the risk of any getting into the interior of the furnace, and cooling it, or injuring the quality of the iron.

(6) One of the most important of the improvements made has been in the driving engines. Those first put up gave constant trouble, and have been replaced by others of greater power, more readily handled, approachable at all points for examination, and less expensive in repairs. One pair of these have been running for more than three years, and have not yet been even overhauled or repacked.

Six charges, each of 900 lbs. of pig, are worked in each furnace per shift. The melting occupies about fifty minutes. The metal is then "thickened," by means of either slag or water, after which a heavy fire is put on, and the rotation of the furnace quickened. In a short time the charge is in violent ebullition, and within about seventy minutes from charging the metal has become granulated and massed together. Balling is then commenced, and the ball when ready is carried by a fork, suspended from overhead runners, to the squeezer. The cinder is not tapped out, in the course of the operation, as is done by Mr. Danks. The squeezed ball is either

reheated, and then rolled into broad puddled bars, or if small blooms or slabs are required, it is taken from the squeezer to the hammer, hammered flat and cut into suitable pieces, put into a heating furnace for a wash heat, and then returned to the hammer, and hammered to the shape desired.

To do the entire work for ten furnaces, including the breaking and wheeling in of the pig iron and the squeezing of the balls and rolling into puddled bars, a puddler and underhand to each furnace, a labourer to every two furnaces, and eleven other men (in all thirty-six men) are required, besides a mason and helper, and a foreman to oversee the whole.

In thirty-seven days worked, during April and May 1877, 1,941 heats in all were made, with nine furnaces, working single shift; 1,746,900 lbs. of pig were charged, and 1,693,010 lbs. of puddled bars made, the loss being 3.08 per cent.; the total amount of coal used, in working the furnaces during these two months, was 40·34 bushels, per ton of puddled bar made; the ore employed for fettling was 484 52 lbs. per ton of bar; the scrap used was 49.89 lbs. per ton, and the scale 55 lbs. per ton. It is easy to obtain a greater yield of bar than that above stated, but this can be done only at the expense of the quality of the metal produced. Further economy in working will be obtained when arrangements are completed for melting the fettling in a separate furnace, and the pig iron in a cupola.

The quality of the iron obtained is most satisfactory. It may be rolled into sheets 10 inch in thickness; bars 2 inches by 8 inches, or up to 3 inches square, may be bent close without showing a crack; and bars with a fibre as fine as silk may be bent close, without breaking, after nicking them with a chisel, while other bars may be made from the same pig, showing a fracture as granular as coarse steel and yet also capable of being bent down close, when not nicked, without fracture. Copies are appended of several reports of tests of the iron, in bars and in plates. Bars 0.96 inch by 0.9 inch, and 6 inches long between datum points, had an ultimate tensile strength of about 60,000 lbs. per square inch, and stretched about 1.8 inch before fracture.

W. H.

On a New Furnace for Mercury Ores. By A. PICHLER.

(Zeitschrift des berg- und hüttenmännischen Vereines für Kärnten, vol. ix., pp. 332-338.) At the mines of the Illyrian Quicksilver Company, near Neumarktl, in Carinthia (of which the Author is the Director), mercury ores occur interspersed through limestone in such a manner that the average assay of the rock treated does not exceed 0.8 per cent. (18 lbs. of mercury per ton of rock). As this is not rich enough to bear the cost of complex methods of dressing and smelting, the furnace to be described has been designed with a view of working continuously,

and without any special requirements in the way of machinery or skilled labour. It consists essentially of a cylindrical shaft, 30 feet high, and 4 feet 3 inches internal diameter, formed of an outer shell of boiler plate 0.315 inch thick, with an internal lining of firebrick-work 6.3 inches thick; an intermediate space of 0.4 inch, allowed for expansion of the brickwork, being filled with graphite. This is mounted on a pedestal of masonry, and is provided with three openings, placed radially and equidistant at the bottom, for drawing the exhausted rock, and a charging and gas-collecting arrangement at the top, all of which can be closed hermetically when the furnace is at work. The charging apparatus consists of a cast-iron funnel, formed by a cone with a cylindrical tube descending some distance into the gas-collecting space in the furnace, closed by a movable cone, or “ bell," at the bottom of the tube, a parallel cone at the neck of the funnel, and a flat plate covering the outer rim, secured by a water joint. Each of these is manipulated by an independent system of suspension chains and counterpoises, so that in charging, two of the three stoppers are always closed. The gases are taken off by a side tube and downcome as in an ordinary blast furnace, and pass subsequently through a pair of serpentine condensers, formed of cast-iron tubes kept cool by an external circulation of water, two large chambers, and an inclined system of wooden labyrinth flues, terminating in a tower with a waterfall blast or trompe. The total length travelled by the smoke before it is discharged is about 350 feet, the draught being regulated by a damper placed at the end of the flue between it and the water blast.

