Campbell, The Manufacture and Properties of Iron and Steel (1907)

Books The Manufacture and Properties of Iron and Steel

Covers H. H. Campbell, "The Manufacture and Properties of Iron and Steel" (4th ed., 1907, 678 PDF pages). Pages below are the book's printed folios, which are the PDF page less 32. PDF pages 1-5, 10, 14, 34, 66, 209, 452, 542, 672-675 and 678 carry no text layer (front matter, plates, the index end); among them folio 510 (PDF 542), in Part III, was not read. Figure VI-A, the 18-ton vessel (PDF 133-134), was read from the image; the other figures are read from the text that describes them. Several tables (VI-A, VI-E, the heat balances of Ch. X) are OCR-garbled in places; the figures quoted were read where the row is legible. Read by topic: the introduction (pp. 3-31), blast-furnace gas and spiegel (pp. 73-83), puddling, cementation and crucible steel, acid and basic Bessemer, the open hearth and its regenerators, fuel and producers, heating furnaces, the acid and basic open-hearth processes and the special methods (Ch. III-XII, pp. 85-233), steel castings (pp. 409-413), and spot pages of Part III (pp. 488, 520-521, 536, 541-542, 581, 597). The blast furnace chapter beyond pp. 73-83, segregation, hot working, heat treatment, test pieces, the effects of the elements, specifications, welding, most of Part III and the statistics were skipped. Subject steel by the Bessemer, open-hearth and crucible processes, with puddling read for numbers: the converter, the open hearth, the crucible heat, recarburising, the ladle and the gas producer.

What the book says

Pig and the processes in outline (pp. 3-31)

  • Pig carries 3.5-4.0 % carbon and usually 1-2 % silicon (p. 4).
  • Acid Bessemer: silicon supplies most of the heat; carbon gives "barely sufficient heat for the bath to hold its own". About 1 % Si is enough with rapid working, 2 % in the old slow days (p. 8).
  • Manganese is added so the steel is "tough while hot" in the rolls; ferro at about 80 % Mn for soft steel, liquid spiegel at about 12 % Mn for rail steel, "for every ten tons of steel about one ton of this spiegel" (p. 8). Blown iron without it is red-short (pp. 8, 22).
  • The steel is poured into a ladle and tapped from its bottom; the slag floats and the separation is perfect (p. 9).
  • Basic Bessemer: dolomite lining, pig of about 2 % P, and more heat needed, which the phosphorus gives (pp. 9-10).
  • The open hearth must be regenerative; gas and air are heated to a yellow heat. "The problem ... is not to reach the desired temperature, but to control the temperature and prevent the roof and walls from melting down" (pp. 11-12).
  • Acid open hearth: "Nothing is added to form a slag"; the slag is about half silica, from the pig's silicon and the sand bottom. Ore is fed after melting, 500-3000 lb over 3-4 hours; ferro is thrown into the ladle at tapping (pp. 12-13). Basic: lime added, dolomite or magnesite lining (pp. 15-17).
  • Open-hearth steel is held the more reliable; acid Bessemer and acid open-hearth steel cannot be told apart by ordinary means (pp. 14-15, 28). Bessemer for buildings and highway bridges, open hearth for railway bridges, boilers and locomotive forgings (pp. 24-25). Steel castings are acid open hearth, with the side-blown Tropenas converter for small work (p. 26). The crucible process is "a century old" (p. 7).
  • Spiegel runs 10-26 % Mn; ferromanganese up to 80 % Mn with up to 7 % C (p. 82). Blast-furnace gas carries 89-106 BTU per cu ft, about 790-945 cal/m3 (p. 76).

