US3791819A

US3791819A - Production of stainless steels

Info

Publication number: US3791819A
Application number: US00775041A
Authority: US
Inventors: R Loutzenhiser; E Aukrust; H Meyer
Original assignee: Jones and Laughlin Steel Corp
Current assignee: Ltv Steel Co Inc; Jones and Laughlin Steel Inc
Priority date: 1968-11-12
Filing date: 1968-11-12
Publication date: 1974-02-12
Anticipated expiration: 1991-02-12

Abstract

Disclosed is a method of making stainless steel that has remarkable advantages over the electric-furnace practices now used. The method uses an oxygen furnace, yielding large savings in the time, capital investment, and labor required to produce a given quantity of steel. The practical difficulties that attend the use of an oxygen vessel to make stainless steel are overcome, principally by using a slag of high basicity ratio (3.5 minimum) operating the furnace so that the melt stays below 3100* F during most of the refining, and reducing the slag in a separate vessel, the contents of which are returned to the oxygen vessel.

Description

United States Patent Loutzenhiser et a1.

1451 Feb. 12, 1974 1 PRODUCTION OF STAINLESS STEELS 3,323,907 6/1967 Kurzinski 75/52 x 3,336,132 8/1967 McCoy 75/60 X [751 f g Dem)", 3,366,474 1/1968 Akita c1111. 75/1305 Mlch; Egll Bethel Park; 3,396,014 8/1968 Keyscr 75/1305 Harry W. Meyer, Mount Lebanon, both of Pa. OTHER PUBLICATIONS [7331 Assigneez Jones & Laughlin Sted Corporation, Basic Open Hearth Steelmaking, Second Edition,

Pittsburgh, Pa. A.I.M.E., New York, 1951, pages 196-197.

[22] filed: 1968 Primary Examiner-L. Dewayne Rutledge [2,1] Appl. No.: 775,041 Assistant Examiner-M. .I. Andrews Related Us. Application Data Attorney, Agent, or F1rm-G. R. Hams; T. A.

. Zalensk1 [63] Contmuanon-m-part of Ser. No. 464,900; June 18,

1965, abandoned,

[57] ABSTRACT U-S. v Disclosed is a method of making tainless teel that 75/52 75/6O, 75/133-5 has remarkable advantages over the electric-furnace Clpractices now used The method uses an oxygen fur; I Field of Search-m3 60, 1. nace, yielding large savings in the time, capital invest- 75/l33-5 ment, and labor re uired to produce a given quantit 9 Y of steel. The practical difficulties that attend the use References Cited of an oxygen vessel to make stainless steel are over- UNITED STATES PATENTS come, principally by using a slag of high basicity ratio 3 507 642 4/1970 Shaw 75/52 minimum) Operating the furnace so that the melt 2:847:30l 8/1958 Shaw 75/1305 stays below 3100 F during most of the refining, and 3,172,758 3/1965 .Iandras 75/60 X reducing the slag in a separate vessel, the contents of 3,198,62 3/ 5 Bell t -m 5/ 6 X which are returned to the oxygen vessel. 3,252,790 5/1966 Krivsky 75/60 Y v I 3,260,591 7 1966 Brown, Jr. et a1. 75/1305 X 1 Claim, 2 Drawing Figures LOW CARBON AND/OR STAINLESS STEEL SCRAP FERRO CHROMIUM 60-70 %CR, 3% C.

CU POLA (LIME) MOLTEN METAL 8-30/ CR, 3'6 %C, BOF 0.1-1.5 lo SI, BAL. Fe

CHROMIUM 01912,) I

SCRAP HOT METAL REDUCED (SLAGI SLAG FERROCHROMIUM SILICON OR FERROSILICON LA DLE PAIENIEII F551 LOW CARBON AND/OR STAINLESS STEEL SCRAP FERRO CHROMIUM 60-70%CR, 3%C.

