Percy, Metallurgy: Iron and Steel (1864)

Books Metallurgy: Iron and Steel, The Manufacture of Iron, The Manufacture and Properties of Iron and Steel

Covers J. Percy, "Metallurgy: Iron and Steel" (1864, 995 PDF pages). The PDF has a text layer on its first page only, so the text was read from OCR of every page. PDF pages 3, 4, 6, 782, 784, 786 and 974-993 OCR empty (front matter, the Ebbw Vale plan plates, the back index). The scan is poor in places: folios 623, 770, 772, 830 and 834 and the casting-shop plates (831-832) were read from rendered images, and small fractions (bar thickness, charcoal mesh, the converting furnace's tonnage on p. 772) are illegible in this copy and are not quoted. Pages below are printed folios: the PDF page less 20 for the refinery, puddling and forge chapters (PDF 641-744), less 22 at the Swedish blast furnace (PDF 395), less 26 for the steel chapters (PDF 790-890). Read: the refinery (pp. 621-626), puddling - furnace, manipulation, Ebbw Vale practice, the boiling process, yields (pp. 641-647, 654-658, 670, 720-722), working of the ball - forge hammers, steam hammers and their patents (pp. 693-701), cementation (pp. 764-773), puddled steel (pp. 790-794), Bessemer and Mushet (pp. 814-824), casting steel - Huntsman, the casting shop, pots, the heat, Krupp (pp. 828-837), steel in the reverberatory and in the cupola (pp. 838-839), Heath's manganese (pp. 840-841), shear steel (p. 859), and the Swedish charcoal blast (p. 373).

What the book says

What the book says

Cementation (pp. 764-773)

  • An old process, "well described by Reaumur in 1722"; the converting furnaces of his engravings are "similar in all essential respects to those which have so long been in operation at Sheffield" (p. 768).
  • The furnace: an oblong fire-brick chest open at the top inside an arched chamber, a grate beneath, flues rising round the chest into the arch, short chimneys at the spring, the whole inside a brick cone like a glass-house; two chests side by side (pp. 768-769, Figs. 208-211).
  • The charge: flat bars about 3 in. broad, a little shorter than the chest; ground charcoal on the bottom, a course of bars flat side down, a layer of charcoal, a course of bars, and so to the top, where a thicker layer of charcoal is sealed with "wheelswarf" (grindstone dust and steel) or clay "as impervious to air as possible" (pp. 769-770). Fresh charcoal is mixed measure for measure with the last heat's, sifted and washed; fresh charcoal alone "acts too finely", so the furnace must run cooler and conversion takes longer (p. 772).
  • The firing: "evenly kept at glowing redness" for a time set by the temper wanted - "Steel for melting purposes requires from 9 to 10 days, for shearing 8 days, and for spring temper 7 days". Trial bars protrude through plugged holes and are drawn and broken to judge the conversion by fracture (p. 771).
  • The fire is then withdrawn and the furnace cools for several days, "during which a further degree of carburization occurs"; the man-holes are opened after three or four days, the bars "broken across, and duly assorted according to the temper indicated by the appearance of their fractured surfaces" (p. 771). Chests that are too large heat unevenly and the bars toward the centre carburise less (p. 772).
  • White-ash coal that does not clinker, and careful stoking, keep the temperature even (p. 772).
  • The converted bar is blistered, "brittle, and may be easily broken across", the fracture dull and laminated (p. 772). The blisters come from cinder in the bar (p. 772); Henry blamed sulphur (p. 773).
  • Macintosh's gas cementation (1825) worked in hours but no chamber stayed air-tight at the heat (p. 773).

