Harbison-Walker, A Study of the Open Hearth (1909), and Dichmann, The Basic Open-Hearth Steel Process (1915)

Books A Study of the Open Hearth, The Basic Open-Hearth Steel Process

Covers Harbison-Walker Refractories Co., "A Study of the Open Hearth" (1909, 105 PDF pages), cited "H-W"; and C. Dichmann, "The Basic Open-Hearth Steel Process" (English edition 1915, 356 PDF pages), cited "Dichmann". Pages below are printed folios. H-W has no text layer past PDF page 1; it was rendered at 300 dpi and read by OCR (tesseract), and garbled figures were checked against crops of the page image. Its folio is the PDF page less seven; the two plates (pp. 10-11) were read as images; PDF pages 99-105 are blank and the text ends at p. 91. Dichmann's folio is the PDF page less sixteen, and PDF pages 2-4, 6, 12, 335, 337-340, 346, 353 and 356 carry no text layer. Many of Dichmann's tables are garbled by OCR, and figures are read from the text that cites them. H-W was read whole. Dichmann was read by topic: the introduction and the buoyancy chapter (pp. 1-16), combustion temperatures, preheat and waste heat (pp. 104-138), slag and the boil (pp. 148-195), the heat numbers (pp. 196-206), ore and flame work (pp. 207-228), judging and finishing the heat (pp. 229-244), the processes and their times (pp. 245-294) and the comparison with the converter (pp. 302-312). The producer chemistry chapters (pp. 17-103), the phosphorus valuation, the gas-analysis tables and most worked slag analyses were skipped.

What the books say

The process and its place

Crucible steel dates from 1740, Bessemer from 1855 and the open hearth of the Siemens brothers from 1861 (H-W p. 12). In the United States the open hearth passed the Bessemer in tonnage in 1908, and the Bessemer, "the leading process of steel manufacture, is unquestionably giving way", with "small chance" of new Bessemer plants (H-W pp. 12-13). The open hearth's advantage is control: the bath is sampled at frequent intervals, less metal is over-oxidised and the ingot yield is higher (H-W pp. 13-14). It exists to take the scrap and the off-grade pig the converter cannot (Dichmann p. 3), and a works can lower its output in bad trade by idling furnaces (Dichmann p. 310). Iron lost to the process is 1.1 % against the converter's 2.6 % (Dichmann p. 312).

The furnace, concretely

The acid and the basic furnace are the same furnace; only the bottom differs (H-W p. 14). H-W's plates (pp. 10-11) draw a 1909 fixed furnace in plan and in three sections; Dichmann's Figs. 1A-1B (pp. 5-6) draw the whole plant: a row of producers, a gas main about 9 m up, two reversing valves, the chambers and a 40 m chimney.

  • Hearth. A rectangular bath, open at each end for the ports (H-W p. 14). At most 15 ft wide, the distance a man can throw patching material to the back wall; 2 to 2 1/2 times as long as it is wide; the bath 15 to 20 in. deep (H-W pp. 18-19). A 60 t furnace is the usual size and takes 65-90 t a heat; the range runs from 15 t to 75 t (H-W p. 16). The bottom sits in a riveted steel pan carried on I-beams and piers, independent of the regenerators (H-W p. 19); the plate shows open arches under it.
  • Sides. The charging side carries the doors; the tapping side is the opposite long wall (H-W plate p. 10). Dichmann's 30 t furnace has three charging doors, served by a charging crane (Dichmann p. 6).
  • Tap hole. In the centre of the tapping-side wall near the base, an opening 15 to 20 in. square arched with silica brick (magnesia or chrome in basic practice), closed from inside with lump anthracite and rammed with fire clay from outside (H-W pp. 25, 27).
  • Roof. Silica brick of the highest grade, 9 in. thick, 12 to 13 1/2 in. on large furnaces, run until it burns down to 3 in.; built high, because the old low roofs burned away; sprung from skewbacks on channels of the steel frame, so the walls carry only themselves; some roofs are also arched from port to port (H-W pp. 16-18). Buckstays and tie-rods with turnbuckles hold the whole (H-W p. 15).
  • Ports. At each end, in most large modern furnaces, one gas port and one air port; older furnaces ran two gas and two air, or one and two, or two and three (H-W pp. 31-32). The air port sits above and outside the gas port; both are pitched down to meet about 3 ft above the metal. Higher cuts the roof, lower chills the flame on cold stock; air above keeps the heat off the roof and gas below keeps the air off the charge (H-W pp. 32-33; Dichmann pp. 128-129). The ports and uptakes are "an ordinary gas burner on a large scale" (H-W p. 32). Bulkheads between are an air-cooled steel frame faced with magnesia or chrome (H-W pp. 30-31).
  • Uptakes and slag pockets. Below each port an uptake drops to a slag pocket, an enlargement of the flue that catches dust and takes slag when the bath froths, with clean-out doors (H-W pp. 38-39).
  • Chambers. Four: a gas chamber and an air chamber at each end (H-W p. 14; Dichmann p. 6). The air chamber is at least 1 1/3 times the gas chamber; tall rather than long, 13 to 18 ft high for a 60 t furnace, about 90 cu ft of chamber per ton of steel; 9 in. checker brick stacked into vertical flues 3 1/2 to 4 in. across, over girder-tile flues at the bottom (H-W pp. 35-36). H-W builds them independent of the furnace, beside and below the yard level at each end (plate p. 10); the older type under the furnace worked quicker, but breakouts ran metal into the checkers and cracks leaked gas (H-W p. 34). Dichmann's plant still has them under the furnace (p. 6).
  • Flues, valves, stack. Horizontal flues run slag pocket to chamber and chamber to stack (H-W p. 39). Gas and air enter through throttle valves; a reversing valve, most often the Siemens butterfly, has a casing with "three openings, one to the stack and one each to the two chambers, gas and air" (H-W pp. 39-40). Dampers are sliding plates (H-W p. 41). Waste gas reaches the valves at 400-700 F, 1000-1200 F when the furnace runs badly (H-W p. 40).