CUPOLA (LIME) MOLTEN METAL 830% CR,3-6%C, 801- O.||.5%S|,BAL.Fe

CHROMIUM ORE, SCRAP Flg. I.

HOT METAL REDUCED (SLAG) G FERROCHROMIUM SILICON OR P FERROSILICON LADLE INVENTORS.

RUSSELL H. LOUTZENHISER EGIL AUKRUST 0nd HARRY W MEYER ATTORNEYS 1 PRODUCTION OF STAINLESS STEELS CROSS-REFERENCES TO RELATED APPLICATIONS This application is a continuation-in-part of our copending application Ser. No. 464,900, filed June 18, 1965, now abandoned.

BACKGROUND OF THE INVENTION 1. Field of the Invention:

This invention relates to an oxygen blowing process for making stainless steel.

2. Description of the Prior Art:

Although several patents disclose making stainless steel with the use of an oxygen vessel, the various difficulties that attend such process have, before our invention, prevented those skilled in the art from obtaining a commerically feasible process.

One problem that besets a converter process for making stainless steel is that in chromium-bearing impure ferrous metal melts at temperatures below 3100F, the oxygen reacts with the chromium in preference to the carbon and siliconthat need to be removed from the hot metal, causing the development of a substantial volume of slag that is rich in chromium oxide. Before this invention, the feasibility of recovering the chromium values from the slag was in doubt. The alternative is to operate the process at melt temperatures such as 3100F to 3600F, but this creates lining prob lems. Refractories of higher quality need to be used, lining lives are shorter, and maintaining the structural integrity of the lining, where its individual bricks become heated through about half of their bulk or more and then cooled, is also a problem.

Another difficulty concerns getting the required chromium values into the product steel at reasonable cost. With the electric furnace practice most commonly in commerical use, chromium values are ob-. tained from chrome ore and from high-carbon ferrochromium, which are both relatively inexpensive. In that practice, low-carbon ferrochromium, which is considerably more costly, is used, if at all, only in small amounts at the end of the process, to adjust the composition but slightly.

In the electric furnace process for making stainless steel, the slag basicity (ratio of basic components such as C210 and MgO to acidic components such as SiO is usually maintained at about 1.5, since such a slag basicity is most favorable for removal of sulfur and phosphorus impurities in the hot metal.

Among the patents in the prior art that are most pertinent to the present invention, specific notice is to be taken of Bell et al. US. Pat. No. 3,198,624, which relates to the making of stainless steel with the use of an oxygen vessel, indicating the possibility of using a step wherein slag generated on top of the hot metal in the converter is decanted into a separate vessel and there reacted with suitable reductant to produce a chromium-rich hot metal, which is then mixed with the metal remaining in the converter to obtain a product of desired high final chromium value. The patent does not, however, mention the improvement in chromium recovery that is obtainable by returning the entire contents of the separate reaction vessel to the converter vessel, whereby, as the chromium-rich hot metal is returned to the converter vessel, further reaction occurs,

increasing the amount of reduction of chromium from the chromium-rich slag that takes place. The patent refers, moreover, to a process for making stainless steel that is distinguishable fromthat of the Applicants in other respects, namely, in the idea that, according to Bell et al, that the refining should be done with the hot metal bath having a temperature such as 3100F to 3600F, so that the refractory brick lining the furnace are necessarily exposed to temperature thathigh for a substantial period of time, becoming heated to such a temperature to a considerable depth, such as about half of their thickness, with the concomitant expansion and contraction possibly endangering the structural integrity of the furnace lining. Moreover, the patent to Bell et al. does not give any inkling of the necessity of using a high slag basicity ratio, such as that called for in accordance with the process of the Applicants.