Puddling (pp. 6, 85-91)

  • Forge pig carries 3.0-4.0 % C (p. 86). About three-quarters of the phosphorus goes, from 46 % on hot charges to 91 % on cold (pp. 87-88).
  • "Boilings" are drawn mid-heat while the metal is high in carbon; this hastens the work, cuts the bottom less and makes balling easier. The "tappings" are sometimes tapped from the bottom, "often allowed to remain, or, if tapped, is charged with the next heat". The fettling is iron ore, too infusible to form cinder early (p. 88).
  • Mill cinder carries 49-61 % Fe and goes back to the blast furnace (p. 89). Slag in the finished bar is 0.19-2.26 % (p. 90). Scrap, turnings and borings enter puddling charges (p. 6).
  • No puddling loss percentage is printed.

Cementation and crucible steel (pp. 94-99, 521)

  • Cementation penetrates about 1/8 in per 24 hours; saturation is about 1.5 % C, the average bar about 1 %; the bars are graded by fracture (p. 94).
  • Seebohm: the best razor steel is melted from evenly converted blister; "It will not do to mix hard and soft steel ... or to melt it from pig let down with iron" (p. 94).
  • Charcoal with bar iron in the pot is "almost universal in America"; wrought iron plus pig, and Swedish pig plus ore, are also charged (pp. 94-95). Manganese ore with carbon is sometimes put in the pot (p. 95).
  • Target 0.60-1.40 % C, with P at most .02, S at most .04, Mn at most .20. A little glass gives a passive slag cover (p. 95).
  • Silicon is reduced from the pot walls to .20-.40 % until the metal lies "quiet and 'dead'", and then the pot is pulled. Carbon is picked up from a graphite pot, or from the coke mixed into a clay pot (p. 95).
  • "The crucible lasts from four to six heats, and the weight of a melt is about 80 pounds when the crucible is new" (p. 95). The ingot is 4 in. square (p. 98).
  • One lot of crucible steel ran .85-1.05 % C: "uniformity can, by no means, be assumed" (p. 99, Table V-D).
  • Sheffield is where the crucible process was born (p. 521).

The acid Bessemer vessel (pp. 100-102; Fig. VI-A)

  • Rotation by rack and pinion. The vessel is turned down "as soon as the charge is decarburized": the metal lies in the belly with the tuyeres above it and the blast can stop without the tuyeres filling. A stationary bottom-blown vessel must keep blowing while its tap-hole is opened (p. 100).
  • Sizes: 5 tons in small plants, 10-20 tons in large (p. 100). The 18-ton Sparrow's Point vessel of Fig. VI-A measures about 11 ft 5 in outside and 9 ft 0 in inside diameter, with a concentric nose opening of about 3 ft 2 in and an overall height of about 16 ft 8 in (the last figure blurred); turned down, the bath lies in the belly.
  • Lining about 1 ft thick; brick tuyeres 20-26 in long with 3/8-1/2 in holes, 2.0-2.5 sq in of tuyere area per ton of charge (p. 102).

The blow (pp. 102-112)

  • Table VI-A (Illinois Steel, 1890; 22,500 lb of pig; blast 27 psi), p. 102, flame by time:

    timeflameC %
    2 m 0 ssilicon flame2.94
    3 m 20 sbrightening2.71
    6 m 3 smoderate carbon flame1.72
    8 m 8 sfull carbon flame0.53
    9 m 10 sflame drops0.04

    Mn falls 0.43 to 0.01; the steel holds 0.45 % C after spiegel.

  • Blast 30 psi at first, tending to 20 psi or less, and cut to 10 psi after the flame "breaks through" when the slag is sloppy. Heats are blown in 7-12 minutes (p. 102).

  • Silicon and carbon exchange preference at a critical temperature: a hot bath burns carbon first. Scrap or steam cools a hot bath; blowing on after the carbon flame drops burns silicon or iron (p. 103).

  • Swedish practice: direct metal since 1857; the blow is stopped high in carbon for tool steel, on pig of 1.5-4 % Mn, and the steel is graded after it is made (pp. 104, 108-109).

  • A cold charge is heated by blowing with the vessel partly tipped, at the cost of iron and lining. Cold charges come from low silicon, a low starting temperature or a newly repaired vessel. A pair of 10-ton vessels makes eight heats an hour (p. 106).