Casting steel (pp. 828-837)

  • Before Huntsman steel was never homogeneous: cementation carries the bar-iron's cinder into the steel; fusion gives "ingots of the metal ... of perfectly uniform composition throughout" (p. 828).
  • Huntsman: born 1704, died 1776 at Attercliffe; experiments at Handsworth; moved to Attercliffe about 1770 (p. 829). No year is given for the invention.
  • "The present method of casting steel is substantially the same as that practised by Huntsman, the difference ... being that now two crucibles instead of one are heated in each furnace" (p. 830).
  • The shop (pp. 830-832, Figs. 214-217): furnaces are "simple rectangular chambers, each having its own flue in the stack common to a row of furnaces"; their tops level with the casting-shop floor, a cellar below; covered with 1 in. cast-iron plates, each hole with a moveable fire-brick cover in a wrought-iron frame; coke fuel; lined with fire-clay brick, ganister or millstone grit. The drawn shop has two rows of six holes, each row on one tall stack, an annealing grate and an open fire beside them, a pot room, a mixing room and a coke shed.
  • Pots (pp. 833-835): pressed in a cast-iron pot-mould round a wooden plug; lids and stands of the same clay. One Sheffield house's pot weighs 25 lb: Derbyshire (Edensor) clay 15, Stannington clay 7, old ground pot 2, ground cinders 1; coke dust 1.5-2 lb per 20 lb of dry clay; trodden barefoot 8-10 hours. 24 lb of clay makes a pot for 28 lb of steel; pots hold 28-30 or 45-50 lb; a man makes 20 a day; they dry 8 days.
  • The heat (pp. 835-836): about twenty pots are "annealed, i.e. gradually heated to redness" mouth down on the grate overnight, then set upright on their stands in furnaces charged with live coal; "brought up" about 20 minutes; charged through a sheet-iron funnel with 30-40 lb of "broken up and assorted blister-steel"; lidded; the furnace filled with coke; refilled after 3/4 hour; after another 3/4 hour the foreman lifts the lid and adds coke by judgment. "When manganese is used it is added along with the steel."
  • Teeming (p. 836): the "puller out" lifts the pot to the teeming hole, a cast-iron box floored with coke dust "so that the pot may be prevented from coming in contact with the iron and cracking"; the lid goes onto the pot just teemed, "which is then immediately returned to the furnace". Slag at the lip is held back with the "flux stick", a rod with a ball of slag on its end; sand and an iron stopper cap the ingot.
  • "A pot will last three meltings, but each time with a diminished charge: thus, in the 1st charging, it may receive 45 lbs, in the 2nd 40 lbs, and in the 3rd only 36 lbs. The furnace needs repairing once a month." Ingot moulds are split cast iron, smoked with coal tar; the steel "eats into the bottom", honeycombing 6 in. moulds (p. 836).
  • A Siemens regenerative crucible furnace at Kapfenberg heats 20 pots of about 60 lb at once on brown coal, 3 of coal to 1 of steel (pp. 836-837). Krupp melts puddled steel in pots of 70 lb (p. 837).
  • "Cupola furnaces cannot be employed for the purpose, even if the purest fuel ... were used; because the metal would be materially changed in composition by contact with solid carbonaceous matter" (p. 838). Sudre melts steel in a reverberatory under "a flux of common bottle-glass, or of blast-furnace cinder" (p. 839).
  • Heath's patent of 1839 puts 1-3 % "carburet of manganese" into the crucible with the broken blister, or with malleable iron and carbonaceous matter (p. 840). Lucas (1791) and Mushet (1800) patented ore and carbon melted in pots (pp. 764-765).
  • Shear steel: blister broken to 16 in. lengths, drawn at a red heat, piled, welded in a hollow coke fire under clay dust, drawn to 2 in. square; broken and rewelded for double shear; named for cloth-shear blades (p. 859).

The refinery (pp. 621-626)

  • A rectangular coke hearth with three water tuyeres a side at 3 psi cold blast, water-block walls, a tap plate in front. 1-1.2 t of pig and 30 lb of hammer slag melt in about 1.5 h; tapped about 2 h after charging, having been blown about 1/2 h; about 4 cwt of coke. Metal and cinder run out together along a water-cooled bed; ridges of cinder across the bed leave lines to break the plate along (p. 623).
  • Pentwrch: 9 cwt of mottled pig to 8 cwt of refined metal (p. 625).