Another patent that has some pertinence to the practices of the present invention is the Shaw U.S. Pat. No. 2,847,301. It teaches that, in the production of stainless steel in an electric furnace, it is particularly desirable, in order to make a clean steel, to add to the steel after oxidizing the melt, in order to remove impurities therefrom, a quantity of a master alloy containing at least twice as much manganese as silicon. The idea is that such a master alloy forms, with the oxygen in the metal bath, a complex manganese-silicon-oxide that is low-melting and will float out of the bath, leaving the product steel cleaner by removing oxygen therefrom before ferrosilicon, which is used to reduce the chromium oxides in the slag, accidentally comes into contact with the metal bath. The Shaw patent teaches the use, in connection with this process, of slag having BRIEF SUMMARY OF THE INVENTION Stainless steel of commerical quality, having a carbon content of, for example, 0.08 percent maximum, is obtained by refining in an oxygen converter vessel an impure chromium-bearing ferrous hot metal, using a vertical oxygen lance that supplies oxygen thereto at such a rate that the refining is completed within, for example, about V2 to 2 hours. During a substantial portion of the oxygen blowing, more than half thereof, the temperature of the ferrous melt in the converter vessel is maintained at less than 3100F, so that problems with lining life and the structural integrity of the lining are, to a great extent, avoided. This leaves, of course, the known difficulty that, when oxygen is brought into contact with a melt of such temperature, most of the chromium contained therein is oxidized and, unless measures to recover the same are taken, lost to the slag. A further feature of the invention is, accordingly, that the slag produced during the oxygen blowing step is decanted into a separate vessel where it is treated with a suitable reductant, such as ferrosilicon or ferrochromium silicide, with the entire contents of the separate reduction vessel being returned to the oxygen converter; when the entire contents of the separate reaction vessel are returned to the oxygen furnace, as a result of the flowing of the chromium-rich hot metal produced in the separate reaction vessel through the slag, further reduction is obtained. This yields chromiumrecovery values substantially greater than those obtained in any other way and lends appreciably to the commerical feasibility of the process. Yet another point of difference between the method of the invention and that of prior art is that there is maintained, in the slag above the ferrous metal in the oxygen converter vessel during the refining operation, a slagbasicity ratio of at least 3.5, and preferably higher, e.g., 5, as compared with a slag-basicity ratio of about 1.5 used in the known electric-furnace practice for the making of stainless steel and in the other oxygen steelmaking practice of which Applicants are aware. This use of a higher slag-basicity ratio improves both the fluidity and the reactivity of the slag involved, as respects the removal of such impurities as sulfur and phosphorus. A high basicity ratio also promotes improved reduction of the chromium values from the slag during the slag-reduction step.

DESCRIPTION OFTHE DRAWINGS DESCRIPTION OF THE PREFERRED EMBODIMENTS The process for producingstainless steel in accordance with the inventionmay be divided in three main steps, namely, melting, refining, and reducing. These will be considered separately as follows.

MELTING As shown in FIG. 1, low carbon and/or stainless steel scrap is charged into a cupolafurnace along with high-carbon ferrochromium containing 60-75 percent chromium, more than 3 percent carbon, and less than 2 percent silicon. The cupola 10 may be of the coldblast or hot-blast type. If an acid lining is used in the cupola, the iron will contain an appreciable amount of silicon (i.e., 0.4-1.5 percent). As an example, a cold-blast acid-lined cupola with a hearth diameter of 29 inches was charged with the following:

Coke, 78 pounds 93.7% fixed carbon Ferrochromium, 165 pounds 1.4% silicon, 4.9%

carbon, and 68% chromium Steel Scrap, 344 pounds 0.15% silicon, 0.25% carbon Limestone, 18 pounds vA total of eight such charges were fed to the cupola. The cold-blast rate was 1500 cubic feet per minute, and this was supplemented by 4000 cubic feet per minute of 99.5 percent purity oxygen. The heat produced 4267 pounds of metal having a composition as follows:

Chromium 19.8%

Silicon 1.08%

Carbon 5.2%

remainder iron and usual impurities.

The metal temperature in the ladle at the end of the heat was 2660F.