  • Vessel slags run 47-69 % SiO2 (p. 106).

  • Loss: 10-11 % on cupola iron, nearly 10 % on direct metal, 8 % at best; shot 0.75 %, iron in the slag 1.8 %, ejected 0.5 %, about 2 % of the iron burned. "120 tons of vessel slag for every 1000 tons of pig-iron", 12 % (p. 107). Small converters lose about 20 % (p. 411).

  • Table VI-E: pig enters at about 1400 C and the blow ends near 1679 C, a rise of about 280 C. Surplus heat per ton: silicon 46,860, iron 14,820, carbon 40,700 (pp. 108-109).

  • Direct metal became practicable through steadier furnace practice and the mixer, an enlarged receiving ladle (pp. 108-110). The cupola is 6-8 ft across and at least 20 ft high, melts 11-15 lb of iron per lb of coke, picks up sulphur, and loses 0.5 % Si and 0.42 % Fe (p. 110).

  • Silicon now runs 1.0-1.5 %, feasible at 0.6-0.8, owing to fast running, hot ladles and vessels, and quick bottom changes (p. 110).

  • 100 lb of scrap more or less in a 7-ton charge shows, and is worth 13 C: about 18 C per 1 % of the charge (p. 111). Ore cools harder: 400 lb of ore cools as much as 1000 lb of scrap (p. 224).

  • Recarburising: solid ferro is thrown into the ladle while it fills. Soft steel is blown to .05 C; .60 Mn added leaves .40, 1.30 added leaves .90. Rail steel is stopped at .10 C and 1.10 Mn added leaves .90. Spiegel is melted and run into the vessel or the ladle (p. 112).

Basic Bessemer (pp. 113-121, 542)

  • Dolomite and tar lining; lime goes in first. The phosphorus after-blow shows no flame change, so it is judged by fracture tests (p. 117). Final slag 16-20 % P2O5; loss 10-17 %; the slag is decanted before recarburising against rephosphorisation (pp. 113-121).
  • A lining lasts 220 heats and a bottom 45-50; a blow takes about 12 minutes, three heats an hour, a bottom change every 16 hours, relining 15 hours plus 6 hours firing (p. 542).

The open hearth furnace (pp. 122-157)

  • To start, the chambers are heated with a wood fire until red; cold producer gas risks explosions (p. 122).
  • The action is cumulative, each reversal hotter; the regenerators run at about 1000 C (1800 F) (p. 123). The flame is never neutral, and runs as a CO flame at the end (pp. 124-125).
  • Checkers: at least 50 cu ft per ton of capacity, better 100; the air chamber is larger than the gas chamber (pp. 126, 129-131).
  • Air at Steelton is blown by a centrifugal fan at very low pressure (p. 132). Gas and air are "pushed in by the upward force of the white-hot vertical port", not pulled; a fan is auxiliary for horizontal chambers. Gas enters below the air and they meet about 5 ft from the metal (p. 144).
  • The Campbell tilting furnace runs the slag through the port joint and repairs holes by tilting (pp. 132-142).
  • Reversal "at least twice every hour" (p. 145).
  • Temperature is judged by eye through blue glass to about 13 C. Chamber hot ends are bright yellow; the cold ends sit about 100 C above the incoming gas (p. 148). Hot gas is "to some extent ... an error": its heat goes out in hotter waste gas (p. 148).
  • Heat balance: 500 lb of coal per ton of steel; 21.6 % lost in the producer; of the rest, half to radiation, a quarter to the stack, a quarter to melting. Internal combustion is one-seventh of the gas heat. Waste gas leaves at 680 C; each 100 C saved is 6 % (pp. 149-156).
  • Producer gas: CO2 5.7, CO 22.0, CH4 2.6, C2H4 0.6, H 10.5, N 58.2 %; 1260 cal/m3; it leaves the producer at 655 C (pp. 153-154).