Puddling (pp. 641-658, 670, 720-722)

  • The furnace is tapped through a hole under the fore-plate into wheeled tap-waggons (p. 646). The rabble is an 8 ft, 40 lb bar; the paddle a lighter bar with a chisel edge (p. 646). A furnace lasts about six months and is repaired every week (p. 647).
  • The sides are lined with "puddling mine" or "bulldog" (roasted tap-cinder) plastered with haematite paste; the bottom is oxide of iron; about 1 cwt of scrap is worked into a ball each 12-hour turn to coat the bottom (p. 655).
  • A Bromford heat, 21 July 1859 (pp. 655-656): 1 cwt refined iron, 3 cwt forge pig and about 1 cwt of hammer slag and scale charged at 1.50; melted by 2.25 with the damper up; the damper let down once melted; raised for the boil at 2.36, "jets of blue flame escaping everywhere"; "come to nature" by 2.53; at 3.05 the first ball out, the damper partly down and the stopper-hole shut, "the body of the furnace should be full of smoky or reducing flame" to stop waste; six balls of about 80 lb by 3.14, "immediately after which the cinder was tapped off".
  • Worked "cold", the iron is red-short (p. 657). A ball drawn too early and put back is a "cobble" (p. 657).
  • Fuel: about 24 cwt (of 120 lb) of Thick coal to 20 cwt (of 112 lb) of puddled bar, 2 cwt more of New Mine coal; 21.5-22 cwt of pig to 20 cwt of bar; a heat takes 1 to 1.75 h (p. 657).
  • Ebbw Vale (p. 658): 4.5 cwt charges of white pig and refined metal half and half, hammer slag and "mill scale (i.e. scale from the rolls)" added by the pig; six balls a heat, seven heats in 12 h, six on grey pig. Cinder is tapped every heat on grey pig, every other heat on white, and once a turn on refined metal, whose cinder builds until it is cut out.
  • Hall is credited with pig-boiling, the cinder bottom and bulldog (p. 670). Cort's puddling patent is 1784, his rolls 1783 (pp. 627-629).
  • Yields (pp. 721-722): at Dudley 24 cwt of pig to 22 cwt of puddled bar, 1 t 1 cwt 3 qr to the ton "good work", 1 t 4 cwt with an unskilful man; puddled bar to finished bar about 1 t 2 cwt 1 qr to the ton.

Working the ball (pp. 693-701)

  • No ball "whether produced in the charcoal finery or the puddling furnace" has been treated otherwise than hammered or squeezed and then rolled; the Forge hammers and draws puddled bar in grooved rolls, the Mill makes merchant iron (p. 693).
  • Tilt hammers 1-5 cwt; a 10 t belly helve at the Mersey works lifts 18-24 in. at up to 60 blows a minute and forged a 27 t gun (p. 694). A hammer face lasts three to seven days (p. 695).
  • The steam hammer, proposed by Watt (1784), sketched by Deverell (1806), "first reduced to a practical form by Nasmyth in 1842"; the 9 June 1842 patent "contains the first drawings of the perfected single-acting hammer" (pp. 695, 701). Helves "have to a considerable extent been replaced in the majority of English forges" by it (p. 695).
  • "Balls, which would crumble to pieces under the helve, may, by gently regulated blows under Nasmyth's hammer, be easily wrought into solid blooms or slabs"; the helve cannot "coax" (p. 695).
  • Hammers of 1.5-5.5 t make puddled blooms and slabs; one 50 cwt hammer serves 11-15 puddling furnaces, "the necessary steam being obtained from the waste heat of two reheating furnaces" (p. 700).