If a basic lining is used in the cupola, a similar metal is produced, but the metal contains appreciably less silicon. For example, use of a charge of 78 pounds of coke, 140 pounds of ferrochromium, 380 pounds of scrap, and 18 pounds of limestone produced a metal containing 17.8 percent chromium, 5.10 percent carbon, 0.22 percent silicon, the remainder iron and impurities. I

The use of abasic lining in the cupola ispreferable since a lower silicon content in the hot metal produces a lower slag volume in the converter and decreases the quantity of reductant that must be added to the reaction ladle. In addition, the sulphur content of the hot metal is lower when a basic slag is used in the cupola. lron of this composition can be used as a basis for the manufacture of stainless steels of the AIS! 200, 300 and 400 series. It is preferable, however, to displace some of the low-carbon scrap and/or ferrorchromium in the cupola charge with chromium-nickel containing scrap or pig iron when manufacturing stainless steels of the 200 and 300 series. Chromium-containing scrap may also be used in the production of stainless steels of the 400 series. It should be understood, however, that a cupola having a water-cooled shell without any lining in the stack may be employed for melting with good re sults.

- REFINING As shown in FIG. 1, the output of the cupola is moltenmetal containing 8-30 percent chromium, 3-6 percent carbon, 0.l-l.5 percent silicon, and the balance iron. The molten metal is charged into a top-blown converter-type furnace 12, one form of which'is shown in FIG. 2. It comprises an outer steel shell 14 having a basic refractory lining 16 provided on its inner surface. The molten metal from the cupola, together with lime, chromium ore and possibly scrap are charged into the vessel through an opening in a nose cone l7; and the molten metal, after refining, is poured from the vessel through the aforesaid opening.

Above the furnace 12 is a water-cooled oxygen lance 18 which, when the furnace is in an upright position as shown by the dotted outline in FIG. 2, is lowered downwardly into the vessel and terminates above the surface of a layer of slag which overlies the molten metal within the converter. When oxygen of high purity (99.5 percent minimum) is projected from the lance 18 onto the surface of the slag and the molten metal therebeneath, a turbulent action is produced while the oxygen reacts with the various impurities in the molten metal to form oxides which pass off as gases or accumulate in the slag layer. It is to be understood that the oxygen blowing rate is sufficiently high that the reaction between the oxygen and the impurities and the molten ferrous metal in the converter takes place sufficiently rapidly that the carbon and silicon therein are removed rapidly, within /-Z hours, to such an extent as to yield a stainless steel of acceptable composition. Almost invariably, the carbon content of the product stainless steel is less than 0.15 percent, and in most instances it is 0.08 percent or below. This constitutes a substantial diminution in the carbon content from that contained in the cupolaproduced hot metal mentioned above, which, as aforesaid, contains 3-6 percent of carbon.

The invention is not limited to the use of commerically pure (99.5% oxygen. For example, a gas containing percent oxygen and 10 percent nitrogen may be used, and with such gas, steel containing 0.04. percent nitrogen can be obtained. This is commerically important, since the production of tonnage oxygen of' 99.5% purity requires a final liquid fractionaldistillation step that raises the oxygen content from about 95 percent to over 99.5 percent. It is desirable, when possible, to avoid the expense of this step. In making stainless steel, particularly the austenitic grades, the nitrogen added to the steel, if such lowerpurity oxygen is used, is not at all harmful, and is sometimes considered helpful. This is pointed out because when the oxygen steelmaking process is used to make the plain-carbon steels, it has hitherto been considered essential in most instances to avoid the use of oxygen less than 99.5 percent pure, since otherwise the nitrogen content of the steel rises to the point where the steel becomes too brittle for most, if not all, uses. Accordingly, it is fair to state that in accordance with the invention, there is used a gas containing about 90 percent or more of molecular oxygen. We have discovered that where oxygen of high purity (99.5 percent minimum) is available, e.g., in connection with apparatus to be used for making either plain-carbon or stainless steel, there is obtained a stainless steel of relatively low nitrogen content, about 0.015-0.025 percent vs. the 0035-0050 percent that is usual in the electricfurnace-produced stainless steels. Such low-nitrogen stainless steel has, as those skilled in the art will appreciate, substantially superior resistance to stresscorrosion cracking, and the production of such lownitrogen stainless steels has, prior to this invention, entailed particular expense and care that are, with our invention, no longer required. The advantage, however, is quite a small one, compared with that obtained by being able to make stainless steel in an oxygen converter, rather than an electric furnace, under conditions economically feasible.