A plan of one end and the hearth, read from the plates:

      charging side (doors)
  +----------------------------+
  |air|                    |air|   air port above the gas port,
  |gas|   bath 15-20 in.   |gas|   both pitched down to meet ~3 ft
  +----------------------------+   above the metal
      tapping side, tap hole at centre, base

  end section:  port -> uptake -> slag pocket -> flue -> checker chamber -> flue -> reversing valve -> stack

Firing and reversal

The flame plays over the charge; heat enters from the top, so a quiet bath heats slowly and a boiling one well (Dichmann pp. 4-5). Gas is under pressure from the air blown into the producer bottom; the air enters by its own buoyancy "in the great majority of furnaces", but old furnaces with choked checkers go short of air, "the working of the furnace becomes feeble ... production falls off", and a fan is added; driven too hard, "the flame blazes out of the doors" (Dichmann pp. 7-16). The chimney must not draw gas in (Dichmann pp. 14-15). Reversal is every 15 to 20 minutes, kept up until the charge is ready to tap, and the temperature rises with each; "the practical limit is merely what the brickwork of the furnace will stand" (H-W pp. 14-15). Long intervals overheat the outlet chambers (Dichmann p. 126); bad melters reverse at ever shorter intervals, where the cure is less gas (Dichmann p. 131).

Gas and air are preheated to about 1200 C, "a preheating which may practically be well endured without harm by all furnaces"; the flame reaches about 1800 C and leaves the far ports at 1700 C (Dichmann pp. 113-119). Pushed higher, production falls and "the checkers melt and even the arches of the chambers threaten to sink"; cooling openings are cut into the ports to let cold air in (Dichmann p. 117). Waste gas leaving at 700 C still carries over 40 % of the fuel's energy (Dichmann p. 126).

The furnace is started with a wood fire in one set of chambers until they are dull red, then gas and air (H-W p. 14); gas is never let into a cold furnace (Dichmann p. 138). The melter admits gas until the flame laps from the doors, raises it as the charge melts and at the end must "hold the heat"; he watches through coloured glass, and when the roof and port blocks stop standing out darker than the flame, the roof is burning (Dichmann pp. 129-130). Gas and air are cut during a reversal (Dichmann pp. 136-137).

Fuel

Natural gas is best, 1000-1100 BTU against producer gas at 130; it is not preheated, because its hydrocarbons coked the checkers, so only the air goes through the chambers (H-W p. 43). A producer is a steel cylinder 8-12 ft across and 10-15 ft high, holding a 6 ft fuel bed, blown with air and a little steam; the steam "does not actually add to the calorific efficiency ... but it serves the purpose of keeping the fire at a low enough temperature to reduce the clinkering" (H-W pp. 44-46). Bituminous gas coal is the fuel; anthracite gas is unsatisfactory because it lacks the hydrocarbons that make the flame radiate (H-W pp. 47-48). Dichmann's producers burn coal and lignite (p. 21), lose 7.5-12 % of the gas's value in the flues and chambers (p. 104) and quote Campbell that hot gas leaves the regenerator less to do (p. 125). Sulphur in the producer coal goes into the steel through the flame, most when the furnace works cold (H-W pp. 72, 84; Dichmann p. 168).