Bessemer (pp. 814-824)

  • Mushet's patent of 22 September 1856 adds molten spiegeleisen to the blown metal. It was "very difficult, if not impracticable, ... to ascertain with certainty when decarburization has proceeded to the right extent", so the metal is blown out and spiegel added; "spiegeleisen has given a value to Bessemer's process ... which certainly it did not previously possess" (p. 814).
  • Blown iron is "highly crystalline on fracture", its ingots porous (p. 817).
  • Sheffield, 1863 (p. 823): 3-4 t charges, 15 psi blast; "The blowing is continued from 12 to 20 minutes"; then "from a small furnace a specific quantity of molten spiegeleisen, usually from 5 to 10 %, is run in; the blowing is then resumed for a period of five minutes", and the whole goes to the ladle.
  • Loss "about 15 %" in conversion, plus "about 7 1/2 %" melting the pig in the air furnace; a vessel blows about 250 t before relining, tuyeres last about 10 t (p. 824).

Charcoal blast (p. 373)

  • In Sweden the charcoal blast "is always much less than in furnaces worked with coal or coke", 1/2 to 1 1/2 psi; hard charcoal takes the most, light charcoal the least.

Against other books

Percy set beside Overman, The Manufacture of Iron (1851), and Campbell, The Manufacture and Properties of Iron and Steel (1907).

  • Pot life. Percy's Sheffield clay pot lasts "three meltings" with charges of 45, 40 and 36 lb (p. 836). Campbell gives "four to six heats" and 80 lb "when the crucible is new" (p. 95), American practice of 1907. Both books shrink the charge as the pot ages.
  • The carbon trim. Percy's blister is "broken up and assorted" (pp. 771, 835); the Sheffield addition is Heath's manganese, 1-3 % (p. 840); carbon goes into the pot only with iron charges (pp. 764-765, 840). Campbell grades by fracture and puts charcoal in the pot, which Percy shows as American (pp. 94-95).
  • Uniformity. Percy has "perfectly uniform composition throughout" (p. 828) and coke dust in the pot clay (p. 834). Campbell has silicon and carbon taken up from the pot, a lot running .85-1.05 % C (pp. 95, 99); the coke in the clay is where that pickup comes from.
  • Puddling loss. Overman gives 4-5 % on good white metal (pp. 301-302). Percy gives 8 % for good work and 17 % for bad at Dudley (p. 721), and 7-9 % at Bromford (p. 657).
  • Puddling cinder. Overman and Campbell have it tapped or left for the next heat. Percy has it tapped into tap-waggons (p. 646): every heat on grey pig, every other on white, once a turn on refined metal (p. 658).
  • Bessemer blow. Campbell gives 7-12 min, corrected by scrap and steam (pp. 102-108). Percy gives 12-20 min in 1863, the stop not being judgeable, so the blow was run out and spiegel added (pp. 814, 823).
  • Spiegel. Campbell has it added in the vessel or the ladle (p. 112). Percy runs it into the converter and blows five minutes more, then all goes to the ladle (p. 823).
  • Weak balls under the hammer. Overman says the steam hammer "crumbles those balls which have been carelessly put together" (p. 341). Percy says the helve crumbles them and the steam hammer's regulated blows save them (p. 695).
  • Charcoal blast. Percy's 1/2-1 1/2 psi, "always much less" than coke (p. 373), agrees with the other books.
  • Reheat burn-off. Campbell has 1-2 % to scale (p. 170). Percy has puddled to finished bar at 1 t 2 cwt 1 qr per ton, about 10 %, and gas reheating wasting less (pp. 720-722).

Gaps

  • The year of Huntsman's invention is not given (p. 829).
  • The Sheffield pot clay is Derbyshire and Stannington clay with ground old pot and cinders and coke dust (p. 834); the book does not name Stourbridge clay.
  • The book names no damper in the crucible heat: the pot is pulled "as soon as the steel is perfectly molten" (p. 835).
  • The converting furnace's tonnage (p. 772) and the bar thickness and charcoal mesh are illegible in this copy.