In accordance with the invention, hot metal is charged directly into the furnace from the cupola, but no scrap is added at this time. With the furnace in an upright position as shown in the'dotted outline in FIG. 2, the lance 18 is lowered and the oxygen is turned on, and refining commences. When the rate of carbon monoxide plus carbon dioxide evolution from the bath has reached a level of approximately 100 percent utilization of lance oxygen for carbon removal, chromium ore is gradually charged into the furnace. This normally occurs after about one-quarter of the total refining time has elapsed. The chromium ore ordinarily contains 1.2-9.1 percent silicon dioxide, 12.4-28.4 percent aluminum trioxide, 0.1-1 .0 percent calcium oxide, 10.6-18.7 percent magnesium oxide, 11.9-16.4 percent ferrous oxide, and 32.4-49.9 percent chromium trioxide. The quantity of chromium ore which is added is dependent upon the desired finishing temperature at the end of refining and also on the size of the furnace being used. For example, in a converter making a 60- ton heat, up to 150 pounds of chromium ore per net ton of steel may be used; and if coke is charged, even larger amounts of chromium ore can be used. In smaller converters, the quantity of chromium ore which can be used is decreased, since a larger fraction of the heat available from the refining reactions is then used to make up for heat losses. The chromium oxide in the slag and ore is reduced by the carbon in the bath. This is illustrated by the following Table I:

TABLE 1 Time of Cr 7r C Bath lbs. otCr Heat Blowing in in Temp. Reduced per nel No. (min.) Bath Bath (Fl ton of Steel Note that as the time of blowing increases, the chromium content of the bath also increases; however as the blow is continued beyond about 26 minutes, some of the chromium is oxidized and returns to the slag.

Calcinated lime or dolomitic lime is added at the beginning of refining so as to produce a basicity (i.e., a ratio of weight of Ca0 MgO to weight of SiO of not less than 1.5, for example. As soon as the metal bath in the furnace during refining has reached a temperature of about 2800F, additional lime is added during oxygen blowing in order to produce a slag basicity of at least 3.5. This is a distinct departure from conventional electric-furnace practice, wherein the basicity of the slag at the completion of refining is normally less than 1.4. While the basicity ordinarily will be determined by the amounts of CaO, MgO and S10 in the slag, it should be understood that in accordance with the broader aspects of the invention, and as used in the following claims, basicity means the ratio of the weight of any basic fluxing agent, or combination of basic fluxing agents, to the weight of any one or more acidic fluxing agents.

When the carbon in the bath has been reduced to the desired level, normally under 0.10 percent, the temperature of the metallic bath will be in the range of about 3200F to 3600F, depending upon the chromium and carbon contents desired in the final product. At this time, oxygen refining is complete.

' Another point that is to be noted about the refining operation is that, as will be seen from certain data hereinafter included, a very substantial portion, more than half, of the carbon removal is done when the temperature of the molten metal in the converter 12 is at a temperature of less than 3100F. It is true that, for short periods of time, the temperature of the molten metal in the converter 12 may exceed 3100F, but this is not permitted to occur for any length of time sufficiently lengthy that the bricks forming the lining of the vessel 12 become heated to such a temperature through any more than a relatively thin surface area thereof, rather than becoming heated, for example, as to half of their thickness or more, as would occur if the refining operation was permitted to continue with the molten metal at such a high temperature for a sufficiently length of time. The use of this lower melt temperature during the greater part of refining has two effects. One is that, as a result, much of the chromium charged to the furnace is oxidized into the form of chromium oxide and begins to appear in the slag. As indicated above, it is known that at melt temperatures less than 3100F, chromium has a greater affinity for oxygen than carbon does, so that oxygen treatment of the chromium-bearing ferrous metal bath tends to cause the chromium to be oxidized into the slag, rather than remaining in the hot metal, as occurs when the bath is at a temperature of 3100F or greater. On the other hand, it is known that at temperatures such as 3100F or greater, though oxidation of the chromium into the slag is avoided, there develop serious problems concerning the life and the structural integrity of the furnace lining. If, for example, temperatures such as 3100F3600F are used for a substantial portion of the refining operation, i.e., one-half or more,