Working a heat

  • Charge. Acid practice removes only carbon, silicon and manganese, so the charge's sulphur and phosphorus must already be under the specification (H-W p. 50). Charges run 25 % scrap abroad and 50-75 % in the United States; some pig is essential, because scrap alone oxidises; too much high-silicon pig makes a heavy slag, too much scrap leaves the metal unprotected (H-W pp. 51-55). Dichmann's typical charge is 30 % pig and 70 % scrap (p. 187). Pig goes on the bottom to protect it, scrap on top (H-W p. 56).
  • Melting. "A good charge, of full weight, melts in three to four hours; soon thereafter the boil begins ... highest point after about half-an-hour ... In about a further half-hour the boil becomes quite faint" (Dichmann p. 248). By melt-down silicon and manganese are gone and 0.60-0.80 % C remains (H-W p. 57).
  • The boil. Carbon leaving as gas stirs the bath and brings cold bottom metal to the flame; "a regular boil is of the greatest importance" (Dichmann pp. 164-166). It fades near 0.1 % C, the bath lies quiet and its bottom cools, so "the charge must be tapped as soon as the carbon content has been brought down" (Dichmann p. 167). When the boil stops, tap (Dichmann p. 229). A charge without carbon "cannot boil" and cannot be worked (Dichmann p. 206). Ore in lumps is fed by judgement, about 250 lb per 1 % C in a normal heat; calculating it is "absolutely useless" (H-W pp. 59-62).
  • Pig and scrap. The oldest and commonest process; "all reactions proceed practically spontaneously, also the regulation of the heat, and the entire process requires only little superintendence and aid" (Dichmann p. 249). Too little pig and the heat will not boil (a "set charge"), cured with spiegel or pig; somewhat too much pig is better, the excess burnt off with a few shovelfuls of ore (Dichmann p. 250). Output about 95 % with ideal pig and scrap, 88 % with dirty scrap (Dichmann pp. 250-251). One works on bad scrap ran 7-8 h heats on 12-13 t; a failed heat stayed in 14 h (Dichmann pp. 251-252).
  • Scrap and carbon. All scrap with coke or anthracite laid on the bottom works: 315 minutes against 280 for pig and scrap in the same 20 t furnace (Dichmann pp. 253-254).
  • Molten charge. Molten pig saves heat (Dichmann p. 259); with ore it grows the charge, and the output exceeds the pig charged by 4-7 % (Dichmann pp. 260, 263, 266, 276). Heat 8996 took 26.2 t of molten pig in 260 minutes (Dichmann pp. 272-276). A furnace makes 17-18 t in 4 h on pig and scrap against 30 t on molten pig (Dichmann p. 290); 30 t fixed furnaces make 130-150 t of ingots a day (Dichmann p. 311).
  • Slag. Acid slag "will usually run somewhat under 10 per cent. of the total weight of the charge"; too thick, lime is added; too thin, the metal runs wild (H-W pp. 62-63). Dichmann's basic heat makes about 11 % (p. 188). Frothing slag is kept in by opening the doors to cool the bath (H-W pp. 38-39); in the Donetz practice the door sills are heaped with limestone, and the froth rises when the flame is cut for a reversal and sinks when it returns (Dichmann pp. 269-270).
  • Finishing. Ferromanganese loses about 40 % of its manganese in the furnace and 20-30 % in the ladle (H-W p. 68); a hard steel is caught "coming down" on carbon, a soft one boiled low and brought back with pig or carbon in the ladle (H-W pp. 69-71). Recarburising is often unnecessary; nickel and chromium go in after deoxidation, aluminium and ferro-silicon in the ladle (Dichmann pp. 239-240).
  • Tapping and judging. A bar is driven through the tap hole; a frozen hole took 10-36 h of sledging (H-W pp. 71-72). Samples by spoon, quenched and broken, read within 2-3 points (H-W p. 73); temperature by eye, by the bar's end after stirring, by the colour of slag and flame (H-W p. 74; Dichmann pp. 229-230).

Special processes

The tilting furnace rolls about its port axis or tilts away, with the tap hole above the metal (H-W pp. 41-42). Talbot's 200 t tilting furnace taps 25-30 % at a time, about 50 t every 6 h (H-W p. 89; Dichmann pp. 261-263); Bertrand-Thiel runs two furnaces in series and the duplex process finishes blown metal in a basic open hearth (H-W pp. 90-91; Dichmann p. 266).