The ore, as prepared for smelting, is divided into coarse and fine classes, the former including pieces from double the size of the fist, down to that of a walnut, and the latter everything smaller. There is no constant relation between the two, the large varying from ninetenths to one-half of the total produce of the mine. The working is generally similar to that of an ordinary continuous kiln. The charges are introduced in the proportion of 11.3 cubic feet of ore to 1.75 cubic feet of charcoal, the draught being so arranged that seven to eight charges pass through daily. In working, a quantity of ore equal to half a charge is constantly kept in the top funnel, and at charging time is dropped into the tube above the lower bell by raising the top valve. The top cover is then opened and the charcoal is dropped into the tube, and upon it the second half-charge. The top valve is then closed, a third half-charge is thrown into the top funnel, the cover lowered and luted; the whole operation requiring about four minutes; after which the lower valve is opened, and the materials in the tube are allowed to drop into the furnace. After this an equal volume of spent rock is drawn at the lower passages, the operation being repeated at intervals of one hour and a half to one hour and three-quarters. As the yield of mercury depends very greatly upon the management of the draught, great care is required in regulating the opening of the damper. this purpose gold plates are inserted at the end of the condenser

For

flue and in the drawing passages. If the draught is too strong, mercury is carried over and whitens the plate at the end of the flue; if it is too weak, the metal shows itself on those in the drawing holes; but when properly regulated, none of the exposed plates are in any way tarnished. The actual opening of the damper varies from 0.6 inch to 3.15 inches, according to the proportion of small ore in the charge. The auxiliary water blast has not yet been found necessary in practice. The capacity of the furnace corresponds to twenty-three charges; each charge, therefore, is forty-six hours in passing through, which time is found to be amply sufficient to exhaust the mercury in the stone. As the latter contains carbonate of lime, no flux is necessary. The mercurial soot, containing 50 per cent. of mercury, obtained in cleaning out the condensers is worked into balls with lime and charged with the ore.

During a working period of five months, 9,966 lbs. of mercurial products were obtained from the various parts of the condensers. Of this the first length of 70 feet gave 4,382 lbs. during the working, and 3,535 lbs. on cleaning up; or together about 80 per cent. of the total. A second length of 100 feet, including the first half of the condenser, gave 1,599 lbs.; a third, of about the same length, being the second half of the condenser, 407 lbs.; and the ultimate system of soot chambers and tail flues only 42 lbs. The proportion of liquid mercury to soot obtained was as 1.8 to 1 during the working, and 1 to 3.5 on clearing up, and the whole produce 39 of the former to 61 of the latter.

H. B.

Protecting the Linings of Blast Furnaces. By J. D. WEEKS.

(The Metallurgical Review, New York, vol. i., pp. 87-92.)

In modern closed-top blast furnaces, especially those working rich hard ores, with coke for fuel, and making a large amount of iron per week, the upper part of the lining has been found to wear out long before the rest, and to require constant watchfulness and repair. This is chiefly due to the battering and abrading action of the charge. Where the ordinary bell or cone plan of charging is adopted, the large lumps are thrown against the lining, not only pounding the brickwork, but making the abrasion, as the charge descends, much greater, on account of the increased size of the lumps that rub against the sides.

The upper part of the lining, when thus worn out, is frequently rebuilt while the furnace is still in blast, but the rebuilt portion is never so strong as the original lining; and much attention has been given to plans for preventing this rapid wear.

Alterations in the mode of charging, so as to distribute the charge thoroughly, without the battering action on the brickwork,