so that the refractory bricks comprising the furnace lining are heated to such a temperature to more than onehalf their thickness, problems concerning the structural integrity of the furnace lining arise. It is important, accordingly, to control the temperature of the metal bath during the refining operation so that operating temperatures so high as to yield the above-indicated disadvantage are not encountered, and to this end, additions at periodic intervals of scrap or other suitable coolant may be made to the material contained in the converter 12, or, considering the size of the converter vessel used, the flow of oxygen used may be adjusted appropriately.

As indicated above, however, it is intended that the oxygen flow rate be sufficiently high that complete refining of the impure chromium-bearing hot metal is ob tained within a reasonably short time, such as about 2 hours or less. In most instances, the time required for the refining will be on theorder of 1 hour.

REDUCTION fluid due to its high temperature and basicity (i.e., above 3.5 is poured from the converter 12 into a ladle 20 (FIG. 2) which preferably has a height-to-diameter ratio greater than 2. In the ladle 20, the slag poured from the converter 12 is mixed with a suitable reducing agent such as ferrochromium silicide and/or ferrosilicon to recover the chromium therefrom. Preferably, the ferrochromium silicide and/or ferrosilicon is fed continuously into the tapping stream 22 during the time when the slag is being tapped from the converter 12. This may be achieved, for example, by storing the ferrochromium silicide and/or ferrosilicon in a hopper 24 provided with a conduit 26 for conveying the granular ferrosilicon into the tapping stream 22 during a tapping operation. Alternatively, however, molten ferrochromium silicide and/or ferrosilicon can be introduced into the ladle 20 either during tapping or thereafter. Finally, it should beunderstood that any means for thoroughly mixing the reducing agent with the slag may be The slag, which at theend of the oxygen blow is very employed in the ladle 20. Feeding of the reducing agent in granular form into the tapping stream, however, is considered most desirable.

In the tapping operation, some of the metal in the 5 converter 12 is unavoidably tapped into the reaction ladle 20. Rapid reduction of the slag results in the chromium and iron oxides being reduced by the silicon in the reductant (i.e., ferrochromium silicide or ferrosilicon). Some manganese oxide in the slag is normally reduced also. The reaction ladle, following .completion of reduction, contains a metal rich in chromium and a slag containing less than percent chromium, 25-40 percent silicon dioxide, and 30-60 percent calcium oxide. If the reduction is effectively carried out, the chromium content of this slag will ordinarily be less than 5 percent. The metal in the reaction ladle may contain 2 to 10 times the maximum chromium specification in the final steel. As shown in FIG. 1, the mixture of metal and slag from the reaction ladle 20 is then returned to the converter 12; and as the high silicon metal pours through the slag as it is transferred back into the converter, further slag reduction occurs. This is evidenced by the fact that the slag above the chromium-rich hot metal in the slag-reduction ladle 20 is black, but the slag above the metal in the converter 12 is light green. Those skilled in the art will realize that this change in slag color is indicative'that a substantial further reduc-' tion in the chromium content of the slag occurs during the pouring from ladle 20 to converter 1.2. Black slag has a chromium oxide content over 10 percent; lightgreen slag has a chromium oxide content of 3-7 percent.