Fuel and heating furnaces (pp. 158-178)

  • Producers burn bituminous coal, anthracite in special cases; a steam jet is run "to keep down the temperature" (pp. 160-161). The water-seal producer gives 2.5-10 % CO2; a hot fire makes clinker and too much steam kills it (pp. 161-162).
  • Soft-coal gas carries tar and soot; the tar gives the flame its luminosity and scrubbing it out reduces the gas's value seriously (pp. 162-163).
  • "The sensible heat of producer gas is wholly wasted" in a regenerative furnace (p. 164).
  • Anthracite gas is far inferior for intense heat and needs steam or the producer grows unmanageably hot (p. 166). Von Juptner ran with no steam jet; the Steelton jet carries 10 % of the coal's energy (pp. 151-152).
  • Water gas gives 2600-2800 cal/m3, about half of retort coal gas (p. 168); producer gas is so about a quarter of coal gas.
  • Heating furnaces: soaking pits; regenerative ingot furnaces at 40 lb of coal per ton or less; coal-fired reverberatories for cold blooms at about 20 times the theoretical fuel, smoky late in the heat, where 1-2 % of the steel lost to scale outweighs any fuel saved; waste heat to a boiler at 25-30 hp per ton of coal per 12 hours; continuous furnaces (pp. 170-173).

The acid open-hearth heat (pp. 179-189)

  • Hearth: nine inches of clay brick, then layers of sand each set for 10 minutes, 18-24 in thick. The bath is 12-15 in deep at 5 tons, 18-24 in at 30-50 tons (p. 179). Slag sand comes "from the tap-hole and ladle-linings" (p. 183).
  • Pig protects the scrap; the bath after melting holds 0.75-1 % C. Too little pig gives a viscous, pasty bath that oxidises and scorifies the bottom (pp. 179-180). Pig is spread over the scrap (p. 180); in America pig is charged last, and a 25-ton furnace is hand-charged in 11-30 minutes (preface to the second edition).
  • Ore: 1500 lb takes 1 % C down to .05, fed in steps of 500 lb (.80, .40, finish). The steel can weigh as much as or more than the pig charged (p. 185).
  • One to two hours of fully liquid bath is enough (p. 187). Every molecule of phosphorus stays in the metal (p. 187).
  • A 1 in square test ingot is chilled and broken; carbon is read within 10 % (pp. 187-188).
  • Ferro in furnace or ladle: 1.00 Mn added leaves .60, .50 leaves .40 (p. 188).

Basic open hearth and special methods (pp. 190-233, 581, 597)

  • Burnt lime is better than limestone (p. 191). Slag of 12 % SiO2 to 48 % CaO; CaO plus MgO at most about 55 %, SiO2 at least 10 % (pp. 194, 197). The first slag is removed (p. 204).
  • Finely divided carbon added in the ladle gives open-hearth rail steel regular results; molten spiegel is impracticable because a cupola run on irregular charges chills (p. 206).
  • Acid open-hearth steel at Steelton went largely into boiler plate (p. 207).
  • Pig-and-ore process: the froth "resembles soda water"; yield 103.1 % of the pig cold and 104.95 % liquid; the slag took 7.3 % of the iron (pp. 211-213). An ore charge of 500 lb per ton of pig would cool the bath 770 C; practice runs 593-643 lb per ton (pp. 219-225).
  • Talbot: tap to tap 3 h 50 m - 4 h 55 m, 92 tons per 24 hours (pp. 213-215). Bertrand-Thiel: 4 h 50 m in the first furnace plus 2 h 20 m in the second (pp. 216-218).
  • Basic slag weighs about twice the burnt lime and carries about 16 % Fe (pp. 228-229).
  • The duplex process is judged expensive (pp. 231-233); an open hearth finishing a duplex charge takes about 3 hours (p. 581).
  • In castings aluminium is three times and silicon four times as effective as manganese at deoxidising (p. 413). In Sweden Bessemer output falls and open-hearth output rises, 207,450 tons in 1900 (p. 597).