The remainder of the process consists of making additions of stainless or carbon steel scrap, nickel, or manganese, depending upon which grade of stainless steel is being produced. These additions are, in the normal manner, calculated based on the composition of the metallic bath and the specification of the final steel and on the temperature of the metallic bath in the converter which. is normally above 320091 The addition of cool-off material is made immediately after the slag is tapped into the reaction ladle.

The effect of the process of the invention in recovering chromium values from slag can best be understood by reference to Table II in which the data were derived from three ton blows.

TABLE II 300 Series 400 Series Heat No. S-15 S-20 S-22 S-l6 S-18 S-19 S-24' Cupola Metal:

Initial Lime charge to 42.9 58.0 56.7 63.4 59.0 60.9 54.7 converter (lbs/NT cupola metal) Lime charge during blow 85.8 87.0 85.0 95.0 88.4 91.0 82.1 (lbs/NT of cupola metal) Basicity of Slag End of 0 5.9 7.4 9.8 5.4 5.5 8.7 7.9 blow Converter metal after blow 13.02 13.05 12.30 14.92 10.36 10.18 11.74 %Cr 1 Chromium yield during 0, 59.7% 71.6% 59.5% 73.1% 51.2% 59.4% 66.0% blow Final metal after reduction in 18.93 19.41 18.01 17.99 17.15 15.94 18.65 reaction ladle %Cr Ferrorchromium silicide I18 I35 118 126 122 130 .112 added in reaction ladle N (lbs/NT of steel) during the blow varies from a low of 82.1 pounds per net ton of cupola metal for heat No. 8-24 to a high of 95 pounds per net ton of cupola metal for heat No. 8-16.

The results of Table 11 show that in all cases the percentage of chromium in the final metal after reduction of the slag increases markedly from that in the metal immediately after the oxygen blow. For example, in heat No. S- the percentage of chromium in the metal increased from 13.02 percent to 18.93 percent; and in heat No. S-l8 increased from 10.36 percent to 17.15 percent. The ferrochromium silicide added in the reaction ladle varies in the examples given in Table 11 from 112 pounds per net ton of steel in the converter for heat No. S-l5 to 135 pounds per net ton of steel for heat No. 8-22. The chromium lost in the final slag expressed as a per cent of total chromium charged varies from a low of 0.8 percent for heat No. 8-15 to 7.5 percent for heat No. 8-24. This, of course, is far below that achieved in conventional electric arc melting furnaces.

Table 111 illustrates the deleterious effect of lime addition to the reaction ladle on slag reduction, as compared with addition to the converter.

' TABLE 1IContinued 300 Series 400 Series Heat No. 8-15 540 s-22 S-l6 S-18 S49 S44 Chromium reduced by slag 44 "'5'1" 34 74 7 33 reduction (lbs/NT of steel) llrBmium lost in final slag .a% 1.6% 4.3% 2.4% 3.6% 6.8% 7.5%

expressed as a percent of total chromium charged "Titdnf 1H 'all d'f'the examples given in Tablellfihe deg'rrda TABLElII the converter 12 was poured into a separate reaction 15 No, ladle 20 while ferrochromium silicide and/or ferrosili-' 7 con was introduced into the tapping stream as shown Cupm' 2;: gig; in FIG. 2. Examples are given for the 300 Series stain- %c 5.12 5.37 less steels containing both chromium and nickel as well 'z g'z z f f g (lbs/NT' 39 as the 400 Series containing chromium alone. 20 Lime charge during blow (lbs/NT of cupola 40 56 h metal) The lnltlal lime charge to the converter before oxyt me tion to eact n ladle s/NT of 27 26 upo a metal) gen 15 blown onto the surface of the metal bath ranges Comm after blow Cr in M2 between 42.9 pounds per net ton of cupola metal for Chromium yield during 0 blow 54.0% 67.4% heat 845 to 34 p d p net ton of cupola 1321 metal after reduction ln reactlon ladle 15.70 18.19 metal for heat No. 8-16. Generally speaking, the initial 25 Ferrochromium silicide added in reaction ladle 143 i b v (lbs/NT of steel) I cha ge should e abo e 35 pounds per ton of Chromium reduced by slag reduction (lbs/NT 37 34 cupola metal for the three ton blows from which the fsg l) data of Table 11 was derived. The lime charge added Chromium 1051i" final Slag expressed as 277% 166% percent of total chromium charged NetTori w in the exemplbf Table 111, the slag was poured from the converter 12 into a separate reaction ladle 20 in the same manner as in the case of the examples given in Table 11, and substantially the same amount of ferrochromium silicide was added to the tapping stream. However, due to the fact that a portion of the lime is added to the reaction ladle during reduction, the chromium reduced by the slag reduction is, on the average, much lower than that for the examples given in Table 11. Moreover, the chromium lost in the final slag is much higher, notwithstanding the addition of lime to the reaction ladle after the slag has been poured therein. This is probably due to the fact that the addition of lime to the reaction ladle quenches the slag and impedes the reduction. Thus, whereas the chromium lost in the final slag varies between 16.6 and 27.7 percent for the examples in Table 111, the maximum chromium lost in the final slag for the examples in Table II is 7.5 percent, indicating that the lime must be added to the converter only, and not to the reaction ladle.

A comparison of Table 11 with the following Table IV 'will demonstrate the effect of carrying out the slag reduction in a separate vessel.

TABLE 1V Heat No. 8-1 S-3 S-5 'Cupola Metal: %Cr 13.80 15.81 15.01 %Si 0.23 0.04 0.13 %C 5.12 5.12 4.92 lnitial lime charge to converter (lbs/NT of 2 13 13 cupola metal) Lime addition to the converter during 0 blow 49 53 64 (lbs/NT of steel) Converter metal after blow %Cr 11.38 9.10 9.15 Chromium yield during 0, blow 57.4% 38.2% 47.1% Final metal after reduction of slag %Cr 13.4 14.6 14.8 Ferrochromium silicide added to the converter 73 73 91 (lbs/NT of steel) Chromium reduced by slag reduction (lbs/NT 26 27 18 of steel) Chromium lost in final slag expressed as a per 27.30% 30.20%

cent of total chromium charged Net Ton In the examples given in Table IV, the conditions for lime addition are substantially the same as those uti lized in conventional electric furnace practice, the ferrochromium silicide being added to the converter itself and an attempt made to mix the ferrochromium silicide with the slag in the converter by rabbling without transfer to a separate raction ladle. Note that the chromium lost in the final slag is generally greater than that lost in conventional electric arc furnaces, probably due to the fact that no external source of heat is available in the BOF during rabbling as is the case in an electric furnace. Thus, rabbling is not an effective means for slag. reduction in the BOF.

Although the invention has been shown in connection with certain specific examples, it will be readily apparent to those skilled in the art that various changes in composition and process steps can be made without departing from the spirit and scope of the invention. We claim as our invention: i V I 1. In the manufacture of stainless steel, the steps of producing in a cupola furnace an alloy containing about 8-30 percent chromium, 3-6 percent carbon and 0.1-l .5 percent silicon, the remainder substantially all iron, transferring the hot metal from the cupola into a top-blown converter-type furnace, refining the hot metal within the converter by blowing oxygen onto the surface thereof, adding calcinated lime at the beginning of refining to produce a basicity in the slag on the bath of at least 1.5, adding additional calcinated lime when the temperature of the bath reaches about 2800F to produce a slag having a basicity of not less than 3.5; pouring the slag formed on the surface of the metal bath in the top-blown converter into a separate reaction ladle having a height-to-diameter ratio greater than two, feeding a material selected from the group consisting of ferrochromium silicide and ferrosilicon continually into the tapping stream as slag is poured from the top-blown converter into the ladle to thereby effect complete mixing of the said material with the slag and produce in the ladle molten metal rich in chromium beneath a layer of reduced slag, and thereafter returning the contents of the ladle back to the con- VC rter.